FI104952B - Process for the preparation of a soluble functional derivative of ICAM-1 (intercellular adhesion molecule) - Google Patents

Description

, 104952

A process for the preparation of a soluble functional derivative of ICAM-1 (intercellular adhesion molecule). This patent application is a partial continuation of US patent application Ser. 045,963 (filed May 4, 1987), 07 / 115,798 (filed November 2, 1987), 07 / 155,943 (filed February 16, 1988), 07 / 189,815 (filed May 3, 1988), 07 / 250,446 (filed) September 28, 1988) and 07 / 324,481 (filed March 16, 1989). This invention was made in part with the support of the (country) government. The government has certain rights to this invention.

The present invention relates to a process for the preparation of a soluble functional derivative of ICAM-1 (intercellular adhesion molecule), which differs from ICAM at least in that it does not contain the transmemranium region of natural ICAM-1, which is soluble in physiological conditions.

White cells must be able to attach to cellular substrates to properly defend the host against foreign invaders such as bacteria or viruses. An excellent review of • the defense system is provided by Eisen, H.W. ("Microbiology1 ', 3rd ed., Harper & Row, Philadelphia, PA, 1980, pp. 290-295 and 381-418). They must be capable of attaching to endothelial cells in order to move from the bloodstream to the site of ongoing inflammation. In addition, they must attach to antigen-containing cells in order for a normal specific immune response to occur, and finally they must attach to appropriate target cells for lysis in virus-infected cells or tumor cells.

Recently, white blood cell surface molecules involved in the mediation of such attachments were identified using 352104952 hybridoma technology. Briefly, monoclonal antibodies to human T cells (Davignon, D., et al., Proc.Natl.Acad.Sei. USA 78: 4535-4539 (1981)) and mouse spleen cells (Springer, T., et al., et al ..

5 Eur. J. Immunol. 9, 301-306 (1979), which bound to white cell surfaces and inhibited the attachment-related functions described above (Springer, T., et al., Fed. Proc. 44, 2660-2663 (1985)). The molecules recognized by these antibodies were designated Mac-1 and lymphocyte fusion antigen 1 (LFA-1). Mac-1 is a heterodimer found on cancer cell, granule cell surfaces, and large granule lymphocyte surfaces. LFA-1 is a heterodimer that is found on most lymphocytes (Springer, T.A. et al., Immunol. Rev. 68: 111-135 (1982)). These two molecules, as well as the third molecule, p50.95 (having a tissue distribution similar to that of Mac-1), are involved in cell attachment (Keizer, G., et al., Eur. J. Immunol. 15 (1985) 1142-1147). ).

The white blood cell molecules described above were found to belong to a group of 20 related glycoproteins (Sanchez-Madrid, F.

et al., J.Exoer.Med. 158: 1785-1803 (1983); Keizer, G.D. et al., Eur.J.Immunol. 15 (1985) 1142-1147). This group of Gly co-proteins consists of heterodimers with one alpha chain and one beta chain. Although the alpha chains of these anti-genes differ, the beta chain was found to be highly conserved (Sanchez-Madrid, F., et al., J.Exper.Med. 158 (1983) 1785-1803). The glycoprotein group (sometimes called "CD18") was found to have a Mole beta weight of 95 kd, while alpha chains were found to vary from 150 kd to 180 kd (Springer, T., Fed. Proc. 44, 2660 (1985)). Although the alpha subunits of the membrane proteins do not share a broad homology among the beta units, close analysis of the alpha subunits of the glycoproteins has revealed that there are significant similarities between the two subgroups of LFA-1 alpha-glycoproteins. of the subunits 3 of 104952 are reported by Sanchez-Madrid, F., et al., (J.Exper.Med.

158: 586-602 (1983); J.Exper.Med. 158: 1785-1803 (1983).

A group of individuals have been identified who are unable to express normal amounts of any member of this adhesive protein group on the cell surfaces of their white blood cells (Anderson, D.C., et al., J.Infect.Des. 152: 668-689 (1985)). The lymphocytes of these patients showed similar deficiencies in vitro to normal counterparts with LFA-1 group 10 molecules opposed to antibodies. Further, these individuals were unable to elicit a normal immune response because their cells were unable to attach to cellular substrates (Anderson, DC, et al., Fed. Proc. 44, 2671-2677 (1985); Anderson, DC, et al., J.Infect. Dis. 152 15: 668-689 (1985). These data indicate that immune responses are impaired when the lymphocytes are unable to attach normally due to the lack of LFA-1 functional attachment molecules.

Thus, in summary, the ability of lymphocytes to maintain the health and viability of an animal requires that the lymphocytes be able to attach to other cells (such as endothelial cells). This attachment has been found to require cell-to-cell contacts involving specific receptor molecules present on the cell surfaces of lymphocytes. These receptors allow the lymphocyte to attach to other lymphocytes or endothelial cells and other non-vascular cells. Cell surface receptor molecules have been found to be very similar to each other. Individuals lacking these cell surface receptor molecules in bone of lympho-30 suffer from: chronic and recurrent inflammation, and other clonic symptoms, including deficient antibody responses.

Because lymphocyte attachment is involved in the process of identifying and rejecting export-35 heavy tissue, understanding this process is of considerable importance in the fields of organ transplantation, tissue transplantation, hypersensitivity, and tumor science.

This invention relates to intercellular adhesion molecule-1 (ICAM-1) and functional derivatives thereof.

The method of obtaining ICAM-1 in substantially pure form according to the invention comprises the step of culturing a host organism transformed with a soluble, functional derivative of ICAM-1 encoding a recombinant vector containing at least one altered amino acid group, or its chemical derivative, in the deletion of 10 ICAM-1 molecule, under conditions where said derivative is expressed and secreted into the culture medium, followed by isolation of the derivative.

Figure 1 is a graph showing the attachment between a normal cell 15 and a cell lacking LFA-1.

Figure 2 is a diagram showing the normal / normal cell attachment process.

20

Figure 3 shows the kinetics of cell aggregation in the absence (X) or in the presence (50) of 50 ng / ml PMA.

Figure 4 shows the co-aggregation of LFA-1 'and LFA-1 + cells. ESV-transformed cells (104), labeled with carboxyfluorescein diacetate, as shown in the figure, were mixed with 105 unlabeled autologous cells (solid columns) and JY cells (open columns) in the presence of PMA. After 1.5 hours, labeled cells, in aggregates or free, were counted using the qualitative assay of Example 2. The percentage of labeled cells in the aggregates is indicated. One representative 35 experiment of two performed is presented.

104952 5

Figure 5 shows the immune precipitation of ICAM-1 and LFA-1 from JY cells. Triton X-100 lysates of JY cells (lanes 1 and 2) or control lysis buffer (lanes 3 and 4) were immunoprecipitated with an antibody capable of binding to ICAM-1 (lanes 1 and 3) or with antibodies are capable of binding to LFA-1 (zones 2 and 4). Panel A shows the results under reducing conditions; panel B shows 10 results obtained under non-reducing conditions. Molecular weight standards were run in zone S.

Figure 6 shows the kinetics of IL-1 and interferon-gamma effects on ICAM-1 expression in human skin connective tissue formation cells. Human skin connective tissue-forming cells were grown to a density of 8 x 10 4 cells / 0.32 cm 2 / well. IL-1 (10 U / ml, solid circles) or recombinant gamma inter-20 Ferron (10 U / ml, open squares) were added, followed by cooling of the assay mixture at 4 ° C at determination of commitment. The standard deviation did not exceed 10%.

25

Figure 7 shows the concentration dependence of the effects of IL-1 and interferon gamma on ICAM-1. Human skin connective tissue formation cells were grown to a density of 8 x 10 4 cells / 0.32 cm 2 / well. IL-2 (open 30 circles), recombinant human IL-1 (open · · · squares), recombinant mouse IL-1 (solid squares), recombinant human gamma interferon ( solid circles) and recombinant interferon beta (open triangles) were added as a marked 35 dilution and the mixtures were incubated for 4 hours (IL-1) or 16 hours (interferon beta and gamma). The results shown are the average of 104952 of 6 times the assay; the standard deviation did not exceed 10%.

Figure 8 shows the nucleotide and amino acid 5 sequences of ICAM-1 cDNA. The first ATG is in position 58.

The translated sequences corresponding to the ICAM-1 trypsin peptides are underlined. The hydrophobic putative signal peptide and transmembrane sequences, respectively, are strongly underlined. The N-linked glycosylation sites are indicated in the box. The polyadenylation signal AATAAA at position 2976 is indicated by the above line. The sequence of the HL-60 cDNA clone is shown. The endothelial cell cDNA was sequenced over almost its entire length and showed only minor anomalies.

Figure 9 shows regions of ICAM-1 homology and relationship to immunoglobulin supergene group. (A) 5 homologous-20 region parallel alignment (DI-5). Two or more residuals stacked in a row are marked with a box. Residues that had remained two or more times in the NCAM regions, as well as residues that had remained in the C2 and Cl series, were placed alongside the internal repeats of ICAM-1. The position of the predicted beta strands in the ICAM-1 region is indicated by the bar lines and lower case letters over the alignment 30, and the known position of the beta strands in the immunoglobulin C regions is indicated by the bar lines and upper case letters below the alignment. The position j of the putative disulfide bridge in the ICAM-1 region is designated S-S. (B-D) Alignment of protein 13 regions 1 homologous to ICAM-1 regions; proteins were provisionally aligned by performing 7,104,952 searches on NBRF databases using the FASTP program. Protein sequences include MAG, NCAM, T cell receptor alpha subunit V region, IgM my chain and alpha 1-B glycoprotein.

5

Figure 10 is a graphical comparison of the secondary structures of ICAM-1 and MAG.

Figure 11 shows LFA-1-positive EBV-transformed-10 and B-early stem cell cells that bind to ICAM-1 in platelets.

Figure 12 shows LFA-1 positive T-early cells and T-lymphoma cells binding to ICAM-1 in mu-15 bound vesicles.

Figure 13 shows the inhibition of JY-B early lymphocyte binding to ICAM-1 in plastic-bound vesicles by pretreatment of cells or vesicles with monoclonal antibodies.

Figure 14 shows the effect of temperature on the binding of T-early lymphocytes to ICAM-1 in plastic-bound vesicles.

'' 25 m ·

Figure 15 illustrates the need for divalent cations to bind T early cells to ICAM-1 in plastic-bound vesicles.

Figure 16 shows the effect of anti-attachment antibodies on the ability of peripheral blood mononuclear cells to proliferate in response to recognition of the T-cell-associated antigen OKT3. T'OKT3 "denotes antigen addition.

Figure 17 shows the effect of anti-attachment antibodies on the ability of peripheral blood mononuclear cells to proliferate in response to the recognition of the non-specific T-cell mitogen, concavallin A. "CONA" means an increase in concanavallin A.

Figure 18 shows the effect of anchorage inhibiting antibodies on the ability of peripheral blood mononuclear cells to proliferate in response to recognition of a keyhole ventral hemocyanin antigen. The insertion of keyhole vial hothemocyanin into cells is labeled "KLH".

Figure 19 shows the effect of anti-attachment antibodies on the ability of peripheral blood mononuclear cells to proliferate in response to tetanus toxoid antigen-15 gene recognition. Addition of tetanus toxoid antigen to cells is designated "AGN".

Figure 20 shows the binding of monoclonal antibodies RR1 / 1, R6.5, LB2, and CL203 to the ICAM-1-20 deletion mutants.

Figure 21 shows the binding of ICAM-1 deletion mutants to LFA-1.

Figure 22 shows anti-ICAM-1 monoclonal antibodies. .

epitopes recognized by substances RR1 / 1, R6.5, LB2 and CL203.

Figure 23 shows the binding ability of ICAM-1 region 2 mutants to LFA-1.

«

Figure 24 shows the binding capacity of ICAM-1 region 3 mutants to LFA.

Figure 25 shows the binding capacity of ICAM-1 region 1 mutants to LFA-1.

104952 9

Figure 26 shows the alignment of the amino terminus regions of the ICAM.

One aspect of the present invention pertains to the discovery of the native LFA-1 binding ligand. Molecules such as those of the LFA-1 'moiety that are involved in the cell attachment process are called "attachment molecules". The native binding ligand of this invention is called "intercellular attachment molecule-1" or "ICAM-1". ICAM-1 is a 76-97 kd glycoprotein. ICAM-1 is not a heterodimer. The present invention relates to ICAM-1 and its "functional derivatives". A "functional derivative" of ICAM-1 is a compound having biological activity (either functional or structural) substantially similar to the biological activity of ICAM-1. The term "functional derivatives" is intended to include "fragments", "variants", "analogs" or "chemical derivatives" of a molecule. By "frament" of a molecule-20, such as ICAM-1, is meant any polypeptide subgroup of molecules. Particularly preferred are fragments of ICAM-1 having ICAM-1 activity and soluble (i.e., not membrane bound). Particularly preferred are fragments of ICAM-1 that have ICAM-1 activity and are soluble (i.e., not membrane-bound). By "variant" of a molecule such as ICAM-1 is meant a molecule that is substantially similar in structure and function to either the entire molecule or fragment thereof. A molecule is said to be "substantially similar" to another molecule if the structures of the two molecules are substantially similar or if both molecules have the same biological activity. Thus, provided that the two molecules have similar activity, both are considered variants of each other in the sense used herein, even if the molecular structure of one does not exist in the other, or if the amino acid residue sequences are not identical. By "analog" of a molecule such as ICAM-1 is meant a molecule that has substantially the same function as either the whole molecule or a fragment thereof. As used herein, a molecule is referred to as a "chemical derivative" of another molecule if it contains additional chemical groups that are not normally part of the molecule. Such groups may improve the solubility, absorption, biological half-life of the molecule, etc. Alternatively, the groups may reduce the toxicity of the molecule, eliminate or attenuate any undesirable side effect of the molecule, etc. The group capable of producing such effects is disclosed in Remington's Pharmaceutical Sciences, 1980. molecules form a special class of "chemical derivatives". A "toxin derivative" molecule is a molecule (such as ICAM-1 or antibody) that contains a toxin group. Binding of such a molecule to the cell brings the toxin moiety close to the cell and thus promotes cell death. Any suitable toxin group may be used; However, toxins such as ricin toxin, diphtheria toxin, radioisotope toxins, membrane pathway toxins, etc. are preferably used. Procedures for coupling such groups to a molecule are well known in the art.

* '25 • ♦

Antigenic molecules such as ICAM-1 or LFA-1 moieties are naturally expressed on lymphoid cells. Thus, if such cells are introduced into a suitable animal, for example, by intraperitoneal injection, etc., the result is the generation of antibodies capable of binding to ICAM-1 or LFA-1 group molecules. If desired, serum from such an animal may be taken and used as a source for obtaining polyclonal antibodies capable of binding to these molecules. Preferably, however, such animals are spleen cells which are fused to a myeloma cell line, after which these fusion cells are allowed to form a hybridoma cell that secretes 10495211 monoclonal antibodies capable of binding to ICAM-1 or to an LFA-1 group molecule.

The hybridoma cells obtained as described above can be screened by various methods to identify the desired hybridoma cells that secrete an antibody capable of binding either to ICAM-1 or to one of the LFA-1 group molecules. In a preferred screening assay, such molecules are identified for their ability to inhibit the aggregation of cells transformed with Epstein-10 Barr virus. Antibodies capable of inhibiting such aggregation are then further screened to determine whether they inhibit said aggregation by binding to ICAM-1 or an LFA-1 group molecule. In such screening, any means of separating ICAM-1 from LFA-1 group molecules can be used. Thus, for example, antibody-bound antigen can be analyzed, for example, by immunoprecipitation and polyacrylamide gel electrophoresis. If the bound antigen belongs to the LFA-1 group molecules, the im-20 co-precipitated antigen is detected as a dimer, whereas if the bound antigen is ICAM-1, only one molecular weight species is immunoprecipitated. Alternatively, it is possible to distinguish between antibodies that bind to LFA-1 moieties and those that bind to ICAM-1 by screening for the ability of the antibody to bind to cells such as granular cells expressing LFA-1: n but not ICAM-1. The ability of the antibody (known to inhibit cellular aggregation) to bind to granule cells implies that the antibody is capable of binding to LFA-1. The absence of such binding represents an antibody that recognizes ICAM-1. The ability of an antibody to bind to a cell, such as granulosa, can be ascertained by conventional means well known to those of ordinary skill in the art. Such means include immunoassays, cell agglutination, filter-binding assays, antibody precipitation, etc.

Alternatively, the anti-aggregation antibodies of this invention may be identified by measuring their ability to differentially bind to cells expressing ICAM-1 (such as activated endothelial cells) and their inability to bind to cells that do not express ICAM-1. Even one of ordinary skill in the art will readily appreciate that the above assays may be modified or carried out in a different order, resulting in a number of different possible screening assays, each capable of identifying or distinguishing between ICAM-1 and LFA-1 clusters. antibodies.

The anti-inflammatory agents of the present invention may be obtained by natural processes (such as inducing an animal, plant, fungus, bacterium, etc., to produce a non-immunoglobulin antagonist of ICAM-1, or inducing an animal to produce polyclonal antibodies capable of binding to ICAM). -l ratio); by synthetic methods (such as using the Merrifield method for the synthesis of polypeptides to further synthesize ICAM-1, functional derivatives of ICAM-1 or protein antagonists of ICAM-1 (either immunoglobulin or non-immunoglobulin); hybridoma technology; (such as to produce monoclonal antibodies capable of binding to ICAM-1); or recombinant DNA technology (such as to produce anti-inflammatory agents of the present invention in various hosts (i.e., yeast, bacteria, fungi, mammalian cell) -30 in cultures, etc.) or recombinant plasmid or viral vectors · The choice of method used will depend on factors such as convenience, desired yield, etc. It is not necessary to use only one of the methods, processes or techniques described above to produce the specific anti-inflammatory agent; processes described, methods and techniques may be combined to provide a particular anti-inflammatory agent.

104952 13 A. Identification of LFA-1 Binding Pair (ICAM-1) I. Assays for LFA-1-Dependent Aggregation Multiple cells transformed with Epstein-Barr virus show aggregation. This aggregation can be enhanced by the presence of phorbol esters. Such homotypic aggregation (i.e., aggregation involving only one cell type) was found to be blocked by anti-LFA-1 antibodies (Rothlein, R., et al., J.Exper.Med. 163: 1132-1149 (1986)). , which reference is incorporated herein by reference. Thus, the extent of LFA-1-dependent binding can be determined by determining the extent of spontaneous or phorbol ester-dependent aggregate formation.

An agent that interferes with LFA-1-dependent aggregation can be identified using an assay that can determine whether it interferes with spontaneous or phorbol ester-dependent aggregation of cells transformed with Epstein-Barr virus. Most Epstein-Barr virus-transformed cells can be used in such an assay as long as the cells are capable of expressing the LFA-1 receptor molecule. Such cells can be prepared by Springer, T.A., et al., J. Exper.Med. 160: 1901-1918 (1984), the disclosure of which is incorporated herein by reference. While any such cell may be used in the LFA-1-dependent binding assay of the present invention, JY cell line cells are preferably used (Terhost, C.T., et al., Proc. Natl.Acad.Sci. USA 73, 910 (1976)). Cells may be incubated in any suitable medium; However, most preferably, cells are cultured in RPMI 1640 medium supplemented with 10% Vascular cover serum and 50 micrograms / ml gentamycin (Gibco Laboratories, NY). Cells should be cultured under conditions suitable for 35 mammalian cell proliferation (i.e., generally at 37 ° C, 5% CO 2, 95% RH, etc.).

14 104952 2. LFA-1 binds to ICAM-1

Individuals whose lymphocytes lack the LFA-1 receptor molecule group have been identified (Anderson, D.C., et al., 5 Fed. Proc. 44 (1985) 2671-2677 / Anderson, D.C., et al.).

J.Infect.Dis. 152: 668-689 (1985). Such individuals are said to suffer from white cell attachment deficiency (LAD). EBV-transformed cells of such individuals are unable to aggregate, neither spontaneously nor in the presence of phorbol esters in the aggregation assay described above. When such cells were mixed with cells expressing LFA-1, aggregation was observed (Rothlein, R., et al., J.Exper.Med.

163: 1132-1149 (1986) (Figure 1). Significantly, these aggregates were not formed if these cells were incubated in the presence of anti-LFA-1 antibodies. Thus, although aggregation required LFA-1, the ability of LFA-1-deficient cells to form aggregates with LFA-1-containing cells indicated that the LFA-1 binding pair is not LFA-1, but rather a previously unknown cell adhesion. 20 chemolecule. Figure 1 shows the mechanism of cell attachment.

B. Intercellular Adhesive Molecule-1 (ICAM-11) A novel intercellular adhesion molecule, ICAM-1, was first identified and partially characterized by Rothlein, R., et al., J. Immunol. 137, 1270-1274 (1986), which is incorporated herein by reference. To detect the ICAM-1 molecule, monoclonal antibodies were prepared from 30 spleen cells of mice after immunization with cells from individuals with genetically deficient LFA-1 expression, and the resulting antibodies were screened for their ability to inhibit LFA. Aggregation of cells expressing -1 (Figure 2) .In detail, mice were immunized with EBV-transformed B cells from LAD patients not expressing LFA-1 to detect ICAM-351. - 104952 spleen cells were then fused to myeloma cells and allowed to form monoclonal antibody-producing hybridoma cells. Formulated B cells derived from normal individuals expressing LFA-1 were then incubated in the presence of a monoclonal antibody produced by a hybridoma cell to identify a potential monoclonal antibody capable of inhibiting phorbol ester of EBV-transformed B cells. mediated, LFA-1-dependent spontaneous aggregation.

Since hybridoma cells were obtained from cells that had never "familiarized" with LFA-1 antigen, no monoclonal antibody against LFA-1 was formed. Thus, any antibody found to inhibit aggregation will necessarily be capable of binding to an antigen which, although not LFA-1, is involved in the LFA-1 attachment process. Although any method can be used to obtain such monoclonal antibodies, preferably, ICAM-1 binding monoclonal antibodies are obtained by immunizing BALB / c mice using the pathways and time frames described by Rothlein, R., et al. (J. Immunol. 137: 1270-1274 (1986)) with peripheral blood mononuclear cells transformed with Epstein-Barr virus from individuals lacking LFA-1. Such cells are disclosed by Springer, T.A., et al., (J.Exper.Med. 160: 1901, 25- 1918 (1984)).

• φ

In a preferred method of generating and detecting antibodies capable of binding to ICAM-1, mice are immunized with either EBV-transformed B cells expressing both ICAM-1 and LFA-1, or more preferably TNF- activated endothelial cells expressing ICAM-1 but not LFA-1. In the most preferred method of generating hybridoma cells that produce anti-ICAM-1 antibodies, BALB / c mice were immunized sequentially with JY cells and differentiated U937 cells (ATCC CRL-1593). From such animals, spleen cells are removed and fused to myeloma cells and allowed to produce antibody-producing hybridoma cells. The antibodies are screened for their ability to inhibit LFA-1-dependent, phorbol ester-5-induced aggregation of the EBV-transformed cell line, such as JY cells expressing both the LFA-1 receptor and ICAM-1, as reported by Rothlein, R., et al. (J. Immunol. 137: 1270-1277 (1987)), antibodies capable of inhibiting such aggregation are then tested for their ability to inhibit phorbol ester-induced aggregation of a 10 cell line, such as SKW3 (Dustin, M. ., et al., J.Exoer.Med. 165 (1987) 672-692), whose ability to spontaneously aggregate in the presence of phorbol ester is inhibited by an antibody capable of binding to LFA-1 but not inhibiting anti-ICAM-1 antibodies. -15 agents. Antibodies capable of inhibiting phorbol ester-induced aggregation of cells such as JY cells but unable to inhibit phorbol ester-induced aggregation of cells such as SKW3 are likely anti-ICAM-1 antibodies. Alternatively, antibodies capable of binding to ICAM-1 may be identified by screening for antibodies that are capable of inhibiting LFA-1-dependent aggregation in LFA-expressing cells (such as JY cells), but are unable to bind to cells, expressing LFA-1 but only *, little or no ICAM-1 (such as normal granule cells) or being able to bind to cells expressing ICAM-1 but not LFA-1 (such as TNF -activated endothelial cells). Another alternative is immunoprecipitation from cells expressing ICAM-1, LFA-1 or both using antibodies that inhibit cells such as JY cells. LFA-1-dependent aggregation, whereby some molecular properties of the antibody precipitated by the antibody are determined by SDS-PAGE or the like. If the property is the same as that of ICAM-1, the antibody can be assumed to be 35 against ICAM-1.

104952 17 Using the monoclonal antibodies prepared as described above, the ICAM-1 cell surface molecule was purified and characterized. ICAM-1 was purified from human cells and tissue using monoclonal antibody-affinity chromatography. In such a method, the monoclonal antibody reacting with ICAM-1 is coupled to an inert column matrix. Any method may be used to obtain such a coupling; however, it is preferred to use Oettgen, H.C., et al., J.Biol.Chem. 259: 10 12034 (1984). When the cell lysate is applied through the matrix, the ICAM-1 molecules present are adsorbed and remain in the matrix. By altering the pH or ion concentration of the column, the bound ICAM-1 molecules can be eluted from the column. Although any suitable matrix may be used, preferably Sepharose gel (Pharmacia) is used as the matrix material. The formation of column matrices and their use for protein purification are well known in the art.

In a manner well known to those of ordinary skill in the art, the assays described above can be used to identify compounds that are capable of attenuating or inhibiting the rate or extent of cell attachment.

ICAM-1 is a cell surface glycoprotein expressed on non-haematopoietic cells such as vascular endothelial cells, thymus epithelial cells, certain other epithelial cells and connective tissue forming cells, and hematopoietic cell lymphocytes, B cells and neurons; importer branch cells in the tonsils, lymph nodes, and Peyer's lymph nodes. ICAM-1 is strongly expressed on vascular endothelial cells in T cell regions in lymph nodes and in the tonsils, which show reactive hyperplasia. ICAM-1 35 is expressed at low levels in peripheral blood lymphocytes. Forbol ester-stimulated differentiation of some myelomonocyte cell lines strongly increases ICAM-1 expression. Thus, ie 104952 ICAM-1 is primarily expressed at sites of inflammation and is not generally expressed in "silent" cells. ICAM-1 expression on dermal connective tissue-forming cells increases from 3-fold to 5-fold with both interleukin-I and interferon-gamma 5 at levels of 10 U / ml over four and 10 hours, respectively. Induction depends on protein and mRNA synthesis and is retroactive.

ICAM-1 exhibits molecular weight heterogeneity in different cell types with a molecular weight of 97 kd on connective tissue-forming cells, 114 kd on myelomonocyte cell line U937, and 90 kd on B-early stem cell JY. An approximately 73 kd intracellular precursor has been found to be involved in the biosynthesis of ICAM-1. The non-N-glycosylated form obtained by treatment with tunicamycin (which inhibits glycosylation) has a molecular weight of 55 kd.

The major product 20 of ICAM-1 isolated from phorbol ester-stimulated U937 cells or connective tissue formation cells is the same and has a molecular weight of 60 kd after chemical de-glycosylation. ICAM-1 monoclonal antibodies interfere with the attachment of parent phytohemagglutinin cells to cell lines lacking LFA-1. By pretreating the fibrous tissue forming cells, but not the lymphocytes, with the monoclonal *. Antibodies capable of binding to ICAM-1 inhibit lymphocytoplasmic cell adhesion. By pretreating lymphocytes, but not connective tissue formation cells, anti-LFA-1 antibodies have also been found to provide inhibition of lymphocyte tissue formation cell attachment.

ICAM-1 is thus a binding ligand of the CD18 complex in leukocytes. It is inducible by connective tissue-forming cells and endothelial cells in vitro by inflammatory mediators such as IL-1, interferon gamma, and tumor necrosis factor, over a time period suitable for infiltration of lymphocytes into inflammatory sites <D 104952: 245-254 (1986); Prober, J.S., et al., J. Immunol. 137 (1986) 1893-1896). Further, ICAM-1 is expressed on non-haematopoietic cells such as vascular endothelial cells, thymus epithelial cells, other epithelial cells 5 and connective tissue forming cells, and on gonadal cells, lymph nodes and Peyer's lymph nodes (Dustin, ML, et al., 10 J. Immunol. 137, 245-254 (1986)). ICAM-1 is expressed on keratinocytes in benign inflammatory lesions such as hypersensitivity dermatitis, lichen skin, rash associated with inflammatory disease, urticaria and vesicular disease. Allergic skin reactions caused by applying hapten to the skin of the patient to whom he is hypersensitive also exhibited strong ICAM-1 expression on keratinocytes. On the other hand, toxic skin patches on the skin did not induce ICAM-1 expression on keratinocytes. ICAM-1 is present in keratinocytes derived from tissue samples derived from skin lesions associated with 20 different skin disorders and iCAM expression is induced in hypersensitivity patch lesions, whereas keratinocytes in toxin patch lesions did not: ICAM-1.

ICAM-1 is thus a cellular substrate to which lymphocytes can attach, allowing lymphocytes to migrate to sites of inflammation and / or perform various effector functions that contribute to this inflammation. Such functions include antibody production, lysis of virus-infected target cells, etc. As used herein, "inflammation" refers to reactions of specific or non-specific defense systems. As used herein, the term "specific defense system" refers to that part of the immune system that responds to the presence of specific antigens. Tu-35 is said to be the result of a specific defense system response to, if caused by, or mediated by, or associated with a specific defense system 104952 reaction. Examples of inflammation resulting from a specific defense system response include response to antigens such as rubella virus, autoimmune diseases, delayed type T cell mediated hypersensitivity response 5 (such as in individuals with a "positive" Mantaux test), psoriasis, etc.

"Non-specific defense system response" is a reaction mediated by leukocytes incapable of immunological memory. Granulocytes and macrophages belong to such cells. As used herein, inflammation is said to be due to a reaction of a non-specific defense system if the inflammation is due to, mediated by, or associated with a reaction of a non-specific defense system. Example-15 causes of inflammation, at least in part, due to a reaction by a non-specific defense system, are inflammations associated with conditions such as: asthma; Adult respiratory distress syndrome (ARDS), or multiple organ damage syndrome following septicemia or trauma; injury caused by repeated perfusion of the myocardium or other tissues; acute glomerulonephritis, reactive arthritis; dermatoses with acute inflammatory components; acute purulent meningitis or other inflammatory diseases of the central nervous system; thermal injury; hemodialysis; leukafe-25; ulcerative colitis; Crohn's disease; necrotizing etherocolitis; granulocyte transfusion-related syndromes; and cytokine-induced toxicity.

According to the present invention, a non-specific defense system can be used to treat such reactions; or for use in therapy, functional derivatives of ICAM-1, and in particular such derivatives containing ICAM-1 S fragments or mutated variants having regions 1, 2 j and 3. Further preferred for such treatment or therapy are ICAM-1. 1 fragments or mutant variants 1 containing ICAM-1 region 2. Most preferred in such treatment or therapy are ICAM-1 104952 21 fragments or mutant variants containing ICAM-1 region 1.

C. Cloning of ICAM-1 gene Any of a number of procedures can be used to clone the ICAM-1 gene. In one such method, a vector vector library consisting of cDNA inserts (obtained from an ICAM-1 expressing cell) is analyzed for the presence of 10 inserts containing the ICAM-1 gene.

Such an assay can be performed by transfecting cells with a vector and then determining ICAM-1 expression. According to a preferred method, the amino acid sequence of the ICAM-1 molecule is determined to clone this gene. To accomplish this task, the ICAM-1 protein can be purified and analyzed by an automatic sequencer. Alternatively, the molecule may be fragmented with cyanogen bromide, or proteases such as papain, chymotrypsin or trypsin (Oike, Y., et al., J. Biol. Chem. 257: 9751-209758 (1982); Liu, C., et al., Int.J.Pept.Protein Res. 21 (1983) 209-215). While it is possible to determine the entire amino acid sequence of ICAM-1, it is preferred to determine the sequence of peptide fragments of the molecule. If the peptides are more than 10 amino acids in length, the sequence information is generally sufficient to clone a gene such as the ICAM-1 gene.

The sequence of the amino acid residues in the peptide is herein designated using either the commonly used 3-letter 30 or their 1-letter designation. A list of these ·· 3- and 1-letter designations can be found in manuals such as Biochemistry, Lehninger, A., Worth Publishers, New York, NY, 1970. When such a sequence is vertically labeled, the amino terminus residue is at the top of the list and the carboxyl terminus the lowest residue in the list. Similarly, when the label is horizontal, the amino terminus is to the left and the 104952 22 carboxyl terminus to the right. The amino acid residues contained in the peptide can be separated by dashes. The purpose of these hyphens is solely to clarify the appearance of the sequence designation. By way of illustration only, the amino acid sequence designated: -Gly-Ala-Ser-Phe- means that the Ala residue is linked to the carboxyl-10 group of Gly, and that the Ser residue is linked to the carboxyl group of the Ala residue and Phe residue. The label further indicates that the amino acid sequence contains the tetrapeptide Gly-ALa-Ser-Phe. The purpose of the designation is not to restrict the amino acid sequence to this one tetrapeptide, but is intended to comprise (1) a tetrapeptide wherein one or more amino acid residues are linked to either an amino or carboxyl terminus, (2) a tetrapeptide wherein one or more amino acid residues are linked carboxyl end, (3) a tetrapeptide which does not contain additional 20 amino acid residues.

Once one or more suitable peptide fragments have been sequenced, DNA sequences capable of encoding them are contemplated. The degenerate nature of the genetic code *. Because of this, more than one codon may be used to encode a particular amino acid (Watson, Molecular Biology of the Gene, 3rd edition, W.A. Benjamin, Inc., Menlo Park, CA, 1977, pp. 356-357). Peptide fragments are analyzed to identify amino acid sequences capable of. 30 encoding oligonucleotides having degenerate. degree is the lowest. This is preferably accomplished by identifying sequences containing amino acids encoded by a single codon. Although sometimes such an amino acid sequence can be encoded by only one oligonucleotide, in most cases the amino acid sequence is capable of encoding. more than one set of similar oligonucleotides. Importantly, when all members of the series contain 104952 23 oligonucleotides capable of encoding a peptide fragment, and thus possibly contain the same nucleotide sequence as the gene encoding the peptide fragment, only one member of the series contains a nucleotide sequence identical to that gene. with respect to the nucleotide sequence. Because this member is present in series and is capable of hybridizing to DNA in the presence of up to 100 other members, it is possible to use a non-fractionated oligonucleotide sequence in the same manner as would use a single oligonucleotide to clone a gene encoding a peptide.

In a completely analogous manner to the above, an oligonucleotide (or sequence of oligonucleotides) whose nucleotide sequence is complementary to an oligonucleotide sequence, or a sequence of sequences capable of encoding a peptide fragment may be used.

A suitable oligonucleotide, or suitable oligonucleotides capable of encoding the ICAM-1 gene fragment (or capable of complementing such an oligonucleotide, or oligonucleotide sequence), is identified (using the procedure described above), synthesized and synthesized in the art. DNA, or more preferably, a cDNA preparation derived from human cells capable of expressing the ICAM-1 gene sequences. Nucleic acid hybridization techniques are disclosed by Maniatis, T., et al., Molecular Cloning, A Laboratory Manual, Coldspring Harbor, NY, 1982, and Haymes, B.D., et al., Nucleic Acid Hybridization. A Practical Approach. IRL Press, Washington, 30 DC, 1985, the references of which are incorporated herein by reference. Use! the DNA or cDNA source is preferably enriched for ICAM-1 sequences. Such enrichment can most easily be obtained from cDNA obtained by extraction of RNA from cells cultured under conditions that induce 35 ICAM-1 synthesis (such as U937 grown in the presence of phorbol ester, etc.).

24 104952

Techniques described above or the like have been successfully able to clone human aldehyde dehydrogenase genes (Hsu, L.C., et al., Proc.Natl.Acad.Sei.

USA 82 (1985) 3771-3775), the fibronectin gene (Suzuki, S., 5 et al., Eur.J.Mol.Biol.Oran.J.4 (1985) 2519-2524), the human estrogen receptor gene (Walter, P.). , et al., Proc. Natl.Acad.Sci. USA 82 (7889-7893) (1985), the tissue-type plasminogen activator gene (Pennica, D., et al. Nature 301: 214-221 (1983)) and alkaline human placental phosphate. -10 again cummy DNA (Kam, W., et al., Proc.Natl.Acad. Sci. USA 82: 8715-8719 (1985)).

According to a preferred alternative method, an expression vector library is prepared by cloning a DNA or more preferably a cDNA from a cell capable of expressing ICAM-1 to an expression vector for cloning the ICAM-1 gene. The library is then screened for its members, which are capable of expressing a protein that binds to an anti-ICAM-1 antibody and has a nucleotide sequence that is capable of encoding polypeptides having the same amino acid sequence as ICAM-1 or ICAM-1. l fragments.

The cloned ICAM-1 gene obtained by the methods described above can be operably linked to an expression vector and introduced into bacterial cells or eukaryotic cells to produce ICAM-1 protein. Techniques for such performance are disclosed by Maniatis, T.,! et al., direct, and are well known in the art.

30 D. Using Assays for LFA-1 Influential Aqqration I g •

The assay described above, which is capable of measuring LFA-1-dependent aggregation, can be used to identify agents that act as antagonists by inhibiting the degree of LFA-35-dependent aggregation. Such antagonists may act by interfering with the ability of LFA-1 or ICAM-1 to mediate aggregation. Thus, such agents include immuno-104952 globulins, such as an antibody capable of binding to either LFA-1 or ICAM-1. In addition, non-immunoglobulin (i.e., chemical) agents can be assayed using the assay described above to determine if they are LFA-1 aggregate-5 antagonists.

E. Uses of Antibodies Capable of Binding to ICAM-1 Receptor Proteins 1. Anti-Inflammatory Agents Monoclonal antibodies to CD18 complex members inhibit many anchorage-dependent functions of white blood cells (Haskard, 15 D., et al., J. Immun. 137: 2901-2906 (1986), homotypic attachments (Rothlein, R., et al., J.Exo.Med. 163: 1132-1149 (1986)), antigen- and mitogen-induced proliferation of immune cells (Davignon , D., et al., Proc. Natl. Acad.Sci. USA 78: 4535-4539 (1981), antibody formation 20 (Fischer, A., et al., J. Immunol. 136: 3198-86 (1986)). 3203), and effector functions of all white blood cells such as cytotoxic lysis activity on T cells (Krensky, AM, et al.).

J. Immunol. 132: 2180-2182 (1984), cancer cells (Strassman, G., et al., J. Immunol. 136: 4328-4333 (1986)), 25 and all cells involved in antibody-dependent cellular cytotoxicity reactions (Kohl, S. ., et al ..

J. Immunol. 133: 2972-2978 (1984)). In all of the above functions, the antibodies inhibit the ability of the white blood cell to attach to a suitable cellular substrate, which in turn inhibits the final result.

«9

As discussed above, the binding of ICMA-1 molecules to LFA-1 moieties is of paramount importance for cell attachment. Through the attachment process, lymphoid cells 35 are able to continuously monitor the animal for the presence of foreign antigens. While these processes are normally desirable, they can also cause organ transplant rejection, tissue transplant rejection, and many autoimmune diseases. Thus, any means capable of attenuating or inhibiting cell attachment would be highly desirable to transplant recipients, tissue transplant recipients 5, and autoimmune patients.

Monoclonal antibodies capable of binding to ICAM-1 are very well suited as anti-inflammatory agents in a mammalian subject. Significantly, such agents differ from general antiinflammatory agents in that they are capable of selectively inhibiting adhesion and do not produce other side effects, such as renal toxicity, as found with conventional agents. Monoclonal antibodies capable of binding to ICAM-1 can therefore be used to prevent organ or tissue rejection and to modify autoimmune responses without fear of such side effects in a mammalian subject.

Significantly, the use of 20 monoclonal antibodies capable of recognizing ICAM-1 is capable of potentially transmitting even HLA-incompatible individuals.

2. Delayed Type Hypersensitivity Reaction Suppressors 25

Because ICAM-1 molecules are predominantly expressed at sites of inflammation, such as those exhibiting a delayed-type hypersensitivity reaction, antibodies (particularly monoclonal) capable of binding to ICAM-1 have 30 therapeutic uses for such reactions ;; to weaken or eliminate it. This therapeutic potential can be exploited in either of two ways. First, a composition comprising a monoclonal antibody to ICAM-1 may be administered to a patient with a delayed-type hypersensitivity reaction. For example, such compositions may be administered to an individual who has been exposed to antigens such as 104952 27 venom murine, poison oak, etc. . In another embodiment, a monoclonal antibody capable of binding to ICAM-1 is administered to a patient in combination with an antigen to prevent a subsequent inflammatory reaction. Thus, additionally administering an antigen with an ICAM-1 binding monoclonal antibody can thus temporarily confer resistance to a person subsequently exposed to that antigen.

10 3. Therapy for chronic inflammatory diseases

Since LAD patients lacking LFA-1 do not produce an inflammatory response, it is believed that antagonism of the natural LFA-1 ligand, ICAM-1, also prevents the inflammatory response. The ability of anti-ICAM-1 antibodies to inhibit inflammation is the basis for their therapeutic use in the treatment of chronic inflammatory and autoimmune diseases such as lupus erythematosus, autoimmune thyroiditis, experimental brain and spinal cord hypersensitivity 20 (EAE), multiple syndrome, multiple syndrome. Such antibodies may also be used as therapy in the treatment of psoriasis. In general, monoclonal antibodies capable of binding to ICAM-1 molecules can be used in the treatment of diseases currently treated with steroid therapy.

4. Diagnostic and Prognostic Uses Since ICAM-1 is predominantly expressed at sites of inflammation, monoclonal antibodies capable of binding to ICAM-1 can be used as a means of imaging or visualizing the site of infection and inflammation in a patient. In such use, monoclonal antibodies are labeled with detectable labels using radioisotopes, affinity labels (such as biotin, avidin, etc.), fluorescent labels, para-magnetic atoms, etc. Procedures for performing such labeling are well known in the art. The clinical use of antibodies in diagnostic imaging is reviewed by Grossman, H.B., Urol.Clin. North Amer. 13: 465-474 (1986); Unger, E.C., et al., Invest.Radioi. 20: 5,693-700 (1985), and Khaw, B.A., et al., Science 209: 295-297 (1980).

The presence of inflammation can also be detected using binding ligands such as mRNA, cDNA, or DNA that binds to ICAM-1 gene sequences, or ICAM-1 mRNA sequences, in cells expressing ICAM-1. Techniques for performing such hybridization assays are described by Maniatis, T. (Direct).

15 Finding the location of such detectable-labeled antibodies is a site of inflammation or tumor development. In one embodiment, this inflammatory detection is performed by taking tissue or blood samples and incubating such samples with a detectable label in the presence of Lei-20 matt antibodies. In a preferred embodiment, this technique is implemented in a non-invasive manner using magnetic imaging, fluorography, etc. Such a diagnostic test can be used to monitor transplant recipients for early signs of tissue rejection. Such assays may also be performed to determine a patient's tendency to rheumatoid arthritis or other chronic inflammatory diseases.

5. Adjuvant for administration of antigenic material for therapeutic or diagnostic purposes.

Immune responses to therapeutic or diagnostic agents such as bovine insulin, interferon, tissue-type plasminogen activator or murine monoclonal antibodies significantly reduce the therapeutic or diagnostic value of such agents and may, in fact, cause diseases such as serum disease. Such a 104952 29 fertilizer may be alleviated by the use of antibodies of the present invention. In this embodiment, such antibodies would be administered in combination with a therapeutic or diagnostic agent. The addition of antibodies prevents 5 recipients from recognizing the agent, thereby preventing the recipient from triggering an immune response against it. As a result of the absence of such an immune response, more therapeutic or diagnostic agent may be administered to the patient.

F. Intercellular Adhesive Molecule-1 (ICAM-1) Acid JCAM-1 is a LFA-1 binding pair. As such, ICAM-1 or functional derivatives thereof can be used interchangeably with antibodies capable of binding to LFA-1 in the treatment of a disease. Thus, in soluble form, such molecules can be used to prevent inflammation, organ rejection, graft rejection, etc. Similarly, ICAM-1 or functional derivatives thereof can be used as anti-ICAM antibodies to reduce the immunogenicity of therapeutic or diagnostic agents.

ICAM-1, its functional derivatives and its antagonists can be used to block metastasis or proliferation of tumor cells expressing ICAM-1 or LFA-1 surfaces. Numerous methods can be used to achieve such an objective. For example, migration of hematopoietic cells requires LFA-1-ICAM-1-30 binding. Antagonists of such binding therefore inhibit this migration and block the formation of metastases from tumor cells in white blood cell lines. Alternatively, toxin derivative molecules capable of binding to either ICAM-1 or LFA-1-35 group molecule may be administered to the patient. Upon binding of such toxin derivative molecules to tumor cells expressing ICAM-1 or the? 30 104952 LFA-1 group molecule, the toxin kills the tumor cell, thereby inhibiting tumor growth.

G. ICAM-1-Dependent Non-Immunoglobulin 5 Antagonist Adhesion Pathways ICAM-1-dependent adherence can be prevented by non-immunoglobulin antagonists capable of binding to either ICAM-1 or LFA-1. One example of a non-10 immunoglobulin antagonist of ICAM-1 is LFA-1. An example of a non-immunoglobulin antagonist that binds to LFA-1 is ICAM-1. Using the assays described above, other non-immunoglobulin antagonists can be identified and purified. Non-immunoglobulin-15 antagonists of ICAM-1-dependent attachment may be used for the same purpose as anti-LFA-1 or anti-ICAM-1 antibodies.

H. Administration of the Compositions of the Invention The therapeutic effects of ICAM-1 may be obtained by administering to the patient the entire ICAM-1 molecule, or any therapeutically active peptide fragment thereof.

ICAM-1 and its functional derivatives can be obtained either synthetically, by recombinant DNA technology, or proteolytically. The therapeutic benefits of ICAM-1 can be enhanced by using functional derivatives of ICAM-1 supplemented with additional amino acid residues to facilitate coupling to the carrier or to enhance the activity of ICAM-1. Functional derivatives of ICAM-1 which lack certain amino acid residues or which contain altered amino acid residues are intended to be encompassed within the scope of this invention, as long as such derivatives demonstrate the ability to effect cell attachment.

• f 104952 31

Both the antibodies of the present invention and the ICAM-1 molecule disclosed herein are said to be "substantially free of intrinsic impurities" if the preparations containing them are substantially free of the materials with which these products normally and naturally occur.

The present invention encompasses antibodies, and biologically active fragments thereof (whether polyclonal or monoclonal) capable of binding to ICAM-1. Such antibodies can be produced either in an animal or in tissue culture or by recombinant DNA technology.

When administering antibodies, or fragments thereof, capable of binding to ICAM-1, or administering to a patient receiving ICAM-1 (or a fragment, variant or derivative thereof), the dosage of the agent administered will vary depending on factors such as age, weight, In general, it is desirable to provide the recipient with an antibody dosage of about 1 pg / kg to 10 mg / kg (patient body weight), although lower or higher doses may be administered. When administering ICAM-1 molecules or functional derivatives thereof to a patient, preferably such molecules are administered at a dosage of about 1 pg / kg to 10 mg / kg (patient body weight), although lower or higher doses may be administered. As shown below, the therapeutically effective dose may be reduced if anti-LAM-1 antibody is administered in addition to anti-LFA-1 antibody. As used herein, a compound is said to be given in adjunct to another compound when the two compounds are administered in such close proximity that both compounds can be detected in the patient's serum at the same time.

«32 104952

Both the antibody capable of binding to ICAM-1 and ICAM-1 itself can be administered to patients intravenously, intramuscularly, subcutaneously, or via the gastrointestinal tract. For administration of the antibody or ICAM-1 by injection, administration may be by continuous fluid delivery or by single or multiple dose administration.

The anti-inflammatory agents of the present invention are to be administered to the receiving subjects in an amount sufficient to suppress the inflammation. The amount is said to be sufficient to "suppress" inflammation if the dosage, route of administration, etc. of the agent are sufficient to attenuate or prevent inflammation.

The anti-ICAM-1 antibody, or fragment thereof, may be administered alone or in combination with one or more other immunosuppressive agents (particularly in an organ or tissue transplant recipient). Administration of such compound (s) may have the purpose of either "prophylactic" or "therapeutic". For prophylactic purposes, the immunosuppressive compound (s) is administered prior to, during, or shortly after, any inflammatory reaction or symptom, but prior to any organ rejection symptoms. The prophylactic administration of Compound 25 (s) is intended to prevent or attenuate any subsequent inflammatory reaction (such as rejection of the transplanted organ or tissue, etc.). For therapeutic purposes, the immunosuppressive compound (s) is administered at the onset (or shortly after) of the actual inflammation (such as organ or tissue rejection). The therapeutic administration of the compound (s) is intended to attenuate the actual inflammation (such as rejection of the transplanted organ or tissue). Thus, the anti-inflammatory agents of the present invention can be administered either before the onset of inflammation (to suppress the expected inflammation) or after the onset of inflammation.

104952 33

The composition is said to be "pharmacologically acceptable" if its administration can be tolerated by the recipient patient. Such a substance is said to be administered in a "therapeutically effective amount" if the amount administered is physiologically significant.

A substance is physiologically significant if a change in the physiology of the receiving patient is observed as a result of its presence.

The antibody of the present invention and ICAM-1-Mole-10 kills may be formulated according to known methods to prepare pharmaceutically usable compositions, whereby these materials, or functional derivatives thereof, are incorporated into a mixture with a pharmaceutically acceptable carrier. Suitable carriers and their composition, including other human proteins, e.g., human serum albumin, are described, for example, in Remington's Pharmaceutical Sciences, 16th ed., Osol, A., ed. Mack, Easton PA, 1980. To form a pharmaceutically acceptable composition suitable for effective administration, such composition will contain an effective amount of an anti-ICAM antibody or ICAM-1 molecule, or a functional derivative thereof, together with an appropriate amount of a carrier.

: * 25 Other pharmaceutical methods may be used to control the duration of action. Controlled release preparations can be obtained using polymers to complex or absorb ICAM-1 antibody or ICAM-1, or functional derivatives thereof.

Controlled release can be achieved by selecting appropriate macromolecules (for example, polyesters, polyamino acids, polyvinylpyrrolidone, ethylene vinyl acetate, methylcellulose, carboxymethylcellulose, or protamine sulfate) and methods for controlling their content and content. Another possible method for controlling the duration of action when using controlled release formulations is to incorporate anti-ICAM-1 antibody or ICAM-1 molecules, or functional derivatives thereof, into polymeric material particles, such as polyesters, polymoacids, hydro-5 gels. , poly (lactic acid) or ethylene-vinyl acetate copolymers. Alternatively, instead of incorporating these agents into polymer particles, it is possible to encapsulate these materials in microcapsules made, for example, by coacervation techniques or interfacial polymerization, for example, hydroxymethylcellulose or for example, liposomes, albumin microspheres, microemulsions, nanoparticles and nanocapsules, or macroemulsions. Such techniques are disclosed in Remington's Pharmaceutical Sciences, 1980.

Having now described the invention in general terms, it will be more readily understood by reference to the following examples, which are provided by way of illustration, and are not intended to limit the present invention unless expressly mentioned.

EXAMPLE 1 25 Culturing Mammalian Cells

In general, EBV-transformed cells and hybridoma cells of the present invention were stored in RPMI 1640 medium supplemented with 30 mM L-glutamine, 50 micrograms / ml gentamycin, and 10% fetal bovine (or fetal calf). ) serum. Cells were cultured at 37 ° C in a 5% carbon dioxide atmosphere at 95% relative humidity.

To obtain 35 Epstein-Barr virus (EBV) transformants, 106 T-cell depleted peripheral blood mononuclear cells per ml of RPMI 1640 medium supplemented with 20% fetal calf serum (FCS) and 50 micrograms / ml gentamycin was incubated for 16 hours with EBV-containing supernatant of B95-8 cells (Thorley-Lawson, DA, et al., J.Exper.Med. 146: 495 (1977)).

5 aliquots of 0.2 ml cells were placed in 10 microtiter plate wells. The medium was replaced with RPMI 1640 medium (supplemented with 20% fetal calf serum and 50 micrograms / ml gentamycin) until cell growth was observed. Cells grew in most wells and propagated in the same medium. Phytohemagglutinin (PHA) mother cells were plated at 106 cells / ml in RPMI 1640 medium (supplemented with 20% fetal calf serum) containing 1: 800 dilution of PHA-P (Difco Laboratories, Inc., Detroit, MI). PHA cell-15 lines were propagated in medium treated with interleukin-2 (IL-2) and treated weekly with PHA (Cantrell, D.A., et al., J.Exper.Med. 158 (1895)). The preceding procedure was disclosed by Springer, T., et al., J.Exper.Med. 160: 1901-1918 (1984), which is incorporated herein by reference. Cells obtained by the above procedure are then screened with anti-LFA-1 antibodies to determine whether they express LFA-1 antigen. Such antibodies are disclosed by Sanchez-Madrid, F., et al., J.Exper.Med. 158: 1785 (1983).

25 EXAMPLE 2

Assays for cell aggregation and attachment

Aggregation assays were used to determine the degree of cell attachment. The cell lines used in these assays were washed twice with RPMI 1640 medium containing 5 mM Hepes buffer (Sigma Chemical, Co., St. Louis) and resuspended to 2 x 10 6 cells / ml. Flat-bottom 96-well microtiter plates 35 (No. 3596, Costar, Cambridge, MA) were supplemented with 50 microliters of the appropriate monoclonal antibody supernatant or 50 microliters of complete medium, containing or without purified monoclonal antibodies. -liter of complete medium containing 200 ng / ml phorbol ester phorbol myristate acetate (PMA) and 100 microliters of cells at a concentration of 2 x 10 6 cells / ml complete medium. This resulted in a final concentration of 50 ng / ml PMA and 2 x 10 5 cells / well. The cells were allowed to settle freely, and the degree of aggregation was evaluated at different times. Values ranged from 0 to 5+, with 0 indicating that there were essentially no cells in ry-10 membranes; 1+ indicated that less than 10% of the cells were in aggregates; 2+ indicated that less than 50% of the cells were aggregated; 3+ indicated that the cells were 100% in small, loose clusters; 4+ indicated that the cells were 100% aggregated into larger clusters; 15 and 5+ indicated that the cells were 100% in large, very dense clusters. To obtain a more quantitative estimate of cell attachment, reagents and cells were added to 5 ml polystyrene tubes in the same order as above. The tubes were placed in a rack on a rotary shaker 20 at 37 ° C. The tubes were shaken for one hour at about 200 rpm, after which 10 microliters of cell suspension was placed on a hemocytometer and the number of free cells was quantified.

% Aggregation was calculated from the following equation:% free cells aggregation - 100x (1 ------------------------------) Number of cells used 30 The number of cells used in the above formula is the number of cells per ml of control tube contents containing only cells and complete medium and not incubated. The number of free cells in the above equation is the same as the number of non-aggregated cells per milliliter of the tube content used in the assay. The foregoing procedures are described in Rothlein, R., et al., J.Exoer.Med. 163: 1132-1149 (1986).

37 104952 EXAMPLE 3 LFA-1-dependent cell aggregation

The qualitative aggregation assay described in Example 2 was performed using the Epstein-Barr transformed cell line JY. When PMA was added to the medium, microtiter plates showed cell aggregation. Time-lapse video recordings showed that the JY cells at the bottom of the microtiter plates were motile, showing active 10 membrane wrinkling and false foot movement. Contact between false cells of neighboring cells often resulted in cell-to-cell attachment. If adhesion was stable, the cell contact region was transferred to the uropod. Contact could be maintained despite intense cell movement and cell retraction in opposite directions. The main difference between PMA treated and untreated cells appeared to be the stability of these contacts once formed. PMA formed clusters of cells that grew in size as more cells adhered to their periphery.

Another method for measuring attachment was the quantitative assay described in Example 2. Cellus suspensions were shaken at 200 rpm for 2 '25 hours, then transferred to a hemocytometer and counted for the number of non-aggregated cells. In the absence of PMA, 42% (SD = 20%, N = 6) of JY cells were aggregated after 2 hours, whereas JY cells incubated in identical conditions at 50 ng / ml PMA, 87 % (SD = 8%, N = 6) was in the aggregates. Kinetic studies of aggregation showed that PMA enhanced the rate and extent of aggregation at all time points tested (Figure 3).

«104952 38 EXAMPLE 4

Inhibition of Cell Aggregation Using Anti-LFA-1 Monoclonal Antibodies To investigate the effects of anti-LFA-1 monoclonal antibodies on PMA-induced cell aggregation, these antibodies were added to cells incubated as in the qualitative aggregation assay of Example 2. Monoclonal antibodies were found to inhibit the formation of cellular aggregates in the presence and absence of PMA. Both the F (ab ') 2 and Fab' fragments against the LFA-1 alpha chain of monoclonal antibodies were able to inhibit cellular aggregation. When substantially 100% of the cells formed aggregates in the absence of anti-LFA-1, less than 20% of the cells were found to be aggregated when the antibody was used. The results of this experiment are described by Rothlein, R., et al. fJ.Exper.Med. 163: 1132-1149 (1986)).

20 EXAMPLE 5

Cell aggregation requires the presence of the LFA-1 receptor! EBV-transformed early cells were prepared in patient! Such cells were screened for monoclonal antibodies capable of recognizing LFA-1, and LFA-1 was found to be absent from the cells.

The qualitative aggregate thio assay described in Example 2 was used using cells described above lacking * LFA-1. Such cells did not spontaneously aggregate even in the presence of PMA.

104952 39 EXAMPLE 6 Finding ICAM-1

Cells lacking LFA-1 in Example 5 were labeled with carboxy-5-cifluorescein diacetate (Patarroyo, M., et al., Cell. Immunol. 63: 237-248 (1981)). Labeled cells were mixed at 1:10 with autologous cells or JY cells and the percentage of fluorescein-labeled cells in the aggregates was determined according to Rothlein, R., et al., J.Exoer.Med. 163: 1132-1149 (1986). Cells lacking LFA-1 were found to be able to co-aggregate with cells expressing LFA-1 (Fig. 4).

To determine whether LFA-1 was only relevant in aggregate formation or (also) in their maintenance, preformed aggregates described above were added to LFA-1 capable of binding. The antibody addition was found to severely violate the finished agg-20 regatta. Time-lapse video recording confirmed that the addition of monoclonal antibodies to the ready aggregates began to induce disruption within 2 hours (Table 1). After the addition of anti-LFA-1 monoclonal antibodies, the pseudo-foot movement and changes in aggregate form of the individual cells continued unchanged. The individual cells ruptured little by little on the edge of the aggregate; After 8 hours, the cells were predominantly lysed. According to the video time recording, the disruption of the preformed aggregates by LFA-1 monoclonal antibodies seemed to be equivalent to the · · ·· aggregation process in the absence of LFA-1 monoclonal antibody over time.

104952 40 TABLE 1 Ability of Anti-LFA-1 Monoclonal Antibodies to Destroy Preformed PMA-Induced JY Cell Aggregates-5

Aggregation points Experiment 18h 2 ha -mAb + mAb 10 1 4+ 4+ l + b 2 3+ 4+ l + c 3 5+ 5+ l + d 15 Aggregation in a qualitative microtiter plate assay was evaluated visually. When anti-LFA-1 antibody was present throughout the assay period, aggregation was less than 1+.

The amount of 20? A aggregation just prior to the addition of the monoclonal antibody at 2 hours.

bTS1 / 18 + TS1 / 22 • m 25 CTS1 / 18 dTS1 / 22 EXAMPLE 7 · 30 Need for Divalent Ions of LFA-1-Dependent Aggregation LFA-1-dependent attachments between cytotoxic T cells and targets require the presence of magnesium; 35 (Martz, E., J. Cell. Biol. 84: 584-598 (1980)). The dependence of PMA-induced JY cell aggregation on divalent 104952 41 cations was tested. The JY cells did not aggregate (using the assay of Example 2) in medium containing no calcium or magnesium ions. Addition of divalent magnesium supported aggregation at as low as 0.3 mM ion concentration. Addition of calcium ions alone had little effect. However, calcium ions were found to increase the ability of magnesium ions to support PMA-induced aggregation. When I, 25 mM calcium ions were added to the medium, low magnesium ion concentrations of 0.02 mM were found to support aggregation. These results indicate that LFA-1-dependent aggregation of cells requires magnesium ions, and that calcium ions, though insufficient on their own, can act synergistically with magnesium ions to promote aggregation.

15 EXAMPLE 8

Isolation of Hybridoma Cells Capable of Expressing Monoclonal Anti-ICC-1 Antibodies Monoclonal antibodies capable of binding to ICAM-1 were isolated by Rothlein, R., et al., J. Immunol. 137: 1270-1274 (1986), wherein said reference is incorporated herein by reference. Thus, 3 BALB / c mice were immunized intraperitoneally with EBV-transformed 25 peripheral blood mononuclear cells obtained from an individual lacking LFA-1 (Springer, T.A., et al.).

J. Exoer.Med, 160 (184) 1901). About 107 cells per ml of RPMI 1640 medium were used for each immunization. Immunization injections were given 45, 29, and 4 days 30 before splenocytes were harvested from mice to produce the desired hybridoma lines. On day 3, prior to spleen cell intake, mice were given a further 107 cells in 0.15 ml medium (intravenously).

Spleen cells isolated from the animals described above were fused to P3X73Ag8.653 myeloma cells in a 4: 1 ratio according to the procedure of Galfren, G., et al., Nature 266: 550 (1977). Sample 104952 of 42 hybridoma cells obtained were plated in 96-well microtiter plates. Hybridoma supernatants were screened for aggregation inhibition and one inhibitory hybridoma (600 wells tested) was cloned and subcloned using the limiting dilution technique. This subclone was named RR1 / 1.1.1 (hereinafter referred to as "RR1 / 1").

Monoclonal antibody RR1 / 1 was found to reproducibly inhibit LFA-1 expressing cell line JY PMA stimulated aggregation. The monoclonal antibody RR1 / 1 inhibits aggregation to the same degree, or slightly less, than some monoclonal antibodies to LFA-1 alpha / beta subunits. In contrast, control monoclonal-15 antibodies against HLA, which is highly expressed on JY cells, did not inhibit aggregation. Antigen bound by monoclonal antibody RR1 / 1 is defined as intercellular adhesion molecule-1 (ICAM-1).

EXAMPLE 9 Use of Monoclonal Anti-ICAM-1 Antibodies to Characterize ICAM-1 to Determine the Nature of ICAM-1, and in particular to Determine if ICAM-1 differs from LFA-1, cellular proteins were immunoprecipitated using monoclonal antibody RR1 / 1. Immunoprecipitation was performed by Rothlein, R., et al.

(J. Immunol. 137, 1270-1274 (1986)). JY cells were lysed at 5 x 10 7 cells / ml in 1% Triton X-100, 0.14 M NaCl, 10 mM Tris, pH 8.0 and just before use. add 1 mM phenylmethylsulfonyl fluoride using 0.2 units / ml of trypsin inhibitor aprotin (in lysis buffer) for 20 minutes at 4 ° C. The lysates were centrifuged at 10,000 x g for 10 minutes, after which they were pre-clarified by adding 50 microliters of a 50% suspension. roliters of CNBr-activated glycine-quenched Sepharose C1-4B gel for 1 h at 4 ° C.

104952 43

One milliliter of lysate was immunoprecipitated by the addition of 20 microliters of a 50% suspension of RR1 / 1 monoclonal antibody coupled to Sepharose Cl-4B resin (1 mg / ml) overnight at 4 ° C: 5 (Springer, TA, et al., J.Exper.Med. 160: 1901 (1984)). Sepharose-bound monoclonal antibody was prepared using CNBr activation of Sepharose Cl-4B in carbonate buffer Marchin, S., et al. (Anal. Biochem. 60 (1974) 149). Washed immunostaining was performed by SDS-PAGE and silver staining by Morrissey, J.H., Anal.

Biochem. 117, 307 (1981).

After elution of the proteins with SDS sample buffer (Ho, MK, et al., J. Biol. Chem. 258: 636 (1983), 100 ° C), the samples were split in half and subjected to electrophoresis (SDS-8% PAGE) (Fig. 5A) or under non-reducing (Fig. 5B) lanes with a molecular weight of 50 kd and 25 kd correspond to heavy and light chains of immunoglobulins monoclonal antibody-Sepharose (Fig. 5A, band). 3.) Various amounts of other bands were also observed in the 25 to 50 kd weight range, but were not present in the fine hairy leukemia cells yielding only the 90 kd molecular weight band. subunit and 95 kd beta-25 subunits were found to migrate differently than ICAM-1 under both reducing (Figure 5A, zone 2) and non-reducing (Figure 5B, zone 2) conditions.

The quantitative aggregation assay described in Example 2 was used to determine the effect of monoclonal antibody RR1 / 1 on PHA early 30 lymphocyte aggregation. Thus, T cell mother cells were stimulated for 4 days with PHA, after which they were extensively washed and then grown for 6 days in IL-2 treated medium. PHA 35 was found to be internalized during this 6 day culture and did not contribute to the aggregation assay. In three different assays on different T cell-derived cell lines, ICAM-1 monoclonal antibodies inhibit aggregation in a reproducible manner (Table 2).

TABLE 2 Inhibition of PMA-Stimulated PHA Early Cell Aggregation by RR1 / 1 Monoclonal Antibody3

Test PMA Mab Capture% Inhibition% 10 lc - Control 9 + Control 51 0 + HL-A, B 58 -14b + LFA-1 alpha 31 39 + ICÄM-1 31 39 15 2e - Control 10 + Control 78 0 + LFA-1 beta 17 78 + ICAM-1 50 36 3f - --- 7 20 + Control 70 + HLA-A, B 80 -14 + LFA-3 83 -19 + LFA-1 alpha 02 97 + LFA-1 beta 03 96 25 + ICAM-1 34 51 Aggregation of aPHA-induced early lymphocytes stimulated with 50 ng / ml PMA was quantitated indirectly by microscopic counting of non-aggregated cells (as described in Example 2).

% inhibition compared to cells treated with PMA and X63 monoclonal antibody.

104952 45 cAggregation was measured one hour after the addition of the monoclonal antibody and PMA. Cells were shaken at 175 rpm.

5 A negative number indicates a percentage increase in aggregation.

eAggregation was measured one hour after the addition of the monoclonal antibody and PMA.

Cells were pelleted at 200 x g for 1 min, incubated at 37 ° C for 15 min, then carefully resuspended and shaken for 45 min at 100 rpm.

The cells were pretreated with PMA for 4 hours at 37 ° C. After the addition of the monoclonal antibody, the tube contents were incubated at 37 ° C without shaking for 20 minutes and then shaking for 100 minutes at 75 rpm.

Repeatedly, LFA-1 monoclonal antibodies were more potent inhibitors than ICAM-1 monoclonal antibodies, whereas ULA A, B and LFA-3 monoclonal antibodies had no effect. These results indicate that of the monoclonal antibodies tested, only those capable of binding to LFA-1 or ICAM-1 were able to inhibit cell attachment.

EXAMPLE 10 Preparation of Monoclonal Antibody to ICAM-1 Immunization BALB / c mouse was immunized intraperitoneally (ip) 35 by injection of 0.5 ml of 2 x 10 7 JY cells in RPMI medium 103 and 24 days before fusion. . On days 4 and 3, 104952 46 pre-fusion mice were immunized i.p. 107 PMA differentiated U937 cells in 0.5 ml RPMI medium.

U937 Cell Differentiation I 5 U937 cells (ATCC CRL-1593) were differentiated by incubation at 5 x 10 µl in RPMI containing 10% fetal bovine serum, 1% glutamine and 50 micrograms / ml gentamycin (complete medium). 10 2 ng / ml phorbol-12 myristate acetate (PMA) in a sterile polypropylene container. On day 3 of this incubation, half of the medium volume was removed and replaced with fresh complete and PMA-containing medium. On day 4, the cells were removed, washed and prepared for immunization.

Fusion

Spleen cells from immunized mice were fused to P3x63. 20: 8: 1 to Ag8.653 myeloma cells according to Galfren et al., (Nature 266: 550, 1977). After fusion, cells were plated in 96 well flat bottom microtiter plates. concentration of 105 spleen cells / well.

Selection of positive cells as anti-ICAM-1

After one week, 50 microliters of supernatant was screened in the qualitative assay of Example 2 using both JY and SKW3 cells as aggregating cell lines. Cells from supernatants which inhibit JY cell aggregation but not SKW3 cell aggregation were selected and cloned 2 times using the limiting dilution method.

This assay identified and subsequently cloned three separate hybridoma cell lines that produced anti-ICAM-1 monoclonal antibodies. The antibodies produced by these hybrids were IgG2a, IgG2b and IgM, respectively. The hybridoma cell line that produced IgG2a antibody ICAM-1 was named R6,5'D6'E9, B2. The antibody produced by the preferred hybridoma cell line was named R6T5'D6'E9'B2 (herein "R6-5-D6").

EXAMPLE 11 ICAM-1 Expression and Regulation A radioimmunoassay was developed to measure 10 ICAM-1 expression. In this assay, purified RR1 / l was iodinated using iodogen with a specific activity of 10 microcurie units / microgram. Endothelial cells were grown in 96-well plates and treated as described for each experiment. The plates were cooled to 4 ° C by placing them in a cold room for 0.5-1 hours and not (i.e.) immediately on ice. The cell monolayers were washed 3x with cold complete medium and then incubated for 20 minutes at 4 ° C with 6 µl-RR1 / L. The cell monolayers were then washed 3x with complete medium. Bound * 251 was released using 0.1 N sodium hydroxide solution and counted. The specific activity of β1-RR1 / lin was adjusted using unlabeled RR1 / I to obtain a linear signal * 25 at the antigen densities observed in this study. Non-specific binding was determined in the presence of 10000 x excess of unlabeled RR1 / L and the value obtained was subtracted from total binding to obtain specific binding.

30 !! ICAM-1 expression, as measured by the radioimmunoassay described above, is increased by human umbilical vein endothelial cells (HUVEC) and human lower limb vein endothelial cells (HSVEC) by the action of IL-1, TNF, LPS and IFN-gamma (Table 3). Lower limb dermal endothelial cells were used in this study to confirm the results obtained with polar venous endothelial cells grown in large glass 48 104952 Kimo endothelial cells obtained from adult I tissue. The basic expression of IGAM-1 is 2x higher in lower limb skin vein endothelial cells than in polar venous endothelial cells. Upon exposure of umbilical vein endothelial-5 cells to recombinant IL-1-alpha, -IL-1-beta, and -TNF-gamma, ICAM-1 expression increased 10-20-fold. IL-1-alpha, TNF and LPS were the most potent inducers and IL-1 was less effective in weight loss and also in response saturation concentrations (Table 3). IL-1 beta at 10,100 ng / ml increased ICAM-1 expression 9-fold on HUVEC cells and 7.3-fold on HSVEC cells, resulting in a semi-maximal growth at 15 ng / ml. rTNF at 50 ng / ml increased ICAM-1 expression 16-fold in HUVEC cells and 11-fold in HSVEC cells, resulting in a half-maximal effect at 0.5 ng / ml. Interferon gamma produced a significant increase in ICAM-1 expression, which was 5.2-fold in HUVEC cells or 3.5-fold in HSVEC cells at 10,000 U / ml.

The effect of LPS at a concentration of 10 micrograms / ml was in the same order of magnitude as that of rTNF. Pairing of these mediators resulted in additive or slightly lower than additive effects on ICAM-1 expression (Table 3). Cross-titration of rTNF with rIL-1-beta or rIFF-gamma showed no synergism between these sub-optimal or optimal concentrations.

Because LPS increased ICAM-1 expression on endothelial cells at concentrations that sometimes occur in culture medium, it was possible that basic ICAM-1 expression was due to LPS in the tut-30 kit. When testing multiple batches of serum, it was found that V. low endotoxin serum gave 25% lower ICAM-1 base expression. All results reported herein were obtained with endothelial cells grown in low-endotoxin-containing rum. However, with the addition of LPS-neutralizing antibiotic at 10 micrograms / ml polymyxin-B, ICAM-1 expression was reduced by only 25% (Table 3). The increase in ICAM-1 expression-it-104952 49 treated with IL-1 or TNF was not affected by the presence of 10 micrograms / ml polymyxin B, consistent with the low levels of endotoxin contained in these preparations (Table 3).

5 TABLE 3 Monoclonal Antibodies to ICAM-1 10 Treatment (16 h) Specific Binding 125I (CPM)

HUVEC HSVEC

Control 603 ± 11 - 1132 ± 31 100 ng / mL yIL-1 beta 5680 ± 638 9x 8320 ± 766 7.3x 50 ng / mL yIL-1 alpha 9910 ± 538 16x 15 50 ng / mL yTNF alpha 9650 * 1500 16x 12690 ± 657 H, 2x 10 pg / ml LPS 9530 * 512 16iI 10459 1388 9 · 2 * 3120 ± 308 5.2x 4002 ± 664 3.5x 10 ng / ml ylFN gamma yIL-1 beta + TNF 1469 ± 1410 24x 16269 ± 660 14x yIL-1 beta + LPS 13986 ± 761 23x 10870 ± 805 lOx 20 yIL-1 beta + ylFN gamma 7849 ± 601 13x 8401 ± 390 7'4x yTNF + LPS 15364 * 1241 24x 16141 ± 1272 14x tt 13480 ± 1189 22x 13238 ± 761 12x yTNF + IFN gamma 10206 ± 320 17x 10987 ± 668 10x LPS + IFN gamma Polymyxin B (10 pg / ml) 480 ± 23 25 Polymyxin B + yIL-1 5390 ± 97 llx Polymyxin B + yTNF 9785 ± 389 20x 1 pg / ml LPS 7598 * 432 13x ",.." TTW, 510 ± 44 1, lx

polymyxin B + LPS

30 ·· Adjustment of ICAM-1 Expression on HVEC and HUVEC Cells - HUVEC or HSVEC cells were plated in 96-well plates at a 1: 3 ratio of a single cell monolayer and allowed to grow together. The cells were then treated with the aforementioned materials or medium for 16 hours, followed by an RIA assay as described in the protocols. All assay readings are from 4-fold duplicate assays.

EXAMPLE 12 Kinetics of Interleukin-1 and Interferon-Gamma Induction by ICAM-1

The kinetics of the effects of interleukin-1 and interferon-gamma on ICAM-1 expression in dermal connective tissue-forming cells were examined using a goat: 25 L-labeled anti-mouse IgG and antibody binding assay, Dustin, M.L., et al.

i (J. Immunol. 137: 245-254 (1986); which reference is incorporated herein by reference). To perform this binding assay, human skin connective tissue-forming cells were grown in a 15 96-well microtiter plate to a density of 2 to 8 x 10 4 cells / well (0.32 cm 2). The cells were washed twice with RPMI 1640 medium supplemented as described in Example 1. The cells were washed again with Hank's Balanced Saline (HBSS) containing 10 mM 20 Hepes, 0.05% NaN 3, and 10% heat-inactivated fetal bovine serum. . Washing with this binding buffer was performed at 14 ° C. To each well was added 50 microliters above! binding buffer and 50 microliters of a suitable hyb-! Ridoma supernatant, wherein X63 and W6 / 32, respectively, were used as negative and positive control.

The contents of the wells were incubated for 30 minutes at 4 ° C with gentle shaking, after which the wells were washed twice with binding buffer and another 50 µCi / 100 microliters of goat S1-labeled anti-mouse IgG was added. A goat-labeled anti-mouse antibody was prepared using Iodogen (Pierce) Frakerin, PJ et al (Biochem.Bioohvs.Res.Commun.

8-0 (1978) 849). After 30 minutes at 4 ° C, the cells were washed twice with 200 microliters of binding-35 buffer and the cell layer was solubilized by the addition of 100 microliters of 0.1 N sodium hydroxide solution. This batch of 100 microliters was counted on a Beckman Model 5500 gamma-

M

51 104952 with calculator. Specific binding scintillation / min was calculated from the formula [scintillation / min with monoclonal antibody] - [scintillation / min by X63]. All steps, including induction with specific reagents, were performed 4 times in duplicate.

The effect of interleukin-1 with a half-life of 2 hours in ICAM-1 induction was faster than that of interferon-gamma with a half-life of 3.75 hours (Figure 6). The time-10 curve for return to ICAM-1 resting levels appeared to be dependent on cell cycle or cell growth rate. In dormant cells, the effects of interleukin-1 and interferon-gamma are stable for 2 to 3 days, whereas in log-phase cultures, ICAM-1 expression remains near baseline for 2 to 15 days after removal of these inducers.

Figure 7 shows dose response curves for ICAM-1 induction with recombinant mouse and recombinant human interleukin-1 and recombinant human gamma interferon.

Interferon gamma and interleukin-1 were found to have similar concentration-dependencies with almost identical effects at 1 ng / ml. Human-murine recombinant interleukin-1 also had similar profiles, but is much less potent than human interleukin-1 preparations in inducing ICAM-1 expression.

Cyclohexamide, which inhibits protein synthesis, and actinomycin D, which inhibits mRNA synthesis, exterminate the effects of both interleukin-1 and interferon-gamma on ICAM- ·; 1 expression on connective tissue-forming cells (Table 4).

In addition, tunicamycin, which inhibits N-linked glycosylation, inhibits the effect of interleukin-1 by only 43%. These results indicate that protein and 35 mRNA synthesis, but not N-linked glycosylation, are prerequisites for interleukin-1 and gamma-interferon-stimulated increases in ICAM-1 expression.

• I

104952 52 TABLE 4

Effects of cycloheximide, actinomycin D and tunicamycin on induction of ICAM-1 by IL-1 and gamma IFN in human skin connective tissue formation cells® 5

Specific Binding 125I Goat Anti-Mouse IgG (cpm)

Treatment with anti-ICAM-1 anti-HLA-A, B, C

Control (4 h) 1425 ± 140 11928 ± 600 10 + cycloheximide 1513 ± 210 10678 ± 471 + actinomycin D 1590 ± 46 12276 ± 608 + tunicamycin 1461 ± 176 12340 ± 940 UL 1 (10 U / ml (4 h) 4264 ± 249 12155 ± 510 + cycloheximide 1619 ± 381 12676 ± 446. 15 + antinomycin D 1613 ± 88 12294 ± 123 j + tunicamycin 3084 ± 113 13434 ± 661 IFNγ (10 U / ml) (18 h) 4659 ± 109 23675 ± 500 + cycloheximide 1461 ± 59 10675 ± 800 + antimycin 1326 ± 185 12089 ± 550 20 human connective tissue-forming cells were grown to a density of 8 x 10 4 cells / 0.32 cm 2 well. Treatments were performed in a final volume of 50 µl containing said reagent.

reagents. Cycloheximide, actinomycin D and tunicamycin were added at a concentration of 20 micrograms / ml, 10

and 2 micrograms / ml, respectively, with cytokines. All readings are averages of 4-fold duplicate determinations, +/- SD.

• * • * i 30 EXAMPLE 13 ICAM-1 tissue distribution

Histological studies were performed using frozen human organ tissue to determine the distribution of ICAM-1 in 35 thymus, lymph nodes, intestine, skin, kidney, and liver. To perform such an analysis, frozen tissue sections (thickness 4 micrometers) from normal human tissues were fixed in acetone for 10 minutes and then stained with 5 monoclonal antibodies using the RR1 / 1 immunoperoxidase technique using the Avidin Biotin Complex Laboratory, ) described by Cerf-Bensussan, N., et al.

(J. Immunol. 130: 2615 (1983)). After incubation with antibody, sections were incubated sequentially with biotinylated equine anti-mouse IgG and avidin-biotinylated peroxidase complexes. The sections were finally dipped in a solution of 3-amino-9-ethylcarbazole (Aldrich Chemical Co., Inc., Milwaukee, WI) to develop a color reaction. The sections were then fixed in 4% formaldehyde solution for 5 minutes, after which they were counterstained with hematocycline. Sections incubated with non-related monoclonal antibodies instead of RR1 / I antibody were used as controls.

The distribution of ICAM-1 was found to be very similar to that of the major Tissue Compatibility Complex (MHC) class II anti-genes. The majority of blood vessels (both small and large) in all tissues showed endothelial cell dye-Y 25 retention with ICAM-1 antibody. Vascular endothelial staining was more pronounced in the inter-follicular (paracortical) areas of the lymph nodes, tonsils and Peyer's lymph nodes than in the blood vessels of the kidneys, liver and normal skin. Liver staining was predominantly confined to capillary cortex cells; liver cells and endothelial cells in most of the portal veins and arteries in the walls were not stained.

Diffusion of large cells and a dendritic staining pattern were observed in the thymus. In the cortex, the staining pattern was local and predominantly dendritic. The thymus cells did not stain. In peripheral lymphoid tissue 104952 54, the lymph node germ cells of the secondary lymphocytes were strongly stained. In some lymphoid cells, the staining pattern was predominantly dendritic, with no apparent staining of the lymphocytes. Poor staining of cells in the mantle band was also observed. In addition, dendritic cells (interdigitation reticulocytes), which contain cytoplasmic extensions, and a small proportion of lymphocytes in the intercellular or paracortical regions were stained with ICAM-1 binding antibody.

10

Phagocytes resemble the cells were stained in the lymph nodes and small intestine in the middle layer. Fibrous tissue-like (helical) cells and dendritic cells dispersed in the basal tissue of most organs examined were stained with ICAM-1 binding antibody. No staining was observed in Langerhans / indeterminant cells in the oropharynx. No staining was observed in smooth muscle tissue.

Epithelial cell staining was observed in 20 reproducible ways in the tonsil mucosa. Although hepatocytes, bile duct-hytepiteeli, intestinal epithelial cells, and tubular epithelial cells in kidney did not stained in most cases, normaalimunuaiskudosleikkeet obtained from kidney solusyöpäliitteisestä nephrectomy specimen showed! V 25 ICAM-1 staining of many proximal tubule cells. These tubular epithelial cells were also stained with anti-HLA-DR binding antibody.

In summary, ICAM-1 is expressed in non-haematopoietic cells such as vascular endothelial cells and in haematopoietic cells such as tissue cancer cells and mitogen-stimulated T lymphocyte stem cells. ICAM-1 was found to be expressed at low concentrations in peripheral blood lymphocytes.

104952 55 EXAMPLE 14 Purification of ICAM-1 by Monoclonal Antibody Affinity Chromatography [0121] The overlay diagram of ICAM-1 was purified from human cells or tissue using monoclonal antibody affinity chromatography. The monoclonal antibody, RR1 / 1, which reacts with ICAM-1, was first purified and then coupled to an inert column matrix. This antibody is described by Rothlein, R., et al., J. Immunol. 137: 1270-1274 (1986) and Dustin, M.L., et al. (J. Immunol. 137: 245 (1986)). ICAM-1 was solubilized from cell membranes by lysing into cells in a nonionic detergent, Triton X-100, at a pH near neutral. The cell lysate containing the dissolved ICAM-1 was then used in pre-columns designed to remove materials that bind non-specifically to the column matrix material and then through the monoclonal antibody-20 column to allow binding of ICAM-1 to the antibody. The antibody column was then washed with a series of detergent-wash buffers whose pH increased to 11.0. During these washes, ICAM-1 remained bound to the antibody matrix, while non-binding and poorly bound. Bound ICAM-1 was then specifically eluted from the column using a pH 12.5 detergent buffer.

Purification of Monoclonal Antibody RR1 / 1 and Hard-30 Plate Coupling to Sepharose C1-4B Gel «· Anti-ICAM-1 Monoclonal Antibody RR1 / 1 was purified by intraperitoneal anesthesia or hybridoma culture supernatants from hybridoma mice using conventional techniques. protein A affinity chromatography (Ey, et al., Immunochem. 15: 429 (1978)). Purified IgG, or rat 104952 56

IgG (Sigma Chemical Co., St. Louis, MO) was hardly coupled to Sepharose C1-4B gel (Pharmacia, Uppsala, Sweden) using the modification of Marchin et al. (Anal. Biochem. 60, 149 (1974)). Briefly, Sepharose C1-4B was washed with distilled water, activated with 40 mg / ml CNBr in 5M potassium hydrogen phosphate solution (pH about 12) for 5 minutes and then washed extensively with 0.1 mM hydrochloric acid at 4 ° C. The filtered activated Sepharose was resuspended in an equal volume of purified antibody-10 (2-10 mg / ml in 0.1 M sodium bicarbonate solution, 0.1 M NaCl). The slurry was incubated for 18 hours at 4 ° C with gentle rotation. The supernatant was then assayed for the presence of unbound antibody by measuring absorbance at 280 nm, and the remaining reactive centers of activated Sepha-15 Rose were saturated by the addition of glycine to a concentration of 0.05 M. The coupling efficiency was usually> 90%.

Solubilization of human spleen membranes with 20 detergents

All operations were performed at 4 ° C. Frozen human spleens (in 200 g fragments) from patients with hairy cell leukemia were thawed in ice in 200 ml of Tris saline (50 mM Tris, 0.14 M NaCl, pH 7.4, 4 ° C). ) with 1 mM phenylmethylsulfonyl fluoride (PMSF), 0.2 U / ml aprotin and 5 mM iodoacetamide. The tissue was cut into small pieces and homogenized at 4 ° C in a Tekmar power homogenizer. The volume 30 was then filled to 300 ml by addition of Tris saline and 100 ml of a 10% solution of Tween 40 (polyoxyethylene sorbitan monopalmitate) in Tris saline was added to give a final concentration of 2.5% Tween 40 to prepare membranes. the homogenate was extracted using three strokes on a Dounce, or more preferably Teflon Potter Elvejhem homogenizer, followed by centrifugation at 1000 xg for 15 minutes. The supernatant was recovered and the pellet was repeatedly extracted with 200 ml of 2.5% Tween 40 in Tris saline. The mixture was centrifuged at 1000 x g for 15 minutes, after which the supernatants from each extract were pooled and centrifuged at 150,000 x g for 1 hour to pellet the membranes. The membranes were washed by re-slurry in 200 ml of Tris saline solution by centrifugation at 150,000 x g for 1 hour. The membrane pellet was resuspended in 10 200 ml Tris saline and homogenized in a motor driven homogenizer using a Teflon pestle until the slurry was uniformly cloudy. The volume was then filled to 900 ml by addition of Tris saline and N-laoyl sarcosine was added to a final concentration of 1%. The mixture was stirred at 4 ° C for 30 minutes, after which the insoluble material in the detergent lysate was removed by centrifugation at 150,000 x g for 1 hour. Triton X-100 was then added to the supernatant to a final concentration of 2% and the lysate was stirred at 4 ° C for 1 hour.

Detergent Lysing of 20 JY-B Ear Stimulated Cells EBV-transformed B-early stem cell line JY was grown in RPMI 1640 medium containing 10% fetal calf serum (FCS) and 10 mM Hepes at a cell density of about 0.8 to 1.0 x 10 6 cells / ml. To increase cell surface expression of ICAM-1, phorbol-12-myristate-13-acetate (PMA) was added to 25 ng / ml for 8-12 hours before harvesting the cells. Sodium vanadate (50 microM) was also added to the cultures. Cells were pelleted by centrifugation at 500 x g for 10 minutes and then washed twice with Hank's equilibrated saline (HBSS) by resuspension and centrifugation. Cells (approximately 5 g in 5 liters of culture) were lysed in 50 ml of lysis buffer (0.14 NaCl, 50 mM Tris, pH 8.0, 1%

Triton X-100, 0.2 U / ml aprotin, 1 mM PMSF, 50 microM sodium vanadate) with stirring at 4 ° C for 30 minutes.

104952 58

The non-lysed nuclei and insoluble waste were centrifuged at 10,000 x g for 15 minutes, after which the supernatant was centrifuged at 150,000 x g for 1 hour and the filtrate was filtered through Whatman 3 mm filter paper.

ICAM-1 affinity chromatography for structural studies A column containing 10 ml of RR1 / 1-Sepharose C1-4B gel (coupling 2.5 mg antibody / ml gel) was used to purify 10 ICAM-1 on a large scale for structural studies. ) and two precolumns containing 10 ml of CNBr-activated glycine-quenched Sepharose C1-4B15 gel and rat IgG coupled to Sepharose C1-4B gel (2 mg / ml). The columns were coupled in series and pre-washed with 10 column volumes of lysis buffer, 10 column volumes of pH 12.5 buffer (50 mM triethylamine, 0.1% Triton X-100, pH 12.5, 4 ° C), then equilibrated with 10 column volumes. lysis buffer. One liter of human spleen detergent lysate was charged at a flow rate of 0.5 to 1.0 ml / min. Two pre-columns were used to remove non-specifically bound material from the lysate prior to use on the RR1 / 1-Sepharose column.

7 25

After loading, the column containing RR1 / 1-Sepharose and ICAM-1 bound thereto was washed successively at a flow rate of 1 mL / min with at least 5 column volumes of each of the following solutions: 1) Lysis buffer, 2) 20 mM Tris, pH 8.0 / 0 , 14 M NaCl / 0.1% Triton X-100, 3) 20 mM glycine, pH 10.0 / 0.1% Triton X-100, and 4) 50 mM triethylamine, pH 11.0 / 0.1 % Triton X-100. All wash buffers contained 1 mM PMSF and 0.2 U / ml aprotin. After washing, the remaining bound ICAM-1 was eluted with 5 column space-35 volumes of elution buffer (50 mM triethylamine / 0.1%

Triton X-100 / pH 12.5, 4 ° C) at a flow rate of 1 ml / 3 min. The eluted ICAM-1 was collected in 1 mL fractions and neutralized 104952 59 immediately by adding 0.1 mL of 1 M Tris pH 6.7. Fractions containing ICAM-1 were identified by SDS-polyacrylamide gel electrophoresis in 10 microliter aliquots (Springer et al., J.Exp.Med. 160: 1901 (1984)) followed by silver staining (Morrissey, J.H., Anal.Biochem.

117: 307 (1981). Under these conditions, the vast majority of ICAM-1 eluted at about 1 column volume and had a purity of greater than 90% on silver-stained electropherograms (the main impurity was a small amount of IgG spilled from the affinity matrix). Fractions containing ICAM-1 were pooled and concentrated approximately 20-fold using Centricon-30 microconcentrators (Amicon, Danvers, MA). Purified ICÄM-1 was quantified by performing a Lowry protein assay on an ethanol precipitated pool sample: 200 g of human spleen produced about 500 micrograms of pure ICAM-1.

About 200 micrograms of purified ICAM-1 was subjected to a second purification step using preparative SDS-polyak-20 aryl amide gel electrophoresis. The band corresponding to ICM-1 was visualized by soaking the gel in 1M potassium chloride solution. The gel region containing ICAM-1 was then excised and subjected to electroelution by Hunkapiller et al., Meth. Enzymol. 91: 227-236 (1983), according to method 25. The purity of the purified protein was> 98% as judged by SDS-PAGE and silver staining.

ICAM-1 Affinity Purification for Functional Studies 30 • t * · ICAM-1 for use in functional assays was purified from JY cell detergent as described above, but on a smaller scale (1 mL column with RR1 / 1-Sepharosea), and using the following modifications. All solutions contained 50 microM sodium vanadate. After washing the column with pH 11.0 buffer containing 0.1% Triton X-100, column 104952 60 was again washed with five column volumes of the same buffer containing 1% n-octyl-beta-D-glucopyranoside ( octyl glucoside) instead of 0.1% Triton X-100. The octyl glucoside detergent replaces Triton X-100 bound to ICAM-1 and, unlike Triton X-100, can subsequently be removed by dialysis. ICAM-1 was then eluted with pH 12.5 buffer containing 1% octylglucoside instead of 0.1% Triton X-100, then analyzed and concentrated as described above.

EXAMPLE 15

Characteristics of Purified ICAM-1 Human spleen purified ICAM-1 migrates on SDS-polyacrylamide gels with a broad band of Mr 72,000-91,000. ICU-1 purified from JY cells also migrates with a broad band of Mr 7,600 These values for Mr are from various cell sources for the immunoprecipitated ICAM-1 in the reported region: Mr = 90,000 for JY cells, 114,000 for myelomonocyte cell line U937, and 97,000 for connective tissue formation cells (Dustin et al.). J. Immunol. 137 (1986) 245). This broad Mr region is considered to be due to a strong but variable degree of glycosylation. The non-glycosylated pre-: 25 cursor has a Mr of 55,000 (Dustin et al.). Protein purified from either JY cells or human spleen retains its antigenic activity as evidenced by its ability to re-bind to the original affinity column and immunoprecipitation on RR1 / 1-Sepharose gel and SDS-polyacrylamide-dielectrophoresis.

To produce ICAM-1 peptide fragments, about 200 micrograms was reduced using 2 mM dithiothreitol / 2% SDS followed by alkylation using 5 mM iodo-35 acetic acid. The protein was precipitated with ethanol and redissolved in 0.1 M NH4CO3 / 0.1 mM CaCl2 / 0.1% Zwittergent 3-14 (Calbiochem) followed by digestion with 104952 61 at 1% w / w trypsin at 37 ° C. At 4 ° C for 4 hours, followed by further digestion with 1% trypsin for 12 hours at 37 ° C. Trypsin peptides were purified by reverse phase HPLC using a 0.4 x 15 cm C4 column (Vydac). Peptides were eluted using a linear gradient of 0% to 60% acetonitrile / 0.1% trifluoroacetic acid. Selected peptides were subjected to sequence analysis with a gas-phase micros sequencer (Applied Biosystems). Sequence information obtained from this study is presented in tau-10, Table 5.

TABLE 5 Amino acid sequences of ICAM-1-trpsin peptides 15 Amino-Peptide acid ------------- 50a 50b 46a 46b X 45 K AA J U O Ml residue

1 [T / V] A (V / A) EV S LEAL VL

2 F SQ PEFNLGLT L / E

20 3 L IT ALPPDSGL P / (G)

4 T SF AATLVI P

5V LP PPVRLE G / Y

6 Y GL LNTPVT N / L

7 P WP PVYQTP (N) 25 8 T P I I T / I G G C P / V (Q) - 9 S F G (G) L - L S K (E) Γ 10 E E (Q) - D E T (D).

11 A S D / P K SLS

12 G / S V V P F F C

30 13 A T DQSED

14 G V W V / L A - Q

15 I K T P

16 SK

'17 A

35 18 P

19 X

20 Q

21 L

62 104952 () = low probability sequence.

[] = very low probability sequence.

/ = denotes sequence ambiguity; the most likely amino acid is marked first.

5a = High Peptide b = Low Peptide EXAMPLE 16 Cloning of the ICAM-1 gene: The 10 ICAM-1 gene can be cloned using any number; different procedures. For example, the amino acid sequence information obtained by sequencing ICAM-1 trypsin; nif fragments (Table 5) can be used to identify an oligonucleotide sequence corresponding to the ICAM-1 gene. Alternatively, the IGAM-1 gene may be cloned using an anti-ICAM-1 antibody to identify clones producing ICÄM-1.

Cloning of the ICÄM-1 gene using oligonucleotide probes Using the genetic code (Watson, JD, Molecular Biology of the Gene, 3rd edition, WA Benjamin, Inc., Menlo Park, CA, 1977), one or more different oligonucleotides can be identified, each capable of coding for kood 25 countries ICAM-1 trypsin peptides. The likelihood that a particular oligonucleotide is actually an actual coding sequence for ICAM-1 can be estimated by looking at the abnormal base pairing ratios and the density at which a particular codon is actually used (to encode a given amino acid) in eukaryotic cells.

«! Such "codon codes" are disclosed by Lathe, R., et al., J.Molec.Biol. 183: 1-12 (1985). Using Lathe's "codon protocol" identifies a single oligonucleotide sequence, or set of oligonucleotides, containing the theoretical "35 most likely" nucleotide sequence (i.e., the lowest abundance nucleotide sequence) capable of encoding ICAM-1 .

104952 63

An oligonucleotide, or set of oligonucleotides containing a theoretical "most probable" sequence capable of encoding ICAM-1 fragments, is used to identify a complementary oligonucleotide or sequence of oligonucleotides that is capable of hybridizing to the "most probable" sequence. An oligonucleotide containing such a complementary sequence can be used as a probe for the identification and isolation of the ICAM-1 gene (Maniatis, T., et al., Molecular Cloning -10A Laboratory Manual, Cold Spring Harbor Press, Cold Spring Harbor, NY, 1982).

As described in Part C, supra, it is possible to clone the ICÄM-1 gene from eukaryotic DNA preparations which are believed to contain this gene. To identify and clone the gene encoding ICAM-1, the DNA library is screened for its ability to hybridize to the oligonucleotide probes described above. Because it is likely that a normal diploid cell will contain only two copies of the ICAM-1 gene, and because it is possible that the ICAM-1 gene may contain large non-transcriptionable intermediate sequences (introns) that are unwanted to be cloned, preferably ICAM -1 coding sequences are isolated from a cDNA library prepared from mRNA of 25 cells producing ICAM-1, rather than from genomic DNA. A suitable DNA or cDNA preparation is enzymatically cleaved or randomly ligated and ligated into recombinant vectors. The ability of these recombinant vectors to hybridize to the oligonucleotide probes described above is then measured. Procedures for carrying out hybridization are described, for example, in publications

r V

; Maniatis, T., Molecular Cloning - A Laboratory Manual, Cold

Spring Harbor Press, Cold Spring Harbor, NY, 1982, or Haymes, B. T., et al., Nucleic Acid Hybridization - A Practical Approach. IRL Press, Oxford, UK, 1985. Vectors capable of such hybridization are then analyzed to determine the amount and quality of the ICAM-1 sequences they contain. From a purely statistical point of view, any 64,104,952 genes, such as the gene encoding the ICAM-1 molecule, could be uniquely identified (by hybridization screening) using only 18 oligonucleotide probes.

Thus, in summary, the actual identification of ICAM-1 peptide sequences allows for the identification of a theoretical "most likely" DNA sequence, or sequence of such sequences, capable of encoding such a peptide. By constructing 10 oligonucleotides that complement this theoretical sequence (or by constructing a set of oligonucleotides that complement a set of "most probable" oligonucleotides), a DNA molecule (or series of DNA molecules) capable of serving as a probe for ICAM-1 is obtained. to identify the gene and isolate it.

Using the ICAM-1 peptide sequences of Table 5, the sequence of the oligonucleotide having the "most probable" sequence capable of encoding AA and J peptides was identified (Table 6 and 7, respectively). Oligonucleotides complementing these sequences were synthesized and purified for use as probes to isolate ICAM-1 gene sequences. Suitable size-selected cDNA libraries were constructed from poly (A) + - RNA derived from '25 PMA-induced HL-60 cells or PS-stimulated umbilical vein endothelial cells. A size-selected cDNA library was prepared using poly (A) + - RNA obtained from PMA-induced HL-60 cells, Gublerin, U., et al. (Gene 25: 263-269 (1983)) and Corbin, A., et al.

30 (EMBO J. 6, 4023-4028 (1987)), which references are incorporated by reference; herein by reference, according to the method.

A size-selected cDNA library was prepared using 5 micrograms / ml poly (A) + RNA derived from umbilical vein endothelial-35 bones stimulated with PS for 4 hours. RNA was extracted by homogenizing the cells in 4 M guanidine-phencyium isothiocyanate and ultracentrifuging the supernatant to 65 104952 using a CsCl gradient (Chirgwin, J.M., et al., Biochem. 18 (1979) 5294-5299). Poly (A) + RNA was isolated from total RNA species mixture using oligo- (dT) cellulose chromatography (Type 3, Collaborative Research) 5 (Aviv, H., et al., Proc.Natl.Acad.Sei. USA). 69: 1408-1412 (1972)).

66 1 0 4 9 5 2 TABLE 6

An oligonucleotide that encapsulates the most likely nucleotide sequence capable of encoding the ICAM-1-AA peptide of ICAM-1-AA Most Amino Acid ICAM-1 AA peptide Complementary group AA-peptide coding sequence 5 '3'

162 Glu G C

The T

G C

163 Leu C G

T A

G C

164 Asp G C

The T

C G

165 Leu C G

T A

G C

166 Arg C G

G C

G C

167 Pro C G

C G

C G

168 Gin C G

The T

G C

169 Gly G C

G C

j CG

170 Leu C G

T A

G C

171 Glu G C

The T

G C

172 Leu C G

\ TA

G C

173 Phe T A

T A

T A

174 Glu G C

The T

G C

3 '5' 104952 67 TABLE 6 (cont'd)

An oligonucleotide complementary to the most likely nucleotide sequence capable of encoding the ICAM-1-AA peptide of ICAM Most Amino Acid ICAM-1 AA Peptide Complementary Group AA Peptide Encoding Sequence

175 As n A T

The T

C G

176 Thr A T

C G

C G

177 Ser U A

C G

A

3 '5' W 9 - »9 104952 68 TABLE 7

An oligonucleotide that encapsulates the most probable nucleotide sequence capable of encoding the ICAM-1-J peptide ICAM The most amino acid acid ICAM-1 AA peptide The cohesive moiety AA peptide encoding sequence 5 '3'

19 Or G C

T A

G C

20 Thr A T

C G

C G

21 Cys T A

G C

C G

22 Ser T A

C G

C G

23 Thr A T

C G

C G

24 Ser T A

C G

C G

25 Cys T A

G C

T A

'2 6 Asp G C

AT C G

27 Gin C G

The T

G C

28 Pro C G

C G

C G

2 9 Lys A T

AT 3 '5' 104952 69

The first cDNA strand was synthesized using 8 micrograms of poly (A) + RNA, avian myeloblastosis virus reverse transcriptase (Life Sciences), and oligo (dT) primer. The DNA-RNA hybrid was digested with RNase H (BRL) and the other strand 5 was synthesized using DNA polymerase I (New England Biolabs). The product was methylated with EcoRI methylase (New England Biolabs), and the flattened ends were ligated to EcoRI (New England Biolabs), followed by digestion with EcoRI and size selection on a low melting point agarose gel. CDNAs larger than 500 bp in size were ligated into lambda-gt10 pre-digested with EcoRI and dephosphorylated (Stratagene). The ligation product was then packaged (Stratagene Gold).

The umbilical vein endothelial cell and HL-60 cDNA libraries were then plated to a concentration of 20,000 cfu / 150 mm plate. The recombinant DNA was transferred in duplicate onto nitrocellulose filters, denatured using 0.5 M NaOH / 1.5 NaCl, neutralized by addition of 1 M Tris buffer, pH 7.5 / 1.5 M NaCl, and baked at 80 ° C. for 2 hours (Benton, WD, et al., Science 196: 180-182 (1977)). The filters were pre-hybridized and then hybridized in 5XSSC containing 5X Denhardt's solution, 50 mM NaPO 4 and 1 microgram / ml salmon sperm DNA. The prehydrogenation was carried out at 45 ° C and allowed to proceed for 1 hour.

32 bp (5-TTGGGCTGGTCACAGGAGGTG-GAGCAGGTGAG) or 47 bp (5'-GAGGTGTTCTCAACAGCTCCAGGCCCTGG-30 GGCCGCAGGTCCAGTCTC) were used for the non-information-based oligonucleotides with the 1-peptide-based and 1-I respectively, AA (Tables 7 and 6) (Lathe, R., J. Molec. Biol. 183, 1-12 (1985)). The ends of the oligonucleotides were labeled with gamma (32 P) ATP using T4 polynucleotide-35 kinase and conditions recommended by the manufacturer (New England Biolabs). The filters were hybridized overnight, then washed twice with 2XSSC / 0.1% SDS solution 30 104952 for 70 minutes at 45 ° C. The phages were isolated from colonies that had undergone hybridization, and then purified by plating and repeatedly screening.

Cloning of the ICAM-1 gene using anti-ICAM-1 antibody Alternatively, the ICAM-1 gene may be cloned using 10 anti-ICAM-1 antibodies. DNA, or more preferably cDNA, is extracted and purified from a cell capable of expressing ICAM-1. Purified cDNA is fragmented (by digestion, endonuclease digestion, etc.) to obtain a pool of DNA or cDNA fragment. The DNA or cDNA-15 fragments of this pool are then cloned into an expression vector to produce an expression vector genomic library, each of which contains a single cloned DNA or cDNA fragment.

EXAMPLE 17 Analysis of 20 cDNA Clones

Phage DNA from positive clones was digested with EcoRI and examined by Southern analysis using single-clone cDNA as a probe. Cross-hybridizing maxi; The 25 micron cDNA inserts were subcloned into the EcoRI site of the plasmid vector pGEM4Z (Promega). HL-60 subclones containing both directions of the cDNA were destroyed by digestion with exonuclease III (Henikoff, S., Gene 28: 351-359 (1984)) following the manufacturer's recommendations (Erase-a-Base, Prome-30 ga). . Gradually, the destroyed cDNAs were then cloned and subjected to dideoxynucleotide chain access sequencing (Sanger, F., et al., Proc.Natl.Acad.

Sci. USA 74: 5463-5477 (1977) according to the manufacturer (Sequenase, U.S. Biochemical). The 5 'and size 35 regions of the HL-60 cDNA were completely sequenced from each strand and the 3' region was sequenced approximately 70% from each strand. The representative endothelial cDNA was sequenced by virtually the entire 104952 71 "shotgun" cloning of 4 bp recognition restriction enzyme fragments.

The cDNA sequence of one HL-60 and one endothelial cell cDNA was determined (Figure 8). The 3023 bp sequence contains a short 5'-untranslated region and a 1.3 bp 3'-untranslated region, at position 2966 a consensus suspolyadenylation signal. The longest open reading region starts at the first ATG codon at position 58 and ends at TGA-Lope-10 at position 1653. Since the translated amino acid sequence and the eight (8) different trypsin peptides, a total of 91 amino acids (underlined in Figure 8), the sequences determined were similar, isolated authentic ICAM-1 cDNA clones. The amino acid sequences of ICAM-1-trypsin-15 nipeptides are shown in Table 8.

TABLE 8 Amino acid sequences of ICAM-1 trypsin peptides 20 Peptide Groups Sequence

J 14-29 X G S V L VTCSTSCDOPK

U 30-39 L L G I E T P L (P) (K)

50 78-85 (T) F L T V Y X T

X 89-95 V E L A P L P

; 25 AA 161-182 X ELDLRPOGLE -

L F E X T S A P X Q L

K 232-246 LNPTVTYGXDSFSAK

45 282-295 S F P A P N V (T / I) L X K P Q (V / L) 30 - indicates that the sequence continues on the next line; ; the underlined sequences were used in the preparation of oligonucleotide probes.

According to the results of the hydrophobicity analysis (Kyte, J., et al., J. Molec Biol. 157, 35, 105-132 (1982)), a signal sequence of 27 residues is present. The + 1-glutamine placement fits in with the failure to obtain the N-terminal sequence for three different 104952 72 ICAM-1 protein preparations; glutamine may possibly be cyclized to pyroglutamic acid to form a blocked N-terminus. The translated sequence 1-453 is predominantly hydrophilic, followed by a hydrophobic putative membrane-permeable region of the 24 residues 5. This permeable membrane-permeable region is immediately followed by multiple charge residues located in the putative cytoplasmic region of 27 regions.

The predicted size for the mature polypeptide chain is 55,219 dalits, which is well suited to the size 55,000 obtained for highly deglycosylated ICAM-1 (Dustin, M.L., et al., J. Immunol. 137, 245-254 (1986)). 8 N-linked glycosylation centers are predicted. The lack of asparagine in trypsin peptide sec-15 sequences at two of these centers ensures their glycosylation and extracellular orientation. Assuming a high mannose N-linked carbohydrate size of 2500 Daltons, the size prediction of the ICAM-1 precursor is 75,000 Daltons as compared to the observed 20 73,000 Daltons (Dustin, ML, et al., J. Immunol. 137: 245-86 (1986)). 254). When a high-mannose product is converted to a complex carbohydrate, the mature ICAM-1 glycoprotein is 76 to 114 kd in size, depending on the cell type (Dustin, ML, et al., J. Immunol. 137 (1986) 25255-254). . Thus, ICAM-1 is a highly glycosylated but otherwise typical integral membrane protein.

EXAMPLE 18 ICAM-1 is an Integrin Binding Member of Immunoglobulin Supergenic Group 30 * Internal replicates of ICAM-1 were inserted using parallel | Microgenie protein orientation program (Queen, C., et al.,

Nucl.Acid Res. 12 (1984) 581-599) and the result was reviewed. ICAM-1 was stepped alongside IgM, N-CAM and MAG using Microgenie and ALIGN (Dayhoff, M.O., et al., Meth. Enzymol. 91, 524-545 (1983)). Computer-aided searches for protein sequence similarities were performed in four protein sequences databases maintained by the National Biomedical Research Foundation (Lipman, DJ, et al., 5 Science 227 (1985)). ) 1435-1439).

Because ICAM-1 is an integrin ligand, its membership in the immunoglobulin supergene family was unexpected. However, examination of the ICAM-1 sequence shows that it meets all the criteria set for belonging to the immunoglobulin supergene group. These criteria, in turn, are discussed below.

The whole extracellular region of ICAM-1 consists of 5 homologous immunoglobulin-like regions shown in parallel in Figure 9A. Regions 1-4 are 88, 97, 99 and 99 residues in length, respectively, and thus represent a typical Ig region size; area 5 is truncated to 68 residues. When performing a computer search on the NBRF database, 20 FASTP programs revealed significant homologies to members of the immunoglobulin supergene family, including the IgM and IgG-C regions, the T-cell receptor alpha subunit variable region, and the alpha-1-beta glycoprotein (Fig. 9B-D).

; Using the above information, the amino acid sequence of ICAM-1 was compared to the amino acid sequences of other members of the immunoglobulin supergene family.

Three types of Ig superfamily regions,, 30V, Cl and C2 are distinguished. Both the V and C regions are composed of two beta-plates which interconnect the intracellular disulfide bond; The V regions contain 9 anti-parallel beta strands, while the C regions contain 7. The constant regions were divided into Cl and C2 according to the 35 typical residues shown in Figure 9A. The Cl group includes proteins related to anti-gene recognition. The C2 group comprises a number of Fc receptors and proteins involved in the binding of sol 74104952, including CD2, LFA-3, MAG and NCAM. ICAM-1 regions were found to be extremely homologous to C2 group regions, indicating ICAM-1 to this group; this is reflected in residues that are more potent in similarity to C2 than Cl regions as shown for beta strands B to F in Figure 9. Similarly, ICAM-1 regions were much better aligned with C2 beta strands A and G than these strands V- and Cl areas along the chest to provide good breast alignment across the 10 dimensions of the C2 region. Breast alignments of C2 regions with NCAM, MAG, and alpha-1 beta glycoprotein are shown in Figures 9B and 9C; similarity ranged from 28% to 33%. Breast alignments with T cell receptor V-alpha, 27% similarity, and IgM-C region-3, 34% similarity, are also shown (Figures 9B, 9D).

One of the most important properties of the immunoglobulin regions is the disulfide-linked cysteines, which bridge the B and F beta strands, which stabilizes the beta sheet to the 20 "sandwich" form; ICAM-1 retains the cysteines in all cases except for the 4 region f, where there is leucine, which may have turned - towards the "sandwich" structure and may possibly stabilize the contact, as shown for some of the other V and C2 upland regions. Distances between cysteines (43, 50, 52 and 37 residues ) are as described for the C2 group.

To investigate the presence of intracellular disulfide bonds in ICAM-1, endothelial cell ICAM-1 was subjected to SDS-PAGE under reducing and non-reducing conditions.

*

Endothelial cell ICAM-1 was used because it shows less glycosylation heterogeneity than JY or hairy cell spleen ICAM-1 and because it is more sensitive to Mr transitions. Thus, ICAM-1 was purified from 16 h LPS (5 35 micrograms / ml) stimulated cultures of umbilical vein endothelial cells by immuno affinity chromatography as described above. Acetone-precipitated ICAM-1 re-104952 75 was slurried in sample buffer (Laemmli, U.K., Nature 227: 680-685 (1970)) using 0.25% 2-mercaptoethanol or 25 mM iodoacetamide and warmed to 100 ° C for 5 minutes. The samples were subjected to SDS-PAGE 4670 run and 5 silver stains 4613. The apparent Mr of endothelial cell ICAM-1 was 100 kd under reducing conditions and 96 kd under non-reducing conditions, indicating strongly the presence of intracellular disulfides in the native ICAM-in.

Using the primary sequence to predict the secondary structure (Chou, PY, et al., Biochem. 13, 211-245 (1974)), 7 predicted beta strands in each ICAM-1 region, identified by a through g in Figure 9A, were found, prediction of the immunoglobulin region and corresponds to the positions of filaments A-H in immunoglobulins (Fig. 9A, bottom). Region 5 lacks A and C filaments, but since these formed the edges of the plates, the plates may still be formed, possibly with filament D positioned at the location of filament C as shown in some other C2 regions; 20 whereby the characteristic disulfide bond between the B and F strands would be retained. Thus, the criteria for region size, sequence homology, persistent cysteines forming the putative intracellular disulfide bond, presence of disulfide bonds, and predicted beta-plate structure are all met with respect to the reading of ICAM-1 in immunoglobulin.

ICAM-1 was found to be most homologous to C2 series glycoproteins NCAM and MAG. This is of particular interest because both NCAM and MAG mediate cell-to-cell attachment. NCAM is important in nerve cell-nerve cell and nerve-muscle interactions (Cunningham, BA, et al., Science 236: 799-806 (1987)), while MAG is important in nerve cell oligodendrocyte and oligodendrocyte interactions during M ., et al., J. Cell Biol. 105 (1987) 1893-1899). Cell surface expression of NCAM and MAG becomes regulated by development 76 104952 during nerve formation and myelination, respectively, analogous to regulated induction of ICAM-1 during inflammation (Springer, TA, et al., Ann.Rev.Immunol. 5 (1987)). 223-252). ICAM-1, NCAM (Cun-5 nigham, BA, et al., Science 236: 799-806 (1987)) and MAG (Salzer, J.L., et al., J. Cell Biol. 101 (1987) 957-965). ) are similar in structure and homology, since each is an integral membrane glycoprotein consisting of five C2 regions that form the extracellular region of the N-terminus, although some extra non-Ig-like sequence is present in NCAM and between the transmembrane region. ICAM-1 sits along its entire length, including transmembrane and cytoplasmic regions, alongside MAG with a 21% similarity; A similar 15% similarity is found when comparing the five domains of ICAM-1 and NCAM-1. Figure 10 is a diagram; na Secondary structures of ICAM-1 and MAG. Comparison of the region by region shows that the degree of homology between the internal regions of the ICAM-1 and NCAM molecules (x +/- sd; 21 20 +/- 2.8% and 18.6 +/- 3.8% respectively) corresponds to the homology level when comparing ICAM-1 regions with NCAM and MAG regions (20.4 +/- 3.7 and 21.9 +/- 2.7, respectively). Although there is evidence that an alternate looping of the C-terminal regions of NCAM occurs (Cunnigham, BA, et al. Science f: 25 216 (1987) 799-806; Barthels, D., et al. , EMBO J. 6: 907-914 (1987), as well as MAG (Lai, C., et al.,

Proc.Natl.Acad.Sei. USA 84: 4377-4341 (1987), no to-1 derivatives have been found here by sequencing endothelial "or HL-60-ICAM-1 clones or by examining the ICAM-1 protein backbone and precursor in various cell types (Dustin , ML, et al., J. Immunol. 137: 245-254 (1986)).

W '^' ICAM-1 acts as a ligand for LFA-1 in lymphocyte interactions! 'In a variety of cell types. Lymphocytes bind to Γ35 ICAM-1, which is internalized in artificial membrane bilayers, and this requires the presence of LFA-1 on lymphocytes, which directly indicates the interaction of LFA-1 with 77 104952 ICAM-1 ( Marlin, SD, et al., Cell 51: 813-819 (1987). LFA-1 is a white blood cell integrin and has no immunoglobulin-like features. White cell tetrins comprise one class of integrins. The other two 5 classes mediate cell matrix interactions and recognize G in their ligands fibronectin, vitronectin, collagen and fibrinogen (Hynes, RO, Cell 48: 549-554 (1987); Ruoslahti, E., et al. Science 238: 491-497 (1987)). leukocytes, and participate in cell-cell interactions, with the only known ligands being ICAM-1 and iC3b, which is a complement-component C3 fragment that does not exhibit any immunoglobulin staa Mac-1 (Kishimoto, T.K. et al., Leukocyte Typing III; McMichael, M. (Ed.), Springer-Verlag, New York (1987); Springer, T.A., et al., Ann.Rev. Immunol. 5: 223-252 (1987); Anderson, D.C., et al., Ann. Rev.Med. 38: 175-194 (1987). Possible peptides belonging to the ICAM-1-20 sequence identified by LFA-1 as derived from sequencing analysis are shown in Table 9.

TABLE 9 Peptides contained in the ICAM-1 sequence possibly identified by LFA-1: '25 -LRGEKEL- -RGEKELKREP- -LRGEKELKREPAVGEPAE- -PRGGS-30 -PGNNRK- -QEDSQPM- -TPERVELAPLPS-RRD-RRD is the first example of an immunoglobulin supergene group member that binds to integrin. Although both of these groups play an important role in cell attachment, finding an interaction between them was unexpected. In contrast, interactions within the immunoglobulin gene super group are quite common. It is quite possible that further examples of interactions between integrin and immunoglobulin groups will be found later. LFA-1 recognizes a non-ICAM-1 ligand (Springer, TA, et al., Ann.Rev.Immunol. 5: 223-252 (1987)) and white blood cell integrin Mac-1 recognizes a neutrophil of a non-C3bi. neutrophil attachment (Anderson, DC, et al., Ann.Rev.Med. 38: 175-194 (1987)). In addition, purified MAG-containing vesicles bind to MAG-containing nerve branching branches, so that the MAG must be able to interact with any particular receptor (Poltorak, M., et al., J. Cell Biol. 105 (1987) 1893-1899). .

The proportion of NCAM in neuronal-neuronal and neuronal-muscle cell interactions has been reported to be due to homophilic NCAM-NCAM interactions (Cunnigham, B.A., et al.,

Science 236: 799-806 (1987)). The important role of MAG in the interactions between Schwann cells surrounding the long branch of nerve cells during myelin sheath formation is possibly due to interaction with a particular receptor, or due to homogeneous MAG-MAG interactions. The homology to NCAM and the frequent occurrence of regional interactions in the immunoglobulin supergene group raise the possibility that ICAM-1 may be involved in homophilic interactions as well as ICAM-1-LFA-1-hetero-J helical interactions. However, the binding of B-early lymphocytes co-expressing similar LFA-1 and ICAM-1 densities to ICAM-1 in artificial phases or cell monolayers can be completely prevented by pretreating 35 B-early lymphocytes with LFA-1-MAb when again, attachment is not affected by pretreatment of B early neurons with ICAM-1-MAb.

f [Pre-treatment with a single layer of ICAM-1-MAb completely destroys the site (Dustin, ML, et al., J. Immunol. 137: 245-254 (1986); Marlin, SD, et al., Cell 51 ( 1987) 813-819). These findings indicate that if any ICAM-1 homophilic interactions are present, they must be less potent than the heterophilic LFA-1 interactions.

The possibility that leukocyte integrins recognize ligands in fundamentally different ways fits into the presence of a sequence of 180 residues in their alpha-subunits, which may be of importance for ligand binding and not present in RGD-recognizing integrins, A. (Corbi, et al., EMBO J. 6: 4023-4028 (1987). Although Mac-1 has been shown to recognize the RGD sequence present in iC3b-5086, ICAM-1 does not contain the RGD sequence (Figure 8). This is consistent with the inability of the fibronectin peptide GRGDSP and the control peptide GRGESP to inhibit ICAM-1-LFA-1 attachment (Marlin, S.D., et al. Cell 51: 813-819 (1987)). However, related sequences such as PRGGS and RGEKE are present in ICAM-1 in regions predicted to form a loop between the beta strands a and b of strand 2 and between strands c and d of strand 2 respectively (Fig. 9) and are thus available for identification. Interestingly, the homologous MAG molecule contains an RGD sequence between regions 1 and 2 (Poltorak, M., et al., 25 J.Cell.Biol. 105 (1987) 1893-1899; Salzer, JL, et al. , Cell Cell Biol. 104: 957-965 (1987).

EXAMPLE 19

Southern and Northern blots 30 9

Southern blots were performed using 5 micrograms of genomic DNA extracted from three cell lines: BL2, Burkitt's lymphoma cell line (gift from Dr. Gilbert Lenoir); JY and Er-LCL, which are EBV-transformed 35 B early cell lines.

104952 80 DNAs were digested with BamHI and EcoRI recommended by the 5X manufacturer (New England Biolabs). The digestion products were subjected to electrophoresis on a 0.8% agarose gel, after which the DNAs were transferred to nylonkal-5 butter (Zeta Probe, BioRad). The filter was pre-hybridized and then hybridized following standard procedures using ICAM cDNA derived from HL-60 and labeled with alpha (32P) dXTPs using random priming (Boehringer Mannheim). Northern blots were performed using 20 micrograms of total RNA or 6 micrograms of poly (A) + - RNA. The RNA was denatured and subjected to electrophoresis run on a 1% agarose-formaldehyde gel, after which the products were electronically transferred to a Zeta Probe membrane. Filters were pre-hybridized and then hyb-15 ligated as previously described (Staunton, DE, et al., EMBO J. 6. (1987) 3695-3701) using an HL-60 cDNA probe comprising 32P-labeled oligonucleotide probes (described above). ).

20 Southern blots using a 3 bp cDNA probe and genomic DNA digested with BamHI and EcoRI showed a single major hybridizing fragment of 20 and 8 bp respectively, which refers to a single gene and that most of the code information • 25 is contained in an 8-ep piece. Blot counts of three different cell lines revealed no evidence of restriction fragment polymorphism.

EXAMPLE 20 Expression of the ICAM-1 gene An "expression vector" is a vector which (due to the presence of appropriate transcription and / or translation control sequences) is capable of expressing a DNA or (cDNA) molecule cloned into a vector and thereby producing a polypeptide or prot . Cloned sequences are expressed when the expression vector is introduced into a suitable host cell. If a prokaryotic expression vector is used, any suitable prokaryotic cell capable of expressing the cloned sequences would be a suitable host cell. Similarly, if a eukaryotic expression vector is used, any suitable eukaryotic cell capable of expressing the cloned sequences would be a suitable host cell. Importantly, because eukaryotic DNA may contain intervening sequences and because prokaryotic cells are unable to properly process such sequences, cDNA derived from a cell capable of expressing ICAM-1 is preferably used to form a prokaryotic genomic expression vector library. . Procedures for preparing cDNA and producing a genomic library are disclosed by Maniatis, T., et al. (Molecular Cloning - A Laboratory Manual. Cold Spring Harbor Press, 15 Cold Spring Harbor, NY, 1982).

The expression vector genome library described above is used to form a host cell bank (where each cell contains one member of the library). The expression vector can be introduced into 20 host cells in any number of different ways (i.e., transformation, transfection, protoplast fusion, electroporation, etc.). The library of expression vector-containing cells is amplified by cloning and its members are individually determined (using immunoassay) to determine whether they produce a protein capable of binding to anti-ICAM-1 antibody.

The expression vectors of the cells that produce the protein capable of binding to the anti-ICAM-1 antibody are then further analyzed to determine whether they express (and thus contain said) the ICAM-1 gene as a whole, and thus contain) only the ICÄM-1 gene fragment, or do they express (and thus contain) a gene whose product, although immunologically related to ICAM-1, is not ICAM-1. While such analysis can be performed in any convenient manner, the nucleotide sequence of the DNA or 10495282 cDNA fragment cloned into the expression vector is preferably determined. These nucleotide sequences are then examined to determine if they are capable of encoding polypeptides having the same amino acid sequence as the trypsin cleavage fragments of ICAM-1 (Table 5).

An expression vector containing a DNA or cDNA molecule encoding the ICAM-1 gene can thus be identified by (i) the ability to direct expression of a protein capable of binding to an anti-ICAM-1 antibody; and (ii) the presence of a nucleotide sequence capable of encoding each I ICAM-1 trypsin fragment. A DNA molecule cloned into such an expression vector can be removed from the expression vector and isolated in pure form.

15 EXAMPLE 21 (Functional activities of purified ICAM-1

In cells, ICAM-1 normally functions as a surface protein associated with the cell membrane. Therefore, the function of purified ICAM-1 was examined after reconstitution of the molecule with artificial lipid membranes (liposomes I or vesicles) by dissolving the protein in liposoluble detergent, followed by removal of the detergent. 25 dialysis. ICAM-1, purified from JY cells and eluted with detergent-octylglucoside as described above, was reconstituted in the vesicles, and the vesicles containing ICAM-1 were fused in glass coverslips or plastic culture wells to allow protein binding of the cells.

Manufacture of planar membranes and plastic-bound blisters

Blisters were prepared by the method of Gay et al. (J. Immunol. 136: 352026 (1986)). Briefly, egg phosphatidylcholine and cholesterol were dissolved in chloroform and mixed in a 7: 2 molar ratio. The lipid mixture was dried to a thin film 104952 83 by rotation in a stream of nitrogen gas and then freeze-dried for at least one hour to remove residual chloroform.

The lipid film was then dissolved in a solution of 1% octyl glucoside / 0.14 M NaCl / 20 mM Tris (pH 7.2) to a final concentration of 0.1 mM phosphatidylcholine. About 10 micrograms of purified ICAM-1 or human glycophorin (Sigma Chemical Co., St. Louis, MO) as control membrane glycoprotein was added per 10 ml of dissolved lipids. The protein-lipid detergent solution was then dialyzed at 4 ° C using 3 turns and 200 volumes of 20 mM Tris / 0.14 M NaCl, pH 7.2 and one change of HBSS.

Plane membranes were prepared by Brian et al., Proc.Natl. Acad. USA 81: 6159 (1984).

The glass top plates (11 mm diameter) were boiled for 15 minutes in a 1: 6 dilution of 7X detergent solution (Linbro), then washed overnight in distilled water and then soaked in 70% ethanol solution and air dried. An 80 microliter drop of blister solution containing either ICAM-1 or glycophorin was placed at the bottom of the wells in a 24-well cluster plate, and the prepared glass slides were carefully placed over them. After incubation for about 20 to 30 minutes at room temperature, the wells were filled with HBSS and the top plates turned, whereupon the planar membrane came up. The wells were then washed with copious amounts of HBSS to remove non-bound vesicles.

The planar membrane surface was cautious of exposure to air.

•

In experiments on glass membrane-fused flat membranes, ICAM-1-containing vesicles were found to bind directly to the plastic surface of multi-well tissue culture plates and retain their functional activity as demonstrated by specific cell binding. Such blisters are hereinafter referred to as "plastic-bound 104952 84 blisters" (PBV) because the nature of the plastic-bound lipid vesicles has not been elucidated. Plastic-bound vesicles were prepared by adding 30 microliters of the vesicle suspension directly to the bottom of the wells with 96-well tissue culture-5 plates (Falcon), after which the plates were incubated and the wells were washed as described for flat membranes.

Cell Fixation Assays 10 Cell fixation assays using plasma membranes or plastic-bound vesicles were performed essentially in the same manner, except that cell numbers and volumes in the PBV assays were 1/5 of those used in the plasma membrane assays.

15 T lymphocytes from normal controls and a patient suffering from a white cell attachment deficiency (LAD) and thus unable to express LFA-1 (Anderson, DC, et al., J.Infect.Dis. 152: 668 (1985)) were prepared by culturing peripheral blood. mononuclear cells at 1 microgram / ml concanavallin A (Con-A) in RPMI 1640 medium supplemented with 20% FCS at 5 x 10 / ml for 3 days. The cells were then washed twice with RPMI and once with 5 mM methyl alpha-D-mannopyranoside solution to remove lectin residue from the cell surface. Cells were grown in RPMI / 20% FCS containing 1 ng / ml recombinant IL-2 and used on day 10-22 of culture initiation.

For the detection of cell binding to plasma membranes or PBV, 30 Con-A parental cells, T-lymphoma SKW3 cells, and EBV transformed; a number of early B-cell line JY cells (LFA-1 positive) and LFA-1-deficient early cell line (BBN) cells (obtained from patient 1, Springer, TA, et al., J.Exoer, Med. 160 (1986) 1901-1918) were radiolabeled by incubating 1 x 10 7 cells per ml with RPMI.

1640/10% FCS with 100 microcurie units of Na 5 Cr 2 O 4 for 1 hour at 37 ° C, then washed 4 times with RPMI

8 5 104952 1640 medium to remove non-internalized stamp. In monoclonal antibody assays, cells or plastic-bound vesicles were pretreated with 20 micrograms / ml purified antibody in RPMI-1640/10% FCS at 4 ° C for 30 minutes, followed by washing 4 times to remove unbound antibody. In experiments that examined the effects of divalent cations on cell binding, cells were washed only once with HBSS containing no Ca 2+, Mg 2+ ions but containing 10% dialyzed FCS, and 10 CaCl and MgCl were added to the labeled concentrations.

In all experiments, cells and platelets or PBV were pre-equilibrated at a suitable temperature (4 ° C, 22 ° C, or 37 ° C) in a suitable assay buffer).

15 To measure cell binding to purified ICAM-1, 5Cr-labeled cells (2 x 105 EBV transformants in plaque membrane assays; 1 x 105 EBV transformants or SKW3 cells, 2 x 155 Con-A parent cells in PBV assays) - was centrifuged at 25 xg for 2 minutes in flat membranes 20 or PBV, followed by incubation at 4 ° C, 22 ° C or 37 ° C for 1 hour. After incubation, unbound cells were removed by performing 8 rounds of loading and aspirating buffer equilibrated to the appropriate temperature. Bound cells were quantitated by dissolving the wells in 0.1 N NaOH / 1% Triton X-100 and counting in a gamma counter. Percent cell binding was determined by dividing the scintillation / min from the bound cells by the cell-associated scintillation / min value used. Scintillations / min 30 used in flat membrane experiments were corrected for the ratio of the surface area of the top plates to the area of the culture wells.

In these assays, EBV-transformed B-early lymphocytes, SKW3-T lymphoma cells, and Con-A-T lymphocytes bound specifically to ICAM-1 on artificial membranes (Figures 11 and 12). The binding was specific because the cells were very poorly bound to control level membranes or 8 6 104952 vesicles containing equivalent amounts of another human cell surface glycoprotein, glycophorin. In addition, LFA-1-positive EBV transformants and Con-A parent cells bound, whereas their LFA-1-negative counterparts did not bind significantly, indicating that the binding was dependent on the presence of LFA-1 in the cells.

Both cell specificity and dependence on cell LFA-1 were confirmed in blocking experiments with monoclonal antibodies (Figure 13). JY cell binding was prevented by 97% when ICAM-1 containing PBVs were pretreated with anti-ICAM-1 monoclonal antibody RR1 / 1. Pretreatment of cells with the same antibody had little effect. In contrast, anti-LFA-1-15 monoclonal antibody TS1 / 18 inhibited binding by 96%, but only when cells, not PBV, were pretreated. The control antibody T2 / 9, which reacted with LFA-3 (another lymphocyte surface antigen), had no significant inhibitory effect when either cells or PBV 20 were pretreated. This experiment demonstrates that, in artificial membranes, ICAM-1 itself, rather than a minor impurity, mediates the observed cell attachment and that attachment is dependent on the presence of LFA-1 on the binding cell.

The binding of cells to ICAM-1 by artificial membranes also demonstrated two other features of the LFA-1-dependent attachment system: temperature dependence and the need for divalent cations. As shown in Figure 14, Con-A parental cells bound to ICAM-1 in PBVs very efficiently at 37 ° C, partially at 22 ° C, and very poorly at 4 ° C.

* I

· As shown in Figure 15, the binding was completely dependent on the presence of divalent cations. At physiological concentrations, Mg 2+ alone was sufficient for maximal cell binding, whereas Ca 2+ alone produced very low binding levels. However, maximal binding was obtained with Mg 2+ at 1/10 of the normal concentration when combined with Ca 2+ having a synergistic effect.

In summary, the specificity, specific inhibition and temperature dependence of monoclonal antibodies on purified ICAM-1 incorporated into artificial membranes, as well as dependence on divalent cations indicate that ICAM-1 is an LFA-1-dependent ligand binding system.

10 EXAMPLE 22 Expression of ICAM-1 and HLA-DR in Hypersensitivity and Toxic Patch Test Reactions 15 Skin samples of five normal individuals were examined for their ICAM-1 and HLA-DR expression. It was found that while endothelial cells normally expressed ICAM-1 in some blood vessels, ICAM-1 was not expressed by normal skin-derived keratinocytes. HLA-DR staining was not detected with any of the keratinocytes derived from normal skin samples. The kinetics of the expression of ICAM-1 and class II anti-genes in cells obtained from hypersensitivity or toxicity lesions were then examined. Half of the six individuals examined were found to have cer-25 tinocytes expressing ICAM-1, four hours after hapten application (Table 10). The percentage of individuals expressing ICAM-1 with their keratinocytes increased as a function of the exposure time to hapten, thereby also increasing the staining intensity per collinocyte, indicating an increase in ICAM-1 expression - up to 48 hours. In fact, at this time, a certain percentage of keratinocytes in all tissue samples were positively stained with ICAM-1. At 72 hours (24 hours after hapten removal), 7 out of 35 individuals still expressed ICAM-1 on their keratinocytes, however, the expression of ICAM-1 in one individual disappeared between 48 and 72 hours.

es 104952 TABLE 10 Kinetics of induction of ICAM-1 and HLA-DR by keratinocytes from hypersensitivity patch samples 5

Time Glass- Biopter Plants ICAM-1 &

on the amount of ICAM-1 HLA-DR HL-DR

10 Normal Skin 5000

Patch Allergy Test 4 6 3a 0 0 8 9 3 0 0 15 24 8 7 0 0 48b 8 5 0 3 72 8601 aSamples were considered positive if at least a small 20 keratinocyte clusters had been stained.

bAll patches were removed at this time.

Histopathologically, the ICAM-1 staining pattern on keratinocytes from tissue samples taken four hours after hapten application was usually in small clusters. At 48 hours, ICAM-1 was expressed on the majority of keratinocytes on the surface, with no difference between the mid and peripheral regions of the lesion. Staining in-; The intensity decreased with keratinocytes approaching the stratum corneum. This was found in tissue samples from both central and peripheral regions of the lesion. Similarly, at this time the patch test was positive (bleeding, redness and blisters). No difference in 35 expression of ICAM-1 was observed when different hapten were used in susceptible individuals. In addition to ceramic 891 0 4 9 5 2, ICAM-1 was also expressed on some mononuclear cells and endothelial cells at the lesion site.

HLA-DR expression on keratinocytes in hypersensitivity skin lesions was less common than that of ICAM-1. None of the individuals studied had lesions in which keratinocytes were positively stained for HLA-DR until 24 hours after the use of the hapten. In fact, only four tissue samples contained keratinocytes expressing HLA-DR, and none of the tissue samples had keratinocytes positive for HLA-DR but not ICAM-1 (Table 10).

In contrast to the hypersensitivity patch test lesion, the toxicant patch 15 lesion induced by croton oil or sodium lauryl sulfate showed keratinocytes that showed little if any ICAM-1 on their surfaces at all times tested (Table 11). In fact, 48 hours after application of the patch, which time point-20 was optimal for ICAM-1 expression in hypersensitivity-patch subjects, only one of the 14 toxin patch subjects had ICAM-1 expressing keratinocytes in their lesions. In contrast to hypersensitivity patch specimens, the keratinocytes did not: 25 express any HLA-DR in the toxic patch test lesions.

These results indicate that ICAM-1 is expressed in immune-based inflammation but not in toxic-based inflammation, and thus ICAM-1 expression can be used to distinguish between immune-based and toxic-based inflammation in function of, e.g., renal renal failure, it is difficult to determine whether the impairment is due to rejection or to the renal toxicity of the therapeutic agent that weakens the immune response.

104952 90

By taking a tissue sample and determining the up-regulation of ICAM-1 expression, an immune-based rejection and a non-immune-toxic reaction can be distinguished.

5 TABLE 11 Kinetics of ICAM-1 and HLA-DR Induction by Keratinocytes from Toxic Patch Specimens 1 10

Time for patch application - Biopi 's on pigs ICAM-1 &

(h) the amount of ICAM-1 HLA-DR HL-DR

15 4 4 0 0 0 8 3 Sat 0 0 24 3100 48b 14 1 0 0 72 3101 20 - aSamples were considered positive if at least small keratinocyte clusters had been discolored.

bAll patches were removed at this time.

25 EXAMPLE 23 ICAM-1 and HLA-DR Expression in Benign Skin Diseases 30 Cells taken from skin lesion samples of patients with different types of skin inflammatory diseases were examined for their ICAM-1 and HLA-DR expression. Certain keratinocytes in samples from hypersensitivity contact dermatitis, vesicle-like rash / vesicle-35 dermatitis, and cutaneous lichen dermatitis expressed ICAM-1. Skin red lichen samples showed severe discoloration, which was similar or even more pronounced than that observed with 48-hour hypersensitivity patch samples (Table 12). Consistent with the results obtained in the hypersensitivity-patch test, the most intense ICAM-1 staining was observed at sites that had undergone heavy mononuclear cell bleeding. In addition, 8 of the 11 red lichen samples tested positive for HLA-DR expression on keratinocytes.

ICAM-1 expression on keratinocytes obtained from tissue samples from patients with inflammatory dermatitis or urticaria was less evident. Only four of the 7 patients studied who had either of these conditions had keratinocytes that expressed ICAM-1 at the site of injury. HLA-DR expression was observed in only one patient and occurred with ICAM-1.

Endothelial cells and part of the mononuclear cell bleed all of the studied benign skin inflammatory diseases expressed ICAM-1 to varying degrees.

TABLE 12 Expression of ICAM-1 and HLA-DR by keratinocytes derived from benign skin diseases

Cases Only ICAM-1 & Diagnosis Rate ICAM-1 HLA-DR HL-DR

Allergic 30 Touch Schema 5 3a 0 2 • <

Lichen Planus 11 308%

Pemphigoid / 2,200 35 Pemphigus

Exanthema 3,200

Utricaria 4 101 40 104952 92 aSamples were considered positive if at least small keratinocyte clusters were discolored.

EXAMPLE 24 The expression of ICAM-1 on keratinocytes derived from dermatitis skin lesions of ICAM-1 on skin samples from five patients suffering from dandruff was investigated before and periodically during treatment with PUVA-10.

Samples were taken from 5 patients with tissue-confirmed conventional dandruff. Samples were taken in series before and during PUVA treatment at the indicated time. PUVA was given 15 to 3-4 times a week.

Five patients were sampled from the periphery of the dandruff centers, and four of these patients were clinically normal skin.

20

The received skin tissue samples were frozen and stored in liquid nitrogen. Six micron cryostat slices were air-dried overnight at room temperature, then fixed in acetone for 10 minutes and then either: immediately stained or wrapped in aluminum foil and stored at -80 ° C until staining.

Staining was performed as follows. The slices were incubated with the monoclonal antibody and stained by the three-step immunoperoxidase method (Stein, H., et al., Adv.Cancer Res. 42: 67-147 (1984)) using a diaminobenzidine-hydrogen peroxide substrate. . Tissues and lymph nodes were used as positive controls for ICAM-1 and HLA-DR staining. Tissue samples stained in the absence of 35 primary antibodies served as negative controls.

104952 93 Monoclonal antibodies against HLA-DR were purchased from Becton Dickinson (Mountainview, California). The monoclonal anti-ICAM-1 antibody was R6-5-D6. Peroxidase-conjugated rabbit anti-mouse Ig and peroxy-5 dase-conjugated porcine anti-rabbit Ig were purchased from Dakapatts manufacturer, Copenhagen, Denmark. Diaminobenzidine tetrahydrochloride was purchased from Sigma (St. Louis, MO).

The results of the study show that endothelial cells in some blood vessels express ICAM-1 in both diseased and normal skin, but the intensity of staining and the number of blood vessels expressing ICAM-1 had increased in dermatitis skin lesions. In addition, the expression pattern of ICAM-1 in 15 keratinocytes from untreated dementia skin lesions in five patients varied only from very small cell clusters to many stained keratinocytes.

During treatment with PUVA, ICAM-1 expression in two patients (patients 2 and 3) showed a significant reduction that preceded or occurred at the time of clinical recovery (Table 13). Patients 1, 4 and 5 had decreases and elevations in ICAM-1 expression during PUVA treatment, which correlated with: 2525 clinical recovery and respiratory disease, respectively. ICAM-1 expression was not present on normal skin keratinocytes before or after PUVA treatment. This indicates that PUVA does not induce ICAM-1 in normal skin keratinocytes.

30

Of note was the observation that the density of mononuclear cell bleeding correlated with the amount of ICAM-1 expression on keratinocytes. This included both a decrease in the number of mononuclear cells in lesions during PUVA treatment, whereby ICAM-1 expression was also decreased, and an increase in the number of mononuclear cells during PUVA treatment, where the expression of ICAM-1 in keratinocytes was more pronounced. Endothelial cells and skin mononuclear cells were also ICAM-104952 94 1 positive. In clinically normal skin, expression of ICAM-1 was restricted to endothelial cells whereby keratinocytes were not labeled.

The expression of HLA-DR on keratinocytes varied. No HLA-DR positive tissue samples were found that were also not ICAM-1 positive.

Taken together, these results indicate that pre-treatment ICAM-1 expression is evident on keratinocytes and correlates with dense mononuclear cell leakage. During PUVA treatment, a marked decrease in ICAM-1 staining is observed with clinical healing. Histopathologically, skin bleeding I was also reduced. When clinical re-morbidity was evident during treatment, ICAM-1 expression on keratinocytes increased, as did When clinical recovery was observed during treatment, there was a concurrent decrease in keratinocyte ICAM-1 staining as well as a decrease in skin bleeding.

Thus, expression of ICAM-1 on keratinocytes correlates with the density of skin mononuclear cells. These results indicate that Γ as a clinical response to PUVA treatment, there was a clear decrease in ICAM-1 expression on keratinocytes, with a more moderate reduction in the number of mononuclear cells. This demonstrates that the expression of ICAM-1 on keratinocytes is responsible for initiating and maintaining skin leakage, and that PUVA treatment down-regulates ICAM-1, which in turn decreases skin leakage and inflammatory response. The results also show that HLA-DR expression on keratinocytes varied during PUVA treatment.

ICAM-1 expression on keratinocytes in lesions associated with dandruff disease correlated with the clinical severity of the lesion and the extent of skin leakage. Thus, ICAM-1 plays a central role in dandruff, whereby its expression and / or interaction with mononuclear cells of the CD18 complex will inhibit the effective treatment of this disease. In addition, monitoring of ICAM-1 expression on keratinocytes will provide an effective tool for the diagnosis and prognosis of dementia and for the assessment of the progress of therapy. 5 TABLE 13 ICAM-1 expression by keratinocytes in dandruff lesions (PS) and clinically normal skin (N) as a function of time 10 before and during PUVA treatment

Time before Treatment of Patient No. PUVA and its _} ___ ^ __ ^ ____ 5

15 after p $ ps N PS N PS N PS N

0 ++ - ++ - ++ - +++ 1 day + 1 week + + - - ++ - + - o 20 2 weeks ++ + - + - + - 3 weeks ++ * o 4 weeks ++ + --- ++ * * 25 5-6 weeks - - - o 7 weeks (++) (+) +++ * * 10 weeks (+) 30 +++ Lots of positive keratinocytes ++ Moderately positive keratinocytes + Locally positive keratinocytes (+) Distributed only very few positive keratinocytes · · 35 charges • <

No positive staining * Clinical recovery o Clinical relapse 96 1 0 4 9 5 2 EXAMPLE 25 Expression of ICÄM-1 and HLA-DR in Malignant Skin Disorders 5 (Table 14). Of the 23 skin T cell lymphomas examined, ICAM-1 positive keratinocytes were found in only 14 cases. Kera-10 tinosytes derived from mycosis fungoides lesion samples tended to lose their ICAM-1 expression in the advanced stages of the disease. However, ICAM-1 expression was found in varying proportions of mononuclear cell bleeding from most skin T-cell lymphoma lesions. Of the remaining 15 lymphomas examined, four of the eight had ICAM-1 expressing keratinocytes. Of the 29 patients studied with malignant skin diseases, five had keratinocytes expressing HLA-DR without expressing ICAM-1 (Table 14).

20 TABLE 14 Expression of ICAM-1 and HLA-DR by Keratinocytes from Malignant Skin Diseases 25

Cases Only ICAM- &

Diagnosis rate of ICAM-1 HLA-DR HLA-DR

CTCL, MFI 8 Sat 0 4 30 CTCL, MFII-III 10 1 2 5 CTCL, SS 3102 • · 'CTCL, Large Cell 2020 35 CBCL 2 0 0 1

Skin leukemia 3111 40 Histocytosis Xl 0 0 0 104952 97 aSamples were considered positive if at least small keratinocyte clusters had been stained.

EXAMPLE 26 5 Effect of anti-ICAM-1 antibodies on proliferation of human peripheral blood mononuclear cells

Human peripheral blood mononuclear cells are induced to proliferate by the presence and recognition of antigen or mitogen. Certain molecules, such as the mitogen concanavallin A, or the T-cell binding antibody OKT3, cause non-specific proliferation of peripheral blood mononuclear cells.

Human peripheral blood mononuclear cells are heterogeneous in that they consist of subpopulations of cells that are capable of recognizing specific antigens. When a peripheral blood mononuclear cell capable of recognizing a specific antigen encounters that antigen, an increase in the corresponding mononuclear cell subpopulation is induced. Rigid 20 cross-toxin and keyhole ventricle hemocyanin are examples of antigens that are recognized by peripheral blood mononuclear cell subpopulations but not recognized by all peripheral blood mononuclear cells in acclimated individuals.

25

The ability of the anti-ICAM-1 monoclonal antibody R6-5-D6 to inhibit human peripheral blood mononuclear cell proliferation responses in systems known to require cell-to-cell attachments was examined.

• Peripheral blood mononuclear cells were purified using Ficoll-Paque (Pharmacia) gradients according to the manufacturer's instructions. The cut-off phase was harvested and the cells were washed three times with RPMI 1640 medium, then cultured in flat-bottom 96-well microtiter plates at 106 cells / ml in RPMI 1640 medium supplemented with 10 8% bovine serum, with 2 mM glutamine and gentamicin (50 micrograms / ml).

Antigen, either T cell mitogen, Concanavallin A (0.25 µg / ml); T-cell binding antibody, OKT3 (0.001 µg / ml) / keyhole vial haemocyanin (10 g / ml) or tetanus toxin (1: 100 dilution from source) was added to cells cultured as described above either in ICAM anti-R6-5-D6; final concentration 10 g / ml). Cells were cultured for 3.5 days (concanavallin A assay), '2.5 days (OKT3 assay), or 5.5 days (keyhole-well-ventricle-haemocyanin and tetanus toxin assays), after the tests were completed.

For 18 hours before the end of the experiment, 2.5 microcurie units of 3 H-thymidine were added to the cultures. Cell division was determined by measuring thymidine transfer to the DNA of peripheral blood mononuclear cells. The displaced thymidine was collected and counted in a liquid scintillation counter (Merluzzi et al., J. Immunol. 139, 166-168 (1987)). The results of these experiments are shown in Fig. 16 (Concanavallin A test), Fig. 17 (OKT3 assay), Fig. 18 (Keyhole vase compartment haemocyanin assay) and Fig. 19 (tetanus toxin assay).

25

It was found that the anti-ICAM-1 antibody inhibited reproductive responses to the non-specific T-cell mitogen, Con-A; non-specific T cell-associated antigen, OKT-3; and for specific antigens, keyhole vase haemo-cyanine and tetanus toxin; mononuclear cells. Inhibition with anti-ICAM-1 antibody was comparable to inhibition with anti-LFA-1, suggesting that ICAM-1 is a functional ligand for LFA-1 and that ICAM-1 antagonism inhibits the specific defense system. 35 responses.

104952 99 EXAMPLE 27 Effect of Anti-ICAM-1 Antibody on Mixed Cell Reaction 5 As discussed above, ICAM-1 is essential for efficient cellular interactions in immune responses mediated by LFA-1-dependent cell attachment. Induction of ICAM-1 in connection with immune responses or in inflammatory disease allows white blood cells to interact with each other and endo-10 thymus cells.

When lymphocytes from two non-related individuals are cultured in the presence of each other, parental cell formation and cell division of the lymphocytes is observed. This response to the response of one lymphocyte population to the presence of another lymphocyte population - known as the Mixed Cell Reaction (MLR) - is analogous to the response of lymphocytes to insertion of mitogen. . 453-458).

20

Experiments were performed to determine the effect of anti-ICAM monoclonal antibodies on human MLR. These experiments were performed as follows. Peripheral blood was taken from normal healthy donors from a venous puncture -... 25 la. Blood was collected in heparin-treated test tubes where it was diluted 1: 1 at room temperature with Puck's G - (Gibco) Balanced Saline (BSS). The blood mixture (20 ml) was layered in 15 ml Ficoll / Hypaque density gradient (Pharmacia, density = 1.078, room temperature 30) and centrifuged at 1000 x g for 20 minutes.

·· The cut-off phase was then recovered and washed with 3X Puck's G

little potential. Cells were counted in a hemacytometer and resuspended in RPMI 1640 medium (Gibco) containing 0.5% gentamycin, 1 mM L-glutamine 35 (Gibco) and 5% heat-inactivated (56 ° C, 30 min.) Human AB. serum (Flow Laboratories) (hereinafter referred to as "RPMI medium").

104952 100 A murine anti-ICAM-1 antibody (R6-5-D6) was used in these experiments. All monoclonal antibodies (prepared from intraperitoneal fluid at Jackson Immuno Research Laboratories, Boston, MA) were used as purified IgG preparations.

Peripheral blood mononuclear cells (PBMCs) were cultured in medium at 6.25 x 105 cells / ml in Linbro round-bottom microtiter plates (# 76-013-05). Stimulator cells from another donor 10 were irradiated at 1000R and then cultured at the same concentration as the response cells. The total volume of the culture was 0.2 ml. In control! response cells alone and stimulator cells alone. The culture plates were incubated at 37 ° C in a humidified atmosphere containing 5% CO 2 for 5 days. The wells were pulsed with 0.5 microcurie units of tritiated thymidine (3HT) (New England Nuclear) for the last 18 hours of culture. In some cases, a bidirectional MLR was performed. The order of the procedure was the same except that cells from another donor were not inactivated by irradiation.

Cells were harvested on glass fiber filters using an automated sample collector (Skatron, Norway), then rinsed with water and methanol. The filters were oven dried and then counted in Aquasol liquid on a Beckman (LS-3801) liquid scintillation counter. The results are expressed as mean twilight / min +/- s.d. sixth; from different cultures.

30

• I

Table 15 shows that purified anti-ICAM-1 monoclonal antibodies inhibited MLR in a dose-dependent manner, with significant suppression at 20 ng / ml. Purified mouse IgG had! ' 35 little or no suppressive effect.

MLR suppression with anti-ICAM-1 monoclonal 104952101 occurs when the antibody is added during the first 24 hours of culture (Table 16).

TABLE 15 5 Effect of Anti-ICÄM-1 Antibody on One-Way Lymphoid Reaction

Reactive, n 1 2HT incorporation

10 Watts Stimulates- J

cells1 thorium cellsb Antibodyec (CPH) - - - 445d ± 143 - + - 148 ± 17 + - - 698 ± 72 15 + + - 42,626 ± 1.579 + + mlgG (10.0 pg) 36.882 ± 1.823 (14%) + + mlgG (0.4 pg) 35,500 ± 1.383 (17%) + + mlgG (0.02 pg) 42.815 ± 1.246 (0%) + + R6-5-D6 (10.0 pg) 8.250 ± 520 (81%) ) 20 + + R6-5-D6 (0.4 pg) 16.142 ± 858 (62%) + + R6-5-D6 (0.03 pg) 28.844 ± 1.780 (32%)

Response Cells 6.25 x 10 3 / ml bStimulator Cells 6.25 x 10 3 / ml irradiated with 1000 25 R • · 2 Purified Monoclonal Anti-cICAM-1 (R6-5-D6) or Purified Mouse IgG (mlgG), final concentrations (micrograms / ml).

3 dAverage +/- s.d. 5-6 cultures; the numbers in parentheses il-30 indicate the percentage inhibition of MLR.

• <102 TABLE 16 104952 Time of addition of anti-ICAM-1 antibody

5 Ra Sb Additions0 3HT Subscription (CPM

Medium or Antibody Addition Time Day 0 Day 1 Day 2 Tray 205d ± 14,476 ± 132,247 ± 75 - + Tray 189 ± 16 nd® nd] ^ q + _ Tray 1, 850 ± 615 nd nd ++ Tray 41,06312 , 940 45,95512,947 50,94313,072 R6-5-D6 17,78111,293 38,40911,681 47,30812,089 (57%) (16%) (7%) a Response Cell 6.25 x 105 pc / ml 15 bStimulator cells 6.25 x 105 pc / ml irradiated with 1000 R medium or purified monoclonal anti-ICAM-1. a list of antibody (R6-5-D6) at a concentration of 10 micrograms / ml was added from day 0 at 24 hour intervals.

20 mean +/- s.d. 4-6 cultures; "end = not set

Percentage Inhibition ^ ♦ ··· In summary, the ability of an anti-ICAM-1 antibody to inhibit 25 MLR indicates that anti-ICAM-1 monoclonal antibodies have therapeutic use in acute graft rejection. Monoclonal antibodies to ICAM-1 are also associated with therapeutic use. immune (system)-mediated disorders that are dependent on LFA-1 / ICAM-1 regulated cell-cell interactions.

The experiments described herein show that the addition of anti-ICAM-1 monoclonal antibodies inhibited the mixed-35 cell reaction (MLR) when added during the first 104952103 of the reaction. In addition, ICAM-1 is upregulated in human peripheral blood monocytes in vitro culture.

In addition, ICAM-1 was found not to be expressed on dormant human peripheral blood lymphocytes or monocytes. ICAM-1 is upregulated in monocytes cultured alone and in cells co-cultured with unrelated donor cells in a mixed cell reaction, whereby measurements are performed using conventional flow cytometry assays. This upregulation of ICAM-1 by monocytes can be used as an indicator of inflammation; especially if ICAM-1 is expressed on fresh monocytes from individuals with acute or chronic inflammation.

The specificity of ICAM-1 on activated monocytes and the ability of an anti-ICAM-1 antibody to inhibit MLR suggest that anti-ICAM-1 monoclonal antibodies have diagnostic and therapeutic potential for acute transplant rejection and associated immune responses (systemic 20). system) in mediated disorders that require cell-to-cell interactions.

EXAMPLE 28 Synergistic Effects of Collaboration of ICAM-1 and LFA-1 Antibodies 25

As shown in Example 27, MLR is inhibited by anti-ICAM-1 antibody. MLR may also be inhibited by an anti-LFA-1 antibody. To determine whether co-administration of anti-ICAM-1 and anti-LFA-1 antibodies had a potentiating or synergistic effect, an MLR assay (as described in Example 27) was performed in the presence of varying concentrations of these two antibodies.

35 This MLR assay showed that the combination of anti-ICAM-1 antibody and anti-LFA-1 antibody - at concentrations at which neither antibody alone specifically inhibited MLR - was significantly more effective in MLR. response (Table 17). This result demonstrates that therapies in addition to administration of anti-ICAM-1 antibody (or fragments thereof) and 5 anti-LFA-1 antibody (or fragments thereof) have the capacity to provide more effective anti-inflammatory therapy. Using such enhanced therapy, lower doses of the antibody may be administered than would otherwise be therapeutically effective, which is relevant under conditions where high concentrations of the individual antibodies induce an anti-idiotype response.

Table 17 Effect of varying doses of anti-ICAM-1 and anti-LFA-1 (R3.1) on the mixed cell reaction

Birth control-%

Concentration (pg / ml) 20 Anti-ICAM-1 (R6-5-D6)

Anti-LFA-1 0 .004 .02 .1 .5 2.5 0.0 0 7 31 54 69 70 0.0008 1 7 28 48 62 71 0.004 0 13 30 50 64 72 25 0, 02 29 38 64 75 84 86 0, 1 92.5 90 91 92 92 92 0, 5 93 90 90 92 93 91 • · • · 104952 105 EXAMPLE 29

Additive Effects on MLR of Combined Administration of Optimal Subcutaneous Anti-ICAM-1 Doses and Other Immune Response Agents 5

As shown in Example 28, MLR is inhibited by combinations of anti-ICAM-1 and anti-LFA-1 antibodies. To determine whether co-administration of anti-ICAM-1 and other immunosuppressive agents (such as dexamethasone-10, azathioprine, cyclosporin A and steroids (such as prednisone, etc.)) would also have potentiated effects, MLR assays were performed. at sub-optimal concentrations (i.e., concentrations below the optimal concentration at which the substance alone would be administered to the subject) R6-5-D6 in combination with other immunosuppressive agents, in which case the procedure was as in Example 27.

The results show that the inhibitory effects of R6-5-D6 are at least additive to the inhibitory effects of suboptimal doses of dexamethasone (Table 18), azathioprine (Table 19) and cyclosporin A (Table 20). This suggests that anti-ICAM-1 antibodies may have an effect in lowering the necessary doses of known immuno-suppressive agents, thereby reducing their toxic side effects. Using an anti-ICAM-1 antibody (or fragment thereof) to achieve such an immune response attenuation, it is possible to combine administration of the antibody (or fragment thereof) with either one additional immune response suppressant.

II

.. or a combination of more than one immunosuppressant.

106 TABLE 18 104952

Effect of anti-ICAM-1 and dexamethasone on human MLR 5 3HT

Inhibitor subscriptions- Inhibi-

Group (ng / ml) nasal (CPM) thio%

Medium - 15 6 - 10 Stimulators (S) - 101

Counter Producers (R) - 4,461 R x S - 34,199 R x S R6-5-D6 (8) 26,225 23 R X S Dex (50) 14,158 59 15 R x S R6-5-D6 (8) + Dex (50) 7,759 77

Dex: Dexamethasone TABLE 19 20

Effect of Anti-ICAM-1 and Azathioprine on Human Ani-MLR 3HT Inhibitor Additions-

Group (ng / ml) nasal (CPM) thio%

Media - 78 -

Stimulators (S) - 174

Responsive Producers (R) - 3,419 - .. 30 R x S - 49,570 - RXS R6-5-D6 (8) 44,374 11 R x S Azathioprine (1) 42,710 14 RXS R6-5-D6 (8) + Azathi-34,246 31 prine (1) iti & i 107 TABLE 20 104952

Effect of anti-ICAM-1 and cyclosporin A on human MLR 5 3HT

Inhibitor subscriptions- Inhibi-

Group (ng / ml) nasal (CPM) thio%

Medium - 87 - 10 Stimulators (S) - 206

Counter Producers (R) - 987 R x S - 31,640 R X S R6-5-D6 (8) 26,282 17 R x S CyA (10) 23, 617 25 15 R X S R6-5-D6 (8) + CyA (10) 19,204 39

CyA: cyclosporin A

EXAMPLE 30 Effect of Anti-ICAM-1 Anti-ICAM-1 Anti-ICAM-1 Anti-ICAM-1 Anti-ICAM-1 Anti-ICAM-1 Anti-ICAM-1

Cynomolgus monkeys were transplanted from another specimen of the same species according to the method described by Cosimi et al. (Transplant.Proc. 13: 499-503 (1981)) with the modification that valium and ketamine were used as anesthetics.

• ·

• I

Thus, kidney transplantation was performed essentially as follows. Heterotrophic kidney transplants were performed on Cynomolgus monkeys weighing 3-5 kg, essentially as in Figure 35 nut Marquet (Marquet et al., Medical Primatoloov, Part II,

Basel, Karger, 1972, p. 125), when the monkeys were first anesthetized with 104952 108 valium and ketamine. The donor kidney veins were made with junctions where the end of one structure was connected to the other side, aortic or hollow vein, using a 7-0 Prolene stitch. Donor ureter 5 was spatulated and implanted in the bladder according to an extravesical approach (Taguchi, Y., et al., Dausset et al. (Eds.): Advances in Transplantation, Baltimore, Williams & Wilkins, 1968, p. 393) Renal function was evaluated weekly or twice per week by determining serum creatinine. In addition, graft specimens were collected for histological examination and complete autopsies were performed on all recipients who died in the test. Most recipients underwent bilateral renal transplantation at the time of transplantation, and the time of subsequent 15 deaths due to urinary poisoning was considered as the end of transplantation survival. Some recipients underwent unilateral native kidney removal and bilateral urinary bladder binding at the time of transplantation. When graft rejection occurred, autologous ureteral suture was removed, resulting in recovery of normal renal function and the possibility of continued immunological monitoring of the recipient animal.

Monoclonal antibody R6-5-D6 was administered daily ... 25 for 12 days starting two days prior to transplantation at a dose of 1-2 mg / kg / day. Serum creatinine levels were periodically determined to monitor rejection. The effect of anti-ICAM-1 antibody on immune rejection of kidney from the same species is shown in Table 30 21.

109 TABLE 21 104952 Activity of R6-5-D6 in Transplantation of a Kidney from a Species of Specimens 5 When Used in Cynomolgus Monkey3

Survival Date R6-5-D6 Dose After Treatment with Monkey (mg / kg) 10 Control 1-8

Control 2 11

Control 3-11

Control 4 10

Control 5 - 9 15 Control 6 - 10 M15 1.0 20

Ml 9 1.0 T

M17 1.0 30 20 M25 1.5 29 M2 3 1.0 llc M27 2.0 34 M7 0.5 22 MH 0.5 26 25 Ml 0 0, 5 22 M8 0.5__26 * aApinos were given R6-5 -D6 for 12 consecutive days starting 2 days prior to shipment.

30 bCause of death unknown; evidence of latent malaria was found.

cDeath cause Kidney necrosis.

35 dThe animal was still alive on 15 August 1988.

104952 110

The results show that R6-5-D6 effectively prolonged the life span of monkeys transplanted from another individual of the same species.

5 EXAMPLE 31 Effect of Anti-ICÄM-1 Antibody on Acute Suppression of Transplanted Organ

To demonstrate that the anti-ICAM-1 antibody is telo-potent in the acute transplant rejection model, R6-5-D6 was also tested in the therapeutic or acute renal rejection model. In this model, monkey kidneys were transplanted (using the protocol described in Example 30) and given 15 mg / kg of 15 cyclosporin A (CyA) i.m. on both sides of the procedure. until stable renal function was achieved. The CyA dose was then reduced twice weekly at 2.5 mg / kg until clotting occurred, as evidenced by an increase in blood creatinine levels. R6-5-D6 was then administered for 10 days and survival time was monitored. It is important to note that in this protocol, the dose of CyA remains suboptimal because it does not change after the acute sweating phase has begun. In this model, tissue-matched controls (N = 5) without antibody protection survive the sweating phase for 5-14 days. 25 resignations. To date, six animals have been tested using R6-5-D6 according to this protocol (Table 22). Two of these animals are still alive (M12, 31 days; and M5, 47 days based on R6-5-D6 dose). Two animals, 38 and 55 days after initiation of R6-5-D6 therapy, respectively, and two animals died due to causes other than acute rejection (one animal died from CyA toxicity and the other died when anesthetized with R6-5-D6). O). This model more closely resembles the (previous) clinical situation where R6-5-D6 was originally administered online.

111 TABLE 22 104952 Activity of R6-5-D6 in prolonging survival of a transplanted kidney from another specimen of the same species in a Cynomolgus monkey3

Selviämispäivä

Monkey Sweat step on day b after treatment

Control0 14-98 5-14 10 M24 41 38 M21 34 4d M3 41 55 M9 12 15 M12 37> 31f M5 2 6> 47f AaPin were given 1-2 mg / kg R6-5-D6 for 10 consecutive days after the onset of rejection calculated.

20 bDay when creatinine levels increased as a result of the reduction in CyA dose and when R6-5-D6 treatment was started. c Five animals were tested using the above therapeutic protocol except that no protective therapy was used. "Survival Days After Treatment" is the number of survival days calculated from the onset of the rise in creatinine levels.

dThe animal died of anesthesia; creatinine levels were low. The animal died from CyA toxicity; creatinine levels were low.

30 The animal was still alive on 15 August 1988.

104952 EXAMPLE 32, Genetic Construction and Expression of Truncated ICAM-1 Derivatives In its native state, ICAM-1 is a cell membrane-bound protein that contains an extracellular region, a transmembrane region, and a cytoplasmic region consisting of five immunoglobulin-like regions. We wanted to construct functional derivatives of ICAM-1 Ξ lacking the transmembrane region and / or the cytoplasmic region to obtain a soluble, secreted form of ICAM-1. These functional derivatives were constructed by oligonucleotide-directed mutation of the ICAM-1 gene, followed by expression of the mutant gene by transfection into monkey cells.

ICAM-1 gene mutants comprising amino acid substitutions and / or truncated derivatives were generated according to the method of Kunkel, T., (Proc.Natl.Acad.Sei. USA 82: 488-492 (1985)). ICAM-1 cDNA was prepared as described above and digested with restriction endonucleases

Sa-11 and Kpn1, followed by the resulting 1.8 bp DNA fragment subcloned into the plasmid vector CDM8 (Seed, B., et al., Proc.Natl.Acad.Sei. USA M (1987) 3365-3369). This construct, designated pCD1.8C, was then transformed with 25 E. coli duf'unq 'strains. One transformant was recovered from the transformants. j internal uracil-containing template using infection with helper phage R408 - (StratageneR). Mutant ICAM-1 cDNA strands were then generated by initializing the synthesis of the second strand with an oligonucleotide containing incompatible base pairs and then transforming the unq + host * (MC1061 / P3) with the resulting heteroduplex. The mutants were isolated by screening for newly created endonuclease restriction centers introduced by the mutant oligonucleotide. Mutant ICAM-1 protein was expressed by transfecting Cos-35 cells with mutant DNA in eukaryotic expression vector CDM8 using DEAE-dextran constant protocols Selden, R.F., et al. , Current Protocols in Molecular Biology inia! ΓΓ: ΐ31 104952 113

Ausubel, F.M., et al., Eds., Ss. 9.2.1.-9.2.6. (1987).

A truncated functional derivative of ICAM-1 lacking the transmembrane and cytoplasmic regions but containing all five extracellular regions of immunoglobulin-like regions was prepared. Using a 30 bp mutant oligonucleotide (CTC TCC CCC CGG TTC TAG ATT GTC ATC ATC), the codons for the tyrosine (Y) and glutamic acid (E) amino acids at positions 452 and 453 were changed to phenyl-10 alanine (F) and the translation stop codon ( TAG). The mutant was isolated by its single Xba I restriction site and was designated "Y452E / F, TAG".

To express the mutant protein, COS cells were transfected with three mutant clones ("2, # 7, and # 8). Three days after transfection with these three mutant clones, culture supernatants and cell lysate were analyzed by immunoprecipitation with 1 SDR / ICAM-1 was precipitated from the culture supernatants of cells transfected with mu-20 mutant subclones # 2 and # 8, but not from the detergents of these cells, and the molecular weight of ICAM-1 detected in the culture supernatant was about 6 kd lower than that of ICAM. The membrane form of 1, which fits the mutant DNA with the expected size. This functional derivative of ICAM-1 is secreted as a soluble protein. In contrast, ICAM-1 was not immunoprecipitated with native ICAM-1 transfected control cell culture supernatants, indicating that the membrane form of ICAM-1 is not secreted from Cos cells, and that ICAM-1 was not fake transfected cells, which acted as inactive controls, in both culture supernatants and cell lysates.

Truncated ICAM-1 secreted by transfected cells was purified by immunoaffinity chromatography with ICAM-1 specific antibody (R6-5-D6) and subsequently tested for functional activity in a cell binding assay 104952,114. Preparations containing native ICAM-1 or truncated excreted form were purified in the presence of detergent octylglucoside, then diluted to a final concentration of 0.25% octylglucoside (5% below the critical micelle concentration of detergent). These ICAM-1 preparations were allowed to bind to the well surfaces of plastic 96-well well plates (Nunc) to obtain ICAM-1 bound to the solid phase. The non-bound material was washed away, after which it was found that about 75-80% and 83-88%, respectively, of SKW-3 cells, which contained LFA-1 on their cell pins, bound specifically to ICAM-1. in native and truncated form. These results indicate that the secreted, truncated, and soluble functional derivative of ICAM-1 retained both immunological reactivity and the ability to mediate ICAM-1-dependent attachment, characteristics of which are required for ICAM-1.

A functional ICAM-1 derivative lacking only the cyt-20 toplasma region was prepared by similar methods. A 25 bp oligonucleotide (TC AGC ACG TAC) was used to convert amino acid 476 (Y) to a TAG translation stop codon.

CTC TAG AAC CGC CA). The mutant was named "Y476 / TAG". Immuno-I by precipitation and SDS-PAGE on Cos cells I.15 transfected with the mutant was found in the membrane of ICAM-1; a bound form with a molecular weight of about 3 kd lower than native ICAM-1. Indirect immunofluorescence of Cos cells transfected with the mutant was found to be the same; spotted staining pattern than native ICAM-1 expressed on LPS-stimulated human endothelial cells. Finally, cells transfected with mutant DNA specifically bound to purified LFA-1 on plastic surfaces in the same manner as Cos cells transfected with native ICAM-1 DNA (Table 23).

35 TABLE 23 115 104952 Ability of ICAM-1 or ICAM-1 Functional Expression Cells to bind to LFA-1 5% of LFA-1 binding cells expressing ICAM-1 in the presence of: 10 TRANSFECTING No antibody RR1 / 1

Invalid 0 0

Native ICAM-1 20 0 Y476 / TAG 20 0 15-EXAMPLE 33 Mapping Functional Regions of ICAM-1 Examination of 20 ICAM-1 revealed that the molecule comprises 7 regions. Five of these regions are extracellular (with region 1 closest to the cell surface and region 1 farthest from the cell surface), one region is the transmembrane region, and another region is the cytoplasmic region (i.e., within the cell). Epitope mapping studies can be used to determine which regions contribute to the ability of ICAM-1 to bind to LFA-1. To carry out such experiments, various deletion mutants are prepared which are characterized by their ability to bind to LFA-1. Alternatively, assays may be performed using an anti-ICAM antibody known to interfere with the ability of ICAM-1 to bind to LFA-1. Examples of such suitable antibodies include RR1 / 1 (Rothlein, R., et al., J. Immunol. 137: 1270-1274 (1986)); 35 R6.5 (Springer, TA, et al., U.S. Patent Publication Serial No. 07/250, 446), LB-2 (Clark, EA, et al.) Leukocyte Typing I, A. Bernard, et al. ed., Springer-Verlag, 1984, pp. 339 -.

104952 116 346) or CL203 Staunton, D.E., et al., Cell 56 (1989) 849-853).

ICAM-1 deletion mutants can be generated in any of a number of different ways. Preferably, however, such mutants are generated by site-directed mutation or otherwise by recombinant gene engineering (such as by constructing gene sequences expressing ICAM-1, from which sequences encoding certain protein regions have been eliminated. Procedures well known in the art for producing such mutants are well known in the art. three ICAM-1 deletion mutants were prepared, the first mutant missing amino acid residues F185 -P284 (i.e., deletion of region 3), the second mutant missing amino acid residues P284 - R451 (i.e. deletion of regions 4 and 5), the third mutant missing Y476. (i.e., cytoplasmic region deletion) The results of these studies indicate that regions 1, 2 and 3 are mainly involved in the interactions between ICAM-1 and anti-ICAM-1 antibody or LFA-1.

EXAMPLE 34 Effect of Mutations in ICAM-1 on LFA-1 Binding ·: The ability of ICAM-1 to interact and bind to LFA-1 is mediated by ICAM-1 amino acid residues located in regions 1 of the ICAM-1 molecule. 8, 9 and 10). However, amino acids located in ICAM-1 regions 30 and 2 contribute to such interactions. Thus, preferred functional derivatives of the present invention include soluble ICAM-1 molecule fragments containing ICAM-1 regions 1, 2 and 3. More preferred are soluble ICAM-1 molecule fragments containing ICAM-1-35 regions 1 and 2 Most preferred are soluble ICAM-1 molecule fragments containing multiple amino acid residues from the first ICAM-1 region involved in the interaction between ICAM-1 region 1. ICAM-1 and LFA-1. Replacement of these amino acids with other amino acids alters the ability of ICAM-1 to bind to LFA-1. These amino acid residues and said substitutions are shown in Figure 25. Figure 25 also shows the effects of such mutations on the ability of the resulting mutant ICAM-1 to bind to LFA-1. In Figures 23-25, the residues are designated by the amino acid single-letter code followed by the position of the residue in the ICAM-1 molecule. Thus, for example, "E90" denotes glutamic acid-10 residue at position 90 of ICAM-1. Similarly, "E90V" denotes a dipeptide consisting of glutamic acid residue at position 90 and an intermediate residue at position 91. The substituent sequence is indicated by a right ("/"). Residues V4, R13, Q27, Q58 and D60S61 of ICAM-1 are involved in 15 LFA-1 binding.

Substitution of these amino acids altered the ability of ICAM-1 to bind to LFA-1. For example, replacement of V4 with G yields a mutant ICAM-1 molecule which is less capable of binding to LFA-1 (Figure 25). Replacing the R13 residue of ICAM-1 with E yields a mutant molecule with significantly less ability to bind to LFA-1 (Figure 25). Replacing the Q58 residue H of ICAM-1 with H yields a mutant molecule with essentially normal ability to bind to LFA-1 (Figure 25).

. Replacing D60S residues of ICAM-1 with KL yields a mutant thymic molecule with substantially less ability to bind to LFA-1 (Figure 25).

Secondary glycosylation centers are also involved in LFA-1 binding (Figure 23). Substitution of N103 for K, or A155N for SV, results in a mutant ICAM-1 molecule that is substantially unable to bind to LFA-1. In contrast, replacement of the glycosylation site N175 by A did not appear to have a significant effect on the ability of mutant ICAM-1 to bind to LFA-1.

104952 118

Mutations in the third ICAM-1 region did not significantly alter the LFA-1 binding of ICAM-1 (Figure 24).

5 EXAMPLE 35 ICAM-1 Multimers with Increased Biological Half-Affinity and "Survival" 10

Chimeric molecules are constructed in which regions 1 and 2 of ICAM-1 are attached to the hinge region of the immunoglobulin heavy chain. In preferred embodiments, the C-terminus of the ICAM-1 region 2 is attached to an immunoglobulin heavy chain gene segment 15 immediately adjacent to the hinge region at the N-terminus, whereby the elasticity of the segment can be transferred to the hinge region. ICAM-1 regions 1 and 2 thus replace the Fab fragment of the antibody. By attaching the structure to the IgG-class heavy chains and then the animal cells produce a 20-chimeric molecule. The production of molecules containing heavy chains derived from IgA or IgM results in higher molecular weight molecules containing 2-12 ICAM-1 molecules. By co-expressing the J chain gene in animal cells that produce the chimeric ICAM-1 heavy chain molecules, IgA and IgM multimers are assembled correctly, yielding mainly IgA molecules containing 4-6 ICAM-1 molecules and IgM- molecules containing about 10 ICAM-1 molecules. These chimeric molecules may have several advantages. First, the Ig molecules are designed to be long-lived in the bloodstream, potentially increasing the biological half-life.

In addition, the multimeric nature of these constructed molecules allows them to interact with a higher affinity for the rhinitis virus and cell surface LFA-1, depending on the therapeutic context, whereby the amount of recombinant protein required to obtain an effective dose is greatly reduced. IgA and IgM are highly glycosylated and molecules normally found in mucosal secretions, for example in the nose.

5

Their highly hydrophilic nature contributes to the retention of the bacteria and viruses they bind to in the mucosa, thereby preventing the latter from attaching to the cells and preventing the passage of the epithelial cell membrane barrier. Thus, their therapeutic efficacy may have increased. IgM, and in particular IgA, are stable in mucosal environments and can increase the stability of ICAM-1 structures. If such a functional ICAM-1 derivative is administered to the bloodstream, it may also increase the biological half-life. IgA 15 does not adhere to complement and would thus be ideal for applications where this is fatal. If IgG-H chain chimeras are desired, the regions involved in complement attachment and Fc receptor interactions could be mutated.

EXAMPLE 36 Formation of ICAM-1 Mutants

Oligonucleotide cut mutation The coding region of the 25 ICAM-1 cDNA in a 1.8 bp SaiI-Kpn1 fragment was subcloned into the expression vector CDMB (Seed, B., et al., Proc.Natl.Acad.Sci. USA 84 (1987) 3365- 3369). Kunkelin, T., (Proc.Natl.Acad.Sei. USA 82: 488-492 (1985)), and Staunton, D., et al., (Staunton, D.E., et al.).

Cell 52 (1988) 925-933), this structure (pCD1.8) was used to generate a single-stranded uracil-containing template for use in oligonucleotide-directed mutation.

35

Easily, pCD18.8 transformed E. coli strain XS127. Individual colonies were grown in 1 ml of Luria broth 104952 120 (LB) medium (Difco) containing 13 micrograms / ml ampicillin and 8 micrograms / ml tetracycline close to the saturation point. 100 microliters of the culture was inoculated with R408 helper phage (Stratagene) as multiplication (10) of 5 (MOI) followed by addition of 10 ml of LB medium containing ampicillin and tetracycline and grown for 16 hours at 37 ° C. The culture was centrifuged at 10,000 rpm for 1 minute, after which the supernatant was filtered as a 0.22 micro-m filter and 10 phage suspensions were inoculated with E. coli strain BW313 / P3, which was then plated on LB-agar medium (Difco). ) on cup pins with growth medium supplemented with ampicillin and tetracycline. The colonies were picked and grown in 1 ml of LB medium containing ampicillin-15 and tetracycline close to the saturation point, which; afterwards, the cultures were infected with an helper phage as an MOI 10 infection. The culture volume was then increased to 250 ml and the cells were cultured overnight. The single-stranded DNA was isolated by standard phage extraction.

20

Mutant oligonucleotides were phosphorylated and used in addition to the pCD18 template in the second strand synthesis reaction ((Staunton, D., et al. Cell 52: 925-933 (1988))).

·; Transfection COS cells were inoculated into 10 cm tissue culture dishes; in an amount such that the culture had grown to 50% within 16 to 24 hours. COS cells were then washed once with '30 TBS and incubated for 4 hours in 4 ml RPMI containing 9% 10% serum (Collaborative), 5 micrograms / ml chloroquine, 3 micrograms mutant plasmid and 200 micrograms DEAE. dextran sulfate. The cells were then washed with 10% DMSO / PBS followed by PBS, followed by culturing for 16 hours in culture medium. The medium was replaced with fresh medium and at 48 hours post infection, COS cells were slurried with trypsin / EDTA-104952121 (Gibco) treatment and plated in 2, 10 cm dishes and 24-well tissue culture plates for HRV binding. At 72 hours, cells were harvested from 10 cm dishes using 5 mM EDTA / HBSS and treated with attachment to LFA-1 coated plastic and immunofluorescence assay.

Binding of LFA-1 to HRV 10 LFA-1 was purified from SKW-3 lysates by immunoaffinity chromatography on TS2 / 4-LFA-1-MAb-Sepharose gel and eluted at pH 11.5 with 2 mM MgCl2 and In the presence of 1% octylglucoside. LFA-1 (10 micrograms / 200 microlitres / 6 cm dish) was bound to bacteriological petri dishes by diluting with 0.1% octyl glucoside in PBS (phosphate buffered saline) containing 2 mM MgCl 2 and incubating overnight at 4 ° C. C. Plates were blocked with 1% BSA (bovine serum albumin) and stored in PBS containing 2 mM MgCl 2, 0.2% BSA, 0.025% azide and 50 micrograms / ml gentamicin.

5Cr-labeled COS cells in PBS containing 5% FCS (fetal calf serum), 2mM MgCl2, 0.025% azide (buffer: 25 µl) were incubated with the addition of 5 micrograms / ml RR1 / L and R6 .5, or without making said addition, on LFA-1-coated microtiter plates at 25 ° C for 1 hour. Non-adherent cells were removed by washing 3 times with buffer. The adherent cells were eluted by adding EDTA to a concentration of 10 mM and gamma counting was performed.

SCORE

Antibodies to ICAM-1 such as RR1 / 1, 35 R6.5, LB-2 or CL203 have been identified. If these antibodies are capable of blocking the ICAM-1 function, they may necessarily be able to bind to a particular site in the ICAM-1 molecule, which is also important for the ICAM-1 function. Thus, by making the ICAM-1 deletion mutants described above and determining the extent to which the anti-ICAM-1 antibodies are capable of binding to the deletion, it can be determined whether the deleted regions are important for the function.

ICAM-1 is an integral membrane protein whose extracellular region is predicted to consist of five Ig-like C regions. To identify regions involved in LFA-1 binding, region 3 and regions 4 and 5 (carboxyl terminus) were deleted using oligonucleotide-directed mutation, followed by expression in COS cells and functional assays. In addition, the entire cytoplasmic region was deleted to verify its potential effect on ICAM-1 interactions. As the predicted cytoplasmic region deletion, Y476 / *, showed no loss of RR1 / 1, R6.5, LB-2, and CL203 reactivity, whereas the deletion of region 3, F185 to R451, resulted in a corresponding decrease in CL203 reactivity ( Figure 20). Thus, the CL203-20 epitope appears to be located in region 4, while RR1 / 1, R6.5 and LB-2 appear to be located in the 2 amino-terminal region.

All 3 deletion mutants show wild-type LFA-1 attachment levels (Figure 21). Amino acid substitutions ποιον. 25 were also introduced into the predicted beta inversions of regions 1, 2 and 3, the structures were expressed in COS cells and tested for functionality. The R6.5 epitope was thus located in the sequence E111GGA in region 2, and may also be associated with E39 in region 1, while RR1 / 1 and LB-2 are each dependent on R13 in region 1 (Figure 22).

«

In addition, mutations in the sequence D71GQS reduce RR1 / 1 binding. As a result of mutations that eliminate N-linked glycosylation centers in N103 and N165, the LFA-1-HRV binding of RR1 / 1, R6.5 and LB-2 is decreased. These mutations appear to affect processing to form ICAM-1 dimers.

104952 123

Other mutations in regions 2 or 3 did not result in a change in LFA-1 attachment (Figures 23 and 24). The amino acids R13 and D60 of region 1 each participate in LFA-1 binding (Figure 25).

Thus, binding of LFA-1 and HRV appears to be a function of the amino terminus of the IgAM-1 region of ICAM-1. Figure 26 shows the alignment of the ICAM amino-terminal regions.

Although the invention has been described in connection with specific embodiments thereof, it should be noted that other modifications may be made thereto, the object of which is to cover all modifications, uses or applications of the invention which generally follow the principles of the invention and depart from this disclosure. is within the scope of the known or conventional practice in the field of the invention and to which the essential features set forth hereinabove can be applied within the scope of the appended claims.

20

Claims (3)

A process for preparing a soluble, functionalite derivative of ICAM-1 (intercellular adhesion molecule), which differs from ICAM-1 by the absence of at least transmembrane area (Figure 8) for natural ICAM-1, which derivative is soluble under physiological conditions, characterized in that a host organism transformed with an ICAM-I soluble functional derivative encoding recombinant vector is cultivated, which derivative contains at least one converted amino acid group, or is a chemical derivative thereof, in a deleted ICAM-1 molecule, under conditions in which said derivative is expressed and secreted in the culture medium, after which the derivative is isolated. 15 Method according to claim 1, characterized in that the functional derivatives of ICAM-1 lack the cytoplasmic region of ICAM-1. 20 Method according to claim 1 or 2, characterized in that the functional derivatives of ICAM-1 lack the amino acid groups after the region 1 (Figure 25) or the regions 1 and 2 (Figure 23) or the regions 1, 2 and 3 (Figure 24). ). »· I