IE83907B1 - Inhibition of lymphocyte adherence to vascular endothelium utilizing a novel extracellular matrix receptor-ligand interaction - Google Patents

Description

INHIBITION or LYMPHOCYTE ADHERENCE TO VASCULAR ENDOTHELIUM UTILIZING A NOVEL EXTRACELLULAR MATRIX RECEPTOR-LIGAND INTERACTION . INTRODUCTION The present invention relates to the use of antibodies binding to the d4Blreceptor and inhibiting the adhesion of one cell to another for the preparation of a pharmaceutical composition for use in mammals to suppress an immune response. It is based on the discovery that the d4Bl extracellular matrix receptor promotes adhesion of lymphocytes to endothelial cells Via attachment to a defined peptide sequence. In the embodiments of the invention, monoclonal antibodies may be used to inhibit binding of lymphocytes to endothelial cells, thereby preventing lymphocyte entrance into tissue and suppressing the immune response. 2. BACKGROUND OF THE INVENTION ‘these techniques. Using monoclonal antibodies, Wayner and class II was involved in cell adhesion only to native of extracellular matrix receptors. The ECMRS are members Jof the integrin family of cell adhesion molecules and Antisera to the VLA—B subunit'were found to block cell radhersion to fibronectin or laminin (Takada et al., 1987, .Nature 326: 607-610). Antibodies to VLA4 have been shown to 'lists some of the members of the integrin family described supra, and Table II lists a number of monoclonal antibodies that recognize various ECMRS.

Anti — ECMR Antibodies ______________._______ (Wayner et a1., 1987 J. Cell .2. FIBRONECTIN Fibronectin is a protein found in the ).

Cellular and plasma fibronectin heterodimers comprise similar but not identical polypeptides. The variability in the structure of fibronectin subunits derives from variations in fibronectin mRNA primary sequence due to alternative splicing in at least 2 regions of the pre-fibronectin mRNA (the ED and IIICS regions).

The binding of fibronectin to cell surfaces may be competitively inhibited by fragments of fibronectin (Akiyama et al., 1985, J. Biol. Chem. 260:13256-13260).

Using synthetic peptides, a sequence of what was thought to be the only minimal cell-recognition site was identified as rapidly on a 75kDa fragment representing the RGDS containing cell-binding domain, whereas B16-F10 melanoma cells did not appear to spread on the 75kDa fragment, but, instead were observed to spread on a 113kDa fragment sequences have not been identified. ?be extremely important in the immune system; in which the localization of immune mediator cells is likely to be due, at least in part, to adhesive interactions between cells.

Recirculation of lymphoid cells is non-random (Male et al., in "Advanced Immunology", J. B. Lippincatt Co., Philadelphia, p. 14.4 — 14.5); lymphocytes demonstrate a preference for the type of secondary lymphoid organ that they will enter. In trafficing through a secondary lymphoid organ, lymphocytes must first bind to the vascular endothelium in the appropriate post-capillary venules, then open up the tight junctions between endothelial cells, and finally migrate into the underlying tissue. recirculating lymphocytes from blood into specific lymphoid tissues, called homing, has been associated with complementary adhesion molecules on the surface of the lymphocytes and on the endothelial cells of the high endothelial venules. -_.- According to one study, antibodies to a murine lymphocyte recepter ,/LPAM—l, inhibited lymphocyte leming to Peyer's patch HEV/Holzmann et al., 1989 cell S6:37—46, Likewise, the adherence of polymorphonuclear leukocytes to vascular endothelium is believed to be a key Migration of In addition, studies with specific anti- glycoprotein antibodies in patients with immune deficits indicated that one or more components of the CD18 complex are required for effective neutrophil chemotaxis and other The physiologic importance of leukocyte adhesion T cells appear to be proteins (supra) is underscored by the existence of a human‘ genetic disease, leukocyte adhesion deficiency (LAD: Anderson et al., 1985, J. Infect. Dis. l52:668; Arnaout et ; :668; Arnaout et al., 1985, Fed. Proc. gg:2664).

ECMRS have also been observed to be associated with functions outside of the immune system. Loss of the region of fibronectin.g Furthermore, we have recently shown that neurite formation on laminin or fibronectin can be inhibited by antibodies to ECMRS.

. SUMMARY or THE INvENTfoN The present invention relates to the use of an antibody or fragment or derivative thereof, which binds to the 0481 recepter and which inhibits the adherence of nucleated lematopoietic cells to vascular endothelial cells for the preparation of a pharmaceutical composition for use in a mammal to suppress an immune response. The antibody or fragment or derivative thereof interferes with the interaction between the extracellular matrix recepter and its ligand.

The invention is based upon the discovery that the a4fll extracellular matrix receptor promotes adhesion of lymphocytes to endothelial cells via attachment to a defined peptide sequence. Prior to the present invention, the ligand of the a4fi1 receptor had not been identified, nor had the function of the a4p1 receptor in lymphocyte attachment been known. By preventing the interaction between the a4fll receptor and its ligands using antibodies» the present invention enables, for the first time, specific intervention in the migration of lymphocytes through the vascular endothelium and into tissues. The present invention, therefore, has particular clinical utility in suppression of the immune response; in various specific embodiments of the invention, the adherence of lymphocytes to endothelium may be inhibited systemically, or may, alternatively, be localized to particular tissues or circumscribed areas. Accordingly, the present invention provides for treatment of diseases involving autoimmune responses as well as other chronic or relapsing activations of the immune system, including allergy, asthma, and chronic inflammatory skin conditions. .1. ABBREVIATIONS Peptide sequences defined herein are represented by the one-letter symbols for amino acid residues as follows: A (alanine), R (arginine), N (asparagine), D (aspartic acid), C (cysteine), Q (glutamine), E (glutamic acid), G (glycine), H (histidine), I (isoleucine), L (leucine), K (lysine), M (methionine), F (phenylalanine), P (proline), S (serine), T (threonine),'W"?tryptophan), Y (tyrosine), V (valine).

. DESCRIPTION OF THE FIGURES Figure 1. Adhesion of T lymphocytes (Molt 4), KS62—l, RD or HTl080 cells to plasma fibronectin, inhibition with PlD6 monoclonal antibody and cell surface expression of a5fil. 51Cr-labeled cells (105 cells/ml) were incubated with PlD6 monoclonal antibody (50 pg/ml) for 60 minutes at 4° C and allowed to attach to fibronectin-coated (20 pg/ml) plastic surfaces in the presence of PlD6 (solid bars) or mouse IgG (open bars) for 30 min (HT1080 or RD) or 4 hr (Molt 4 or K562) at 37' C. Adhesion to plasma fibronectin (pFN) is cr cpm bound to the plastic surfaces. Cell expressed as surface expression of aSfil was determined by flow cytometry by staining of cells in suspension with PlD6 monoclonal antibody. Log PlD6 fluorescence (striped bars) is expressed as mean channel number (0-255) above background.

Figure 2. Immune precipitation of lymphocyte fibronectin receptor from HT10890, Molt 4 or chronically activated CD8+ T (LAK) cell detergent extracts. gin the absence of serum for 1 hour at 37' C. 1251-labeled Molt 4, LAK or HT1080 cells were extracted with 1% Triton X-100 in the presence of phenylmethyl sulfonyl fluoride (1 mM), N—ethylma1eimide (1 mM), leupeptin (1 pg/ml) and diisopropyl fluorophosphate (1 mM)_as protease inhibitors. Aliquots of these extracts were immune precipitated with monoclonal antibodies directed to .1351 (P185), azpi (P1H5) and (1451 (P3E30. precipitated antigens were run on 7.5% SQS-PAGE gels in the The three bands immune precipitated with P3E3 from T lymphocytes are The immune absence of 2—ME and visualized by autoradiography. indicated (arrows).

Figure 3. Identification of lymphocyte specific fibronectin receptor as Integrin a4fi1V:V¢ 1251-surface labeled Jurkat cells were extracted with 0.3% CHAPS in the presence of 1 mM CaCl2, 1 mM diisopropyl—fluorophosphate, 1 mM phenylmethyl sulfonyl fluoride, 1mM N—ethylmaleimide, 1 pg/ml leupetin and 2 pg/ml soybean trypsin inhibitor. Aliquots of the extracts were then immune precipitated with myeloma (SP2) culture supernatant or with monoclonal antibodies P3E3, P4C2, P4G9 or with PID6 (anti-a5fi1).

SDS-PAGE gels in the absence of reducing agent and visualized The immune precipitates were run on 8% by autoradiography. Molecular weight markers are shown on the left—hand side. are the bands present in immune precipitates prepared with P323, P4C2 and P4G9 (arrows).

Localization of a4fi1 and asfll in focal adhesions on fibronectin-coated surfaces.

The a5 and 51 subunits are indicated as Figure 4.

RD cells were trypsinized and allowed to adhere to silanized and fibronectin—coated (20 pg/ml) glass cover slips At the end of this time the cells were prepared for localization of receptors in focal adhesions as described (Experimental Procedures). Panels A and C show focal adhesions (arrows) visualized by interference reflexion microscopy when RD cells r-v-I \./‘ N V1 are adhered to fibronectin. Panel B shows the reorganization of the RGD prototype fibronectin receptor a55l stained with antibody AB33 to the focal adhesions (arrows). Panel D shows the reorganization of a4fll stained with P4G9 (FITC) also to the focal adhesions when RD cells are adhered to fibronectin (arrows). Panels A and B are the same field and Panels C and D are the same field.

Figure 5A. Domain structure of human plasma fibronectin (pFN) showing the origin of the fragments used in this study. B. SDS-PAGE gel analysis (10% acrylamide) demonstrating the purity of the fragments.

The 80 kDa fragment had the N—terminal amino acid sequence SD()VPSPR()LQF, and therefore begins at position 874 of the fibronectin molecule (Kornblihtt et al., 1985, EMBO J. 4:17S5-1759). This fragment contains the cell binding domain (Cell) and the RGDS sequence of fibronectin (*). The 58 kDa and 38 kDa fragments had the N-terminal amino acid sequence TAGPDQTEMTIEGLQ. Both fragments contain the C-terminal Heparin binding domain (Hep II) and result from a different cleavage of the two fibronectin chains by trypsin. The 38 kDa fragment comprises the first 67 amino acid residues of the alternatively spliced connecting segment of fibronectin (IIICS) (Garcia-Pardo, 1987, Biochem. J., 241:923—928) and it is therefore derived from the A chain. The 38 kDa fragment does not contain the REDV adhesion site recognized by B16-F10 melanoma cells (Humphries et al., 1986, sgpgg: Humphries et al., 1987, ggpga). The 58 kDa fragment is also derived from the 8 chain of fibronectin and lacks the IIICS region (Garcia-Pardo, et al., 1989, EMBO J., submitted). The 58 kDa fragment also contains the C-terminal fibrin binding domain of fibronectin (Fib II), and is similar to previously reported fragments from this region of plasma fibronectin (Click, E. M., and Balian, G. 1985, Biochem., 24:6685—6696).

The bands are visualized by a silver strain. ha \,/‘I Figure 6. Adhesion of hematopoietic cells to plasma fibronectin and the purified 38 kDa and 80 kDa tryptic fragments of plasma fibronectin.

Cr-labeled K562 (erythroleukemia), Jurkat (CD3+ T lymphocyte) and YT (CD3- T lymphocyte) cells (105/well) were allowed to adhere to plastic surfaces that had been coated with intact plasma fibronectin (pFN) or the purified 80 kDa and 38 kDa tryptic fragments at the indicate concentrations for 2 hours at 37° C. At the end of this time non-adherent cells were washed off and the bound cells were solubilized in SDS/NaOH and quantitated. lCr cpm.

The results are expressed as bound Figure 7. Effect of the mondélohal antibodies PlD6 and P4C2 to the integrin receptors c551 and a4fll (respectively) on adhesion of T lymphocytes to intact plasma fibronectin (pFN) or the purified 80 kDa and 38 kDa tryptic fragments. 51Cr—labeled Molt 4 cells were incubated with purified P1D6 or P4C2 monoclonal antibodies (50 pg/ml) or purified mouse IgG (50 pg/ml) for 1 hour at 4° C. They were then allowed to adhere to plastic surfaces that had been coated with intact plasma fibronectin, or the 80 kDa and 38 kDa tryptic fragments at the indicated concentrations for 1 hour. At the end of this time the non-adherent cells were washed off and the adherent cells were solubilized and bound Slcr cpm were quantitated in a gamma counter. The results are expressed as bound cpm.

Figure 8. Effect of CS-1 B12 peptide on T lymphocyte adhesion to IL-15 Activated HUVE cells.

Figure 9. (a) Diagram of the III CS and CS—1 regions. (b) Amino acid sequence of CS-1, A13, and B12. _l7..

. DETAILED DESCRIPTION OF THE INVENTION In experiments designed to examine the function of a5fl1 in lymphocytes, it was observed that resting peripheral blood and cultured T lymphocytes (Holt 4 or Jurkat) expressed an affinity for fibronectin independent of the prototype fibronectin receptor, a5fll. Although these cells attached to fibronectin-coated surfaces they expressed low or undetectable levels of a5fll recognized by the functionally defined monoclonal antibody, PlD6g(Wayner et al., 1988, J.

Cell Biol. lgZ:l88l-1891). adhesion to fibronectin could only be partially inhibited by Furthermore, T lymphocyte PID6 or RGD containing peptides suggesting the involvement of other receptors for fibronectin in the adhesion process.

Alternatively, adhesion of other cells to fibronectin, such as malignant or transformed fibroblasts and activated T lymphocytes (LAK cells) could be completely inhibited by PlD6. This suggested that resting peripheral blood T lymphocytes and cultured T cell leukemias express multiple independent and functional fibronectin receptors.

According to the present invention, an alternative fibronectin receptor was identified by preparing monoclonal antibodies that specifically inhibited the adhesion of T lymphocytes but not other cells to fibronectin. This receptor was identical to the integrin receptor, a4fl1, and mediated the attachment of peripheral blood lymphocytes, cultured T cell lines and RD cells to plasma fibronectin.

Furthermore, T lymphocytes expressed a clear preference for a 38 kDa tryptic fragment of plasma fibronectin (Garcia—Pardo et al., 1987, Biochem. J., 241:923-928) containing the Heparin II domain and 67 amino acid residues of the type III connecting segment (IIICS) spanning the CS-1, CS—2 and CS-3 regions defined by Humphries et al., 1986, J. Cell. Biol., 103:2637-2647; Humphries et al., 1987, J. Biol. Chem., 262:6886-6892). According to the present invention, T lymphocytes were found to attach only to CS—1 and monoclonal /‘ (2 -18.. antibodies to a4fll (P3E3, P4C2 P4G9) completely inhibited T lymphocyte adhesion to the 38 kDa fragment and to CS-1. T lymphocytes were also found to attach (with much lower affinity) to a site present in the Heparin II domain and monoclonal antibodies to a4fll also inhibited this interaction.

The functionally defined monoclonal antibodies to a4fil did not inhibit T lymphocyte adhesion to an 80 kDa tryptic fragment of plasma fibronectin containing the RGD sequence, whereas antibodies to a5fl1 (the prototype fibronectin receptor) completely inhibited this interaction.

In addition, the present invention relates to the discovery that the a4fll receptor mediates the interaction between lymphocytes and endothelial cells. According to the invention, antibodies or peptides can be used to block the adhesion of lymphocytes to endothelial cells.

For purposes of clarity of disclosure, and not by way of limitation, the present invention will be described in the following subsections. 3 i) Preparation of antibodies to extracellular matrix receptors (ECMRS); ii) Characterization of the ECMR-ligand interaction; iii) Methods of intervening in cell adhesion; iv) Utility of the invention; and v) Peptides and antibodies of the invention. .1. PREPARATION OF ANTIBODIES TO EXTRACELLULAR MATRIX RECEPTORS Preparation of antibodies to extracellular matrix receptors may be performed using any method for generating antibodies known in the art. extracellular matrix receptor (ECMR) may be used as Intact cells, or purified run immunogen. Immunization of a host is preferably performed using immunogen obtained from a xenogeneic source.

Antibodies may be polyclonal or monoclonal.

Various procedures known in the art may be used for the production of polyclonal antibodies to epitopes of a given ECMR. can be immunized by injection with an ECMR protein, or a For the production of antibody, various host animals synthetic protein, or fragment thereof, or, alternatively, intact cells may be used. Various adjuvants may be utilized to increase the immunological response, depending on the host species, and including but not limited to Freund’s (complete and incomplete), mineral gels such as aluminum hydroxide, surface active substances such as lysolecithin, pluronic polyols, polyanions, peptides, oil emulsions, keyhole limpet hemocyanins, dinitrophenol, and potentially useful human adjuvants such as BCG (bacille Calmette-Guerin) and Corynebacterium parvum.

A monoclonal antibody to an epitope of a ECMR can be prepared by using any technique which provides for the production of antibody molecules by continuous cell lines in culture. These include but are not limited to the hybridoma techniques originally described by Kohler and Milstein (1975, Nature 256:495-497) and Taggart and Samloff (1983, Science glg:1228-1230), and the more recent human B cell hybridoma technique (Kozbor et al., 1983, Immunology Today 4:72) and EBV-hybridoma technique (Cole et al., 1985, Monoclonal Antibodies and Cancer Therapy, Alan R. Liss, Inc., pp. 77-96}.

The monoclonal antibodies for therapeutic use may be human monoclonal antibodies or chimeric human—mouse (or other species) monoclonal antibodies. Human monoclonal antibodies may be made by any of numerous techniques known in the art (g;g;, Teng et al., 1983, Proc. Natl. Acad. sci.

U.S.A. 80:7308-7312; Kozbor et al., 1983, Immunology Today 4:72-79; Olsson et al., 1982, Math. Enzymol. 92:3—16).

Chimeric antibody molecules may be prepared containing a mouse . _20_ antigen-binding domain with human constant regions (Morrison et al., 1984, Proc. Natl. Acad. Sci. U.S.A. 8l:6851, Takeda et al., 1935, Nature 314:452).

A molecular clone of an antibody to an ECMR epitope Recombinant DNA methodology (see e.g., Maniatis et al., 1982, Molecular can be prepared by known techniques.

Cloning, A Laboratory Manual, Cold spring Harbor Laboratory, Cold Spring Harbor, New York) may be used to construct nucleic acid sequences which encode a monoclonal antibody molecule, or antigen binding region thereof. K Antibody fragments which contain the idiotype of the molecule can be generated by known techniques. For example, such fragments include but are not limited to: the F(ab')2 fragment which can be produced by pepsin digestion of the antibody molecule: the Fab’ fragments which can be generated by reducing the disulfide bridges of the F(ab’)2 fragment, and the 2 Fab or Fab fragments which can be generated by treating the antibody molecule with papain and a reducing agent.

Likewise, antibodies which are reactive with ECMRs produced by the above methods may be identified and selected by any technique known in the art. For example, antibodies may be shown to bind to and/or immunoprecipitate a known ECMR which has been purified or otherwise separated from other proteins, as in a polyacrylamide el. Alternatively, antibodies to ECMRS may be identified by their ability to compete with previously known ECMR antibodies for binding to ECMRS. Antibodies which bind to ECMRS may also be identified by their ability to block an ECMR/ligand interaction. For example, and not by way of limitation, cells bearing an ECMR pad \.I'1 _21_ receptor which binds to fibronectin (which need not, itself, have been identified or characterized, but merely functionally defined) may be shown to adhere to a substrate coated with fibronectin. If an antisera or hybridoma supernatant may be shown to inhibit the adherence of cells to the substrate, the antibodies contained in antisera or supernatant may recognize the ECMR receptor.

According to the invention, antibodies which recognize the a4fll receptor may be prepared by the methods outlined supra. In a preferred embodiment of the invention, monoclonal antibodies directed toward a4fi1 may be produced as follows: mice obtained from RBF/DN mice may be immunized with about 100 pl of packed T lymphocytes;'their spleens may subsequently e removed and fused with myeloma cells, for example, NS-1/FOX—NY myeloma cells, as described by 0i and Herzenberg (1980, in "Selected Methods In Cellular Immunology," Mishell and Shiigi, eds, Freeman and Co., San Francisco, pp. 351-373) and Taggart and Samloff (1983, Science 2_l2: 1228-1230) .

RPM1 1640 media supplemented with Viable heterokaryons may then be selected in adenine/aminopterin/thymidine. Hybridomas producing antibody directed toward lymphocyte ECMRS may be screened by adhesion to fibronectin-coated surfaces and cloned by limiting dilution. In particular, antibodies directed toward a4fll may be identified, for example, by the ability to block adherence of lymphocytes to substrate coated with CS—1 peptide or its derivatives, or to endothelial cells. Antibodies which recognize a4fll will not, however, inhibit the binding of cells bearing the a5fl1 receptor to RGD-peptide coated substrate.

Alternatively, antibodies directed toward a451, may be identified by their ability to i) competitively inhibit the binding of known anti-a451, antibodies (such as P402 or P4C10), or ii) bind to the same protein as known anti—a4fl1, antibodies (e.g. in a protein gel, Western blot, or in sequential immunoprecipitation experiments). you .2. CHARACTERIZATION OF THE ECMR/LIGAND INTERACTION The interaction between an extracellular membrane receptor may be characterized, for example, and not by way of limitation, by the following methods: i) Determination of receptor distribution and function; ii) Intervention in receptor/ligand binding; iii) Isolation and chemical characterization of receptor and/or ligand.

These methods will be described more fully in the three following subsections.

DETERMINATION OF RECEPTOR DISTRiBUTION AND FUNCTION According to the methods of the invention, receptor .2.1. distribution may be determined using any method known in the art. For example, and not by way of limitation, cell populations bearing the ECMR may be identified using monoclonal antibodies directed toward the ECMR of interest.

Binding of antibody to the ECMR may be detected using immunohistochemical techniques such as immunofluorescence and immune peroxidase staining. Alternatively, populations of cells bearing the ECMR of interest may be collected using fluorescence-activated cell sorting techniques.

Because there appears to be a specific reorganization of cell surface adhesion receptors to the focal adhesions when cells are grown on the appropriate ligands (Burridge et al., 1988, Ann. Rev. Cell Biol. g:487—525), one method for characterizing the functional interaction between a given receptor and a potential ligand involves determining whether the ECMR of interest distributes into the focal adhesions formed between cell and ligand substrate. For ‘example, and not by way of limitation, a4fl1 may be shown to interact with fibronectin in a receptor/ligand relationship by the following method (see also section 6.2.3., infra).

Lymphocytes may be allowed to adhere to a fibronectin substrate, and the focal adhesions between cells and substrate may be visualized by interference reflexion microscopy (Izzard et al., 1976, J. Cell Sci. gi:129—159). Antibodies which recognize a4fll, such as P4G9 or P4C10, may be used to show, using standard immunohistochemical techniques, e.g. fluoro-iso thiocyanate, that in the absence of serum, a4fl1 redistributes into the focal adhesions.

Interaction between ECMR and ligand may also be characterized by testing for the ability of the ECMR to adhere to a variety of different substrates. type of interest or an ECHR of interst may be tested for the ability to bind to substrates consisting of purified components of the extracellular matrix; such as fibronectin, For example, a cell collagen, vitronectin or laminin. In a specific embodiment of the invention, cells bearing the a4fl1 may be shown to adhere to fibronectin, but not to collagen or laminin substrates as a result of the c451/fibronectin interaction.

Further where any ECMR of interest is shown to bind to a particular protein ligand, substrates bearing subfragments of the protein ligand may be tested for the ability to bind to an ECMR on the surface of cells, thereby permitting the localisation of the binding site between ECMR and ligand.

Specifically,where the receptor is d481, which has been determined to bind to fibronectin (supra), substrates bearing subfragments of fibronectin may be tested for their ability to bind d4Bl—bearing cells, as exemplified in Section 6, infra. Although lymphocytes attached to the 80 kDa cell binding domain of fibronectin bearing the d481, receptor (Figure 5a) they demonstrated a clear preference for an non—RGD containing region located on a 38 kDa fragment.

\. G -24..

On a molar basis, the 38 kDa fragment was three times more efficient than the 58 kDa fragment in mediating T lymphocyte adhesion. As shown in Figure 5A the 38 kDa and 58 kDa fragments were derived from the A and B chains of plasma They therefore differ in the plasma fibronectin. .2.2. INTERVENTION IN RECEPTOR/LIGAND BINDING The ECMR/ligand relationship may be further characterized by identifying and evaluating agents which interfere with receptor/ligand binding.

For example, antibodies directed to an ECMR of interest may be used to inhibit ligand/receptor binding.

Given the observation that a particular cell type adheres to a given ligand or cellular substrate, it may be of interest to identify the ECMR involved in the interaction. A panel of monoclonal antibodies, each directed toward a different ECMR, 'may be tested for the ability to block the adherence of cells to substrate. Inhibition of binding by a particular antibody would suggest that the ECMR recognized by that antibody is involved in the adhesive interaction. In a specific embodiment of the invention, lymphocyte adherence to endothelial cells in culture may be inhibited by antibodies directed toward a4fll, variety of other ECMRS (see Section 7, that a4p1, is necessary for lymphocyte adhesion to endothelial In contrast, these kDa fragment containing the prototype adhesion sequence arg- gly—asp (RGD). Adhesion of T cells to the 80 kDa fragment could be completely inhibited by a monoclonal antibody to a5fl1 (PlD6) or by RGDS. P1D6 and RGDS failed to inhibit T lymphocyte adhesion to the 38 and 58 kDa fragments or to CS~ 1. Together, these data show that a4fll functions as the receptor for the carboxy terminal adhesion domain of plasma fibronectin receptor for alternative adhesion sequences in IIICS (CS-1) and possibly Hep II.

Further. the ECMR/ligand relationship may be characterized by determining the structure of the ligand. In particular, the ability of agents to compete with ligand in the ECMR/ligand interaction may be evaluated. For example, where the ligand is a protein, various fragments of the protein may be tested for ,their ability to competitively inhibit receptor/ligand binding. In particular because lymphocytes are observed to bind to endothelial cells as well as to fibronectin, peptide fragments of fibronectin may be tested for the ability to completely inhibit the binding of lymphocytes to endothelial cell substrate. As exemplified in Section 7, infra, CS-l peptide, and, in particular, the peptide EILDVPST was able to competitively inhibit the binding of lymphocytes to fibronectin and to endothelial cells, thereby localizing the binding site on the ligand to a region identical or homologous to EILDVPST. .3. METHODS OF INTERVENING IN CELL ADHESION According to the invention, adherence of one cell to another may be inhibited by intervening in the ECMR/ligand interaction. In a particular embodiment of the invention, the binding of lymphocytes to endothelial cells may be inhibited by interfering with the binding of a4fi1 to its ligand. This may be accomplished by using antibodies directed toward the ECMR, or, alternatively, to its ligand (antibodies may be generated toward ligand in a manner analogous to that described in Section 5.1).

In a specific embodiment of the invention, anti-a4fl1 antibody, or a fragment or derivative thereof, may be used to inhibit the binding of lymphocytes bearing a4fil receptors to vascular endothelial cells. In preferred embodiments, the antibody is a monoclonal antibody, in particular antibody P4C2 (a4fl1) or P4C10 (fil), or fragments or derivatives thereof, including chimeric antibodies with the same binding specificities. particular antibody P4C2 (a4fl1) or P4C10 (51), or fragments or derivatives thereof, including chimeric antibodies with the same binding specificities. .4. UTILITY OF THE INVENTION ___________w____________ According to the invention, the adherence of one cell to another may be inhibited by interfering in the binding between the ECMR and its ligand. In the embodiments of the invention, the adherence of lymphocytes to endothelial cells may be inhibited by interfering with the binding of a4fl1 on lymphocytes to its ligand on the endothelial cell surface. According to the invention, the interaction of additional ECMR with endothelial cell ligands, and the inhibition of adhesion of these cells to endothelium by interfering with the ECMR/endothelial cell interaction is envisioned. For example, the adhesion of macrophages to the endothelium may also be inhibited by intervention in the macrophage ECMR/endothelial cell interaction. melanoma cells, which also recognize the CS—1 peptide, may be inhibited from metastasizing and entering tissues using the Likewise, peptides or antibodies of the invention.

The method used in the invention is therefore useful in preventing the egress of lymphocytes through the vascular endothelium and into tissue. Accordingly, the present invention provides for a method of suppressing the immune response in human patients in need of such treatment. In particular embodiments, the present invention provides for methods of treatment of diseases associated with chronic or relapsing activation of the immune system, including collagen vascular diseases and other autoimmune diseases (such as systemic lupus erythematosis and rheumatoid arthritis), multiple sclerosis, asthma, and allergy, to name but a few.

The present invention is also for use in the treatment of relatively acute activations of the immune system in patients in need of such treatment, including, for example, and not by way of limitation, graft versus host disease, allograft rejection, or transfusion reaction.

Depending on the nature of the patient's disorder, it may be desirable to inhibit lymphocyte migration into tissues systemically or, alternatively, locally. For example, in diseases involving multiple organ systems, such as systemic lupus erythematosis, it may be desirable to inhibit lymphocyte adhesion systemically during a clinical exacerbation. However, for a localized Contact dermatitis, it may be preferable to restrict migration of lymphocytes only into those tissues affected.

Control of systemic versus localized use of the methods of the present invention may be achieved by modifying the compositions of antibodies or peptides administered or by altering the structure of these agents or their pharmacologic compositions. For example, the antibodies or peptides of the invention may be administered by any route, including subcutaneous, intramuscular, intravascular, intravenous, intraarterial, intranasal, oral, intraperitoneal, rectal, intratracheal, or intrathecal. However, to achieve local inhibition of lymphocyte adhesion to endothelium, it may be desirable to administer the antibodies or peptides of the invention, in therapeutic amounts and in a suitable pharmacologic carrier, subcutaneously or intramuscularly.

Alternatively, to achieve systemic inhibition of lymphocyte adhesion, it may be desirable to administer the antibodies or peptides intravenously.

In various embodiments of the invention it is advantageous to use a pharmacologic carrier which facilities delivery of the antibodies<3f the invention For example, when antibodiesare to be delivered to the skin (e.g. for the treatment of chronic -29..

In additional embodiments, the antibodies of the invention may be conjugated to antibodies or other ligands which might direct the antibodies to specific tissues.

Furthermore, antibodies may be produced which mimic the ECMR, and thereby attach to endothelial cell ligands, blocking lymphocyte adhesion. _30..

ANTIBODIES OF THE INVENTION Antibodies of the invention, produced and defined as described supra, include monoclonal as well as polyclonal antibodies and fragments and derivatives thereof, the F(ab’)2, Fab’, and Fab fragments. including EXAMPLE: INDENTIFICATION AND CHARAZTERIZATION OF THE LYMPHOCYTE ADHESION RECEPTOR FOR AN ALTERNATIVE CELL ATTACHMENT DOMAIN IN PLASMA FIBRONECTIN The following experiments have described a new fibronectin receptor which appears to be identical to the integrin receptor a4fl1 (flemler et al., 1987, supra), preferentially expressed by nucleated hematopoietic cells.

Identification of a4fll as a specific fibronectin receptor was based on (i) inhibition of cell adhesion to fibronectin by monoclonal antibodies (P4C2, P3E3 and P469), and (ii) specific reorganization and concentration of a4fi1 into fibronectin-dependent focal adhesions. These findings suggest that a4flI and a551, the prototype fibronectin receptor function together as primary mediators of cell adhesion to fibronectin.

MATERIALS AND METHODS 6.1.1. REAGENTS Phenylmethyl sulfonyl fluoride, n—ethylmaleimide, leupeptin, diisopropyl fluorophosphate, 2—mercaptoethanol, bovine serum albumin (BSA), Triton X-100, Protein A—Agarose, soybean trypsin inhibitor, and V8 protease (from Staphylococcus aureus, strain V8, protease type XVII) were .1. purchased from Sigma Chemical Co. (St. Louis, MO).

Lactoperoxidase and glucose oxidase were from Calbiochem (San Diego, CA).

Malvern, PA. Fluorescein-conjugated (goat) anti-mouse IgG and IgM (H and L chains) or rhodamine-conjugated (goat) anti-rabbit IgG and IgM (H and L chains) were obtained from R-phycoerythrin-conjugated TPCK—trypsin was from Cooper Biomedical, Tago, Inc. (Burlingame, CA). _31_ Rabbit anti-mouse IgG (H+L) antiserum was obtained from Cappel 51 strepavidin was from Biomeda (Foster City, CA).

(Cooper Biomedical, Malvern, PA). Cr—sodium chromate was from New England Nuclear. 125 Hts., IL). Human recombinant interleukin-2 (IL-2) was a I was from Amersham (Arlington generous gift from Dr. D. Urdal (Immunex Corp., Seattle, WA).

Laminin was purchased from Collaborative Research, Inc.

Bedford, MA) and purified plasma fibronectin and collagen Types I and III were prepared as previously described (Wayner, E. A. and Carter, W. G., 1987, sgpra and Wayner et al., 1988, supra).

Peripheral blood from normal human donors were prepared as described (Kunicki et al., 1988, supra; Wayner et al., 1988, supra). blood cells from patients with acute lymphocytic, large Peripheral granular lymphocyte (LGL) or myelogenous leukemia were obtained from Dr. I. Bernstein and Dr. T. Loughran (Fred Hutchinson Cancer Research Center). Human lymphokine (500 U/ml IL-2) activated killer (LAK) cells and the monoclonal HLA 87 specific human cytotoxic T lymphocyte (CTL) cell line, C1C4, were prepared according to standard protocols (Grimm et al., 1982, J. Exp. Med., 155:1923-1941; Glasebrook, A. L. and Fitch, F. W., 1980, J. Exp. Med., 15l:876-895; Brooks, 1983, Nature, 305:155—158: wayner, E. A. and Brooks, C. G., 1984, J. Immunol., 132:2135-2142; Wayner, E. A. and Carter, W. G., 1987, J. Cell Biol., 105:1873~1884). The EBV transformed B lymphocyte cell line (BLCL), ST-1, was derived from the donor spleen used in the production of the C1C4 CTL line. All _32, other cell lines and cell culture conditions were as previously described (Wayner, E. A. and Carter, W. G., 1987, supra: Wayner et al., 1988, supra). .1.3. ANTIBODIES A rabbit polyclonal antibody, AB33, prepared against the cytoplasmic domain of the fibronectin receptor, a5fil, used to detect asfll in focal adhesions. Monoclonal antibodies A1A5, against the common integrin (Hynes, R. 0., 1987, supgg) 51 subunit of the VLA family of receptors (Hemler, M. E., 1988, ggprg) and BS-G10 to the VLA 4 a subunit (Hemler et al., , supra) were obtained from Dr. Martin Hemler of the Dana—Farber Cancer Inst., Boston, MA).— Monoclonal antibodies to the integrin receptors a3fi1 (P185), a2fl1 (P1H5) and a5fll (P106) have been described and were developed in this laboratory. PIHS and PID6 inhibit fibroblast and platelet adhesion to collagen and fibronectin-coated substrates, EBBEE7 Kunicki et al., 1988, supra; Wayner et al., 1988, supra).

Monoclonal antibodies to lymphocyte adhesion respectively (Wayner, E. A. and Carter, W. G., 1987, receptors were produced by the methods of Oi and Herzenberg (Oi, V.T. and Herzenberg, L.A. 1980, Immunoglobulin producing hybrid cell lines. In: Selected Methods in Cellular Immunology. Ed. by B.B. Mishell and S.M. Shiigi, W.H.

Freeman and Co., San Francisco, pp. 351-373) and Taggart and Samloff (Taggart, R.T. and Samloff, I.M., 1983, Science, 219, 1228-1230) as described (Wayner and Carter, 1987; Wayner, et al., 1988). spleens from RBF/DN mice immunized with 100 pl f packed T lymphocytes were removed and fused with NS—1/FOX-NY myeloma cells. Viable heterokaryons were selected in RPMI 1640 supplemented with adenine/aminopterin/thymidine (Taggart and Samloff, 1983). Hybridomas producing antibody directed to lymphocyte adhesion receptors were screened by specific inhibition of lymphocyte adhesion to fibronectin-coated surfaces and cloned by limiting dilution. standard biochemical criterion.

INHIBITION OF CELL ADHESION TO INTACT FIBRONECTIN AND FIBRONECTIN FRAGMENTS Antibodies that would alter cell adhesion to .1.4. purified plasma fibronectin, tryptic fragments and CS peptides were identified as previously described (Wayner and 1987). Briefly, 48 well virgin styrene plates were coated with human plasma fibronectin (5 pl/ml). The plates were blocked with PBS supplemented with 10 mg/ml heat denatured BSA (HBSA). T lymphocyte or HT108O cells were labeled with Naz 51Cr04 (50 plci/ml for 2-4 hr), washed, and 5 X 104 HT1080 or cultured T cells or 5 X 1010 PBL/well were incubated with hybridoma culture supernatants (1:2 dilution in PBS supplemented with 1 mg/ml heat denatured BSA) or control myeloma cell culture supernatant for 15 minutes at The cells were allowed to adhere to the Carter, room temperature. protein-coated surfaces in the presence of the hybridoma supernatants for 15-30 minutes (HT1080) or 2-4 hours (lymphocytes) at 37'C. Non-adherent cells were removed by 'washing with PBS, and the adherent cells were dissolved in SDS/NaOH and bound 51Cr-cpm were quantitated in a gamma counter.

- IMMUNE PRECIPITATION, AND SEQUENTIAL IMMUNE PRECIPITATION, V8 PROTEASE PEPTIDE MAPPING AND POLYACRYLAMIDE GEL ELECTROPHORESIS Viable cells were surface labeled with 125-iodineflas .1.5.

Polyacrylamide slab gels containing sodium ).

PREPARATION OF TRYPTIC FRAGMENTS FROM HUMAN PLASMA FIBRONECTIN AND SYNTHESIS OF CS PEPTIDES .1.6. fibronectin (CS-1 and CS-2) were synthesized and coupled to rabbit IgG as described (Humphries et al., 1986, and Humphries et al., 1987, J. Biol. Chem. 262:6886-6892). run- \/1 .1.7. FLUORESCENCE ANALYSIS OF RECEPTOR EXPRESSION Expression of ECMRS on cells in suspension was analyzed by one or two color flow cytometry on an EPICS 750 dual laser cell sorter (Coulter, Hialeah, FL). Positive fluorescence was determined on a 3 decade log scale and fluorescence intensity (log F1) was expressed as mean channel number (O-255). mouse IgG negative control was determined for each cell Background fluorescence for a non-immune population and subtracted. Adherent cells were trypsinized and allowed to recover for 15 minutes at 37'C in the presence of serum before use for flow cytometry. cells in suspension were For one or two-color fluorescence measurements, 106 incubated for 30 minutes with protein‘C3Sepharose purified goat IgG (20 pg/ml) and then with first stage antibodies at 4'C for 60 minutes, washed in Hanks Balanced Salt solution containing 10 mg/ml HBSA and 0.02% sodium azide (Hanks/BSA/SA) and incubated with FITC-conjugated rabbit anti—mouse IgG for 60 minutes at 4'C in Hanks/BSA/SA. They were washed and fixed in cold 2% paraformaldehyde (prepared fresh) in PBS. For two—color fluorescence purified and biotinylated monoclonal antibody was then added to the FITC- stained and fixed cells to a final concentration of 1 ug/ml in Hanks/BSA/SA and incubated at 4’C for 60 min. Prior fixation with 2% paraformaldehyde had little effect on expression of lymphocyte integrin receptors. The fixed cells were washed and incubated in 0.5 ml Hanks/BSA/SA containing phycoerythrin—conjugated strepavidin (Bionetics) at 1/50 dilution for 30 min at 4‘C. Finally the stained cells were washed and fixed again in 2% paraformaldehyde in PBS and held at 4’C in the dark for analysis on the EPICS flow cytometer. .1.8. LOCALIZATION OF RECEPTORS IN FOCAL ADHESIONS Adherent cells were trypsinized, washed in RPMI supplemented with 1 pg/ml BSA plus 100 ug/ml soybean trypsin inhibitor and allowed to adhere to acid washed and silanized glass cover slips coated with fibronectin, laminin or collagen (20 ug/ml) in the absence of serum for 1-4 hour as At the end of the incubation non—adherent cells were removed and adherent cells were fixed in 100 mM sodium cacodylate, 100 mM sucrose, 4.5 mM Caclz, 2% formaldehyde for 20 min. permeabilized with 0.5% Triton X-100 for 5 minutes, then washed and blocked with 25% goat serum in PBS. The permeabilized cells were stained with antibodies to specific described (Carter and wayner, in preparation).

They were receptors (60 minutes at room temperature), washed and incubated with either FITC-conjugated goat anti-mouse or rhodamine—conjugated goat anti—rabbit IgG (45 minutes at room temperature) and washed again. The c6Vér~slips,were inverted onto glass slides for fluorescence and interference reflexion microscopy (IRM) as described (Izzard, S.C. and Lochner, L.R., 1976, J. Cell. Sci., 21, 129-159). .1.9. TISSUE STAINING All sections versa. .2. RESULTS IDENTIFICATION OF AN ALTERNATIVE FIBRONECTIN RECEPTOR Cultured T lymphocytes (Molt 4), K562, RD (rhabdomyosarcoma) and HT108O (fibrosarcoma) cells and .2.1. freshly derived PBL (not shown) adhered to fibronectin-coated surfaces (Figure 1: However, Molt 4 and RD cells expressed low or undetectable levels of the prototype fibronectin receptor (integrin a5fll) recognized by monoclonal antibody P1D6 (Figure 1: striped bars). Consistent with this, adhesion of Holt 4 and RD cells to fibronectin could not be completely inhibited by P1D6 (Figure 1: solid bars).

Alternatively, adhesion of cells to fibronectin that expressed abundant a5fl1 (HT1080 and K562) could be effectively inhibited by P1D6. Furthermore, the synthetic peptide RGDS did not completely inhibit T lymphocyte adhesion to plasma fibronectin (S0—70% for Molt 4 or Jurkat cells versus 80-90% open bars). for fibroblasts and 100% for K562-1 cells). Together, these data suggested that some cells, such as T lymphocytes, express fibronectin adhesion receptors other than a5Bl.

We attempted to identify other putative fibronectin receptors by preparing monoclonal antibodies to cultured T lymphocytes and screening them for their ability to specifically inhibit lymphocyte but not fibroblast adhesion to fibronectin-coated surfaces. Using this protocol several monoclonal antibodies (P4C2, P3E3, P4G9) were identified that inhibited cultured T lymphocyte but not HT1080 cell adhesion to fibronectin (Table 3). _38_ TABLE III SPECIFIC INHIBITION OF LYMPHOCYTE ADHESION TO PLASMA FIBRONECTIN BY MONOCLONAL ANTIBODIES P3E3, P4C2 AND PPG49 FIBRONECTIN ADHESION (% or CONTROL) CELLS SP2 PID6 P3E3 P4C2 P4G9 (a5B1) ISPBL 100% 43% 38% 10% 52% Jurkat 100% 22% 33% 12% 48% Molt 4 100% 18% 12% 8% 39% HTIOBO 100% 5% 98% 93% 104% _39_ Immune precipitation from Triton®X-100 detergent lysates prepared with 125 HTl080 (Figure 2) cells showed that the inhibitory monoclonal antibodies (data shown for P3E3) reacted with a single protein present in lymphocyte extracts that migrated at M 150,000 (p150) in the presence (not shown) or absence (Figure 2) of reducing agent. Under these immune precipitation conditions p150 lacked an apparent a-fl subunit structure and did not migrate with either the a or p subunit of the integrin receptors azfil or a3fl1 (Figure 2). The antigen immune precipitated from'Triton®x—1oo detergent extracts prepared with chronically activated CD8+ killer T lymphocytes (LAK) or CTL (not shown) contained, invaddition to p150, relatively large quantities of two smaller proteins that migrated at M 80,000 and 70,000 in the presence (not shown) or absence of reducing agent. V8 protease peptide mapping revealed that p80 and p70 were proteolytic fragments of p150 (not shown). These lower molecular weight forms could be immune precipitated from chronically activated T cells even when detergent extracts were prepared in the presence of multiple protease inhibitors (Figure 2 legend). p80 and p70 were virtually absent from extracts prepared with resting PBL, cultured T (Holt 4, Jurkat) or B cell leukemias and RD cells.

The biochemical characteristics of p150 suggested that it might be related to the VLA 4 antigen described by Hemler (Hemler et al., 1987). This was confirmed by sequential immune precipitation (not shown) with a VLA 4 specific monoclonal antibody, B5—G10. p150 was established as an a subunit of the integrin super family by its association with pl when immune precipitations were carried out after CHAPS detergent (0.3%) solubilization of 1251 surface labeled T lymphocytes in the presence of 1 mM Ca++ Under these conditions a4 was precipitated as a The identity of pl was confirmed by V8 (Figure 3). heterodimer with fil.

I—surface labeled PBL (not shown), Molt 4 or pfotease peptide mapping (not shown). The a4Bl heterodimer immune precipitated from T lymphocytes with the inhibitory monoclonal antibodies (P3E3, P4C2 and P4G9) was shown to be distinct from the prototype fibronectin receptor, a5fll, immune precipitated with P106 by three criteria. 1) The relative quantities of a4fll and a5fil present in detergent extracts of T lymphocytes were distinct with higher levels of This was in agreement with ) In a4fil being present (Figure 3). the data we obtained using flow cytometry (Figure 1). sequential immune precipitation experiments, monoclonal antibodies to a4fll did not preclear a5fl1 (now shown). 3) The V8 protease peptide maps derived from the a4 and a5 subunits precipitated with monoclonal antibodies F323 and PlD6 were clearly distinguishable (not shown). conditions (0.3% CHAPS and 1mM Caclz) used to solubilize the Furthermore, under the conjugate of a4p1 from Jurkat cells (Figure 3) another protein of higher molecular weight (p180) also reacted with the monoclonal antibodies or co—precipitated with a4fll. p180 was absent from extracts prepared with PlD6 monoclonal antibody (Figure 3), non-lymphoid cells or Triton®X—100 detergent extracts prepared in the absence of Ca++. The relationship of p180 to other integrins is not known. Since a4 could be immune precipitated without 51 after solubilization of T cells with Triton®X-100 in the absence of Ca++ this revealed that the inhibitory monoclonal antibodies recognized epitopes present on the a4 subunit (Figure 2).

DISTRIBUTION OF a4fl1 AND a5fi1 IN CULTURED CELLS AND TISSUE As has been previously reported (Hemler, supra), a4fil was widely distributed on nucleated hematopoietic cells ’(Table IV). .2.2. #41, TABLE IV DISTRIBUTION OF THE FIBRONECTIN ‘ RECEPTORS a4fl1 AND asfll ON HUMAN CELLS Relative Flourescence Intensity Cells W __ Hematopoietic Cells PBL +++ LGL (CD3-, CD16+) +++ Monocytes (CD16+) ++ Granulocytes - 2oPlate1ets - Spleen +++ Tonsil +++ ALL (T or B) +++ LGL Leukemia (CD3+, CD4+) +++ ZSAML +++ BLCL ++ Holt 4 (CD3+, CD4+) +++ Jurka (CD3+, CD4+) +++ YT (CD3—) ++ 30PHA blasts (CD4+) ++++ CTL (CD3+, CD8+) ++++ LAK (CD3+, CD8+) ++++ HL—6O ++ U937 - ++ +/- or — +/- or - +/- +4- +++ +++ Cells c451 K562-1 - Fibroblasts HFF (p5) + HT1080) + RD ++ VA13 + Egithelial Cells OC-1 OVCAR-47 T47D QG56 Relative Flourescence Intensity a5fi1 lymph node and tonsil _43._ Two-color flow cytometry revealed that all lymphocyte subpopulations derived from spleen, tonsil and peripheral blood expressed abundant a4fl1. In addition, peripheral blood monocytes, freshly derived acute lymphocytic (T or B) leukemias, all large granular lymphocytic (LGL) and myelogenous leukemias and cultured T and B lymphocyte cell lines we examined expressed abundant a4fll (Table IV).

Normal human blood platelets and granulocytes were negative for a4fi1 (Table IV). granulocytes were negative for a4fil.

Normal human blood platelets and In contrast, the only hematopoietic cell populations that expressed a5fil were activated T cells, platelets, monocytes and granulocytes, acute lymphocytic (T or B) and myelogéhofis leukemias and cultured K562, HL-60 and U937 cells. Some cultured T (Molt 4 or Jurkat) and B (ST—1) cell lines expressed low levels of a5fl1 as detected by P1D6 monoclonal antibody. In some normal individuals, a subpopulation of PBL were positive for PlD6 fluorescence detected by flow cytometry. we are investigating the nature of this subpopulation of PBL which express a5B1. TY cells, a CD3 T cell lymphoma, were completely negative for PlD6 by flow cytometry. These results show that the major fibronectin receptor constitutively expressed by resting T lymphocytes is a4fil and as we have previously reported (Wayner et al., 1988) expression of a5fi1in T lymphocytes is restricted to leukemic or activated cultured cells. Interestingly, most fibroblast cell lines expressed low levels of a4fl1 while large vessel endothelial cells (HUVES) and cultured epithelial cells were negative for a4fll by flow cytometry.

In tissue, a4fl1 was present in adult spleen, and essentially absent from all other tissues we examined. In addition, the relative quantities of the fibronectin adhesion receptors expressed by cells in specific tissue domains varied dramatically. For example, PBL and lymphocytes in tonsil and cortex and germinal -44.. genter areas expressed large quantities of a4Bl but virtually no a5fi1. a4fil was also found in epithelial regions in adult lymphatic tissue, but whether this was the result of lymphocyte infiltration of these areas or expression of a4fll by lymphatic epithelial cells was unclear. .2.3. a4fl1 LOCALIZES IN FIBRONECTIN-DEPENDENT FOCAL ADHESIONS There is a specific reorganization of cell surface adhesion receptors to the focal adhesions when cells are grown on the appropriate ligands in the absence of serum (reviewed by Burridge et all: i988, Ann. Rev. Cell Biol., 4, 487-525). investigated whether this receptor would distribute into Since some fibroblasts express c451 we focal adhesions when fibronectin was used as the adhesion substrate. primary focal contact sites or focal adhesions could be As can be seen in Figure 4 (A and C), the visualized by interference reflexion microscopy (Izzard, S.C. L.R., 1976, J. Cell. SCi., 21, 129-159) when RD cells were grown in fibronectin. As we and others and Lochner, cells. !\3 Q1 -45..

’ The presence of both receptors in focal contacts suggested the possibility that a4fll and a551 bind to distinct adhesion sequences in fibronectin. In fact, evidence for this was obtained when P4C2 and P1D6 were used simultaneously to inhibit cell adhesion to intact plasma fibronectin. PlD6 and P4C2 when used together completely inhibited adhesion of T lymphocytes and partially inhibited adhesion of RD cells to intact plasma fibronectin (Table V).

TABLE V COMBINED EFFECT OF MONOCLONAL ANTIBODIES PlD6 AND P4C2 ON T LYMPHOCYTE AND RD CELL ADHESION TO FIBRONECTIN ADHESION CELLS ANTIBODY SPECIFICITY (% OF CONTROL 1 SD) RD IGG - 100% P1D6 1:551 31 1 11 P4C2 a4fll 99 1 7 P1D6 + ~~A:~ 36 1 8 P4C2 Jurkat IGG - 100% P1D6 a5fll 26 1 9 P4C2 a4fll 38 1 14 P1D6 + O P4C2 {U K,/W «lymphocytes.

Interestingly, unlike T lymphocytes, neither PlD6 nor P4C2 alone were good inhibitors of RD cell adhesion to intact plasma fibronectin. RD cell adhesion to fibronectin could be efficiently inhibited by P1D6 and P4C2 only when used together. .2.4. a4fl1 FUNCTIONS AS THE RECEPTOR FOR AN RGD INDEPENDENT ALTERNATIVE ATTACHMENT SITE IN FIBRONECTIN The preceding results (Table III, Table V, Figure 1, Figure 4) clearly indicated that attachment of some cells to plasma fibronectin was mediated by two independent cell surface receptors, qgpl_and a5fll. It has been well documented that the ligand for asfil in fibronectin is the 80 kDa cell—binding domain which contains the RGD sequence (Pytela, R., Pierschbacher, M.D.,, and Ruoslahti, E., 1985, Cell, gQ:191-198). To determine the region of fibronectin that interacts with a4Bl we examined the adhesion of cultured T lymphocytes to various proteolytic fragments of plasma fibronectin (see Figure 5A and B), as well as the effect of monoclonal antibodies PID6 and P4C2 on lymphocyte adhesion to these fragments. As shown in Figure 6, Jurkat, YT and Molt 4 cells attach to a 38 kDa fragment containing the Heparin (Hep) II domain much more efficiently than to an RGD-containing fragment (80 kDa). Jurkat and Holt 4 cells also attach in a dose dependent manner to another Hep II domain containing fragment of 58 kDa. Maximum cell attachment to the 58 kDa fragment, however, reached only 30% of that achieved by the 38 kDa fibronectin fragment. This suggests that the 38 kDa fragment contains a high affinity attachment site for T T lymphocytes did not adhere to the N- terminal 29 kDa fragment containing the Hep I domain of plasma fibronectin. In general, freshly derived PBL showed a similar pattern of attachment as Jurkat or Molt 4 cells mm: \,l'\ agd the ability of freshly derived PBL to bind to the 80 Other hematopoietic cell lines such as K562 cells (Figure 6) kDa fragment correlated with expression of a5fi1. exhibited a clear preference for the 80 kDa fragment of plasma fibronectin while RD cells expressed promiscuous adhesion to all the gragments of plasma fibronectin tested except the N-terminal 29 kDa fragment. RDGS (1 mg/ml) partially inhibited (50%) Jurkat cell adhesion to intact fibronectin and completely (100%) inhibited their adhesion to the 80 kDa fragment. Jurkat cell adhesion to the 38 kDa fragment was unaffected by RGDS (up to 1 mg/ml).

As we have previously shown (Table 3 and Figure 1), monoclonal antibodies to a4fi1 and a5fi;_partially inhibited T lymphocyte adhesion to intact plasma fibronectin (Figure 7, top). As expected, P1D6 completely inhibited adhesion of T cells to the 80 kDa fragment which contains the RGD adhesion sequence (Figure 7, middle).

PlD6 did not inhibit T lymphocyte adhesion to the 38 kDa (Figure 7, bottom) or 58 kDa fragments. In contrast, P4C2 completely inhibited T lymphocyte adhesion to the 38 kDa fragment and had no effect on adhesion to the 80 kDa fragment (Figure 7). Furthermore, adhesion of T lymphocytes to the 58 kDa fragment which also contains Hep II could be inhibited by P4C2. In every case other T lymphocyte cell lines which express both a4fi1 and a5p1 (such as Jurkat cells) behave exactly as Molt 4 cells (Figure 7). As seen in Table 4, K562 cells express only a5p1. Adhesion of K562 cells to the 38 (Figure 6) and 58 kDa fragments was greatly reduced when compared to their adhesion to the 80 kDa fragment (Figure 6). Adhesion of these cells to intact plasma fibronectin (Figure 1) or the ‘80 kDa fragment could be completely inhibited by P1D6. on the other hand, YT cells which do not express aspl (Table IV) adhere poorly to intact plasma fibronectin and the 80 kDa fragment (Figure 6). These cells require 2-3 times .5 a1., 1987, J. Biol. Chem. 262:6886-6892). ..49- longer to adhere to plasma fibronectin—coated surfaces than Jurkat or Molt 4 cells. efficiently and in a dose dependent manner to the 38 kDa YT cells, however, adhere fragment (Figure 6) and adhesion of these cells to the 38 kDa fragment could be completely inhibited by P4C2. data indicate a direct correlation between expression of a4fll and the ability to attach to fragments of plasma fibronectin containing the Hep II and IIICS regions.

Furthermore, these data show unequivocally that a4fll functions as the receptor for this alternative cell These adhesion domain. .2.5. a4fi1 IS THE LYMPHOCYTE RECEPTOR FOR CS-1 this interaction.

? T lymphocytes (Jurkat or Molt 4 cells) recognize and attach to CS-1 (rabbit IgG conjugate) -coated plastic surfaces (Table VI). T lymphocytes (Jurkat) do not attach to CS-2 (rabbit IgG conjugate) coated surfaces or to plastic surfaces coated with rabbit IgG alone. monoclonal antibodies to a4fi1 (P4C2) completely inhibited T lymphocyte adhesion to CS-1 while antibodies to a5fl1 (P106) had absolutely no effect (Table VI). shown antibodies to a4fi1 completely and specifically inhibited T lymphocyte adhesion to the 38 kDa fragment (Table VI) while antibodies to osfil specifically inhibited Furthermore, As we have previously adhesion to the RGD containing 80 kDa fragment. -51..

‘E’ TABLE VI INHIBITION 0? T LYMPHOCYTE ADHESION To cs-1 PEPTIDE WITH MONOCLONAL ANTIBODIES TO a4fll 1 ANTIBODY LIGAND IgG p4c2 P1D6 80 kDa 8580 1 214 7154 i 398 202 i 105 38 kDa 22680 i 1014 114 1 78 24917 i 352 cs—1 44339 1 513 841 3 555 42397 i 728 cs-2 2576 1 214 535 3 258 435 1 168 ‘V’! .3. DISCUSSION Using monoclonal antibody technology (wayner, E.A., Carter, W.G., Piotrowicz, R. and T.J. Kunicki, 1988, J.

Cell Biol., 10: 1881-1891) we have identified a new fibronectin receptor a4fi1. and P4G9 recognized epitopes on the a4 subunit and completely inhibited the adhesion of peripheral blood and cultured T lymphocytes to a 38 kDa tryptic fragment of plasma fibronectin containing the carboxy terminal Heparin II domain and part of the type III connecting segment (IICS). The ligand in IIICS for a4fi1 was the CS-1 region previously defined as an adhesion site for melanoma cells.

The functionally defined monoclonal antibodies to a4 partially inhibited T lymphocyte adhesion to intact plasma fibronectin and had no effect on their attachment to an 80 kDa tryptic fragment containing the RGD adhesion sequence.

Monoclonal antibodies (P106 and PIFB) to the previously described fibronectin receptor, a5fll, completely inhibited T lymphocyte adhesion to the 80 kDa fragment but had no effect on their attachment to the 38 kDa fragment or to CS-1. fibroblasts which express these receptors were grown on Both c451 and a5fi1 localized to focal adhesions when Monoclonal antibodies P3E3, P4C2 N U) f&bronectin—coated surfaces. These findings demonstrated a specific interaction of both receptors with fibronectin at focal contacts.

Recently, Bernardi et al., 1987, sgpra; Liao et al., 1987, Exp. Cell, Res., 171:306-320; Liao et al., 1989, Exp. Cell Res., l81:348—361 reported that some B lymphocyte cell lines bind to a region of plasma fibronectin located within the carboxy terminal Hep II domain. 1987, sgprg identified an integrin-like receptor on B cells. However, it is not clear whether the protein they described was a4fl1, c251 or a5fil. Bernardi et al., 1987, sgpra also identified fibronectin receptors expressed by B Interestingly, in this study, B cells which Liao et al., lymphocytes. attached to fragments containing Hep II expressed a receptor similar to a4fl1 while cells which attached to the RGD containing cell adhesion domain expressed a receptor similar to aspl. However, from these data it was also not possible to clearly identify the receptor involved in binding. Together, the results of these previous reports and the present findings provide clear evidence in support of i) the existence of an alternative adhesion domain present in the carboxy terminal region of plasma fibronectin and ii) a role for a4fi1 as the receptor for this alternative adhesion site. It will be interesting to determine the precise amino acid sequences responsible for a4fl1 interaction with fibronectin. Since neither the 38 or kDa fragments or CS-1 contain an RGD sequence (Kornblihtt et al., 1985, supra: Garcia—Pardo, 1987, supra; Humphries et al., 1986, supra; and Humphries et al., 1987, ggpgg), it is clear that characterization of the ligand for a4fll will identify a new amino acid sequence important for cell adhesion to fibronectin. Since the 38 kDa fragment does not contain CS-5 (Garcia-Pardo, 1987, ggpra) the minimal amino acid sequence responsible for T lymphocyte ....4 \./‘1 N VI .._54._ dfihesion to 38 kDa and therefore the ligand for a4fll in these cells is not arg—g1u-asp-val or REDV (Humphries et 1986, supra).

Like a251, the Q4 subunit is weakly associated with the 51 subunit. and our previous findings (Wayner, E. A. and Carter, W. G., 1987, supra and Wayner et al., 1988, supra) show that the functionally defined monoclonal antibodies to azfil and a4fl1 al., The data presented here (Figure 2) selectively interact with epitopes present on the a subunits, based on immune precipitated of a2 or a4 without 51 after subunit dissociation. These results suggest that the unique a subunit is responsible for determining the This concept is now further support by the observations ligand-binding specificity of each aévb complex. presented here that a5 and a4, which are both complexed with fil, mediate adhesion to distinct sites on fibronectin.‘ This is not to suggest that the fl subunit is not important in binding, but that the specificity of receptor-ligand interactions is determined by a or a unique a-fi complex.

It is interesting that while LAX cells expressed abundant cell surface a4fll it did not appear to be a functional receptor;l P1D6 completely inhibited LAK cell adhesion to fibronectin. The reason for this could be that LAK cells express a degraded form of a4 (see Figure 2). In addition, because they are activated, LAX cells over express a5fi1 when compared to resting peripheral blood or leukemic T cells (Table VIII). express larger quantities of a5fil relative to c451 (K562-1 and HT1080) adhesion to the 80 kDa RGD containing domain via a5fl1 is dominant (see K562-1 cells, Figure 6). This implies that regulation of receptor expression determines the ability of a cell to recognize and bind to different sites on fibronectin. Furthermore, it is also possible that co—expression of the two receptors for fibronectin could increase the avidity of cell binding, for example, In other cells which Jurkat and RD cells express relatively promiscuous adhesion to fibronectin when compared to YT cells which express only a4fll.

The regulation of cell adhesion of fibronectin isgpotentially complex even under the simplest possible conditions, which assume that a5p1 and a4fll function independently of each other and do not overlap during interaction with the two binding sites on fibronectin.

Variation from this simple state provides opportunities for exquisitely sensitive regulation of cell adhesion. At the least complex level, this regulation can be roughly categorized as i) processes that control the synthesis and/or exposure of the binding sites‘onTthe ligand and (ii) regulation of functional expression of the receptors.

Examples of regulation at both levels are currently available and include, the observation that lymphokines andv specific antigen induce a5fll expression on T lymphocytes followed by increased cell adhesion to fibronectin (Wayner ; et al., 1988, supra). In addition, the control of mRNA splicing in the IIICS region of fibronectin (Kornblihtt et al., 1985, sgprg) during wound healing or inflammation may dictate the specificity of receptor—ligand binding in resting or activated T cells. Variations from the simple state are intriguing but require additional experimentation to even begin to identify the multitude of potential mechanisms.

In conclusion, these findings show clearly that cultured T lymphocytes use two independent receptors during attachment to firbonectin and that i) a5fll is the receptor for the RGD containing cell adhesion domain, and ii) a4fl1 is the receptor for a carboxy terminal cell adhesion region containing the Heparin II and IIICS domains. Furthermore, these data show that T lymphocytes express a clear preference for a region of molecular heterogeneity in IIICS fits-1) generated by alternative splicing of fibronectin pre-mRNA and that a4fil is the receptor for this adhesion site.

. EXAMPLE: LYMPHOCYTE ADHESION TO ACTIVATED ENDOTHELIUM Is MEDIATED BY THE BINDING or THE INTEGRIN RECEPTOR 0,451 TO CS-I IN THE TALTERNATIVELY SPLICED IIICS REGION OF FIBRONECTIN The following experiments demonstrated the role of the aafll receptor and its ligand, CS-1, in mediating T cell adhesion to cultured large vessel endothelial cells and endothelial cells which had been activated with a variety of cytokines associated with the inflammatory response, including IL-1, tumor necrosis factor alpha (TNFa), and tumor necrosis factor beta (TNFfi). In addition, the ability of monoclonal antibodies and peptide fragments to block adherence of lymphocytes to endothelium via the aafil receptor was demonstrated. .1. MATERIALS AND METHODS 7.1.1. REAGENTS Reagents used were as described in Section .1.1, supra. .1.2. CELLS AND CELL CULTURE Jurkat (Human T cell leukemia) was obtained from Dr. Paul Conlon (Immunex. Corp., Seattle, WA), Ramos (Human B cell Leukemia) was obtained from the American Type Culture Collection (Rockville, MD). The LAD (leukocyte adhesion deficient) and ST-1 B cell lines were prepared by Epsten-Barr virus tansformation of human B lymphocytes.

The LAD cell line was developed from the B cells of a patient with a deficiency in the $2 integrin family of adhesion receptors and was obtained from Dr. John Harlan (Harborview Medical Center, Seattle, WA). Human umbilical _57._. t . vein endothelial cells (HUVES) were purchased from Cell Systems, Seattle, WA. HUVES were maintained in defined (serum-free) media also purchased from Cell Systems (CS—100® media).

ACTIVATION OF HUVES WITH INFLAMMATORY CYTOKINES HUVES were incubated with IL-1 5 (1 ng/ml) or in some experiments with TNF a (10 ng/ml) for 6-24 hours. At the end of this incubation the HUVE monolayers were washed .1.3. and used in the adhesion assay. .1.4. SYNTHESIS OF CS PEPTIDES Peptides derived from the CS-1 region of plasma fibronectin were synthesized and HPLC‘purified according to standard protocols by Dr. James Blake at the Oncogen Corp., Seattle, WA. serum albumin or KLH also according to standard protocols The RGDS control peptide was obtained The CS—1 peptide was conjugated to rabbit by Dr. James Blake. from Peninsula Laboratories (Belmont, CA). .1.5. MONOCLONAL ANTIBODIES you \./w -58..

Human umbilical vein endothelial cells (HUVES) were cultured in 48 well plates as described (supra). To measure adherence of lymphocytes to HUVE monolayer cultures, lymphocytes were labeled with Naz 51CrO4 (50 pCi/ml) for 2-4 hours), washed, and then 105 lymphocytes were incubated with HUVE monolayers in the presence or absence of inhibitory antibodies or CS—1 derived peptides.

The lymphocytes were allowed to adhere for 30 minutes at 37'. Non-adherent cells were subsequently removed by washing with PBS, and the adherent cells were dissolved in SDS/NaOH . counter. activated prior to the adhesion assay by incubating with IL-1 3 (1 ng/ml) or TNF-fl (10 ng/ml) for 6-24 hours in defined CS—lOO®media (Cell Systems, Seattle, WA).

Bound Slcr-cpm were quantitated in a gamma In some experiments, endothélial cells were .2. RESULTS SURFACE PHENOTYPE OF LYMPHOCYTES FROM NORMAL AND LAD PATIENTS In order to establish what mechanism lymphocytes .2.1. use during extravasation we first determined the surface phenotype of normal and LAD lymphocytes with respect to the integrin receptors. These data are in Table VII and show clearly that the LAD cells possess a normal cell surface phenotype with respect to the 51 containing integrins. since, as expected, the B cells derived from the patient with LAD were negative for 52, this strongly suggests that LAD lymphocytes use the pl containing integrins during their adhesion to and passage through the endothelium. cpm = counts per minute. The data are ffbm a Single representative experiment. -61..

TO RESTING AND ACTIVATED ENDOTHELIAL CELLS Chromium-labeled lymphocytes from various cell lines were tested for their ability to adhere to either resting or activated endothelial cells (Table VII).

Although all the cell lines tested were found to adhere to some extent to resting endothelium, adhesion of T and B lymphocytes to endothelium activated by either IL-1 or TNF was observed to be much greater, by a factor of as much as ten-fold. endothelium was not found to be significantly different Adhesion of lymphocytes from LAD patients to from that observed for ST-1 cell lines derived from normal B cells (ST-1). cell lines between adhesion to IL-1 versus TNF activated There was no differencejobserved among endothelium.

EFFECTS OF ANTI-RECEPTOR ANTIBODIES ON LYMPHOCYTE ADHERENCE TO ENDOTHELIUM When the ability of chromium-labeled lymphocytes .2.3. to adhere to endothelium was tested in the presence of hybridoma supernatants, only monoclonal antibodies directed toward aafil or 51 were found to inhibit adhesion; monoclonal antibodies directed toward other receptors such as the prototype fibronectin receptor and the a3fl1 receptor were found to have virtually no inhibitory effect (Table 8). In the presence of monoclonal antibodies P4C2 (directed toward aafll) and P4C10 (directed toward 51), adhesion of labeled lymphocytes to endothelium was completely abrogated. Interestingly, adhesion of LAD cell adhesion was also inhibited by anti-aapl antibody [P4C2], indicating that the CD18 receptor is not involved in the observed adherence properties. In addition, although a251, a5fll (Table VIII) and 32 (not shown) are expressed by lymphocytes, antibodies to these receptors did not inhibit lymphocyte adhesion to either basal or activated HUVEs (see Hand \./'\ /\/:/ integrin receptor and binding of an inhibitory antibody to it does not necessarily lead to inhibition of lymphocyte binding to endothelium.

These data show that surface expression of an This implies a specific role for a451 in mediating lymphocyte adhesion to the endothelium as the first step in extravasation. Furthermore, since the antibodies against a4fll inhibited lymphocyte adhesion to endothelial cells this suggested that the ligand for a4fll, the amino acid sequence EILDVPST (see Table XII) might also be involved in lymphocyte diapedesis via binding of a4fll to this sequence present in a ligand expressed on the surface of the endothelium. g Table IX Effect of Inhibitory Monoclonal Antibodies on Lymphocyte Adhesion to HUVE Monolayers (IL-lfi Activated) Adhesion (cpm) Ce1l‘Line Antibody Specificity Basal +IL—1fi LAD (B) SP2 - 19542 104672 P1D6 asfll 15688 113696 P185 azfil, 19064 90912 ,0 p4c2 (1451 6458 38132 P4C1O p1 6360 52552 Ramos (B) SP2 - 972 12157 P1D6 asfil 808 11196 ,5 PIBS .1251 124 10028 P4C2 a4fl1 456 3688 P4C1O 51 604 3152 Jurkat (T) SP2 - 83924 372159 P1D6 asfll 83956 417588 P185 azfll 66580 489952 P4C2 a4fl1 23108 136632 P4C10 51 36892 230416 (U U\ _64_.

THE ROLE or cs-1 AS LIGAND TO a ,9 IN LYMPHOCYTE ADHESION TO ENooTHELfUi°x The ability of synthetic CS-1 and derivative .2.4. peptides to inhibit adherence of chromium-labeled lymphocytes to activated endothelial cells was evaluated using various peptides. The synthetic CS-1 peptide was a strong inhibitor of T or B lymphocyte adhesion to basal or activated endothelial cell monolayers (Tables X and XI).

Interstingly, the EILDVPST sequence was the minimal peptide also required to inhibit lymphocyte adhesion to resting or activated HUVEs (Tables X and XI). with the Ramos B cell line, adhesion or these cells to HUVEs could be completely abrogated withtthe EILDVPST In control experiments (Table XI), the RGDS sequence which is the ligand for the prototype fibronectin In some cases, such as peptide. receptor, asfll, did not inhibit lymphocyte adhesion to resting or activated HUVES.

%L Table X ‘‘ Effect of cs—1 and Cs~1 Derived peptides on Lymphocyte Adhesion to HUVE’s Cell Line Peptide # Sequence Basal LAD (B) 293A Unrelated 20856 344 cs—1» 17500 ,0 350 VpST NDa 352 EILDVPST ND 354 GPEILDVPSI_, ND Ramos (B) 293A Unrelated 4856 344 cs—1 1660 350 VPST ND 352 EILDVPST ND 354 GPEILDVPST ND 20Jurkat (T) 293(A) Unrelated 58084 344 CS-1 29568 350 VPST ND 352 EILDVPST ND 354 GPEILDVPST ND aND=ndt determined Adhesion (cpm) +IL-15 42728 29484 27219 4568 2584 2265 127544 93056 897 Table XI Effect of CS-1 and CS-1 Derived Peptides or RGDS on Lymphocyte Adhesion to HUVES Adhesion (cpm) Ce11 Line Peptide # Sequence Basal Jurkat (T) — — 161092 - RGDS 298688 344 CS-1 82404 ,0 350 VPST 203716 351 LDVPST 166948 352 EILDVPST,,_:h 84456 LAD (B) - - 44860 ,5 - RGDS 70652 344 CS-1 22976 350 VPST 38176 351 LDVPST 39700 352 EILDVPST 29964 Ramos (B) - RGDS 16920 344 CS-1 1844 350 VPST 4168 351 LDVPST 3532 352 EILDVPST 16 +IL-1B 248976 322208 3262CC 234796 51560 98860 92792 58784 5160 15320 15092 49Table XII INHIBITION OF LYMPHOCYTE ADHESION TO FIBRONECTIN WITH PEPTIDES DERIVED FROM CSB12 PEPTIDE SEQUENCE INHIBITION CS-1 +++ A13 DELPQLVTLPHPN - B13 LHGPEILDVPST’ +++ 10350 VPST — 351 LDVPST - 352 EILDVPST "*‘t. +++ 354 GPEILDVPST +++ -68.. z 7.3. DISCUSSION Polymorphonucleated leukocytes (neutrophils) from patients with leukocyte adhesion deficiency (LAD) have a defect in expression of the 32 integrin subunit and therefore cannot use the 52 containing receptors (LFA-1, Mac-1 or p 150/95) in their adhesion to the vascular endothelium. Neutrophils from these patients therefore, do not leave the blood stream to pass into peripheral tissues.

LAD lymphocytes, however, do undergo diapedesis to pass through the endothelium and can be found in tissues derived from patients with this disorder. This, therefore, implies that lymphocytes use a mechanism distinct from the 52 containing integrins during their passage from the blood séream into the peripheral tissues. The following series of experiments comprises our attempts to fully understand the mechanisms utilized by peripheral blood lymphocytes during diapedesis. _ The experiments described supgg have clearly shown the important role played by the aafll receptor in the adhesion of lymphocytes to vascular endothelial cells.

All lymphocyte cell lines tested were shown to express aafll and/or a5fl1 by fluorescence analysis, and were observed to adhere to cultured human umbilical*vein endothelial cells. This adhesion was found to be blocked only by monoclonal antibodies directed toward aafll: antibodies directed toward other receptors were not found to have essentially any inhibitory effect, revealing the importance of the a4fl1 receptor in the adhesive interaction between lymphocytes and endothelium.

In addition, synthetic CS-1 and derivative peptides (Tables 9, 10, and 11) were found to inhibit adhesion of lymphocytes to endothelium. The amino acid sequence EILDVPST was found to be particularly important to the interaction. It must be emphasized that it has not been determined whether the lymphocyte ahfil receptor is, in fact, interacting with fibronectin on the endothelial cell surface. It is also possible that aafll is recognizing the peptide EILDVPST or a similar sequence, in the context of another, non-fibronectin protein. -70.. $e 8. DEPOSIT 0? CELL LINES The following cell lines have been deposited with the ATCC, Rockville, MD, and have been assigned the following accession numbers: Cell Line Accession Number P4C2 HB—10215 P4G9 H8-10213 P3E3 HB—l0212 P4C1O HB—10214

Claims (16)

IS CLAIMED IS:

1. Use of an antibody, or fragment or derivative thereof which binds to the a4B1 receptor and inhibits the adherence of nucleated hematopoietic cells to vascular endothelial cells for the preparation of a pharmaceutical composition for use in a mammal to suppress an immune response.

2. The use according to claim 1 wherein the antibody is a monoclonal antibody.

3. The use according to claim 1 or 2 wherein the antibody binds to the a4 subunit of the d4Bl receptor.

4. The use according to claim 2 or 3 wherein the monoclonal antibody is P4C2, produced by the hybridoma deposited with the American Type Cultufe7€ollection and having the accession number HB10215.

5. The use according to claim 1 or 2 wherein the antibody binds to the B1 subunit of the d4B1 receptor.

6. The use according to claim 2 or 5 wherein the monoclonal antibody is P4C10, produced by the hybridoma deposited with the American Type Culture Collection and having the accession number HB102l4.

7. The use according to any of the preceding claims wherein the antibody fragment is a Fab, Fab’ or F(ab)2.

8. The use according to any of the preceding claims wherein the antibody derivative is a chimeric antibody or a recombinant antigen binging region.

9. The use according to any of the preceding claims wherein the suppression of the immune response is a treatment for allergy, asthma, or a chronic inflammatory skin condition.

10. The use according to any of the preceding claims wherein the pharmaceutical composition is formulated for subcutaneous, intramuscular, intravascular, intravenous, ingraarterial, intranasal, oral intraperitoneal, intratrachael, or intrathecal administration.

11. The use according to any one of the preceding claims wherein the nucleated hematopoietic cell is a lymphocyte or a monocyte.

12. The use according to any of the preceding claims wherein the mammal is human.

13. Process for the preparation of a pharmaceutical composition for suppressing an immune response in a mammal characterised in that as an essential constituent of said pharmaceutical composition an antibody or fragment or derivative thereof as defined in any one of claims 1 to 12 is used.

14. Use according to Claim 1 substantially as hereinbefore described by way of Example.

15. The process of Claim 13 substantially as hereinbefore described by way of Example.

16. The produce of a process according to claim 13 or 15 substantially as hereinbefore described by way of example.