KR100463567B1 - Peptides and Zero-Glycan Inhibitors of Selectin-Mediated Inflammation - Google Patents

Abstract

Tyrosine sulphate of PSGL-1, in particular one or more residues (46, 48, 51), work together with sialated and fucoated glycans, most preferably Thr-57 to mediate high affinity binding with P-selectin do. PSGL-1 O-glycans were found to be composed of Galsyl → 4GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH in the disialized or neutral form of Core-2 tetrasaccharide. A small number of O-glycans are two main species containing sialyl Lewis X antigens (one species being disalylated, monofucose glycan of formula (I), and the other species being formula (II) Α, 1,3 fucose produced as monosialated, trifucose glycan having a polylactosamine backbone of:

[Formula I]

(I)

[Formula II]

(II)

Where

R is H, OH, another sugar or aglycone such as an amino acid.

Description

Peptides and 0-glycan inhibitors of selectin mediated inflammation

The present invention relates to inhibitors of selectin mediated inflammation, in particular to inhibitors derived from PSGL-1, ligands for P-selectin, and novel fructose derived from PSGL-1.

US Government Richard D. Cummings and Roger P. The United States National Institute of Hygiene Research Grant Nos. PO1 HL54804 and RO1 HL 43510 to McEver have the rights of the present invention.

Selectin is a family of three Ca ²⁺ -dependent cell membrane-binding lectins that initiate the attachment of leukocytes to platelets or endothelial cells under shear forces found in the circulating veins [Raski, (1992) Science 258, 964-969; Bebilacua group, (1993) J. Clin. Invest. 91, 379-387; McEver (1994) Curr. Opin. Immunol. 6, 75-84]. L-selectin expressed in white blood cells binds to constitutive ligands or induced ligands of endothelial cells. E-selectin expressed by cytokine-active endothelial cells, thrombin-activated platelets and P-selectin expressed by endothelial cells bind to ligands of a subset of bone marrow cells and lymphocytes.

P-selectin is a calcium-dependent carbohydrate-binding protein expressed on the surface of activated platelets and endothelial cells in response to thrombin and other agonists [MacEver et al., (1995) J. Biol. Chem. 270, 11025-11028; Barki, a. (1994) Proc. Natl. Acad. Sci., United States, 91, 7390-7397; Springer, T.A. (1995) Annu. Rev. Physiol. 57, 827-872. Through binding of leukocytes to glycoconjugate-based counterreceptors, P-selectin mediates rolling binding of these cells in activated platelets and endothelial cells (Lawrence and Springer (1991) Cell 65, 852-873; Moore et al. (1995) J. Cell. Biol. 128, 661-671. Sialic acid and fucose are both components of the P-selectin counter-receptor of leukocytes [Coral Group, (1990) Biochem. Biophys. Res. Comm. 172, 1349-1356; Moore Group, (1992) J. Cell. Biol. 118, 445-456; Saco et al., (1993) Cell 75, 1179-1186).

Although the selectin weakly interacts with small sialated, fucose fructose such as sialyl Lewis X (sLe ^x ; NeuAcα2-3Galβ1-4 [Fucα1-3] GlcNAc) [explosion party, (1992) J. Cell Biol. 117, 895-902; Barki, (1992) Curr. Opin. Cell Biol. 257, 257-266], a limited number of glycoproteins [Moore party, (1992) J. Cell Biol. 118, 445-456; Raski et al., (1992) Cell 69, 927-938; Levinovic Group, (1993) J. Cell Biol. 121, 449-459; Walcheck Group, (1993) J. Exp. Med. 178, 853-863; Baumheuter group, (1993) Science 262, 436-438; Binding with high affinity with the glycans of Berg et al., (1993) Nature 366, 695-698] or proteoglycans (Norgard-Siumnitz et al., (1993) Science 261, 480-483). Determining factor of sialyl Lewis X-containing fructose to a (sLe ^x), NeuAcα2-3Galβ1-4 [Fucα1-3] GlcNAcβ1-R of the white blood cell surface inhibits leukocyte adhesion to P- selectin [poly party, (1991 ) Proc. Natl. Acad. Sci. US 88, 6224-6228; Explosives party, (1992) J. Cell Biol. 117, 895-902] However, since non-myeloid cells expressing high levels of sLe ^x bind insufficiently to P-selectin compared to bone marrow cells, expression of sLe ^{x on} the cell surface binds with high affinity for P-selectin. Not enough to [Ju et al. (1991), J. Cell Biol. 115, 557-564. High affinity ligands are potentially important as physiological mediators of selectin-mediated leukocyte adhesion in the presence of inflammation. Thus, understanding the structural basis for the high affinity recognition of the sugar binders by selectin has attracted much attention.

A subset of the high affinity selectin ligands consists of glycine-like glycoproteins (MacEver et al., (1995)). Sialomusine ligands for P-selectin are expressed by human neutrophils and human promyelocytic cell line HL-60 [Mourning, (1992); Moore et al. (1994) J. Biol. Chem. 269, 23318-23327. Leukocytes express one high affinity ligand for P-selectin, namely P-selectin glycoprotein ligand-1 (PSGL-1) [Mourde, (1995); Moore group, (1992); Sako et al. (1993); Nordgard, et al. (1993) J. Biol. Chem. 268, 12764-127748; Moore et al. (1994) J. Biol. Chem. 269, 23318-23327. The binding of P-selectin to the ligand is Ca ²⁺ dependent and inhibited by treating the ligand with sialidase.

PSGL is a homodimer with two disulfide bonded subunits having a molecular weight of about 120,000 as measured by SDS-PAGE (Moore et al., 1992). Each subunit has only three N-glycans, but many clustered sialated O-glycans [Moore party, (1992); Nordgard, et al. (1993) J. Biol. Chem. 268, 12764-12774. The extracellular domain of PSGL-1 is very extended, which is characteristic of mucin-like proteins. PSGL-1 comprises a continuous 10-mer repeat (15-mer repeat for human medullary cells HL-60 and 16-mer repeat for human leukocytes) and has a type of extracellular domain containing many serine, threonine and proline 1 cell membrane protein [Sako et al., (1993) Cell 75, 1179-1186; Moore et al. (1995) J. Cell Biol. 128, 661-671; Veldmann's group, (1995) J. Biol. Chem. 270, 16470-16475. Following the 18-residue signal peptide is a 19-41 residue propeptide, which is synthesized in leukocytes and then removed from PSGL-1 [Sako et al., (1993); Bacino party, (1995) J. Biol. Chem. 270, 21966-21974. The extracellular domain of the processed mature protein is between 42 and 318 residues, followed by a 25-residue permeable membrane domain in the 69-resid cytoplasmic tail. The protein has many unmodified sialic residues as well as the sLe ^x antigen. The ligand contains one or more PNGaseF-sensitive N-linked glycans that P-selectin is not required to recognize. In contrast, the protein has a clustered sialated O-linked fructose that makes the polypeptide backbone sensitive to degradation by the O-sialglycoprotease enzyme. Treatment of intact HL-60 cells with the enzyme removes the high affinity binding site of P-selectin without affecting the overall surface expression of sLe ^{x [} Yushiyama et al. (1993), J. Biol. Chem. 268, 15229-15237], preventing the attachment of cells to immobilized P-selectin (Norgard et al., (1993) J. Biol. Chem. 268, 12764-12774; Steininger group, Biochem. Biophys. Res. Commun. (1992) 188, 760-766. The data suggest that sialomucin with only a small portion of the cell surface sLe ^x corresponds to a functionally important high affinity binding site for P-selectin in human bone marrow cells.

Sarko et al. (1993) isolated cDNA derived from human HL-60 cells encoding PSGL-1. The cDNA-derived sequence for each subunit of PSGL-1 predicts a Type 1 permeable membrane protein of 402 amino acids. The extracellular domain has an N-terminal signal peptide of residues 1-18 and putative propeptide of residues 19-41. When this propeptide is cleaved, the extracellular domain of the mature protein is 42-308 residues. The sequence eventually has 25 residues of the permeable membrane domain and 69 residues of the cytoplasmic tail. The extracellular domain is rich in serine and threonine, which are potential sites for O-glycosylation. The extracellular domain has one cysteine capable of promoting dimer formation and three potential positions for the addition of N-linked fructose, as well as three potential tyrosine sulfide positions at 46, 48 and 51 residues.

Although conventional studies require sialization and fucoseization of PSGL-1 to bind P-selectin, post-translational modification of PSGL-1 may also be important. PSGL-1 is highly O-glycosylated and contains sialated and fucoated O-linked poly-N-acetyllactosamines, including several glycans ending at sLe ^x (Moore et al., 1994, J.). Biol. Chem. 269, 23318-23327). sLe ^x or related glycans are not sufficient for PSGL-1 to bind high-friendly to P-selectin. For example, sulfate compounds without sialic acid or fucose inhibit the attachment of leukocytes to P-selectin [Norgard-Schnitz et al., (1993) Science 261, 480-483; Nelsson group, (1993) Blood 82, 3253-3258; Seconi Group, (1994) J. Biol. Chem. 269, 15060-15066; Skinner group, (1989) Biochem. Biophys. Res. Comm. 164, 1373-1379. The data suggest that sulfidation of PSGL-1 may be required for high friendly binding to P-selectin. Recombinant PSGL-1 expressed in COS cells with fucosyltransferases interacts with P-selectin in cell adhesion experiments (Saco et al., 1993). However, the similar degree of fructose of recombinant PSGL-1 and fructose of the original glycoprotein ligand is unknown.

Although carbohydrates of PSGL-1 are important in binding to selectin, the detailed chemical structure of glycans is not available. Much information has been obtained on the glycosylation of molecules through enzymatic treatment of the original ligand and the study of recombinant PSGL-1 expressed in various cell types. While these indirect methods can provide valuable information about the critical determinants of the ligand, detailed structural information about O-glycans from the original PSGL-1 identifies the glycans that are important for the function of the ligand, and PSGL While -1 is a ligand for P- and E-selectin, other mucins such as CD43 are essential in providing a clearer understanding of what is not.

Accordingly, the present invention generally relates to a method for producing glycated and sulfided PSGL-1.

The present invention also relates to methods and reagents for inhibiting the binding of PSGL-1 to P-selectin and other selectins.

Moreover, the present invention relates to novel O-glycans useful for modifying the binding to P-selectin.

Summary of the Invention

Binding to P-selectin through studies confirming the important amino acid residues at the N-terminus of PSGL-1, the importance of sulfation of tyrosine residues, and the structure and role of O-glycan carbohydrates coupled with threonine in the N-terminal region The sulfidation, glycated peptide and O-glycan structures unique to PSGL-1 have been found.

The examples show that PSGL-1 is sulfided at one or more tyrosines, primarily at residues 46, 48, 51, which are required for high affinity binding to P-selectin. These studies show that PSGL-1 synthesized in human HL-60 cells can be metabolically labeled with [ ³⁵ S] sulfate and treated with PSGL-1 with bacterial arylsulfatase to release sulfate from tyrosine to P-selectin Decreases binding to. Moreover, binding of PSGL-1 to P-selectin is inhibited by antiserum against synthetic peptides comprising the sulfidation sites of three putative tyrosines near the N-terminus of PSGL-1.

The study also found that PSGL-1 expression in PSGL-1 in recombinantly treated mammalian cells expressing two enzymes, fucoosyltransferase (FTIII) and core 2 β1-6-N-acetylglucosaminyltransferase. -1 is required for proper saccharification. Comparative experiments showed that the glycosylation of PSGL-1 expressed in cells expressing the enzyme was performed by recombinant PSGL-1 expressed in COS or CHO cells that did not express core 2 β1-6-N-acetylglucosaminyltransferase. It is more similar to the glycosylation of the original PSGL-1 than glycosylation. Ser / Thr-binding O-glycans of PSGL-1 synthesized by HL-60 cells were radiolabeled by metabolism with ³ H-sugar precursors. In a control study, O-glycans of CD43 (leucosialin), a mucin-like glycoprotein expressed by HL-60 cells, were also analyzed and compared to PSGL-1. O-glycans were isolated from Ser / Thr residues by mild treatment of purified glycoproteins with base / boronide, and glycan structures were determined by combining techniques. Contrary to expectations, PSGL-1 is not much fucoseylated; Most O-glycans are dissylated or neutral forms of Galβ1 → 4GlcNAcβ1 → 6 (Gaβ1 → 3) GalNAcOH. A small number of O-glycans are the two main species containing sialyl Lewis X antigens (one species is disialated, monofucose glycan) And the other species monosialated, trifucose glycan with polylactosamine backbone In which R is H, OH, another sugar or aglycone such as an amino acid).

The results indicate that PSGL-1 contains a unique fucose O-glycan involved in the high affinity interaction between PSGL-1 and selectin.

1A, 1B and 1C show anions using NaH ₂ PO ₄ , pH 3.0 gradient (dashed line) for PSGL-1 digested with enzymes indicating that PSGL-1 contains tyrosine sulfate sensitive to arylsulfatase A graph showing retention time in exchange chromatography. 1A shows ³⁵ S-PSGL-1 hydrolyzed with a strong base; FIG. 1B shows ³⁵ S-PSGL-1 hydrolyzed with a strong base and sham-treated with 1000 mU boiled arylsulfatase; 1C shows ³⁵ S-PSGL-1 hydrolyzed with strong acid and treated with 1000mU of active enzyme. The retention times of tyrosine, tyrosine sulphate and free sulphate are shown. Free sulfate eluted at 40.0 minutes. Sulfated monosaccharides (Gal-6-sulfate, GlcNAc-6-sulfate, GalNAc-6-sulfate and GalNAc-4-sulfate) eluted at similar retention times of 14-15 minutes.

FIG. 2 is a graph showing the rate of decrease of Ca ²⁺ dependent binding of ¹²⁵ I-PSGL-1 to [ ³⁵ S] sulfate and P-selectin isolated from intact ³⁵ S-PSGL-1 by arylsulfatase, Tyrosine sulfate is important for the binding of P-selectin.

3A and 3B are graphs showing that anti-peptide serum for specific PSGL-1 interferes with the binding of PSGL-1 to P-selectin. ¹²⁵ I-PSGL-1 is antiserum (anti-42) against 42-56 residues of normal rabbit serum (NRS) or PSGL-1 in microtiter plates containing immobilized recombinant water-soluble P-selectin or human serum albumin (HSA). -56 amino acids) and binding was measured as a function of serum concentration as shown in FIG. 3A. Platelet-induced P-selectin was incubated with HL-60 cells in the presence of 20% NRS or anti-42-56 amino acid serum. Binding of P-selectin is measured by indirect immunofluorescence and flow cytometry, and FIG. 3B shows the cell number for P-selectin.

4A-4E show that PSGL-1 fraction P-10-2b contains glycans of minimum size with sLe ^x crystallites. Fraction P-10-2b derived from ³ H-GlcN-PSGL-1 was treated with exoglycosidase and analyzed by falling paper chromatography for 24 hours. β-depleted glycans were first desialized with neuraminidase and neutral glycans were separated from sialic acid released by ion-exchange chromatography. (A) chromatography of desialylated glycans; (B) desialylated glycans treated with β-galactosidase; (C) desialylated glycans treated with Streptomyces α1,3 / 4-fucoosidase; (D) desialylated glycans treated with β-galactosidase, β-N-acetylhexosaminidase and streptomyces α1,3 / 4-fucoosidase; (E) β-depleted glycans derived from ³ H-Fuc-PSGL-1 fraction P-10-2b were desialized with enzymes and blank chromatographed in the same experiment. The true standard shift is indicated: 5, Galβ1 → 4 (Fucα1 → 3) GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH; 4, Galβ1 → 4GlcNAβ1 → 6 (Galβ1 → 3) GalNAcOH; 3 ^* , GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH; 2, Galβ1 → 3GalNAcOH; 1, GlcNAc.

5A and 5B show that PSGL-1 fraction P-10-2a consists of sialized and trifucoseized polylactosamine-containing O-glycans. Fraction P-10-2a derived from ³ H-GlcN-PSGL-1 was treated with exoglycosidase and analyzed by falling paper chromatography for 24 hours. β-depleted glycans are first desialized with neuraminidase and the released sialic acid is endo-β-galactosidase, or a combination of β-galactosidase and β-N-acetylhexaxaminidase. It was isolated from neutral glycans before and after treatment with. (B) prior to chemical defucosation of desialylated glycans and treatment with an endo-β-galactosidase or a combination of β-galactosidase and β-N-acetylhexosaminidase; It was analyzed later. The true standard shift is indicated: 3, Galβ1 → 4GlcNAcβ1 → 3Galβ1; 3 ^* , GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH; 2, Galβ1 → 4GlcNAc; 1, GlcNAc.

6 shows that P-10-3 glycans contain neutral core-2 tetrasaccharide. ³ H-GlcN-P-10-3 was treated with neuraminidase and released sialic acid was separated from neutral glycans by QAE-Sephadex chromatography. Desialylated glycans were subjected to down-paper chromatography for 18 hours before and after treatment with β-galactosidase alone or with a combination of β-galactosidase and β-N-acetylhexaxaminidase. Analyzed. The true standard shift is indicated: 4, Galβ1 → 4GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH; e ^* , GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH; 2, Galβ1 → 4GlcNAc; 1, GlcNAc.

7 shows the structure and relative proportions of O-glycans in PSGL-1.

Pharmaceutical and Diagnostic Peptides

Based on the information obtained from the following examples, PSGL-1, as described in the above formulas and references for the following dimensions and examples, using chemical or enzymatic sulfidation and glycosylation followed by recombinant techniques or peptide synthesizers. Peptides useful for inhibiting the binding of P-selectin and other selectins to can be synthesized. The peptide preferably comprises at least one of three tyrosines (46, 48 and 51 residues of SEQ ID NO: 1) at the amino terminus of the protein, at least one of which is sulfided (-SO ₄ ), and preferably O One or more threonine or serine suitable for binding glycosylation, such as the threonine residue of amino acid residue 57 of SEQ ID NO: 1. Examples of useful minimum peptides correspond to the 48 to 57 residues of SEQ ID NO: 1 as well as conservatively substituted peptides that do not alter the binding of the peptide to P-selectin. As demonstrated in the examples, it is important that the peptide comprises Core-2, sialated, fucose O-glycans.

Amino acid sequence modifications belong to one or more of the following three classes: substitutions, insertions, or removals. Insertion includes fusion of amino and / or carboxy termini as well as intrasequence insertion of one or multiple amino acid residues. Insertions will typically be fewer insertions, such as 1 to 4 residues, than insertion of amino or carboxy terminal fusions. Immunofusion protein derivatives, such as those described in the Examples, are prepared by fusing polypeptides large enough to immunize a target sequence by cross-linking in vitro or by recombinant cell culture transformed with DNA encoding the fusion. Removal is characterized in that one or more amino acid residues are removed from the protein sequence. Typically, only about 2 to 6 residues are removed at any one position in the protein molecule. These variants are generally prepared by site specific mutagenesis at the nucleotides of the DNA encoding the protein, thereby producing the DNA encoding the variant and then expressing the DNA in recombinant cell culture. Techniques for making substitution mutations at predetermined positions of DNA having a known sequence are known, such as M13 primer mutagenesis. Amino acid substitutions are usually single residue substitutions; Insertion will generally be about 1 to 10 amino acid residues; Removal will range from about 1 to 30 residues. Removal and insertion preferably takes place in contiguous pairs, ie the removal of two residues or the insertion of two residues. The final structure can be made by substitution, removal, insertion or a combination thereof. Mutations should not put the sequence out of the reading frame, and preferably should not create complementary regions that can form secondary mRNA structures. Substitution variants are those in which one or more residues have been removed and other residues inserted in their place. Such substituents are generally prepared according to Tables 1 and 2 below and are called conservative substitutions.

TABLE 1

Acronym of Amino Acid

TABLE 2

Amino acid substitutions

Substantial changes in function or immunity can be achieved by selecting substituents that are less conservative than the examples in Table 2, i.e. (a) the structure of the polypeptide backbone at the substitution region, such as a sheet or helix structure, (b) the charge or hydrophobicity of the molecule at the target (c) The effect of maintaining the bulk of the side chains is more pronounced by selecting other residues. In general, substitutions that are expected to cause the largest change in protein properties include (a) hydrophilic residues such as serine or threonine replaced by hydrophobic residues such as leucine, isoleucine, phenylalanine, valine or alanine or vice versa; (b) cysteine or proline is substituted with another moiety or vice versa; (c) residues with positive side chains, such as lysine, arginine or histidine, are substituted with negatively charged residues such as glutamic acid or aspartic acid or vice versa; Or (d) residues having a bulk side chain, such as phenylalanine, having no side chain, such as glycine in this case or vice versa; (e) Substitution resulting from increasing the number of sulfided and / or glycosylated positions.

Substitution or elimination mutagenesis can be used to insert a site for N-glycosylation (Asn-X-Thr / Ser) or O-glycosylation (Ser or Thr). It is also desirable to remove cysteine or other susceptible residues. Removal or substitution of potential proteolytic sites, such as Arg, is accomplished by, for example, removing one of the basic residues or substituting with glutamine or histidine residues.

Post-translational modifications are the result of the action of recombinant host cells on the expressed polypeptide. Glutamine and asparagine residues are frequently deamidated after translation to the corresponding glutamic acid and aspartic acid. Alternatively, the moiety is deamidated under mildly acidic conditions. Other post-translational modifications include hydroxylation of proline and lysine; Phosphorylation of hydroxyl groups at serine or threonine residues; Methylation of o-amino groups in lysine, arginine and histidine side chains [T. Cretone, Protein: Structure and Molecular Properties, W. H. Freeman & Co., San Francisco, pp 79-86 (1983); Acetylation of N-terminal amines; And in some cases amidation of the C-terminal carboxy.

Peptide Preparation Method

Fabides can be prepared according to known techniques, as generally described in the examples of cited documents, and these techniques are specifically described herein. Since preferred peptides are all glycosylated and sulfided, it is preferred to express the peptides in a suitable mammalian host cell as described below.

Chemical synthesis

Preferred synthetic methods are Merrifield [J. Amer. Chem. Soc., 85, 2149-2154 (1963), to prepare peptides according to the solid phase synthesis technique first described. Other techniques are for example M. Peptide Synthesis (2nd edition, John Wiley & Sons, 1976) in Bodantsky's group can be found in other documents known to those skilled in the art.

Suitable protecting groups and their abbreviations that can be used in these synthesis are described above, as well as J. F. double oil. McOmi's Protective Groups in Organic Chemistry (Plenum Press, New York, 1973). Common protecting groups used in the present invention are t-butyloxycarbonyl (Boc), fluorenylmethoxycarbonyl (FMOC), benzyl (Bzl), tosyl (Tos), o-bromo-phenylmethoxycarbonyl (BrCBZ Or BrZ), phenylmethoxycarbonyl (CBZ or Z), 2-chloro-phenylmethoxycarbonyl, (2-Cl-CBZ or Cl-Z), 4-methoxy-2,3,6-trimethylbenzene Sulfonyl (Mtr), formyl (CHO) and tert-butyl (t-Bu). Trifluoroacetic acid (TFA), methylene chloride (CH ₂ Cl ₂ ), N, N-diisopropylethylamine (DIEA), N-methylpyrrolidone (NMP), 1-hydroxybenzotriazole (HOBT) , Dimethyl sulfoxide (DMSO), acetic anhydride (Ac ₂ O).

General methods for synthesizing peptides by solid phase peptide synthesis using N ^α -Boc protection:

General synthetic methods for solid phase peptide synthesis using N ^α -FMOC protection:

N-terminal acetylation of the unprotected N ^α -amino group of the peptide synthesized using the Boc or FMOC method was treated with 10% Ac ₂ O and 5% DIEA in NMP and then the peptide resin was subjected to NMP and / or CH ₂ Cl By washing with _two .

Expression system

As demonstrated in the examples below, the sulfurized and glycated PSGL-1 peptides described herein that bind P-selectin are peptides having one or more sulfates and suitable glycosylation sites that bind tyrosine residues. Studies comparing recombinant PSGL-1 expressed in mammalian cells (COS cells) expressing fucoosyltransferase and purified original PSGL-1 demonstrated that glycosylation was actually different. Thus a new expression system has been developed that glycosylates PSGL-1 more naturally. This expression system is based on a mammalian expression system that expresses two or more glycosyltransferases, namely fucosyltransferase (FTIII) and core 2 β1-6-N-acetylglucosaminyltransferase. Preferred embodiments are based on well characterized expression lines such as COS or CHO cells in which the cell line has been transfected with cDNA encoding the two enzymes. cDNAs are specified in SEQ ID NO: 2 and SEQ ID NO: 3, respectively. SEQ ID NO: 2 encodes the human α (1,3 / 1,4) fucoosyltransferase (FTIII) described by Kukosca-Ratalo group (Genes & Devel. 4, 1288-1303, 1990). SEQ ID NO: 3 shows core 2 β1-6-N-acetylglucosaminyltransferase (EC as reported by Bierhuizen and Fukuda (Proc. Natl. Acad. Sci. US 89, 9326-9330, 1992). 2.4.1.102). Previous studies have not recognized the importance of the Core 2 6-N-acetylglucoacylaminyltransferase for the modification of PSGL-1 to bind P-selectin. Alternatively, an expression system can of course be selected based on the expression of either or all enzymes, and the second enzyme is provided by transfection if necessary.

cDNAs are introduced into cells under the control of suitable promoters and, optionally, enhancers and selection sequences.

Preferred promoters that regulate transcription from vectors in mammalian host cells are a variety of sources, such as viruses such as Simian virus 40 (SV40), adenoviruses, retroviruses, hepatitas-B viruses and most preferably cytomegaloviruses. From the genome or from heterologous mammalian promoters such as the beta actin promoter. Early and late promoters of the SV40 virus are readily obtained as SV40 restriction fragments that also contain the origin of replication of the SV40 virus. Peer Group, Nature, 273: 113 (1978). The immediate early promoter of human-cytomegalovirus is readily obtained as a HindIII restriction fragment. P. J. Greenaway Group, Gene 18 355-360, (1982). Of course, promoters of host cells or related species are also useful.

Transcription of the DNA encoding the desired protein by higher eukaryotic cells is increased by inserting an enhancer sequence into the vector. Enhancers are cis-active elements of DNA, typically about 10 to 300 bp, that act on the promoter to increase transcription initiation capacity. Enhancers as well as the 5 '(L. lymines group, Proc. Natl. Acad. Sci. 78: 993) and 3' (M. L. Rusky group, Mol. Cell Bio. 3: 1108, 1983) of the transcription unit. Rather, it is a relatively orientation and position independent element found in the coding sequence itself (T. F. Osborne Group, Mol. Cell Bio. 4, 1293, 1984). Many enhancer sequences are known from mammalian genes (globin, elastase, albumin, α-fetoprotein and insulin). Examples include the SV40 enhancer at the second half of the origin of replication (100-270 bp), the initial promoter enhancer of cytomegalovirus, the polyoma enhancer at the second half of the origin of replication and the adenovirus enhancer.

Expression vectors used in eukaryotic host cells (yeast, fungi, insects, plants, animals or nucleated cells) may also contain sequences necessary for transcription termination to affect mRNA expression. These regions are transcribed as polyadenylated fragments at untranslated sites of mRNA encoding the desired protein. The 3 'untranslated region also includes a transcription termination position.

Expression vectors may contain a selection gene or selection marker. Examples of suitable markers for mammalian cells include dihydrofolate reductase (DHFR), thymidine kinase or neomycin. If the selection marker successfully enters the mammalian host cell, these transformed mammalian host cells will survive when placed under selection pressure. There are two widely used selection categories. The first category is based on the use of mutant cell lines that cannot grow without cell metabolism and adjuvant medium. Two examples are DHFR-negative CHO cells, mouse LTK cells. These cells cannot grow without nutrients such as thymidine or hypoxanthine. Because the cells lack the specific genes required for the complete nucleotide synthesis pathway, these cells cannot survive without the lacking nucleotides provided in the adjuvant medium. An alternative to the supplemental medium is to introduce intact DHFR or TK genes into cells lacking each gene to alter their growth requirements. Cells not transformed with DHFR or TK genes will not be able to survive in non-adjuvant media.

The second category is dominant selection, a selection method that is used in all cell types and does not require the use of mutant cell lines. These methods usually use drugs that arrest the growth of the host cell. These cells with new genes will express proteins that are drug resistant and will survive in selection. Examples of drugs used by the dominant selection include neomycin (P. Sudden and P. Berg, J. Molec, Appl. Genet. 1: 327, 1982), mycophenolic acid (R. S. Mulligan and P. Berg). , Science 209: 1422, 1980) or hygromycin (B. Sugden et al., Mol. Cell. Biol, 5, 410-413, 1985). The three examples each use bacterial genes under eukaryotic control to confer resistance to the suitable drug G418 or neomycin (geneticin), xgpt (mycophenolic acid) or hygromycin.

The term "amplification" refers to the increase or replication of an isolated region in the chromosomal DNA of a cell. Amplification is carried out using methotrexate (MTX), which inactivates selected reagents such as DHFR. Amplification or continuous replication of the DHFR gene produces higher amounts of DHFR by higher amounts of MTX. Amplification pressure is applied by adding even larger amounts of MTX to the medium despite the presence of endogenous DHFR. Desired genes can be amplified by cotransfecting mammalian host cells with plasmids having DNA encoding the desired protein and DHFR or amplification genes to enter together. The investigator is convinced that the cells require more DHFR and this requirement is met by replication of the selection gene, selection of cells that can grow at much higher MTX concentrations. When a gene encoding the desired heterologous protein enters with the gene of choice, the replication of the gene results in a copy of the gene encoding the desired protein. The result is increased replication of the gene encoding the desired heterologous protein, ie the amplified gene expresses more of the desired heterologous protein.

Preferred suitable host cells for expressing vectors in higher eukaryotic cells include monkey kidney CVI cell lines transformed by SV40 (COS-7, ATCC CRL 1651); Human fetal kidney cell line (293, F. L. Graham et al., J. Gen Virol. 36: 59, 1977); Baby hamster kidney cells (BHK, ATCC CCL 10); Chinese hamster ovary cell-DHFR (CHO, Urlaub and Chasin, Proc. Natl. Acad. Sci. USA, 77, 4216, 1980); Mouse sertoli cells (TM4, J. P. Matter, Biol. Reprod. 23: 243-251, 1980); Monkey kidney cells (CVI ATCC CCL 70); African green monkey kidney cells (VERO-76, ATCC CRL-1587); Human cervical tumor cells (HELA, ATCC CCL 2); Dog kidney cells (MDCK, ATCC CCL 34); Buffalo rat hepatocytes (BRL 3A, ATCC CRL 1442); Human lung cells (W138, ATCC CCL 75); Human hepatocytes (hep G2, HB 8065); Mouse breast cancer (MMT 060562, ATCC CCL 51); And TRI cells (J. P. Matter's group, Annals N.Y. Acad. Sci 383; 44-68, 1982). Host cells can be transformed with expression vectors and cultured in conventional nutrient media modified to be suitable for introducing promoters, selecting transformants, or amplifying genes. Culture conditions, such as temperature and pH, are the conditions used in the host cell selected for expression.

Preparation of diagnostic and therapeutic agents derived from protein or carbohydrate components of glycoprotein ligands for P-selectin

Glycoprotein ligands for P-selectin described above have a variety of applications as diagnostic reagents and potentially in the treatment of numerous inflammatory and thrombotic diseases.

Diagnostic reagents

Antibodies to ligands can be used for the detection of human diseases resulting from defects in P-selectin ligands. This disease is often observed in patients with increased susceptibility to infections in which leukocytes cannot bind to active platelets or endothelial cells. Experimental cells, generally leukocytes, are collected and screened by standard medically approved techniques. Detection systems include ELISA methods, binding radiolabeled antibodies to immobilized active cells, flow-cytometry, immunoperoxidase or immunogold assays, or other methods known to those of skill in the art. .

Antibodies to specific protein or carbohydrate components of the ligand may be used to identify defects in the expression of the core protein or in glycosyltransferases and / or modifying enzymes that form suitable fructose chains of the protein. Antibodies can also be used to screen cells and tissues in addition to leukocytes that express a protein or carbohydrate component of a ligand for P-selectin.

The complementary DNA clones encoding the protein component of the ligand can be isolated and sequenced. These cDNA probes can be used as diagnostic reagents for testing the expression of RNA transcripts of ligands in leukocytes and other tissues by standard methods such as Northern blotting of RNA isolated from cells and in situ hybridization of resected tissue.

Similar methods can be used to identify qualitative or quantitative diseases of P-selectin itself. Glycoprotein ligands, carbohydrates, or suitable derivatives thereof are labeled to test their ability to bind to P-selectin of activated platelets obtained from patients with diseases due to defects in P-selectin.

The ligand, or component thereof, can also be used for the analysis of P-selectin, which binds to a screen for compounds that interfere with the ligand's interaction with P-selectin.

Clinical application

Clinically, compounds that interfere with the binding of P-selectin and / or other selectins, including L-selectin, since P-selectin has several functions related to leukocyte adhesion, inflammation, tumor metastasis and coagulation, Carbohydrates, for example, can be used to control the reaction. These compounds include antibodies to PSGL-1 or fragments thereof, PSGL-1 or fragments thereof. For example, glycoprotein ligands, or compounds thereof, particularly carbohydrate fragments, can be used to inhibit leukocyte adhesion by competitively binding to P-selectin expressed on the surface of activated platelets or endothelial cells. Similarly, antibodies to ligands can be used to interfere with P-selectin mediated cell adhesion by competitively binding to P-selectin ligands of leukocytes or other cells. The therapy is useful in sudden situations where effective but temporary suppression of leukocyte-mediated inflammation is required. In addition, treatment of chronic diseases can also be achieved by continuous administration of reagents, such as subcutaneous or oral administration.

Inflammatory responses can damage the host if not checked because white blood cells release many toxic molecules that can damage normal tissue. These molecules include proteolytic enzymes and free radicals. Examples of pathological situations in which leukocytes can cause tissue damage include injury from ischemia and reperfusion, bacterial sepsis and diffuse intravascular coagulation, adult respiratory distress syndrome, tumor metastasis, rheumatoid arthritis and arteriosclerosis.

Reperfusion injury is a major problem in clinical cardiology. Therapies that reduce leukocyte adhesion in ischemic myocardium can greatly enhance the therapeutic efficacy of thrombolytics. Treatment of thrombolysis with reagents such as tissue plasminogen activators or streptokinase can reduce coronary artery occlusion in many patients with severe myocardial ischemia prior to irreversible cardiomyocyte death. However, many of these patients still suffer cardiomyopathy despite blood flow recovery. This "reperfusion injury" is probably associated with the attachment of leukocytes to vascular endothelial cells in the ischemic region, partly because of the activity of platelets and endothelial cells by thrombin and cytokines attached to leukocytes (Romson's party). , Circulation 67: 1016-1023, 1983). These adherent leukocytes can travel through endothelial cells and destroy the ischemic myocardium, which is rescued by blood flow recovery.

Ischemia and reperfusion seizures; Mesenteric and peripheral vascular diseases; Organ transplantation; And numerous other clinical diseases that cause organ damage mediated by the attachment of leukocytes to the blood vessel surface, including circulatory shock (in which case many organs may be damaged after blood flow recovery).

Bacterial sepsis and diffuse vascular coagulation often exist simultaneously in emergency patients. It is associated with the production of thrombin, cytokines and other inflammatory mediators, the activity of platelets and endothelial cells, and adhesion of leukocytes and platelet aggregation through the vascular system. Leukocyte-dependent organ damage is an important feature of this condition.

Adult respiratory distress syndrome is a destructive lung disease that occurs in patients with sepsis or subsequent trauma and is associated with extensive attachment and aggregation of leukocytes in the pulmonary circulation. This causes large amounts of plasma to extravasate into the lungs and destroy lung tissue, all of which are mostly mediated by white blood cell products.

Two often fatal associated lung diseases are in immunosuppressed patients who have undergone allogeneic bone marrow transplantation and cancer patients who suffer from complications due to general vascular leakage resulting from treatment of interleukin-2 treated LAK cells (lymphokine-active lymphocytes). LAK cells are known to adhere to the vessel wall and release products that are toxic to endothelial cells. Although the mechanism by which LAK cells attach to endothelial cells is unknown, the cells will probably release molecules that activate endothelial cells and bind to endothelial cells in a similar mechanism to that acting on neutrophils.

Tumor cells of many malignant tumors (including carcinomas, lymphomas and sarcomas) can metastasize to distant locations through the pulmonary system. The attachment of tumor cells to endothelial cells and their subsequent migration mechanisms are unknown, but in at least some cases will be similar to the mechanism of white blood cells. Specifically, certain carcinoma cells include E-selectin (Rice and Bevilacqua, Science 246: 1303-1306, 1991) and P-selectin (Arupo et al., Proc. Natl. Acad. Sci, USA, 89: 2292-2296). , 1992; Stone and Wagner, I. Clin. Invst. 92: 804-813, 1992). The association between platelets and metastatic tumor cells is well documented, suggesting the role of platelets in the spread of some cancers. Because P-selectin is expressed in activated platelets, it is believed that P-selectin is involved in the association of platelets with at least some malignancies.

Platelet-leukocyte interaction is thought to be important in atherosclerosis. Platelets will serve to augment monocytes into atherosclerotic plaques; The accumulation of monocytes is the first detectable sign of atheromatous formation. Rupture of fully mature plaques leads to the activation and enhancement of platelet deposition and thrombus formation, as well as early gradation of neutrophils into the ischemic region. P-selectin is also inappropriately expressed on the surface of endothelial cells of atheromatous masses, where it mediates the initial adhesion of monocytes to migrate into the atheromatous mass (Johnson-Tidi et al., Am. J. Path. 144: 952-961, 1994). .

Another possible application is the treatment of rheumatoid arthritis and other autoimmune or inflammatory diseases.

In these clinical applications, the peptides or isolated O-glycans, or fragments thereof, described herein can be administered to prevent selectin-dependent interactions by competitively binding to P-selectin expressed in active cells. In addition, antibodies to carbohydrate components of proteins and / or ligands, or fragments thereof, may be administered. The antibody is preferably one that is of human origin or modified to remove a site that is susceptible to causing an immunogenic response.

The peptides described herein also include inorganic acids such as hydrochloric acid, hydrobromic acid, perchloric acid, nitric acid, thiocyanic acid, sulfuric acid and phosphoric acid; By reaction with organic acids such as formic acid, acetic acid, propionic acid, glycolic acid, lactic acid, pyruvic acid, oxalic acid, malonic acid, succinic acid, maleic acid and fumaric acid or inorganic bases such as sodium hydroxide, ammonia hydroxide and potassium hydroxide; And pharmaceutically acceptable acid- or base-addition salts formed by reaction with organic bases such as mono-, di-, trialkyl and aryl amines and substituted ethanolamines.

The carbohydrate component (O-glycan structure or component thereof) of the ligand or antibody in a suitable medical carrier is preferably administered intravenously when immediate relief is needed. Carbohydrates can also be administered to the muscle, abdominal cavity, subcutaneous, oral cavity as a carbohydrate itself, conjugated to a carrier molecule or by a drug transport system. Carbohydrates can be chemically modified to increase half-life in vivo.

Carbohydrates may be isolated from cells expressing carbohydrates as a result of genetic engineering, as described in the examples of naturally or transfected mammalian cells, or preferably by synthetic methods. Such methods are known to those skilled in the art. In addition, a number of additional glycosyltransferases have been cloned (J. C. Paulson and K. J. Coli, J. Biol. Chem. 264: 17615-17618,1989). Thus, one of ordinary skill in the art can use a combination of synthetic chemistry and enzyme synthesis to prepare pharmaceutical or diagnostic reagents.

Carbohydrates with biological activity are carbohydrates that inhibit the binding of white blood cells to P-selectin. Suitable medical administration devices for the patient are known to those skilled in the art. For parenteral administration, carbohydrates are usually dissolved or suspended in sterile water or saline. For oral administration, carbohydrates will be incorporated into tablet, liquid or capsule inert carriers. Suitable carriers may be starch or sugar and may include lubricants, flavors, binders, and other materials of the same nature. Carbohydrates may also be administered locally to the wound or inflamed area by topical application of a solution or cream.

Alternatively, carbohydrates may be administered in, on or as part of liposomes or microspheres (or microparticles). Methods for preparing liposomes and microspheres for patient administration are known to those skilled in the art. U.S. Patent 4,789,734 describes a method for encapsulating a biological material in liposomes. In essence, the material is dissolved in aqueous solution, suitable phospholipids and added lipids with surfactant and dialysis or sonicated material, if necessary. A good overview of known methods is given. It is described in Chapter 14, "Liposomes" in the book Gregoridis, "Drug Carriers in Biology and Medicine" (pp. 287-341, Academic Press, 1979). Microspheres made of polymers or proteins are known to those skilled in the art and can be prepared to enter the bloodstream directly from the gastrointestinal tract. Alternatively, carbohydrates can be incorporated and the microspheres, or a mixture of microspheres, can be injected to release slowly over a long period of time over days to months. See, eg, US Pat. Nos. 4,906,474, 4,925,673 and 3,625,214.

Carbohydrates must be active when administered parenterally in an amount of at least about 1 μg / kg body weight. In the treatment of most inflammatory diseases, the dosage will be between 0.1 and 30 mg / kg body weight. In some embodiments, the dosage of some characteristic carbohydrates may be 70 mg / kg.

Criteria for evaluating responses according to the type of treatment using antibodies or carbohydrates will follow certain conditions and will generally follow standard medical practice. For example, the criteria for an effective dosage to prevent the expansion of myocardial infarction will be determined by one of skill in the art by observing marker enzymes of myocardial necrosis in plasma or by monitoring electrocardiograms, important signals and clinical responses. In the treatment of acute respiratory distress syndrome, improvements in dissolution of intraarterial oxygen, pulmonary infiltrates, and clinical improvement measured by reduced respiratory distress and resuscitation will be investigated. In the treatment of shock (hypertension) patients, the effective dosage will be based on the clinical response after recovery of blood pressure and the specific measurement of important organs such as liver and kidney function. Nerve function will be monitored in seizure patients. Special tests are used to monitor the function of the transplanted organs, such as serum creatinine, urine flow, and serum electrolyte in kidney transplant patients.

The invention will be further understood by reference to the following non-limiting examples:

Example 1 Measurement of Structure and Function of Fructose Bound to PSGL-1

The effect of various glycosidases on the structure and function of PSGL-1 from human neutrophils was measured. Unlike recombinant PSGL-1 expressed in COS cells with FTIII, the molecule is usually resistant to the treatment of endo-α-N-acetylgalactosaminiase, which binds to serine or threonine. Simple core 1 Galβ1-3GalNAc disaccharide has little to no effect. Human neutrophil PSGL-1 mainly reveals sLe ^x and its bisialated counterpart Le ^x to O-linked poly-N-acetyllactosamine. These data indicate that PSGL-1 in human neutrophils is a series of sialated, fucose-linked poly-N-acetyllactoses that P-selectin preferentially recognizes over recombinant PSGL-1 expressed in COS cells with FTIII Implies the provision of citizens. To further characterize the structure and function of the attached fructose, the effect of glycosidase on radioiodide PSGL-1, purified from human neutrophils, was investigated. PSGL-1 was found to have poly-N-acetyllactosamine, only some of which could be removed using endo-β-galactosidase. Most of the Le ^x and sLe ^x structures were on the endo-β-galactosidase-sensitive chain. Treatment of peptide: N-glycosidase F (PNGaseF) removed two or more of the three possible N-linked fructose from PSGL-1. The expression of Le ^x and sLe ^x did not change to an appreciable level by treatment with PNGaseF, indicating that these structures are mainly present in O-linked poly-N-acetyllactosamine. Endo-β-galactosidase-treated PSGL-1 had the ability to bind to P-selectin, in which some of the fructose recognized by P-selectin was enzyme-resistant poly-N-acetyllactosamine Present in or on a chain deficient in poly-N-acetyllactosamine. PNGaseF treatment did not affect the ability of PSGL-1 to bind P-selectin, demonstrating that the fructose required for P-selectin recognition is O-linked. These data demonstrated that PSGL-1 from human neutrophils exhibits complex sialated and fucoseylated O-linked poly-N-acetyllactosamines that enhance high affinity binding to P-selectin.

Materials and methods

Abbreviations used are as follows:

CHO, Chinese Hamster Ovary;

Con A, Concanavalin A;

HSA, human serum albumin;

Le ^x , Lewis X;

PBS, phosphate-buffered saline;

PAGE, polyacrylamide gel electrophoresis;

PNGaseF, peptide: N-glycosidase F;

PSGL-1, P-selectin glycoprotein ligand;

sLe ^x , Sialyl Lewis X;

tES, truncated soluble E-selectin;

tPS, cleaved soluble P-selectin;

WGA, wheat germ agglutinin.

Materials: Triton X-100, Brij-58, and sialidase from Arthrobacter ureafaciens (75 U / mg, EC 3.2.1.18) were obtained from CalBio Chem-Behring Corp. (La Jolla, Calif.) Obtained from. Trade names Iodobeads, trade name 3M Emphaze Biosupport Medium, and protein A Sepharose CL4B were obtained from Pierce Chemical Co. (Rockford, Ill.). Endo-β-galactosidase (5U / mg, EC 3.2.1.103) of Escherichia freundii was obtained from V-Labs, Inc. (Covington, Los Angeles, USA). Α1,3 / 4-L-fucoosidase of Streptomyces sp. 142 (EC 3.2.1.51) was obtained from PanVera Corp. (Madison, WI) and Jack Bean's β-N-acetylglucoosaminidase (EC 3.2.1.30, 53 U / mg) was obtained from Sigma Chemical Co. (St. Louis, MO). Recombinant peptide: N-glycosidase F (EC 3.2.2.18, 25,000 U / mg, PNGaseF) and endo-α-N-acetylgalactosaminiase (EC) of Diplococcus pneumoniae 3.2.1.97, trade name O-glycanase) was purchased from Genzyme Corp. (Cambridge, Mass.). Concanavalin A Sepharose (Con A, 10 mg / ml resin) was obtained from Pharmacia / LKB (Uppsala, Sweden). Wheat germ agglutinin (WGA) agarose (7.6 mg / ml resin) and tomato lectins were obtained from Vector Laboratories, Inc., Burlingham, Vermont, USA. Tomato lectin was coupled with cyanogen bromide-activated sepharose at a density of 2 mg / ml resin in the presence of 7.5 mg / ml chitotriose.

Antibodies: Anti-human P-selectin mAb (monoclonal antibodies) S12 and G1 were prepared and characterized as described by McEver and Martin (1984), Wong et al. (1990). Anti-human E-selectin mAb CL2 and CL37 (Mulligan et al., J. Clin. Invest. 88, 1396-1406, 1991) were Provided by Wayne Smith (Baylor College of Medicine, Houston, Texas). LeuM1 (CD15, IgM) was purchased from Becton-Dickinson & Co. (San Jose, CA). CSLEX-1 hybridomas (HB 8580) were purchased from ATCC, and IgM mAb was sepa- cal in PBS with boric acid precipitation and pH 7.4, as described by Shatyl et al. (J. BIol. Chem. 260, 11107-11114, 1985). Purification from ascites was carried out by gel filtration on Rhodes CL4B. Goat anti-mouse μ-chain specific IgG and MOPC104E (IgM) were purchased from Cappel-Organon Technika (Durham, North Carolina).

Peptide Antiserum Preparation and Immunoprecipitation: Peptides corresponding to residues 42-56 (QATEYEYLDYDFLPE C ) (SEQ ID NO: 1) and 354-376 ( C ISSLLPDGGEGPSATANGGLSK) (SEQ ID NO: 1) of the cDNA-derived amino acid sequence of PSGL-1 are described in Applied Biosystems Model 431 Synthesis was carried out in a peptide synthesizer. Peptides 42-56 were coupled to maleimide-activated keyhole limpet hemocyanin (manufactured by Pierce Chemical Co.) via an added cysteine and injected into New Zealand white rabbits (provided by the Montana State University Animal Resources Center). . Immune serum was collected and used to immunoprecipitate ¹²⁵ I-PSGL-1. Protein beads (25 μl, 50% suspension) were previously incubated with 1 μl of 1 or 4 dilutions of normal or immune rabbit serum at 37 ° C. for 90 minutes. The beads were then washed once with 0.1 M NaCl, 20 mM MOPS, pH7.5, 0.1% Triton X-100 (MBS / 0.1% Triton). Purified ¹²⁵ I-PSGL-1 or radioiodide-WPG eluate was then incubated with beads in the presence or absence of 0.2 mg / ml immunopeptide or PSGL-1 peptide 354-376 acting as a negative control. After 90 minutes of incubation at 37 ° C., the beads were washed five times with MBS / 0.1% Triton and eluted by boiling for 5 minutes with SDS sample buffer (Laemley, Nature 227, 680-685, 1970). Immunoprecipitates were analyzed by SDS-PAGE, followed by autoradiography.

Production of CHO D ^- cells stably expressing the soluble form of E-selectin: To modify the 5 'and 3' ends of the E-selectin to produce high levels of soluble recombinant E-selectin in CHO D ^- cells, An expression plasmid containing full length cDNA for human E-selectin (ELAM-1: BBG 57, British BioTechnology Products, Ltd.) was used as a template in the polymerase chain reaction. Oligonucleotide pairs used in polymerase chain reaction (5 'GTC CCT CTA GAC CAC CAT GAT TGC TTC ACA GTT T + 5' TCC CGG TCG ACT TAG GGA GCT TCA CAG GT) (SEQ ID NOS: 4 and 5, respectively) Introduce the Xba I position and the complete Kozak sequence (Kozak, Nucleic Acids Res. 9, 5233-5252, 1981) before the initiation codon, and the 6th consensus repeat and the full-length E-selectin It was designed to introduce the Sal I position following the termination codon introduced after amino acid 550 located between the transmembrane regions. DNA sequencing confirmed the introduction of residue cDNA of E-selectin as well as the changes caused by polymerase chain reaction. The truncated E-selectin (tES) gene was then introduced as an Xba I-Sal I fragment into a mammalian expression vector pDSRα2 (European Patent Application A20398753) modified to include unique Xba I and Sal I restriction sites. tES expression vectors were transfected into a CHO cell line (DHFR-minus CHO) deficient in dihydrofolate reductase (Urlaub et al., Proc. Natl. Acad. Sci. USA, 77, 4216-4220, 1980). The transformants were selected in medium lacking hypoxanthine and thymidine (Burdrel's group, Protein Exp. Purif. 4, 130-140, 1993). RNase protection assays were used to screen for transformants expressing high levels of E-selectin-specific mRNA (Turner et al., Blood 80, 374-381, 1992).

Protein Purification: PSGL-1 was purified from human neutrophil cell membranes prepared by the following modification method, as described by Moore et al. (1992). Neutrophil cell membranes extracted with 5% Triton X-100, 20 mM MOPS, pH 7.5, 0.02% NaN ₃ in 0.1 M NaCl were added to WGA-agarose, and the bound protein was eluted with 0.5 M N-acetylglucosamine. (WGA eluate). After extensive dialysis with 0.1M NaCl, 20mM MOPS, pH7.5, 0.02% NaN ₃ , 0.02% Brij-58, 2mM CaCl ₂ , 2mM MgCl ₂ (equilibration buffer), the WGA eluate was digested with P-selectin ( tPS) -Emphaze column (10 mg lectin / ml resin). The column was washed extensively with equilibration buffer and eluted with 5 mM EDTA. Mono Q PC pooled with ligand-containing fractions and equilibrated with 0.1 M NaCl, 20 mM MOPS, pH 7.5, 2 mM EDTA, 0.02% NaN ₃ , 0.02% Brij-58 using the SMART SMART Sepatation System (Pharmacia / LKB) It was added to a 1.6 / 5 column. The column was developed with a 2-ml linear gradient (0.1-1.0 M NaCl) at 50 μml / min. Some of the fractions were iodized with Na ¹²⁵ I using the trade name Iodobeads according to the manufacturer's instructions. Purity of the fractions was analyzed by SDS-PAGE and autoradiography of the iodide fractions.

Soluble recombinant form P-selectin (tPS) cleaved after the ninth consensus repeat was purified by immunoaffinity from media treated with 293 cells permanently transfected as described by Yushiyama Group (1993). As measured by sedimentation rate and equilibrium analysis, tPS was an asymmetric monomer having a molecular weight of 103,600 Da. E ₂₈₀ ^1% for tPS was 12.3.

Cleaved recombinant soluble E-selectin (tES) was purified by immunoaffinity from media treated with CHO D ⁻ cells stably secreting tES. The cells were cultured in the same amount as Dulbecco's modified Eagle's medium (DMEM) and Ham's F-12 nutrient mixtures supplemented with 10% fetal bovine serum, 10 U / ml penicillin and 10 mg / ml streptomycin. In the confluence state, the cells were transferred to serum-free medium and the treated medium was harvested every 7 days.

The treated medium was clarified and sterilized using two Sartorius filters (10 μm and 0.2 μm) connected in series and concentrated 50 times (Filtron Maximate, 10 kDa MW _CO ). The anti-E-selectin mAb H18 / 7 (Bevilaquaa et al., Proc. Natl. Acad. Sci. USA, 84, 9238-9242, 1987) is a cyanogen bromide-activated cepharo at a density of 6 mg / ml resin. Coupling with Os. The H18 / 7-Sepharose column was equilibrated at 4 ° C. with PBS at pH 7.5 containing 1 mM CaCl ₂ and 1 mM MgCl ₂ , and 500 ml of concentrated medium was added to the column (V _{T) at} a flow rate of 10 cm / hour. = 200 mL). The column was eluted with 0.1 M glycine at pH 2.7 and the fractions were neutralized rapidly by adding Tris-HCl at pH 7.5 to a final concentration of 0.1 M. The eluate was dialyzed with 25 mM sodium phosphate at pH 7.5 and added to a Q-High Performance anion exchange column (V _T = 200 mL, Pharmacia) equilibrated with the same buffer. The column was developed with a linear gradient of NaCl (0.17-0.22M). Purity was assessed by SDS-PAGE and N-terminal sequence analysis. Protein concentrations were determined using E ₂₈₀ ^1% = 12.5 calculated using the method of Jill and Bonn Hippel (Anal. Biochem. 182, 319-326, 1989). The yield of tES was 250 mg per 500 ml of concentrated treatment medium.

Sedimentation Equilibrium Analysis of Soluble tES: Sedimentation equilibrium experiments were performed using a four-hole titanium rotor, a single-column Kel-f centerpiece, and a sapphire window in a Beckman Optima XLA analyte ultracentrifuge with on-line Rayleigh interference optics. . Data were obtained at rotor speeds of 10,000, 15,000, 20,000 and 25,000 rpm at 20 ° C. using television-based cameras and on-line capture and analysis systems [T.M. Laue, 1981, PhD Thesis, University of Connecticut, Storres, CT; Laue et al., 1992, Analytical Ultracentrifugation in Biochemistry and Polymer Science (ed by S. Harding, A. Loewe and J. C. Horton) pp. 90-125, Royal Society of Chemistry, London. Samples were added at concentrations of about 1.0, 0.5, 0.25 and 0.13 mg / ml in PBS at pH 7.4. After the estimated time for equilibrium, the equilibrium was checked from time to time by collecting data and subtracting successive scans (D. E. Pantis, 1964, Biochemistry, 3, 297-317). The data in the optical window was selected using the REEDIT program (supplied by David Ifantis). Data were analyzed to estimate the monomer molecular weight of tES using the NONLIN program (D. E. Pantis, 1964, Biochemistry, 3, 297-317). Three or more settling equilibrium data obtained at different loading concentrations, spinning positions and angular velocities of tES were fitted by simultaneous nonlinear least-squares method. To calculate molecular weight, the partial specific volume of tES was estimated to be 0.688 ml / g (error ± 0.03 ml / g), accounting for the uncertainty in the carbohydrate composition of tES (Laue et al., 1992).

Sedimentation Rate Analysis of Soluble tES: Sedimentation rate experiments were performed on a Beckman Optima XLA analyte ultracentrifuge using Rayleigh interference optics. Experiments were carried out at 20 ° C., 60,000 rpm using a 4-hole titanium rotor, 2-channel, charcoal-filled, epon centerpiece and sapphire window. Samples were loaded at a concentration of 0.5 mg / ml in PBS at pH 7.4. Sedimentation coefficient is W. It was measured by the time derived from the concentration graph as described by Stafford (1992, Anal. Biochem. 203, 295-301).

Protein Iodide: mAb G1 and CL2 (200 μg) were iodinated with 400 μCi of carrier-free Na ¹²⁵ I (ICN Biomedical, Inc., Costa Mesa, Calif.) Using the tradename Iodobeads. Free ¹²⁵ I was removed by gel filtration through a Sephadex G-25 column (PD-10, Pharmacia) equilibrated with 0.1 M NaCl, 20 mM MOPS, pH 7.5, 0.1% Triton X-100. The labeled antibody was then centrifuged for 30 minutes at 90,000 × g in a TL-100 ultracentrifuge (manufactured by Beckman, Palo Alto, CA). The concentration of labeled protein was determined by Micro BCA Protein Assay Kit (Pierce) using each unlabeled protein as reference. The range of specific activity of the labeled antibodies was 0.5-1.0 μCi / μg protein.

PSGL-1 (about 0.5 μg) or WGA eluate (about 100 μg) was iodinated with 400 μCi Na ¹²⁵ I and salts were removed as described above. Radiolabeled PSGL-1 was concentrated to about 200 μl using a Centicon-30 device (Pharmacia / LKB) and 0.15 M NaCl, 20 mM MOPS, pH7.5 using the SMART SMART Sepatation System (Pharmacia / LKB). , Gel filtered on a Superose 6 PC 3.2 / 30 column equilibrated with 0.02% NaN ₃ , 0.1% Brij-58. The range of specific activity of the radiolabeled ligand was 14-30 μCi / μg protein. Labeled proteins were generally higher than 95% which could precipitate with trichloroacetic acid.

Binding of ¹²⁵ I-PSGL-1 to fixed tPS or tES: tPS or tES (50 μl, 10 μg / ml) in HBSS (HBSS / Az) with 0.02% NaN ₃ was microtiter plate (trade name Immulon I Removawell). Strips, Dynatech Laboratories, Inc., Chantilly, Va.) Was incubated overnight at 4 ° C. After washing three times with HBSS / Az, the wells were blocked with HBSS / Az / 1% human serum albumin (HSA) at 22 ° C. for 2 hours. Wells were then washed once with HBSS / Az and ¹²⁵ I-PSGL-1 diluted with HBSS / Az / 0.1% HSA was added (50 μl, 5,000-10,000 cpm / well). After 1 hour at 22 ° C., the wells were quickly washed five times with HBSS / Az using a 12-channel plate washer (trade name Nunc-Immuno Wash 12, Nunc Inc., Naperville, Ill.). Each well was then counted in a γ-counter. All analyzes were performed in four replicates. In some assays the binding of ¹²⁵ I-PSGL-1 to immobilized tPS was measured with increasing concentrations of liquid tPS or tES.

Binding of ¹²⁵ I-PSGL-1 to Immobilized Anti-Carbohydrate Antibodies: Goat anti-mouse μ-chain specific IgG (50 μl, 10 μg / ml) in HBSS / Az at 4 ° C. overnight at Immulon I Removawell Strips Incubated. Wells were washed and blocked as described above. CSLEX-1, LeuM1 or MOPC104E was then added (50 μl, 5 μg / ml). After 1 hour incubation at 22 ° C., the wells were washed three times with HBSS / Az and ¹²⁵ I-PSGL-1 (50 μl, 5,000-10,000 cpm / well) in HBSS / Az / 0.1% HSA was added. After 1 hour, the wells were washed as described above and each well was counted in a γ-counter. All analyzes were performed in four replicates.

Enzymatic digestion of ¹²⁵ I-PSGL-1: sialidase (0.2 U / ml, 15 h), endo-β-galactosidase (2 U / ml, 15 h), and α1,3 / 4-L-fuco Decomposition using an ossidase (0.32 mU / ml, 15 h) was performed at 0.1 M NaCl, 50 mM sodium acetate, pH5.5. Decomposition with PNGaseF (40 U / ml, 15 h) was performed at 0.1 M sodium phosphate, pH8.6. PSGL-1 generally does not preprocess because PNGaseF has a similar effect regardless of SDS pretreatment. Subsequent degradation with sialidase (0.2 U / ml, 2 hours) followed by endo-α-N-acetylgalactososaminidase (0.1 U / ml, 15 hours) was achieved with 0.1 M NaCl, 50 mM sodium acetate, It was carried out at pH6.0. Degradation with all glycosidases was carried out at 37 ° C. in the presence of 20 μM leupetin, 30 μM antipinein, 1 mM benzamidine and 0.02% NaN ₃ .

Lectin affinity chromatography: Concanavalin Con A Sepharose column equilibrated with 0.1M NaCl, 20 mM Tris, pH7.5, 2 mM CaCl ₂ , 0.1% Brij-58, 0.02% NaN ₃ (V _T = 1 mL) was added ¹²⁵ I-PSGL-1. The column was washed with 5 column volumes of equilibration buffer and eluted with 0.5M α-methylmannoside in equilibration buffer preheated to 65 ° C. Tomato lectin chromatography equilibrates the column (V _T = 1 ml) with 0.1 M NaCl, 20 mM MOPS, pH7.5, 0.1% Brij-58, 0.02% NaN ₃ , and 20 mg / ml chitotriose in equilibration buffer. It was carried out similarly except eluting with. More than 85% of the count loaded on the lectin column was obtained.

Position Density Determination: mAb for P-selectin (G1) or E-selectin (CL2), which inhibits interaction with bone marrow cells, was used for position density measurements. Wells were coated with tPS or tES (50 μl, 10 μg / ml) and blocked as described above. Saturated concentrations of ¹²⁵ I-labeled mAb (2 μg / ml) in HBSS / Az / 0.1% HSA were added to the wells and incubated at 22 ° C. for 1 hour. The wells were washed five times and then counted in a γ-counter. Unbound was defined as cpm bound to the selectin-coated well minus cpm bound to the HSA-coated well. Site density was calculated assuming monovalent binding of the antibody at saturation. All analyzes were performed in four replicates.

Neutrophil Adhesion Assay: Human neutrophils were collected from spontaneous donors by dextran sedimentation, hypotonic degradation, and Ficoll-Hippark density gradient centrifugation as described by Zimmerman's group (1985, J. Clin. Invest. 76, 2235-2246). Heparin was isolated from treated blood by agreement. Quantification of neutrophil attachment to tPS or tES (50 μl, 10 μg / ml) immobilized on the brand name Immulon I Removawell microtiter plate was carried out as described by Zune et al. (1990).

result

Antisera against PSGL-1 peptide recognized P-selectin glycoprotein ligand from human neutrophils. Polypeptide cores of sialomusine ligands for P-selectin characterized in human neutrophils and HL-60 cells were immunized with glycoprotein ligand PSGL-1 for P-selectin identified by cloning expression from HL-60 cDNA library To determine if it is scientifically relevant, antisera against peptides corresponding to 42-56 residues of the cDNA-derived amino acid sequence of PSGL-1 prepared in rabbits were examined to determine that the antiserum was subjected to P-selectin affinity chromatography. It was examined whether the glycoprotein ligand separated by the antibody could be recognized. The radiochemical purity of the ¹²⁵ IP-selectin glycoprotein ligand used in this study indicated that the ligand had a relative molecular weight of about 250,000 under non-reducing conditions and about 120,000 under reducing conditions. Thus, the radioiodinated glycoprotein contained two disulfide bound subunits of M _r 120,000, which is consistent with previous results using protein blotting and metabolic labeling. As mentioned above, the small regions of the dimeric glycoproteins were reduced to resistance under the conditions used. Immune serum, but not normal rabbit serum, precipitated purified ¹²⁵ IP-selectin ligand. Precipitation by antiserum was specific for the 42-56 peptide sequence since the precipitation was completely inhibited by including the peptide rather than the control peptide corresponding to amino acids 354-376 of PSGL-1. The specificity of the antiserum was further demonstrated by selectively immunoprecipitating ligands from a crude mixture of radiolabeled neutrophil cell membrane proteins eluted from a WGA column.

Trypsin digestive peptides were derived from P-selectin glycoproteins purified from human neutrophils. The sequences His-Met-Tyr-Pro-Val-Arg and Pro-Gly- Leu- Thr-Pro-Glu-Pro of these peptides are amino acids 340-345 and 380- in the cDNA-derived sequence of PSGL-1 (SEQ ID NO: 1). Same as 386 respectively . These data and results obtained in immunoprecipitation experiments established the polypeptide backbone of the P-selectin ligands studied in both groups.

Purified ¹²⁵ I-PSGL-1 specifically bound P-selectin. ¹²⁵ I-PSGL-1 isolated from human neutrophils bound in time-dependent fashion to tPS immobilized on microtiter plates. Binding increased as a function of the amount of tPS coated in the wells. Binding was Ca ²⁺ -dependent and inhibited by mAb G1 to P-selectin, which inhibits the attachment of leukocytes to P-selectin, but not by mAb S12, which does not interfere with adhesion. In addition, 84.6 ± 2.4% (mean ± standard deviation, n = 4) of radiolabeled glycoproteins bound back to recombinant soluble P-selectin (tPS) and eluted with buffer containing EDTA. This result demonstrated that the function of ¹²⁵ I-PSGL-1 was substantially unchanged by iodide.

PSGL-1 contained substituted poly-N-acetyllactosamines. PSGL-1 in human bone marrow cells contains clustered sialated O-linked fructose released from the polypeptide backbone by β-removal (Norgard et al., 1993). To determine if the O-linked glycan has a simple structure, after treatment with ¹²⁵ I-PSGL-1 with sialidase, non-sialated Galβ1 rather than the more complex O-linked glycan from serine or threonine residues Treatment with endo-α-N-acetylgalactoosaminidase to remove -3GalNAc Core 1 disaccharide (Umemoto group, 1977, J. Biol. Chem. 252, 8609-8614). Treatment with sialidase showed the electrophoretic mobility of PSGL-1, confirming conventional results. Subsequent addition of endo-β-N-acetylgalactosamidase only slightly increased the mobility of PSGL-1 than sialidase-treated PSGL-1, indicating that very few O-linked fructose It suggests a simple structure sensitive to enzymes.

It is fucose because PSGL-1 contains sialated fucose tetrasaccharide antigen sLe ^x (Norgarde et al., 1993). Treatment of glycoproteins with α1,3 / 4 fucoosidase, which removes the fucose attached to the second GlcNAc at the end, affected the electrophoretic mobility, although it removed the fucose residues that are part of the Le ^x epitope of the molecule. Did not. The ligand is then an endo-β-galactosidase that hydrolyzes the internal β1-4 bond (Galβ1-4GlcNAc) between galactose and N-acetylglucosamine in the extended straight chain poly-N-acetyllactosamine chain. Treated with an aze (Scooter party, 1983, Biochem. J. 213, 485-494; Scooter party, 1984, J. Biol. Chem. 259, 6586-6592). The enzyme accelerated the mobility of PSGL-1 and reduced the heterogeneity of electrophoresis, suggesting that PSGL-1 contains some poly-N-acetyllactosamine.

In order to confirm the presence of poly-N-acetyllactosamine in PSGL-1, the binding capacity of ¹²⁵ I-PSGL-1 to a column containing immobilized tomato lectin that binds well to poly-N-acetyllactosamine was tested. Inspected. In this system, a single poly-N-acetyllactosamine chain is sufficient to allow the glycoprotein to bind to tomato lectin (Merkle et al., 1987, J. Biol. Chem. 262, 8179-8189). More than 95% of PSGL-1 bound to the column, which mainly means that all molecules contained at least one poly-N-acetyllactosamine (Table 3). Treatment of PSGL-1 with endo-β-galactosidase moderately reduced the amount of material bound to the column to 84%. This suggests that, if not all, most poly-N-acetyllactosamines can be removed from some of the PSGL-1 molecules by endo-β-galactosidase. Since the substitution of poly-N-acetyllactosamine with sialic acid and / or fucose inhibits the efficiency of the enzyme, PSGL-1 may be converted to sialidase and fuco prior to addition of endo-β-galactosidase. Pretreatment with ossidase. Treatment with sialidase and fucoosidase did not affect the binding of PSGL-1 to the column, indicating that sialic acid or fucose substitution affected the binding of poly-N-acetyllactosamine to tomato lectin. It is consistent with the fact that it is not crazy. However, subsequent addition of endo-β-galactosidase reduced the glycoprotein binding to the column by 69%. Because tomato lectins weakly bind to terminal GlcNAc residues exposed by endo-β-galactosidase treatment (Cummings, 1994, Methods Enzymol. 230, 66-86), to remove the terminal residues, β-N- Acetylglucoosaminidase was added. Only 53% of the material bound to tomato lectins. These data clearly demonstrated that PSGL-1 contained poly-N-acetyllactosamine. Moreover, at least some of the chains were resistant to endo-β-galactosidase due to sialic acid and / or fucose substitutions. The failure of sialidase, fucoosidase, endo-β-galactosidase and β-N-acetylglucoosaminidase to eliminate the binding of all PSGL-1 molecules to tomato lectins is a cluster of O-linked glycans Enzyme resistance can be reflected due to the nature, the presence of internal fucose residues, and / or the presence of sulfates in additional substituents such as poly-N-acetyllactosamine.

TABLE 3

Effect of Glycosidase on the Binding of ¹²⁵ I-PSGL-1 to Tomato Lectin Affinity Columns

¹²⁵ I-PSGL-1 was treated with the indicated glycosidase and added to a fixed tomato lectin column (V _T = 1 ml, 2 mg lectin / ml resin). After washing, the column was eluted with 20 mg / ml chitotriose. The binding rate of PSGL-1 was defined as the radioactivity eluted with chitotriose divided by the total radioactivity of the column through, wash and eluate.

PSGL-1 contained a limited number of heterogeneous N-linked fructose. PNGaseF treatment slightly increased the electrophoretic mobility of [ ¹²⁵ I] PSGL-1, consistent with previous studies that all molecules contained a limited number of N-linked glycans sensitive to the enzyme. The cDNA-derived sequence of PSGL-1 indicates that the molecule contains only three potential positions for the binding of N-linked fructose (Sako et al., 1993). In order to explore the possible heterogeneity of the N-linked structure, Siamese-treated and PNGaseF-treated [ ¹²⁵ I] PSGL-1 were added to the Concanalvaline A column, which was subjected to high mannose N- Plant lectins that bind very well to binding glycans, bind well to complex 2-antennary and hybrid N-linked fructose, and bind very weakly to complex 3-antennae and 4-antennae N-linked chains. (Ogata group, 1975, J. Biochem. Tokyo, 78, 687-696). 41.9 ± 4.2% of Siamese-treated ligands bound to Con A, whereas only 3.7 ± 1.0% of PNGaseF-treated materials bound (mean ± standard deviation, n = 4). More than 85% of the count loaded on the column was obtained. This result demonstrated that PNGaseF eliminated the N-linked glycans recognized by Con A from PSGL-1.

The PSGL-1 fraction not bound by Con A moved slightly slower than the fraction bound by Con A. To confirm that the Con A-unbound material also contained N-linked fructose, both bound and unbound fractions were treated with PNGaseF. The enzyme reduced the molecular weights of the Con A-unbound and Con A-binding fractions to 14.5 kDa and 13.7 kDa, respectively. Thus, all fractions had PNGaseF-sensitive N-linked glycans. These data suggested that the Con A-non-binding fraction contained 3-antennary and / or 4-antennary N-linked glycans that Con A did not recognize. The largest N-linked glycan (molecular weight about 6,600) that can be described in bone marrow cells is the disialized 4-antenna structure with poly-N-acetyllactosamine sequences at each antenna and external and core fucose (spoons) 1984, J. Biol. Chem. 259, 4792-4801). Thus, the change in electrophoretic mobility observed after PNGaseF treatment was consistent with the removal of two or more and optionally three complex N-linked glycans. Even after PNGaseF treatment, Con A-unbound material migrated slightly slower than Con A-bound material. This may indicate that the Con A-non-binding fraction has one N-binding chain that is resistant to PNGaseF. Alternatively, there may be other structural differences between the Con A-linked and Con A-unbound fractions, most likely the heterogeneity in O-linked glycosylation.

Treatment of sialidase or endo-β-galactosidase had a similar effect on the electrophoretic mobility of the Con A-non-binding and Con A-binding fractions, presumably affecting mainly O-linked fructose rich in these enzymes. Because it gave. Endo-β-galactosidase similarly increased the electrophoretic mobility of pre-treated PSGL-1 with PNGaseF and PSGL-1 containing all N-linked fructose, which at least some poly-N-acetyllactosamine Suggested that it is in bound glycans.

PSGL-1 expressed Le ^x and sLe ^x in O-linked poly-N-acetyllactosamine. To determine if PSGL-1 expresses Le ^x and sLe ^x in poly-N-acetyllactosamine, it is directed to immobilized monoclonal IgM antibodies against sLe ^x (CSLEX-1) or Le ^x (LeuM1), or An assay was developed to measure the binding of ¹²⁵ I-PSGL-1 to immobilized irrelevant IgM mAb (MOPC104E). PSGL-1 bound CSLEX1 but not MOPC104E, confirming conventional immunoblotting data that the glycoprotein expresses sLe ^x . More than 85% of ¹²⁵ I-PSGL-1 molecules bound to the CSLEX-1 affinity column, indicating that virtually all molecules contained sLe ^x . PSGL-1 also bound LeuM1, indicating that PSGL-1 contained Le ^x , although desialylation may have produced the structure during the purification of the glycoprotein. Treatment of PSGL-1 with sialidase eliminated binding to CSLEX-1 and increased binding to LeuM1, consistent with the known specificity of these antibodies. Treatment of PSGL-1 with fucoosidase had no effect on binding to CSLEX-1 and continued inhibition of the enzyme by terminal sialic acid (Sano et al., 1992, J. Biol. Chem. 267, 1522). -1527; K. Maemura and M. Fukuda, 1992, J. Biol. Chem. 267, 24379-24386). In contrast, fucoosidase eliminated the binding of PSGL-1 to LeuM1.

After establishing the specificity of the assay, it was determined whether the treatment of PSGL-1 with endo-β-galactosidase affected the expression of Le ^x or sLe ^x . The enzyme eliminated PSGL-1 binding to immobilized LeuM1 and reduced binding to CSLEX-1 to 64 ± 14% (mean ± standard deviation, n = 4). The data clearly showed that poly-N-acetyllactosamine cleaved by endo-β-galactosidase brought some, perhaps all Le ^x as well as some sLe ^x to PSGL-1. The remaining sLe ^x structure may be in a side chain or substituted poly-N-acetyllactosamine that is resistant to the enzyme or in a chain deficient in poly-N-acetyllactosamine.

In contrast to the effects of endo-β-galactosidase, PNGaseF did not affect the binding of PSGL-1 to immobilized LeuM1 or CSLEX-1, which was Le ^x or sLe to PNGaseF sensitive N-linked fructose. Indicates that almost no ^x structures are present. Regarding other data, this observation indicates that O-linked fructose brings most, perhaps all Le ^x or sLe ^x structures. Moreover, at least some of these structures were in O-linked poly-N-acetyllactosamines.

Effect of Glycosidase on the Binding of PSGL-1 to P-Selectin. The effect of glycosidase on the binding of ¹²⁵ I-PSGL-1 to immobilized tPS was then measured. Treatment of sialidase to remove the binding to the antibody, and wherein ^x -sLe improve binding to the anti--Le ^x antibodies it was completely inhibited the binding of PSGL-1 to tPS. In contrast, treatment with fucoosidase to eliminate binding to anti-Le ^x antibodies but not anti-sLe ^x antibodies did not affect the interaction of PSGL-1 with tPS. These data are consistent with the results of other conventional assays that the high-affinity binding of P-selectin to bone marrow cells or PSGL-1 requires sialic acid in the ligand.

Treatment of PSGL-1 with endo-β-galactosidase only moderately reduced binding to fixed tPS. In four independent experiments, endo-β-galactosidase reduced binding to 25 ± 11% (mean ± standard deviation, n = 4). These data suggested that some, but not all, sialylation chains required for P-selectin recognition are present in the enzyme-sensitive poly-N-acetyllactosamine. The remaining glycans necessary for recognition may be in deficient fructose which is poly-N-acetyllactosamine or poly-N-acetyllactosamine resistant to endo-β-galactosidase. Endo-β-galactosidase further inhibited the binding of PSGL-1 to anti-sLe ^x antibodies rather than inhibiting binding to tPS. The level of sLe ^x interacting with tPS cannot be precisely related because the assay does not know how much sLe ^x structure is required for PSGL-1 to bind to antibodies or tPS.

Treatment of PSGL-1 with PNGaseF does not affect the binding capacity for tPS, which supports the conclusion that N-linked fructose plays little role in P-selectin recognition.

The structure and function of PSGL-1 in human neutrophils was further elucidated using analysis with purified radioiodide glycoproteins. The results suggested that the molecule provides a series of sialated and fucose O-linked poly-N-acetyllactosamines that form a high-affinity binding structure for P-selectin.

Although most fructose of PSGL-1 is O-linked, there are some N-linked glycans that have been found to be heterogeneous. All PSGL-1 molecules contained PNGaseF-sensitive N-linked fructose, but Con A is a lectin that binds well to high mannose N-linked glycans and less well to complex 2-antennaic and hybrid N-linked chains. Only 42% of the glycoprotein molecules bound. Based on the visually reduced molecular weight observed in the SDS gel, PNGaseF removed two or more of the three possible N-linked glycans from the Con A-bound and Con A-unbound fractions of PSGL-1. In the Con A-non-binding fraction of PSGL-1, one chain is resistant to PNGaseF because, after PNGaseF treatment, the Con A-non-binding material still moves slightly slower on the SDS gel than the Con A-binding material. . Alternatively, PNGaseF may have removed all N-linked chains from the Con A-linked and Con A-unbound fractions, in which case the difference in mobility may reflect heterogeneous O-linked glycosylation. PNGaseF-sensitive N-linked glycans resulted in little or no Le ^x and sLe ^x , and did not contribute to the binding interaction with P-selectin to a measurable degree. The results strongly suggest that O-linked glycans represent the relevant structures recognized by P-selectin. In subgroups of PSGL-1, a single PNGaseF-resistant N-linked glycan is unlikely to be involved in P-selectin recognition.

Since at least some O-linked fructose of PSGL-1 results in (α2-3) sialic acid and (α1-3) fucose, they must have a relatively large side chain structure. In line with this expectation, it was found that PSGL-1 rarely results in Ser / Thr-linked Galβ1-3GalNAc core 1 disaccharides that endo-α-N-acetylgalactosaminiase degrades following removal of terminal sialic acid. . These data are consistent with previous observations that less than 10% of the ³ H-labeled O-linked fructose released by β-removal from PSGL-1 was small when evaluated by gel filtration.

O-linked glycans of PSGL-1 include heterogeneous groups of poly-N-acetyllactosamine. As measured by mobility in SDS gels and binding to tomato lectins, endo-β-galactosidase cleaves several poly-N-acetyllactosamines at internal β1-4 binding between Gal and GlcNAc. . The enzyme treatment also reduced the amounts of Le ^x and sLe ^x , indicating that some of their structures are in poly-N-acetyllactosamine. As measured by binding to tomato lectins, endo-β-galactosidase only cleaves other poly-N-acetyllactosamines after treatment with sialidase and fucoosidase, indicating that the molecule It was demonstrated that it had sialic acid and fucose which prevented the access of the endo-β-galactosidase to the terminal or near the terminal. Since a significant amount of the PSGL-1 molecule is still bound to tomato lectin after all three enzymes have been processed, other O remain resistant to endo-β-galactosidase due to branching or additional internal substitution. There may be a -binding poly-N-acetyllactosamine chain. The structure will be at least partly involved in the ability of PSGL-1 to bind tPS after cleavage with endo-β-galactosidase.

Previous studies have shown that human bone marrow cells express poly-N-acetyllactosamine on some O-linked glycans. These structures occur almost exclusively on branches attached to C-6 of GalNAc as part of core 2 O-linked fructose (Fukuda et al., 1986, J. Biol. Chem. 261, 12796-12806; Hanish et al., 1989). J. Biol. Chem. 264, 872-883). Of the about 80 O-linked glycans of the major sialomuscins of human bone marrow cells, only one or two contain poly-N-acetyllactosamine, about half of which have terminal sLe ^x . In contrast, PSGL-1 of neutrophils appears to contain a higher proportion of core 2 poly-N-acetyllactosamine. Direct structural analysis will be required to determine the specific monosaccharides, linking structures and variants that make up the O-linked glycans of PSGL-1. However, current data suggest that some of these glycans may differ from those of leucosialin, suggesting that human bone marrow cells may differently glycate the polypeptide backbone of two distinct sialomycins. This different glycosylation is functionally important because P-selectin has a high affinity linkage with PSGL-1 but not leucosialine.

Most carbohydrate ligands need to be sialated and fucose to be recognized by selectin. Similarly, PSGL-1 requires sialic acid and fucose to interact with P-selectin. Although PSGL-1 expresses terminal sLe ^x tetrasaccharide, it is not clear whether this structure itself is required for recognition. The simple model is to provide a number of terminal sLe ^x which increase the binding capacity of PSGL-1 to interact with P-selectin. However, conventional observation indicates that high concentrations of cell surface sLe ^x are not sufficient to optimally interact with P-selectin. Additional modifications of each fructose chain appear to be essential for optimal recognition. The particular properties of some fructose in the vicinity will also provide a clustered sugar that enhances high affinity binding to P-selectin.

Sarco group (1993) binds to P-selectin when expressed in cotransfected COS cells with α1-3 / 4 fucosyltransferase (FTIII), but not COS cells lacking fucosyltransferase. Recombinant PSGL-1 was prepared. The relative molecular weight of the recombinant molecule was similar to that of PSGL-1 of human myeloid cell HL-60, and the authors clearly concluded that the structure and function of recombinant PSGL-1 are identical to the original molecule. However, the specific fructose structures bound to the original PSGL-1 and glycosyltransferases necessary to form these structures are not known. Furthermore, α1-3 / 4 fucoosyltransferase (FTIII) expressed in COS cells producing recombinant PSGL-1 is different from that of human bone marrow cells (Goelz et al., 1990, Cell 63, 1349-1356; Sasaki et al., 1994, J. Biol. Chem. 269, 14730-14737). Thus, it is not clear whether the glycosylation and resulting function of recombinant PSGL-1 is identical to the original glycoprotein of human bone marrow cells. In addition, after cleavage of the recombinant PSGL-1 with sialidase, the addition of endo-α-N-galactoosaminidase substantially increased the mobility of the glycoprotein, in contrast to the original glycoprotein. Recombinant PSGL-1 expressed in COS cells with FTIII shows that it contains a lot of simple O-binding core 1 Galβ1-3GalNAc disaccharide sensitive to the enzyme.

Sarco et al. Revealed that recombinant PSGL-1 no longer binds to P-selectin after treatment with sialidase and endo-β-N-galactosamiminidase, and the O-linked fructose was induced by P-selectin. We concluded that it is important for recognition. Based on the data provided, the conclusion seems to be correct. However, Sarco's experiments cannot be used to support the role of O-linked glycans in recognition. It is necessary to extend the sialidase treatment to remove the binding to the original PSGL-1, and the sialidase treatment conditions for recombinant PSGL-1 used by Sarco companion reduced the binding of P-selectin but eliminated it. I could not. The loss of binding of P-selectin after addition of endo-α-N-galactosaminiase, which releases only non-sialated core 1 Galβ1-3GalNAc disaccharides, is associated with sialidase, which removes residual sialic acid required for recognition. It appeared to be due to previously observed contamination of the enzyme.

Sarco et al. Also suggested that N-linked fructose contributes to binding to P-selectin since recombinant P-selectin did not reprecipitate all PNGaseF-treated recombinant PSGL-1. However, these authors did not examine whether the additional precipitation step increased the recovery of PSGL-1 and did not prove whether the reprecipitation assay was quantitative. The same population also observed that bone marrow cells HL-60 treated with tunicamycin, an inhibitor of N-linked glycosylation, did not adhere to P-selectin expressing cells, suggesting that N-linked glycosylation is important for P-selectin recognition. Conclusions (Larsen et al., 1992, J. Biol. Chem. 267, 11104-11110). However, tunicamycin exerts indirect effects on cells, such as inhibiting the exit of some proteins from the endoplasmic reticulum, making it difficult to interpret the effects on cell adhesion. The study of the present application provides no evidence that the N-binding glycans of PSGL-1 of human neutrophils play a role in recognition by P-selectin.

Example 2: Expression of PSGL-1 in CHO Cells Transfected with cDNA Encoding Fucoosyltransferase III and Core 2 β1-6-N-acetylglucosaminyltransferase

Functional recombinant forms of PSGL-1 were expressed that had many of the properties of the original PSGL-1 purified from human neutrophils. This study demonstrated that PSGL-1 requires sialated and fucose O-linked fructose containing a branched Core 2 structure to bind P-selectin.

PSGL-1 was expressed in selected CHO cells because N- and O-linked fructose was very well characterized on the cells. In particular, Ser / Thr-binding fructose of secreted and surface glycoproteins is a simple, core 1, mono- and disialylated derivative of Galβ1-3GalNAcα1-Ser / Thr (Sasaki et al., J. Biol. Chem. 262: 12059-12076 , 1987; Seguchi party, Arch. Biochem. Biophys. 284: 245-256, 1991). DHFR-minus CHO cells encode cDNA encoding Foucault siltransferase III (FTIII) (Cucoska-Latalo et al., 1990), core 2 β1-6-N-acetylglucosaminyltransferase cDNA (Vierhuizen and Fukuda, Proc. Natl. Acad. Sci. USA, 89: 9326-933, 1992; Vierhuizen et al., J. Biol. Chem. 269: 4473-4479, 1994), or the two above Stably transfected with cDNA. These cDNAs were linked to pRc / RSV or pCDNA3 expression vectors (Invitrogen). After transfection by calcium phosphate precipitation, stably transfected clones were selected in medium containing G418. Stable clones expressing one or all glycosyltransferases were identified using specific glycosyltransferase assays performed on cell lysates. Cells expressing FTIII also expressed high levels of sialyl Lewis X antigen (NeuAcα2-3Galβ1-4 [Fucα1-3] GlcNAcβ1-4-R) on the cell surface, which was measured by binding of CSLEX1 monoclonal antibodies. (Fukushima group, Cancer Res. 44: 5279-5285, 1984).

CDNA encoding full length PSGL-1 (Moore et al., J. Cell Biol. 128: 661-671, 1995) was inserted into the pZeoSV expression vector (Invitrogen). CHO cells expressing glycosyltransferase were transfected with PSGL-1 cDNA using lipofectamine reagent (Gibco BRL Life Technologies). In transient expression assays, cells were analyzed by flow cytometry 48 hours after transfection. Stable clones were selected in medium containing both G418 and Zeocin (Invitrogen). Analysis of cell-binding antibodies by flow cytometry was performed as described by Moore et al. (1995). Analysis of cell-bound platelet-induced P-selectin by flow cytometry was performed as described by Moore and Thompson (Biochem. Biophys. Res. Commun. 186: 173-181, 1992).

In transient expression studies, PSGL-1 was co-expressed in FTIII, core 2 β1-6-N-acetylglucosaminyltransferase, or most transfected cells expressing both glycosyltransferases. . Coexpression was demonstrated by strong binding of the anti-PSGL-1 monoclonal antibody PL1 or PL2 (Mourning, 1995) measured by flow cytometry. In contrast, flow cytometry revealed that P-selectin only binds to CHO cells expressing FTIII, Core 2 β1-6-N-acetylglucoosaminyltransferase and PSGL-1. P-selectin did not bind to CHO cells expressing both one or two glycosyltransferases in the absence of PSGL-1. P-selectin also did not bind to CHO cells expressing FTIII or Core 2 β1-6-N-acetylglucosaminyltransferase in the presence of PSGL-1. Thus, PSGL-1, a functional form capable of binding P-selectin, was expressed only in CHO cells that also expressed FTIII and core 2 β1-6-N-acetylglucosaminyltransferase.

CHO cell clones stably expressing FTIII and Core 2 β1-6-N-acetylglucosaminyltransferase were also selected. These cells expressed PSGL-1 at high levels as detected by the binding of PL1 and PL2 by flow cytometry. P-selectin also bound well to these stably selected cells. Thus, high affinity binding of P-selectin to CHO cells requires co-expression of FTIII, Core 2 β1-6-N-acetylglucosaminyltransferase and PSGL-1 by the cells. In cells expressing PSGL-1 transiently or stably, binding of P-selectin is interrupted by inhibitory monoclonal antibody G1 to P-selectin, but not non-inhibiting monoclonal antibody S12 to P-selectin The binding of P-selectin was specific (Jean et al., Nature, 343: 757-760, 1990). The binding was also Ca ²⁺ -dependent, which is consistent with the need for Ca ²⁺ to bind the selectin to carbohydrate ligands. P-selectin also specifically bound to PSGL-1 in cells, and PL1, not PL2, completely disrupted the binding of P-selectin. Moreover, the following anti-42-56 serum, but not the control rabbit serum, interfered with the binding of P-selectin.

The data indicate that CHO cells can be genetically engineered to express a recombinant form of PSGL-1 that specifically and highly affinity binds to P-selectin, similar to the original PSGL-1 of human leukocytes. Inhibition of P-selectin binding by PL1 and anti-42-56 sera indicates that P-selectin binds to the same region of recombinant PSGL-1 and original PSGL-1 obtained from human leukocytes. The need for core 2 enzymes in CHO cells consists of branched core 2 O-linked fructose, in which the key component of the P-selectin recognition site in PSGL-1 is Galβ1-3 (Galβ1-4GlcNAcβ1-6) GalNAc-Ser / Thr backbone It is shown. Sialation and fucoseization of PSGL-1 are also required for high affinity binding to P-selectin; These modifications must occur in the core 2 O-linked glycans that are important.

Example 3 Determination of Regions Containing Tyrosine Sulfate Important for Binding of P-Selectin in PSGL-1.

PSGL-1 has an extracellular region of 308 residues. The first 18 residues make up the signal peptide. The 19-41 residues constitute putative propeptide that will be cleaved after synthesis, possibly in trans-Golgi network. Thus, the putative N-terminus of the mature protein starts at 42 residues. The three tyrosines in the consensus sequence for tyrosine sulphate are 46, 48 and 51 residues near the N-terminus. Rabbit polyclonal antibodies to synthetic residues encoding 42-56 residues completely inhibited the binding of PSGL-1 to P-selectin, suggesting that one or more important recognition sites for P-selectin exist in this region. do. All references to residue numbers relate to SEQ ID NO: 1.

Two IgG monoclonal antibodies PL1 and PL2 against PSGL-1 were previously described by Moore et al. (J. Cell Biol. 128: 661-671, 1995). PL1, but not PL2, completely inhibits the binding of PSGL-1 to P-selectin and eliminates the attachment of leukocytes to P-selectin under static and shear conditions. In order to map the region where the epitopes for PL1 and PL2 are located in PSGL-1, a series of fusion proteins containing overlapping regions of the extracellular region of PSGL-1 were expressed in bacteria. The PSGL-1 sequence of the fusion protein is as follows: fragment (1), 18-78 residues; Fragment (2), 63-123 residues; Fragment (3), 109-168 residues; Fragment (4), 154-200 residues; Fragment (5), 188-258 residues; Fragment (6), residue 243-308. The DNA encoding each fragment was fused to the frame of the DNA encoding the dihydrofolate reductase and 6X His tag in the expression vector pQE-40 (Qiagen). Each fragment was expressed in bacteria and purified on a 6 × His tag bound nickel affinity column. Reactivity of each fusion protein with the antibody was evaluated by western blotting under reducing conditions.

PL1 bound strongly to fragment (1) and not to other fragments. This indicates that the PL1 epitope is located at 18-78 residues. Given that the propeptide is removed from the original mature PSGL-1, the epitope is located at 42-78 residues. Since PL1 does not bind to fragment (2) (63-123 residues), the epitope is likely to be located at 42-70 residues. Thus, PL1 binds to an N-terminal epitope near the epitope for polyclonal anti-42-56 serum. This result further suggests that an important component of the P-selectin recognition site is in a small region at the N-terminus of PSGL-1. In contrast, the noninvasive monoclonal antibody PL2 bound to fragments (5) and (6) (but not to other fragments), indicating that the PL2 recognizes an epitope within the 243-258 residues present in both fragments. Suggested.

These results demonstrated that P-selectin binds to specific regions of the original PSGL-1 and recombinant PSGL-1. This region is located near the N-terminus of the mature molecule. High affinity binding of PSGL-1 to P-selectin requires tyrosine sulphate and sialated, fucoseized core 2 side chain O-linked fructose. Fragments of PSGL-1, or derivatives thereof, that contain a binding site for P-selectin can be prepared. These fragments may be short enough to encode 42-58 residues; This fragment contains not only three tyrosines that can be sulfided but also two threonines that can be the attachment sites for core 2 O-linked fructose. Longer fragments that extend from the 42 residues will also bind highly friendly to P-selectin.

Experimental procedure

Chemicals and Reagents

Carrier-free [ ³⁵ S] sulfate (1100-1600 Ci / mmol) was purchased from DuPont / NEN (Bilmington, Germany). The trade name Emphaze affinity support resin was purchased from Pierce Biochemicals (Rockford, Ill.). Aerobacter aerogenes arylsulfatase, neuraminidase of Artrobacter urea faciens and sulfided monosaccharides were obtained from Sigma Chemical Co. (St. Louis, Missouri, USA). Recombinant peptide: N-glycosidase F was purchased from Schöllinger Mannheim (Indianapolis, Indiana). All reagents for cell culture were obtained from GIBCO / BRL (Grand Island, NY). True tyrosine sulfate standards were synthesized as described by Hutner (1984, Meth. Enzymol. 107, 200-224) using concentrated H ₂ SO ₄ and tyrosine. Other chemicals are ACS grade or better and were obtained from Fisher Scientific.

Isolation of Radiolabeled PSGL-1

PSGL-1 was purified from human neutrophils and radiolabeled with Na ¹²⁵ I as described by Moore group (1994). HL-60 cells (1-2 × 10 ⁶ cells / ml) were treated with 100 μCi / ml [ ³⁵ S] in 37 ° C., sulfate-deficient medium (S-MEM) containing 10% of dialyzed fetal bovine serum. Label for 48 hours. ³⁵ S-PSGL-1 was purified on a column containing the recombinant soluble P-selectin (Ushiyama et al., 1993, J. Biol. Chem. 268, 15229-15237) coupled with the trade name Emphaze at a density of 5 mg / ml. Purification was carried out using chemical chromatography (Mourde, 1994). Concentrated samples eluted with EDTA were dialyzed with Ca ²⁺ -containing buffer and rechromatated on a P-selectin column. Purified ³⁵ S-PSGL-1 was treated with reducing or non-reducing SDS sample buffer and electrophoresed on 7.5% polyacrylamide gel (Laemley, 1970). The gel was dried and then exposed to Fuji RX X-ray film at -80 ° C.

Detection of Tyrosine Sulfate

Tyrosine sulfate of PSGL-1 was determined by base hydrolysis of ³⁵ S-PSGL-1. After confirming ³⁵ S-PSGL-1 by autoradiography, the dried gels were aligned with the exposed X-ray film and the non-reducing lanes were hydrobase hydrolyzed at 110 ° C., 0.1 M NaOH for 24 hours as described. (Hortin group, J. Biol. Chem. 261, 15827-15830, 1986). The hydrolyzate was passed through a Dowex 50 (H ⁺ ) column washed with water, lyophilized and anion exchange chromatography using a Varian AX-5 column (4 mm x 30 cm) eluted with a gradient of NaH ₂ PO ₄ at pH3.0. Analysis by graphy.

Enzyme Treatment of PSGL-1

All arylsulfatase treatments were carried out using aerobacter aerogenes arylsulfatase (EC 3.1.6.1) in about 500 μl of 0.01 M Tris-HCl at pH 7.5 overnight at 37 ° C. (Fowler And Lammers, 1964, Biochemistry 3, 230-237). After cleavage, the enzyme reaction was terminated by boiling for 15 minutes. Siamese treatment was carried out using 1000mU of suspended enzyme. ³⁵ S-tyrosine sulfate was prepared from ³⁵ S-PSGL-1 by base hydrolysis as described above. Base hydrolysates were treated with 50mU or 1000mU of active arylsulfatase, or 1000mU boiled arylsulfatase and analyzed by anion exchange chromatography.

The specificity of Aerobacter aerogenes arylsulfatase was measured for sulfided monosaccharides GlcNAc-6-sulfate, Gal-6-sulfate, GalNAc-6-sulfate and GalNAc-4-sulfate. Each sulfided monosaccharide (100 nmol) was boiled at 1000 mU or treated separately with active enzyme. After incubation overnight, the reaction mixture was boiled for 15 minutes, lyophilized, resuspended in water and analyzed by high pH anion exchange chromatography and PAD detection using a Dionex CarboPac PA-1 column (Silatipard). And Cummings, 1995, Glycobiology 5, 291-297).

Intact ³⁵ S-PSGL-1 or ¹²⁵ I-PSGL-1 was treated with aerobacter aerogenes arylsulfatase as described above. [ ³⁵ S] sulfate separated from ³⁵ S-PSGL-1 by arylsulfatase was quantified by BaSO ₄ precipitation using 100 μl of saturated Na ₂ SO ₃ as carrier (Silatipard, 1993, J. Virol. 67, 943-945). In the graph showing the data, the sulphate-treated samples were corrected to set the sham-treated samples to 0% by correcting the sulphate away from the sham-treated samples. The actual rate of radioactivity released from the sham-treated samples was in the range of 1.3-5.0%.

In a control experiment, ¹²⁵ I-PSGL-1 was treated with 1000mU of Arturobacter ureapaciens neuraminidase in 0.05M sodium acetate at pH5.0 overnight at 37 ° C. prior to addition to the CSLEX-1 antibody column. CSLEX-1 chromatography was performed as previously described (Moore One, 1994).

³⁵ S-PSGL-1 was treated with peptide: N-glycosidase e as described by Schilatipard (1993). ³⁵ S-PSGL-1 was denatured by stopping at 50 μl of 50 mM sodium phosphate, pH7.5, 50 mM 2-mercaptoethanol, 0.5% SDS for 5 minutes. The SDS was then diluted to a final concentration of 0.2% using 1.5% NP-40. 10 U of enzyme was added to denatured ³⁵ S-PSGL-1 and incubated overnight at 37 ° C.

Recombination of PSGL-1 Treated with Arylsulfatase to P-Selectin

PSGL-1 treated with arylsulfatase was added to a P-selectin affinity column in Ca ²⁺ -containing buffer and the bound radioactivity eluted with EDTA. EDTA was used to calibrate the ratio of ¹²⁵ I-PSGL-1 eluted from the column to set the sham-treated sample to 100% (actual recombination rate ranged from 87-99%).

Immunoprecipitation and P-Selectin Binding Assay

Immunoprecipitation of ³⁵ S-PSGL-1 was performed by rabbit antiserum (anti-42) against peptides (Sacco, 1993) corresponding to 42-56 residues (QATEYEYLDYDFLPE) of PSGL-1, as described by Moore et al. (1994). -56) (Moore party, 1995; Moore party, 1994) or normal rabbit serum (NRS). Immunoprecipitates were analyzed by SDS-PAGE under non-reducing conditions, followed by autoradiography. Solid phase P-selectin binding assays were performed as described by Moore et al. (1994). In brief, 42-56 of NRS or PSGL-1 in microtiter plates containing immobilized recombinant soluble P-selectin (50 μl, 10 μg / ml) or human serum albumin (HSA) (50 μl, 1% HSA) ¹²⁵ I-PSGL-1 was cultured in the presence of antiserum (anti-42-56) to the residue. Binding of platelet-induced P-selectin to HL-60 cells was measured by flow cytometry as described in the presence of 20% NRS or 20% anti-42-56 serum.

Results and discussion

To confirm that PSGL-1 is sulfided after transcription, human promyelocytic leuchemia HL-60 cells are labeled with [ ³⁵ S] sulfate by metabolism, and PSGL-1 is affinity for the P-selectin column. Purification was performed from the eluate of the cells by chromatography. Proteins labeled with [ ³⁵ S] sulfate were detected to migrate in SDS gels at a relative molecular weight of 120,000 under reducing conditions and 240,000 under non-reducing conditions.

The mobility was consistent with the disulfide-linked homodimer structure of PSGL-1 (Moore party, 1992). In addition, proteins labeled with [ ³⁵ S] sulfate were immunoprecipitated with specific rabbit antiserum against synthetic peptides encoding 42-56 residues, the extracellular region of PSGL-1. Supernatant analysis confirmed that anti-42-56 precipitated all ³⁵ S-PSGL-1, whereas normal rabbit serum (NRS) did not precipitate ³⁵ S-PSGL-1 at all. These results demonstrated that PSGL-1 was sulfided after transcription.

Sulfates may be tyrosine sulfate (Hutner's group, 1988, Mod. Cell Biol. 6, 97-140) or sulfide carbohydrates (Pette's group, 1991, Cell 65, 1103-1110; xelen and lindal, 1991, Ann. Rev. Biochem. 60, 443-475), can be incorporated into eukaryotic glycoproteins. In basic studies, sulfide carbohydrates were not detected in PSGL-1 when using other techniques to detect sulfide carbohydrate structures in glycoproteins (Shilatipard, (1995, 1993)). The cDNA sequence of PSGL-1 predicts four extracellular tyrosine residues, three of which are clustered at positions 46, 48 and 51 within the predicted consensus sequence for tyrosine sulfidation (Huttner and Baieu, 1988). To confirm that PSGL-1 has tyrosine sulphate, a gel piece containing 240 kDa of ³⁵ S-PSGL-1 was hydrolyzed with a strong base and analyzed by anion exchange chromatography. A single radioactive peak covalently extracted with tyrosine sulfate rather than a sulfided sugar standard was obtained (FIG. 1A).

Then, the importance of tyrosine sulfate in the binding of PSGL-1 to P-selectin was examined. Although the function of tyrosine sulphate in proteins is not clear, some proteins require this modification for optimal activity (Pitman et al., 1992, Biochemistry 31, 3315-3325; et al., 1994, Biochemistry 33, 13946- 13953). A common method for evaluating the importance of tyrosine sulfate is to inhibit the sulfidation of newly synthesized proteins with chemical inhibitors or to replace tyrosine with phenylalanine by site-directed mutagenesis. Other methods of enzymatic removal of sulphate from intact PSGL-1 tyrosine were also examined. The ability of Aerobacter aerogenes arylsulfatase to remove sulfate from ³⁵ S-tyrosine sulfate was also measured. 50mU by a small amount of the aryl alcohol Zapata dehydratase was quantitatively remove the ^[35 S] sulfate from ³⁵ S- tyrosine sulfate derived from PSGL-1 (Fig. 1b, 1c). The removal was specific because 1000mU of the enzyme did not remove sulfate from Gal-6-sulfate, GlcNAc-6-sulfate, GalNAc-4-sulfate and GalNAc-6-sulfate.

The ability of arylsulfatase to remove sulfate from intact ³⁵ S-PSGL-1 was examined. [ ³⁵ S] sulfate released by the enzyme was quantified by insoluble BaSO ₄ precipitation. Up to 50% of [ ³⁵ S] sulfate in PSGL-1 was removed by 500 mU of arylsulfatase; Increasing the amount of enzyme did not release more radioactivity (FIG. 2).

The functional significance of tyrosine sulphate in PSGL-1 was assessed by treating ¹²⁵ I-PSGL-1 with arylsulfatase and measuring recombination of the treated ligand to the P-selectin column. Binding of ¹²⁵ I-PSGL-1 treated with arylsulfatase to P-selectin was dose-dependently reduced; Reduced binding and sulfate removed from ³⁵ S-PSGL-1 were inversely related to each other (FIG. 2). Since PSGL-1 treated with 1000mU of arylsulfatase bound quantitatively to an affinity column containing the monoclonal antibody CSLEX-1 of sLe ^x , the PSGL-1 to P-selectin after arylsulfatase treatment Reduced binding was not due to released sialic acid and / or fucose, which contaminated exoglycosidase.

Aryl alcohol Zapata azepin had remove approximately 50% ^[35 S] sulfate from ³⁵ S-PSGL-1, and reduced to the same ¹²⁵ I-PSGL-1 to the recombination of P- selectin. Fractions of ³⁵ S-PSGL-1 that recombine P-selectin after treatment with arylsulfatase are likely to have important tyrosine sulfate residues, whereas fractions that do not recombine P-selectin lost tyrosine sulphate. . ³⁵ S-PSGL-1 was treated with 1000 mU of arylsulfatase and added to the P-selectin column. Binding fractions and unbound fractions were analyzed by SDS-PAGE. A band corresponding to ³⁵ S-PSGL-1 was observed in the binding fraction, while no band was observed in the unbound fraction. Radioactivity of the unbound fraction indicated free sulphate. When the band of ³⁵ S-PSGL-1 was hydrolyzed in the binding fraction, radioactivity was obtained in tyrosine sulfate. In the control experiment, ¹²⁵ I-PSGL-1 was treated with 1000 mU of arylsulfatase and added to the P-selectin column. SDS-PAGE analysis confirmed that intact ¹²⁵ I-PSGL-1 was recovered in both unbound and bound fractions. Thus, arylsulfatase quantitatively removed sulfate from a subset of PSGL-1 that no longer bound P-selectin, while the enzyme did not degrade PSGL-1.

Tyrosine sulphate remaining in P-selectin bound to a subset of PSGL-1 is directed against arylsulfatase either because of the inaccessibility of arylsulfatase or because of some other feature of PSGL-1 that interferes with the action of the enzyme. It may be resistant. ^{When 35} S-PSGL-1 is partially eliminated by the peptide: N-glycosidase F and neuraminidase of Artrobacter ureapaciens, subsequent arylsulfatase treatment results in up to 70% [ ³⁵ S] sulfate was removed. Since neuraminidase also eliminates the binding of PSGL-1 to P-selectin, it is determined whether increased elimination of [ ³⁵ S] sulfate further reduces the binding of PSGL-1 to P-selectin. Could not be. The results suggest that extensive glycosylation of PSGL-1 may explain the inaccessibility of some tyrosine sulfate sites for arylsulfatase.

To further demonstrate the importance of tyrosine sulphate on the function of PSGL-1, antisera against peptides encoding 42-56 residues, the extracellular region of PSGL-1, was used. As noted above, the region contains three potential tyrosine sulfide positions at 46, 48 and 51 residues. The antiserum inhibited the binding of ¹²⁵ I-PSGL-1 to immobilized P-selectin (FIG. 3A) and also eliminated the binding of P-selectin to HL-60 cells as measured by flow cytometry. (FIG. 3B). Thus, antiserum against the 42-56 peptide binds to the original PSGL-1 and effectively interferes with recognition by P-selectin.

The results demonstrated that PSGL-1 contains tyrosine sulfate, which is required for high affinity binding with P-selectin. It has conventionally been found that PSGL-1 contains sLe ^x determinants in O-linked fructose and that both sialic acid and fucose are required for the binding of PSGL-1 to P-selectin. Tyrosine sulfate may be important because it enhances the proper expression of glycans that binds directly to P-selectin. Alternatively, tyrosine sulfate may interact directly with P-selectin. Since sulfatide and sulphide fructose are known to bind P-selectin, the latter seems to be more likely. Sulfates are also important determinants of the binding of GlyCAM-1 to L-selectin, but GlyCAM-1 contains sulfates in Gal-6-sulfate and GlcNAc-6-sulfate rather than monosaccharides.

The results suggest a model in which PSGL-1 provides both carbohydrates and tyrosine sulphate as components of important recognition sites for P-selectin. The position is assumed to be in the region of the N-terminus, cell membrane terminus of PSGL-1, near three potential tyrosine sulfate residues. Consistent with these assumptions, the monoclonal antibody PL1, which recognizes epitopes at the cell membrane end in PSGL-1, inhibited the binding of liquid P-selectin to leukocytes under static and shear conditions and eliminated the adhesion of neutrophils to P-selectin. . In contrast, the monoclonal antibody PL2, which recognizes an epitope near the cell membrane in PSGL-1, does not inhibit binding to P-selectin. The concept that P-selectin recognizes local positions in mucin-type glycoproteins is evident from the model that selectins recognize multiple, clustered O-binding glycans attached across the entire polypeptide (Raski et al., 1992, Cell 69, 927). -938; Baumhutter Group, 1993, Science 262, 436-438).

Example 4: Determination of O-glycan structure in PSGL

The O-glycan structure of PSGL-1 synthesized by metabolically radiolabeled HL-60 cells was determined using ³ H-sugar precursors. The glycosylation of PSGL-1 and the glycosylation of CD43 were compared to determine whether two sialomuscins expressed in the same cell differ O-glycosylated. The HL-60 cell line was used in this study because it is possible to compare post-transcriptional modifications known to be important for binding to both P- and E-selectin of both HL-60 and neutrophil PSGL-1.

This study demonstrated that most O-glycans of PSGL-1 are disialized or neutral forms of Core-2 tetrasaccharide. Only some O-glycans are fucoseized and they contain structural determinants for the sLe ^x antigen. These results demonstrated that PSGL-1 glycated differently from CD43 and contained unique O-glycans that could be important for high affinity interactions with P- and E-selectin.

The abbreviations used are:

sLe ^x , Sialyl Lewis X;

NDV, Newcastle Disease Virus;

CHO, Chinese Hamster Ovary;

GlcN, glucosamine;

GalN, galactosamine;

GalNOH, galactosaminitol;

GalNAcOH, N-acetylgalactososaminitol;

FT, fucoosyltransferase;

C2GnT, core 2 β1,6 N-acetylglucosaminyltransferase;

HPAEC, high pH anion exchange chromatography;

I. Experimental Procedure

material

Proteins A-Sepharose, QAE-Sephadex, Neuraminidase of Artrobacter ureapaciens, Jack bean β-galactosidase, Jackbin β-N-acetylhexosaminidase and Galβ1 → 3GalNAc Was obtained from Sigma. Pronase and TPCK-treated trypsin were purchased from Worthington Biochemicals. Escherichia prefundi endo-β-galactosidase was obtained from V-labs. Α1,3 / 4-fucoosidase of Streptomyces sp. 142 was purchased from Takara. Neuraminidase of Newcastle Disease Virus was obtained from Oxford Glycosystems. D- [6- ³ H] glucosamine hydrochloric acid (20-45 Ci / mmol), D- [2- ³ H] mannose (20-30 Ci / mmol) and L- [6- ³ H] fucose (70- 90 Ci / mmol) was purchased from Dupont / NEN and NaB [ ³ H] ₄ (40-60 Ci / mmol) was purchased from ARC. Rabbit anti-mouse IgG1 was purchased from Zymed. The trade name Emphaze affinity support resin was obtained from Pierce Biochemicals. Bio-Gel P-4 and P-10 resins and molecular weight markers were obtained from BioRad. All cell culture reagents were obtained from GIBCO / BRL. Other chemicals are ACS grade or higher and were obtained from Fisher Scientific.

Isolation of Radiolabeled PSGL-1 by Metabolism

Radiolabeling by ³ H-GlcN, ³ H-Man and ³ H-Fuc metabolism of HL-60 cells was performed as described (Mourning, 1992, J. Cell. Biol. 118, 445-456; Wilkins Et al., 1995, J. Biol. Chem. 270, 22677-22680). ³ H-PSGL-1 is a fixed recombinant soluble P-selectin (Ushiyama et al., 1993, J. Biol. Chem. 268, coupled with a trade name Emphaze at a density of 5 mg / ml (bed volume of 1.0 ml). 15229-15237) was purified from the cell extracts using affinity chromatography on the column (Moore party, 1994, J. Biol. Chem. 269, 23318-23327). Concentrated samples eluted with EDTA were dialyzed with Ca ²⁺ -containing buffer and rechromatated on a P-selectin column. Purified ³ H-PSGL-1 was treated with reducing or non-reducing SDS sample buffer (Laemley, UK, 1970, Nature 227, 680-685) and electrophoresed on 7.5% polyacrylamide gels. The gel was dried and then autoradiated by exposure to a Fuji RX X-ray film at -80 ° C. About 250 kDa of ³ H-PSGL-1 was analyzed in gel pieces obtained from non-reducing lanes.

Immunoprecipitation of CD43

CD43 was immunoprecipitated from HC-60 cells labeled with ³ H-sugars using CD43-specific mAb, H5H5 (IgG1). H5H5 hybridoma cell line T. Produced by Dr. Auguste and obtained from Developmental hybridoma Bank (The Johns Hopkins University School of Medicine). CD43 was first immunoprecipitated by preloading 100 μg of rabbit anti-mouse IgG1 antibody into 50 μl of protein A-Sepharose beads for 1 hour at 37 ° C. After washing, CD43 specific antibodies were taken up in the beads by adding 500 μl of undiluted H5H5 culture medium at 37 ° C. for 1 hour. After washing, preloaded beads were added to HC-60 cell extracts labeled with ³ H-sugars and incubated overnight at 4 ° C. The beads were washed five times with Hanks balanced salt solution, then boiled in non-reducing SDS sample buffer for 5 minutes and analyzed on a 10% SDS-polyacrylamide gel. 120 kDa purified CD43 was cut out from the gel fragments identified after autoradiography and analyzed.

³ H-PSGL-1 and ³ Β-removal of radiolabeled O-glycans from H-CD43 and chromatography of glycans in Bio-Gel and QAE-Sephadex

Gel pieces containing radiolabeled PSGL-1 and CD43 were treated directly with mild base / borohydride as described (Yer et al., 1971, Arch. Biochem. Biophys. 142, 101-105; Cummings party, 1983 J. Biol. Chem. 258, 15261-15273). The radiolabeled glycans that were released were removed by chromatography on a Dowex 50 (H ⁺ type) column and further analyzed for glycans. Chromatography was performed on a Bio-Gel P-4 (medium mesh) on a 1 × 90 cm column and a Bio-Gel P-10 (medium mesh) on a 1 × 48 cm column in 0.1 M pyridyl / acetate buffer at pH5.5. And 1 ml of fractions were collected. Anionic properties and sialation of glycans were partially assessed by chromatography on a QAE-Sephadex column, step eluting with 2 mM Tris-base containing 20, 70, 140, 200 and 250 mM NaCl.

Preparation of Radiolabeled O-Glycan Standards

Simple core -1 O- glycan Galβ1 → 3GalNAc was Galβ1 → 3GalNAcOH ^[3 H] of NaB ^[3 H] was radiolabeled by reduction in the presence of ₄ (line Li, 1978, J. Biol. Chem. 253 to form , 7762-777). ³ H-GlcN-glycans of standard cover were prepared from total glycoproteins of the HL-60 cells labeled with ³ H-GlcN after the step (Carlson line, 1986, J. Biol. Chem. 261, 12787- 12795). The standard was NeuAcα2 → 3Galβ1 → 4GlcNAcβ1 → 6 (NeuAcα2 → 3Galβ1 → 3) GalNAcOH, neutral core-2 tetrasaccharide Galβ1 → 4GlcNAcβ1 → 6 (3Galβ1 → 3) GalNAcOH, and trisaccharide GlcNAcβ1 → 6 → 3) GalNAcOH. To prepare the standard, cell extracts labeled with ³ H-GlcN were dialyzed at 0.01 M NaHCO ₃ , pH7.5 and treated overnight at 37 ° C. with 100 μg / ml TPCK-treated trypsin. The dried trypsin digest was treated with mild base / boron hydride for β-removal, and the separated O-glycans were fractionated according to size on a Bio-Gel P-4 column. The collected fractions corresponding to radioactive peaks were analyzed for charge distribution by QAE-Sephadex column chromatography. Each glycan structure was then confirmed by sensitivity to Artrobacter neuraminidase, β-galactosidase, and β-N-acetylhexosaminidase by falling paper chromatography as described below. .

Core-2 fucoated pentasaccharide standard Galβ1 → 4 (Fucα1 → 3) GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH is ³ H-GlcN- in the presence of 50 μg of extract from COS7 cells stably transfected with human FTIV gene. Glucose was prepared by incubating Galβ1 → 4GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH with unlabeled GDP-Fuc (100 μM). Said α1,3FT adds the Fuc of α1,3 binding to the GlcNAc residue of the receptor containing the terminal N-acetyllactosamine sequence (Goels et al, 1990, Cell 63, 1349-1356; Loewe et al, 1991, J Biol.Chem. 266, 21777-21783; Kumar et al., 1991, J. Biol.Chem. 266, 21777-21783). The reaction product migrated like pentose as expected and converted to the corresponding tetrasaccharide by treatment with Streptomyces α1,3 / 4-fucoosidase.

To label Man: Fuc and provide standard high mannose type N-glycans, radiolabeled glycans were prepared from HL-60 cell extracts labeled with [ ³ H] Man. To prepare glycopeptides, the cell extracts were precipitated with TCA and treated with pronase, as described by Cummings party (1983, J. Biol. Chem. 258, 15261-15273). From these glycopeptides, high mannose type N-glycan standards (Man ₉ GlcNAc ₁ , Man ₈ GlcNAc ₁ , Man ₇ GlcNAc ₁ , Man ₆ GlcNAc ₁ , Man ₅ GlcNAc ₁ ) were endo-β-N-acetylglucosa Prepared after treatment with minidase H [R.D. Cummings, 1993, Glycobiology: A Practical Approach (ed by M. Fukuda and A. Kobata), pp. 243-289, IRL Press at Oxford University Press, Oxford, UK.

[ ³ Relative molar ratio of radiolabeled monosaccharides in glycans labeled with H] GlcN

NeuAc, GlcNAc and GalNAcOH residues are radiolabeled in [ ³ H] GlcN-labeled disialylated hexose NeuAcα2 → 3Galβ1 → 4GlcNAcβ1 → 6 (NeuAcα2 → 3Galβ1 → 3) GalNAcOH prepared in HL-60 cells. , 1994, Meth. Enzymol. 230, 16-32). The relative molar ratio of the residues at equilibrium radiolabeling should be close to one. In order to determine the relative molar ratios for GlcNAc and GalNAc residues, the ³ H-GlcN-hexasaccharide was desialized to prepare ³ H-GlcN-saccharide; The tetrasaccharide was hydrolyzed in strong acid and the released radioactivity was confirmed by HPAEC as described below. All radioactivity was obtained at ³ H-GlcN and ³ H-GalNOH at a ratio of ³ H-GlcN: ³ H-GalNOH = 1.0: 0.8. This ratio was used as a correction factor in calculating the relative molar ratios of the isolated glycans, ie the radioactivity recovered in the ³ H-GalN (OH) of the sample after hydrolysis was divided by 0.8. The relative molar ratio is consistent with other studies on HL-60 cells radiolabeled by metabolism with ³ H-GlcN precursors (K. Maemura and M. Fukuda, 1992, J. Biol. Chem. 267, 24379-24386 ). To determine the relative molar ratios for NeuAc and GlcNAc residues, ³ H-GlcN-hexasaccharide was desialized with neuraminidase and the radioactivity released in NeuAc was measured. Radioactivity recovered from GlcN and NeuAc was 1.0: 1.4, respectively. Since the radioactivity of NeuAc was derived from the two residues of the desialized hexasaccharide, this gives a final value with a relative molar ratio GlcN: NeuAc = 1.0: 0.7.

Determination of Monosaccharide Composition

The ratio of GlcN: GalN and Man: Fuc in PSGL-1 and CD43 was measured after hydrolysis of the excised gel pieces containing purified PSGL-1 and CD43 with strong acid. The gel pieces were treated with 250 μl 2N trifluoroacetic acid (TFA) at 121 ° C. for 2 hours. The radiolabeled monosaccharides released in the hydrolyzate were identified by eluting with pH anion exchange chromatography (HPAEC) in a CarboPac PA-1 column (4 × 250 mm) and 16 mM NaOH for 30 minutes in a Dionex system. Fractions (0.3 min) were collected and radioactivity was measured with a scintillation counter. Radiolabeled monosaccharides were identified by comparison with retention times of standard monosaccharides. The relative molar ratio of GlcN: GalN to glycans was calculated by dividing the radioactivity in the GlcN peak by the radioactivity in the GalN peak and correcting for the difference in inactivation of ³ H-GlcN: ³ H-GalN. ^The Man: Fuc ratio in ³ H-Man-glycans was also determined using acid as described above, using a ratio of 1.0: 1.0 of Man: Fuc commonly observed after equilibrium radiolabeling of cells with [2- ³ H] Man. Measured after hydrolysis.

Other procedure

endo-β-gal of β-N-acetylhexosaminidase, β-galactosidase, arterobacter neuraminidase, streptomyces α1,3 / 4-fucoosidase, and Escherichia prepundy Enzymatic treatment of glycans with lactosidase is described in Cummings et al. (1989, Methods in Cell Biol. 32, 142-180); It was performed as described by Sribatic acid party (1992, J. Biol. Chem. 267, 2019 6-20203). Treatment of NDV-neuramidase was carried out at 20 [deg.] C. with 20mU of enzyme for 24 hours at 37 [deg.] C., followed by 20mU of phosphate, pH7.0, followed by further addition of 20mU of enzyme and further incubation for 24 hours. O-glycans are known as Al. As described by Cummings (1993, Glycobiology), analysis and purification by falling paper chromatography on Whatman filter papers for a defined time period using a pyridine / ethylacetate / water / acetic acid (5: 5: 1: 3) solvent system It was. Glycans were chemically defucoseized by treatment with 0.1 N TFA at 100 ° C. for 1 hour, as described by Spooners et al. (1984, J. Biol. Chem. 259, 4792-4801).

II. result

Purification of PSGL-1 and CD43

PSGL-1 and CD43 were purified from HL-60 cells metabolically labeled with ³ H-GlcN. PSGL-1 of HL-60 cell extracts labeled with ³ H-GlcN was purified by affinity chromatography on immobilized P-selectin followed by rechromatography of the bound EDTA-elution fraction. In fact, all radiolabels (90-99%) bound to P-selectin in the two purified material. This two-step process was required to remove about 120 kDa contaminating glycoprotein under non-reducing conditions remaining after the first step. Purified PSGL-1 behaved as a dimer of about 250 kDa under non-reducing conditions and about 120 kDa under reducing conditions. Purified PSGL-1 was dissolved in SDS-PAGE in 7.5% polyacrylamide under reducing or non-reducing conditions and analyzed by autoradiography.

CD43 ³ H-GlcN-labeled HL-60 cell extracts were immunoprecipitated with CD43-specific mAbs and secondary rabbit anti-mouse IgG1 antibodies or secondary antibodies alone. The immunoprecipitates were electrophoresed under non-reducing conditions on 10% acrylamide gels and then analyzed by autoradiography. A single band of 120 kDa for CD43 was observed under both reducing and non-reducing conditions.

Composition of radiolabeled sugars in PSGL-1 and CD43

In the initial evaluation of PSGL-1 and CD43 glycosylation, the ratio of GlcN: GalN of the purified glycoprotein was measured. ³ H-GlcN was metabolized with GlcNAc, GalNAc and sialic acid radiolabeled by animal cells. Gel pieces containing ³ H-GlcN-glycoprotein were treated with strong acid (destroying sialic acid) and radiolabeled GlcN and GalN were identified by Dionex HPAEC. The GlcN: GalN ratio is about 2: 1 for PSGL-1 and about 1: 1 for CD43. The GlcN: GalN ratio of about 1: 1 for CD43 is consistent with published evidence that most glycans of CD43 have a simple core-2 motif Galβ1 → 4GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH. The results demonstrated that the glycans of PSGL-1 contained more GlcNAc compared to GalNAc than the glycans of CD43.

The presence of the sLe ^x determinant in PSGL-1 leads to the prediction that PSGL-1 is heavily fucoseized. The amount of Fuc present in PSGL-1 and CD43 isolated from HL-60 cells metabolically radiolabeled with [2- ³ H] Man was analyzed. The labeled precursor by the metabolism of the cell is converted into 2- ³ H-Fuc, after equilibrium labeling relative specific activity of Man and Fuc is the same. ³ H-Man-PSGL-1 and ³ H-Man-CD43 were isolated by SDS-PAGE and autoradiography. The corresponding bands were hydrolyzed with strong acid and the separated monosaccharides were separated by HPAEC in the Dionex system. Man: Fuc ratio was determined to be 3: 5 for PSGL-1 and 3: 2 for CD43. As a control, the Man: Fuc ratio was also measured for total glycoprotein unpurified from HL-60 cells and found to be 3: 1. PSGL-1, CD43, and total HL-60 glycoproteins in the ³ H-Man and ³ H-Fuc relative amount of ^{3 H-Man-PSGL-1} , 3 H-Man-CD43 , and ³ H-labeled HL-60 of Man- Total glycoproteins of the cells were measured after hydrolysis with acid. Acid hydrolysates were analyzed by HPAEC using a Carbo-Pac PA-1 column eluted with 16 mM NaOH. The total radioactivity of each hazy peak was measured. PSGL-1 contains more Fuc residues than CD43 and more Fuc residues than the average glycoprotein of HL-60 cells.

The information can be used to estimate the number of Fuc residues in PSGL-1. The cDNA sequence of PSGL-1 predicts that PSGL-1 has three potential N-glycosylation sites. PSGL-1 only contains complex forms of N-glycans, each of which should have three Man residues. Thus, three complex forms of N-glycans of PSGL-1 represent 9 Man residues per mole, thus 15 Fuc residues per mole of PSGL-1. In contrast, CD43 contains only one N-linked glycan and has much less Fuc compared to PSGL-1. Collectively, compositional analysis of ³ H-Man-glycoprotein demonstrated that PSGL-1 was glycosylated differently than CD43.

Β-removal of O-glycans from PSGL-1 and CD43 and chromatography on Bio-Gel P-4 and P-10

O-glycans of PSGL-1 and CD43 were dropped directly by β-removal of the gel fragment containing purified glycoproteins with a mild base-borohydride and fractionated on a Bio-Gel P-4 column. ³ H-GlcN-glycans separated from PSGL-1 and CD43 were recovered in three fractions showing total radioactivity of 47%, 41% and 11%, respectively. Compared with CD43, PSGL-1 contained a relatively large glycan at a high rate.

P-4-I samples from PSGL-1 and CD43 were further purified by chromatography on a Bio-Gel P-10 column. Radioactivity of the P-4-I fraction from PSGL-1 was separated into two major peaks in Bio-Gel P-10, referred to as P-10-I and P-10-2. The population of glycans is larger than the disialylated core-2 hexasaccharide standard NeuAcα2 → 3Galβ1 → 4GlcNAcβ1 → 6 (NeuAcα2 → 3Galβ1 → 3) GalNAcOH. The main material in the P-4-II fraction of PSGL-1 was eluted at Bio-Gel P-10 at a position slightly smaller than the dialsylated Core-2 hexasaccharide standard (called P-10-3). In contrast, in all P-4-I and P-4-II fractions of CD43, most of the material was eluted the same in Bio-Gel P-10 as in the sialylated Core-2 hexasaccharide standard.

Glycans recovered in P-4-I from CD43 were analyzed using exoglycosidase treatment and anion exchange chromatography. The glycan was shown to be the expected sialylated core-2 hexasaccharide. A small fraction of the P-4-I sample from CD43 was recovered as a larger size glycan in Bio-Gel P-10, which is a polylacto with a small fraction of O-glycans expanded from the Core-2 motif from CD43. Consistent with previous studies of having a amine structure. Since the glycan structure of CD43 is described, it was not further analyzed.

Composition Analysis of P-10-1, P-10-2 and P-10-3 Glycans from PSGL-1

O-glycans released from Ser / Thr residues by β-elimination should contain ³ H-GalNAcOH at the reducing end which can be recovered as ³ H-GalNOH by subsequent strong acid hydrolysis. To determine which glycans in the mixture of PSGL-1 provide O-glycans, the GalNOH, GalN and GlcN contents of each glycan from Bio-Gel P-10 are ³ H-GlcN-PSGL-1 P-10 -1, P-10-2 and P-10-3 fractions were hydrolyzed with strong acid and then measured by HPAEC of Dionex system. Acid hydrolysates were analyzed by HPAEC using a Carbo-Pac PA-1 column eluted with 16 mM NaOH. The total radioactivity of each peak was measured. More than 95% of the total GalN (OH) was recovered as GalNOH in the PSGL-1 P-10-2 and P-10-3 fractions, indicating that β-elimination was effective and that O-glycans in PSGL-1 It was to prove that it appeared. Glycans of P-10-1 lack GalNOH and did not show O-glycans released by β-elimination. Instead, the P-10-1 fraction contained N-glycans still attached to the peptide. This was confirmed by the presence of ³ H-Man recovered from the glycan from ³ H-Man-PSGL-1. Small amounts of GalN recovered in the P-10-1 sample may be from GalNAc residues present in N-glycans or from small amounts of residue O-glycans still bound to the peptide and not released during the β-removal process. Could have been.

Sialation of P-10-1, P-10-2 and P-10-3 Glycans from PSGL-1

Sialization of O-glycans of P-10-2 and P-10-3 and N-glycans of P-10-1 was carried out by anion-exchange column chromatography in QAE-Sephadex before and after neuraminidase treatment. Determined. In this system, glycans with -1 charge (1 sialic acid) elute with 20 mM NaCl, glycans with -2 charge (2 sialic acid) elute with 70 mM NaCl, and -3 charges (3 Sialic acids) were eluted with 140 mM NaCl [24, 25]. P-10-1, P-10-2 and P-10-3 fractions of ³ H-GlcN-PSGL-1 were added to the QAE-Sephadex column, which was indicated before or after treatment with NDV neuraminidase. Eluted with NaCl at the concentration (mM). The glycans of the P-10-1 fraction were heterogeneously charged, consistent with the production of N-glycans in the glycopeptides of the fraction. After treatment with NDV neuraminidase, radioactivity eluted with 20 mM NaCl was analyzed by falling paper chromatography for 20 hours. True NeuAc moved 25 cm. P-10-2 glycans were mono- and disialized species and P-10-3 glycans were a mixture of neutral and monosialized species.

To determine if the anionic properties of the glycans were due to sialic acid and to define the binding of sialic acid, a portion of the O-glycans labeled with ³ H-GlcN was treated with NDV neuraminidase. The enzyme exhibits high specificity for α2,3-linked sialic acid residues and will not cleave efficiently other bonds of sialic acid [Paulson et al., 1982, J. Biol. Chem. 257, 12734-12738. After NDV neuraminidase treatment, the glycans of P-10-1 were less charged, consistent with the loss of sialic acid. However, the presence of glycans charged at residues indicates the expected profile for the N-glycans of glycopeptides. In contrast, the glycans of P-10-2 and P-10-3 became neutral after NDV neuraminidase treatment, demonstrating that all sialic acids of these glycans are α2,3-linked. Peaks of material eluting with 20 mM NaCl after NDV neuraminidase treatment were quantitatively recovered as free ³ H-NeuAc, as revealed by co-extraction with standard NeuAc in falling paper chromatography. ^It was previously established that sialic acid in PSGL-1 of HL-60 cells labeled with ³ H-GlcN is Neu5Ac. The results indicate that O-glycans of P-10-2 are mono- and disialized species, O-glycans of P-10-3 are neutral and sialated species, and sialic acid is α2,3- with respect to glycan. Proved binding. Moreover, the results demonstrated that O-glycans of PSGL-1 were not sulfided because neutral species occurred after NDV neuraminidase treatment.

Falling Paper Chromatography of P-10-2 Glycans from PSGL-1

Glycans of P-10-2 were further purified using preparative falling paper chromatography to recover two major species. One peak contained a larger sized glycan that slowly moved from the origin (called P-10-2a). Small peaks of faster moving material were also observed (about 24 cm), indicating some residue glycans derived from P-4-II peaks. When anion-exchange chromatography was performed on QAE-Sephadex, P-10-2a material was eluted as monosialized species, while P-10-2b material was eluted as disialized species.

Exoglycosidase Treatment of P-10-2b Glycans

The smaller size of the sialylated glycan in P-10-2b was desalialized by treatment with NDV neuraminidase, and the separated sialic acid was removed by chromatography on QAE-Sephadex. Desialized glycans were analyzed by descending paper chromatography before and after sequential exoglycosidase treatment (FIGS. 4A-E). After treatment with neuraminidase, desialized P-10-2b glycans were fractionated into two species. A few peaks (about 10%) migrated with fucose pentasaccharide standard Galβ1 → 4 (Fucβ1 → 3) GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH, referred to as P-10-2b ₁ . The major peak of the material (about 90%) shifted with Galβ1 → 4GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH per standard Core-2 saccharide, referred to as P-10-2b ₂ .

Desialized P-10-2b glycans were treated with β-N-acetylhexaxaminidase. The treatment did not change the migration of species in the sample, indicating that the glycans lacked terminal GlcNAc residues. However, when the desialylated P-10-2b mixture is treated with β-galactosidase, the P-10-2b ₁ glycan is not affected, while the P-10-2b ₂ glycan is degraded to standard Trisaccharide was shifted with GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH (FIG. 4B). Galβ1 → 3GalNAcOH structure is resistant to jackbin β-galactosidase because the enzyme does not efficiently cleave β1,3-bound terminal galactose residues (Lee et al., 1975, J. Biol. Chem. 250, 6786-6791). This was confirmed by a control experiment in which the standard disaccharide Galβ1 → 3GalNAcOH was not cleaved at the concentration of jackbin β-galactosidase used in the assay. Since P-10-2b ₂ glycans are derived from disialized species, the results show that P-10-2b ₂ glycans are derived from NeuAcα2 → 3Galβ1 → 4GlcNAcβ1 → 6 (NeuAcα2 → 3Galβ1 → 3) GalNAcOH Proved.

When the disialylated ³ H-GlcN-P-10-2b glycan is treated with Streptomyces α1,3 / 4-fucoosidase, the P-10-2b ₁ glycan is lost and all glycans recovered are cores. -2 was shifted with the standard Galβ1 → 4GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH (FIG. 4C). Desialylated P-10-2b glycans when treated with Streptomyces α1,3 / 4-fucoosidase, β-N-acetylhexosaminidase and β-galactosidase, P-10-2b _{Both 1} and P-10-2b ₂ glycans were completely digested to free ³ H-GlcNAc and ³ H-disaccharide Galβ1 → 3GalNAcOH (FIG. 4D). These results demonstrated that P-10-2b ₁ glycan has fucose pentasaccharide structure Galβ1 → 4 (Fucα1 → 3) GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH. Since the original P-10-2b glycan is a disialized species, the P-10-2b ₁ glycan contains sLe ^x antigens NeuAcα2 → 3Galβ1 → 4 (Fucα1 → 3) GlcNAcβ1 → R and has a structure of seven-phase NeuAcα2 → 3Galβ1 → 4 (Fucα1 → 3) GlcNAcβ1 → 6 (NeuAcα2 → 3Galβ1 → 3) GalNAcOH (FIG. 7),

To further confirm the presence of fucose in these glycans and to facilitate the identification of Fuc in other glycans, HL-60 cells were radiolabeled with [ ³ H-Fuc] by metabolism. O-glycans labeled with ³ H-Fuc were recovered by β-elimination as described for ³ H-GlcN-glycans following the same procedure described above. ³ H-Fuc recovered in the desialized P-10-2b fraction migrated with desialized ³ H-GlcN-P-10-2b ₁ glycan (FIG. 4E). Collectively, the results demonstrated that P-10-2b ₁ glycan of PSGL-1 was fucoseized and contained an sLe ^x structure.

Endo- and exoglycosidase treatment of P-10-2a glycans

Because of its relatively large size, P-10-2a glycans were likely to contain polylactosamine [3Galβ1 → 4GlcNAcβ1-] _n . To assess this possibility, P-10-2a glycans were desialized and sialic acid was removed by QAE-Sephadex chromatography. The resulting neutral glycans were treated with endo-β-galactosidase, which cleaves internal β1 → 4 galactose residues in type 2 polylactosamine [Kovata and Takasaki, 1993, Glycobiology: A Practical Approach (M. Fukuda and A. Kobata, pp. 165-185, IRL Press at Oxford University Press, Oxford, UK. The glycan was resistant to the treatment (FIG. 5A). Desialized P-10-2a glycans were also resistant to the combined treatment of β-galactosidase and β-N-acetylhexaxaminidase (FIG. 5A). At this time, there was a possibility that the P-10-2a glycan contained a polylactosamine backbone in which all internal GlcNAc residues were fucoseized. The polyfucosylated polylactosamine structure was resistant to endo-β-galactosidase (Fukuda et al., 1984, J. Biol. Chem. 259, 10925-10935). Complete defucosation was required before endo-β-galactosidase cleaved the chain.

Desialated P-10-2a glycans were chemically defucoseized and treated with endo-β-galactosidase. After removing the fucose, the enzyme quantitatively cleaves P-10-2a glycans to identify Galβ1 → 4GlcNAcβ1 → 3Gal, triglyceride GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH and disaccharide GlcNAcβ1 → 3Gal. Three major compounds were released at approximately the same molar ratio (FIG. 5B). The production of this product by the specific action of endo-β-galactosidase is Predicted for glycans having a backbone structure of which arrows indicate specific cleavage sites for endo-β-galactosidase of afucoseized polylactosamine containing glycans (Kovata and Takasaki) , 1993, Glycobiology). The chain length of polylactosamine can be deduced from the radioactivity recovered in disaccharide GlcNAcβ1 → 3Gal for other fragments. Recovery of trisaccharide GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH after endo-β-galactosidase treatment demonstrated that polylactosamine extends from GlcNAc to GalNAcOH residues by β1-6 binding. Recovery of Galβ1 → 4GlcNAcβ1 → 3Gal demonstrated that the Gal residue is at the non-reducing end of polylactosamine. As expected, desalialized and defucosylated P-10-2a glycans were treated with a combination of β-galactosidase and β-N-acetylhexaxaminidase, resulting in complete cleavage and ³ H- GlcNAc and ³ H-disaccharide Galβ1 → 3GalNAcOH were released (FIG. 4D).

Before fucose is removed, the resistance of desialylated P-10-2a to endo-β-galactosidase indicates that each GlcNAc residue in the glycan has the structure Galβ1 → 4 (Fucα1 → 3) GlcNAcβ1 → 3Galβ1 → 4 ( Fucα1 → 3) GlcNAcβ1 → 3Galβ1 → 4 (Fucα1 → 3) GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH was consistent with the possibility of containing α1,3-linked fucose residues. O-glycans containing incompletely fucoated polylactosamines are sensitive to endo-β-galactosidase.

To confirm the predicted fucose pattern for P-10-2a glycans, ³ H-Fuc-P-10-2a glycans were prepared from ³ H-Fuc-PSGL-1. After treatment with neuraminidase, desialylated ³ H-Fuc-glycans migrated with desialized ³ H-GlcN-P-10-2a glycans. When desialized ³ H-Fuc-P-10-2a glycan was treated with Streptomyces α1,3 / 4-fucoosidase, about one third of the radioactive Fuc fell off. The results were predicted based on the hypothesized glycan structure and the specificity of Streptomyces α1,3 / 4 fucosidase, which is capable of removing Fuc from the second GlcNAc residue at the end within the polyfucoseated glycan (Maemura And Fukuda, 1992, J. Biol. Chem. 267, 24379-24386; Sano et al., 1992, J. Biol. Chem. 267, 1522-1527). Collectively, the data demonstrated that the P-10-2a glycan contains three Fuc residues, one of which is in the terminal lactosamine unit.

P-10-2a glycans are monosialized and may have one of two possible structures (a or b):

To distinguish between these possibilities, ³ H-GlcN-P-10-2a sialated glycan was synthesized with or without β-galactosidase, β-N-acetylhexosaminidase, and streptococcus in the presence or absence of neuraminidase. Treated with a mixture of micros α1,3 / 4-fucoosidase. Structure (a) required neuraminidase along with other enzymes for complete cleavage, while structure (b) was expected to be degraded by exoglycosidase without neuraminidase. ³ H-GlcN-P-10-2a treated with neuraminidase alone released radiolabeled NeuAc. When glycans were treated with a mixture of β-galactosidase, β-N-acetylhexaminidase and streptomyces α1,3 / 4 fucosidase without neuraminidase, no radioactivity was released. Including neuraminidase, along with other exoglycosidases, completely degraded the glycans to release ³ H-NeuAc, ³ H-GlcNAc and ³ H-disaccharide Galβ1 → 3GalNAcOH. Collectively, the results indicate that a single sialic acid residue in the P-10-2a glycan is present at the terminal position of the polylactosamine chain, as shown in structure (a), and these glycans have an sLe ^x structure. Proved (Fig. 7).

Endo- and exoglycosidase treatment of P-10-3 glycans

P-10-3 glycans were mainly neutral or monosialized species. The dominant species (radioactivity of about 80) was bisialized. P-10-3 fractions were treated with neuraminidase and the separated sialic acid was removed by chromatography on QAE-Sephadex. About 90% of the desialylated species migrated with Galβ1 → 4GlcNAcβ1 → 6 (Galβ1 → 3) GalNAcOH with the standard core-2 tetrasaccharide and the rest with Galβ1 → 3GalNAcOH with the core-1 disaccharide (FIG. 6). Treatment of desialized P-10-3 glycans with β-galactosidase shifted the main peak mobility to the peak of the expected trisaccharide standard (FIG. 6). Treatment with the combination of β-galactosidase and β-N-acetylhexosaminidase produced free ³ H-GlcNAc and ³ H-Galβ1 → 3GalNAcOH (FIG. 6). About 20% of the radiolabels in P-10-3 glycans were found to be predominantly monosialized species that are converted to neutral species by neuraminidase treatment. The results demonstrated that P-10-3 glycans are mainly neutral core-2 tetrasaccharide and some monosialated core-2 pentasaccharide (FIG. 7). In addition, some of the neutral glycans had a core-2 disaccharide structure (FIG. 7).

Relative proportion of O-glycans in PSGL-1 of HL-60 cells

The relative ratio of O-glycans in PSGL-1 can be calculated from the amount of radioactivity recovered in each O-glycan and the relative molar ratio of ³ H-GlcN: ³ H-GalNOH: ³ H-NeuAc 1.0: 0.8: 0.7 (FIG. 7). Most of O-glycans in PSGL-1 contained a Core-2 structure. The few glycans recovered in P-10-2b ₁ and P-10-2b ₂ contained sLe ^x crystallites.

III. conclusion

This study demonstrated that PSGL-1 of human HL-60 cells contains O-glycans with Core-2 motif. Most glycans were not fucoseized and were a mixture of neutral and sialated species. A few O-glycans (about 14%) were fucoseized and contained terminal sLe ^x structures. There were two kinds of fucose O-glycans; One type was a disalylated lactose deficient in polylactosamine, and the other type was a unique monosialated, tri-fucoseized glycan containing polylactosamine. The presence of Core-2 O-glycans in PSGL-1 of HL-60 cells was consistent with the results of studies on glycosylation of PSGL-1 in human neutrophils. Desalialization of human neutrophils and HL-60 cells PSGL-1 is only involved in the treatment of endo-α-N-acetylgalactososaminidase (O-glycanase) that cleaves desialylated core-1 glycans. It was resistant to. Direct demonstration of sLe ^x determinants and polylactosamines of O-glycans in PSGL-1 derived from HL-60 provides indirect evidence that the structure is present in O-glycans of PSGL-1 derived from neutrophils. Reinforcement.

Significant differences were observed in O-glycans of PSGL-1 and CD43 in HL-60 cells. Although the two proteins predominantly have Core-2 O-glycans, PSGL-1 had a lot of neutral Core-2 tetrasaccharide, whereas CD43 had mostly disialylated Core-2 hexasaccharide. Core-2 structures are precursors for polylactosamine synthesis in O-glycans, but only PSGL-1 had a significant amount of polylactosamine. This indicates that a core-2 structure is required but not sufficient for polylactosamine addition. CD43 also lacked the two fucose O-glycans found in PSGL-1. Although monofucose O-glycans have been identified in CD43, the species exhibits only 0.5% O-glycans in CD43. The trifucose monosialized O-glycans identified in PSGL-1 were not found in CD43.

As mentioned above, P-selectin can weakly bind to various sulfated glycans, which inhibited the binding of P-selectin to human bone marrow cells. However, the O-glycans of PSGL-1 were not sulfided. Instead, PSGL-1 contained tyrosine sulfate, which is required for interaction with P-selectin but not E-selectin. Three consensus positions for tyrosine sulfidation occurred at the amino acid terminus of PSGL-1 at 46, 48 and 51 residues. PL1, the mAb for PSGL-1, recognized an epitope that spans 49-62 residues that interfere with binding to P-selectin and overlap with the tyrosine sulfidation site. Near the tyrosine sulfidation site there were two Thr residues indicating potential O-glycosylation sites at 44 and 57 residues. When PSGL-1 was coexpressed in COS cells with FTIII or FTVII, mutations at these residues reduced the binding of PSGL-1 to P-selectin. These results suggest that only one or two O-glycans linked with tyrosine sulfate residues may be sufficient to enhance the high affinity binding of PSGL-1 to P-selectin. However, O-glycans in other regions of the presenter may also contribute to the interaction with P-selectin and E-selectin.

Originally mucin-type glycoproteins have been proposed to act as convenient scaffolds where many O-glycans can be clustered to be recognized by selectin. However, the data indicate that mucin-type glycoproteins are glycosylated differently with other studies.

order

(1) General Information:

(i) Applicant: Board of Regents of the University of Oklahoma

(ii) Name of the Invention: Peptide and O-glycan inhibitors of selectin mediated inflammation

(iii) number of sequences: 5

(iv) contact address:

    (A) To: Patrea L. Fabst

    (B) Street: West Peachtree Street 1201 ONE Atlantic Center 2800

    (C) City: Atlanta

    (D) State: Georgia

    (E) Country: United States

    (F) Zip code: 30306-3450

(v) computer readable form:

    (A) Medium form: floppy disk

    (B) Computer: IBM PC Compatible

    (C) working system: PC-DOS / MS-DOS

    (D) Software: PatentIn Release # 1.0, Version # 1.25

(vi) Current application data:

    (A) Application number:

    (B) filing date:

    (C) Classification:

(viii) Attorney / Agent Information:

    (A) Statement: Fabst, Patrea L.

           (B) Registration Number: 31,284

           (C) Reference / Litigation Number: OMRF110CIP7

(ix) Communications Information:

    (A) Telephone: (404) 873-8794

    (B) Fax: (404) 873-8795

(2) Information about SEQ ID NO: 1

(i) Sequence Features:

    (A) Length: 402 amino acids

    (B) Class: Amino Acid

    (C) strands: single

    (D) topology: linear

(ii) Molecular type: peptide

(xi) sequence description:

[SEQ ID NO 1]

(2) Information about SEQ ID NO:

(i) Sequence Features:

    (A) Length: 2042 base pairs

    (B) Type: nucleic acid

    (C) strands: double

    (D) topology: linear

(ii) Molecular Type: cDNA

(xi) sequence description:

[SEQ ID NO 2]

(2) information about SEQ ID NO:

(i) Sequence Features:

    (A) Length: 2102 base pairs

    (B) Type: nucleic acid

    (C) strands: double

    (D) topology: linear

(ii) Molecular Type: cDNA

(xi) sequence description:

[SEQ ID NO 3]

(2) information about SEQ ID NO:

(i) Sequence Features:

    (A) Length: 34 base pairs

    (B) Type: nucleic acid

    (C) strands: single

    (D) topology: linear

(ii) Molecular type: DNA

(xi) sequence description:

[SEQ ID NO 4]

(2) information about SEQ ID NO:

(i) Sequence Features:

    (A) Length: 29 base pairs

    (B) Type: nucleic acid

    (C) strands: single

    (D) topology: linear

(ii) Molecular type: DNA

(xi) sequence description:

[SEQ ID NO: 5]

Claims (35)

Dialylated, monofucoated glycans Wherein R ₁ is H, sugar or aglycone and monosialated, trifucosylated glycan with polylactosamine backbone Wherein R ₂ is H, OH, sugar or aglycone. The purified O-glycan of claim 1, wherein R ₂ is OH. The purified O-glycan of claim 1, wherein R ₂ is aglycone. The purified O-glycan according to any one of claims 1 to 3, wherein R ₁ is aglycone. A pharmaceutically acceptable carrier for administration to a patient in need thereof with an O-glycan according to any one of claims 1 to 3, The treatment may include inhibiting the binding of PSGL-1 to P-selectin, regulating or reducing leukocyte adhesion, regulating or reducing tumor metastasis, blocking cell adhesion, inhibiting or controlling inflammation, inhibiting leukocyte mediated inflammation, leukocyte adhesion in ischemic myocardium. Reduction or control, reduction or control of reperfusion injury, reduction or control of organ damage mediated by leukocyte adhesion, reduction or control of seizures, reduction or control of vascular disease, reduction or control of mesenteric and peripheral vascular disease, in the vascular system Reduction or control of leukocyte adhesion, reduction or control of platelet aggregation in the vascular system, reduction or control of organ damage, reduction or control of leukocyte dependent organ damage, reduction or control of bacterial sepsis, reduction or control of diffuse vascular coagulation, Reduction or control of adult respiratory distress syndrome, reduction or control of leukocyte adhesion in pulmonary circulation, general Reduction or control of phosphorus vascular outflow, reduction or control of atherosclerosis, treatment of rheumatoid arthritis, treatment of autoimmune diseases, treatment or control of inflammatory diseases, prevention of myocardial infarction, treatment of acute respiratory distress syndrome, treatment of shock , Treatment of hypotension, treatment of thrombotic disease, control of coagulation, or reduction or control of circulating shock. A pharmaceutically acceptable carrier for administration to a patient in need thereof with an O-glycan according to claim 4, The treatment may include inhibiting the binding of PSGL-1 to P-cellictin, regulating or reducing leukocyte adhesion, regulating or reducing tumor metastasis, blocking cell adhesion, inhibiting or controlling inflammation, inhibiting leukocyte mediated inflammation, leukocyte adhesion in ischemic myocardium. Reduction or control, reduction or control of reperfusion injury, reduction or control of organ damage mediated by leukocyte adhesion, reduction or control of seizures, reduction or control of vascular disease, reduction or control of mesenteric and peripheral vascular disease, vascular system Reduction or control of leukocyte adhesion, reduction or control of platelet aggregation in the vascular system, reduction or control of organ damage, reduction or control of leukocyte dependent organ damage, reduction or control of bacterial sepsis, reduction or control of interstitial vascular coagulation Reduction or control of adult respiratory distress syndrome, reduction or control of leukocyte adhesion in pulmonary circulation, general Reduction or control of phosphorus vascular outflow, reduction or control of atherosclerosis, treatment of rheumatoid arthritis, treatment of autoimmune diseases, treatment or control of inflammatory diseases, prevention of myocardial infarction, treatment of acute respiratory distress syndrome, treatment of shock , Treatment of hypotension, treatment of thrombotic disease, control of coagulation, or reduction or control of circulating shock. Administering an effective amount of O-glycans sufficient to inhibit binding of PSGL-1 to P-selectin, wherein the O-glycans are disialized, monofucose glycans Wherein R ₁ is H, OH, sugar or aglycone; and Monosialated, Trifucose Glycans with Polylactosamine Backbone Wherein R ₂ is selected from the group consisting of H, OH, sugar or aglycone. 16. A method of inhibiting binding of PSGL-1 to P-selectin in vitro. 8. The method of claim 7, wherein R ₁ is wherein R ₂ is OH. 8. The method of claim 7, wherein R ₂ is aglycone. 10. The method of any one of claims 7-9, wherein R ₁ is aglycone. A therapeutic pharmaceutical composition comprising an O-glycan and a pharmaceutically acceptable carrier, The O-glycans are dissylated, monofucose glycans Wherein R ₁ is H or OH; and Monosialated, Trifucose Glycans with Polylactosamine Backbone Wherein R ₂ is H or OH; In addition The treatment may include inhibiting the binding of PSGL-1 to P-selectin, regulating or reducing leukocyte adhesion, regulating or reducing tumor metastasis, blocking cell adhesion, inhibiting or controlling inflammation, inhibiting leukocyte mediated inflammation, leukocyte adhesion in ischemic myocardium. Reduction or control, reduction or control of reperfusion injury, reduction or control of organ damage mediated by leukocyte adhesion, reduction or control of seizures, reduction or control of vascular disease, reduction or control of mesenteric and peripheral vascular disease, in the vascular system Reduction or control of leukocyte adhesion, reduction or control of platelet aggregation in the vascular system, reduction or control of organ damage, reduction or control of leukocyte dependent organ damage, reduction or control of bacterial sepsis, reduction or control of diffuse vascular coagulation, Reduction or control of adult respiratory distress syndrome, reduction or control of leukocyte adhesion in pulmonary circulation, general Reduction or control of phosphorus vascular outflow, reduction or control of atherosclerosis, treatment of rheumatoid arthritis, treatment of autoimmune diseases, treatment or control of inflammatory diseases, prevention of myocardial infarction, treatment of acute respiratory distress syndrome, treatment of shock A pharmaceutical composition comprising treating hypotension, treating thrombotic disease, controlling coagulation, or reducing or controlling circulating shock. The pharmaceutical composition of claim 11, wherein the treatment comprises inhibiting binding of PSGL-1 to P-selectin. The pharmaceutical composition of claim 11, wherein the treatment comprises regulating leukocyte adhesion. The pharmaceutical composition of claim 11, wherein the treatment comprises controlling inflammation. The pharmaceutical composition of claim 11, wherein the treatment comprises controlling tumor metastasis. The pharmaceutical composition of claim 11, wherein the treatment comprises reducing or controlling coagulation, seizures, or vascular disease. The method of claim 11, wherein the treatment comprises regulating or reducing leukocyte adhesion, regulating or reducing tumor metastasis, blocking cell adhesion, inhibiting inflammation, inhibiting leukocyte mediated inflammation, reducing or controlling leukocyte adhesion in ischemic myocardium, reducing reperfusion injury or Control, reduction or control of organ damage mediated by leukocyte adhesion, reduction or control of seizures, reduction or control of vascular disease, reduction or control of mesenteric and peripheral vascular disease, reduction or control of leukocyte adhesion in the vascular system, vascular system Reduction or control of platelet aggregation, reduction or control of organ damage, reduction or control of leukocyte dependent organ damage, reduction or control of bacterial sepsis, reduction or control of diffuse vascular coagulation, reduction or control of adult respiratory distress syndrome , Reducing or regulating leukocyte adhesion in the pulmonary circulation, reducing or adjusting general vascular outflow , Reduction or control of arteriosclerosis, treatment of rheumatoid arthritis, treatment of autoimmune diseases, treatment or control of inflammatory diseases, prevention of myocardial infarction, treatment of acute respiratory distress syndrome, treatment of shock, treatment of hypotension, thrombosis A pharmaceutical composition comprising treatment of, control of coagulation, or reduction or control of circulating shock. 18. The pharmaceutical composition of any one of claims 11 to 17, wherein the composition comprises an effective amount of O-glycans. A therapeutic pharmaceutical composition comprising an O-glycan and a pharmaceutically acceptable carrier, The O-glycans are dissylated, monofucose glycans Wherein R ₁ is sugar Monosialated, Trifucose Glycans with Polylactosamine Backbone Wherein R ₂ is a sugar; The treatment may include inhibiting the binding of PSGL-1 to P-selectin, regulating or reducing leukocyte adhesion, regulating or reducing tumor metastasis, blocking cell adhesion, inhibiting or controlling inflammation, inhibiting leukocyte mediated inflammation, leukocyte adhesion in ischemic myocardium. Reduction or control, reduction or control of reperfusion injury, reduction or control of organ damage mediated by leukocyte adhesion, reduction or control of seizures, reduction or control of vascular disease, reduction or control of mesenteric and peripheral vascular disease, in the vascular system Reduction or control of leukocyte adhesion, reduction or control of platelet aggregation in the vascular system, reduction or control of organ damage, reduction or control of leukocyte dependent organ damage, reduction or control of bacterial sepsis, reduction or control of diffuse vascular coagulation, Reduction or control of adult respiratory distress syndrome, reduction or control of leukocyte adhesion in pulmonary circulation, general Reduction or control of vascular outflow, reduction or control of atherosclerosis, treatment of rheumatoid arthritis, treatment of autoimmune diseases, treatment or control of inflammatory diseases, prevention of myocardial infarction, treatment of acute respiratory distress syndrome, treatment of shock, A pharmaceutical composition comprising treatment of hypotension, treatment of thrombotic disease, control of coagulation, or reduction or control of circulating shock. The pharmaceutical composition of claim 19, wherein the treatment comprises inhibiting binding of PSGL-1 to P-selectin. The pharmaceutical composition of claim 19, wherein said treating comprises regulating leukocyte adhesion. The pharmaceutical composition of claim 19, wherein the treatment comprises controlling inflammation. The pharmaceutical composition of claim 19, wherein said treating comprises regulating tumor metastasis. The pharmaceutical composition of claim 19, wherein the treatment comprises reducing or controlling coagulation, seizures, or vascular disease. 20. The method of claim 19, wherein the treatment comprises regulating or reducing leukocyte adhesion, regulating or reducing tumor metastasis, blocking cell adhesion, inhibiting inflammation, inhibiting leukocyte mediated inflammation, reducing or controlling leukocyte adhesion in ischemic myocardium, reducing reperfusion injury or Control, reduction or control of organ damage mediated by leukocyte adhesion, reduction or control of seizures, reduction or control of vascular disease, reduction or control of mesenteric and peripheral vascular disease, reduction or control of leukocyte adhesion in the vascular system, vascular system Reduction or control of platelet aggregation, reduction or control of organ damage, reduction or control of leukocyte dependent organ damage, reduction or control of bacterial sepsis, reduction or control of diffuse vascular coagulation, reduction or control of adult respiratory distress syndrome , Reducing or regulating leukocyte adhesion in the pulmonary circulation, reducing or adjusting general vascular outflow , Reduction or control of arteriosclerosis, treatment of rheumatoid arthritis, treatment of autoimmune diseases, treatment or control of inflammatory diseases, prevention of myocardial infarction, treatment of acute respiratory distress syndrome, treatment of shock, treatment of hypotension, thrombosis A pharmaceutical composition comprising treatment of, control of coagulation, or reduction or control of circulating shock. 26. The pharmaceutical composition of any one of claims 19-25, wherein the composition comprises an effective amount of O-glycans. A therapeutic pharmaceutical composition comprising an O-glycan and a pharmaceutically acceptable carrier, The O-glycans are dissylated, monofucose glycans Wherein R ₁ is aglycone; and Monosialated, Trifucose Glycans with Polylactosamine Backbone Wherein R ₂ is aglycone; The treatment may include inhibiting the binding of PSGL-1 to P-selectin, regulating or decreasing leukocyte adhesion, regulating or reducing tumor metastasis, blocking cell adhesion, inhibiting or controlling inflammation, inhibiting leukocyte mediated inflammation, leukocyte adhesion in ischemic myocardium. Reduction or control, reduction or control of reperfusion injury, reduction or control of organ damage mediated by leukocyte adhesion, reduction or control of seizures, reduction or control of vascular disease, reduction or control of mesenteric and peripheral vascular disease, in the vascular system Reduction or control of leukocyte adhesion, reduction or control of platelet aggregation in the vascular system, reduction or control of organ damage, reduction or control of leukocyte dependent organ damage, reduction or control of bacterial sepsis, reduction or control of diffuse vascular coagulation, Reduction or control of adult respiratory distress syndrome, reduction or control of leukocyte adhesion in pulmonary circulation, general Reduction or control of vascular outflow, reduction or control of atherosclerosis, treatment of rheumatoid arthritis, treatment of autoimmune diseases, treatment or control of inflammatory diseases, prevention of myocardial infarction, treatment of acute respiratory distress syndrome, treatment of shock, A pharmaceutical composition comprising treatment of hypotension, treatment of thrombotic disease, control of coagulation, or reduction or control of circulating shock. The pharmaceutical composition of claim 27, wherein the treatment comprises inhibiting binding of PSGL-1 to P-selectin. The pharmaceutical composition of claim 27, wherein the treatment comprises regulating leukocyte adhesion. The pharmaceutical composition of claim 27, wherein the treatment comprises controlling inflammation. The pharmaceutical composition of claim 27, wherein the treatment comprises controlling tumor metastasis. The pharmaceutical composition of claim 27, wherein the treatment comprises reducing or controlling coagulation, seizures, or vascular disease. The method of claim 27, wherein the treatment comprises regulating or reducing leukocyte adhesion, regulating or reducing tumor metastasis, blocking cell adhesion, inhibiting inflammation, inhibiting leukocyte mediated inflammation, reducing or controlling leukocyte adhesion in ischemic myocardium, reducing reperfusion injury or Control, reduction or control of organ damage mediated by leukocyte adhesion, reduction or control of seizures, reduction or control of vascular disease, reduction or control of mesenteric and peripheral vascular disease, reduction or control of leukocyte adhesion in the vascular system, vascular system Reduction or control of platelet aggregation, reduction or control of organ damage, reduction or control of leukocyte dependent organ damage, reduction or control of bacterial sepsis, reduction or control of diffuse vascular coagulation, reduction or control of adult respiratory distress syndrome , Reducing or regulating leukocyte adhesion in the pulmonary circulation, reducing or adjusting general vascular outflow , Reduction or control of arteriosclerosis, treatment of rheumatoid arthritis, treatment of autoimmune diseases, treatment or control of inflammatory diseases, prevention of myocardial infarction, treatment of acute respiratory distress syndrome, treatment of shock, treatment of hypotension, thrombosis A pharmaceutical composition comprising treatment of, control of coagulation, or reduction or control of circulating shock. 34. The pharmaceutical composition of any one of claims 27-33, wherein said composition comprises an effective amount of O-glycans. Inhibits binding of PSGL-1 to P-selectin, regulates or reduces leukocyte adhesion, controls or decreases tumor metastasis, blocks cell adhesion, inhibits inflammation, inhibits leukocyte-mediated inflammation, reduces or regulates leukocyte adhesion in ischemic myocardium, reperfusion Reduction or control of injury, reduction or control of organ damage mediated by leukocyte adhesion, reduction or control of seizures, reduction or control of vascular disease, reduction or control of mesenteric and peripheral vascular disease, reduction of leukocyte adhesion in the vascular system, or Control, reduction or control of platelet aggregation in the vascular system, reduction or control of organ damage, reduction or control of leukocyte dependent organ damage, reduction or control of bacterial sepsis, reduction or control of interstitial intravascular coagulation, adult respiratory distress syndrome Reduction or control of, reduction or control of leukocyte adhesion in the pulmonary circulation, reduction of general vascular outflow, or Regulation, reduction or control of atherosclerosis, treatment of rheumatoid arthritis, treatment of autoimmune diseases, treatment or control of inflammatory diseases, prevention of myocardial infarction, treatment of acute respiratory distress syndrome, treatment of shock, treatment of hypotension, In the pharmaceutical composition for the treatment of thrombotic disease, control of coagulation, or reduction or control of circulating shock, The pharmaceutical composition is a disialized, monofucose glycan Wherein R ₁ is H, OH, sugar or aglycone; and Monosialated, Trifucose Glycans with Polylactosamine Backbone Wherein R ₂ is H, OH, sugar or aglycone.