AGP-antibody fusion proteins and related molecules and methods

Abstract

Disclosed is a method of inhibiting the binding of a cell bearing a cell adhesion protein to a molecule or cell bearing a carbohydrate determinant specific for the cell adhesion molecule. The method involves contacting the cell adhesion protein-bearing cell with an AGP-antibody bearing the carbohydrate determinant. Also disclosed are AGP-antibody fusion proteins to which are covalently bonded carbohydrate moieties which interfere with the antibody portion's ability to fix complement and bind an Fc receptor. The methods of the invention may be used, for example, in any antibody-based therapy, for example, to reduce inflammation.

Description

BACKGROUND OF THE INVENTION

This invention relates to AGP-antibody fusion proteins, DNA, and uses thereof.

ELAM-1 is an integral membrane adhesion protein. It possesses an extracellular domain including an N-terminal lectin-related segment, an epidermal growth factor related repeat, and multiple complement regulatory protein motifs (Bevilacqua et al., Science 243:1160, 1989; Stoolman et al., Cell 56:907, 1989). ELAM-1 is specifically expressed on the surface of endothelial cells activated by the cytokines IL1 and tumor necrosis factor (TNF) (Bevilacqua et al., Proc. Natl. Acad. Sci. USA 84:9238, 1987), or the peptide hormone Substance P (Matis et al., J. Invest. Dermatol. 94:492, 1990). It mediates adhesion of myeloid cells (e.g., neutrophilic granulocytes) to cytokine-activated endothelial cells (Bevilacqua et al., Proc. Natl. Acad. Sci USA 84:9238, 1987). It has been suggested that ELAM-1 is involved in the regulation of inflammatory and immunological events at the interface of the blood and the blood vessel wall (Bevilacqua et al., Science 243:1160, 1989).

SUMMARY OF THE INVENTION

In a first aspect, the invention features a method of inhibiting the binding of a cell bearing a cell adhesion protein to a molecule or cell bearing a carbohydrate determinant specific for the cell adhesion molecule. The method involves contacting the cell adhesion protein-bearing cell with an α 1 -acid glycoprotein (AGP)-antibody fusion protein bearing the carbohydrate determinant.

In a second aspect, the invention features a method of reducing inflammation in a human patient involving administering to the patient a therapeutically-effective amount of an AGP-antibody fusion protein bearing a sialyl-Le X determinant.

In a third aspect, the invention features an AGP-antibody fusion protein to which there is covalently bonded a carbohydrate determinant specific for a cell adhesion protein.

In related aspects, the invention features purified nucleic acid encoding an AGP-antibody fusion protein containing sites for the attachment of a carbohydrate determinant which is specific for a cell adhesion protein; a vector including such nucleic acid; and a recombinant cell including such a vector.

In preferred embodiments of each of the above aspects, the cell adhesion protein is a selectin, such as ELAM-1; the carbohydrate determinant is sialyl-Le X ; the sialyl-Le X determinant is N-linked or O-linked; the AGP-antibody fusion protein contains multiple sialyl-Le X determinants; the AGP-antibody fusion protein includes, as an antibody portion, an IgG1 CH2, CH3, or hinge domain; the AGP-antibody fusion protein includes one or more of the N-linked glycan addition sites of α 1 -acid glycoprotein; the antibody portion of the AGP-antibody fusion protein bears one or more non-naturally occurring carbohydrate determinants; the carbohydrate determinant interferes with the antibody's ability to fix complement and bind an F c receptor (for example, due to a carbohydrate determinant attached to one or more of amino acids 274, 287, or 322 of the sequence shown in FIG. 1 ); and the AGP-antibody fusion protein is soluble.

In a final aspect, the invention features a method of protecting a mammal against any adverse immune reaction (including, without limitation, sepsis, organ damage attributable to inappropriate leukocyte extravasation, adult respiratory distress syndrome, glomerular nephritis, and rhumatoid arthritis), the method including administering to the mammal a therapeutically-effective amount of an AGP-antibody fusion protein. Preferably, the method involves treating a mammal for an adverse immune reaction which is induced by a microbial factor. Such microbial factors include, without limitation, gram-negative bacteria lipopolysaccharides (LPS), peptidoglycans from gram-positive organisms, mannan from fungal cell walls, polysaccharides, extracellular enzymes (e.g., streptokinase) and toxins (e.g., toxic shock enterotoxins of staphylococci). In other preferred embodiments, the method involves treating a mammal for any adverse immune reaction (see supra) which is induced by a host factor. Such host factors include, without limitation, metabolites of complement, kinin, and coagulation systems, factors released from stimulated cells (e.g., cytokines such as interleukin 1 (IL1) and tumor necrosis factor-α (TNF)), enzymes and oxidants from polymorphonuclear leukocytes (PMNs), vasopeptides (e.g., histamine), and products of the metabolism of arachidonic acid. In other preferred embodiments, the adverse immune reaction is induced by recombinant TNF-α or is induced by recombinant IL1. In yet other preferred embodiments, the adverse immune reaction is septic shock or is septicemia.

In other preferred embodiments, the AGP-antibody fusion protein bears one or more carbohydrate determinants capable of interfering with the antibody's ability to fix complement and bind an F c receptor; the antibody portion of the AGP-antibody fusion protein consists of the CH2 and CH3 antibody domains; and the antibody portion of the AGP-antibody fusion protein is an IgG1 domain.

By “cell adhesion protein” is meant a protein, present at some point in its in vivo existence on the cell surface, which mediates a specific interaction with a protein (e.g., a protein bearing a carbohydrate ligand) on the surface of a second cell.

By “carbohydrate determinant” is meant a moiety containing one or more carbohydrate groups which is present on a cell surface (at some point in its in vivo existence) and which interacts in a specific manner with a protein, e.g., a cell adhesion protein, e.g., on the surface of a second cell.

By “selectin” is meant a member of a family of cellular adhesion molecules that are characterized structurally by the presence of a lectin-like domain, an epidermal growth factor-like domain, a series of cysteine-rich repeats, a transmembrane domain and a short cytoplasmic tail.

By “non-naturally occurring” is meant, in this case, any carbohydrate determinant that is not one which is naturally bound to the fusion protein at that amino acid location.

By “inflammation” is meant a pathologic process consisting of cytologic and histologic reactions that occur in the affected blood vessels and adjacent tissues in response to an injury or abnormal stimulation caused by a physical, chemical, or biologic agent.

By “purified nucleic acid” is meant DNA that is free of the genes which, in the naturally-occurring genome of the organism from which the DNA of the invention is derived, flank the gene. The term therefore includes, for example, a recombinant DNA which is incorporated into a vector; into an autonomously replicating plasmid or virus; or into the genomic DNA of a prokaryote or eukaryote; or which exists as a separate molecule (e.g., a cDNA or a genomic or cDNA fragment produced by PCR or restriction endonuclease digestion) independent of other sequences. It also includes a recombinant DNA which is part of a hybrid gene encoding additional polypeptide sequence.

By “N-linked” is meant bonded to the amide nitrogen of an asparagine residue of a protein.

By “O-linked” is meant bonded to the hydroxyl-group oxygen of a serine, threonine, or hydroxylysine residue of a protein.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

The drawings will first briefly be described.

DRAWINGS

FIGS. 1A-1D (SEQ. ID NOS: 1, 2 and 7) shows the sequence of IgG1 and mutations designed to create N-linked glycan addition sites.

FIG. 2 is a graph showing the binding capability of α-ELAM-1 antibodies to a series of ELAM-1 sub-fragments and ELAM-1 fusion proteins.

FIG. 3 is a representation of the domain structure of an ELAM-1-IgG1 fusion protein.

FIG. 4 is a graph showing the degree of binding of a series of mammalian cell lines to an ELAM-1-IgG1 fusion protein.

FIGS. 5A-5F is a graph showing the degree of binding of a series of mammalian cell lines to antibodies directed against the cell surface molecules, CD15, CD63, or sialyl-Le X .

FIG. 6 is a graph showing binding of neuraminidase-treated myeloid cells to ELAM-1 or to α-CD15 antibody.

FIG. 7 is a graph showing inhibition of ELAM-1 bearing cell adhesion by α-sialyl-Le X antibody.

FIG. 8 is a graph showing the effect of amniotic fluid-derived sialyl-Le X determinant on binding of myeloid cells to ELAM-1.

FIG. 9 is a graph showing the ability of in vitro-fucosylated α 1 -acid glycoprotein to bind ELAM-1.

FIG. 10 is a graph showing the effect of IL1 and IL-8 on cell surface expression of sialyl-Le X .

FIG. 11 is a graph showing the effect of IL1 and IL-8 on the adhesion of myeloid cells to ELAM-1.

FIG. 12 is a graph showing the plasma half-life of AGP (dotted line), AGP fused to the CH2 and CH3 domains of human IgG1 (solid line, solid triangles), and AGP fused to the hinge, CH2, and CH3 domains of human IgG1 (solid line, open squares).

FIG. 13 is a protein gel showing an increase in molecular weight of immunoglobulin fusion proteins upon introduction of additional glycosylation sites into the CH2 domain of human IgG1.

FIG. 14 is a bar graph showing inhibition of adhesion of HL-60 cells and T-lymphocytes to ELAM-1 by an AGP-IgG1 fusion protein.

FIGS. 15A and 15B is the nucleotide sequence (Panel A; SEQ ID NO: 3) and amino acid sequence (Panel B; SEQ ID NO: 4) of an AGP-IgG1 fusion protein.

FIGS. 16A and 16B is the nucleotide sequence (Panel A; SEQ ID NO: 5) and amino acid sequence (Panel B; SEQ ID NO: 6) of an CD4-IgG1 fusion protein.

Antibody bearing multiple sialyl-Le X determinants

In one embodiment, the invention features an antibody bearing one or more carbohydrate side chains which mask the CH2 portion of the immunoglobulin molecule and thus inhibit complement fixation and F c receptor binding. Such antibodies are useful for disrupting undesirable interactions between cells or proteins, or, generally, for disrupting an interaction between any two molecules, one of which bears a determinant specifically recognized by an antibody. Because the carbohydrate moieties block the immunoglobulin domain which triggers complement fixation and F c receptor binding, such antibodies do not elicit the undesirable side effects (i.e., those resulting from complement fixation and F c receptor binding) frequently associated with antibody-based therapies. Preferably, the carbohydrate groups serve not only to inhibit undesirable complement fixation and F c receptor binding, but also perform the function of competitively inhibiting a carbohydrate ligand-cell adhesion protein interaction. Where the carbohydrate groups perform this function, the antibody generally does not serve any function arising from its specificity, but serves only as a carrier for the carbohydrate groups. There is described below such a molecule, in which the carbohydrate side chain includes the sialyl-Le X determinant. Sialyl-Le X normally acts to facilitate interaction between cells which bear it (e.g., neutrophils) and cells which bear the protein, ELAM-1 (e.g., endothelial cells, e.g., those lining the blood vessel walls). Disrupting this interaction has therapeutic applications, for example, in minimizing inflammation, such as that which occurs following tissue injury, e.g., myocardial infarction, which is characteristic of diseases such as psoriasis or rheumatoid arthritis, or for preventing or inhibiting septicemia or septic shock which is induced by a microbial- or host-mediated immune reaction.

Because N-linked glycan addition sites are well known to be: N X S/T (where N is asparagine, S is serine, T is threonine, and X is any amino acid except proline), we designed a molecule including several such sites for attachment of sialyl-Le X side chains. Inspection of the IgG1 sequence ( FIG. 1 ; SEQ. ID NOS: 1 and 7) reveals at least five sites at which N-linked glycan addition sites may be introduced into the molecule in advantageous locations, where complement fixing and F c receptor binding ability will be impaired by the process These sites (i.e., beginning at amino acid residues 274, 287, 295, 322, and 335), although these are preferred sites of N-linked glycan addition they are not the only candidates; other useful sites may be identified and incorporated into the IgG1 sequence using, as guidance, the following criteria: (1) The sites are, preferably, located in the CH2 region of the immunoglobulin molecule, i.e., in the portion of the molecule responsible for complement fixation and F c receptor binding; (2) the sites are located in regions of the sequence, predicted by their hydrophilic nature, to be present on the outside of the immunoglobulin molecule and therefore accessible to the enzymes responsible for attachment of carbohydrate side chains; (3) the sites are located in a region which is minimally disruptive to the primary amino acid sequence and, thus, the predicted secondary amino acid structure. For example, a naturally-occurring site which differs from an N-linked glycan addition site by a single amino acid would be preferable to a site requiring two alterations in the amino acid sequence. Moreover, it is preferable to create an N-linked glycan addition site by subsituting amino acids of similar charge or polarity (e.g., substitution of one uncharged amino acid for another). One or more N-linked glycan addition site substitutions may be engineered into the IgG1-encoding sequence; such sequences (i.e., those which encode an antibody molecule to which sialyl-Le X moieties are attached) are termed IgG1-sialyl-Le X or IgG1-Le X .

A particular IgG1 molecule bearing sialyl-Le X moieties is produced as follows. The IgG1 gene is publically available, and its sequence is shown in FIG. 1 (SEQ.ID NO.:1). The gene is mutagenized by standard methods of in vitro site-directed mutagenesis in order to introduce one or more N-linked glycan addition sites (e.g., those described above and shown above the naturally-occurring sequence in FIG. 1 ; SEQ ID NO:2). The gene is then inserted into a vector designed to express the protein in a eukaryotic cell (see, e.g., those vectors described in Gillies et al., U.S. Pat. No. 4,663,281, hereby incorporated by reference). The eukaryotic host cell is preferably a mammalian cell (e.g., a CHO or lec11 cell), and the expression vector containing the mutated IgG1-Le X -encoding sequence is introduced into the host cell by transient or stable transfection using standard techniques. Such host cells are also transfected (transiently or stably) with a vector capable of expressing an α(1,3)fucosyltransferase capable of attaching the sialyl-Le X groups to the antibody molecule at the glycosylation sites. The α(1,3)fucosyltransferase gene may be expressed from a vector distinct from that encoding IgG1-Le X , or both genes may be carried on, and expressed from, a common vector. Mammalian cells are particularly useful hosts for the synthesis of IgG1-Le X because they provide all required precursors for sialyl-Le X production.

An α(1,3)fucosyltransferase cDNA is described in Lowe et al. ( Cell 63:475, 1990). The α(1,3)fucosyltransferase enzyme, encoded by this cDNA, recognizes a sialylated precursor molecule and adds either an α(1,3)- or an α(1,4)-linked fucose moiety to N-acetylglucosamine side chains. The sialyl-Le X determinant is characterized by an α(1,3)-linkage, and, as such, the α(1,3)fucosyltransferase enzyme of Lowe (supra) produces both the desired sialyl-Le X -modified molecules and products bearing α(1,4)-linked fucose which, although not active in binding to ELAM-1, do not interfere with the action of the sialyl-Le X -modified molecules nor produce other undesirable side effects.

Production of IgG1-sialyl-Le X would be more efficient, however, if an α(1,3)fucosyltransferase was utilized which exclusively catalyzed α(1,3) fucose linkages. Such a mammalian enzyme exists, and the cDNA therefor can be isolated as follows. A cDNA library, prepared from mRNA which is isolated from a myeloid cell line (e.g., HL-60), is inserted into a mammalian cell expression vector such as πH3M (see, Simmons et al., Nature 331:624, 1988; Aruffo and Seed, Proc. Natl. Acad. Sci. USA 84:8573, 1987) and transfected into a mammalian cell line, preferably, COS7 cells (as described in Seed and Aruffo, Proc. Natl. Acad. Sci. USA 84:3365, 1987). The proper cDNA clone is isolated by the immunoselection procedure described in Aruffo and Seed, supra; Seed and Aruffo, supra; and U.S. patent application Ser. No. 07/379,076, hereby incorporated by reference. Transfected cells are harvested and incubated with monoclonal antibodies PM-81 (anti-CD15; Medarex, West Lebanon, N.H.), FMC13 (anti-CD15; Sera-Lab/Accurate Chemical and Scientific, Westbury, N.Y.), MC1 (anti-CD15; Sera-Lab. Westbury, N.Y.) and VIM8 (anti-CD65). Following incubation (e.g., for 1 hour), cells are separated from free antibody by centrifugation through a cushion of 2% Ficoll in PBS and allowed to settle on plastic dishes coated with affinity-purified goat antibodies to mouse IgM (as described below). Adherent cells are collected and a Hirt supernatant containing episomal DNA is prepared. The purified Hirt supernantant DNA is transformed (e.g., by electroporation) into E. coli (preferably, E. coli MC1061/p3) by standard techniques and ampicillin- and tetracycline-resistant colonies selected by standard methods. Antibiotic resistant colonies are then pooled and the plasmids amplified (e.g., following addition of spectinomycin hydrochloride overnight). The resulting culture is converted to spheroplasts and the spheroplasts fused to COS7 cells by standard procedures (see, for example, Seed and Aruffo, supra). Cells are allowed to incubate (preferably, 2 to 3days) and are exposed to antibodies as described above. Preferably, two rounds of spheroplast fusion and “panning” (i.e., the procedure described above) are performed, and the bacterial cells resulting from the last round of panning are collected, and their plasmid DNA prepared. This cDNA encodes an enzyme, i.e., an α(1,3)fucosyltransferase, which directs the appearance of the desired CD15 and CD65 determinants, i.e., the sialyl-Le X determinant.

Host cells expressing α(1,3)fucosyltransferase and IgG1-Le X (and thus producing an IgG1 molecule bearing sialyl-Le X determinants) are grown by standard methods and the IgG1-Le X protein is purified from a cell lysate based on its affinity for a Protein A column or any other standard technique of antibody isolation and purification.

Use

For administering such a compound to a patient, the pharmaceutically-pure IgG1-Le X is suspended in an acceptable carrier, e.g., physiological saline, and is delivered to patients intravenously in a single or in multiple doses. Optimally, a sufficient quantity of IgG1-Le X is provided to saturate all ELAM-1-binding sites on an endothelial cell. Typically, this may be achieved with doses of 0.1 mg/kg or greater. The preferred dosage is in the range of 0.1-2.0 mg/kg.

The carrier molecule, for example sialyl-Le X -modified α 1 -acid glycoprotein-antibody fusion proteins (described below) would be produced generally as described herein and would be administered intravenously to patients as described above (i.e., preferably, at a dose sufficient to saturate all cellular ELAM-1 binding sites, e.g., at 0.1 mg/kg or greater).

IgG1-sialyl-Le X or sialyl-Le X -modified a 1 -AGP-antibody fusion proteins may be used, in one example, for the treatment of a patient suffering from a heart attack. Following such a trauma, the endothelial cells lining the blood vessels express ELAM-1 on their surfaces and, without treatment, neutrophils, bearing sialyl-Le X on their surfaces, bind such ELAM-1-bearing endothelial cells, contributing to inflammation. Treatment with a sialyl-Le X bearing molecule would attenuate the inflammation by competitively inhibiting the interaction between the invading neutrophils and the blood vessel endothelial cells in the vicinity of the heart. Compounds such as IgG1 sialyl-Le X or sialyl-Le X -modified α 1 -AGP-antibody fusions may also be used, as described above, for the treatment of septic shock, septicemia, or any disease characterized by chronic inflammatory conditions, e.g., rheumatoid arthritis, psoriasis, or pemphigus vulgaris. In addition, antibodies or antibody fusion proteins according to the invention may be used in conventional techniques of antibody based therapies or in vivo diagnostics, taking advantage of the antibody's specificity to target therapeutic diagnostic sites. Again, attached carbohydrates determinants mask the CH2 domain of the antibody and block the undesirable effects of complement fixation and F c receptor binding.

Experimental Information

Sialyl-Lewis X (sialyl-Le X ) determinants were shown to interact with ELAM-1 and facilitate binding to ELAM-1 bearing endothelial cells by the following experiments. These examples are presented to illustrate, not limit, the invention.

Recognition by ELAM-1 of the Sialyl-Le X Determinant

The ELAM-1 domains necessary for granulocyte-binding activity were localized using two monoclonal anti-ELAM-1 antibodies: H18/7, which effectively blocks leukocyte adhesion to activated endothelium, and H4/18, which does not (Pober et al., J. Immunol. 136:1680, 1986; Bevilacqua et al., Proc. Natl. Acad. Sci. USA 84:9238, 1987). Full length ELAM-1 was expressed from the cDNA carried on plasmid, pELAM-1 (Bevilacqua et al., Science 243:1160, 1989). Carboxyl terminal deletions of the ELAM-1 cDNA were created by polymerase chain reaction to produce the proteins shown in FIG. 2 . Primer sequences for the PCR deletions were designed based on the full-length ELAM-1 sequence of Bevilacqua et al. ( Science 243:1160, 1989). Once generated, ELAM-1 cDNA fragments were then fused, by standard techniques, to the transmembrane and intracellular coding portions of a CD7 cDNA (i.e., nucleotides 501 to 1236 of the CD7 cDNA described in Aruffo and Seed, EMBO. J. 6:3313, 1987). Plasmids bearing the resulting fusions were transfected into COS cells. Reactivity to monoclonal antibodies was determined by indirect immunofluorescence microscopy of fixed, permeabilized cells by the method of Aruffo et al., Cell 61:1303, 1990. The results of this analysis are shown in FIG. 2 and are representative of transfections of three to six independent isolates of the constructs shown. FIG. 2 indicates that H18/7 binding required the lectin-related segment plus the EGF-repeat domains, while H4/18 reactivity required, in addition, the first three complement regulatory protein repeat elements. L indicates the lectin-related segment; EGF indicates the EGF-related repeat segment; and CR1-CR6 indicate complement regulatory protein elements.

To further define the binding site for H18/7, a fragment was exchanged between the ELAM-1 cDNA and the equivalent fragment of the related Leu8 (LECCAM1) cDNA (described in Camerini et al., Nature 342:78, 1989). Leu8 (LECCAM-1)/ELAM-1 chimeras were created by restriction fragment interchange from a conserved BglII site within the lectin domain (i.e., at nucleotide 454 and 475 of the ELAM-1 and Leu8 cDNA sequences, respectively). As shown in FIG. 2 , H18/7 bound to an antigenic determinant located principally in the first 75% of the ELAM-1 lectin domain. Together with the above result (i.e., that both the lectin-like and EGF-repeat-like domains were required for H18/7 binding to truncated ELAM-1) suggests that the EGF-related repeat element may play a role in shaping the structure of the lectin domain.

To study the possible lectin-carbohydrate interactions suggested by the epitope mapping, a soluble ELAM-1 protein chimera was prepared by the fusion of a cDNA fragment encoding the ELAM-1 extracellular domain to a genomic fragment encoding the hinge, i.e., the CH2 and CH3 domains, of human IgG1 (Aruffo et al., Cell 61:1303, 1990; FIG. 3 ). The ELAM-1-IgG1 chimera was prepared as follows. Synthetic oligonucleotides having the sequence: CGGAATTCCAGTACTACTCACCTGGTCCGCCGATGGTCTCCGGGC (SEQ. ID NO.: 8) and CCAGATATACGCGTTGACATTGATTATTGACTAGTTATT (SEQ. ID NO.:9), and corresponding to the splice donor/carboxyl terminus of ELAM-1 and to a location in the vector upstream of the inserted cDNA, respectively, were prepared by standard techniques. Polymerase chain reaction with these oligonucleotides and the ELAM-1 cDNA expression plasmid, pELAM-1, as template, yielded an 1800 bp fragment which was digested with XhoI and EcoRI and subcloned into XhoI/EcoRI-digested expression vector πH3M (Aruffo and Seed, Proc. Natl. Acad. Sci. USA 84:8573, 1987). The subcloned fragment was released by digestion with XhoI and ScaI and ligated to the XhoI/ScaI-digested IgG1 expression plasmid described in Aruffo et al. Cell 61:1303, 1990. The resulting construct was transfected into COS cells (as described in Seed and Aruffo, Proc. Natl. Acad. Sci 84:3365, 1987), and the desired fusion protein, termed ELAM-Rg, was recovered from the supernatant by adsorption to and elution from protein A-agarose as described in Aruffo et al. ( Cell 61:1303, 1990). The initial construct, and a subsequent version in which the majority of the PCR-amplified segment (i.e., nucleotides 1 to 1464 of the ELAM-1 sequence) was replaced by a homologous restriction fragment interchange (to avoid potential mutations introduced during amplification) showed identical binding activity. The soluble protein appeared in the form of disulfide-linked dimers, presumably mediated by the hinge region cysteine residues responsible for the inter-heavy chain linkage of active immunoglobulins.

To determine whether myeloid cells bound soluble ELAM-1, plastic dishes precoated with goat-anti-human IgG antibodies, were incubated with supernatants expressing ELAM-Rg. These experiments were carried out as follows. Human granulocytes were isolated from freshly drawn, heparinized whole blood by centrifugation through Ficoll/sodium diatrizoate (Mono-Poly Resolving Medium, Flow Laboratories, McLean, Va.) for 20 min. at 800×g. Cell lines were obtained from the American Type Culture Collection (ATCC) and were maintained in IMDM with 10% fetal bovine serum as described in Aruffo and Seed ( Proc. Natl. Acad. Sci. USA 84:8573, 1987). Adhesion to ELAM-Rg was carried out in bacterial culture dishes (Falcon 1008, Becton-Dickinson Labware, Lincoln Park, N.J.) to which affinity-purified goat anti-human IgG antibody (Organon Teknika/Cappel, Malverne Pa.) had been allowed to adsorb at 10 μg/ml in 50 mM Tris-HCl, pH 9.5 for at least one hour. Remaining protein binding sites were then blocked by overnight incubation with 1 mg/ml bovine serum albumin in Phosphate Buffered Saline (PBS; 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na 2 HPO 4 .7H 2 O, 1.4 mM KH 2 PO 4 , pH7.3). Dishes were incubated with ELAM-Rg (≈1 μg/ml) for 30 min. at 22°, washed with PBS, overlaid with cells (10 6 cells) for 10 min. at 22°, and washed three times with PBS. The adherent cells per unit area of dish were enumerated with the aid of an ocular reticle and scored as follows: >100 cells, +++++; 100-75 cells, ++++; 75-50 cells, +++; 50-25 cells, ++; and 25-10 cells, +. All values represented the average of triplicate determinations.

The treated plastic acquired the ability to specifically bind granulocytes as well as the myeloid cell lines HL60 and THP1 (FIG. 4 ). Other cell lines of both hematopoietic origin (i.e., U937 and HSB-2) and nonhematopoietic origin (i.e., WIDR) were found to bind to the ELAM-1 coated plastic as well (FIG. 4 ). Dishes coated with CD8 fusion protein (Aruffo et al., Cell 61:1303, 1990) showed negligible affinity (for granulocytes or any of the cell lines tested (Bevilacqua et al., Proc. Natl. Acad. Sci. USA 84:9238, 1987).

To correlate the binding activity with surface phenotype, various monoclonal antibodies recognizing known granulocyte carbohydrate antigens were screened for reactivity with the above cell types. 5×10 5 cells were incubated with the following antibodies (as ascites at 1:250 dilution, or as purified antibody at 4 μg/ml): PM-81 (anti-CD15; Medarex, W. Lebanon, N H; Ball et al., J. Immunol. 130:2937, 1983), CSLEX1 (anti-sialyl-Le X ; Fukushima et al., Cancer Res. 44:5279, 1984); or VIM2 (anti-CD65; Macher et al., J. Biol. Chem. 263:10186, 1988), followed by a fluorescein-conjugated goat anti-mouse IgG+IgA+IgM antibody (Organon Teknika/Cappel, Malverne, Pa.).

Results are shown in FIG. 5 . Sparse dots represent the negative control (no primary antibody); dense dots, anti-CD15 mAb; solid line, anti-CD63 mAb; and broken line, anti-sialyl-Le X mAb. The U937 line used herein lacked sialyl-Le X determinants, unlike the related U937 cell line tested by Terasaki and coworkers (Fukushima et al., Cancer Res. 44:5279, 1984). An initial survey showed that ELAM-1 adhesion potential correlated with the presence of the CD15 determinant (i.e., Le X , or lacto-N-fucopentaose III; Gooi et al., Nature 292:156, 1981; Huang et al., Blood 61:1020, 1983; Magnan et al., Arch. Biochem. Biophys. 233;501, 1984; Gooi et al., Eur. J. Immunol. 13:306, 1983; Tetteroo et al., Eur. J. Immunol. 14:1089, 1984), but not with the determinants associated with CD17 (lactosyl ceramide; Symington et al., J. Biol. Chem. 259:6008, 1984), CD65 (VI 3 NeuAcIII 3 FucnorLcnOse 6 Cer; Macher et al., J. Biol. Chem. 263:10186, 1988) or sulfatides (Fredman et al., Biochem. J. 251:17, 1988).

To further test the correlation between ELAM-1 adhesion potential and the presence of CD15, cells bearing CD15 were treated with neuraminidase, an enzyme known to cleave terminal sialyl groups. HL60 cells (10 6 /plate) were incubated in 50 μl of 0.15M NaCl, 4 mM CaCl 2 , pH 5.5, for 1 hr. at 37° in the presence or absence of 41.5 mU of neuraminidase (from Vibrio cholerae, type II, Sigma, St. Louis, Mo.). Cells were washed three times with PBS and adherence to either ELAM-Rg- or PM-81-coated dishes was scored as described above. Dishes were coated with ELAM-Rg as described above, and with purified PM-81 antibody at 10 μg/ml in 50 mM Tris-HCl pH 9.5. Adherence assays were carried out as described above. Results shown in FIG. 6 are expressed as percent of control and were calculated from the mean±standard deviation for the average of triplicate determinations in three independent experiments.

FIG. 6 indicates that the correlation of ELAM-1 adhesion potential with the presence of CD15 was imperfect because digestion of the cells with neuraminidase abolished binding to ELAM-1 but increased binding to immobilized anti-CD15 antibodies.

Association with CD15 and sensitivity to neuraminidase suggested that the sialylated form of the CD15 carbohydrate antigen might represent the physiological ELAM-1 ligand. To test this idea, CSLEX1 monoclonal antibody was assayed for its ability to inhibit adhesion of HL60 cells to ELAM-Rg. 10 6 HL60 cells were incubated with CSLEX1 (1:50 in PBS) for 30 min. on ice, then crosslinked with affinity purified goat anti-mouse IgM antibody (Organon Teknika/Cappel, Malverne, Pa.) at 20 μg/ml in PBS for 30 min., and fixed with 2% formaldehyde in PBS for 20 min. at 22°. Cells were washed three times in PBS/1% glycine and incubated with ELAM-1-Rg-coated dishes as described above. The data (presented as percent of control) in FIG. 7 represents the mean±standard deviation of triplicate determinations in three independent experiments.

FIG. 7 indicates that there exists a very good correspondence between the surface density of sialyl-Le X and the rank order of the number of cells bound per unit area of ELAM-Rg coated plastic. In addition, anti-sialyl-Le X antibody completely inhibited adhesion of myeloid cells to ELAM-1, whereas anti-CD65 and anti-CD15 antibodies had no activity under identical conditions.

The carbohydrate epitope recognized by the CSLEX1 antibody has been reported to be NeuNAcα2-3Galβ1-4(Fucα1-3)GlcNAcβ1-3Gal, based on motifs common to structurally characterized glycolipids with which the antibody reacts (Fukushima et al., Cancer Res. 44:5279, 1984). Chemical analysis of the fucosylated lactosaminoglycans of neutrophilic granulocytes has shown that both the Le X (CD15) and sialyl-Le X determinants are predominantly represented on tetraantennary asparagine-linked glycans whose individual strands are built up from poly(N-acetyllactosamine) chains bearing variable α(1,3)-linked fucose substitutions (Fukuda et al., J. Biol. Chem. 259:10925, 1984; Spooncer et al., J. Biol. Chem. 259:4792, 1984). Serological evidence supports the existence of the sialyl dimeric Le X determinant on granulocytes as well (Fukushi et al., J. Biol. Chem. 259:10511, 1984; Fukushi et al., Cancer Res. 45:3711, 1985). As such, the residue on the reducing side of the sialyl-Le X group is galactose in all of the granulocyte structures thus far identified. Although antibody CSLEX1 blocks binding, the structure recognized by ELAM-1 might be more complex than the structure recognized by CSLEX1. To establish the minimum glycan structure for ELAM-1 binding, chemically-characterized glycans bearing sialyl-Le X determinants were evaluated for ELAM-1 recognition.

Amniotic fluid is one source of well defined sialyl-Le X determinant which is found in a very different context than the granulocyte cell surface. The sialyl-Le X -bearing carbohydrate of amniotic mucins is joined β1-6 to a 3-substituted N-acetylgalactosamine, which in turn is attached directly to the polypeptide backbone through O-linkage to serine or threonine (Hanisch et al., Carbohydr. Res. 178:29, 1988). Amniotic fluid-derived sialyl-Le X determinants were tested for their ability to block binding of myeloid cells to immobilized ELAM-1. Human amniotic fluid (HAF) was either used without purification, fractionated by centrifugal ultrafiltration (100 kDal nominal cutoff; Centricon 100, Amicon, Danvers Mass.), or fractionated, following phenol extraction, by size exclusion chromatography (Sephacryl S-300 HR) in 4M guanidinium chloride (by the method of Hanisch et al., Carbohydr. Res. 178:29, 1988) to yield purified mucins. Such mucins were used at a protein concentration of approximately 150 μg/ml. Binding to ELAM-Rg coated plastic was performed as described above. Results (expressed as percent of control) shown in FIG. 8 are the average of triplicate determinations in two independent experiments.

FIG. 8 indicates that, despite the dissimilarity between the granulocyte glycans and the amniotic fluid mucins, purified amniotic fluid mucins, as well as unfractionated amniotic fluid (which appears to contain all of its activity in the mucin-rich high molecular weight fraction), efficiently blocked binding of myeloid cells to immobilized ELAM-1.

Another source of sialyl-Le X determinants is fucosylated α 1 -acid glycoprotein (α 1 -AGP) (Biou et al., Biochim. Biophys. Acta. 913:308, 1987; Wieruszeski et al., FEBS Lett. 238:390, 1988). Chemical analysis of human α 1 -AGP has shown that fucose is present on a minor fraction of N-linked glycans (Schmid et al., Biochim. Biophys. Acta. 492:291, 1977; Fournet et al., Biochemistry 17:5206, 1978), but that the asialo protein at least partially blocks the binding of anti-CD15 antibodies (Gooi et al., Eur. J. Immunol. 13:306, 1983; Tetteroo et al., Eur. J. Immunol. 14:1089, 1984). A modest (35±9%) reduction in binding of HL60 cells to ELAM-1 (adsorbed to plastic) was achieved with 200 μg/ml of the protein.

In vitro Production of a Sialyl-Le X Molecule

To extend these results, enzymatically-fucosylated α 1 -AGP was prepared in vitro. The biosynthesis of the sialyl-Le X determinant is controlled by a specific α(1,3)fucosyltransferase (Campbell et al., Cell 35:303, 1983; Campbell et al., J. Biol. Chem. 259:11208, 1984), which adds fucose to the N-acetylglucosamine moiety of terminal N-acetyllactosamine or its 3-sialyl adduct; a genetically and biochemically distinct specific α(1,3)fucosyltransferase is known to only add fucose to the asialyl precursor (Prieels et al., Eur. J. Biochem. 130:347, 1983; Muramatsu et al., Eur. J. Biochem. 157:71, 1986). A third fucosyltransferase is known to form both α(1,3) and a α(1,4) linkages, apparently to unsialylated substrates (Prieels et al., J. Biol. Chem. 256:10456, 1981; Lowe et al., infra). Biosynthesis of the sialyl-Le X determinant proceeds by sequential sialylation followed by fucosylation because α(2,3)sialytransferase cannot recognize the fucosylated terminal N-acetyllactosamine that is CD15 (Holmes et al., J. Biol. Chem. 261:3737, 1986). α 1 -acid glycoprotein is a good substrate for the α(1,3)fucosyltransferase of amniotic fluid (e.g., Hanisch et al., Carbohydr. Res. 178:23, 1988), an enzyme which forms sialyl-Le X from sialylated and nonsialylated precursors, respectively.

Amniotic fluid fucosyltransferase was isolated by affinity chromatography and evaluated for its ability to convert α 1 -AGP into an ELAM-1 ligand as follows. α(1,3)fucosyltransferase was isolated from concentrated amniotic fluid by fetuin-agarose chromatography as previously described in Hanisch et al. ( Carbohydr. Res. 178:23, 1988) and Mitsakos et al. ( Biol. Chem. Hoppe - Seyler 369:661, 1988). 0.8 μCi GDP 14 C-fucose (225 Ci/mole) and 100 μg of bovine α 1 -AGP (Sigma, St. Louis, Mo.) were added to a reaction mix containing 25 mM Tris-HCl pH 7.0, 35 mM MgCl 2 and 1 mM ATP in a final volume of 120 μl. The reaction was allowed to proceed for 24 h. at 37°, at which time approximately 10% of the 14 C-labeled fucose had been incorporated into TCA-insoluble material. Unincorporated label was removed by centrifugal ultrafiltration (Centricon 10, Amicon, Danvers, Mass.). 20 μl of a 1:5 dilution of the labelled material, or 10 μl of a 1:5 dilution of a similarly constituted reaction mixture lacking labelled GDP-fucose, was adsorbed to plastic dishes (as described above) or to 96 well microtiter plates (Falcon 3911, Becton Dickinson, Oxnard, Calif.). Wells were incubated at 220 with ELAM-Rg or CD8-Rg at 1 μg/ml for 1 hr., washed with PBS, and incubated with a radioiodinated goat anti-human IgG antibody (DuPont/NEN, Boston, Mass.) for an additional hr. Following washing, labelled antibody binding was measured in a gamma counter. Results shown in FIG. 9 are expressed as the mean±standard deviation of quadruplicate determinations and are representative of two independent experiments.

FIG. 9 shows that α 1 -AGP incubated with fucosyltransferase in the presence of GDP-fucose bound significantly more ELAM-Rg than did α 1 -AGP alone, or α 1 -AGP incubated with enzyme in the absence of GDP-fucose.

The fucosylated glycans of asialo-α-AGP bear the terminal structure Galβ(1-4)-(Fucα(1-3))GlcNAcβ(1-4)Man, while the nonfucosylated termini of the asialoprotein consist of the N-acetyllactosamine group joined either β(1-4), β(1-2), or α(1-6) to mannose (Fournet et al., Biochemistry 17:5206, 1978). Hence neither any pre-existing sialyl-Le X determinants of α 1 -AGP nor any of the potential fucosyl adducts to N-acetylglucosamine can be joined to galactose. These results, together with the inhibition of ELAM-1 binding by mucin O-linked glycans, indicate that the sialyl-Le X grouping by itself has appreciable affinity for ELAM-1. It remains to be determined, though, whether quantitatively stronger ELAM-1 binding might be promoted by residues neighboring the sialyl-Le X determinants on granulocytes or by steric factors affecting their accessibility.

IL-8 Blocks Myeloid Cell Adhesion to ELAM-1

It has been reported that the release of IL-8 from IL-1-treated endothelial cells causes granulocytes to lose the ability to bind to IL1 induced endothelium (Gimbrone et al., Science 246:1601, 1989). The effect of IL1 and IL-8 on sialyl-Le X surface antigen expression was determined as follows. Granulocytes were incubated with IL1β (10 ng/ml; Pepro Tech, Rock Hill, N.J.) or IL-8 (25 ng/ml; Pepro Tech, Rock Hill, N.J.) for 20 min. at 37°, washed three times, and incubated with a monoclonal antibody to either CD15 (PM-81), CD65 (VIM2) or sialyl-Le X (CSLEX1) (described above) on ice. Results in FIG. 10 are given as the relative mean fluorescence intensity (MFI) determined by flow cytometry, as a percent of the MFI of granulocytes incubated in parallel without cytokines, and are representative of four experiments.

FIG. 10 shows that neither IL1 nor IL-8 caused a substantial reduction in the expression of cell surface sialyl-Le X .

To test the ability of supernatants harvested from granulocytes exposed to IL-8 to block binding to ELAM-1, granulocytes (5×10 7 /ml) were incubated with IL1 or IL-8 (at the concentrations above) for 1 hr. at 37°. Supernatants were collected after centrifugation and incubated with ELAM-Rg-coated dishes. Cells were added after 30 min., and binding was determined as described above. Immunoadsorption was performed with 40 μl of Protein-A agarose beads (Sigma, St. Louis, Mo.) to which 10 μg of affinity purified rabbit anti-mouse IgM were adsorbed, followed by 5 μl of CSLEX1 ascites. Control beads were similarly prepared but were not incubated with CSLEX1. The beads were washed with PBS and incubated with the supernatants for 1 hr. at 4°. Results are shown in FIG. 11 and are expressed as the percent of cells bound, relative to the number bound in the presence of supernatants of granulocytes incubated without cytokine under the same conditions. Data shown are mean±standard deviation of triplicate determinations in three independent experiments.

FIG. 11 shows that supernatants harvested from cultures of granulocytes treated with IL-8, but not IL1, blocked the adhesion of HL60 cells to immobilized ELAM-1, and the binding inhibition could be specifically reversed by adsorption of the supernatants with solid-phase CSLEX1, but not with the immunoadsorption matrix alone.

α 1 -Acid Glycoprotein-Immunoglobulin Fusion Proteins

As discussed herein, antibody fusion proteins according to the invention have important therapeutic and diagnostic uses. Previous work has demonstrated that large amounts of antibody fusion proteins may be generated and secreted transiently from transfected mammalian cells (for example, COS cells). In general, to produce an antibody fusion protein, cDNA encoding a domain of interest is fused in-frame to human IgG domains (for example, constant domains) by standard techniques, and the fusion protein is expressed. The antibody portion of the molecule facilitates fusion protein purification and also prolongs the plasma half-life of otherwise short-lived polypeptides or polypeptide domains. Preferably, antibody fusion proteins are expressed according to the methods disclosed in Seed et al., U.S. Ser. No. 08/483,151 entitled “Fucosyltransferase Genes and Uses Thereof,” filed Jun. 7, 1995 (which is hereby incorporated by reference), e.g., using IgG or IgM antibodies or portions thereof.

Recombinant plasmids expressing particular α 1 -AGP-antibody fusion proteins (e.g., α 1 -AGP-Hinge-CH2-CH3 and α 1 -AGP-CH2-CH3 proteins) were constructed as follows. A cDNA encoding the acute phase α 1 -AGP gene was cloned from a human liver cDNA library by polymerase chain reaction (PCR) using oligonucleotide primers corresponding to the 5 ′ and 3 ′ coding regions of α 1 -AGP (Board et al., Gene 44:127, 1986) according to standard techniques. The 5′ α 1 -AGP primer was designed to contain a HindIII restriction site and the 3′ primer was designed to contain a BamHI restriction site rather than the α 1 -AGP stop codon. The PCR-amplified product was then digested with HindIII/BamHI and cloned into a HindIII/BamHI-cut plasmid expression cassette (see Aruffo et al., Cell 61:1303, 1990) containing constant domains of human IgG1 (i.e., Hinge-CH2-CH3 or CH2-CH3). A nucleotide sequence and amino acid sequence of this AGP-IgG fusion protein is shown in FIGS. 15A and 15B respectively. Resulting expression plasmids were transfected into COS cells (as described in Seed and Aruffo, Proc. Natl. Acad. Sci 84:3365, 1987) for the production of soluble α 1 -AGP antibody fusion proteins or α 1 -AGP.

To prepare an α 1 -AGP-IgG1 fusion protein capable of inhibiting complement fixation and F c receptor binding, we introduced additional sialyl-Le X consensus glycosylation sites (N-X-T/S) into the CH2 domain of human IgG1 as follows. Three sialyl-Le X sites were added to the CH2 domain by creating novel glycosylation consensus sites as described supra; the CH2 domain of human IgG1 naturally contains a single glycosylation site. In particular, consensus N-linked glycosylation sites were introduced at amino acid positions #274, #287, and #322 ( FIG. 1 ; SEQ ID NO:2) by mutating amino acids K to N, A to N, and K to N, respectively, using standard methods of in vitro site-directed oligonucleotide mutagenesis (see, Ausubel et al., Current Protocols In Molecular Biology , Wiley Interscience, 1995). To maximize base pair substitution, a method based on uracil N-glycosidase repair was used to incorporate the degenerate oligonucleotides into the expression vector. The single-stranded DNA produced by this method contained deoxyuracil residues which allowed for efficient selection of mutated strands. Introduced muations were verified by DNA sequence analysis.

To prepare soluble fusion proteins, expression plasmids were transfected into COS cells stably expressing α(1,3/1,4) fucosyltransferase, and the expressed α 1 -AGP-IgG1 fusion proteins were isolated as follows. COS cells were incubated with plasmids carrying a fused α 1 -AGP-IgG1 fusion gene (α 1 -AGP-Hinge-CH2-CH3 or α 1 -AGP-CH2-CH3) in medium containing 10% NuSerum (Collaborative Research), DEAE-dextran, and chloroquine diphosphate. After incubation the culture medium was removed, and the cells were briefly exposed to 10% dimethyl sulfoxide (DMSO) in phosphate-buffered saline (PBS). The cells were then transferred to fresh medium containing 10% NuSerum (Collaborative Research), DEAE-dextran, and chloroquine diphosphate. After 7 to 10 days, the culture supernatant was harvested, and soluble α 1 -AGP-IgG1 fusion protein was purified from the supernatant based on its affinity for protein A agarose beads. The α 1 -AGP-IgG1 fusion protein was recovered from protein A beads by eluting with dilute acetic acid. Eluate fractions containing α 1 -AGP-IgG1 fusion protein were neutralized, and then concentrated by centrifugal ultrafiltration. Purified α 1 -AGP-IgG1 fusion protein was quantified by ELISA using mouse anti-human IgG antisera according to standard methods.

To prepare [ 35 S]-radiolabelled recombinant α 1 -AGP and α 1 -AGP-IgG1 fusion proteins, the above-described procedure was modified as follows. Twelve hours following transfection, a fraction of the COS cells transfected with each construct were seeded into flasks containing fresh culture medium. Thirty-six hours post-transfection, the cells were then washed with PBS and overlayed with cysteine-methionine-free culture medium for 30 minutes. Next, [ 35 S] cysteine and [ 35 S] methionine were added to a final concentration of 150 μCi/ml, and the cells were allowed to incorporate label overnight. After labeling, the culture supernatants were harvested and incubated with protein A beads. The protein A beads were collected by centrifugation and washed. Radiolabelled fusion protein was subsequently eluted from the beads using dilute SDS (sodium dodecyl sulfate). For visual analysis, the eluate was electrophoretically separated on a 6% discontinuous polyacrylamide gel with or without prior exposure to mercaptoethanol.

To assess the effect that the fusion of acute phase protein α 1 -AGP to the 5′ end of the human IgG1 (Hinge-CH2-CH3 or CH2-CH3) has on α 1 -AGP's half-life, we examined the in vivo half-life of recombinant fusion proteins in Balb/c mice. These experiments were carried out as follows. After injecting the mice with radiolabelled [ 35 S] α 1 -AGP-IgG1 fusion proteins or [ 35 S] α 1 -AGP, retroocular blood aliquots were obtained at various time intervals to determine their plasma half-life. Radioactivity of each blood sample was determined in triplicate by standard scintillation spectrometry. As shown in FIG. 12 , α 1 -AGP unfused to IgG1 had a plasma half-life of less than 90 minutes, and was rapidly cleared from the circulation. By contrast, fusion of α 1 -AGP to either the CH2-CH3 domains (resulting in a monomeric protein) or to the Hinge-CH2-CH3 domains (resulting in a dimeric protein) significantly prolonged the plasma half-life of these recombinant molecules to several days. These results demonstrated the enhanced stability of α 1 -AGP when fused to IgG1 constant domains such as CH2-CH3 or Hinge-CH2-CH3.

To determine whether the engineered consensus N-linked glycosylation sites were glycosylated in vivo, we examined the molecular weight of recombinant α 1 -AGP-IgG1 fusion protein produced in COS cells stably expressing α(1,3/1,4) fucosyltransferase. As shown in FIG. 13 , expression of fusion protein in COS cells followed by immunoprecipitation with polyclonal sera against anti-human IgG by standard methods revealed an increase in the molecular weight of the mutated IgG1 (termed #123 in FIG. 13 ) in comparison to an identical fusion protein lacking the three additional sialyl-Le X sites (termed IgG in FIG. 13 ), indicating that the additional glycosylation sites were efficiently utilized. In addition, this hyperglycosylated form of α 1 -AGP-IgG1 retained its affinity for protein A as documented by protein A immunoprecipitation ( FIG. 13 , Protein A panel).

To test for the ability of the α 1 -AGP-IgG1 fusion protein (i.e., α 1 -AGP-Hinge-CH2-CH3) produced in COS cells to inhibit adhesion of human T lymphocytes and myeloid HL60 cells to ELAM-1 (i.e., the adhesion molecule that binds sialyl-Le X ), we analyzed the adhesion of human T lymphocytes and myeloid HL60 cells to immobilized ELAM-1 in the presence of increasing concentrations of fusion protein. These experiments were carried out generally as described above. As shown in FIG. 14 , at approximately 50 μg/ml of fusion protein, there was a >50% inhibition of adhesion for both cell types (compare 0 μg/ml to 50 μg/ml for both HL60 and T lymphocytes shown in FIG. 14 ). By contrast, we observed that purified sialyl-Le X had no effect on adhesion at concentrations up to 200 μg/ml.

Coexpression of the native constant domain of human IgG1 with fucosyltransferase did not result in a molecule capable of binding to ELAM-1.

The introduction of the above-described additional glycosylation sites within the CH2 domain created a molecule that was unrecognized by F c receptor or complement using assays that are standard in the art; exemplary complement fixation assays include Weir et al., Handbook of Experimental Immunology ; and Coligan et al. Current Protocols In Immunology , Wiley Interscience, 1995. This experiment was conducted using an antibody fusion protein hat contained a CD23 domain joined to an antibody portion hat included the sialyl-Le X addition sites described above. Because the constant region of the expressed antibody fusion protein protrudes from the cell surface, it mimics a cell surface-bound antibody and provides a substrate for complement fixation.

In our assay, an antibody fusion protein that included the CH2 domain modified to include sialyl-Le X sites at amino acids #274 (K to N), # 287 (A to N), and #322 (K to N) was compared to an identical fusion protein lacking those sialyl-Le X addition sites. Results indicated that complement fixation was blocked by the presence of the sialyl-Le X moieties on the expressed antibody molecule. The extent of lysis exhibited by cells expressing the sialyl-Le X -masked antibody molecule was less than that exhibited by cells expressing the control antibody molecule and was approximately equivalent to basal levels of cell lysis in the assay. Similar results were obtained in a separate set of analogous complement fixation experiments using antibody-CD4 fusion molecules ( FIGS. 16A and 16B ; SEQ ID NOS: 5 and 6). Deletion of the hinge region of IgG1 also created a molecule that was unrecognized by F c receptor and complement.

Based on the above, any number of recombinant α 1 -AGP-antibody fusion proteins may be designed having long plasma half-lives and the ability to inhibit undesirable cell-cell interactions (for example, the interactions between leukocytes and selectin-bearing cells). To generate molecules with heightened inhibitory potency, candidate molecules are designed and screened using the assays described above. In one particular example, molecules may be screened for their ability to incorporate sialyl-Le X and block the binding of neutrophils to activated endothelial cells; such molecules find use in the inhibition of selectin-dependent inflammatory reactions and tissue injury inflicted by invading leukocytes.

OTHER EMBODIMENTS

Other embodiments are within the claims. For example, the invention encompasses any antibody with decreased ability to fix complement and bind an F c receptor as a result of attached carbohydrate side chains; as described above, these include antibodies useful for their immunologic specificity (e.g., antibodies, e.g., MY904, Todd et al., U.S. Pat. No. 4,840,793) which competitively inhibit cell-cell interactions as well as antibodies which are used only as carriers for therapeutic carbohydrate groups. Because N-linked glycan addition sites are well known to be: Asparagine (N), X, Serine (S) or Threonine (T), where X stands for any residue except proline, one skilled in the art may design a molecule having any number of such sites and thus any number of carbohydrate side chains. Glycosylation sites are incorporated into the antibody sequence, for example, by in vitro site-directed mutagenesis.

Non-complement fixing and non-F c receptor binding antibodies would be used for purposes other than treating inflammation. For example, such an antibody directed against GPII b III a may be used to inhibit platelet aggregation and therefore would be useful for the treatment of myocardial infarction. Antibodies to proteins, such as fibrin or one of the clotting cascade proteins, would be useful for inhibiting thrombotic formation. In general, any antibody (including, without limitation, anti-Mo-1 and anti-CD14) proposed or demonstrated to have a therapeutic use may be improved by the addition of carbohydrate moieties which mask the antibody's ability to fix complement and bind an F c receptor. This characteristic would be particularly important, for example, for immunoglobulin fusion proteins such as IgG fusion proteins (see, for example, Seed et al., U.S. Ser. No. 08/483,151 entitled “Fucosyltransferase Genes and Uses Thereof,” filed Jun. 7, 1995) and IgM fusion proteins (see, for example, Zettlemeisl et al., DNA Cell Biol. 9:347, 1990). In this case, a protein of interest is fused (for example, genetically) to an immunoglobulin molecule to increase the protein's serum half life. Because these fusion proteins have an extended life in the patient, they are more likely to be recognized as foreign antigens, and it is therefore particularly useful for such proteins to evade the patient's F c receptor binding and complement fixation systems.

For the purpose of blocking interactions between cells or proteins, any other appropriate carrier molecule may be utilized. Generally, proteins are preferred because of their relatively long half-lives in serum. One class of carrier proteins are serum proteins such as albumin (e.g., bovine serum albumin or human serum albumin), transferrin, or α-2 macroglobulin. The carrier proteins can contain endogenous glycan addition sites or sites may be introduced into the DNA sequence of the carrier protein (as described above) by, for example, site-directed mutagenesis. The carrier molecule, less preferably, may be a lipid. In one example, the lipid, with one or more attached carbohydrate moieties (e.g., sialyl-Le X determinants), is delivered as a liposome to a target cell wall (e.g., an endothelial cell wall). The liposome may block a cell or protein interaction or may be used to deliver a drug to its appropriate site of action.

Cell adhesion molecules in addition to ELAM-1 may be inhibited by attachment of appropriate carbohydrate recognition moieties to a carrier molecule as described above. Such cell adhesion molecules may include, without limitation, LFA1, LFA-3, ICAM1, PADGEM, Mel14, LAM1, a cadherin, cell-CAM, or an N-CAM. Other glycans may be attached including, without limitation, any N-linked glycan, O-linked glycan, GMP140, Leu8, and phosphatidyl inositol phosphate glycans. In cases where the glycan addition signal is known, it may be introduced into the DNA sequence of an organic carrier molecule as described above. Alternatively, if the precise site is not known, but the general region of such an addition site in some naturally-occurring molecule has been defined, a DNA fragment (from the sequence encoding the naturally-occurring molecule) which includes that region may be cloned into the DNA sequence of the desired carrier molecule (e.g., an IgG1- or α 1 -AGP-encoding sequence).

Production of carrier molecules bearing carbohydrate moieties may be carried out in a cell, preferably, a eukaryotic cell other than yeast. Mammalian cells, e.g., mammalian cell lines, provide particularly suitable hosts. These cells generally synthesize the necessary precursor molecules and produce or can be engineered to produce the enzymes responsible for carbohydrate attachment. For the attachment of sialyl-Le X determinants, mammalian cell lines such as CHO and lec11 are particularly suitable. Alternatively, carbohydrate moieties may be attached to a carrier molecule in vitro, i.e., extracellularly. In one example, α(1,3)fucosyltransferase would be bound to a solid support (e.g., a column) and the desired carrier molecule passed over the bound fucosyltransferase enzyme, under conditions which facilitate attachment of sialyl-Le X groups to their appropriate site or the carrier molecule.

The invention also encompasses the use of sialyl-Lewis X -modified α 1 -AGP-antibody fusion proteins for protecting against, inhibiting, or treating a shock-inducing event, the clinical manifestations of shock, or both which are caused by microbial factors (e.g., lipopolysaccharides (LPS)), microbial toxins (e.g., toxic shock enterotoxins), host mediators (e.g., cytokines), or anti-tumor therapies (e.g., administration of tumor necrosis factor (TNF) or interleukin1 (IL1)), or any combination thereof. For example, an α 1 -AGP antibody fusion protein can be administered to a human patient to alleviate the effects of septic shock induced by microbial LPS. The ability of an α 1 -AGP fusion protein to protect against, treat or inhibit the effects of shock (e.g., septicemia or toxic shock syndrome) is evaluated according to standard methods known in the art (e.g., those described in Libert et al. (1994) J. Exp. Med. 180: 1571-1575).

All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent, and patent application was specifically and individually indicated to be incoporated by reference.

9 2287 base pairs nucleic acid double linear Genomic DNA Coding Sequence 1...300, 361...453, 679...903, 961...972, 1360...1404, 1524...1853, 1950...2270 1AAG CTT ACC ACC ATG GAC TGG ACC TGG AGG TTC CTC TTC TTT GTG GTG 48Lys Leu Thr Thr Met Asp Trp Thr Trp Arg Phe Leu Phe Phe Val Val 5 10 15GCA GCA GCT ACA GGT GTC CAG TCC CAG GTG CAG CTG GTG CAG TCT GGG 96Ala Ala Ala Thr Gly Val Gln Ser Gln Val Gln Leu Val Gln Ser Gly 20 25 30GCT GAG GTG AAG AAG CCT GGG TCC TCG GTG AAG GTC TCC TGC AAG GCT 144Ala Glu Val Lys Lys Pro Gly Ser Ser Val Lys Val Ser Cys Lys Ala 35 40 45TCT GGA GGC ACC TTC AGC AGC TAT GCT ATC AGC TGG GTG CGA CAG GCC 192Ser Gly Gly Thr Phe Ser Ser Tyr Ala Ile Ser Trp Val Arg Gln Ala 50 55 60CCT GGA CAA GGG CTT GAG TGG ATG GGA GGG ATC ATC CCT ATC TTT GGT 240Pro Gly Gln Gly Leu Glu Trp Met Gly Gly Ile Ile Pro Ile Phe Gly65 70 75 80ACA GCA AAC TAC GCA CAG AAG TTC CAG GGC AGA GTC ACG ATT ACC GCG 288Thr Ala Asn Tyr Ala Gln Lys Phe Gln Gly Arg Val Thr Ile Thr Ala 85 90 95GAC GAA TCC ACG 300Asp Glu Ser Thr 100AGCACAGCCT ACATGGAGCT GAGCAGCCTG AGATCTGAGG ACACGGCCGT GTATTACTGT 360GCG AGA GAT AAT GGA GCG TAT TGT AGT GGT GGT AGC TGC TAC TCG GGC 408Ala Arg Asp Asn Gly Ala Tyr Cys Ser Gly Gly Ser Cys Tyr Ser Gly 105 110 115TGG TTC GAC CCC TGG GGC CAG GGA ACC CTG GTC ACC GTC TCT TCA 453Trp Phe Asp Pro Trp Gly Gln Gly Thr Leu Val Thr Val Ser Ser 120 125 130GGTGAGTACT GAATTCTAGC TTTCTGGGGC AGGCCAGGCC TGACCTTGGC TTTGGGGCAG 513GGAGGGGGCT AAGGTGAGGC AGGTGGCGCC AGCAGGTGCA CACCCAATGC CCATGAGCCC 573AGACACTGGA CGCTGAACCT CGCGGACAGT TAAGAACCCA GGGGCCTCTG CGCCTGGGCC 633CAGCTCTGTC CCACACCGCG GTCACATGGC ACCACCTCTC TTGCA 678GCC TCC ACC AAG GGC CCA TCG GTC TTC CCC CTG GCA CCC TCC TCC AAG 726Ala Ser Thr Lys Gly Pro Ser Val Phe Pro Leu Ala Pro Ser Ser Lys 135 140 145AGC ACC TCT GGG GGC ACA GCG GCC CTG GGC TGC CTG GTC AAG GAC TAC 774Ser Thr Ser Gly Gly Thr Ala Ala Leu Gly Cys Leu Val Lys Asp Tyr 150 155 160TTC CCC GAA CCG GTG ACG GTG TCG TGG AAC TCA GGC GCC CTG ACC AGC 822Phe Pro Glu Pro Val Thr Val Ser Trp Asn Ser Gly Ala Leu Thr Ser 165 170 175GGC GTG CAC ACC TTC CCG GCT GTC CTA CAG TCC TCA GGA CTC TAC TCC 870Gly Val His Thr Phe Pro Ala Val Leu Gln Ser Ser Gly Leu Tyr Ser180 185 190 195CTC AGC AGC GTG GTG ACC GTG CCC TCC AGC AGC 903Leu Ser Ser Val Val Thr Val Pro Ser Ser Ser 200 205TTGGGCACCC AGACCTACAT CTGCAACGTG AATCACAAGC CCAGCAACAC CAAGGTG 960GAC AAG AAA GTT 972Asp Lys Lys Val 210GGTGAGAGGC CAGCACAGGG AGGGAGGGTG TCTGCTGGAA GCAGGCTCAG CGCTCCTGCC 1032TGGACGCATC CCGGCTATGC AGCCCCAGTC CAGGGCAGCA AGGCAGGCCC CGTCTGCCTC 1092TTCACCCGGA GCCTCTGCCC GCCCCACTCA TGCTCAGGGA GAGGGTCTTC TGGCTTTTTC 1152CCAGGCTCTG GGCAGGCACA GGCTAGGTGC CCCTAACCCA GGCCCTGCAC ACAAAGGGGC 1212AGGTGCTGGG CTCAGACCTG CCAAGAGCCA TATCCGGGAG GACCCTGCCC CTGACCTAAG 1272CCCACCCCAA AGGCCAAACT CTCCACTCCC TCAGCTCGGA CACCTTCTCT CCTCCCAGAT 1332TCCAGTAACT CCCAATCTTC TCTCTGC 1359AGA GCC CAA ATC TTG TGA CAA AAC TCA CAC ATG CCC ACC GTG CCC 1404Glu Pro Lys Ser Cys Asp Lys Thr His Thr Cys Pro Pro Cys Pro 215 220 225AGGTAAGCCA GCCCAGGCCT CGCCCTCCAG CTCAAGGCGG GACAGGTGCC CTAGAGTAGC 1464CTGCATCCAG GGACAGGCCC CAGCCGGGTG CTGACACGTC CACCTCCATC TCTTCCTCA 1523GCA CCT GAA CTC CTG GGG GGA CCG TCA GTC TTC CTC TAG GGG GGT TTT 1571Ala Pro Glu Leu Leu Gly Gly Pro Ser Val Phe Leu Phe Pro Pro Lys 230 235 240GGG TTC CTG TGG GAG TAC TAG AGG GCC TGG GGA CTC CAG TGT ACG CAC 1619Pro Lys Asp Thr Leu Met Ile Ser Arg Thr Pro Glu Val Thr Cys Val 245 250 255CTG GTG GAC GTG AGC CAC GAA GAC CCT GAG GTC AAG TTC AAC TGG TAC 1667Val Val Asp Val Ser His Glu Asp Pro Glu Val Lys Phe Asn Trp Tyr 260 265 270GTG GAC GGC GTG GAG GTG CAT AAT GCC AAG ACA AAG CCG CGG GAG GAG 1715Val Asp Gly Val Glu Val His Asn Ala Lys Thr Lys Pro Arg Glu Glu 275 280 285CAG TAC AAC AGC ACG TAC CGG GTG GTC AGC GTC CTC ACC GTC CTG CAC 1763Gln Tyr Asn Ser Thr Tyr Arg Val Val Ser Val Leu Thr Val Leu His290 295 300 305CAG GAC TGG CTG AAT GGC AAG GAG TAC AAG TGC AAG GTC TCC AAC AAA 1811Gln Asp Trp Leu Asn Gly Lys Glu Tyr Lys Cys Lys Val Ser Asn Lys 310 315 320GCC CTC CCA GCC CCC ATC GAG AAA ACC ATC TCC AAA GCC AAA 1853Ala Leu Pro Ala Pro Ile Glu Lys Thr Ile Ser Lys Ala Lys 325 330 335GGTGGGACCC GTGGGGTGCG AGGGCCACAT GGACAGAGGC CGGCTCGGCC CACCCTCT 1911GCCCTGAGAG TGACCGCTGT ACCAACCTCT GTCCTACA 1949GGG CAG CCC CGA GAA CCA CAG GTG TAC ACC CTG CCC CCA TCC CGG GAT 1997Gly Gln Pro Arg Glu Pro Gln Val Tyr Thr Leu Pro Pro Ser Arg Asp 340 345 350GAG CTG ACC AAG AAC CAG GTC AGC CTG ACC TGC CTG GTC AAA GGC TTC 2045Glu Leu Thr Lys Asn Gln Val Ser Leu Thr Cys Leu Val Lys Gly Phe 355 360 365TAT CCC AGC GAC ATC GCC GTG GAG TGG GAG AGC AAT GGG CAG CCG GAG 2093Tyr Pro Ser Asp Ile Ala Val Glu Trp Glu Ser Asn Gly Gln Pro Glu 370 375 380AAC AAC TAC AAG ACC ACG CCT CCC GTG CTG GAC TCC GAC GGC TCC TTC 2141Asn Asn Tyr Lys Thr Thr Pro Pro Val Leu Asp Ser Asp Gly Ser Phe 385 390 395TTC CTC TAC AGC AAG CTC ACC GTG GAC AAG AGC AGG TGG CAG CAG GGG 2189Phe Leu Tyr Ser Lys Leu Thr Val Asp Lys Ser Arg Trp Gln Gln Gly400 405 410 415AAC GTC TTC TCA TGC TCC GTG ATG CAT GAG GCT CTG CAC AAC CAC TAC 2237Asn Val Phe Ser Cys Ser Val Met His Glu Ala Leu His Asn His Tyr 420 425 430ACG CAG AAG AGC CTC TCC CTG TCT CCG GGT AAA 2270Thr Gln Lys Ser Leu Ser Leu Ser Pro Gly Lys 435 440TGAGTGCGAC GGCCGGC 2287 442 amino acids amino acid unknown linear protein 2Lys Leu Thr Thr Met Asp Trp Thr Trp Arg Phe Leu Phe Phe Val Val 1 5 10 15Ala Ala Ala Thr Gly Val Gln Ser Gln Val Gln Leu Val Gln Ser Gly 20 25 30Ala Glu Val Lys Lys Pro Gly Ser Ser Val Lys Val Ser Cys Lys Ala 35 40 45Ser Gly Gly Thr Phe Ser Ser Tyr Ala Ile Ser Trp Val Arg Gln Ala 50 55 60Pro Gly Gln Gly Leu Glu Trp Met Gly Gly Ile Ile Pro Ile Phe Gly65 70 75 80Thr Ala Asn Tyr Ala Gln Lys Phe Gln Gly Arg Val Thr Ile Thr Ala 85 90 95Asp Glu Ser Thr Ala Arg Asp Asn Gly Ala Tyr Cys Ser Gly Gly Ser 100 105 110Cys Tyr Ser Gly Trp Phe Asp Pro Trp Gly Gln Gly Thr Leu Val Thr 115 120 125Val Ser Ser Ala Ser Thr Lys Gly Pro Ser Val Phe Pro Leu Ala Pro 130 135 140Ser Ser Lys Ser Thr Ser Gly Gly Thr Ala Ala Leu Gly Cys Leu Val145 150 155 160Lys Asp Tyr Phe Pro Glu Pro Val Thr Val Ser Trp Asn Ser Gly Ala 165 170 175Leu Thr Ser Gly Val His Thr Phe Pro Ala Val Leu Gln Ser Ser Gly 180 185 190Leu Tyr Ser Leu Ser Ser Val Val Thr Val Pro Ser Ser Ser Asp Lys 195 200 205Lys Val Glu Pro Lys Ser Cys Asp Lys Thr His Thr Cys Pro Pro Cys 210 215 220Pro Ala Pro Glu Leu Leu Gly Gly Pro Ser Val Phe Leu Phe Pro Pro225 230 235 240Lys Pro Lys Asp Thr Leu Met Ile Ser Arg Thr Pro Glu Val Thr Cys 245 250 255Val Val Val Asp Val Ser His Glu Asp Pro Glu Val Asn Phe Ser Trp 260 265 270Tyr Val Asp Gly Val Glu Val His Asn Asn Lys Thr Lys Pro Arg Glu 275 280 285Glu Asn Tyr Ser Ser Thr Tyr Arg Val Val Ser Val Leu Thr Val Leu 290 295 300His Gln Asp Trp Leu Asn Gly Lys Glu Tyr Lys Cys Asn Val Ser Asn305 310 315 320Lys Ala Leu Pro Ala Pro Ile Glu Lys Asn Ile Ser Lys Ala Lys Gly 325 330 335Gln Pro Arg Glu Pro Gln Val Tyr Thr Leu Pro Pro Ser Arg Asp Glu 340 345 350Leu Thr Lys Asn Gln Val Ser Leu Thr Cys Leu Val Lys Gly Phe Tyr 355 360 365Pro Ser Asp Ile Ala Val Glu Trp Glu Ser Asn Gly Gln Pro Glu Asn 370 375 380Asn Tyr Lys Thr Thr Pro Pro Val Leu Asp Ser Asp Gly Ser Phe Phe385 390 395 400Leu Tyr Ser Lys Leu Thr Val Asp Lys Ser Arg Trp Gln Gln Gly Asn 405 410 415Val Phe Ser Cys Ser Val Met His Glu Ala Leu His Asn His Tyr Thr 420 425 430Gln Lys Ser Leu Ser Leu Ser Pro Gly Lys 435 440 1894 base pairs nucleic acid double linear cDNA 3ATGGCGCTGT CCTGGGTTCT TACAGTCCTG AGCCTCCTAC CTCTGCTGGA AGCCCAGATC 60CCATTGTGTG CCAACCTAGT ACCGGTGCCC ATCACCAACG CCACCCTGGA CCAGATCACT 120GGCAAGTGGT TTTATATCGC ATCGGCCTTT CGAAACGAGG AGTACAATAA GTCGGTTCAG 180GAGATCCAAG CAACCTTCTT TTACTTCACC CCCAACAAGA CAGAGGACAC GATCTTTCTC 240AGAGAGTACC AGACCCGACA GGACCAGTGC ATCTATAACA CCACCTACCT GAATGTCCAG 300CGGGAAAATG GGACCATCTC CAGATACGTG GGAGGCCAAG AGCATTTCGC TCACTTGCTG 360ATCCTCAGGG ACACCAAGAC CTACATGCTT GCTTTTGACG TGAACGATGA GAAGAACTGG 420GGGCTGTCTG TCTATGCTGA CAAGCCAGAG ACGACCAAGG AGCAACTGGG AGAGTTCTAC 480GAAGCTCTCG ACTGCTTGCG CATTCCCAAG TCAGATGTCG TGTACACCGA TTGGAAAAAG 540GATAAGTGTG AGCCACTGGA GAAGCAGCAC GAGAAGGAGA GGAAACAGGA GGAGGGGGAA 600TCGGATCCCG AGGGTGAGTA CTAAGCTTCA GCGCTCCTGC CTGGACGCAT CCCGGCTATG 660CAGCCCCAGT CCAGGGCAGC AAGGCAGGCC CCGTCTGCCT CTTCACCCGG AGCCTCTGCC 720CGCCCCACTC ATGCTCAGGG AGAGGGTCTT CTGGCTTTTT CCCAGGCTCT GGGCAGGCAC 780AGGCTAGGTG CCCCTAACCC AGGCCCTGCA CACAAAGGGG CAGGTGCTGG GCTCAGACCT 840GCCAAGAGCC ATATCCGGGA GGACCCTGCC CCTGACCTAA GCCCACCCCA AAGGCCAAAC 900TCTCCACTCC CTCAGCTCGG ACACCTTCTC TCCTCCCAGA TTCCAGTAAC TCCCAATCTT 960CTCTCTGCAG AGCCCAAATC TTGTGACAAA ACTCACACAT GCCCACCGTG CCCAGGTAAG 1020CCAGCCCAGG CCTCGCCCTC CAGCTCAAGG CGGGACAGGT GCCCTAGAGT AGCCTGCATC 1080CAGGGACAGG CCCCAGCCGG GTGCTGACAC GTCCACCTCC ATCTCTTCCT CAGCACCTGA 1140ACTCCTGGGG GGACCGTCAG TCTTCCTCTT CCCCCCAAAA CCCAAGGACA CCCTCATGAT 1200CTCCCGGACC CCTGAGGTCA CATGCGTGGT GGTGGACGTG AGCCACGAAG ACCCTGAGGT 1260CAAGTTCAAC TGGTACGTGG ACGGCGTGGA GGTGCATAAT GCCAAGACAA AGCCGCGGGA 1320GGAGCAGTAC AACAGCACGT ACCGGGTGGT CAGCGTCCTC ACCGTCCTGC ACCAGGACTG 1380GCTGAATGGC AAGGAGTACA AGTGCAAGGT CTCCAACAAA GCCCTCCCAG CCCCCATCGA 1440GAAAACCATC TCCAAAGCCA AAGGTGGGAC CCGTGGGGTG CGAGGGCCAC ATGGACAGAG 1500GCCGGCTCGG CCCACCCTCT GCCCTGAGAG TGACCGCTGT ACCAACCTCT GTCCTACAGG 1560GCAGCCCCGA GAACCACAGG TGTACACCCT GCCCCCATCC CGGGATGAGC TGACCAAGAA 1620CCAGGTCAGC CTGACCTGCC TGGTCAAAGG CTTCTATCCC AGCGACATCG CCGTGGAGTG 1680GGAGAGCAAT GGGCAGCCGG AGAACAACTA CAAGACCACG CCTCCCGTGC TGGACTCCGA 1740CGGCTCCTTC TTCCTCTACA GCAAGCTCAC CGTGGACAAG AGCAGGTGGC AGCAGGGGAA 1800CGTCTTCTCA TGCTCCGTGA TGCATGAGGC TCTGCACAAC CACTACACGC AGAAGAGCCT 1860CTCCCTGTCT CCGGGTAAAT GAGTGCGACG GCCG 1894 434 amino acids amino acid unknown linear protein 4Met Ala Leu Ser Trp Val Leu Thr Val Leu Ser Leu Leu Pro Leu Leu 1 5 10 15Glu Ala Gln Ile Pro Leu Cys Ala Asn Leu Val Pro Val Pro Ile Thr 20 25 30Asn Ala Thr Leu Asp Gln Ile Thr Gly Lys Trp Phe Tyr Ile Ala Ser 35 40 45Ala Phe Arg Asn Glu Glu Tyr Asn Lys Ser Val Gln Glu Ile Gln Ala 50 55 60Thr Phe Phe Tyr Phe Thr Pro Asn Lys Thr Glu Asp Thr Ile Phe Leu65 70 75 80Arg Glu Tyr Gln Thr Arg Gln Asp Gln Cys Ile Tyr Asn Thr Thr Tyr 85 90 95Leu Asn Val Gln Arg Glu Asn Gly Thr Ile Ser Arg Tyr Val Gly Gly 100 105 110Gln Glu His Phe Ala His Leu Leu Ile Leu Arg Asp Thr Lys Thr Tyr 115 120 125Met Leu Ala Phe Asp Val Asn Asp Glu Lys Asn Trp Gly Leu Ser Val 130 135 140Tyr Ala Asp Lys Pro Glu Thr Thr Lys Glu Gln Leu Gly Glu Phe Tyr145 150 155 160Glu Ala Leu Asp Cys Leu Arg Ile Pro Lys Ser Asp Trp Tyr Thr Asp 165 170 175Trp Lys Lys Asp Lys Cys Glu Pro Leu Glu Lys Gln His Glu Lys Glu 180 185 190Arg Lys Gln Glu Glu Gly Glu Ser Asp Pro Glu Gly Glu Pro Lys Ser 195 200 205Cys Asp Lys Thr His Thr Cys Pro Pro Cys Pro Ala Pro Glu Leu Leu 210 215 220Gly Gly Pro Ser Val Phe Leu Phe Pro Pro Lys Pro Lys Asp Thr Leu225 230 235 240Met Ile Ser Arg Thr Pro Glu Val Thr Cys Trp Val Asp Val Ser His 245 250 255Glu Asp Pro Glu Val Lys Phe Asn Trp Tyr Val Asp Gly Val Glu Val 260 265 270His Asn Ala Lys Thr Lys Pro Arg Glu Glu Gln Tyr Asn Ser Thr Tyr 275 280 285Arg Trp Ser Val Leu Thr Val Leu His Gln Asp Trp Leu Asn Gly Lys 290 295 300Glu Tyr Lys Cys Lys Val Ser Asn Lys Ala Leu Pro Ala Pro Ile Glu305 310 315 320Lys Thr Ile Ser Lys Ala Lys Gly Gln Pro Arg Glu Pro Gln Val Tyr 325 330 335Thr Leu Pro Pro Ser Arg Asp Glu Leu Thr Lys Asn Gln Val Ser Leu 340 345 350Thr Cys Leu Val Lys Gly Phe Tyr Pro Ser Asp Ile Ala Val Glu Trp 355 360 365Glu Ser Asn Gly Gln Pro Glu Asn Asn Tyr Lys Thr Thr Pro Pro Val 370 375 380Leu Asp Ser Asp Gly Ser Phe Phe Leu Tyr Ser Lys Leu Thr Val Asp385 390 395 400Lys Ser Arg Trp Gln Gln Gly Asn Val Phe Ser Cys Ser Val Met His 405 410 415Glu Ala Leu His Asn His Tyr Thr Gln Lys Ser Leu Ser Leu Ser Pro 420 425 430Gly Lys 2589 base pairs nucleic acid double linear cDNA 5GCCTGTTTGA GAAGCAGCGG GCAAGAAAGA CGCAAGCCCA GAGGCCCTGC CATTTCTGTG 60GGCTCAGGTC CCTACTGGCT CAGGCCCCTG CCTCCCTCGG CAAGGCCACA ATGAACCGGG 120GAGTCCCTTT TAGGCACTTG CTTCTGGTGC TGCAACTGGC GCTCCTCCCA GCAGCCACTC 180AGGGAAACAA AGTGGTGCTG GGCAAAAAAG GGGATACAGT GGAACTGACC TGTACAGCTT 240CCCAGAAGAA GAGCATACAA TTCCACTGGA AAAACTCCAA CCAGATAAAG ATTCTGGGAA 300ATCAGGGCTC CTTCTTAACT AAAGGTCCAT CCAAGCTGAA TGATCGCGCT GACTCAAGAA 360GAAGCCTTTG GGACCAAGGA AACTTCCCCC TGATCATCAA GAATCTTAAG ATAGAAGACT 420CAGATACTTA CATCTGTGAA GTGGAGGACC AGAAGGAGGA GGTGCAATTG CTAGTGTTCG 480GATTGACTGC CAACTCTGAC ACCCACCTGC TTCAGGGGCA GAGCCTGACC CTGACCTTGG 540AGAGCCCCCC TGGTAGTAGC CCCTCAGTGC AATGTAGGAG TCCAAGGGGT AAAAACATAC 600AGGGGGGGAA GACCCTCTCC GTGTCTCAGC TGGAGCTCCA GGATAGTGGC ACCTGGACAT 660GCACTGTCTT GCAGAACCAG AAGAAGGTGG AGTTCAAAAT AGACATCGTG GTGCTAGCTT 720TCCAGAAGGC CTCCAGCATA GTCTATAAGA AAGAGGGGGA ACAGGTGGAG TTCTCCTTCC 780CACTCGCCTT TACAGTTGAA AAGCTGACGG GCAGTGGCGA GCTGTGGTGG CAGGCGGAGA 840GGGCTTCCTC CTCCAAGTCT TGGATCACCT TTGACCTGAA GAACAAGGAA GTGTCTGTAA 900AACGGGTTAC CCAGGACCCT AAGCTCCAGA TGGGCAAGAA GCTCCCGCTC CACCTCACCC 960TGCCCCAGGC CTTGCCTCAG TATGCTGGCT CTGGAAACCT CACCCTGGCC CTTGAAGCGA 1020AAACAGGAAA GTTGCATCAG GAAGTGAACC TGGTGGTGAT GAGAGCCACT CAGCTCCAGA 1080AAAATTTGAC CTGTGAGGTG TGGGGACCCA CCTCCCCTAA GCTGATGCTG AGCTTGAAAC 1140TGGAGAACAA GGAGGCAAAG GTCTCGAAGC GGGAGAAGCC GGTGTGGGTG CTGAACCCTG 1200AGGCGGGGAT GTGGCAGTGT CTGCTGAGTG ACTCGGGACA GGTCCTGCTG GAATCCAACA 1260TCAAGGTTCT GCCCACATGG TCCACCCCGG TGCACGCGGA TCCCGAGGGT GAGTACTAAG 1320CTTCAGCGCT CCTGCCTGGA CGCATCCCGG CTATGCAGCC CCAGTCCAGG GCAGCAAGGC 1380AGGCCCCGTC TGCCTCTTCA CCCGGAGCCT CTGCCCGCCC CACTCATGCT CAGGGAGAGG 1440GTCTTCTGGC TTTTTCCCAG GCTCTGGGCA GGCACAGGCT AGGTGCCCCT AACCCAGGCC 1500CTGCACACAA AGGGGCAGGT GCTGGGCTCA GACCTGCCAA GAGCCATATC CGGGAGGACC 1560CTGCCCCTGA CCTAAGCCCA CCCCAAAGGC CAAACTCTCC ACTCCCTCAG CTCGGACACC 1620TTCTCTCCTC CCAGATTCCA GTAACTCCCA ATCTTCTCTC TGCAGAGCCC AAATCTTGTG 1680ACAAAACTCA CACATGCCCA CCGTGCCCAG GTAAGCCAGC CCAGGCCTCG CCCTCCAGCT 1740CAAGGCGGGA CAGGTGCCCT AGAGTAGCCT GCATCCAGGG ACAGGCCCCA GCCGGGTGCT 1800GACACGTCCA CCTCCATCTC TTCCTCAGCA CCTGAACTCC TGGGGGGACC GTCAGTCTTC 1860CTCTTCCCCC CAAAACCCAA GGACACCCTC ATGATCTCCC GGACCCCTGA GGTCACATGC 1920GTGGTGGTGG ACGTGAGCCA CGAAGACCCT GAGGTCAAGT TCAACTGGTA CGTGGACGGC 1980GTGGAGGTGC ATAATGCCAA GACAAAGCCG CGGGAGGAGC AGTACAACAG CACGTACCGG 2040GTGGTCAGCG TCCTCACCGT CCTGCACCAG GACTGGCTGA ATGGCAAGGA GTACAAGTGC 2100AAGGTCTCCA ACAAAGCCCT CCCAGCCCCC ATCGAGAAAA CCATCTCCAA AGCCAAAGGT 2160GGGACCCGTG GGGTGCGAGG GCCACATGGA CAGAGGCCGG CTCGGCCCAC CCTCTGCCCT 2220GAGAGTGACC GCTGTACCAA CCTCTGTCCT ACAGGGCAGC CCCGAGAACC ACAGGTGTAC 2280ACCCTGCCCC CATCCCGGGA TGAGCTGACC AAGAACCAGG TCAGCCTGAC CTGCCTGGTC 2340AAAGGCTTCT ATCCCAGCGA CATCGCCGTG GAGTGGGAGA GCAATGGGCA GCCGGAGAAC 2400AACTACAAGA CCACGCCTCC CGTGCTGGAC TCCGACGGCT CCTTCTTCCT CTACAGCAAG 2460CTCACCGTGG ACAAGAGCAG GTGGCAGCAG GGGAACGTCT TCTCATGCTC CGTGATGCAT 2520GAGGCTCTGC ACAACCACTA CACGCAGAAG AGCCTCTCCC TGTCTCCGGG TAAATGAGTG 2580CGACGGCCG 2589 630 amino acids amino acid unknown linear protein 6Met Asn Arg Gly Val Pro Phe Arg His Leu Leu Leu Val Leu Gln Leu 1 5 10 15Ala Leu Leu Pro Ala Ala Thr Gln Gly Asn Lys Val Val Leu Gly Lys 20 25 30Lys Gly Asp Thr Val Glu Leu Thr Cys Thr Ala Ser Gln Lys Lys Ser 35 40 45Ile Gln Phe His Trp Lys Asn Ser Asn Gln Ile Lys Ile Leu Gly Asn 50 55 60Gln Gly Ser Phe Leu Thr Lys Gly Pro Ser Lys Leu Asn Asp Arg Ala65 70 75 80Asp Ser Arg Arg Ser Leu Trp Asp Gln Gly Asn Phe Pro Leu Ile Ile 85 90 95Lys Asn Leu Lys Ile Glu Asp Ser Asp Thr Tyr Ile Cys Glu Val Glu 100 105 110Asp Gln Lys Glu Glu Val Gln Leu Leu Val Phe Gly Leu Thr Ala Asn 115 120 125Ser Asp Thr His Leu Leu Gln Gly Gln Ser Leu Thr Leu Thr Leu Glu 130 135 140Ser Pro Pro Gly Ser Ser Pro Ser Val Gln Cys Arg Ser Pro Arg Gly145 150 155 160Lys Asn Ile Gln Gly Gly Lys Thr Leu Ser Val Ser Gln Leu Glu Leu 165 170 175Gln Asp Ser Gly Thr Trp Thr Cys Thr Val Leu Gln Asn Gln Lys Lys 180 185 190Val Glu Phe Lys Ile Asp Ile Val Val Leu Ala Phe Gln Lys Ala Ser 195 200 205Ser Ile Val Tyr Lys Lys Glu Gly Glu Gln Val Glu Phe Ser Phe Pro 210 215 220Leu Ala Phe Thr Val Glu Lys Leu Thr Gly Ser Gly Glu Leu Trp Trp225 230 235 240Gln Ala Glu Arg Ala Ser Ser Ser Lys Ser Trp Ile Thr Phe Asp Leu 245 250 255Lys Asn Lys Glu Val Ser Val Lys Arg Val Thr Gln Asp Pro Lys Leu 260 265 270Gln Met Gly Lys Lys Leu Pro Leu His Leu Thr Leu Pro Gln Ala Leu 275 280 285Pro Gln Tyr Ala Gly Ser Gly Asn Leu Thr Leu Ala Leu Glu Ala Lys 290 295 300Thr Gly Lys Leu His Gln Glu Val Asn Leu Val Val Met Arg Ala Thr305 310 315 320Gln Leu Gln Lys Asn Leu Thr Cys Glu Val Trp Gly Pro Thr Ser Pro 325 330 335Lys Leu Met Leu Ser Leu Lys Leu Glu Asn Lys Glu Ala Lys Val Ser 340 345 350Lys Arg Glu Lys Pro Val Trp Val Leu Asn Pro Glu Ala Gly Met Trp 355 360 365Gln Cys Leu Leu Ser Asp Ser Gly Gln Val Leu Leu Glu Ser Asn Ile 370 375 380Lys Val Leu Pro Thr Trp Ser Thr Pro Val His Ala Asp Pro Glu Gly385 390 395 400Glu Pro Lys Ser Cys Asp Lys Thr His Thr Cys Pro Pro Cys Pro Ala 405 410 415Pro Glu Leu Leu Gly Gly Pro Ser Val Phe Leu Phe Pro Pro Lys Pro 420 425 430Lys Asp Thr Leu Met Ile Ser Arg Thr Pro Glu Val Thr Cys Trp Val 435 440 445Asp Val Ser His Glu Asp Pro Glu Val Lys Phe Asn Trp Tyr Val Asp 450 455 460Gly Val Glu Val His Asn Ala Lys Thr Lys Pro Arg Glu Glu Gln Tyr465 470 475 480Asn Ser Thr Tyr Arg Trp Ser Val Leu Thr Val Leu His Gln Asp Trp 485 490 495Leu Asn Gly Lys Glu Tyr Lys Cys Lys Val Ser Asn Lys Ala Leu Pro 500 505 510Ala Pro Ile Glu Lys Thr Ile Ser Lys Ala Lys Gly Gln Pro Arg Glu 515 520 525Pro Gln Val Tyr Thr Leu Pro Pro Ser Arg Asp Glu Leu Thr Lys Asn 530 535 540Gln Val Ser Leu Thr Cys Leu Val Lys Gly Phe Tyr Pro Ser Asp Ile545 550 555 560Ala Val Glu Trp Glu Ser Asn Gly Gln Pro Glu Asn Asn Tyr Lys Thr 565 570 575Thr Pro Pro Val Leu Asp Ser Asp Gly Ser Phe Phe Leu Tyr Ser Lys 580 585 590Leu Thr Val Asp Lys Ser Arg Trp Gln Gln Gly Asn Val Phe Ser Cys 595 600 605Ser Val Met His Glu Ala Leu His Asn His Tyr Thr Gln Lys Ser Leu 610 615 620Ser Leu Ser Pro Gly Lys625 630 442 amino acids amino acid unknown linear protein 7Lys Leu Thr Thr Met Asp Trp Thr Trp Arg Phe Leu Phe Phe Val Val 1 5 10 15Ala Ala Ala Thr Gly Val Gln Ser Gln Val Gln Leu Val Gln Ser Gly 20 25 30Ala Glu Val Lys Lys Pro Gly Ser Ser Val Lys Val Ser Cys Lys Ala 35 40 45Ser Gly Gly Thr Phe Ser Ser Tyr Ala Ile Ser Trp Val Arg Gln Ala 50 55 60Pro Gly Gln Gly Leu Glu Trp Met Gly Gly Ile Ile Pro Ile Phe Gly65 70 75 80Thr Ala Asn Tyr Ala Gln Lys Phe Gln Gly Arg Val Thr Ile Thr Ala 85 90 95Asp Glu Ser Thr Ala Arg Asp Asn Gly Ala Tyr Cys Ser Gly Gly Ser 100 105 110Cys Tyr Ser Gly Trp Phe Asp Pro Trp Gly Gln Gly Thr Leu Val Thr 115 120 125Val Ser Ser Ala Ser Thr Lys Gly Pro Ser Val Phe Pro Leu Ala Pro 130 135 140Ser Ser Lys Ser Thr Ser Gly Gly Thr Ala Ala Leu Gly Cys Leu Val145 150 155 160Lys Asp Tyr Phe Pro Glu Pro Val Thr Val Ser Trp Asn Ser Gly Ala 165 170 175Leu Thr Ser Gly Val His Thr Phe Pro Ala Val Leu Gln Ser Ser Gly 180 185 190Leu Tyr Ser Leu Ser Ser Val Val Thr Val Pro Ser Ser Ser Asp Lys 195 200 205Lys Val Glu Pro Lys Ser Cys Asp Lys Thr His Thr Cys Pro Pro Cys 210 215 220Pro Ala Pro Glu Leu Leu Gly Gly Pro Ser Val Phe Leu Phe Pro Pro225 230 235 240Lys Pro Lys Asp Thr Leu Met Ile Ser Arg Thr Pro Glu Val Thr Cys 245 250 255Val Val Val Asp Val Ser His Glu Asp Pro Glu Val Lys Phe Asn Trp 260 265 270Tyr Val Asp Gly Val Glu Val His Asn Ala Lys Thr Lys Pro Arg Glu 275 280 285Glu Gln Tyr Asn Ser Thr Tyr Arg Val Val Ser Val Leu Thr Val Leu 290 295 300His Gln Asp Trp Leu Asn Gly Lys Glu Tyr Lys Cys Lys Val Ser Asn305 310 315 320Lys Ala Leu Pro Ala Pro Ile Glu Lys Thr Ile Ser Lys Ala Lys Gly 325 330 335Gln Pro Arg Glu Pro Gln Val Tyr Thr Leu Pro Pro Ser Arg Asp Glu 340 345 350Leu Thr Lys Asn Gln Val Ser Leu Thr Cys Leu Val Lys Gly Phe Tyr 355 360 365Pro Ser Asp Ile Ala Val Glu Trp Glu Ser Asn Gly Gln Pro Glu Asn 370 375 380Asn Tyr Lys Thr Thr Pro Pro Val Leu Asp Ser Asp Gly Ser Phe Phe385 390 395 400Leu Tyr Ser Lys Leu Thr Val Asp Lys Ser Arg Trp Gln Gln Gly Asn 405 410 415Val Phe Ser Cys Ser Val Met His Glu Ala Leu His Asn His Tyr Thr 420 425 430Gln Lys Ser Leu Ser Leu Ser Pro Gly Lys 435 440 45 base pairs nucleic acid single linear cDNA 8CGGAATTCCA GTACTACTCA CCTGGTCCGC CGATGGTCTC CGGGC 45 39 base pairs nucleic acid single linear cDNA 9CCAGATATAC GCGTTGACAT TGATTATTGA CTAGTTATT 39

Claims

1. A method of inhibiting the binding of a cell bearing an ELAM-1 protein to a molecule or cell bearing a sialyl-LeX determinant, comprising contacting said ELAM-1 bearing cell with an α1-acid glycoprotein (AGP)-antibody fusion protein bearing said sialyl-LeX determinant, wherein said inhibition of an ELAM-1-sialyl-LeX-based interaction treats an adverse immune reaction.

2. The method of claim 1, wherein said AGP-antibody fusion protein comprises multiple said sialyl-LeX determinants.

3. The method of claim 1, wherein said antibody portion comprises an IgG1 CH2, CH3, or hinge domain.

4. The method of claim 1, wherein said antibody portion of said AGP-antibody fusion protein bears one or more non-naturally occurring sialyl-LeX determinants.

5. The method of claim 1, wherein the presence of said sialyl-LeX determinant on said AGP-antibody fusion protein interferes with said AGP-antibody fusion protein's ability to fix complement and bind an Fc receptor.

6. The method of claim 5, wherein said sialyl-LeX determinant is attached to amino acid 269 of SEQ ID NO: 2.

7. The method of claim 5, wherein said sialyl-LeX determinant is attached to amino acid 282 of SEQ ID NO: 2.

8. The method of claim 5, wherein said sialyl-LeX determinant is attached to amino acid 317 of SEQ ID NO: 2.

9. The method of claim 5, wherein said AGP-antibody fusion protein is soluble.

10. The method of claim 5, wherein said sialyl-LeX determinant is attached to amino acid 290 of SEQ ID NO: 2.

11. The method of claim 5, wherein said sialyl-LeX determinant is attached to amino acid 330 of SEQ ID NO: 2.