AU4734099A - Mucosal vascular addressins and uses thereof - Google Patents

Description

I I M S F Ref: 388163D1

AUSTRALIA

PATENTS ACT 1990 COMPLETE SPECIRCATION FOR A STANDARD PATENT

ORIGINAL

c

MM

ft..

Name and Address of Applicant: ft .Me ft Actual Inventor(s): Leukosite Inc.

215 First Street Cambridge Massachusetts 02142 UNITED STATES OF AMERICA Michael J. Briskin, Douglas J. Ringler and Dominic Picarella Spruson Ferguson, Patent Attorneys Level 33 St Martins Tower, 31 Market Street Sydney, New South Wales, 2000, Australia Mucosal Vascular Addressins and Uses Thereof Address for Service: Invention Title: The following statement is a full description of this invention, including the best method of performing it known to me/us:- 5845 1 Mucosal Vascular Addressins and Uses Thereof Description Background of the Invention Lymphocyte homing from the circulation to the lymphoid tissues and migration to sites of inflammation is regulated by interaction with receptors expressed in postcapillary venules, including high endothelial venules (HEV) found in secondary lymphoid tissues mesenteric lymph nodes, Peyer's Patches (Bevilacqua, Annu. Rev.

Immunol.. 11. 767-804 (1993); Butcher, Cell, 67: 1033-1036 (1991); Picker, et al., Annu. Rev. Immunol., 10: 561-591 (1992); and Springer, Cell, 76: 301-314 1i (1994)). These interactions are tissue specific in nature.

Inflammation chronic inflammation) is characterised by infiltration of the affected tissue by leukocytes, such as lymphocytes, lymphoblasts, and mononuclear phagocytes. The remarkable selectivity by which leukocytes preferentially migrate to 15 various tissues during both normal circulation and inflammation results from a series of *:oo adhesive and activating events involving multiple receptor-ligand interactions as proposed by 0 •00 *o 0 Butcher and others (Butcher, Cell, 67: 1033-1036 (1991); vonAdrian, et al., Proc. Natl. Acad. Sci.

USA, 88: 7538 (1991); Mayadas, et al., Cell, 74: 541 (1993); Springer, Cell, 76: 301 (1994)). As an initial step, there is a transient, rolling interaction between leukocytes and endothelium, which results from the interaction of selectins (and by a4 integrins in some instances) with their carbohydrate ligands. This interaction which is characterized by rolling in the direction of flow can be assessed by known methods (Lawrence, M.B. and T.A. Springer, Cell, 65: 859 (1991); WO 92/21746, Springer et al., (December 10, 1992)). This is followed by activation events mediated by chemoattractants such as chemokines and their receptors, which cause activation of integrin adhesiveness and influence the direction of migration of leukocytes through vascular walls. Such secondary signals in turn trigger the firm adhesion of leukocytes to endothelium via leukocyte integrins and their endothelial ligands (Ig-like receptors and the ECM), and subsequent transendothelial migration from the circulation across the vascular endothelium.

In secondary lymphoid tissues, such as Peyer's patches (PPs) and lymph nodes peripheral lymph nodes (PLN)), leukocyte trafficking and homing is regulated by 25 interactions of homing receptors on the surface of leukocytes with endothelial cells lining the post-capillary venules, notably high endothelial venules (HEV) (Gowans, J.L. and E.J. Knight, Proc. R. Soc. Lond., 159: 257 (1964)). Receptors termed vascular addressins, which are present on the endothelial cell surface and regulate the migration and subsequent extravasation of lymphocyte subsets. The vascular addressins show restricted patterns of expression and this tissue specific expression makes an important contribution to the specificity of leukocyte trafficking (Picker, L.J. and E.C. Butcher, Annu. Rev.

-3- Immunol., 10: 561-591 (1992); Berg, et al., Cellular and molecular mechanisms of inflammation, 2: 111 (1991); Butcher, Cell, 67: 1033-1036 (1991)).

Mucosal vascular addressin MAdCAM-1 (Mucosal Addressin Cell Adhesion Molecule-l) is an immunoglobulin superfamily adhesion receptor for lymphocytes, which is distinct from VCAM-1 and ICAM-1. MAdCAM-1 was identified in the mouse as a -60 kd glycoprotein which is selectively expressed at sites of lymphocyte extravasation. In particular, MAdCAM-1 expression was reported in vascular endothelial cells of mucosal tissues, including gut-associated tissues or lymphoid organs, such as Peyer's patches and venules of the lamina propria of the small and large intestine, and the lactating mammary gland, but not in peripheral lymph nodes.

15 MAdCAM-1 is involved in lymphocyte binding to Peyer's Patches. (Streeter, et al., Nature, 331: 41-46 (1988); Nakache, et al., Nature, 337: 179-181 (1989); Picker, et al., Annu. Rev. Immunol., 10: 561-591 (1992); Briskin, et al., Nature, 363: 461 (1993); 20 Berg, et al., Nature, 366: 695-698 (1993); Berlin, SC., et al., Cell, 74: 185-195 (1993)). MAdCAM-1 can be induced in vitro by proinflammatory stimuli (Sikorski, et a l J. Immunol., 151: 5239-5250 (1993)).

MAdCAM-1 specifically binds the lymphocyte integrin a4I7 (also referred to as LPAM-1 (mouse), a4fp (mouse)), which is a lymphocyte homing receptor involved in homing to Peyer's patches (Berlin, et al., Cell, 80: 413-422 (1994); Berlin, et al., Cell, 74: 185-195 (1993); and Erle, et al., J. Immunol., 153: 517-528 (1994)). In contrast to VCAM-1 and fibronectin, which interact with both a401 and a467 (Berlin, et al., Cell, 74: 185-195 (1993); Strauch, et al., Int. Immunol., 6: 263 (1994)), MAdCAM-1 is a selective receptor for a47.

Inflammatory bowel disease (IBD), such as ulcerative colitis and Crohn's disease, for example, can be a -4debilitating and progressive disease involving inflammation of the gastrointestinal tract. Affecting an estimated two million people in the United States alone, symptoms include abdominal pain, cramping, diarrhea and rectal bleeding.

IBD treatments have included anti-inflammatory drugs (such as, corticosteroids and sulfasalazine), immunosuppressive drugs (such as, 6-mercaptopurine, cyclosporine and azathioprine) and surgery (such as, colectomy). Podolsky, New Engl. J. Med., 325: 928-937 (1991) and Podolsky, New Engl. J. Med., 325: 1008-1016 (1991).

Some studies have suggested that the cell adhesion molecule, ICAM-l, mediates leukocyte recruitment to inflammatory sites through adhesion to leukocyte surface ligands, Mac-l or LFA-1 (Springer, Nature, 346: 425- 15 434 (1990)). In addition, vascular cell adhesion molecule- 1 (VCAM-1), which recognizes the a4l1 integrin (VLA-4), has been reported to play a role in in vivo leukocyte recruitment (Silber et al., J. Clin. Invest. 93: 1554-1563 (1994)). It has been proposed that IBD can be treated by 20 blocking the interaction of ICAM-1 with LFA-1 or Mac-1, or of VCAM- 1 with a4e 1 WO 93/15764). However, these therapeutic targets are likely to be involved in inflammatory processes in multiple organs, and a functional blockade could cause systemic immune dysfunction.

In contrast to VCAM-1 and ICAM-l, MAdCAM is preferentially expressed in the gastrointestinal tract, binds the a407 integrin found on lymphocytes, and participates in the homing of these cells to mucosal sites, such as Peyer's patches in the intestinal wall (Hamann et al., Journal of Immunology, 152: 3282-3293 (1994)). The use of inhibitors to the binding of MAdCAM to the receptor, a407, in the treatment of diseases such as IBD has not been suggested. Moreover, although human a4 and 67 genes and proteins have been identified (Yuan et al., Int. Immunol., 2: 1097-1108 (1990); Erle et al., J. Biol. Chem., 266: 11009-11016 (1991); Bevilacqua, Annu. Rev. mmunol., 11: 767-804 (1993); Springer, Cell, 76: 301-314 (1994)), human or primate MAdCAM-1 has not been cloned or characterized.

Summary of the Invention The present invention relates to proteins or polypeptides, referred to herein as isolated and/or recombinant essentially pure) primate MAdCAMs.

In one embodiment, primate MAdCAM can selectively bind to cells which express the a4f7 integrin, particularly lymphocytes. The recombinant proteins of the present invention, including variants, can be produced in host cells as described herein. In addition, antibodies reactive with the proteins of the present invention can be produced using a primate MAdCAM or a variant thereof as immunogen, for example. Such antibodies or fragments thereof are useful in therapeutic, diagnostic and research applications. For example, the antibodies can be used in ~the purification and study of MAdCAMs, the identification of cells which express MAdCAM, and the detection or quantitation of MAdCAM in a sample.

::The invention further relates to isolated and/or recombinant essentially pure) nucleic acids which encode a primate MAdCAM, such as human MAdCAMs. In another aspect, the invention relates to recombinant nucleic acid constructs, such as plasmids or retroviral vectors, which contain a nucleic acid which encodes a protein of the present invention or portion thereof. The nucleic acids and constructs can be used to produce recombinant primate MAdCAMs. In another embodiment, the nucleic acid encodes an antisense nucleic acid which can hybridize with a second nucleic acid encoding a primate MAdCAM, and which can inhibit the expression of the protein when introduced into cells).

-6- Also encompassed by the present invention are methods of identifying ligands and/or inhibitors antagonists) of MAdCAM function. For example, primate MAdCAM, including variants, can be used in assays adhesion assays) designed to identify antagonists which block the binding of MAdCAM to the ligand, a407 integrin.

The invention further relates to methods of therapy, including a method of treating an individual suffering from a disease associated with leukocyte (such as lymphocyte or monocyte) recruitment to the gastrointestinal tract or other tissues as a result of binding of leukocytes to gut-associated endothelium expressing the molecule MAdCAM, comprising administering to the individual a mammal, such as a primate) an effective amount of an agent or 15 compound, such as an antibody, which inhibits the binding of leukocytes to endothelial MAdCAM. The antibody is preferably a monoclonal, chimeric and/or humanized antibody or an antigen binding fragment thereof, and inhibits adhesion of leukocytes expressing an integrin containing 20 the f7 chain (such as a4 f7) to endothelium expressing S* MAdCAM. In one embodiment, the monoclonal antibody or antigen binding fragment thereof has the antigenic specificity of a monoclonal antibody selected from the group consisting of FIB 21, FIB 30, FIB 504 and ACT-1.

Inflammatory bowel diseases, such as, but not limited to, ulcerative colitis, Crohn's disease, Pouchitis, celiac disease, microscopic or collagenous colitis, and eosinophilic gastroenteritis can be treated according to the claimed method.

Brief Description of the Drawings Figure 1 is an illustration of the nucleotide sequence (SEQ ID NO:1) determined from subclones of cDNA clone 4 encoding human MAdCAM-1, and the sequence of the predicted protein encoded by the open reading frame (MAdCAM-1; SEQ ID -7- NO:2). The predicted signal peptide and transmembrane region are underlined in bold. Cysteine residues of the two Ig-like domains are boxed, as are potential N-linked glycosylation sites. The mucin domain, containing the PPDTTS(Q/P)E repeat consisting of 71 amino acids is outlined by a thin bold line.

Figure 2 is an illustration of the nucleotide sequence (SEQ ID NO:3) determined from subclones of cDNA clone encoding human MAdCAM-1, and the sequence of the predicted protein encoded by the open reading frame (MAdCAM-1; SEQ ID NO:4). The predicted signal peptide and transmembrane region are underlined in bold. Cysteine residues of the two Ig-like domains are boxed, as are potential N-linked glycosylation sites. The mucin domain, containing the PPDTTS(Q/P)E repeat consisting of 47 amino acids is outlined by a thin bold line.

Figure 3 is an illustration of the nucleotide sequence (SEQ ID NO:5) determined from subclones of cDNA clone 31D encoding macaque MAdCAM-1, and the sequence of the 20 predicted protein encoded by the open reading frame (MAdCAM-1; SEQ ID NO:6). The predicted signal peptide and transmembrane region are underlined in bold. Cysteine residues of the two Ig-like domains are boxed. The mucin domain, which contains a single PPDTTS(Q/P)E repeat, is 25 outlined by a thin bold line.

Figures 4A-4B are histograms illustrating the selective binding of cells transfected with human MAdCAM-1 to lymphocytes expressing a4#7. Figure 4A illustrates the results of an experiment in which RPMI 8866 cells (0.5 X 10 6 /well), which express a4#7 (and not a4l), bound to CHO/P cells expressing murine or human MAdCAM-1, but did not bind to CHO/P cells transfected with human VCAM-l or to CHO/P cells transfected with pcDNA-3. Figure 4B illustrates the results of an experiment in which CHO/P 1 -8cells transfected with human VCAM- 1 bound to Jurkat cells (which express high levels of a4l), but failed to bind to CHO/P cells transfected with murine or human AdCAMi or to CHO/P cells transfected with pcDNA-3 as a control. Binding is shown as the number of bound RPMI 8866 cells per CH/ cell (Figure 4A) or bound Jurkat cells per CHO/P cell (Figure 4B) in an average of at least four fields (0X objective) standard error. Binding reactions included control IgG, anti-44#7 (monoclonal antibody ACT-i), or anti-murine MAdCAM-1 (monoclonal antibody MECA-367) as indicated.

Figure 5 is a histogram illustrating that human experiment with standard deviations as shown.

Figure 6 is an illustration of the deduced domain structures of murine and human MAdCAM-. The two Nterminal immunoglobulin dominns bounded by disulfide bonds (indicated by loops) implicated in cell adhesion S- transmembrane regions and a cytoplasmic tail are present in murine, macaque and human proteins. Huma MAdCAM- 1 hs l o n g er cytoplasmic tail A bhn a Srlonger cytoplasmic tail. An eight-amino acid repeat found i* in the mucin domain is present in 4 or 8 copies in human isoforms, but appears only once in the murine and macaque.

h Figures 7A and 7B are graphic illustrations of histologic scores of inflammatory activity and epithelial injury from left (descending) and right (ascending) colon of mice exposed to 10 days of DSS in their drinking water.

Three groups of mice are shown, consisting of groups receiving an irrelevant rat IgG2a antibody FIB 21 or FIB 30 antibodies.

Figure 8 is a graph of y counts per minute (cpm) 1 SEM) as a percentage of total input counts from mice given DSS in the drinking water for 10 days. Six groups -9consisted of negative controls given water alone, positive controls given DSS alone, test groups given irrelevant rat IgG2a antibody, FIB21, MECA-367, or FIB21 with MECA-367 Figure 9 is a graph depicting the histologic scores 1 SEM) for villus fusion obtained from jejunal biopsy samples of common marmosets before and on the 14th day of treatment with 2 mg/kg/day of ACT-1 monoclonal antibody.

Figure 10 is a graph depicting the histologic scores 1 SEM) for villus atrophy obtained from jejunal biopsy samples of common marmosets before and on the 14th day of treatment with 2 mg/kg/day of ACT-I monoclonal antibody.

Figure 11 is an illustration of the stool consistency t of colitic animals (cotton-top tamarins) treated with

ACT-I

Santibody.

V *15 Figure 12 is an illustration of the inflammatory activity in colitic animals (cotton-top tamarins) treated with ACT-I antibody as assessed histologically.

Figure 13 is a diagram illustrating the experimental S::0 protocol for treatment of chronically colitic cotton-top 20 tamarins with ACT-I antibody.

Figure 14 is a graph illustrating the therapeutic effect on stool consistency of administration of ACT-I antibody or an irrelevant, isotype-matched antibody to chronically colitic cotton-top tamarins.

Figure 15 is a histogram illustrating the therapeutic effect of ACT-I immunotherapy on colonic inflammatory activity in chronically colitic colitic cotton-top tamarins treated with ACT-i antibody or an irrelevant control monoclonal antibody. Each bar represents the mean within a treatment group of the absolute change in the inflammatory activity score 1 SEM, computed for each animal by comparing the score at a particular time point with the same animal's score on Day 0.

Figure 16 is a histogram illustrating the absolute numbers of a4f7+ lymphocytes in the peripheral circulatory pool in colitic cotton-top tamarins treated with ACT-1 antibody. Each bar represents the mean 1 SEM) of the concentration of a4?7+ cells in blood as detected by ACT-i.

Figures 17A-17E are plots illustrating the results of a FACS analysis showing the specific staining of Hut 78 cells with MAdCAM-Ig chimeras. Supernatants from COS cells transiently transfected with a human MAdCAM-Ig chimeric constructs (from four independent transfections) were incubated with HuT 78 cells in the presence of 2 mM Mn++, and bound chimera was detected using a phycoerythrinconjugated antibody specific for human IgGl. Figure 17A, media control; Figures 17B-17C, supernatants from cells transfected with Clone 21 (comprising the entire extracellular domain of human MAdCAM); Figures 17D-17E, 15 supernatants from cells Clone 38 (comprising the two N-terminal Ig domains of human MAdCAM). Binding after preincubation of cells with media alone (at right).

Binding was inhibited by preincubation of cells with anti- 7 MAb FIB 504 (at left).

20 Figure 18 is a graph illustrating the difference in body weight of scid mice reconstituted with 1 X 106 ~T cells relative to control scid mice reconstituted with an equal number of CD45RB° T cells derived from BALB/c spleen. Recipient mice were weighed at weekly intervals to evaluate progression of disease.

o Figure 19 is a histogram illustrating the increased accumulation in the colon of intravenously injected 1"In-labeled mesenteric lymph node cells in scid mice reconstituted with CD45RBhi T cells as compared with the accumulation in colon of scid mice reconstituted with an equal number of CD45RBIO T cells, and the inhibition of accumulation by treatment for 2 weeks with a combination of anti-#7 (FIB 504) and anti-MAdCAM (MECA-367) monoclonal antibodies. Results are expressed as counts per minute -11- (CPM) in colon normalized to CPM in spleen and corrected for background.

Figure 20 is a histogram illustrating the complete inhibition accumulation of "'In-labeled cells (injected intravenously) in the colon of scid mice by treatment for 4 months (starting from the time of reconstitution) with FIB 504, MECA-367, or FIB 504 plus MECA-367. From left to right: scid mice reconstituted with CD45RB'o T cells, receiving irrelevant isotype-matched control rat IgG2a; scid mice reconstituted with CD45RBhi T cells, receiving either irrelevant isotype-matched control rat IgG2a, FIB 504, MECA-367, or FIB 504 MECA-367.

Figure 21 is a histogram illustrating the reduction in S: ~the number of CD4+ T cells in the ascending (right) or 15 descending (left) colon in scid mice treated for 14 days with a combination of FIB 504 plus MECA-367 as compared to mice treated with an isotype-matched control rat IgG2a antibody as determined by staining frozen sections of left and right colon with a rat antibody specific for mouse CD4.

20 Detailed Description of the Invention Proteins and Peptides The present invention relates to isolated and/or recombinant (including, essentially pure) proteins or polypeptides designated primate MAdCAMs (Mucosal Addressin Cell Adhesion Molecules) and variants of primate MAdCAMs.

In a preferred embodiment, the isolated and/or recombinant proteins of the present invention have at least one property, activity or function characteristic of a primate MAdCAM (as defined herein), such as binding function the ability to bind an a407 integrin), and/or cellular adhesion molecule function the ability to mediate cellular adhesion such as a407-dependent adhesion in vitro j -12and/or in vivo), and/or an immunological property as defined herein. For example, some proteins of the present invention can selectively bind to an a4,7 integrin and thereby mediate a4f7-dependent cellular adhesion to cells bearing the a4f7 integrin, such as leukocytes (especially lymphocytes such as T or B cells) in vitro and/or in vivo.

In one aspect, proteins of the present invention can mediate heterotypic cell adhesion of endothelial cells to leukocytes such as lymphocytes).

In another embodiment, proteins of the present invention can bind a primate a4f7 integrin from the same or a different primate species, and/or have cellular adhesion .molecule function the ability to mediate cellular adhesion such as a4f7-dependent adhesion in vitro and/or in 15 vivo). For example, as shown herein, human and macaque MAdCAM-1 proteins, produced in mammalian cells by expression of cDNA clones, can selectively bind to a4f7 ntegrin present on human lymphocytes, and can function as adhesion molecules capable of mediating selective 20 adhesion to cells bearing the a4/7 integrin.

Proteins or polypeptides referred to herein as "isolated" are proteins or polypeptides purified to a state beyond that in which they exist in mammalian cells.

"Isolated" proteins or polypeptides include proteins or 25 polypeptides obtained by methods described herein, similar methods or other suitable methods, including essentially pure proteins or polypeptides, proteins or polypeptides produced by chemical synthesis synthetic peptides), or by combinations of biological and chemical methods, and recombinant proteins or polypeptides which are isolated.

The proteins can be obtained in an isolated state of at least about 50 by weight, preferably at least about 75 by weight, and more preferably, in essentially pure form.

Proteins or polypeptides referred to herein as -13- "recombinant" are proteins or polypeptides produced by the expression of recombinant nucleic acids.

As used herein "primate MAdCAM" refers to naturally occurring or endogenous primate MAdCAM proteins, to proteins having an amino acid sequence which is the same as that of a naturally occurring or endogenous corresponding primate MAdCAM recombinant proteins), and to functional variants of each of the foregoing functional fragments and/or mutants produced via mutagenesis and/or recombinant techniques). Accordingly, as defined herein, the term includes mature primate MAdCAM, S. glycosylated or unglycosylated MAdCAM proteins, polymorphic or allelic variants, and other isoforms of primate MAdCAM produced by alternative splicing or other cellular 15 processes), and functional fragments.

Naturally occurring or endogenous primate MAdCAM proteins includes wild type proteins such as mature MAdCAM, polymorphic or allelic variants and other isoforms which occur naturally in primates humans or other 20 non-human primates, such as macaque, cotton top tamarin) Such proteins can be recovered from a source which naturally produces primate MAdCAM. These proteins and primate MAdCAM proteins having the same amino acid sequence as a naturally occurring or endogenous corresponding 25 primate MAdCAM, are referred to by the name of the corresponding primate. For example, where the corresponding primate is a human, the protein is designated as a human MAdCAM protein a recombinant human MAdCAM produced in a suitable host cell).

"Functional variants" of primate MAdCAMs include functional fragments, functional mutant proteins, and/or functional fusion proteins. Generally, fragments or portions of primate MAdCAM encompassed by the present invention include those having a deletion one or more deletions) of an amino acid one or more amino -14acids) relative to the mature primate MAdCAM (such as N-terminal, C-terminal or internal deletions). Fragments or portions in which only contiguous amino acids have been deleted or in which non-contiguous amino acids have been deleted relative to mature primate MAdCAM are also envisioned.

Generally, mutants or derivatives of primate MAdCAMs, encompassed by the present invention include natural or artificial variants differing by the addition, deletion and/or substitution of one or more contiguous or non-contiguous amino acid residues, or modified polypeptides in which one or more residues is modified, and mutants comprising one or more modified residues.

Preferred mutants are natural or artificial variants of primate MAdCAM differing by the addition, deletion and/or substitution of one or more contiguous or non-contiguous amino acid residues.

A "functional fragment or portion", "functional mutant" and/or "functional fusion protein" of a primate 20 MAdCAM refers to an isolated and/or recombinant protein or oligopeptide which has at least one property, activity and/or function characteristic of a primate MAdCAM, such as binding function the ability to bind an a4/7 integrin), and/or cellular adhesion molecule function the ability to mediate cellular adhesion such as 47-dependent adhesion in vitro and/or in vivo), and/or retains at least one immunological property of a primate MAdCAM.

As used herein, a protein, polypeptide or oligopeptide having "at least one immunological property" of a primate MAdCAM is one which is bound by at least one antibody of a selected epitopic specificity which binds to a naturally occurring or endogenous primate MAdCAM or to a protein having the same amino acid seqence as the naturally occurring or endogenous primate MAdCAM human __1_1 MAdCAM-1), and/or is an immunogen capable of inducing the formation in a suitable animal of an antibody of a selected epitopic specificity which binds to a naturally occurring or endogenous primate MAdCAM or to a protein having the same amino acid seqence as the naturally occurring or endogenous primate MAdCAM. For example, a suitable fragment can cross-react with an antibody which is raised against and/or reactive with isolated primate MAdCAM.

Suitable fragments or mutants can be identified by screening. For example, the N-terminal, C-terminal, or internal regions of the protein can be deleted in a stepwise fashion and the resulting protein or polypeptide can be screened using a suitable binding or adhesion assay, 15 such as an assay described herein. Where the resulting protein displays activity in the assay, the resulting protein ("fragment") is functional. Information regarding the structure and function of murine MAdCAM and other adhesion molecules, and of primate MAdCAMs as shown herein, provides a basis for dividing primate MAdCAM into functional domains (see below).

The term variant also encompasses fusion proteins, comprising a primate MAdCAM mature human MAdCAM-1) as a first moiety, linked to a second moiety not occurring 25 in the primate MAdCAM as found in nature. Thus, the second moiety can be an amino acid, oligopeptide or polypeptide.

The first moiety can be in an N-terminal location, C-terminal location or internal to the fusion protein. In one embodiment, the fusion protein comprises a human MAdCAM or portion thereof as the first moiety, and a second moiety comprising a linker sequence and affinity ligand an enzyme, an antigen, epitope tag).

In another embodiment, the fusion protein is a hybrid immunoglobulin, such as a hybrid comprising a primate MAdCAM moiety fused at its C-terminus, to the N-terminus of -16an immunoglobulin moiety one or more immunoglobulin constant regions, preferably of primate origin), such as those prepared according to Capon et al., U.S. Patent No.

5,428,130 and 5,225,538, the teachings of which are incorporated herein by reference in their entirety). The hybrid immunoglobulin comprises a fusion protein or polypeptide containing at least a portion of an immunoglobulin chain, and preferably at least one complete immunoglobulin domain CH1, hinge). Other examples of "immunoadhesins" have been reported (Watson, et al., Nature, 349: 164-167 (1991); Martin, et al., J.

Virol., 67: 3561-3568 (1993); Staunton, et al., J.

Exp. Med., 176: 1471-1476 (1992); Capon, et al., Nature, 337: 525-531 (1989); Jakubowski et al., J.

15 Immunol., 155: 938-946 (1995)). For example, a fusion protein comprising all or a portion of a primate human) MAdCAM and an immunoglobulin heavy or light chain constant region or portion thereof can be prepared by preparing a nucleic acid which encodes the fusion 20 protein). Typically, the fusion is constructed such that the C-terminal end of the MAdCAM is joined to the N-terminal end of the immunoglobulin constant region.

However, fusion proteins in which the N-terminal end of the MAdCAM is joined to the C-terminal end of the 25 immunoglobulin constant region can be made. Preferably, a portion of primate MAdCAM which is sufficient for binding to a ligand a4f7 integrin), such as the complete extracellular domain or a portion comprising the two N-terminal immunoglobulin domains in which the transmembrane region is deleted, is used (see e.g., Example 3).

A variety of hybrid immunoglobulin molecules can be produced monomeric, homodimeric, heterodimeric, homotetrameric, heterotetrameric), depending upon the type of constant region selected and the portion used -17light chain constant region, heavy chain constant region (such as 71, 72, 73, 74, al, a2, 6, e, and g constant regions obtained from IgG, IgA, IgD, IgE, or IgM), and portions thereof) in the fusion polypeptide, and whether they are assembled into multimeric forms with each other and/or with other hybrid immunoglobluins or immunoglobulin chains (see Capon et al., U.S. Patent No. 5,428,130 and 5,225,538). In a preferred embodiment, the fusion protein comprises a complete heavy chain constant region or at least a functionally active hinge region, CH2 and CH3 domain. A particular constant region IgG1), variant or portions thereof can be selected to tailor effector function. For example, an mutated constant region (variant) can be incorporated into a fusion protein to 15 minimize binding to Fc receptors (Example 3; Winter et al., GB 2,209,757 B; and Morrison et al., WO 89/07142), and/or to fix complement (WO 94/29351, December 22, 1994), etc.

Examples of "primate MAdCAM" proteins include proteins encoded by a human or macaque MAdCAM-1 nucleic acid of the present invention, such as a protein having an amino acid sequence as set forth or substantially as set forth in Figure 1 (SEQ ID NO:2), Figure 2 (SEQ ID NO:4) or Figure 3 (SEQ ID NO:6), and functional portions thereof. In a preferred embodiment, a primate MAdCAM or variant has an 25 amino acid sequence which is at least about 55% similar, more preferably at least about 75% similar, and still more preferably at least about 90% similar, to a protein shown in Figure 1 (SEQ ID NO:2), Figure 2 (SEQ ID NO:4) or Figure 3 (SEQ ID NO:6).

MAdCAM Structure Murine MAdCAM-1, a member of the immunoglobulin supergene family, is a multi-domain molecule, comprising both immunoglobulin-related and mucin-like sequences (Briskin, et al., Nature, 363:461 (1993)). As -18indicated in Figure 6, the murine form contains two aminoterminal immunoglobulin-like domains are homologous to domains of the Ig-like adhesion receptors, ICAM-1 and VCAM-1, and are implicated in integrin binding. The third (membrane proximal) immunoglobulin-like domain, while unrelated to adhesion receptors of this class, shares homology with another mucosal-related immunoglobulin superfamily member, IgA. In addition to the three immunoglobulin-like domains, murine MAdCAM-1 has a serine/threonine-rich mucin-like domain between the second and third Ig-like domains. These structural elements suggest that MAdCAM-1 facilitates more than one function in cell adhesion cascades, and recent studies of murine MAdCAM-1 support a role for MAdCAM-1 in both selectin and 15 integrin binding (Moore, et al., J. Cell. Biol., 118:445 (1992); Bargatze, et al., Immunity, 3:99-108 (1995)). Also in this regard, it has been reported that murine MAdCAM-1, when expressed in mesenteric lymph nodes can present L-selectin binding carbohydrates associated 20 with the peripheral node addressin epitope, MECA-79 (Berg, et al., Nature, 366:695 (1993)).

As described herein human and macaque MAdCAM-1 proteins have two immunoglobulin-like (Ig-like) domains which are homologous to the two amino-terminal immunoglobulin-like integrin binding domains of murine MAdCAM-1 (Figures 1-3, and However, the similarity of sequences within the region homologous to the mucin/IgA domain of murine MAdCAM-1 is much less apparent. The membrane proximal regions of the human and macaque receptors exhibit considerable variation (as compared with each other or murine MAdCAM-1) with respect to the length of the mucin-like sequence and the lack of a membrane proximal Ig (IgA like) domain.

Two isoforms of human MAdCAM-1 have been identified which exhibited single amino acid polymorphisms and j -19variation in the number Of copies of a serine/threonine/proline rich repeat in the mucin region.

These two isoforms appear to be encoded in genomic

DNA,

suggesting allelic variation and/or alternative processing of this sequence. These two isoforms may serve as alternative mechanisms of regulating a407 binding affinity and/or presenting carbohydrates for selectin binding. The presence of these Ig-like and mucin domains in primate MAdCAMs described herein is also consistent with role in selectin as well as integrin binding.

Recent domain swapping experiments in murine MAdCAM-1 have shown that, although domain one of MAdCAM-1 can weakly bind a4 adhesion is poor in the absence of strong integrin activation. The two amino-terminal Ig-like 15 domains (which are similar to domains of ICAM-1 and VCAM-1) are sufficient for a4R7 binding activity in an activation independent manner comparable to that of wild type murine MAdCAM-1.

A short motif (GLDTSL) present in domain one of murine 20 MAdCAM-1, is conserved and required for integrin binding in other Ig-like adhesion receptors, including of domain one of ICAM-1, ICAM-2, and ICAM-3, and domains 1 and 4 of VCAM-1 (Staunton, Cell, 52: 925-33 (1988); Staunton, et al., Nature, 339:61 (1989); Osborn, et al., 25 Cell, 59:1203 (1989); Fawcett, et al., Nature, 360:481 (1992)). This sequence, is located between 6 sheets c and d of these integrin binding domains. The GLDTSL motif was found in the primate MAdCAMs characterized here.

Mutagenesis of E34 (Glu 34 in this motif of domain 1 of ICAM-1 (underlined above) and of D40 (Asp 40 in VCAM-1 (in bold face above) had profound effects on binding of LFA-1 and a4fl, respectively (Osborn, et al., J. Cell.

Biol, 124:601-608 (1994); Renz, et al., J. Cell.

Biol., 125:1395-1406 (1994); Staunton, et al., Cell, 61:243-254 (1990); Vonderheide, et al., j. Cell.

Biol., 125:215-222 (1994)). More recently, a fragment of VCAM-1 comprising the two N-terminal domains was subjected to crystallographic structure determination (Jones,

E.Y.,

et al., Nature, 373:539-544 (1995); Wang, J-H, et al., Proc. Natl. Acad. Sci. USA., 92:5714-5718 (1995)). The conserved motif in VCAM-1 (QIDSPL) appears to be highly exposed on the N-terminal portion of the CD loop of the first Ig domain in a position that appears to be readily accessible to integrins.

A nucleotide substitution in this motif of murine MAdCAM-1, resulting in a change at amino acid 61 from leucine to arginine (L61-R61), abolishes MAdCAM-1 15 interactions with resting lymphocytes expressing a437.

STherefore, murine MAdCAM-I also requires this conserved amino acid motif, GLDTSL, within the computer predicted

CD

loop of its N-terminal domain for binding its integrin ligand, a4f7.

20 Comparisons of human MAdCAM cDNA clones 4 and (Figures 1 and 2) revealed that the amino-terminal 225 .e amino acids are identical in clones 4 and 20. This region comprises a predicted 18 amino acid hydrophobic leader or signal sequence, and two immunoglobulin-like domains. This region can be aligned with primate and murine MAdCAM-1, and displays the following conserved features: a predicted signal peptide (identical in the human proteins, and similar to the macaque and murine signal peptides); two pairs of cysteine residues in the first Ig-like domain, the cysteines of each pair being separated by 3 amino acids; a sequence of nine amino acids (which contains the "LDTSL" motif) in the predicted C-D loop of Ig-like domain 1, and is implicated as a general integrin recognition site (identical in each primate clone); and an uncharacteristically large second immunoglobulin-like -21domain. The size of the second Ig-like domain, with approximately 70 amino acids between cysteine residues would classify it as a (variable) type domain, in contrast with the C2 type (constant) domains which are more typically found in the Ig-like adhesion receptors (Hunkapiller, et al., Adv. in Immunol., 44:1-62 (1989); Williams, et al., Annu. Rev. Immunol., 6:381-405 (1988)). Within this domain is an extended C'-E loop containing an abundance of negatively charged residues, which is common to each primate, murine and human MAdCAM-1 clone characterized, but which is not seen in related adhesion receptors.

The next region found in clones 4 and 20 is analogous to the mucin domain of murine MAdCAM-1, due to a prevalence 15 of serine, threonine and proline (69% for clone 4 and 76% for clone 20) residues (boxed in Figure 1 and Figure 2).

This region, although similar in amino acid composition to murine MAdCAM-1, is highly divergent from murine MAdCAM-1.

Therefore, selection for conservation of the integrin 20 binding Ig-like domains appears greater than that of the mucin sequences. The human MAdCAM-1 domain is 71 amino acids long in clone 4, and 47 amino acids long in clone This region also contains two polymorphisms: a polymorphism at amino acid 240, which is proline in :25 clone 4 and serine in clone 20; and a polymorphism at amino acid 242, which is asparagine in clone 4 and aspartate in clone 20. In addition, the human mucin domains contain a repeat of 8 amino acids consisting of the sequence PPDTTS(Q/P)E, which appears eight times in clone 4 and five times in clone Since the human mucin domain is highly repetitive, truncation of three repeats in clone 20 relative to clone 4 could be the result of processes such as alternative splicing or mutation an aberrant recombination event) that maintain the reading frame, yielding a receptor -22that is functional with respect to integrin binding, and suggesting that some or all of the mucin sequences are dispensable for integrin binding. Consistently, it has been shown that Ig-like domains 1 and 2 of murine MAdCAM-1 are sufficient for activation-independent adhesion to a437, indicating that murine mucin sequences are dispensable for integrin binding. Also of interest in this regard, the macaque clone which was isolated lacks most of the repeat region.

The remaining C-terminal 110 amino acids are identical between clones 4 and 20: 47 amino acids precede a predicted hydrophobic transmembrane segment of 20 amino "acids, which is followed by a cytoplasmic tail of 43 amino acids. The 47 amino acids immediately C-terminal to the 15 mucin region are in a region corresponding to the IgA-like Ig domain of murine MAdCAM-1. Although the human and macaque proteins are similar in this region, they are divergent from murine MAdCAM-1. Compared with murine MAdCAM-1, the human proteins are 59 amino acids shorter in 20 this region, and lack any characteristics of an Ig-like domain. The transmembrane domains of all the receptors are similar, but the cytoplasmic tail is considerably longer (43 amino acids) in human (26 in primate and 20 in the mouse) MAdCAM-1.

25 Method of Producing Recombinant Proteins Another aspect of the invention relates to a method of producing a primate MAdCAM or variant portion) thereof. Recombinant protein can be obtained, for example, by the expression of a recombinant DNA molecule encoding a primate MAdCAM or variant thereof in a suitable host cell, for example.

Constructs suitable for the expression of a primate MAdCAM or variant thereof are also provided. The constructs can be introduced into a suitable host cell, and i -23cells which express a recombinant primate MAdCAM or variant thereof, can be produced and maintained in culture. Such cells are useful for a variety of purposes, and can be used in adhesion assays in an assay to screen for ligands and/or candidate inhibitors of MAdCAM-mediated adhesion), in the production of protein for characterization, isolation and/or purification, affinity purification), and as immunogens, for instance. Suitable host cells can be procaryotic, including bacterial cells such as E. coli, B. subtilis and or other suitable bacteria, or eucaryotic, such as fungal or yeast cells Pichia pastoris, Aspergillus species, Saccharomyces cerevisiae, Schizosaccharomyces pombe, Neurospora crassa), or other lower eucaryotic cells, and cells of higher 15 eucaryotes such as those from insects Sf9 insect cells) or mammals Chinese hamster ovary cells (CHO) o COS cells, HuT 78 cells, 293 cells). (See, Ausubel, F.M. et al., eds. Current Protocols in Molecular Biology, Greene Publishing Associates and John Wiley Sons Inc., 20 (1993)). In one embodiment, host cells capable of expressing membrane-bound mature protein are used. In another embodiment, host cells capable of secreting a soluble MAdCAM soluble MAdCAM, such as MAdCAM lacking the C-terminal transmembrane region and cytoplasmic tail).

Host cells which produce a recombinant primate MAdCAM or variants thereof can be produced as follows. For example, a nucleic acid encoding all or part of the coding sequence for the desired protein can be inserted into a nucleic acid vector, a DNA vector, such as a plasmid, virus or other suitable replicon for expression. A variety of vectors are available, including vectors which are maintained in single copy or multiple copy, or which become integrated into the host cell chromosome.

-24- The transcriptional and/or translational signals of a a MAdCAM-1 gene can be used to direct expression.

Alternatively, suitable expression vectors for the expression of a nucleic acid encoding all or part of the coding sequence of the desired protein are available.

Suitable expression vectors can contain a number of components, including, but not limited to one or more of the following: an origin of replication; a selectable marker gene; one or more expression control elements, such as a transcriptional control element a promoter, an enhancer, terminator), and/or one or more translation signals; a signal sequence or leader sequence for membrane targeting or secretion (of primate origin or from a heterologous primate or non-primate species). In a 15 construct, a signal sequence can be provided by the vector, the primate MAdCAM coding sequence, or other source.

A promoter is provided for expression in a suitable host cell. Promoters can be constitutive or inducible. In the vectors, the promoter is operably linked to a nucleic acid encoding the primate MAdCAM or variant thereof, and is capable of directing expression of the encoded polypeptide.

A variety of suitable promoters for procaryotic lac, tac, T3, T7 promoters for E. coli) and eucaryotic yeast alcohol dehydrogenase (ADH1), SV40, CMV) hosts are available.

In addition, the expression vectors typically comprise a selectable marker for selection of host cells carrying the vector, in the case of replicable expression vector, an origin or replication. Genes encoding products which confer antibiotic or drug resistance are common selectable markers and may be used in procaryotic f-lactamase gene (ampicillin resistance), Tet gene for tetracycline resistance) and eucaryotic cells neomycin (G418 or geneticin), gpt (mycophenolic acid), ampicillin, or hygromycin resistance genes). Dihydrofolate reductase marker genes permit selection with methotrexate in a variety of hosts. Genes encoding the gene product of auxotrophic markers of the host LEU2, URA3, HIS3) are often used as selectable markers in yeast. Use of viral baculovirus) or phage vectors, and vectors which are capable of integrating into the genome of the host cell, such as retroviral vectors, are also contemplated. The present invention also relates to cells carrying these expression vectors.

For example, a nucleic acid encoding a primate MAdCAM or variant thereof can be incorporated into the vector, operably linked to one or more expression control elements, and the construct can be introduced into host cells which are maintained under conditions suitable for expression, 15 whereby the encoded polypeptide is produced. The construct can be introduced into cells by a method appropriate to the host cell selected transformation, transfection, electroporation, infection). For production of a protein, host cells comprising the construct are maintained under 20 conditions appropriate for expression, in the presence of inducer, suitable media supplemented with appropriate salts, growth factors, antibiotic, nutritional supplements, etc.). The encoded protein human MAdCAM-1) can be isolated from the host cells or medium.

25 Fusion proteins can also be produced in this manner.

For example, some embodiments can be produced by the insertion of a primate MAdCAM cDNA or portion thereof into a suitable expression vector, such as Bluescript®II SK (Stratagene), pGEX-4T-2 (Pharmacia), pcDNA-3 (Invitrogen) and pET-15b (Novagen). The resulting construct is then introduced into a suitable host cell for expression. Upon expression, fusion protein can be isolated or purified from a cell lysate by means of a suitable affinity matrix (see Current Protocols in Molecular Biology (Ausubel, F.M.

et al., eds., Vol. 2, Suppl. 26, pp. 16.4.1-16.7.8 (1991)).

-26- In addition, affinity labels provide a means of detecting a fusion protein. For example, the cell surface expression or presence in a particular cell fraction of a fusion protein comprising an antigen or epitope affinity label can be detected by means of an appropriate antibody.

Nucleic Acids, Constructs and Vectors The present invention relates to isolated and/or recombinant (including, essentially pure) nucleic acids having sequences which encode a primate MAdCAM or variant thereof as described herein.

Nucleic acids referred to herein as "isolated'" are nucleic acids separated away from the nucleic acids of the genomic DNA or cellular RNA of their source of origin as it exists in cells or in a mixture of nucleic 15 acids such as a library), and may have undergone further processing. "Isolated" nucleic acids include nucleic acids obtained by methods described herein, similar methods or other suitable methods, including essentially pure nucleic acids, nucleic acids produced by chemical synthesis, by 20 combinations of biological and chemical methods, and recombinant nucleic acids which are isolated (see e.g., Daugherty, B.L. et al., Nucleic Acids Res., 19(9):2471-2476 ~(1991); Lewis, A.P. and J.S. Crowe, Gene, 101: 297-302 (1991)). Nucleic acids referred to herein as "recombinant" 25 are nucleic acids which have been produced by recombinant DNA methodology, including those nucleic acids that are generated by procedures which rely upon a method of artificial recombination, such as the polymerase chain reaction (PCR) and/or cloning into a vector using restriction enzymes. "Recombinant" nucleic acids are also those that result from recombination events that occur through the natural mechanisms of cells, but are selected for after the introduction to the cells of nucleic acids 111_ -27designed to allow and make probable a desired recombination event.

In one embodiment, the nucleic acid or portion thereof encodes a protein or polypeptide having at least one property, activity or function characteristic of a primate MAdCAM (as defined herein), such as binding function the ability to bind an a4#7 integrin), and/or cellular adhesion molecule function the ability to mediate cellular adhesion such as a407-dependent adhesion in vitro and/or in vivo), and/or an immunological property as defined herein.

The present invention also relates more specifically to isolated and/or recombinant nucleic acids or a portion thereof having sequences which encode human or macaque 15 MAdCAM-1 or variants thereof.

The invention further relates to isolated and/or recombinant nucleic acids that are characterized by: their ability to hybridize to a nucleic acid encoding a primate MAdCAM, such as a nucleic acid having a nucleotide sequence as set forth or substantially as set forth in Figure 1 (SEQ ID NO:1), Figure 2 (SEQ ID NO:3), or Figure 3 (SEQ ID NO:5); the complement of any one of or portions of either of the foregoing a portion comprising the open reading frame); or 25 by their ability to encode a polypeptide having the amino acid sequence of a primate MAdCAM SEQ ID NO:2, SEQ ID NO:4, or SEQ ID NO:6); or by both characteristics.

In one embodiment, the nucleic acid shares at least about 50% nucleotide sequence similarity to any one of the nucleotide sequences shown in Figure 1, Figure 2, or Figure 3 (SEQ ID NO:1, 3, or 5, respectively) or to one of the MAdCAM coding regions thereof. More preferably, the nucleic acid shares at least about 75% nucleotide sequence similarity, and still more preferably, at least about -28nucleotide sequence similarity, to any one of the sequences shown in Figure 1, Figure 2, or Figure 3 (SEQ ID NO:1, 3, or 5, respectively) or to one of the MAdCAM coding regions thereof.

Isolated and/or recombinant nucleic acids meeting these criteria comprise nucleic acids having sequences identical to sequences of naturally occurring primate MAdCAMs or variants of the naturally occurring sequences.

Such variants include mutants differing by the addition, deletion or substitution of one or more residues, modified nucleic acids in which one or more residues are modified DNA or RNA analogs), and mutants comprising one or more modified residues.

Nucleic acids of the present invention, including 15 those which hybridize to a selected nucleic acid as described above, can be detected or isolated under high stringency conditions or moderate stringency conditions, for example. "High stringency conditions" and "moderate stringency conditions" for nucleic acid hybridizations are 20 explained at pages 2.10.1-2.10.16 (see particularly 2.10.8- 1 1 and pages 6.3.1-6 in Current Protocols in Molecular Biology (Ausubel, F.M. et al., eds., Vol. i, Suppl. 26, 1991), the teachings of which are hereby incorporated by reference. Factors such as probe length, base composition, percent mismatch between the hybridizing sequences, temperature and ionic strength influence the stability of ""nucleic acid hybrids. Thus, high or moderate stringency conditions can be determined empirically, and depend in part upon the characteristics of the known nucleic acid DNA) and the other nucleic acids to be assessed for hybridization thereto.

Isolated and/or recombinant nucleic acids that are characterized by their ability to hybridize under high or moderate stringency conditions) to a nucleic acid encoding a primate MAdCAM (for example, those nucleic -29acids depicted in Figure. (SEQ ID NO:1), Figure 2 (SEQ ID NO:3), and Figure 3 (SEQ ID NO:5), the complement of such nucleic acids, or a portion thereof, can also encode a protein or polypeptide having at least one property, activity or function characteristic of a primate MAdCAM (as defined herein), such as binding function the ability to bind an a4#7 integrin), and/or cellular adhesion molecule function the ability to mediate cellular adhesion such as a4f7-dependent adhesion in vitro and/or in vivo), and/or an immunological property as defined herein. Preferred nucleic acids have lengths of at least about 40 nucleotides, more preferably at least about a 50, and still more preferably at least about nucleotides.

15 The binding function of a primate MAdCAM or variant thereof which is encoded by a nucleic acid of the present invention can be detected by standard assays for ligand binding assays which monitor formation of a complex between isolated and/or recombinant MAdCAM and an a4/7 20 integrin) or standard adhesion assays in which adhesion between a first cell expressing a recombinant primate MAdCAM, and a second cell bearing an a4f7 integrin is monitored), or other suitable methods. Binding and/or adhesion assays or other suitable methods can also be used in procedures for the identification and/or isolation of nucleic acids which encode a polypeptide of the present invention (see Example The antigenic properties of proteins or polypeptides encoded by nucleic acids of the present invention can be determined by immunological methods employing antibodies that bind to a primate MAdCAM, such as immunoblotting, immunoprecipitation and immunoassay radioimmunoassay,

ELISA).

Nucleic acids of the present invention can be used in the production of proteins or polypeptides. For example, a nucleic acid DNA) encoding a primate MAdCAM can be incorporated into various-constructs and vectors created for further manipulation of sequences or for production of the encoded polypeptide in suitable host cells as described above.

A further embodiment of the invention is antisense nucleic acid, which is complementary, in whole or in part, to a target molecule comprising a sense strand, and can hybridize with the target molecule. The target can be DNA, or its RNA counterpart wherein T residues of the DNA are U residues in the RNA counterpart). When introduced into a cell, antisense nucleic acid can inhibit the expression of the gene encoded by the sense strand.

Antisense nucleic acids can be produced by standard techniques.

15 In a particular embodiment, the antisense nucleic acid is wholly or partially complementary to and can hybridize with a target nucleic acid, wherein the target nucleic acid can hybridize to a nucleic acid having the sequence of the complement of the top strand shown in Figure I (SEQ ID 20 NO:1), Figure 2 (SEQ ID NO:3), or Figure 3 (SEQ ID For example, antisense nucleic acid can be complementary to a target nucleic acid having the sequence shown as the top strand of the open reading frame in Figure 1 (SEQ ID NO:1), Figure 2 (SEQ ID NO:3), or Figure 3 (SEQ ID NO:5), or to a portion thereof sufficient to allow hybridization. In another embodiment, the antisense nucleic acid is wholly or partially complementary to and can hybridize with a target nucleic acid which encodes a primate MAdCAM.

The nucleic acids can also be used as probes in in situ hybridization) to assess associations between inflammatory bowel disease (IBD) (or other conditions) and increased expression of primate MAdCAM in affected tissues.

The nucleic acids can also be used as probes to detect and/or isolate by hybridization with RNA or DNA) polymorphic or allelic variants, for example, in a sample -31inflamed tissue) obtained from a primate. Moreover, the presence or frequency of a particular variant in a sample(s) obtained from one or more affected primates, as compared with a sample(s) from normal primate(s), can be indicative of an association between inflammatory bowel disease (IBD) (or other conditions) and a particular variant, which in turn can be used in the diagnosis of the condition.

As described in the Examples, a cDNA clone encoding macaque MAdCAM-1 was isolated by expression cloning, and the cDNA was used as a probe to screen a human cDNA library. Two distinct nucleic acids encoding human MAdCAM-1 were isolated and characterized. Additional S. human, macaque or other primate genes or cDNAs can be obtained. For example, the genes described here, or sufficient portions thereof, whether isolated and/or recombinant or synthetic, can be used as probes or primers to detect and/or recover additional nucleic acids encoding primate MAdCAMs or variants thereof from a suitable source such as a primate genomic or cDNA library, according to methods described herein or other suitable methods by hybridization, PCR, expression cloning or other suitable techniques).

In one embodiment, nucleic acids encoding primate MAdCAM are producible by methods such as PCR amplification.

For example, appropriate primers a pair of primers or nested primers) can be designed which comprise a sequence which is complementary or substantially complementary to a portion of a primate MAdCAM cDNA described herein. For instance, primers complementary to the or 3'-ends of the coding sequence and/or flanking the coding sequence can be designed. Such primers can be used in a polymerase chain reaction with a suitable template nucleic acid to obtain nucleic acid encoding -32primate MAdCAM, for example. Suitable templates include constructs described herein (such as pcD3PMAd, pcD3HuMAd-4 or pcD3HuMAd-20), a cDNA library or another suitable source of primate human) cDNA or genomic DNA. Primers can contain portions complementary to flanking sequences of the construct selected as template as appropriate.

Additional genes or cDNAs can be used to express primate MAdCAM, with utilities corresponding to those described herein, and can be used in the production of constructs, host cells, and antibodies using methods described herein. The approaches described herein, including, but not limited to, the approaches used to isolate and manipulate macaque and human MAdCAM-1, to 15 construct vectors and host strains, and to produce and use the proteins, to produce antibodies, etc., can be applied to other primates.

Therapeutic Methods and Compositions The present invention also provides antibodies which can bind a "primate MAdCAM" in vitro and/or in vivo; and/or can inhibit an activity or function characteristic of a "primate MAdCAM", such as binding function the ability to bind an a4f7 integrin) and/or cellular adhesion molecule function the 25 ability to mediate cellular adhesion such as a4f7-dependent adhesion in vitro and/or in vivo). Such antibodies include antibodies which can bind a human or macaque MAdCAM encoded by cDNA clone 4, cDNA clone 20 or cDNA clone 31D. Also encompassed are antibodies which can bind a naturally occurring or endogenous primate MAdCAM human MAdCAM). Preferably the antibodies are capable of selective binding of primate MAdCAM in vitro and/or in vivo bind selectively to primate MAdCAM expressed in -33mucosal tissue and/or spleen as assessed immunohistologically)).

In one embodiment, the antibodies can bind primate MAdCAM and inhibit binding of "primate MAdCAM" to an a4f7 integrin human), thereby inhibiting cellular adhesion mediated by MAdCAM, preferably selectively. Such an antibody can inhibit a4?7-dependent cellular adhesion to cells bearing an a407 integrin, such as leukocytes (especially lymphocytes such as T or B cells) in vitro and/or in vivo. For example, eleven hybridomas were identified which produced antibodies which specifically inhibit the adhesion of RPMI 8866 cells to MAdCAM-1 (Example 2, hybridomas designated 10G4, 8C1, 10G3, 9G12, 9E4, 7H12, 10F2, 10A6, 1E5, 2F5, 7G11). Thus, antibodies 15 which can inhibit cellular adhesion of cells bearing an Sa4f7 integrin to vascular endothelial cells in mucosal tissues, including gut-associated tissues or lymphoid organs are encompassed by the antibodies of the present invention.

20 Preferably, the antibodies can bind a primate MAdCAM with high affinity (for example, a Ka in the range of about 1 10 nM, or a Kd in the range of about 1 X 10- 8 to 1 X 10-1 0 mol-).

The antibodies of the present invention are useful in 25 a variety of applications, including processes, research, Sdiagnostic and therapeutic applications. For instance, they can be used to isolate and/or purify primate MAdCAM or variants thereof by affinity purification or other suitable methods), and to study MAdCAM structure conformation) and function.

The antibodies of the present invention can also be used to modulate MAdCAM function in diagnostic in vitro) or therapeutic applications. For instance, antibodies can act as inhibitors of to inhibit (reduce or -34prevent) binding function and/or cellular adhesion molecule function of a primate MAdCAM as described herein.

In addition, antibodies of the present invention can be used to detect and/or measure the level of a primate MAdCAM in a sample tissues or body fluids, such as an inflammatory exudate, blood, serum, bowel fluid, or on cells transfected with a nucleic acid of the present invention). For example, a sample tissue and/or fluid) can be obtained from a primate and a suitable immunological method can be used to detect and/or measure primate MAdCAM levels, including methods such as enzymelinked immunosorbent assays (ELISA), including chemiluminescence assays, radioimmunoassay, and immunohistology. In one embodiment, a method of detecting 15 a selected primate MadCAM in a sample is provided, comprising contacting a sample with an antibody which binds an isolated primate MAdCAM under conditions suitable for specific binding of said antibody to the selected primate MAdCAM, and detecting antibody-MAdCAM complexes which are 20 formed.

SIn an application of the method, antibodies reactive with a primate MAdCAM-1 can be used to analyze normal versus inflamed tissues in human and non-human primates for primate MAdCAM reactivity and/or expression immunohistologically). Thus, the antibodies of the present invention permit immunological methods of assessment of expression of primate human MAdCAM-1) in normal versus inflamed tissues, through which the presence of disease, disease progress and/or the efficacy of antiprimate MAdCAM-1 therapy in inflammatory disease can be assessed.

The present invention also provides "primate MAdCAM" as defined herein, including functional variants, such as soluble primate MAdCAM lacking the all or part of the transmembrane region and cytoplasmic tail, such that .1.1 the protein secreted) and functional fusion proteins hybrid immunoglobulins comprising a primate MAdCAM moiety fused at its C-terminus, to the N-terminus of an immunoglobulin moiety). These molecules are useful in a variety of applications, including processes, research, diagnostic and therapeutic applications.

For example, primate MAdCAM, MAdCAM-Ig fusion proteins or other recombinant soluble primate MAdCAM molecules can be used in assays to identify ligands or inhibitors a blocking antibody) of primate MAdCAM: a47 interaction.

As used herein, an inhibitor is a compound which inhibits (reduces or prevents) the binding of primate MAdCAM-1 to a ligand, including a4f37 integrin, and/or which inhibits the triggering of a cellular response mediated by the ligand.

15 An effective amount is an amount sufficient to achieve inhibition of binding or adhesion to primate MAdCAM-1 and/or signalling an amount sufficient to inhibit adhesion of a cell bearing a primate MAdCAM-1 ligand (including a4f7 integrins, such as human a4/37 integrin, and 20 its primate homologs)) to isolated/and or recombinant primate MAdCAM.

*In one aspect, a method of detecting or identifying a ligand of primate MAdCAM or an agent which binds a primate MadCAM is provided, in which an one or more) agent 25 to be tested (or a candidate ligand) is contacted with an isolated and/or recombinant "primate MAdCAM", including "functional variants", as defined herein under conditions suitable for binding of ligand thereto, and the formation of a complex between said agent and primate MAdCAM is detected. In one embodiment, an agent to be tested is combined with a host cell expressing recombinant primate MAdCAM or a functional variant under conditions suitable for binding of ligand thereto. In one embodiment, the primate MAdCAM or functional variant is labeled with a suitable label fluorescent label, isotope label), and binding is determined by detection of the label.

-36- Specificity of binding can be assessed by competition or displacement, for example, using unlabeled agent, an unlabeled isolated and/or recombinant primate MAdCAM or functional variant, or a second ligand of primate MAdCAM as competitor.

In another aspect, a method of detecting an inhibitor of cellular adhesion mediated by primate MAdCAM is provided. In one embodiment, an agent to be tested is combined with a ligand of primate MAdCAM, and an isolated and/or recombinant primate MAdCAM or functional variant fusion protein) under conditions suitable for binding of ligand thereto. The formation of a complex between the ligand and primate MAdCAM or the functional variant is monitored. A decrease in binding of ligand in 15 the presence of the agent relative to a suitable control binding in the absence of agent) is indicative that the agent is an inhibitor. For example, the fusion proteins and assays described in Example 3 can be used to detect inhibitors. An agent to be tested can also be 20 combined with a first cell expressing a recombinant primate MAdCAM, and a second cell bearing an a407 integrin under conditions suitable for adhesion of said first cell to said second cell. Adhesion between said first and second cells •can be monitored, and decreased adhesion (reduced or abolished) as compared with a suitable control is indicative that the agent is an inhibitor. A cell or cells which naturally express a ligand for MAdCAM-1, such as a leukocyte an a4f7 B lymphocyte, T lymphocyte) or other cell which expresses a ligand for MAdCAM-1 a recombinant cell) can be used.

Assays such as those described in Example 3 can be used to identify compounds which inhibit binding in vitro.

As shown herein, fusion proteins comprising a primate MAdCAM moiety (two chimeric MAdCAM-Ig fusions) can bind to -37a467 positive lymphocytes in solution. Thus, primate MAdCAM, including functional variants, particularly soluble primate MAdCAM molecules and fusion proteins such as the chimeric MAdCAM-Ig fusions described in Example 3, provide candidate inhibitors of a4f7:MAdCAM interaction and of in vivo lymphocyte recruitment to inflammatory sites, which can be useful in therapy as described hereinbelow. The in vivo efficacy of these molecules can be assessed using methods described herein (see Examples 4 and 5) or other suitable methods. For example, primate models such as those described in Example 5 can be used. The CD45RBHi/SCID model provides a mouse model with similarity to both Crohn's disease and ulcerative colitis (Example 4, Powrie, F. et al., Immunity, 1: 553-562 (1994)).

15 Efficacy in this model can be assessed using an experimental protocol similar to the one used for monoclonal antibodies (Example Parameters such as inhibition of recruitment of '"In-labeled cells to the colon and reduction in the number of CD4 T lymphocytes in the lamina proria of the large intestine after administration intravenous intraperitoneally and per oral can be assessed. Knockout mice which develop intestinal lesions similar to those of human inflammatory bowel disease have also been described 25 (Strober, W. and Ehrhardt, Cell, 75: 203-205 (1993)), and NOD mice provide an animal model of insulin-dependent diabetes mellitus.

The invention further relates to the discovery that diseases associated with leukocyte recruitment to the gastrointestinal tract, such as IBD, or other mucosal tissues can be treated by inhibiting MAdCAM binding to the a4?7 integrin and/or triggering of a407-mediated cellular responses. Compounds or agents which inhibit binding include "primate MAdCAM" as defined herein, including -38soluble primate MAdCAM molecules and fusion proteins, as well as antibodies or antigen binding fragments thereof which bind MAdCAM and/or the a4#7 integrin. Antibodies which can be used in the method include recombinant or non-recombinant polyclonal, monoclonal, chimeric, humanized and/or anti-idiotypic antibodies.

Monoclonal antibodies that bind MAdCAM or a407 have been described. For example, MECA-367 is an anti-MAdCAM antibody of the IgG2a subtype and is described in Gallatin et al., Nature, 304: 30 (1983) and Michie et al., Am. J.

Pathol., 143: 1688-1698 (1993). ACT-1 is a monoclonal antibody which binds the 4(37 integrin (Lazarovits et al., J. Immunol., 133: 1857 (1984); Schweighoffer et al., J. Immunol., 151: 717-729 (1993)). FIB 21 binds the 07 15 chain is described and characterized in Berlin et al., Cell, 74: 184-195 (1993); Andrew, D.P. et al., J. Immunol., 153: 3847-3861 (1994)).

Other polyclonal or monoclonal antibodies, such as antibodies which bind to the same or similar epitopes as 20 the antibodies described above, can be made according to methods described herein, methods known in the art or other suitable methods (such as Kohler et al., Nature, 256:495a a 497 (1975), Harlow et al., 1988, Antibodies: A Laboratory Manual, (Cold Spring Harbor, NY) or Current Protocols in Molecular Biology, Vol. 2 (Supplement 27, Summer '94), Ausubel et al., Eds. (John Wiley Sons: New York, NY), Chapter 11 (1991)). Antibodies can also be produced which can compete with any one of the antibodies produced by the hybridoma cell lines designated 10G4, 8C1, 10G3, 9G12, 9E4, 7H12, 10F2, 10A6, 1E5, 2F5, or 7G11 for binding to a cell bearing an a4#7 integrin, preferably human a4#7 integrin.

For example, antibodies can be raised against an appropriate immunogen in a suitable mammal a mouse, rat, rabbit or sheep). Immunogens include, for example, MAdCAM, a4f7, or immunogenic fragments thereof. For -39example, aprimate MAdCAM or a variant thereof can be produced and used as an immunogen to raise antibodies in a suitable immunization protocol.

Antibody-producing cells a lymphocyte) can be isolated from, for example, the lymph nodes or spleen of an immunized animal. The cells can then be fused to a suitable immortalized cell a myeloma cell line), thereby forming a hybridoma. Fused cells can be isolated employing selective culturing techniques. Cells which produce antibodies with the desired specificity can be selected by a suitable assay ELISA) (see e.g., Example 2).

In one embodiment, the immunogen can be an antibody which binds, for example, MAdCAM, a4f7, or immunogenic 15 fragments thereof. The antibody raised thereby can be an anti-idiotypic antibody, which can also be used in the present invention Patent No. 4,699,880).

Single chain antibodies, and chimeric, humanized or primatized (CDR-grafted or resurfaced, such as, according 20 to EP 0,592,406; Padlan et al., April 13, 1994) antibodies, as well as chimeric or CDR-grafted single chain antibodies, comprising portions derived from different species, can also be used in the invention. The various portions of f these antibodies can be joined together chemically by 25 conventional techniques, or can be prepared as a contiguous protein using genetic engineering techniques. For example, nucleic acids encoding a chimeric or humanized chain can be expressed to produce a contiguous protein. See, e.g., Cabilly et al., U.S. Patent No. 4,816,567; Cabilly et al., European Patent No. 0,125,023 Bl; Boss et al., U.S. Patent No. 4,816,397; Boss et al., European Patent No. 0,120,694 Bl; Neuberger, M.S. et al., WO 86/01533; Neuberger, M.S. et al., European Patent No. 0,194,276 Bl; Winter, U.S. Patent No. 5,225,539; and Winter, European Patent No. 0,239,400 Bl. See also, Newman, R. et al., BioTechnology, 10:1455- 1460 (1992), regarding primatized antibody, and Ladner et al., U.S. Patent No. 4,946,778 and Bird, R.E. et al., Science, 242:423-426 (1988)) regarding single chain antibodies.

In addition, functional fragments of antibodies, including fragments of chimeric, humanized, primatized or single chain antibodies, can also be produced. Functional fragments of the foregoing antibodies retain at least one binding function of the full-length antibody from which they are derived and, preferably, retain the ability to inhibit interaction. For example, antibody fragments capable of binding to the a4?7 integrin, MAdCAM or portion thereof include, but are not limited to, Fv, Fab, Fab' and F(ab') 2 fragments. Such fragments can be produced by 15 enzymatic cleavage or by recombinant techniques. For instance, papain or pepsin cleavage can generate Fab or F(ab') 2 fragments, respectively. Alternatively, antibodies can be produced in a variety of truncated forms using antibody genes in which one or more stop codons has been 20 introduced upstream of the natural stop site. For example, a chimeric gene encoding a heavy chain portion can be designed to include DNA sequences encoding the CHI domain and hinge region of the heavy chain.

Antibodies and antigen binding fragments thereof which can be used in the claimed method include antibodies which ;bind to MAdCAM and/or a4/7, such as anti-f7 chain antibodies. For example, antibodies from the group including FIB 21, FIB 30, FIB 504 and ACT-1 and mixtures thereof can be administered. Alternatively or in addition, antigen fragments of these antibodies can be administered.

Compounds or agents which inhibit the binding of MAdCAM and the a407 integrin can be administered according to the claimed method in the treatment of diseases associated with leukocyte lymphocyte, monocyte) -41infiltration of tissues (including recruitment and/or accumulation of leukocytes in tissues) which express the molecule MAdCAM-1. An effective amount of a compound or agent one or more) is administered to an individual a mammal, such as a human or other primate) in order to treat such a disease. For example, inflammatory diseases, including diseases which are associated with leukocyte infiltration of the gastrointestinal tract (including gut-associated endothelium), other mucosal tissues, or tissues expressing the molecule MAdCAM-1 gut-associated tissues, such as venules of the lamina propria of the small and large intestine; and mammary gland lactating mammary gland)), can be treated according to the present method. Similarly, an individual having a 15 disease associated with leukocyte infiltration of tissues as a result of binding of leukocytes to cells endothelial cells) expressing the molecule MAdCAM-1 can be treated according to the present invention.

In a particularly preferred embodiment, diseases which 20 can be treated accordingly include inflammatory bowel disease (IBD), such as ulcerative colitis, Crohn's disease, ileitis, Celiac disease, nontropical Sprue, enteropathy associated with seronegative arthropathies, microscopic or collagenous colitis, eosinophilic gastroenteritis, or pouchitis resulting after proctocolectomy, and ileoanal anastomosis.

Pancreatitis and insulin-dependent diabetes mellitus are other diseases which can be treated using the present method. It has been reported that MAdCAM-1 is expressed by some vessels in the exocrine pancreas from NOD (nonobese diabetic) mice, as well as from BALB/c and SJL mice.

Expression of MAdCAM-1 was reportedly induced on endothelium in inflamed islets of the pancreas of the NOD mouse, and MAdCAM-1 was the predominant addressin expressed by NOD islet endothelium at early stages of insulitis t -42- (Hanninen, et al., J. Clin. Invest., 92: 2509-2515 (1993)). Further, accumulation of lymphocytes expressing a4#7 within islets was observed, and MAdCAM-1 was implicated in the binding of lymphoma cells via a4?7 to vessels from inflamed islets (Hanninen, et al., J.

Clin. Invest., 92: 2509-2515 (1993)).

Examples of inflammatory diseases associated with mucosal tissues which can be treated according to the present method include mastitis (mammary gland), cholecystitis, cholangitis or pericholangitis (bile duct and surrounding tissue of the liver), chronic bronchitis, chronic sinusitis, asthma, and graft versus host disease in the gastrointestinal tract). As seen in Crohn's disease, inflammation often extends beyond the mucosal 15 surface, accordingly chronic inflammatory diseases of the lung which result in interstitial fibrosis, such as hypersensitivity pneumonitis, collagen diseases, sarcoidosis, and other idiopathic conditions can be amenable to treatment.

20 The compound is administered in an effective amount which inhibits binding of MAdCAM to the a407 integrin. For therapy, an effective amount will be sufficient to achieve the desired therapeutic and/or prophylactic effect (such as an amount sufficient to reduce or prevent MAdCAM-mediated binding and/or signalling, thereby inhibiting leukocyte adhesion and infiltration and/or associated cellular Sresponses). The compounds can be administered in a single dose or multiple doses. The dosage can be determined by methods known in the art and is dependent, for example, upon the individual's age, sensitivity, tolerance and overall well-being. Suitable dosages for antibodies can be from 0.1-1.0 mg/kg body weight per treatment.

According to the method, a compound or agent can be administered to an individual a human) alone or in conjunction with another agent. A compound or agent can be

FI___

-43administered before, along with or subsequent to administration of the additional agent. In one embodiment, more than one monoclonal antibody which inhibits the Sbinding of leukocytes to endothelial MAdCAM is administered. Alternatively, a monoclonal antibody which inhibits the binding of leukocytes to endothelial ligands is administered in addition to an anti-MAdCAM or anti-f7 antibody. For example, an antibody that inhibits the binding of leukocytes to an endothelial ligand other than MAdCAM, such as an anti-ICAM-l or anti-VCAM-1 antibody can also be administered. In another embodiment, an additional pharmacologically active ingredient sulfasalazine an antiinflammatory compound, or a steroidal or other nonsteroidal antiinflammatory compound) can be administered in 15 conjunction with the compound or agent the antibody of the present invention).

A variety of routes of administration are possible including, but not necessarily limited to parenteral intravenous, intraarterial, intramuscular, subcutaneous injection), oral dietary), topical, inhalation intrabronchial, intranasal or oral inhalation, intranasal drops), or rectal, depending on the disease or condition to be treated. Parenteral administration is a preferred mode of administration.

*5 Formulation of a compound to be administered will vary according to the route of administration selected solution, emulsion, capsule). An appropriate composition comprising the compound to be administered can be prepared in a physiologically acceptable vehicle or carrier. For solutions or emulsions, suitable carriers include, for example, aqueous or alcoholic/aqueous solutions, emulsions or suspensions, including saline and buffered media.

Parenteral vehicles can include sodium chloride solution, Ringer's dextrose, dextrose and sodium chloride, lactated Ringer's or fixed oils. Intravenous vehicles can include -44various additives, preservatives, or fluid, nutrient or electrolyte replenishers (See, generally, Remington/s Pharmaceutical Science, 16th Edition, Mack, Ed. 1980). For inhalation, the compound can be solubilized and loaded into a suitable dispenser for administration an atomizer, nebulizer or pressurized aerosol dispenser).

EXEMPLIFICATION

The present invention will now be illustrated by the following Examples, which are not intended to be limiting in any way.

Introduction "Hybridization studies using zoo blots with murine MAdCAM DNA probes under low stringency conditions indicated that nucleotide conservation between murine MAdCAM-1 and 15 higher species was poor. A functional expression approach was used to clone primate and human homologs, whereby cells transfected with cDNAs which conferred the ability to adhere to a target lymphocyte cell line expressing high levels of the MAdCAM-l ligand (a407) were identified and the cDNAs recovered. As human tissue sources were scarce, a primate homolog of MAdCAM-1 was first identified.

:For expression cloning, a primate cDNA expression library, derived from mesenteric lymph nodes of a macaque, was made in a eukaryotic expression vector pRSVsport (from 25 Gibco/BRL). A high efficiency transfection system using the CHO/P cell line (Heffernan, M. and J.D. Dennis, Nucleic Acids Res., 19: 85-92 (1991)) was used. The library was separated and individual pools (representing approximately 1,500 clones) were transfected in wells of 24 well tissue culture plates. Cell adhesion assays were performed to identify cDNAs which conferred an adhesive phenotype on T and B cell lines expressing the a417 integrin, a known ligand for MAdCAM-1. Adhesion was identified .I1IXI1( C microscopically by rosetting of the T and B cell lines on the transfected cells. A pool conferring the desired phenotype was subfractionated until a single full-length cDNA clone designated clone 31D was identified.

DNA

sequencing of the amino-terminal portion of the cDNA revealed homology of the macaque clone to murine MAdCAM-1 (Briskin, et al, Nature (Lond.), 363:461-464 (1993)) at both the protein and nucleic acid level.

When introduced into CHO/P cells by transient transfection, the cDNA insert obtained from clone 31D directed the expression of a protein which could mediate binding to two cell lines which express a4,7:

TKI,

a murine T cell lymphoma (Butcher, et al., Eur. J. of Immunol., 10: 556-561 (1980)); and RPMI 8866, a human B 15 cell lymphoma (Erle, et al., J. Immunol., 153: 517- 528 (1994)). Binding of TK1 cells to cells transfected with the macaque cDNA could be blocked by antibodies to either the a4 (MAb PS/2) or the /7 (MAb FIB 504) integrins, and binding of RPMI 8866 to CHO/P cells transiently 20 transfected with macaque cDNA (clone 31D in pSV-SPORT) was blocked by the anti-a47 MAb, ACT-1. In control experiments, a cDNA encoding human VCAM-1 failed to bind the RPMI 8866 human B cell line. Jurkat cells, a T cell line which expresses a4B1 and not a407, was shown to bind VCAM-1, but failed to bind transfectants expressing macaque cDNA.

The cDNA encoding a primate (macaque) homolog of murine MAdCAM-1 was used as a probe to obtain a clone encoding a human homolog by hybridization. To obtain a human MAdCAM-1 clone, two cDNA libraries, one derived from histologically normal human mesenteric lymph node (MLN) and one derived from an inflamed MLN lymph node from a patient with Crohn's disease, were constructed in the XZiplox phage vector from Gibco/BRL. cDNA from the macaque clone was used to screen these libraries. Two different human cDNA -46clones of similar size were isolated. These clones each appeared to be full-length by preliminary sequence analysis. Analysis of human, as well as macaque, MAdCAM-1 cDNAs indicates that each of the encoded proteins has a predicted hydrophobic leader sequence (underlined in Figures with the remaining portions of the proteins corresponding to predicted mature human or macaque MAdCAM-1, respectively.

To assess function, the human cDNA inserts were subcloned into the pCDNA3 expression vector (Invitrogen) and transient expression assays were used to demonstrate function. The human cDNAs can be expressed as functional proteins, and are capable of mediating specific binding to cells expressing a4f7. Accordingly, these two human cDNA 15 clones are designated as human MAdCAM-1 cDNAs.

Stable transfectants of both the primate and human cDNAs were generated in a mouse pre-B cell line, Ll-2 and CHO cells. Ll-2 transfectants were used to immunize mice and generate monoclonal antibodies against human MAdCAM-1.

Antibodies capable of inhibiting the interaction between .MAdCAM-1 and a47 were identified. The production of blocking antibodies directed against human MAdCAM-1 is a significant advance, as previous attempts to produce such blocking antibodies having cross-reactivity with the human 25 homolog using murine MAdCAM-1 have failed.

Example 1. Cloning of Macaque and Human MAdCAM-1 cDNAs RNA isolation and selection of message Total RNA was isolated from primate (macaque) mesenteric lymph nodes (MLN); histologically normal human mesenteric lymph nodes; human mesenteric lymph nodes (inflamed ileal nodes) from a patient with Crohn's disease; and tissue culture cells by use of the CsTFATM -47- (cesium trifluoroacetate).reagent (Pharmacia; Cat. #17-087- 02). Total RNA from mesenteric lymph node was obtained from two species of macaque (Macaca fascicularis, and Macaca mulatta), and was combined prior to isolation of poly-A RNA. Tissue was first snap frozen in liquid nitrogen and subjected to dounce homogenization in a solution consisting of 5.5 M guanidinium isothiocyanate, mM sodium citrate, 0.5% sodium laurel sarcosine and 0.2 M 2-mercaptoethanol, while tissue culture cells (1-5 X 108) were washed once in phosphate buffered saline (PBS) and homogenized by pipetting. A clarified lysate was then layered on a cushion of CsTFA and total RNA was pelleted by centrifugation for 20 hours at 30,000 RPM.

mRNA was selected by the polyATract mRNA isolation 15 system from Promega. The system uses a biotinylated oligo(dT) primer to hybridize (in solution) to poly A tails of eukaryotic messages. The hybrids were captured and washed at high stringency using streptavidin coupled to paramagnetic particles and a magnetic separation stand.

20 mRNA was selected by a single purification in this system and the yields ranged from 1-2% of the total RNA yield.

The integrity of both the total and mRNA was analyzed by gel electrophoresis and ethidium bromide staining.

cDNA synthesis 25 cDNA was synthesized using the SuperscriptTM lambda system (Cat. #18256-016) in conjunction with either the XZiploxTM vector (Gibco/BRL, Gaithersburg, MD, Cat.

#19643-014) in the case of the human libraries, or the pSV-SPORT-1 vector (Gibco/BRL, Cat. #15388-010) in the case of the macaque library. The following modifications from the standard protocol were made. cDNA was labeled only in the first or second strand (but not both) with a3P-dCTP -48and estimates of quantity were made by inspection of ethidium bromide staining of aliquots of cDNA fractions.

DNA Sequencing The entire macaque and human MAdCAM-1 cDNAs were first isolated in the library vectors pSV-SPORT-1 and pZL1 (rescued from XZiploxTM), respectively. Based on restriction mapping, fragments were subcloned into Bluescript' vectors (Stratagene) to facilitate sequencing from internal regions of the cDNAs. After sequence analysis of these clones, oligonucleotide primers were made to complete the sequence. Overlapping sequence of both strands was obtained. Sequence analysis utilized the sequenaseTM 7-deaza-dGTP DNA sequencing kit with sequenase version 2.0 T7 DNA polymerase (United States Biochemical) S* 15 and 3 5S-dCTP (Amersham Life Science and New England Nuclear). The delta TAQ seqeuncing kit (USB) and gamma 32p-ATP (Amersham) G-C rich sequence were also used for G-C rich sequences.

Sequences were entered and analyzed using the Lasergene system (DNASTAR, Inc.). Nucleotide sequence alignments were performed by the Clustal method with Weighted residue weight table, using a gap penalty of and a gap length penalty of 10, and default parameters (Pairwise alignment parameters were: ktuple 2, gap penalty 5, window 4, and diagonals saved 4).

Amino acid sequence alignments were performed by the Clustal method with the PAM250 residue weight table, using a gap penalty of 10 and a gap length penalty of 10 and default parameters (Pairwise alignment parameters were: ktuple 1, gap penalty 3, window 4, and diagonals saved I 1 -49- Preparation of macaque expression library The size fractionation procedure was also modified slightly for construction of the macaque expression library to ensure large kb) inserts. After one round of fractionation, only the first (largest) fraction of cDNA was saved and the remaining fractions were pooled and subjected to a subsequent round of fractionation. The top fraction from the next round was pooled with the top fraction from the previous round and the second fraction from this round was also used. These two fractions were precipitated and put into ligations with the pSV-SPORT-1 vector and a fraction of each ligation was transformed into electrocompetent DH1OB bacteria (Gibco) to estimate both the titer of the library and the average insert size.

Estimates from ligation of only top largest cDNA fraction °o revealed the potential of making up 2.4 million independent clones with an average insert size of 1.9 kb and a median size of 2 kb.

The actual library screened consisted of 150,000 20 independent clones which were plated at a density of 1,500 clones/plate on 100 LB agar plates (to generate 100 pools of 1,500 clones/pool) with ampicillin at 50 gg/ml and grown overnight at 37 0 C. For purification of individual pools, each plate was overlayed with approximately 2 ml of Luria 25 broth and the colonies were scraped off of each plate with a standard tissue culture cell scraper, and bacterial suspensions were transferred to microfuge tubes. Prior to purification, a glycerol stock was generated from each pool. Plasmid DNAs were purified using QIAprep spin columns (QIAGEN) according to manufacturer's instructions.

Transfections CHO/P cells (Heffernan, M. and J.D. Dennis, Nucleic Acids Res., 19:85-92 (1991)) were seeded into 24 well plates approximately 24 hours prior to transfection at a density of 40,000 cells/well. DNAs were transiently transfected using the LipofectAMINET reagent (GIBCO; Cat.

#18324-012), essentially following the recommended protocol with further optimization for 24-well plates as follows: 200 ng of DNA (representing either a plasmid pool or purified control DNAs) was diluted to 20 tl with Opti-MEM 1 reduced serum media (GIBCO) and diluted into 20 Ml of a mixture that consists of 18 pl Opti-MEM 1 and 2 Al of LipofectAMINE T M reagent. This liposome mixture was then incubated for approximately 30 minutes at ambient 15 temperature after which, 200 Al of Opti-MEM 1 was added, and the entire mixture was then overlayed onto a well of CHO/P cells and returned to the incubator. After a hour incubation at 37 0 C, 240 pl of MEM-a (Gibco) media with 20% fetal calf serum (FCS) was added to each well, and the cells were incubated for an additional 18-24 hours at 37 0

C.

The media was then changed to standard MEM-a with 10% FCS, and the adhesion assay was performed approximately 20-24 hours later.

Adhesion assays for expression cloni For the adhesion assays in the expression cloning screen, the murine T cell lymphoma TK1 which expresses high levels of a4,7 (Butcher, et al., Eur. J. Immunol., 556-561 (1980)) was used to detect CHO/P cells transfected with cDNAs capable of conferring an adhesive phenotype. TK1 cells were resuspended at a density of 2 X 10 6 /ml in an assay buffer which consisted of HBSS (Hanks Balanced Salt Solution, without Ca 2 or Mg2+), supplemented with 2% bovine calf serum, 20 mM HEPES, pH -51- 7.3, 2 mM Mg 2 and 2 mM Ca 2 Each well transfected with a DNA pool was preincubated with 0.25 ml of a combined supernatant containing monoclonal antibodies to both human VCAM-1 (MAb 2G7; Graber, et al., J. Immunol., 145:819-830 (1990)) and murine MAdCAM-1 (MAb MECA-367; American Type Culture Collection (Rockville, MD), Accession No. HB9478; Streeter, et al., Nature, 331:41 (1988)); see also, U.S. Patent No. 5,403,919 to Butcher) in order to eliminate adhesion mediated by VCAM-1 (which is expressed at high levels in primate lymph nodes) or any potential contaminating murine MAdCAM-1 expression plasmids. After incubation at 4 0C for 15 minutes, 0.25 ml of the TK1 cell suspension (5 x 105 TKl cells) was added to each well, and incubation on a rocking platform was continued for an additional 30 minutes at 4 0 C. Plates were washed by gently inverting in a large beaker of phosphate buffered saline (PBS) followed by inversion in a beaker of PBS with gluteraldehyde for fixation for a minimum of 1 hour. Wells were then examined microscopically (10X objective) for 20 rosetting of TK1 cells.

Purification of macaque clones Pools yielding one or more TKl rosettes were further subfractionated by the following protocol:

DNA

representative of a positive pool was retransformed into 25 DH 1 OB and plated on ninety-six 100 mm petri dishes at a density of approximately 200 colonies/plate.

Nitrocellulose filters were used to generate replica plates, and one set of each plate was then subjected to DNA purification and subsequent adhesion assays as described above. A replica plate representative of a positive pool was then further subfractionated into pools of 5 colonies, which were replica plated and grown overnight in LB media containing ampicillin. After one more round of DNA -52purification and adhesion assays, individual clones could then be grown up and the clones conferring adhesion of the TK1 cells were identified.

A full-length clone which was shown to encode MAdCAM-i was obtained and designated clone 31D. Clone 31D, constructed in pSV-SPORT-1 (P25), contains a 5'-SalI to NotI-3' cDNA insert. Transformants of E. coli strain DH1OB containing clone 31D were obtained. For expression in stable cell lines, this cDNA was subcloned into expression vector pcDNA-3 (Invitrogen), which carries a neo resistance gene suitable for G418 selection. In particular, insert of clone 31D was released by digestion with EcoRI and NotI, and inserted into pcDNA-3 which had been cleaved with EcoRI and NotI to obtain pcD3pMAd.

15 Results A cDNA expression library, divided into pools of 1,500 independent clones, was constructed from mRNA purified from macaque mesenteric lymph nodes (MLNs). Each pool was transiently transfected into the CHO/P cell line, and 48 hours after transfection, a cell adhesion assay was performed using the murine T cell lymphoma TKl. As VCAM-1 a. is expressed in MLNs, assays were done in the presence of anti-VCAM-l M A b 2G7 (Graber, et al., J. Immunol., 145:819-830 (1990)). Additionally, assays were performed 25 at 4 0 C in order to eliminate adhesion mediated by ICAM cDNAs (TK1 cells express high levels of LFA-1 and LFA-1 is not functional at 40C). Microscopic examination of the assays revealed several wells with noticeable rosetting of TKl cells. Two wells were chosen for further analysis by repeating the transfection and determining whether the binding mediated by the pools could be blocked by anti-f7 or anti-a4 MAbs. TK1 binding to one of the pools was completely inhibited by pre-incubation of TK1 cells with either anti-a4 MAb PS/2 or anti-f7 MAb FIB 504. This pool -53was subjected to three rounds of subfractionation until a single clone, called 31D, was isolated. Purified clone 31D mediated TK1 cell binding which could be inhibited by antia4 or anti-P7 antibodies.

The insert size of clone 31D was approximately 1.8 kb.

Sequencing of the amino-terminus revealed several features consistent with a primate homolog of murine MAdCAM-1. The signal peptides were both 21 amino acids in length.

Although the amino acid similarity was found to be only 48%, identity was 71% if non-conservative substitutions were considered. In addition, the protein encoded by clone 31D had a characteristic unique to Ig-family adhesion receptors: two pairs of cysteines separated by 3-4 (3 in this case) amino acids in the first immunoglobulin domain.

15 Finally, 8 amino acids C-terminal to the first double cysteines is a stretch of 9 amino acids that is identical to a sequence in murine MAdCAM-1. Within this region was the sequence LDTSL, which aligns with a consensus motif for integrin/Ig family member interactions. Although this 20 motif has general conservation with respect to other Ig adhesion receptors such as ICAM-1, ICAM-2, ICAM-3 and VCAM-1 (Osborn, et al., J. Cell. Biol, 124:601-608 (1994); Renz, et al., J. Cell. Biol., 125:1395-1406 (1994)), this exact sequence was previously found only in 25 murine MAdCAM-1. The functional significance of this motif is suggested by the fact that a point mutation which changes the first L (leucine) of the motif at amino acid 61 to an R (arginine) in murine MAdCAM-1 had a dramatic effect on MAdCAM-1: @a4f7 binding (not shown). The results of the functional studies together with these sequence characteristics indicate that clone 31D encodes a primate homolog to murine MAdCAM-1.

_Ilr -54- Screening of a human pha e library and uriication of human clones Human phage cDNA libraries were constructed in the XZiploxTM vector (Gibco/BRL). Human cDNA was made from RNA isolated from either normal or inflamed mesenteric lymph nodes (MLN) as described above. cDNA was synthesized as described above, ligated into the phage vector, and titered on bacterial strain Y1090 (ZL) Ziplox). Duplicate filters from approximately 500,000 independent clones (50,000 clones/filter) from both the normal and the Crohn's MLN phage libraries were screened with 32 P-labeled fulllength macaque MAdCAM-1 cDNA.

prepare the probe, a ~1.7 kb EcoRI-NotI fragment was excised from clone 31D, and isolated using GeneClean :o 15 (BIO 101). The fragment was labeled with a32-dCTP by priming with random hexamers (Maniatis et al., In: Moleculer Cloning (Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1990)).

Screening conditions were as follows: 50,000 phage clones were plated on 150 mm petri dishes containing

NZYCM

agar (Gibco/BRL). After incubation ranging from 7-16 hours, the plates were overlaid with 132 mm nitrocellulose filters (Schleicher and Schuell, Keene, NH) for 2 minutes and then five minutes to transfer first and second (duplicate) lifts of phage clones, respectively. Filters were then soaked for 5 minutes in denaturing solution M sodium chloride, 0.5 N sodium hydroxide) followed by neutralization in 1.5 M sodium chloride, 0.5 M Tris-HC1, pH Filters were air dried for 15 minutes and then baked under vacum for 2 hours at 80 0

C.

Filters were pre-hybridized for 2 hours at 55 0 C in 2M Na 2

HPO

4 0.5% SDS, 5X Denhardt's (lX Denhardt's solution is 0.02% bovine serum albumin, 0.02% ficoll, and 0.02% polyvinyl-pyrolidone), 1 mM EDTA, and 50 pg/ml denatured salmon sperm DNA, and subsequently hybridized overnight at 0 C in the same buffer. Filters were washed once at room temperature in 2X SSC, 0.1% SDS (lX SSC is 0.15 M sodium chloride, 0.015 M sodium citrate), followed by three to four washes at 650C in 0.1X SSC and 0.1% SDS. Filters were monitored with a Geiger counter to see that the background was reduced.

Positive clones were plaque purified, and the plasmid pZLl containing the cDNA inserts was rescued using the CRE LOX recombination system (GIBCO) (plasmid pZLl is containined within the body of the lambda Ziplox vector).

In particular, a purified phage plaque was suspended in 200 pl of phage buffer (20 mM Tris HC1, pH 7.5, 145 mM NaCl, 8 mM MgSO 4 7H20, 0.01% gelatin) for 5 minutes at room 15 temperature. 20 Al of the phage suspension was then added to 100 Al of an overnight culture of DH1OB (ZL) and incubated for an additional 5 minutes. Dilutions of the mixture were then plated on LB plates supplemented with ampicillin at 50 Ag/ml and 10 mM MgCl 2 and incubated overnight at 30 0 Single colonies, now containing the cDNA inserted into the pZL1 vector were grown as standard overnight cultures and plasmids were then purified using Qiagen plasmid purification reagents.

Identification of distinct functional human MAdCAM-1 CDNA clones Two human cDNA libraries from histologically normal human mesenteric lymph nodes, and inflamed mesenteric lymph nodes from a patient with Crohn's disease were screened using the entire macaque MAdCAM-1 cDNA as a probe. One cross-hybridizing clone was isolated from the normal library, and two cross-hybridizing clones were isolated from the Crohn's library. One of the two clones isolated from the Crohn's library was about 1.3 kb, appeared to be -56incomplete at the S'-end, and was not sequenced. The clone from the normal library (clone 4) was slightly larger (1624 bp) than the longer clone (1558 bp) isolated from the Crohn's library (clone 20). Although these two cDNAs differ in size by approximately 100 bp, their 5' and 3' untranslated sequences were almost identical in length.

Each clone appeared full-length, as they both contained an amino-terminal signal sequence that was almost identical to the macaque sequence.

Additionally, preliminary sequencing demonstrated the same distinguishing characteristics of the amino-terminal Ig-like domain as the primate cDNA. Since the differences in the size of these clones could not be attributed to the length of the untranslated sequences, it seemed likely that 15 the variation resided in the coding region.

In order to determine whether each clone encoded functional human MAdCAM-1, the inserts of each clone were subcloned into the pCDNA-3 expression vector (Invitrogen, San Diego, CA), which carries a neo resistance gene suitable for G418 selection. The human cDNAs (which were made using NotI oligo-dT primers at the 3'-end, and SalI adapters at the 5'-end) were ligated into the XZipLox vector, which contains plasmid pZL1. pZL1 vectors with cDNA inserts were rescued as described above. For subcloning, the inserts of clones 4 and 20 were each released by digestion from the pZLl backbone with EcoRI and NotI. The EcoRI-NotI fragments were isolated by Geneclean (Bio 101) following electrophoresis on a 1% agarose gel, and the fragments were ligated into pcDNA-3 which had been cleaved with EcoRI and NotI. The ligation mixture was used to transform a DH10B E. coli Max efficiency strain (GIBCO), and transformants were obtained following selection on LB agar supplemented with 50 gg/ml ampicillin (Amp). Plasmids designated pcD3huMAd4 (insert r _II -57from clone 4) and pcD3huMAd20 (insert from clone 20) were obtained and analyzed by restriction digestion.

Clone pcD3huMAd4 (insert from clone 4) or pcD3huMAd20 (insert from clone 20) was transiently transfected into CHO/P cells. Each clone directed the expression of a functional protein which could mediate binding and adhesion, as assessed by adhesion of CHO/P transfectants to the human B cell lymphoma RPMI 8866 (Figure 5) or to TK1 cells (not shown).

Adhesion of the CHO/P transfectants to RPMI 8866 cells was blocked by preincubation with anti-a407 MAb ACT-1, but not by control IgG. Adhesion of transfectants to TK1 cells was blocked by anti-07 MAb FIB 504. These results indicate that clone 4 (from a normal mesenteric node library) and 15 clone 20 (from a Crohn's library) each encode functional MAdCAM- 1 proteins. To further characterize these distinct cDNAs, both clones were completely sequenced.

Results The cDNAs from Clones 4 and 20, encoding human MAdCAM-1, are 1628 bp and 1543 bp, respectively, in length.

cDNA from Clone 4 (Figure 1; SEQ ID NO:1) contains an open reading frame of 1218 bp encoding a predicted protein of 406 amino acids (SEQ ID NO:2), and a 3' untranslated region of 410 bp, but contains no 5' untranslated region. cDNA 25 from Clone 20 (Figure 2; SEQ ID NO:3) contains 4 bp of untranslated sequence, an open reading frame of 1146 bp encoding a predicted protein of 382 amino acids (SEQ ID NO:4), and a 3' untranslated region of 393 bp. The predicted molecular masses of the encoded proteins, after cleavage of a predicted signal sequence of 18 amino acids are 40,910 (clone 4) and 38,375 (clone 20) daltons.

Multiple alignments were performed to analyze the degree of similarity between the different cloned species of MAdCAM-1. Nucleotide alignments revealed 81.9% sequence I -58similarity between mouse and rat MAdCAM-1 cDNAs, 41.8% similarity between mouse and macaque cDNAs, 42.1% similarity between murine and human (Clone 4) MAdCAM-1 cDNAs, and 41.8% similarity between murine and human (Clone 20) MAdCAM-1 cDNAs. Alignment of the nucleotide sequences of macaque MAdCAM-1 with human Clone 4 and Clone 20 cDNAs revealed sequence similarities of 70.7% and 75.0%, respectively.

The amino acid sequence similarities were determined to be 78.5% between mouse and rat MAdCAM-1, 44.3% between mouse and macaque, and 39% between murine and MAdCAM-1 encoded by human Clone 4.

Comparisons of cDNA clones 4 and 20 revealed a region which is homologous to the mucin domain of murine MAdCAM-1, 15 due to a prevalence of serine, threonine and proline (69% for clone 4 and 76% for clone 20) residues (boxed in Figure 1 and Figure This region, although similar in amino acid composition to murine MAdCAM-1, is highly divergent from murine MAdCAM-1. The domain is 71 amino acids long in 20 clone 4, and 47 amino acids long in clone 20. This region also contains two polymorphisms: a polymorphism at amino acid 240, which is proline in clone 4 and serine in clone 20; and a polymorphism at amino acid 242, which is asparagine in clone 4 and aspartate in clone 20. In addition, the human mucin domains contain a repeat of 8 amino acids consisting of the sequence PPDTTS(Q/P)E, which appears eight times in clone 4 and five times in clone To assess the origin of clones 4 and 20, PCR primers flanking the repeat were used to amplify human genomic DNA.

The following primers were used: TAC TGC CAG GCC ACG-3' (Primer SEQ ID NO:7) CTG GGA GAT CTC AGG G-3' (Primer SEQ ID NO:8) -59- ACG ATG AGG CTG CCT GG-3' (Primer SEQ ID NO:9) GAG CCT GGG CTC CTG GG-3' (Primer SEQ ID The primers were nested primers. In the first reaction, primers 1 and 2 were used. For the second amplification reaction, a 1:1000 dilution of the first reaction was prepared, and 1 pl was used with primers 3 and 4.

Amplification reactions contained either 0.5 ig of genomic DNA, 10 picograms of control plasmids (pcD3HuMAd4 or pcD3HuMAd20), or approximately 1 ng of double-stranded cDNA that was prepared previously for the ZipLox libraries.

Genomic DNA was obtained from three sources (Promega; ClonTech, and by purification from Jurkat cells). The conditions of amplification were: one cycle for 5 minutes at 94 0 C; 25 cycles at 94 0 C for 45 seconds; 60 0 C for 15 seconds and 72°C for one minute followed by one cycle for *99 minutes at 72 0

C.

The amplification reactions from genomic DNA yielded two bands which comigrated with the individual products of PCR reactions using either clone 4 or clone 20 cDNA as template. This data suggests that the two cDNA clones are isoforms encoded by genomic DNA, and are probably generated by alternative splicing or by transcription of two different alleles. Extensive polymorphism and sequence divergence has been documented in other mucin sequences Hilkens, J. et al., Trends, Biochem. Sci, 17: 359-363 (1992)). For example, repetitive portions of intestinal mucins are not well-conserved between rodents and humans (Gum, J.G. et al., J. Biol. Chem., 266: 22733- 22738 (1991)). One caveat is that, based on an analysis of murine genomic structure, the human genomic DNA could contain an intron in this region. If so, the PCR primers used in this experiment would span the intron, and amplification of human genomic DNA would not be expected to produce bands of the same size as those produced by amplification of the cDNA controls. Isolation and analysis of human MAdCAM-1 genomic clones can conclusively exclude the possibility of a cloning artifact. Further assessment of normal and/or inflamed tissue from normal individuals and from patients with IBD, Crohn's disease or other inflammatory conditions can be performed to determine if there is a correlation between the clone 20 isoform and an inflammatory disease or activity.

The comparison of murine, macaque, and two isoforms of human MAdCAM-1 indicates that the amino-terminal portions of these receptors exhibit domain structures likely to be ~involved in recognition of a4/7. In contrast, the regions of these receptors in a location corresponding to the 15 location of the mucin/IgA domain of murine MAdCAM-1 display similar amino acid compositions (serine, threonine, proline-rich mucin regions), but are more divergent from one another.

o Expression of human MAdCAM-1 RNA Northern analysis was carried out using human multiple tissue Northerns I and II (commercially prepared by Clontech, Palo Alto, CA), or 2 Ag of poly A+ RNA from cell ~lines and tissues that were prepared as described above.

RNA was denatured and electrophoresed through a 1% agarose o* 25 formaldehyde gel and transferred to a PVDF (Immobilon, Millipore) membrane by standard capillary blot procedures.

RNA samples were stained with ethidium bromide to initially ensure that the quality and quantity of each cell or tissue RNA was equivalent. After transfer, RNA was fixed to membranes by UV crosslinking (Stratalinker, Stratagene) and this blot and the commercially prepared blots were prehybridized at 68'C for 1 hour in ExpressHyb (Clontech).

The cDNA insert from clone 4 was labeled with a32-dCTP by -61priming with random hexamers (Maniatis et al., In Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, (199)).

Hybridization was performed at 680C for 1 hour in ExpressHyb with denatured probe at a concentration of 2 X 106 cpm/ml.

Blots were then washed in 0.lX SSC, 0.1% SDS for minutes at 650C with one change of wash at 30 minutes. The exposure time was 48 hours with an intensifying screen.

After this exposure, the blot was stripped by washing for minutes in 0.5% SDS and rehybridized under the same conditions with a -actin cDNA. The exposure time was 2 hours.

."Results 15 Northern blots were probed for MAdCAM-1 expression using the entire cDNA insert from clone 4 as a probe. A single RNA species of approximately 1.6 kb was highly expressed in the small intestine and was expressed to a lesser extent in the colon and spleen. No significant 20 expression was observed in other tissues examined under these conditions, including thymus, prostate, ovaries, testes and peripheral blood leukocytes (PBL). This tissuespecific pattern of expression is consistent with studies in the mouse showing restricted expression of MAdCAM-1 in 25 Peyer's Patches, MLN (mesenteric lymph node), intestinal lamina propria and some expression in the marginal sinus around splenic white pulp nodules in the spleen (Hemler, Annu. Rev. Immunol., 8:365 (1990); Berg, et al., Cellular and molecular mechanisms of inflammation, 2:111 (1991); Briskin, et al., Nature, 363:461 (1993)). No significant expression was observed in other tissues examined, including heart, brain, placenta, lung, liver, skeletal muscle, or kidney; however, low levels of -62expression were detected in pancreas. These data indicate that human MAdCAM-1 expression is tissue-specific with expression in mucosal tissues and spleen; a thorough immunohistochemical analysis of tissue distribution can be performed using monoclonal antibodies against human MAdCAM-1 (see below).

Example 2. Characterization of MAdCAM-1 Clones Functional adhesion assays Plasmids: The following plasmids were used in the functional adhesion assays: pSV-SPORT-1 (Gibco/BRL) or pcDNA-3 *i (Invitrogen) were used as controls; murine MAdCAM-1 in pCDM8 (pCDMAD-7; Briskin, et al., Nature, 363:461 (1993)); seven domain human VCAM-1 (Polte, et al.

15 Nucleic Acids Res., 18:5901 (1990)) in pcDNA3 (pCD3VCAM); 9 and human MAdCAM-1 in pcDNA-3 (pCDhuMAd4) (see above).

Monoclonal antibodies: The following monoclonal antibodies (MAb) were used in the functional adhesion assays: anti-murine MAdCAM-1 20 MAb MECA-367 (American Type Culture Collection (Rockville, MD), Accession No. HB9478; Streeter, et al., Nature, 331:41 (1988); and U.S. Patent No. 5,403,919 to Butcher); anti-human VCAM-1 MAb 2G7 (American Type Culture Collection (Rockville, MD); Graber, et al., J.

Immunol., 145:819-830 (1990)); anti-murine a467 MAb DATK 32 (Andrew, et al., J. Immunol., 153:3847-3861 (1994)); anti-murine $7 MAb FIB 504 (Andrew, et al., J. Immunol., 153: 3847 (1994)); anti-human a407 MAb ACT-1 (Lazarovits, et al., J. Immunol., 133:1857 (1984)); anti-human integrin f1 (CD29) (Becton Dickinson; San Jose, CA, Cat. #550034); and murine IgGi and rat IgG2A as irrelevant controls.

-63- Cell Lines: The following cell lines were used in functional adhesion assays: Murine T cell lymphoma TKl (Butcher, et al., Eur. J. Immunol., 10:556-561 (1980); E. Butcher (Stanford, CA); RPMI 8866, a human B cell lymphoma line which expresses a407 (and not a4fl) (American Type Culture Collection (Rockville, MD); Erle, et al., J.

Immunol., 153:517 (1994); a gift from D. Erle);

JURKAT,

a human T cell line which expresses a431 (and not a4f7) (American Type Culture Collection (Rockville, and (4) Ramos, a human (B lymphocytic) Burkitt lymphoma cell line that expresses a4fl3 (and not a4#7) (American Type Culture Collection (Rockville, MA), Accession No. ATCC CRL 1596).

15 Functional adhesion assays: For functional adhesion assays, plasmids encoding various species of MAdCAM-1, human VCAM-1, and control S" plasmids were introduced by transient transfection into o. CHO/P cells as described above (Example 1) with the following modifications. As several wells were to be transfected for antibody inhibition studies, a master liposome mix with multiples of the wells to be transfected was first made for each plasmid. This ensured that the same liposome mixture was transfected into each well.

48 hours after transfection, the medium was removed.

An antibody supernatant (0.25 mls) (containing either antihuman VCAM-1 MAb 2G7 or anti-murine MAdCAM-1 MAb MECA-367), or 0.25 mls of adhesion assay buffer as a control were added, and the mixture was preincubated at 4oC for minutes.

In parallel, lymphocyte cell lines (RPMI 8866 or Jurkat) were spun down and resuspended at a density of 2 X 10 6 /ml in assay buffer consisting of HBSS (without Ca -64or Mg supplemented with 2% bovine calf serum, 20 mM HEPES pH 7.3, 2 mM Mg and 2 mM Ca 0.25 ml aliquots X 105 cells) of these RPMI 8866 or JURKAT cell suspensions were preincubated with a small volume of various purified antibodies or with an equal volume of DATK 32 supernatant at 4 0 C for 15 minutes. Where DATK 32 was used in a preincubation with a cell line, prior to the start of the assay, the supernatant or buffer present in the wells (containing the transfectants) was aspirated in order to obtain volume of 0.5 ml total for the adhesion assay.

For preincubations, purified antibodies (ACT 1, FIB 504 anti-l) and control IgG antibodies were used at concentrations of 20 jg/ml. 0.25 mls of antibody supernatant (used neat) containing anti-human VCAM-1 (MAb 2G7) or anti-murine MAdCAM-1 (MAb MECA-367) were used in preincubations. 0.25 mls of antibody supernatant of DATK 32 were used in the preincubation.

S2 After the preincubations, cell lines (Jurkat or RPMI 8866) were combined with the tranfectants in the wells, and incubation on a rocking platform was continued for an additional 30 minutes at 4 0

C.

Assays were fixed as described above. Plates were washed by gentle inversion in a large beaker of phosphate S. 25 buffered saline (PBS), followed by inversion in a beaker of PBS with 1.5% gluteraldehyde for fixation for a minimum of 1 hour. Adhesion was assessed by counting both lymphocytes and CHO cells in a field at 20X magnification. For each assay, the number of lymphocytes bound per CHO/P cell was averaged as a minimum of four fields with standard error.

Results in each case are from one of three experiments performed with similar results.

Results Murine MAdCAM-1 specifically binds lymphocytesexpressing a437 (and not a4ll). In order to determine the specificity of human MAdCAM-1 lymphocyte interactions, adhesion assays were performed to assess the ability of transiently transfected CHO/P cells expressing human MAdCAM-1 to bind to the RPMI 8866 cell line which only expresses a407 (Erle, et al., J. Immunol., 153:517 (1994)), or to the T cell line Jurkat, which exclusively express a401. Binding of these cell lines was compared to that of transiently transfected CHO/P cells expressing murine MAdCAM-1 and human VCAM-1. The results are presented in Figures 4A-4B.

S" RPMI 8866 cells did not bind to control transfectants, but avidly bound to transfectants expressing human or murine MAdCAM-l. This binding was completely inhibited by preincubation with anti-a4f7 MAb ACT-1 (Figure 4A). VCAM-1 transfectants failed to bind RPMI 8866, which is consistent with the previous demonstration that a407/VCAM-1 interactions are activation-dependent (Postigo, et al., J. Immunol., 151:2471-2483 (1993); Ruegg, et al., J. Cell. Biol., 117:179-189 (1992)). The failure of RPMI 8866 cells to bind VCAM-1 transfectants was not due to lack of expression, as FACS analysis using anti-VCAM-1 MAb 2G7 indicated a transfection efficiency of 60%. Moroever, the same VCAM-1 transfectants were able to bind Jurkat cells, and binding was completely inhibited by preincubation with either anti-VCAM-1 or anti-31 MAbs (Figure 4B). Murine and human MAdCAM-1 transfectants did not bind Jurkat cells (an a43l positive line). These data demonstrate that human MAdCAM-1 can selectively bind to human leukocytes lymphocytes expressing a4f7 integrins.

_8 -66- L1-2 and CHO cell transfectants The mouse L1-2 cell line is derived from a pre-B lymphoma, and was obtained from Dr. Eugene Butcher (Stanford University, Stanford, CA). The genes encoding either the macaque or human cDNAs for MAdCAM-1 were subcloned into the pcDNA-3 vector (Invitrogen) as described above. The resulting plasmids (pcD3HuMAd4, pcD3HuMAd20, or pCD3PMad (macaque)) were introduced into L1-2 cells by transfection as follows: Ll-2 cells were grown to a 0 density of approximately 10 6 /ml. Either 50, 25 or 12.5 million cells were washed in HBSS and then resuspended in a 0.8 ml of a buffer consisting of Hanks balanced salt solution supplemented with 20 mM HEPES, pH 7.05. A solution consisting of 20 g of linearized plasmid, 500 Ag of tRNA and HBSS to bring the final volume to 200 gl was added to the cell suspension to bring the total volume to iml. After a 10 minute incubation at room temperature the cell/DNA mixture was transferred to an electroporation cuvette (BioRad, Richmond, CA) and electroporated at 250 20 volts, 960 mF in a BioRad gene pulser. Following another 1 0 minute incubation at room temperature, the cells were diluted to 25 ml in standard Ll-2 growth media (RMPI 1640, 10% Hyclone fetal bovine serum, 50 U/ml Penicillin/Styreptomycin (Gibco) and 0.29 mg/ml L Glutamine (Gibco) and returned to the incubator at 37C. 48 hours later, the cells were pelleted by centrifugation and resuspended in 50 ml of L1-2 media supplemented with G418 (Geneticin; Gibco) at 0.8 mg/ml. Dilutions of the cell suspension were plated in 96-well microtiter plates and single colonies were grown up analyzed for expression of MAdCAM-1.

L1-2 cell clones expressing MAdCAM-1 could be detected by adherence to TK1 cells. Ll-2 (non-transfected cells) and TK1 cells both grow as single cell suspensions.

-67- Surface expression of MAdCAM-1 can be detected by its ability to mediate adhesion by virtue of its interaction with a4f7 expressed on TK1 cells. Specificity of this interaction was further demonstrated by inhibition by pretreatment of TK1 cells with anti-#7 MAb FIB 504.

CHO cells (Chinese Hamster Ovary Cells; American Type Culture Collection (Rockville, MD)) stably transfected with either the macaque or human MAdCAM-1 clones were prepared by electroporation as described above for the Ll-2 cells 10 with the following exceptions. Media for CHO cell growth was a-MEM with deoxyribonucleosides (Gibco) and 10% fetal calf serum (Gibco) and 50 U/ml Penicillin/Streptomycin (Gibco) and 0.29 mg/ml L Glutamine (Gibco). Selection media consisted of the same media with 0.55 mg/ml G418 15 (Gibco). Single clones were grown up and analyzed for their ability to exhibit a4f7-dependent binding of RPMI 8866 cells using the functional adhesion assay described above (for transients), except that cells were S• plated at 50,000 cells per well in a 24-well plate the day 20 before the assay. Using this criteria, a line called CHO HuMAd 4 was established.

Monoclonal antibodies capable of inhibiting adhesion Monoclonal antibodies against human MAdCAM-1 were generated by immunizing C57BL/6 mice with L1-2 MAdCAM-1 transfectants. Mice were immunized intraperitoneally with million cells resuspended in HBSS three times at two week intervals, and a final fourth immunization (of million cells resuspended in HBSS) was injected intravenously. The first immunization was performed with a mixture of two clones (L1-2 cell clone 23 and clone 19) expressing macaque MAdCAM-1. The remaining boosts were done with a single Ll-2 clone (L1-2 clone HuMAD4/17) expressing human MAdCAM-1.

-68- A successful fusion was performed which generated approximately 5,000 hybridomas. Four days after the final intravenous injection, the spleen was removed and a single cell suspension was prepared in serum free DMEM media.

These cells were fused with the fusion partner according to the method of Galfre et al. (Galfre, et al., Nature, 299:550-552 (1977)). 20 ml of spleen cells and 20 ml of SP2/0 cells were combined, spun at 800 g for minutes and the media was removed by aspiration.

A

1 0 solution of 50% polyethylene glycol 1500 (PEG 1500) (Boehringer Mannheim, Indianapolis, IN) prewarmed to 37 0

C

was added to the cell pellet over 2 minutes, followed by ml of DMEM media over 3 minutes. The cell suspension was spun at 600 g for 3 minutes and the supernatant was 15 removed. The pellet was resuspended gently in DMEM media containing 20% fetal calf serum, 2 mM L-glutamine, 100 U/ml penicillin, 100 g/ml streptomycin sulfate, and HAT selection media (Sigma, St. Louis, MO). Cells were plated into ten 96-well flat bottom microtiter plates at 200 20 Al/well.

Ten days after the fusion, supernatants from the wells were screened for reactivity against CHO human MAdCAM-1 transfectants (CHO HuMAd 4 cells), by fluorescence staining. Staining of 500,000 cells per sample was performed essentially as described, using 50 l of each supernatant and 50 pl cells Harlow and D. Lane, 1989, In: Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY). The secondary antibody was an FITC-labeled anti-murine IgG (H L) (Jackson Labs) that was diluted 1:200. Strong reactivity was judged as a 2-3 log increase in fluorescence of as compared with untransfected CHO cells.

48 antibody supernatants were selected for strong reactivity against CHO HuMAd 4 cells. These antibody supernatants were then screened for their ability to block -69the adhesion of CHO HuMAd 4 cells to RPMI 8866 cells. As a control, the ability of supernatants to inhibit Ramos cell binding to VCAM-1 transfectants was examined, as it should not be affected by a specific anti-human MAdCAM-1 MAb. To identify blocking anti-human MAdCAM-1 monoclonal antibodies, the following assay was performed. To provide control transfectants, CHO/P cells were transfected with pCD3VCAM as described above, and were assayed 48 hours after transfection. 48 hours before the adhesion inhibition assay, 40,000 cells per well of VCAM-1 transient transfectants were plated into 24 well plates. 24 hours before assay, 50,000 cells per well of CHOHuMAd 4 transfectants were plated in 24 well plates. On the day of the assay, each anti-human MAdCAM-1 supernatant (0.25 mls) 15 was added to a well containing either CHOHuMAD 4 transfectants or VCAM-1 transfectants, and the mixture was preincubated at 4 0 C for 15 minutes. Adhesion assays were performed, using RPMI 8866 cells with the MAdCAM-1 transfectants or Ramos cells (a human B cell line that expresses a4f1) with the VCAM-1 transfectants.

In parallel, cells (RPMI 8866 or Ramos) were resuspended at a density of 2 X 10 6 /ml in an assay buffer consisting of HBSS (without Ca or Mg supplemented with 2% bovine calf serum, 20 mM HEPES pH 7.3, 2 mM Mg and 2 mM Ca After the preincubation of the transfectants with antibody, 0.25 mls of the RPMI 8866 or Ramos cell suspensions (5 x 105 cells) were added to each well, and incubation on a rocking platform was continued for an additional 30 minutes at 4 0 C. The wells were washed, fixed and examined as described above to assess inhibition of binding.

Eleven out of 48 of the hybridoma supernatants examined displayed substantial blocking activity, inhibiting the adhesion of RPMI 8866 cells to transfectants expressing MAdCAM-1. Adhesion of Ramos cells to transfectants expressing VCAM-1 was unaffected, indicating selective inhibition of a467-mediated interactions.

Selected blocking hybridomas were subcloned by limiting dilution.

Results Stable cell lines expressing macaque or human MAdCAM-1 were made in the murine pre-B lymphoma L1-2. These cells were used to immunize C57BL/6 mice and prepare hybridomas.

The resulting fusion was screened by immunoflourescence staining of CHO HuMAd 4 transfectants expressing human MAdCAM-1. Screening of approximately 1,000 wells produced 48 supernatants exhibiting strong reactivity against the CHO HuMAd 4 transfectants, while non-transfected CHO cells 15 were negative. These supernatants were subsequently tested for their ability to specifically block adhesion of RPMI 8866 cells to human MAdCAM-1 transfectants.

11 of the 48 hybridoma supernatants examined could specifically inhibit the adhesion of RPMI 8866 cells to MAdCAM- 1 while adhesion of Ramos cells (which express a431) to VCAM- 1 transfectants was unaffected by the same supernatants. These hybridomas were designated 10G4, 8C1 10G3, 9G12, 9E4, 7H12, 10F2, 10A6, 1E5, 2F5, 7G11.

E xample 3. Design and functional analysis of a human MAdCAM-1-IqG chimera Construction of MAdCAM-IqG Chimera Human MAdCAM-1 clone 4 cDNA in pCDNA3 (Invitrogen, San Diego, called pcD3huMAd4 (Example 1) was used as a template for PCR amplification of extracellular regions of human MAdCAM-1 to be fused with the constant region of human IgGl. Primer HUMADIG4/2 (SEQ ID NO:11), which -71contains the 5' end of human MAdCAM-1 coding sequence

(ATG

codon, bold), was synthesized: HindIII 5'-GGAAGCTTCCACCATGGATTTCGGACTGGCCC-3, This 5' primer was used in conjunction with a 3' primer designated HUMADIG2 (SEQ ID NO:12) to amplify regions encoding the two amino-terminal immunoglobulin-like (Ig) domains of human MAdCAM-1. Primer HUMADIG2 (SEQ ID NO:12), which contains a portion complementary to coding strand nucleotides 667-683 of SEQ ID NO:1, has the following sequence: Spe 5'-CCGACTAGTGTCGGGCTGTGCAGGAC-3' Alternatively, the 5' primer was used in conjunction with 15 3' primer HUMADIG3 to amplify a region encoding the entire extracellular domain of human MAdCAM-1 (clone The 3' primer HUMADIG2 (SEQ ID NO:12), which contains a portion complementary to coding strand nucleotides 992-1010 of SEQ ID NO:1, has the following sequence: 20 Spel 5'-GGACTAGTGGTTTGGACGAGCCTGTTG-3' 9 The primers were designed with a 5' HindIII site or 3' Spel sites as indicated. These primers were used to PCR amplify two different MAdCAM fragments, using a PCR optimizer kit from Invitrogen (San Diego, CA). The PCR products were digested with the enzymes HindIII and Spel to generate ends for cloning. The products were subsequently purified by gel electrophoresis using the Glassmax DNA isolation system (Gibco, Bethesda MD).

A -1 kb fragment encompassing the CH1, H (hinge), CH2 and CH3 regions was excised by digestion with Spel and EcoRI from a construct encoding a human immunoglobulin 1y -72heavy chain having an Fc-mutated human constant region.

The antibody encoded by this construct was used as an isotype matched irrelevant control hereinbelow. The human constant region in this construct was originally obtained by PCR amplification of the CAMPATH-1H heavy chain (Reichmann, L. et al., Nature, 322: 323-327 (1988)) as described by Sims, M.J. et al. Immunol., 151: 2296-2308 (1993)) and Waldmann et al. (WO 93/02191, February 4, 1993 (page the teachings of which are each incorporated herein by reference in their entirety. The mutations in the constant region of this construct (Leu 2 34 Ala 234 and Gly 237 Ala 7 were designed to reduce binding to human Fcy receptors, and were produced by oligonucleotide-directed mutagenesis. Thus, the MAdCAM-Ig fusions produced contain o 15 the SpeI-EcoRI constant region fragment described by Sims et al. Immunol., 151: 2296-2308 (1993)) and Waldmann et al. (WO 93/02191), except for the introduction of Leu 34 Ala 2 3 4 and Gly 237 Ala 237 mutations.

The 1 kb SpeI-EcoRI fragment encoding the Fc-mutated IgG1 constant region was isolated by gel electrophoresis using the Glassmax DNA isolation system (Gibco, Bethesda MD). This constant region fragment, the HindIII-Spel fragments containing either the two N-terminal Ig domains of MAdCAM-1 or the entire extracellular domain, 25 were ligated in a three-way ligation to vector pEE12 (Stephens, P.L. and M.L. Cockett, Nucl. Acids Res., 17: 7110 (1989) and Bebbington, C.R. and C.C.G. Hentschel, 1987, The use of vectors based on gene amplification for the expression of cloned genes in mammalian cells, (Academic Press, which had been digested with HindIII and EcoRI. Transformants of the bacterial strain were obtained. Colonies were grown and mini plasmid preps were analyzed by restriction mapping. Three constructs which encode fusion proteins comprising either -73the entire extracellular domain of MAdCAM-1 (construct HuMAdIg2l) or the two N-terminal Ig domains (construct HuMAdIg31 or HuMAdIg38) fused to the Fc-mutated IgGi constant region, were sequenced across the entire MAdCAM-1 portions, confirming proper fusion of segments and the absence of PCR induced mutations.

For initial testing, each construct was transiently transfected onto monolayers of 5 X 107 COS cells in 1 ml of RPMI buffer (no serum) and 25 Ag of plasmid using electroporation with a Biorad Gene Pulser under standard conditions (960 MF, 250 72-96 hours after transfection, supernatants were harvested, passaged through 0.45 A filters and stored at 4 0 C in the presence of 0.05% sodium azide. Production of chimera was confirmed by a 15 sandwich ELISA, using an anti-human IgG1 antibody as capture antibody and the same antibody, which was conjugated to alkaline phosphatase as second antibody for detection. Irrelevant control antibody (having an identical constant region) was used as a standard. The 20 chimera was also analyzed by western blotting using an anti-human MAdCAM-1 monoclonal antibody, and was found to run at approximately 200 kd, consistent with the size of a homodimer.

Soluble human MAdCAM-Iq chimeras specifically bind a487 positive cells Supernatants from four different transfections were assayed for their ability to stain the T cell line HuT 78, which was previously shown to bind MAdCAM-1 only in the presence of Accordingly, each solution used in this assay contained 2 mM HuT 78 cells (a human T cell lymphoma line; American Type Culture Collection, Accession No. ATCC TIB 161) are a4f7-bearing cells. To test the binding specificity of the chimeras, Hut 78 cells were -74preincubated with either media alone (RPMI 1640 with 2% FCS) or media and 10 gg/ml of the anti-f7 antibody FIB 504.

Approximately 100,000 cells were incubated on ice for minutes and then washed with HBSS plus 2% FCS 2 mM Ca++/ 2 mM Cells were then incubated for 20 minutes on ice with media once again or with supernatants from one of four independent transfections, two with a chimera containing the entire extracellular domain of MAdCAM-1 (clone 21) and two with a truncated form of MAdCAM containing the two N-terminal Ig domains (clone 38) for minutes. After washing, cells were then incubated with an anti-human IgG antibody conjugated with phycoerythrin and staining above background was assessed by flow cytometry (FACScan). Only cells incubated with the chimera 15 supernatants stained above background, while preincubation with the g7 MAb reduced this staining to background levels, indicating a specific interaction of the chimera with the a4/7 integrin (Figures 17A-17E).

o* Permanent NSO cell lines secreting human MAdCAM-Ig 20 chimera were selected after transfection by electroporation, by growth in a glutamine free media as previously described (Cockett, et al., Bio/Technology, 8: 662-667 (1990)). Cloned lines were adapted to growth in spinner culture. Supernatants from 25 three of these cloned lines (samples and a partially purified chimera (Clone 21, purified by binding to protein A, sample A) were tested for their ability to support adhesion of the B cell line RPMI 8866. Briefly,

NEN

maxisorb plates were incubated with 100 gl/well of Protein A at 20 Ag/ml in carbonate buffer, pH 9.5 overnight at 4 0 C. Plates were then washed 2X with RPMI 1640 media (no serum). 100 gl of chimera (or serial dilutions in RPMI) were bound to the wells at 370 for 2 hours and then washed once. Wells were then blocked with FCS for 1 hour at 37 0 C, washed once, and then preincubated with tissue culture supernatants containing either an anti-human VCAM-1 MAb (2G7) as a control or the anti-human MAdCAM-1 MAb 10G3 (Example 2G7 and 10G3 MAbs were removed before addition of cells. RPMI 8866 cells were fluorescently labeled by preincubation with BCECF-AM stain (BCECF-AM; 2',7'-bis-(2-carboxyethyl)-5-(and-6)-carboxyflourescein, acetoxymethyl ester; Molecular Probes), 100 pl of cells were added to each well (to a final concentration of 105 cell/well), and incubated on a rotary shaker for 30 minutes at room temperature. Binding of RPMI 8866 cells to immobilized chimeras was assessed by reading flourescence values using a Fluorescence Concentration Analyzer (IDEXX).

Specific binding was demonstrated as only the anti-human 15 MAdCAM-I MAb could block binding of cells to MAdCAM-Ig chimera (Table 1).

These and other such chimeric fusion proteins can be used for assessing the ability of an agent small molecule) to block a407 binding to chimera, to identify inhibitors of a4I7-MAdCAM interaction. Additionally, since chimeric fusion proteins can bind to a4f7 positive *lymphocytes in solution, they provide candidate inhibitors of in vivo lymphocyte recruitment to inflammatory sites.

S.

9amle Anti HUMAd MAbs

A-

a. a a a.s a..

a a a a a. a a a .a a a a a. a TABLE 1 Neat 195527 3860 195527 6526 195527 4566 195527 7350 1:2 195527 3092 195527 3626 195527 4094 30840 6794 1:4 195527 1746 195527 3274 35548 3922 46852 6020 1:8 18560 2200 195528 2978 9570 3492 16270 4510 1:16 4254 482 52056 1648 21782 2436 5474 6548 1:32 3558 564 2932 1518 1926 2566 2656 5122

C

The a4o7 Positive B Cell Line RPMI 8866 Specifically Binds Soluble Human MAdCAM Ig Chimera. The Human MAdCAM-l Ig Chimera Clone 21 that was partially purified over protein A (Sample A) or tissue culture supernatants from different NSO clones (Samples B-D) were immobilized on 96 well plates with protein A, and either preincubated with an anti-VCAM-1 MAb 2G7 (designated under MAbs) as a negative control or with the anti-human MAdCAM MAb 10G3 Purified chimera or supernatants (used either undiluted ("neat") or at serial 1:2 dilutions) was bound to wells via Protein A, and incubated with fluorescently-labeled RPMI 8866 cells on a rotary shaker for 30 minutes at room temperature. After washing with an automated plate washer (EL 404 Microplate Autowasher, BIO-TEK Instruments), bound cells were counted with an automated plate reader (IDEXX). Raw numbers are thus a reflection of numbers of cells bound.

-77- Example 4. Inhibition of Lymphocvte Recruitment to Colon A. DSS-Induction of colitis in mice BALB/c mice were given access to a 5% solution of dextran sodium sulfate (DSS) in their drinking water for a period of 10 days, as previously described (Lab. Invest.

69:238-249, 1993). During this time period, the mice developed clinical symptoms of colitis including softening of stools and bloody diarrhea. Multifocal epithelial injury and ulceration, similar to ulcerative colitis in humans, was evident on histologic examination of colonic mucosa from affected mice. Moreover, affected mice lost 20-30% of their initial body weight by day fe Antibody blockade of B7 and MAdCAM interactions s To determine the efficacy of 7-specific antibodies in 15 blocking the recruitment of lymphocytes to the colon, BALB/c mice were given daily intraperitoneal injections of 100 gg of monoclonal antibodies against 07, consisting of either FIB21 or FIB30 in saline, as previously characterized and described (Berlin, et al., 20 Cell 74:185-195, 1993; Michie, et al., Am. J. Pathol.

143:1688-1698, 1993; Hamann, et al., J. Immunol.

152:3282-3293, 1994) or an isotype-matched control rat monoclonal antibody at the same dose (Andrew et al., supra) over the 10 day course of DSS treatment.

Methods of evaluation Two methods were used to evaluate efficacy of the antibody therapy to inhibit leukocyte infiltration and mucosal injury in the colitic mouse. In the first method, treatment was judged histologically by two blinded observers using a scoring system for the evaluation of epithelial injury and degree of leukocyte cellular infiltration (Table For this assessment, colon tissue -78was first fixed in 10% neutral buffered forinalin, dehydrated, embedded in paraffin, sectioned, and the sections were stained with hematoxylin and eosin prior to examination.

I.

06 0*0* 0 0 00 -79- TABLE 2: PATHOLOGY

EVALUATION

Grade Definition Normal Mild Moderate

INFLAMMATION

Absence of clusters of polymorphonoclear leukocytes (PMNs) or mononuclear cells in the lamina propria; absence of intraepithelial PMNs Focal aggregates of PMNs and/or mononuclear cells in the lamina propria (equivocal or slight) or presence of isolated intraepithelial PMNs in 3 or fewer crypts per cross-section Focal aggregates of PMNs and/or mononuclear cells in the lamina propria (multi-focal or diffuse or intraepithelial PMNs in more than. 3 crypts per crosssection Diffuse infiltration of PMNs or mononuclear cells in the lamina propria (diffuse >5X) or crypt abscesses STRUCTURAL OR EPITHELIAL

ALTERATIONS

Tight crypts, no erosion, columnar epithelial cells Epithelial immaturity; equivocal irregularity of epithelial surface At least two foci of crypt branching or loss of crypts loss of surface epithelium Severe Normal Mild Moderate Severe Diffuse or multifocal branching or loss of crypts fibrosis; complete loss of epithelium (focal) 1 Additional histologic assessment was performed using immunohistochemistry for the detection and semiquantification of lymphocytes expressing 07 integrins and mucosal venules expressing MAdCAM. As previously described (Ringler, et al., Am. J. Pathol., 134: 373- 383 (1989)), colon tissue was first snap-frozen in OCT compound, sectioned while frozen, and the sections were subsequently fixed in acetone for 10 min at 4 0 C. After washing in phosphate buffered saline (PBS), nonspecific antibody binding sites were blocked with 10% normal rabbit serum diluted in PBS for 10 min, followed in sequence with washes by FIB21 antibody at 20 mg/ml in PBS for 30 min at room temperature biotinylated rabbit anti-rat polyclonal antibody, avidin-peroxidase complexes, and 15 finally the chromogen, diaminobenzidine and hydrogen peroxide diluted in Tris buffer.

In the second method, recruitment of lymphocytes to the colon was quantitatively assessed using radiolabeled mesenteric lymph node lymphocytes from syngeneic donor 20 mice. The experimental design of the animal experiments was similar to that described above except that BALB/c mice were placed on 5% DSS for 9 days (instead of 10) and on day 8, mice were given i.p. injections of 100 gg of FIB21 (anti-?7), MECA-367 (anti-MAdCAM), a mixture of both, or an isotype-matched control monoclonal antibody in saline. On day 9, mesenteric lymph node cells were isolated from donor syngeneic BALB/c mice, labeled with 51 Cr, and 5.0 x 106 cells/mouse were incubated for 30 minutes at 37 0 C with 500 pg control antibody, 250 pg of MECA-367, 500 mg FIB21, or both (total amount is 750 pg) in saline. The labeled cells and antibody were then injected intravenously into the DSS-treated recipient mice. Full-length colons were harvested from all experimental animals 1 hour after -81injection, and 7-irradiation was measured using a y-counter.

Data analysis Differences between mean scores obtained for each group of animals were assessed for statistical significance using a paired Student's t-test. Differences between means were considered significant when P 0.05.

Results Histologically, inflammation and epithelial injury to the mucosa were most severe in the descending colon, rectum and cecum. Analysis of frozen tissue sections of colon by immunohistochemistry revealed that the most significant recruitment of 87 lymphocytes was to the right colon. In addition, the level of expression of the mucosal vascular addressin, MAdCAM-1, was found to be expressed only at low levels in vessels in the intestinal mucosa early in DSS treatment (3 days), but increased dramatically after 9 days of DSS treatment, supporting the conclusion that 07 and MAdCAM-1 interactions are relevant to the inflammatory process in the colonic mucosa during DSS-induced colitis.

Histologic evaluation of mice exposed to a ~course of DSS and daily therapy using f7-specific antibodies demonstrated that substantial reductions of leukocyte recruitment (P<0.01 for FIB30 and P<0.001 for FIB21) and epithelial injury (P<O.05) occurred in right (ascending) colon compared to animals receiving a control antibody at the same dose (Figures 7A and 7B).

Furthermore, analysis using immunohistochemistry of frozen sections from these animals suggested that the number of $7 cells recruited to the right colon, but not other sections of colon, during DSS treatment was reduced.

-82- Lymphocyte recruitment to inflamed colon was then quantitatively assessed using radiolabeled mesenteric lymphocytes taken from syngeneic donors. One hour after injection of these cells in DSS-treated recipients, there was a trend towards a reduction in the number of 51 Crlabeled cells recruited to colon in mice that were treated with either p7-specific antibodies or the MAdCAM-specific antibodies, but not in mice treated with the isotopematched control antibodies (Figure 8).

B. Induction of colitis in scid mice and inhibition of recruitment of lymoh node cells to colon Scid mice reconstituted with CD45RBhi CD4 T cells develop colitis and a severe wasting syndrome. The colitis that develops in scid mice reconstituted with CD45RB CD4 15 T cells differs from most other murine models of IBD in that the induced colitis in the scid mouse clearly requires the presence of CD4 T cells for the induction, if not the pathogenesis, of the disease (Powrie, Immunity, 3:171 (1995), the teachings of which are incorporated herein in 20 by reference their entirety)).

A modification of the method of Morrissey et al. and Powrie et al. (Morrissey et al., J. Exp. Med., 178:237 (1993); Powrie et al., Int. Imm., 5:1461 (1993), the teachings of which are both incorporated herein by reference in their entirety) was used to enrich for CD4+ T cells, isolated from BALB/c spleen, by depletion of granulytic leukocytes, CD8 T cells, B220+ cells, I-A cells and MAC-1 macrophages. CD45RB hi cells were selected by cell sorter, gating on the brightest 40-45% of CD4 cells stained with anti-CD45RB. Recipient scid mice were reconstituted by intravenous injection of 1 X 106 or CD45RB 1 O T cells into the tail vein. Four mice -83were reconstituted with CD45RBhi T cells and four mice with T cells.

Reconstituted mice were monitored weekly for changes in weight and the development of fecal occult blood.

Typically, within 4-6 weeks post reconstitution, the difference in body weight of mice reconstituted with T cells relative to control scid mice reconstituted with an equal number of CD45RBO T cells, became statistically significant (Figure 18).

10 Colitis can be induced in this model with as few and X 104 cells. Generally, from 1 5 X 105 is used.

Although the kinetics of onset of disease is not uniform among mice in a given reconstitution, the colitis is of similar severity once the body weight has decreased to 85% of initial weight. Histological observations were consistent with the reports of others, and indicate that the colitis in this model is characterized by massive infiltration of CD4 T cells in the mucosa and sub-mucosa, epithelial immaturity, ulceration, crypt hyperplasia, loss 20 of goblet cells and crypt absesses. Similar to Crohn's SS.° Disease, the IBD in the scid model is also characterized by transmural infiltration with deep fistulas. Unlike the other murine models of colitis, the severity of the disease is not limited to the distal colon but is of equal severity in the transverse and proximal colon.

Antibody blockade of 7 and MAdCAM interactions Anti-murine MAdCAM-1 antibody (MAb MECA-367; American Type Culture Collection (Rockville, MD), Accession No.

HB 9478; Streeter, et al., Nature, 331:41 (1988); see also, U.S. Patent No. 5,403,919 to Butcher) and anti-murine f7 antibody (MAb FIB 504; Andrew, et al., J.

Immunol., 153: 3847 (1994)) were used in these studies.

-84scid mice were reconstituted with 2 X 105 CD45RB" or 0 CD4+ T cells. Five months post-reconstitution, mice were injected for 14 days with 200 ig/day of Rat IgG2a control antibody or a mixture of 100 Ag/day FIB-504 (murine R7-specific) 100 pg/day MECA-367 (murine MAdCAMspecific). Antibody was in PBS. There were five mice in each treatment group. After 14 days, mice were injected intravenously with 5 X 106 mesenteric lymph node cells (BALB/c) labeled with 11 In-oxine. 24 hours after adoptive transfer of labeled cells, tissues were harvested and assessed for radioactivity. Background levels of radioactivity in tissues, contributed by non-specific trapping of cells, were assessed by injection of 5 X 106 labeled cells fixed with 2% PBS-buffered formaldehyde.

15 Results were expressed as counts per minute (CPM) in colon normalized to CPM in spleen and corrected for background.

This quantitative assessment of infiltration to the colon in scid mice reconstituted with CD45RB hi CD4+ T cells S: 20 revealed an increase in localization of 10- to 100-fold as compared with the level observed in scid recipients reconstituted with an equal number of CD45RB

I

CD4+ T cells.

This increased accumulation of labeled cells in the colon was inhibited 50-75% by treatment for 2 weeks with a 25 combination of anti-37 and anti-MAdCAM monoclonal antibodies (Figure 19).

In another experiment, scid mice were reconstituted with 5 X 104 CD45RBhi or CD45RBO CD4 T cells. At the time of reconstitution, mice were treated with either 500 pg of FIB 504 (f7-specific) (6 mice); 500 Mg MECA-367 (MAdCAM-specific) (3 mice); 1 mg isotype-matched control antibody (7 mice); or 1 mg FIB 504 MECA-367 (500 Mg each) (5 mice). Following reconstitution, antibodies were administered at weekly intervals: (a) 250 Ag FIB 504; 250 ig MECA-367; 500 fg isotypematched control; or 500 gg FIB 504 MECA-367 (250 ag each).

After 4 months of treatment, mice were injected with 5 X 106 '"In-labeled mesenteric lymph node cells (BALB/c), and recruitment to the colon was assessed by measuring levels of radioactivity. Results were calculated as described for Figure 19. Treatment of scid mice for 4 months, starting from the time of reconstitution, with FIB 504 and MECA-367, alone or in combination, inhibited the increased recruitment of lymphocytes to the colon by 100% (Figure ~scid mice were reconstituted with 2.0 X 10 5 to X 105 CD45RB hi or CD45RB 0 CD4 T cells. After 4 months, 15 the mice were treated for 14 days with a combination of FIB 504 (07-specific) MECA-367 (MAdCAM-specific) (100 Mg '"of each MAb per day for a total of 200 Mg/day) or an isotype-matched control antibody (200 fg/day). Antibody was in PBS. Each experimental group consisted of 4 mice.

20 Frozen sections of left and right colon were stained with a rat monoclonal antibody specific for mouse CD4 and developed with either Fast Red or AEC (3-amino-9-ethyl- ~carbazole) chromogen. One cross-section of left colon and right colon from each mouse was analyzed for positive 25 staining for CD4 using a Leica Quantimet 500 Image analyzer. Each section was surveyed in its entirety using a 10X objective. Significance was determined using a students t-test. Data represents the mean positive count/tissue area 1 standard deviation.

Histological assessment by immunohistochemistry with a panel of antibodies specific for markers of cell lineage and state of differentiation, suggested that virtually all infiltrating cells in the colons of scid mice reconstituted with CD45RBhi T cells were CD4 T cells. No CD8+ T cells -86or B220 B cells could be identified under the conditions used. Further, treatment of these mice with a combination of and MAdCAM-specific monoclonal antibodies significantly reduced the number of CD4 T cells in the ascending or descending colon relative to the controls (Figure 21). As mesenteric lymph node cells are lymphocytes, these results indicate that the interaction of a4f7 on lymphocytes with MAdCAM is important in the recruitment of lymphoyctes to sites of inflammation in the colon and that agents which block this interaction can reduce inflammation.

Example 5. Resolution of Villus Alterations in the Common Marmoset (Callithrix iacchus) with Malabsorptive Enteritis ~Description of model i 15 Common marmosets (Callithrix jacchus) are a new world nonhuman primate that, under captive conditions at the New England Regional Primate Research Center (NERPRC), develop a steroid-nonresponsive, spontaneous malabsorption syndrome characterized by weight loss, diarrhea, and small 20 intestinal mucosal changes consistent with loss of absorptive capacity. These histologic changes include small intestinal villus atrophy and fusion, and a mononuclear leukocyte infiltrate within the lamina propria similar to Celiac disease (nontropical sprue) in humans.

Retrospective analysis from the pathology archive files at NERPRC demonstrated that up to 80% of common marmosets have, to various degrees, malabsorptive enteritis at the time of postmortem examination.

Antibody therapy protocol Adult common marmosets were selected for study from the colony-at-large at NERPRC. Base-line studies on all -87animals included physical examination, complete blood count (CBC), blood chemistry profile, serum B12, c-reactive protein, and full-thickness jejunal biopsy by laparotomy.

Following recovery from abdominal surgery, the animals were treated for 14 days with 2 mg/kg/day of ACT-1 monoclonal antibody, a blocking monoclonal antibody against a conformational epitope of a407 (Schweighoffer, et al., J. Immunol. 151:717-729, 1993). Previous studies indicated that this antibody cross-reacted to Callithrix a407. All assessments that were performed prior to antibody therapy were repeated between the 10th and 14th day of antibody therapy.

~Analysis of je-unal biopsies Full-thickness jejunal biopsies from each marmoset 15 were evaluated histologically by two independent pathologists, and villus architecture was scored according to the following grading criteria: Villus atrophy 0 normal mucosal thickness and villus height 1 mild atrophy; slight shortening of villi; height approximately 75% of normal 2 moderate atrophy; villi approximately 33-50% Snormal height 3 severe atrophy; short normal) or no observable villi Villus fusion 0 normal; no fusion 1 1-2 villi in specimen fused 2 Between 1-2 and 50% of villi in specimen fused 3 >50% villi in specimen fused -88- Data analysis Differences between mean scores obtained for each group of animals were assessed for statistical significance using a paired Student's t-test. Differences between means were considered significant when P 0.05.

Results The mean scores for villus fusion and atrophy before and after antibody therapy with the ACT-1 monoclonal antibody are shown in Figures 9 and 10, respectively. As demonstrated, there was almost complete resolution of villus atrophy (P<0.01) and a trend for improvement of villus fusion after a two-week course of therapy with the ACT-1 antibody. The effect was not secondary to nonspecific effects of exposure to foreign immunoglobulin 15 since other animals treated with various monoclonal antibodies directed against epitopes other than that recognized by ACT-1 were ineffective in reducing villus fusion and atrophy scores.

Example 6. Resolution of Colitis in the Cotton Top Tamarin 20 Description of model The cotton-top tamarin (CTT) (Saguinus oedipus) is a New World nonhuman primate which develops spontaneous, and often chronic, colitis which is clinically and histologically similar to ulcerative colitis in man (Madara, et al., Gastroenterology, 88: 13-19 (1985)).

Immunotherapy. Clinical Assessment and Mucosal Biopsy An experimental protocol involving clinical assessment, colonic mucosal biopsy, and ACT-1 immunotherapy of colitic CTTs was instituted (Figure 13). ACT-1 is a murine IgGl monoclonal antibody reactive with human a437 (Schweighoffer, et al., J. Immunol., 151: 717-729 ~I -89- (1993); Lazarovits, A.I.,-et al., J. Immunol., 133: 1857-1862 (1984); and Erle, et al., J. Immunol., 153: 517-528 (1994)). ACT-1 was found to cross-react in the tamarin as assessed by immunohistologic staining with ACT-1 antibody of colitic mucosa from affected animals. These initial pilot studies demonstrated that from 40-80% of mononuclear cells within the lamina propria of colon from affected animals were a4?7+, similar to human colitic mucosa. ACT-1 was also found to cross-react with a407 from the CTT using flow cytometry on CTT peripheral blood lymphocytes (PBLs).

CTTs with chronic colitis were chosen from the colony-at-large at the New England Regional Primate Research Center, Southborough, Massachusetts based upon 15 clinical observation of diarrhea and weight loss.

To confirm the presence of colitis (as defined by a histologic inflammatory activity score of 2 or colony animals noted to have clinical emaciation and diarrhea were evaluated for colonic inflammatory activity by routine 20 histological assessment of colonic mucosal biopsy samples on multiple occasions prior to experimental assessment of antibody immunotherapy (Figure 13). Chronically colitic CTTs were screened for colitis inflammatory activity on at least two occasions by examination of mucosal specimens from the terminal descending colon, 2-3 cm from the anus, using a pediatric fiberoptic endoscope. Inflammatory activity scores were based upon the relative numbers of neutrophils within the lamina propria, crypt lumena, crypt epithelium, and surface epithelium. In particular, a histopathologic scoring system of acute and chronic inflammatory activity was used (described by Madara,

J.L.,

et al., Gastroenterology, 88: 13-19 (1985)). All biopsy samples were scored and categorized into four groups, with 0 representing normal mucosa and 3 representing the most severe and inflamed mucosa. Scores of 0 and 1 do not represent symptomatic colitis, while scores of 2 to 3 represent mild to severe colitic activity. Animals selected for study had: moderate (grade 2) or severe (grade 3) structural alterations of surface and crypt epithelium on the first biopsy specimen, suggestive of colitis of a chronic nature, and moderate (grade 2) or severe (grade 3) inflammatory activity on at least two biopsy samples taken 3-7 days apart prior to immunotherapy.

Biopsy samples satisfying these criteria were characterized by the presence of crypt branching and/or loss with polymorphonuclear leukocyte (PMNs) infiltration to either the lamina propria and/or epithelial compartment.

Thus, animals selected for study had repeated evidence of colonic inflammatory activity and clinically-relevant 15 colitis of a chronic nature with no recent evidence of S remission. Moreover, persistence of diarrhea to the first day of administration of monoclonal antibody was requisite for the animal to be included in the study. Within 5 days of confirmation of colitis, the animals began immunotherapy 20 with ACT-1 monoclonal antibody.

ACT-

1 antibody was produced by culture in a hollow fiber cell fermenter using a sterile pyrogen-free flowpath, purified by protein A affinity chromatography, and diluted in sterile 0.9% NaCl prior to use in vivo. Because CTTs are an endangered species, ACT-1 was also demonstrated to cross-react to a 4 07 on PBLs from a related species, the common marmoset (Callithrix jacchus) in order to perform a pharmacokinetic analysis of the antibody prior to administration to colitic CTTs. In this component of the study, ACT-1 was administered to two normal adult common marmosets, first as a single intravenous infusion and then as a single intramuscular injection 24 hrs later.

Intravenous administration of 2.0 mg/kg of ACT-1 in these animals yielded an estimated serum half life of approximately 50 hours, with continued absorption of

I

-91antibody from 2-24 hours after a single intramuscular injection. Using this dosing regime, peak serum concentrations of antibody were approximately 60 gg/ml, while trough concentrations were 18.0 g/ml. No adverse clinical effects were observed in marmosets given ACT-1.

In view of these observations, half of the cotton top tamarins satisfying the study requirements (n 4; collective age 31 years) were given a single intravenous bolus of ACT-1 at a dose of 2.0 mg/kg the first day, and 7 subsequent intramuscular injections of the same amount every 24 hours, for a total of 8 days of immunotherapy. The other half of the chronically colitic control animals (n 4; collective age 26 years) received antibody 86D, a murine monoclonal antibody (IgGl) to sheep 15 TCR 76 (Mackay, et al., Eur. J. Immunol., 19: 1477-1483 (1989)), which does not cross-react in CTTs (data not shown). This irrelevant, isotype-matched antibody was produced, purified, and administered under identical conditions as ACT-1.

Colonic mucosal biopsies were again obtained at the time of the first antibody infusion (Day 0) and on days and 20. The biopsies were evaluated by an independent pathologist. Additional colon biopsies were frozen for immunohistology. For histologic analyses, colonic mucosal biopsy specimens, taken 2-3 cm for the anus, were immediately snap-frozen in OCT compound, and duplicate samples taken from the adjacent area were fixed in phosphate-buffered formalin, processed by routine histological techniques, embedded in paraffin, cut at a thickness of 6.0 im, and sections stained with hematoxylin and eosin. The formalin-fixed samples were then examined histopathologically. Acetone-fixed, frozen sections were used to detect murine IgGi administered in vivo by eliminating the primary antibody in the sequence of a previously described avidin-biotin peroxidase __1__11 -92immunohistochemical technique (Ringler, et al., Clin.

Immunol. Immunopathol., 49: 349-364 (1988)).

Animal caretakers were blinded as to therapeutic regime (ACT-1 vs. isotype-matched irrelevant monoclonal antibody), and evaluated stool consistency in each animal on a daily basis by categorizing stool as diarrhea, semisolid, or solid. Scores were assigned as follows: 0, formed, solid stool; 1, liquid stool with some solid components (semi-solid); or 2, liquid stool (diarrhea).

Animals were weighed every other day, while blood was drawn at the same intervals for flow cytometry, hematology, and storage of serum or plasma for further analyses, such as antibody concentration, anti-mouse IgG titer, clinical chemistry, or acute phase proteins.

15 Computer-assisted morphometric image analysis Quantitative, computer-assisted, morphometric analysis of mucosal biopsy sections was performed using a Leica Quantimet 500 Image Analyzer. First, immunohistochemical analysis of mucosal sections was performed to delineate 20 specific leukocyte cell types using an avidin-biotin peroxidase technique, as previously described (Ringler, et al., Clin. Immunol. Immunopathol., 49: 349-364 (1988)). Paraformaldehyde-, acetone-, or formalin-fixed, frozen sections were used to identify neutrophils, f7+ 25 cells, and monocyte/macrophages respectively, by using, as primary reagents in the sequence, a sheep anti-elastase polyclonal antibody (Biodesign, Kennebunk, ME) to identify neutrophils, FIB21 monoclonal antibody (rat IgG2a) to identify the 87 chain (Andrew, et al., J.

Immunol., 153: 3847-3861 (1994)), and HAM-56 monoclonal antibody (mouse IgM) to identify macrophages (Dako Corp., Carpinteria, CA). Examination of stained tissue sections using the elastase antibody documented that this reagent only recognized polymorphonuclear cells in CTT colonic -93mucosa. Formalin-fixed, paraffin-embedded tissue sections were used to enumerate T and B cells, using a rabbit polyclonal antibody to human CD3 (Dako Corp., Carpinteria, CA) and L26 monoclonal antibody (mouse IgG2a) (Dako Corp., Carpinteria, CA), respectively, as primary reagents in the sequence. For detection of primary antibodies, speciesand isotype-specific secondary reagents were used in order to eliminate recognition of ACT-1 or irrelevant murine IgGI antibody in tissues. After immunohistochemical procedures, 10 each cell population was enumerated on 2-4 random fields/section of mucosa. Cells were selected based on the color wavelength generated from the brown diaminobenzidine reaction product, and color selection criteria were identical on all sections analyzed for each cell-specific marker. Because of frozen section morphology and the high relative density of lymphocytes and macrophages, :quantification of these cell types was evaluated as the immunoreactive fractional area of mucosa, while all other leukocyte cell types were enumerated as the cell 20 number/mucosal area. The values were expressed as the mean 1 SEM) percent of the pretreatment (day 0) value within a treatment group, obtained by comparing the value from each animal's biopsy sample at a particular timepoint to the value obtained from the same animal at day 0. Thus, values less than 100% (shown in bold in Table 3 below) represent a decrease of leukocyte cell density compared to the pretreatment samples, while values greater than 100% represent an increase of mucosal leukocyte cell density.

Significance was determined using a paired Student's t-test and comparing mean raw scores of cell density at a particular time point to those at pretreatment.

Differences between means were considered significant when P 0.05.

-94- Hematology and Flow Cytometry Lymphocytes expressing a4f7 integrin were enumerated by flow cytometry and the ACT-1 monoclonal antibody using methods previously described (Mackay, et al., Eur. J.

Immunol., 19: 1477-1483 (1989)). Because saturating serum concentrations of ACT-I were achieved with the infusion protocol, exogenously-administered ACT-1 in the serum could be used to enumerate the number of a4a7+ lymphocytes in the blood. Briefly, whole blood from each animal at each blood collection was analyzed by diluting 100 gl of EDTAanticoagulated blood with PBS/10% rabbit serum/5% human AB serum for 20 min at 4 0 C. After removal of the blocking serum, in the case of animals treated with ACT-1, blood cells were directly incubated with either 100 gl of 15 fluorescein-conjugated, rabbit anti-mouse IgG (Dako *Corporation, Carpinteria, CA), or in the case of the animals given irrelevant antibody or pretreatment blood samples, ACT-1 was added to blood at 10 jg/ml followed then by the secondary antibody. For each sample, a minimum of 20 10,000 cells was analyzed. Routine blood cell counts and differential analyses were performed using a Baker 5000 hematology analyzer and appropriate gating for cotton-top 0 tamarin red cells, white cells, and platelets. From the hematology analysis and flow cytometry results, absolute 25 numbers of a407+ lymphocytes per gl of blood were calculated.

Results/Progress Serum concentrations Serum concentrations of ACT-1 and an irrelevant isotype-matched antibody were generally both 2 10 jg/ml for the first 10 days of the study. On days 2-10 of the study, biotinylated ACT-1, used in whole blood at a concentration of 10 ig/ml, failed to significantly label peripheral lymphocytes as assessed by flow cytometry in animals

I

M

treated with ACT-1, while on day 0 and in animals treated with an irrelevant antibody, the same antibody recognized between 70-90% of the peripheral lymphocyte pool, similar to the staining profile of ACT-1 on human lymphocytes.

Collectively, these results suggested that the therapeutic protocol for ACT-1 in colitic CTTs resulted in saturation of the a4l7 integrin on lymphocytes in the peripheral circulation.

The ability of ACT-1 to recognize extravascular a4f7 10 cells within the lamina propria of colonic mucosa of colitic CTTs was also assessed. Immunohistochemical techniques were used to detect murine IgG1 in colonic mucosal biopsies from the study animals, and ACT-1 was observed on cell membranes of mononuclear cells within the 15 lamina propria on all biopsy time points for the first days of the study in animals treated with ACT-1, but not, as expected, from Day 0 prior to antibody infusion. No labeling of lamina propria cells was observed in animals given irrelevant antibody. Therefore, the dosing regime 20 utilized in the study resulted in neutralizing serum S" concentrations of ACT-1, and concomitant extravascular recognition and labeling of immune cells within colitic mucosa. ACT-1 antibody localized to the target site, namely lymphocytes within the peripheral blood and specifically to the extravascular compartment within colitic mucosa.

Clinical effect All four test animals maintained either a grade 2 or 3 colitic inflammatory activity in both the pre-treatment and Day 0 biopsy samples, which for 3 animals was separated by days. In addition, changes within the mucosal architecture of all four animals demonstrated that these four animals had colitis of a long-lasting nature.

-96- Therefore, all animals appeared to have a chronic disease course.

Histopathologic analysis of colonic mucosal biopsies was performed. The results from representative CTTs (animals Sgo 326-84 and Sgo 17-85) before and 5 days after immunotherapy with ACT-1 illustrated the therapeutic effect of ACT-1 immunotherapy on microscopic changes to the colonic mucosa in colitic CTTs. Prior to immunotherapy, there was a purulent exudate within the epithelial compartment and crypt lumen, epithelial immaturity characterized by loss of fully-differentiated goblet cells, and the lamina propria was expanded by a mononuclear and purulent inflammatory infiltrate (Sgo 326-84, muscularis mucosae). After ACT-1 immunotherapy, ACT-1 was localized 15 to membranes of mononuclear cells within the lamina propria using immunohistochemical techniques (Sgo 17-85, muscularis mucosae), and the neutrophilic component of the inflammatory infiltrate had resolved, fully-differentiated goblet cells were observed, and the lamina propria were no 20 longer expanded by mononuclear cells and/or neutrophils (Sgo 326-84).

The clinical effect of ACT-1 on stool consistency in colitic CTTs was striking (Figure 14). An improvement in diarrhea to at least a semi-solid stool consistency was observed in all animals within 24 hours after the first dose, while complete resolution to solid stool occurred in all animals by 72 hours. Control animals did not improve and were observed to have diarrhea for the entire study period, showing, in addition, that the preselection criteria for this group of animals effectively eliminated those with spontaneous remissions.

All animals maintained solid stools for approximately 1 week after termination of antibody injections (Figure 11). With respect to individual animals, one animal (Sgo 63-93) had solid stool from Day 4 until the end of the -97protocol at Day 20 (Figure 11). Two animals (Sgo 129-91 and Sgo 17-85) had slight relapses to semi-solid stools after Day 14 in the study (Figure 11). The fourth animal (Sgo 326-84) showed a persistent improvement/ resolution of diarrhea from Day 6 to Day Similarly, leukocyte infiltrates in the colon were markedly attenuated in CTTs given ACT-1. Compared to pretreatment biopsies, histologic analysis of colonic mucosa (formalin-fixed biopsy specimens of colonic mucosa) from animals treated with ACT-1 showed an improvement in inflammatory activity and associated structural alterations of the mucosa. Using a histologic scoring system of colonic inflammatory activity (Madara, et al.

Gastroenterology, 88: 13-19 (1985)), animals treated with 15 ACT- 1 had marked decreases in inflammatory activity scores at all time points compared to baseline pretreatment scores, while scores from animals given the control antibody did not change (Figure 15). There were no changes in inflammatory scores for the irrelevant treatment group 20 on Day 20 (Figure 15). Mean raw scores of inflammatory activity at all time points in the ACT-1-treated group were statistically lower than those from the same animals at Day (Days 5 and 10, P 0.05; Day 20, P 0.01).

With respect to individual animals, all four test 25 animals showed improvement in inflammatory activity during or after ACT- 1 immunotherapy. The colitis in two animals (Sgo 129-91 and Sgo 17-85) completely resolved by Day (Figure 12). Another animal (Sgo 63-93) did not show complete abrogation of colitis activity until Day (Figure 12), while mucosal biopsy scores from the fourth animal (Sgo 326-84) showed improvement during the entire study period (Figure 12; two biopsies on day 20 in Sgo 326- 84 were scored as 0 and Furthermore, animal 326-84 gained 20% of its original body weight during the study period.

-98- In order to provide a more quantitative assessment of efficacy, computer-assisted morphometric image analysis of leukocyte subsets within colonic mucosa from all study animals was performed. Table 3 illustrates the results of an analysis of the effect of ACT-1 immunotherapy on colonic inflammatory activity in chronically colitic CTTs (expressed as a percent of pretreatment values). ACT-1 immunotherapy resulted in significant reductions in the densities of mucosal leukocytes compared to baseline numbers established prior to antibody administration, while the control group generally had either similar or increased numbers of mucosal leukocytes after administration of irrelevant antibody. Ten days after the first dose of

ACT-

1 antibody, there were approximately 30% fewer mucosal 15 mononuclear leukocytes expressing f7 integrins compared to retreatment values. This decrease was not attributed to decreases in the numbers of a4/37+ lymphocytes in the peripheral circulatory pool (Figure 16), nor was it related to manipulation or nonspecific effects since the numbers of leukocytes in colonic mucosa from the control animals either increased or remained largely unchanged. Similarly, mucosal T cells were reduced by approximately 50% at day and by about 25% by day 10 in animals treated with ACT-1, while mucosal T cells in the irrelevant antibody group at the same time points did not change. Comparable reductions in mucosal B cells were seen in the ACT-1-treated group but not in the control group. Interestingly, ACT-1 immunotherapy also reduced the density of mucosal leukocytes which have either little or no expression of a4f7. Neutrophils were reduced by 40-45% at days 10 and in animals treated with ACT-1, yet reductions in PMNs were not observed in the control group. Similarly, mucosal macrophages were reduced in all post-treatment biopsy samples by 30-45% in animals given ACT-1, while macrophages in the control group did not change. Interestingly, using -99inuunohistochemistry and a cross-reactive monoclonal antibody specific for cloned human MAdCAM (lOG3, Example no change in expression was observed in colitic CTTs treated with ACT-i or control antibody.

0* too to

S

S S S S *5 S S 55 5 5*S S 5.6 5 S S S

S

S 55 5 5 TABLE 3 Quantitative analysis of leukocyte density in colonic mucosa of CTTs treated with ACT-I or control monoclonal antibodies

ACT-I

PMNs 07+ Cells T Cells B Cells

M

Day 5 83.0± 11.3 69-9± 12.0 50.4±5.91 53.1±11.8t 70.3±6.4t Irrelevant IgGi PMNs 037+ Cells T Cells B Cells

M

757±78 3.±0.0 104.0±8.0 256.1±67.7 141.8±21.6 148.3 ±49.3 118.8±27.6 105.6±9.6 357.0±108.4 80.7±9.8 Day 10 57.5 ±13.7 f 68.4±7.1t

M%

75.9±4.4§ 82.1±17.6 58.5±4.31 Day 20 (M1 61.1±11.4§ 59.8±12.6§ 96.4±10.4 61.3±13.4§ .98.5±5.0§ 339.6+94.1 113.7±28.1 72.4 ±11.3 t 209.0±51.0 174.5 ±48.0 t P 0.05;, SP 0. 02; §P 0.01; SP 0.001.

I

-101- No overt toxicity was observed in the study animals that could be attributed to ACT-1 administration. None of the study animals showed changes in clinical chemistry assessments of liver and renal function (data not shown).

The ACT-1 antibody is a nonlytic monoclonal reagent (Lazarovits, et al., J. Immunol., 133: 1857-1862 (1984)), and leukopenia was not observed in any animal during the study. Indeed, there was a trend for neutrophilia (peak numbers in peripheral blood approaching 40 x 10 3 /1i; CTT normal range is 1.4 12.0 X 103/1i) in all study animals, including the control animals, during the first week of study when daily anesthesia/manipulation was used to administer the antibodies. There was also a trend for lymphocytosis in the animals given ACT-i, with 15 absolute numbers of lymphocytes in peripheral blood approaching 18 X 103/1l (CTT normal range is 0.6 5.7 x 103/gl). Therefore, decreased recruitment of any leukocyte cell type to the colon in the ACT--treated group could not be attributed to changes in the number of leukocytes in the 20 peripheral circulatory pool.

Summary When administered to chronically colitic cotton-top tamarins, a monoclonal antibody to a4f7 integrin rapidly resolved diarrhea and colonic inflammatory activity, indicating efficacy in improving colitis. There appeared to be a good correlation between histologic inflammatory activity scores and stool consistency. The observation that stool consistency generally improved in 1-2 days in animals receiving ACT-1 antibody is noteworthy.

Furthermore, the relative density of mucosal leukocyte subsets was greatly attenuated in response to immunotherapy with anti-a4f7 antibodies. These results also demonstrate -102an efficacious therapy for an inflammatory process which may be organ- or tissue specific (mucosal-specific).

The therapeutic effect of ACT-1 in colitic CTTs may be mediated by inhibition of lymphocyte recruitment to gut.

Alternatively, or in addition, the therapeutic effect observed may reflect alterations in other cell interactions or signalling events mediated by a407 integrin. These results indicate that ACT-I antibody is an effective antagonist of a407 integrin function, and that inhibition of a4f7 integrin function can be an organ- or tissue-specific treatment modality in the clinical management of individuals with inflammatory bowel disease.

Further, the results indicate a role for a407 integrin in the pathogenesis of inflammatory bowel disease. a4,7 15 integrin provides a potentially organ-specific, therapeutic target for inflammatory bowel disease.

Equivalents Those skilled in the art will be able to recognize, or be able to ascertain, using no more than routine 20 experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the following eoc* claims.

-103- SEQUENCE LISTING GENERAL INFORMATION:

APPLICANT:

NAME: LeukoSite, Inc.

STREET: 215 First Street CITY: Cambridge STATE/PROVINCE: Massachusetts COUNTRY: U.S.A.

POSTAL CODE/ZIP: 02142 TELEPHONE: (617) 621-9350 TELEFAX: (617) 621-9349

APPLICANT/INVENTOR:

NAME: Michael J. Briskin STREET: 28 Harbell Street CITY: Lexington STATE/PROVINCE: Massachusetts COUNTRY: U.S.A.

POSTAL CODE/ZIP: 02173

APPLICANT/INVENTOR:

NAME: Douglas J. Ringler STREET: 382 Ocean Avenue, #1008 CITY: Revere STATE/PROVINCE: Massachusetts S COUNTRY: U.S.A.

POSTAL CODE/ZIP: 02151

APPLICANT/INVENTOR:

NAME: Dominic Picarella STREET: 2 North Bennet Court, #4 CITY: Boston STATE/PROVINCE: Massachusetts COUNTRY: U.S.A.

POSTAL CODE/ZIP: 02113

APPLICANT/INVENTOR:

NAME: Walter Newman STREET: 3 Durham Street, #3 CITY: Boston STATE/PROVINCE: Massachusetts COUNTRY: U.S.A.

POSTAL CODE/ZIP: 02115 (ii) TITLE OF INVENTION: Mucosal Vascular Addressins and Uses Thereof (iii) NUMBER OF SEQUENCES: 13 (iv) CORRESPONDENCE

ADDRESS:

ADDRESSEE: Hamilton, Brook, Smith Reynolds, P.C.

STREET: Two Militia Drive CITY: Lexington STATE: Massachusetts COUNTRY: U.S.A.

ZIP: 02173-4799 -104- 0 00** 0 0* .00* COMPUTER READABLE FORM: MEDIUM TYPE: Floppy disk COMPUTER: IBM PC compatible OPERATING SYSTEM: PC-DOS/MS-DOS SOFTWARE: PatentIn Release Version #1.30 (vi) CURRENT APPLICATION DATA: APPLICATION NUMBER: FILING DATE:

CLASSIFICATION:

(vii) PRIOR APPLICATION DATA: APPLICATION NUMBER: US 08/523,004 FILING DATE: 01-SEP-1995 (vii) PRIOR APPLICATION DATA: APPLICATION NUMBER: US 08/386,857 FILING DATE: 10-FEB-1995 (viii) ATTORNEY/AGENT INFORMATION: NAME: Brook, David E.

REGISTRATION NUMBER: 22,592 REFERENCE/DOCKET NUMBER: LKS94-04A2 PCT (ix) TELECOMMUNICATION INFORMATION: TELEPHONE: 617-861-6240 TELEFAX: 617-861-9540 INFORMATION FOR SEQ ID NO:1: SEQUENCE CHARACTERISTICS: LENGTH: 1624 base pairs TYPE: nucleic acid STRANDEDNESS: double TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (ix) FEATURE: NAME/KEY: CDS LOCATION: 1..1218 (xi) SEQUENCE DESCRIPTION: SEQ ID NO:1: *40 0 S. 0 000 00 00 0 ATG GAT TTC GGA CTG GCC CTC CTG CTG GCG GGG Met Asp Phe Gly Leu Ala Leu Leu Leu Ala Gly 1 5 10 CTT CTG GGG CTC CTC Leu Leu Gly Leu Leu CTC GGC CAG Leu Gly Gln CCG GTG GTG Pro Val Val CTC CAG GTG AAG CCC CTG CAG GTG GAG Leu Gln Val Lys Pro Leu Gln Val Glu CCC CCG GAG Pro Pro Glu ACC TGC CGC Thr Cys Arg GCC GTG GCC TTG Ala Val Ala Leu GGC GCC TCG CGC CAG CTC Gly Ala Ser Arg Gln Leu 40 -105- CTG GCC TGC C GAC CGC GGC GCC TCG GTC CAG TGG CGG GGC CTG CAC Leu Ala Cys Ala Asp Arg Gly Ala Ser Val Gin Trp Arg Gly Leu Asp so 55 60

ACC

Thr AGC CTG GC Ser Leu Gly 0CC GTG Ala Val 70 CAC TCC GAC ACG Gin Set Asp Thr CGC ACC GTC CTC ACC Arg 8cr Val Leu Thr GTG CGC AAC CC Val Arg Asn Ala CTC TCG C GCC CCC Leu Ser Ala Ala Gly ACC TTC CAC CAC ACC Thr Phe Gin His Thr 105 ACC CCC GTG TOC Thr Arg Val Cys GTG GCC Val Gly TCC TGC GCC 8cr Cys Gly GCC TTC CCC Ala Phe Pro 115 CCC CC Gly Arg 100 GTC CAG CTC Val Gin Leu CTT CTG TAC Leu Val Tyr 110 GTG CCT GCT Val Pro Gly GAC CAG CTG ACC GTC Asp Cmn Leu Thr Val 120 TCC CCA GCA GCC Set Pro Ala Ala 240 288 336 384 432 480 528 576 CAC CCC Asp Pro 130 GAG GTC CCC TGT Glu Val Ala Cys ACG 0CC Thr Ala 135 CAC AAA GTC ACG CCC GTG GAC CCC His Lys Val Thr Pro Val Asp Pro 140 9 0

CCOS

S. eC

C

AAC

Asn 145 C CTC TCC TTC Ala Leu Ser Phe CTG CTC GTC CCC Leu Leu Val Gly CAC GAA CTG GAG Gin Clu Leu Clu CC CAA CCC CTG Ala Gin Ala Leu CCC GAG GTG CAG Pro Glu Val Gin GAG GAG Glu Glu 170 GAG GAG GAG Glu Glu Giu CCC CAG Pro Gln CCC GAC GAG GAC GTG CTG TTC AGO Gly Asp Glu Asp Val Leu Phe Arg 180 ACA GAG CCC TGC Thr Glu Arg Trp CCC CTC CCC Arg Leu Pro 190 CCC ACG ATC Ala Thr Met e.g.

C

e5 C CCC CTC 0CC Pro Leu Gly 195 ACC CCT GTC CCC Thr Pro Val Pro CCC CTC TAC TC Ala Leu Tyr Cys AGO CTG Arg Leu 210 CCT GGC TTG GAG Pro Gly Leu Glu ACC CAC CCC CAG 8cr His Arg Gin ATC CCC CTC CTG Ile Pro Val Leu AGC CCC ACC Set Pro Thr TCC CCC Ser Pro 230 GAG CCT CCC GAC Ciu Pro Pro Asp ACC TOC CCC GAG Thr Set Pro Glu 0 00 CCC AAC ACC ACC Pro Asn Thr Thr CCC GAC ACC ACC Pro Asp Thr Thr 260 CCC GAG TCT CCC Pro Glu Set Pro ACC ACC TCC CCC Thr Thr Set Pro GAG TCT Giu Set TCC CAG GAG CCT Set Gin Giu Pro GAC ACC ACC Asp Thr Thr TCC CAG GAG CCT Set Gin Glu Pro 270 TCC CCC GAG CCT Set Pro Giu Pro 285 CCC CAC ACC ACC TCC CAC GAG Pro Asp Thr Thr Set Gin Glu 275 CCT CCC CAC ACC ACC Pro Pro Asp Thr Thr 280 -106- CCC GAC AAO Pro Asp Lys 290 ACC TCC CCG Thr Ser Pro GAG CCC GCC CCC CAG CAG GGC TCC ACA CAC Glu Pro Ala Pro Gin Gin Gly Ser Thr His 295 300 ACC CCC Thr Pro 305 CAG OCT Gin Ala CCT GCG Pro Ala AGG AGC CCA Arg ser Pro GGG CCC ACG Gly Pro Thr 325 GOT GAC CAG Giv Asn Gin TCC ACC AGO Ser Thr Arg GOA GAA GTG Gly Giu Vai ACT CGC CGC CCT GAG ATC TCC Thr Arg Arg Pro Glu Ile Ser 315 320 ATC CCA ACA GOC TCG TCC AAA Ile Pro Thr Giy Ser Ser Lys 330 335 CTG TOG ACC AGC ACT GCG GTG Leu Trp Thr Ser Ser Ala Val CTG CCC OCG Leu Pro Ala CTG GGA CTG Leu Giy Leu 355

OCT

Ala 345

CCC

Pro CTC CTG GCC Leu Leu Ala ACG TAT CAC Thr Tyr His

CTC

Leu 365

GCT

Ala 350 TOO AAA CGC Trp Lys Arg TCT CTG AGO Ser Leu Arg TGC CGG CAC CTG GCT GAG GAC GAC ACC CAC CCA CCA Cys Arg His Leu Ala Giu Asp Asp Thr His Pro Pro 370 375 380 912 960 1008 1056 1104 1152 1200 1248 1308 1368 1428 1488 1548 1608 1624 CTT CTG CCC CAG GTG TCG GCC TGG GCT GGG TTA AGO GOG ACC GGC CAG Leu Leu Pro Gin Val Ser Ala Trp Ala Gly Leu Arg Gly Thr Gly Gin 385 390 395 400 GTC GGG ATC AGC CCC TCC TGAOTGOCCA GCCTTTCCCC CTGTGAAAGC Val Oiy Ile Ser Pro Ser 405

AAAATAGCTT

CTCAAAGTCA

GCTCATCAGA

CTCCCTGAGT

ATGTCTCACG

CTGTCGTCAG

AAAAAAAAAA

OOACCCCTTC AAGTTGAGAA CTOOTCAGOG TCCCTCTGCT CACAGAGATG GATGCATOTT AACTCAAAAG AAGGCCACTG TTTGTCTCAC GGTCCCCACC TTTCTGGACG OAACCACGTA TCTCCCTAAA AATGCGTAAO ACCAAGCTGT GACCTCCTOA OGCTTTGGCA AATAAACCTC

AAAAAA

CAAACCTGCC

CTGATTGCCT

CTACCCATGA

CTTTTTACAT

OCCCTGACCA

CTAAAATGAT

TCCCATTCTA

CTTTGGAGAA

CCTGAAGCCC

ACATTGATTC

CCCTGGGCCC

AAAAAAAAAA

INFORMATION FOR SEQ ID NO:2: SEQUENCE CHARACTERISTICS: LENGTH: 406 amino acids TYPE: amino acid TOPOLOGY: linear (ii) MOLECULE TYPE: protein (xi) SEQUENCE DESCRIPTION: SEQ ID NO:2: Met Asp Phe Oly Leu Ala Leu Leu Leu Ala Gly Leu Leu Oly Leu Leu 1 5 10 Leu Gly Gin Si Pro Val Val A: Leu Ala Cys A] Thr Ser Leu G] Val Arg Asn Al Ser Cys Gly Gi Ala Phe Pro As 115 Asp Pro Glu Va 130 Asn Ala Leu Sei 145 Ala Gin Ala Let.

Gly Asp Glu As 1 18c Pro Leu Gly Thr 195 Arg Leu Pro Gly 210 His Ser Pro Thr 225 Pro Asn Thr Thr Pro Asp Thr Thr 260 Pro Asp Thr Thr 275 Pro Asp Lys Thr 290 Thr Pro Arg Ser 305 Gin Ala Gly Pro Pro Ala Gly Asp 340 ar Leu Gin V La Val Ala LE La Asp Arg a] .y Ala Val Gl 70 a Ser Leu Se y Arg Thr Ph 0 p Gin Leu Th 1 Ala Cys Th: 13~ r Phe Ser Lea 150 i Gly Pro Glt 165 Val Leu Phe Pro Val Pro Leu Glu Leu 215 Ser Pro Glu 230 Ser Pro Giu 245 Ser Gin Glu Ser Gin Giu Ser Pro Glu 295 Pro Gly Ser 310 Thr Gin Gly 325 Gin Leu Pro -107al Lys Pro Leu Gin Val Glu 25 eu Gly Ala Ser Arg Gin Leu 40 45 Ly Ala Ser Val Gin Trp Arg 5 60 .n Ser Asp Thr Gly Arg Ser 75 r Ala Ala Gly Thr Arg Val 90 e Gin His Thr Val Gin Leu 105 r Val Ser Pro Ala Ala Leu 120 125 r Ala His Lys Val Thr Pro 5 140 az Leu Val Gly Gly Gin Giu 155 Val Gin Glu dlu Glu Giu G 170 Arg Val Thr Glu Arg Trp A~ 185 1 Pro Ala Leu Tyr Cys Gin A 200 205 Ser His Arg Gin Ala Ile P 220 Pro Pro Asp Thr Thr Ser P 235 Ser Pro Asp Thr Thr Ser P~ 250 Pro Pro Asp Thr Thr Ser G 2652 Pro Pro Asp Thr Thr Ser P1 280 285 Pro Ala Pro Gin Gin Gly Se 300 Thr Arg Thr Arg Arg Pro Gi 315 Giu Val Ile Pro Thr Gly Se 330 Ala Ala Leu Trp Thr Ser Se 345 35 Pro Pro Giu Thr Cys Arg Gly Leu Asp Val Leu Thr Cys Val Gly Leu Val Tyr 110 Val Pro Gly Val Asp Pro eu GlU Gly 160 1u Pro Gin 175 ~rg Leu Pro .la Thr met ro Vai Leu ro Giu Pro 240 ro Glu Ser 255 In Giu Pro *o Glu Pro *r Thr His U Ile Ser 320 Ser Lys 335 Ala Val 0 -108- Leu Gly Leu Leu Leu Leu Ala Leu Pro Thr Tyr His Leu 355 360 365 Cys Arg His Leu Ala Glu Asp Asp Thr His Pro Pro Ala 370 375 380 Leu Leu Pro Gin Val Ser Ala Trp Ala Gly Leu Arg Gly 385 390 395 Trp Lys Arg Ser Leu Arg Thr Gly Val Gly Ile Ser Pro Ser 405 INFORMATION FOR SEQ ID NO:3: SEQUENCE CHARACTERISTICS: LENGTH: 1539 base pairs TYPE: nucleic acid STRANDEDNESS: double TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (ix) FEATURE: NAME/KEY: CDS LOCATION: 1..1146 (xi) SEQUENCE DESCRIPTION: SEQ ID NO:3: ATG GAT TTC GGA CTG GCC CTC CTG CTG GCG GGG Met Asp Phe Gly Leu Ala Leu Leu Leu Ala Gly 1 5 10 CTT CTG GGG Leu Leu Gly CTC CTC Leu Leu CTC GGC CAG TCC CTC CAG GTG AAG CCC Leu Gly Gin Ser Leu Gln Val Lys Pro 25 CTG CAG GTG GAG Leu Gin Val Glu CCC CCG GAG Pro Pro Glu ACC TGC CGC Thr Cys Arg CCG GTG GTG Pro Val Val GCC GTG GCC TTG Ala Val Ala Leu GCC TCG CGC CAG Ala Ser Arg Gin CTG GCC Leu Ala TGC GCG GAC CGC Cys Ala Asp Arg GGG GCC Gly Ala 55 TCG GTG CAG Ser Val Gin CGG GGC CTG GAC Arg Gly Leu Asp AGC CTG GGC GCG Ser Leu Gly Ala CAG TCG GAC ACG Gln Ser Asp Thr CGC AGC GTC CTC Arg Ser Val Leu GTG CGC AAC GCC Val Arg Asn Ala CTG TCG GCG GCC Leu Ser Ala Ala ACC CGC GTG TGC Thr Arg Val Cys GTG GGC Val Gly TCC TGC GGG Ser Cys Gly GGC CGC Gly Arg 100 ACC TTC CAG Thr Phe Gln ACC GTG CAG CTC Thr Val Gin Leu CTT GTG TAC Leu Val Tyr 110 -109- GCC TTC CCG GAC CAG CTG ACC GTC TCC CCA GCA GCC CTG GTG CCT GGT Ala Phe Pro Asp Gin Leu Thr Val Ser Pro Ala Ala Leu Val Pro Gly 115 120 125 GAC CCG Asp Pro 130 GAG GTG GCC TOT Glu Val Ala Cys GCC CAC AAA GTC Ala His Lys Val ACG CCC GTG GAC CCC Thr Pro Val Asp Pro 140 384 432 480 528

AAC

Asn 145 GCG CTC TCC TTC Ala Leu Ser Phe CTG CTC GTC GG Leu Leu Val Gly

GGC

Gly CAG GAA CTG GAG Gin Glu Leu Glu GCG CAA GCC CTG Ala Oln Ala Leu CCG GAG GTG CAG Pro Olu Val Gln

GAG

Glu 170 GAG GAG GAG GAG Glu Glu Glu Glu CCC CAG Pro Cln 000 GAC GAG GlY Asp Olu CCC CTG GG Pro Leu Cly 195 GAC GTG Asp Val 180 CTG TTC AGO Leu Phe Arg

GTG

Val 185 ACA GAG CCC TG Thr Glu Arg Trp CGO CTG CCG Arg Leu Pro 190 GCC ACG ATG Ala Thr Met ACC CCT GTC CCC Thr Pro Val Pro CCC GCC Pro Ala 200 CTC TAC TGC Leu Tyr Cys AGO CTO Arg Leu 210 CCT GOC TTG GAG Pro Oly Leu Glu ACC CAC COC CAC Ser His Arg Gln ATC CCC GTC CTG Ile Pro Val Leu

CAC

His 225 AGC CCG ACC TCC Ser Pro Thr Ser GAG CCT CCC Glu Pro Pro GAC ACC Asp Thr 235 ACC TCC CCG GAG Thr Ser Pro Glu 576 624 672 720 768 816 864 912 CCC GAC ACC ACC Pro Asp Thr Thr CCG GAG TCT CCC Pro Glu Ser Pro ACC ACC TOC CAG Thr Thr Ser Gin GAO CCT Glu Pro CCC GAC ACC Pro Asp Thr 0CC CCC CAG Ala Pro Gin 275 TOC CCC GAG CCT Ser Pro Glu Pro GAC AAO ACC TCC Asp Lys Thr Ser CCC GAG CCC Pro Glu Pro 270 GOC TCC ACC CAG GCC TCC ACA Gin Gly Ser Thr ACC CCC AGO ACC Thr Pro Arg Ser AGO ACT CCC CCC CCT GAG ATC TCC CAC OCT 000 CCC ACO CAG OGA OAA Arg Thr Arg Arg Pro Glu Ile Ser Gin Ala Oly Pro Thr Gin Gly Glu 290 295 300 ATC CCA ACA GOC Ile Pro Thr Gly TCC AAA CCT 0CC Ser Lys Pro Ala CAC CAC CTO CCC Asp Gin Leu Pro OCT CTC TGO ACC Ala Leu Trp Thr ACT OCO OrG CTG Ser Ala Val Leu CTG CTG CTC CTO Leu Leu Leu Leu GCC TTG Ala Leu 960 1008 1056 CCC ACC TAT CAC CTC TGG AAA CC Pro Thr Tyr His Leu Trp Lys Arg 340 TOC CGC CAC CTC OCT GAG GAC GAC Cys Arg His Leu Ala Glu Asp Asp 345 350 -110- ACC CAC CCA CCA GCT TCT CTG AGG CTT CTG CCC CAG GTG TCG Thr His Pro Pro Ala Ser Leu Arg Leu Leu Pro Gin Val Ser 355 360 365 GCT GGG TTA AGG GGG ACC GGC CAG GTC GGG ATC AGC CCC TCC Ala Gly Leu Arg Gly Thr Gly Gin Val Gly Ile Ser Pro Ser 370 375 380 TGAGTGGCCA GCCTTTCCCC CTGTGAAAGC AAAATAGCTT GGACCCCTTC CTGGTCAGGG CAAACCTGCC TCCCATTCTA CTCAAAGTCA TCCCTCTGTT C GATGCATGTT CTGATTGCCT CTTTGGAGAA GCTCATCAGA AACTCAAAAG TTTGTCTCAC CTACCCATGA CCTGAAGCCC CTCCCTGAGT GGTCCCCACC T GAACCACGTA CTTTTTACAT ACATTGATTC ATGTCTCACG TCTCCCTAAA i ACCAAGCTGT GCCCTGACCA CCCTGGGCCC CTGTCGTCAG GACCTCCTGA G AATAAACCTC CTAAAATGAA AAAAAAAAAA AAA GCC TGG Ala Trp

LAGTTGAGAA

~ACAGAGATG

LAGGCCACTG

TTCTGGACG

.ATGCGTAAG

GCTTTGGCA

1104 1146 1206 1266 1326 1386 1446 1506 1539 9e** 9* INFORMATION FOR SEQ ID NO:4: SEQUENCE CHARACTERISTICS: LENGTH: 382 amino acids TYPE: amino acid TOPOLOGY: linear (ii) MOLECULE TYPE: protein (xi) SEQUENCE DESCRIPTION: SEQ ID NO:4: 9* Met Asp Phe Gly Leu Ala Leu Leu Leu Ala 1 5 10 Leu Gly Gin Ser Leu Gin Val Lys Pro Leu 20 25 Pro Val Val Ala Val Ala Leu Gly Ala Ser 40 Leu Ala Cys Ala Asp Arg Gly Ala Ser Val 55 Thr Ser Leu Gly Ala Val Gin Ser Asp Thr 70 Val Arg Asn Ala Ser Leu Ser Ala Ala Gly 90 Gly Leu Leu Gly Leu Leu Gin Val Giu Pro Pro Giu Arg Gin Leu Thr Cys Arg Gin Trp Arg Gly Leu Asp Gly Arg Ser Val Leu Thr 75 Thr Arg Val Cys Val Gly Ser Cys Gly Gly Arg Thr Phe Gin His Thr Val Gin Leu Leu Val Tyr 100 105 110 Ala Phe Pro Asp Gin Leu Thr Val Ser Pro Ala Ala Leu Val Pro Gly 115 120 125 9*

C

Asp Pro Glu Val Ala 130 Asn Ala Leu Ser Phe 145 Ala Gln Ala Leu Gly 165 Gly Asp Glu Asp Val 180 Pro Leu Gly Thr Pro 195 Arg Leu Pro Gly Leu 210 His Ser Pro Thr Ser 225 Pro Asp Thr Thr Ser 245 Pro Asp Thr Thr Ser 260 Ala Pro Gin Gin Gly 275 Arg Thr Arg Arg Pro C 290 Val Ile Pro Thr Gly S 305 3 Ala Leu Trp Thr Ser S 325 Pro Thr Tyr His Leu T 340 Thr His Pro Pro Ala S 355 Ala Gly Leu Arg Gly TI 370 Cys Thr Ala 135 Ser Leu Leu 150 Pro Glu Val Leu Phe Arg Val Pro Pro 200 Glu Lau Ser 215 Pro Glu Pro 230 Pro Glu Ser Pro Glu Pro ~er Thr His 280 ~lu Ile Ser 295 er Ser Lys ~10 er Ala Val I rp Lys Arg C 3 er Leu Arg L 360 hr Gly Gin V 375 -111- His Lys Val Thr 140 Val Gly Gly Gin 155 Gin Glu Glu Giu 170 Val Thr Glu Arg 185 Ala Leu Tyr Cys His Arg Gin Ala 220 Pro Asp Thr Thr 235 Pro Asp Thr Thr 250 Pro Asp Lys Thr 265 Thr Pro Arg Ser Gln Ala Gly Pro 'J 300 ?ro Ala Gly Asp G 315 ~eu Gly Leu Leu L 330 ~ys Arg His Leu A '45 eu Leu Pro Gin V 3 al Gly Ile Ser P.

380 Pro Val Giu Leu Glu Glu Trp Arg 190 Gin Ala 205 Ile Pro Ser Ser Ser ?ro 'hr Iln eu la al ro Pro Gin Pro 270 Gly Gin Leu Lau Glu 350 Ser Ser Giu Giu 255 Giu Ser Gly Pro Al a 335 Assp kla Ser 240 Pro Pro Thr Giu Ala 320 Leu Asp Trp Asp Pro Glu Gly 160 Pro Gin 175 Leu Pro Thr Met Val Leu INFORMATION FOR SEQ ID fi) SEQUENCE CHARACTERISTICS: LENGTH: 1721 base pairs TYPE: nucleic acid STRANDEDNESS: double TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA -112- (ix) FEATURE: NAME/KEY: CDS LOCATION: 1038 (xi) SEQUENCE DESCRIPTION: SEQ ID NO:S: AGC ATG GAT CGG GGC CTG GCC CTC CTG CTG GCG Met Asp Arg Gly Leu Ala Leu Leu Leu Ala 1 5 10 CTC CAG CCG GGC TGC GGC CAG TCC CTC CAG GTG Leu Gin Pro Gly Cys Gly Gin Ser Leu Gin Val 25 GGG CTT CTG GGG CTC Gly Leu Leu Gly Leu AAG CCC CTG Lys Pro Leu CAG GTG Gin Val GAG CCC CCG GAG Giu Pro Pro Giu CCG GTG GTG GCC Pro Val Val Ala GCC CTG GOC GCC Ala Leu Gly Ala be&.

TCT CGC CAG Ser Arg Gin GTG CAG TGG Val Gin Trp CTC ACC TGC Leu Thr Cys so CGC CTG GAC TGC GCG GAC Arg Leu Asp Cys Ala Asp 55 COC GGG GCC Arg Gly Ala CGG GGC Arg Gly CTG GAC ACC AGC Leu Asp Thr Ser CTG GGC GCG GTG CAG TCG Leu Gly Ala Val Gin Ser 70 75 GAC GC GGC CGC Asp Ala Gly Arg 48 96 144 192 240 288 336 384 432 GTC CTC ACC GTG Val Leu Thr Val AAC GCC TCG CTG Asn Ala Ser Leu GCG GCC GGG Ala Ala Gly ACC CGT Thr Arg GTG CGG Val Arg 110 GTG TGC GTG Val Cys Val CTC CTT OTO Leu Leu Val GGC TCC Gly Ser 100 TGC GGG GCC CGC Cys Giy Gly Arg TTC CAG CAC ACC Phe Gin His Thr GCC TTC CCG GAC Ala Phe Pro Asp CTG ACC ATC TCC Leu Thr Ile Ser CCC GCA GCC Pro Ala Ala 125 AAA GTC ACG Lys Val Thr CTG GTG CCT GGT Leu Val Pro Giy 130 GAC CCC GAG Asp Pro Glu GCC TGT ACG CC Ala Cys Thr Ala CCT GTG Pro Val 145 GAC CCC AAT GCG CTC TCC TTC TCC CTG CTC CTG GGG GAC CAG Asp Pro Asn Ala Leu Ser Phe Ser Leu Leu Leu Gly Asp Gin 150 155 CTG GAG GGG CC Leu Giu Gly Ala GCT CTG GGC CCG Ala Leu Gly Pro GTC GAG GAG GAG Val Glu Giu Glu GAG GAG CCC CAC Giu Giu Pro Gin GAG GAG GAC CTG Giu Clu Asp Val TTC AGG GTC ACA Phe Arg Val Thr GAG CGC Glu Arg 190 TGG CGG CTG CCG ACC CTG GCA ACC Trp Arg Leu Pro Thr Leu Ala Thr 195 CCT GTC CTG CCC C CTC TAC TGC Pro Val Leu Pro Ala Leu Tyr Cys 200 205 -113- CAG GCC ACG Gin Ala Thr 210 ATG AGO CTG Met Arg Leu CTG CAC GGC Leu His Gly ATC CCG Ile Pro 225 TCC CCG Ser Pro 240 GGC TCC Gly Ser

GTC

Val CCT GGC Pro Gly 215 CCG ACC Pro Thr 230 OCG ACC Ala Thr AGG AGC Arg Ser TTG GAG CTC AGC Leu Olu Leu Ser CAC CGC CAG GCC His Arg Gin Ala 220 CCC GAC ACG ACC Pro Asp Thr Thr GAA CCC CGG Giu Pro Arg TCC CGG GAG Ser Arg Giu TCC CCG GAG Ser Pro Giu 250 CCG GGC TCT Pro Giy Ser ACC CCC CAG Thr Pro Gin ACA CAC Thr His

AGC

Ser 260

CAG

Gin ACC AGO ACT Thr Arg Thr TOC CC Cys Arg 270 CCA ACA Pro Thr CCT GAO ATC Pro Giu Ile OCT GGG CCC Ala Gly Pro OGA OAA OTO Gly Glu Val GOC TCG TCC AAA Gly Ser Ser Lys 290 CCT ACO GOT Pro Thr Giy CAG CTG CCC OCO Gin Leu Pro Ala CTG TOG ACC Leu Trp Thr AGC AGT GCO OTO CTO GGA CTG CTG CTC CTG OCT TTO CCC ACC TAC C Ser Ser Ala Vai Leu Gly Leu Leu Leu Leu Ala Leu Pro Thr Tyr H 305 310 315 CTC TOG AAA COT TOC COG CAC CTG OCT GAO GAC GOC 0CC CAC CCA C Leu Trp Lys Arg Cys Arg His Leu Ala Olu Asp Oly Ala His Pro P 320 325 330 3 OCT TCT CTO ACT AOC CAG CCC TTC CCC CTO TOAAOOOAAA ATAGGTTGGA Ala Ser Leu Ser Ser Gin Pro Phe Pro Leu 340 345 CCCCTTCAAG CTGAGAACTG GTCGGGGCAA ACCTGCCTCC CATTCTATTC AAAGTCY

AC

ATCG

CTCTGGTCAC AGAGAGOGAC CTCAAAAOAA GOTOATCGTT GAGTGACCCC TGACTTTCTO TCATATCTCC CTAAAATGCG CAOGACCTCC TGAGGCTTTO GCCGOCOCG GTGGCTCAAG CACGAGGTCA GGAOATCGAG AAATACAAAA ATTAOCCGOO TOAGOCAGGA GAATGGCGTG ACTOCACTCC AGCCTGGGGG

AAA

GCACATTCTO ATTOCCTCCT TTGGAAAGGC

TOTCCCOCCT

GACGOAACCA

TAAAACCAGC

GCAAATAAAC

CCTGTAATCC

ACCATCCTGG

AGCGGTGGCG

AACCCGGGAG

ACAOAGCGAG

ACCCGTGACC

ACOTACTTCT

TGTGCCCCGA

CTCCTAAAAG

CAGCACTTTO

CTAACCCGTG

GGCOCCTGTA

GCOOAOCTTG

ACTCCGTCTC

TOOAAOCCCC

TACATATATT

CCACCTTGGO

GATAGAAACT

GOAGOCCOAG

AAACCCCGTC

GTCCCAGCTA

CAGTOAGCTO

AAAAAAAAAA

TCATCAGAAA

CGCCCCOCTC

GATTCATGTG

CCCCTGCCAT

GAAACTTGTG

GTGOTGGAT

TCTACTAAAA

CTCGGGAGGC

AGATCCOGCC

AAAAAAAAAA

912 960 1008 1058 1118 1178 1238 1298 1358 1418 1478 1538 1598 1658 1718 1721 -114- INFORMATION FOR SEQ ID NO:6: SEQUENCE CHARACTERISTICS: LENGTH: 345 amino acids TYPE: amino acid TOPOLOGY: linear (ii) MOLECULE TYPE: protein (xi) SEQUENCE DESCRIPTION: SEQ ID NO:6: Met Asp Arg Gly Leu Ala Leu Leu Leu Ala Gly Leu Leu Gly Lau Leu 1 5 10 Gin Pro Gly Cys Gly Gin Ser Leu Gin Val Lys Pro Leu Gin Val Glu 25 Pro Pro Glu Pro Val Val Ala Val Ala Leu Gly Ala Ser Arg Gin Leu 40 Thr Cys Arg Leu Asp Cys Ala Asp Arg Gly Ala Thr Val Gin Trp Arg 55 Gly Leu Asp Thr Ser Leu Gly Ala Val Gin Ser Asp Ala Gly Arg Ser 70 75 Val Leu Thr Val Arg Asn Ala Ser Leu Ser Ala Ala Gly Thr Arg Val 90 9..*Cys Val Gly Ser Cys Gly Gly Arg Thr Phe Gin His Thr Val Arg Leu .9.100 105 110 Leu Val Tyr Ala Phe Pro Asp Gin Leu Thr Ile Ser Pro Ala Ala Leu 115 120 125 Val Pro Gly Asp Pro Glu Val Ala Cys Thr Ala His Lys Val Thr Pro .9.130 135 140 9Val Asp Pro Asn Ala Leu Ser Phe Ser Leu Leu Leu Gly Asp Gin Glu 145 150 155 160 *Leu Glu Gly Ala Gin Ala Leu Gly Pro Glu Val Giu Giu Giu Glu Glu 9165 170 175 Glu Pro Gin Giu Glu Giu Asp Val Leu Phe Arg Val Thr Glu Arg Trp 180 185 190 Arg Leu Pro Thr Leu Ala Thr Pro Val Leu Pro Ala Leu Tyr Cys GIn 195 200 205 Ala Thr Met Arg Leu Pro Gly Leu Glu Leu Ser His Arg Gln Ala Ile 210 215 220 Pro Val Leu His Gly Pro Thr Ser Arg Glu Pro Pro Asp Thr Thr Ser 225 230 235 240 Pro Glu Pro Arg Ala Ala Thr Ser Pro Giu Thr Thr Pro Gin Gin Gly 245 250 255 -115- Ser Thr His Ser Pro Arg Ser Pro Gly Ser Thr Arg Thr Cys Arg Pro 260 265 270 Glu Ile Ser Gin Ala Gly Pro Thr Gin Gly Glu Val Ile Pro Thr Gly 275 280 285 Ser Ser Lys Pro Thr Gly Asp Gln Leu Pro Ala Ala Leu Trp Thr Ser 290 295 300 Ser Ala Val Leu Gly Leu Leu Leu Leu Ala Leu Pro Thr Tyr His Leu 305 310 315 320 Trp Lys Arg Cys Arg His Leu Ala Glu Asp Gly Ala His Pro Pro Ala 325 330 335 Ser Leu Ser Ser Gln Pro Phe Pro Leu 340 345 INFORMATION FOR SEQ ID NO:7: SEQUENCE

CHARACTERISTICS:

LENGTH: 18 base pairs TYPE: nucleic acid 0 STRANDEDNESS: single S TOPOLOGY: unknown (xi) SEQUENCE DESCRIPTION: SEQ ID NO:7: CTCTACTGCC AGGCCACG 18 INFORMATION FOR SEQ ID NO:8: S(i) SEQUENCE CHARACTERISTICS: LENGTH: 19 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: unknown (xi) SEQUENCE DESCRIPTION: SEQ ID NO:8: AGCCTGGGAG ATCTCAGGG 19 INFORMATION FOR SEQ ID NO:9: SEQUENCE CHARACTERISTICS: LENGTH: 20 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: unknown (xi) SEQUENCE DESCRIPTION: SEQ ID NO:9: GCCACGATGA GGCTGCCTGG -116- INFORMATION FOR SEQ ID SEQUENCE CHARACTERISTICS: LENGTH: 20 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: unknown (xi) SEQUENCE DESCRIPTION: SEQ ID GTGGAGCCTG GGCTCCTGGG INFORMATION FOR SEQ ID NO:11: SEQUENCE CHARACTERISTICS: LENGTH: 32 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:11: GGAAGCTTCC ACCATGGATT TCGGACTGGC CC INFORMATION FOR SEQ ID NO:12: SEQUENCE CHARACTERISTICS: LENGTH: 26 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear 0 (xi) SEQUENCE DESCRIPTION: SEQ ID NO:12: CCGACTAGTG TCGGGCTGTG CAGGAC o• 26 INFORMATION FOR SEQ ID NO:13: SEQUENCE CHARACTERISTICS: LENGTH: 27 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO:13: GGACTAGTGG TTTGGACGAG CCTGTTG

Claims (90)

1. An isolated nucleic acid which encodes a primate MAdCAM.

2. The isolated nucleic acid of Claim 1, wherein the isolated nucleic acid is recombinant.

3. An isolated nucleic acid of Claim 1, wherein said nucleic acid hybridizes under stringent conditions with a second nucleic acid, the second nucleic acid 10 having a nucleotide sequence as shown in Figure 1 (SEQ ID NO:1), Figure 2 (SEQ ID NO:3), or Figure 3 (SEQ ID

4. The isolated nucleic acid of Claim 3, wherein said nucleic acid is essentially pure. 15 5. An isolated nucleic acid of Claim 1, wherein said nucleic acid encodes the polypeptide shown in Figure 1 (SEQ ID NO:2), the polypeptide shown in Figure 2 (SEQ (SEQ ID NO:4 the polypeptide shown in Figure 2 (SEQ ID NO:4), the polypeptide shown in Figure 3 (SEQ ID NO:6), or the corresponding mature proteins.

6. The isolated nucleic acid of Claim 5, which is a recombinant nucleic acid.

7. The isolated nucleic acid of Claim 5, wherein said nucleic acid is essentially pure. -118-

8. The isolated nucleic acid of Claim 5 having a nucleotide sequence selected from the group consisting of a nucleotide sequence as shown Figure 1 (SEQ ID NO:1), a nucleotide sequence as shown Figure 2 (SEQ ID NO:3), a nucleotide sequence as shown Figure 3 (SEQ ID and a portion of any of the foregoing comprising the coding sequence.

9. A recombinant nucleic acid construct comprising a nucleic acid of Claim 1.

10. The recombinant nucleic acid construct of Claim 9, wherein the recombinant nucleic acid is operably linked to an expression control sequence.

11. The recombinant nucleic acid construct of Claim 9 Scomprising a nucleic acid, wherein said nucleic acid 15 encodes a polypeptide having an amino acid sequence as set forth in Figure 1 (SEQ ID NO:2), Figure 2 (SEQ ID NO:4), Figure 3 (SEQ ID NO:6).

12. The recombinant construct of Claim 11, wherein the nucleic acid is operably linked to an expression 20 control sequence.

13. An isolated primate MAdCAM.

14. The isolated primate MAdCAM of Claim 13, wherein the primate MAdCAM is a human MAdCAM, encoded by a nucleic acid which hybridizes under stringent conditions to a second nucleic acid, the second nucleic acid having a nucleotide sequence as shown in Figure 1 (SEQ ID NO:1) or Figure 2 (SEQ ID NO:3). -119- The isolated primate MAdCAM of Claim 13, wherein the primate MAdCAM is a human MAdCAM as shown in Figure 1 (SEQ ID NO:2), Figure 2 (SEQ ID NO:4), or the corresponding mature protein of either of the foregoing.

16. The isolated primate MAdCAM of Claim 13, wherein the primate MAdCAM is a macaque MAdCAM, encoded by a nucleic acid which hybridizes under stringent conditions to a second nucleic acid, the second nucleic acid having a nucleotide sequence as shown in Figure 3 (SEQ ID S

17. The isolated primate MAdCAM of Claim 13, wherein the primate MAdCAM is a macque MAdCAM as shown in Figure 3 (SEQ ID NO:6) or the corresponding mature protein.

18. The isolated primate MAdCAM of Claim 15 having essentially an amino acid sequence consisting of amino acids 19-406 of Figure 1 (SEQ ID NO:2), 19-382 of Figure 2 (SEQ ID NO:4) or 22-346 of Figure 3 (SEQ ID NO:6). e

19. An isolated primate MAdCAM having one or more functions selected from the group consisting of binding to a407 integrin and mediation of cellular adhesion. The isolated primate MAdCAM of Claim 19, wherein cellular adhesion is a417 integrin-dependent.

21. The isolated primate MAdCAM of Claim 20, wherein binding is selective for a4f7. i -120-

22. A host cell containing a recombinant nucleic acid of Claim 2.

23. The host cell of Claim 22, wherein the nucleic acid is operably linked to an expression control sequence, whereby primate MAdCAM is expressed when the host cell is maintained under conditions suitable for expression.

24. A fusion protein comprising a primate MAdCAM.

25. The fusion protein of Claim 24, comprising a first moiety and a second moiety, wherein said first moiety is a primate MAdCAM and said second moiety is at least a portion of an immunoglobulin chain or variant thereof.

26. The fusion protein of Claim 25, wherein said first 15 moiety is joined at its C-terminal end to the N-terminal end of the second moiety.

27. The fusion protein of Claim 25, wherein the first moiety is selected from the group consisting of a fragment of human MAdCAM containing the entire extracellular domain and a fragment of human MAdCAM containing two N-terminal immunoglobulin domains.

28. The fusion protein of Claim 25, wherein the second moiety is at least a portion of an immunoglobulin heavy chain constant region or variant thereof.

29. The fusion protein of Claim 28, wherein the immunoglobulin heavy chain is of the IgG class. -121- The fusion protein of Claim 28, wherein the second moiety comprises hinge, CH2 and CH3 domains of an immunoglobulin heavy chain.

31. A hybrid immunoglobulin comprising a fusion protein of Claim

32. A hybrid immunoglobulin comprising a fusion protein of Claim 31, wherein said hybrid immunoglobulin is a homodimer.

33. A nucleic acid construct comprising a nucleic acid 1 0 containing a coding sequence which encodes a fusion protein of Claim 24, wherein optionally the coding sequence is operably linked to an expression control sequence.

34. A nucleic acid construct, comprising a nucleic acid 15 containing a sequence which encodes a fusion protein of Claim 25, wherein optionally the coding sequence is operably linked to an expression control sequence. A method for producing a primate MAdCAM comprising: introducing into a host cell a nucleic acid 20 construct comprising a nucleic acid which encodes a primate MAdCAM, whereby a recombinant host cell is produced having said coding sequence operably linked to at least one expression control sequence; and maintaining the host cells produced in step (a) in a suitable medium under conditions whereby the nucleic acid is expressed.

36. The method of Claim 35, further comprising the step of isolating primate MAdCAM. -122-

37. A method for producing a primate MAdCAM comprising maintaining a host cell containing a recombinant nucleic acid encoding a primate MAdCAM under conditions suitable for expression of the nucleic acid, whereby primate MAdCAM is produced.

38. The method of Claim 37 further comprising the step of isolating primate MAdCAM.

39. An antibody or functional portion thereof which binds primate MAdCAM. 10 40. The antibody of Claim 39, wherein said antibody can inhibit one or more functions of a primate MAdCAM.

41. The antibody of Claim 39, wherein said antibody can selectively inhibit a47-dependent adhesion. :42. The antibody of Claim 40, wherein the primate is a 15 human.

43. A method of detecting a selected primate MadCAM in a sample comprising: a) contacting a sample with an antibody which binds an isolated primate MAdCAM under conditions suitable for specific binding of said antibody to the selected primate MAdCAM; and c) detecting antibody-MAdCAM complexes.

44. A method of detecting or identifying a ligand of or an agent which binds a primate MAdCAM comprising combining an agent to be tested with an isolated primate MAdCAM under conditions suitable for binding of ligand thereto, and detecting or measuring the -123- formation of a complex between said agent and primate MAdCAM. A method of detecting or identifying a ligand of or an agent which binds a primate MAdCAM comprising: a) combining an agent to be tested with a host cell expressing recombinant primate MAdCAM under conditions suitable for binding of ligand thereto; and b) detecting or measuring the formation of a complex 1 0 between said agent and the primate MAdCAM.

46. A method of detecting an inhibitor of binding of primate MAdCAM to a ligand thereof comprising: a) combining an agent to be tested with a ligand of primate MAdCAM and a composition comprising 15 isolated and/or recombinant primate MAdCAM under conditions suitable for binding of ligand thereto; and b) detecting or measuring binding between primate SMAdCAM and ligand, whereby decreased binding as compared with a suitable control indicates that the agent is an inhibitor.

47. The method of Claim 46, wherein the isolated and/or recombinant primate MAdCAM is a fusion protein.

48. The method of Claim 46, wherein the composition comprising isolated and/or recombinant primate MAdCAM contains a host cell expressing recombinant primate MAdCAM. -124-

49. A method of detecting an inhibitor of cellular adhesion mediated by MAdCAM, comprising: a) combining an agent to be tested, a first cell expressing a recombinant primate MAdCAM, and a second cell bearing an a407 integrin under conditions suitable for adhesion of said first cell to said second cell; and b) detecting or measuring adhesion between said first and second cells, whereby decreased adhesion as compared with a suitable control indicates that the agent is an inhibitor.

50. The method of Claim 49 wherein the agent an antibody or antibody fragment.

51. A method of treating an individual having a disease 15 associated with leukocyte infiltration of tissues expressing the molecule MAdCAM, comprising administering to the individual an effective amount of an antibody which can inhibit the binding of leukocytes to MAdCAM. 20 52. The method of Claim 51 wherein the disease is a disease associated with leukocyte recruitment to the gastrointestinal tract or other tissues as a result of binding of leukocytes to gut-associated endothelium expressing the molecule MAdCAM, and the antibody can inhibit the binding of leukocytes to endothelial MAdCAM.

53. The method of Claim 52 wherein antibody is a monoclonal antibody or an antigen binding fragment thereof. -125-

54. The method of Claim 53 wherein the monoclonal antibody or antigen binding fragment thereof inhibits adhesion of leukocytes expressing an integrin containing the 67 chain and endothelium expressing MAdCAM.

55. The method of Claim 54 wherein the monoclonal antibody or antigen binding fragment thereof binds a407 integrin.

56. The method of Claim 55 wherein the monoclonal antibody or antigen binding fragment thereof binds 07. 10 57. The method of Claim 56 wherein the monoclonal antibody :is or antigen binding fragment thereof binds MAdCAM. S

58. The method of Claim 54 wherein the monoclonal antibody or antigen binding fragment thereof has the antigenic specificity of a monoclonal antibody selected from the 15 group consisting of FIB 21, FIB 30, FIB 504 and ACT-1.

59. The method of Claim 58 wherein the monoclonal antibody or antigen binding fragment thereof is selected from the group consisting of FIB 21, FIB 30, FIB 504 and ACT-1 or antigen binding fragments thereof. 20 60. The method of Claim 59 wherein the monoclonal antibody is ACT-1.

61. The method of Claim 54 wherein the monoclonal antibody is selected from the group consisting of a chimeric antibody and a humanized antibody.

62. The method of Claim 54 wherein the leukocytes are lymphocytes. -126-

63. The method of Claim 54 wherein the leukocytes are monocytes.

64. The method of Claim 54 wherein the disease is inflammatory bowel disease.

65. The method of Claim 64 wherein the disease is ulcerative colitis.

66. The method of Claim 64 wherein the disease is Crohn's disease.

67. The method of Claim 64 wherein the disease is Celiac 0 disease, enteropathy associated with seronegative arthropathies, microscopic or collagenous colitis, eosinophilic gastroenteritis, or pouchitis.

68. The method of Claim 64 wherein the monoclonal antibody or antigen binding fragment thereof binds a47. 15 69. The method of Claim 64 wherein the monoclonal antibody or antigen binding fragment thereof binds MAdCAM. The method of Claim 64 wherein the monoclonal antibody or antigen binding fragment thereof has the antigenic specificity of a monoclonal antibody selected from the group consisting of FIB 21, FIB30, FIB 504 and ACT-1.

71. The method of Claim 70 wherein the monoclonal antibody or antigen binding fragment thereof is selected from the group consisting of FIB 21, FIB30, FIB 504 and ACT-1 or antigen binding fragments thereof.

72. The method of Claim 71 wherein the monoclonal antibody is ACT-1. -127-

73. The method of Claim 64 wherein the monoclonal antibody is selected from the group consisting of a chimeric antibody and a humanized antibody.

74. The method of Claim 64 wherein more than one monoclonal antibody which inhibits the binding of leukocytes to endothelial MAdCAM is administered. The method of Claim 64 wherein more than one monoclonal antibody which inhibits the binding of leukocytes to endothelial ligands is administered. 1 0 76. The method of Claim 75 wherein at least one monoclonal antibody inhibits the binding of leukocytes to an endothelial ligand other than MAdCAM. fee

77. A method for treating inflammatory bowel disease in an individual comprising administering to the individual an effective amount of an antibody which binds o* ^endothelial MAdCAM or the a407 integrin.

78. The method of Claim 77 wherein antibody is a monoclonal antibody or an antigen binding fragment thereof. 20 79. The method of Claim 78 wherein the monoclonal antibody or antigen binding fragment thereof binds a4,7 integrin. The method of Claim 79 wherein the monoclonal antibody or antigen binding fragment thereof binds P7.

81. The method of Claim 78 wherein the monoclonal antibody or antigen binding fragment thereof binds MAdCAM. -128-

82. The method of Claim 78 wherein the monoclonal antibody or antigen binding fragment thereof has the antigenic specificity of a monoclonal antibody selected from the group consisting of FIB 21, FIB 30, FIB 504 and ACT-1.

83. The method of Claim 82 wherein the monoclonal antibody or antigen binding fragment thereof is selected from the group consisting of FIB 21, FIB 30, FIB 504 and ACT-1 or antigen binding fragments thereof.

84. The method of Claim 83 wherein the monoclonal antibody 10 is ACT-1.

85. The method of Claim 78 wherein the monoclonal antibody .is selected from the group consisting of a chimeric antibody and a humanized antibody.

86. The method of Claim 78 wherein the disease is 15 ulcerative colitis.

87. The method of Claim 78 wherein the disease is Crohn's disease.

88. The method of Claim 78 wherein the disease is Celiac disease, enteropathy associated with seronegative 20 arthropathies, microscopic or collagenous colitis, eosinophilic gastroenteritis, or pouchitis.

89. A method of treating a primate having a disease associated with leukocyte infiltration of tissues expressing the molecule MAdCAM-1, comprising administering an effective amount of an antibody having specificity for a primate MAdCAM. -129- The method of Claim 89, wherein the antibody can inhibit the interaction a primate MAdCAM with an a407 integrin.

91. The method of Claim 89, wherein the disease is a disease associated with leukocyte infiltration of tissues as a result of binding of leukocytes to gut- associated endothelium expressing the molecule MAdCAM.

92. A method for treating inflammatory bowel disease in a primate, comprising administering to the primate an effective amount of an antibody which binds a primate g'e MAdCAM

93. The method of Claim 92, wherein the antibody can inhibit the interaction a primate MAdCAM with an a437 integrin.

94. A method of treating a primate having a disease associated with leukocyte infiltration of tissues expressing the molecule MAdCAM-1, comprising administering an effective amount of a primate MAdCAM or a hybrid immunoglobulin comprising a primate 20 MAdCAM.

95. The method of Claim 94, wherein the primate MAdCAM or a hybrid immunoglobulin comprising a primate MAdCAM can inhibit the interaction a primate MAdCAM with an a4#7 integrin.

96. The method of Claim 94, wherein said hybrid immunoglobulin contains a fusion protein, comprising a first moiety and a second moiety, wherein said first moiety is a primate MAdCAM and said second moiety is at least a portion of an immunoglobulin chain. I 130

97. The method of claim 94, wherein the disease is a disease associated with leukocyte infiltration of tissues as a result of binding of leukocytes to gut-associated cndothelium expressing the molecule MAdCAM.

98. A method for treating inflammatory bowel disease in a primate, comprising administering to the primate an effective amount of a primate MAdCAM or a hybrid imm uLnologlobulin comprising a primate MAdCAM.

99. The method of claim 98, wherein the primate MAdCAM or a hybrid immunologlobulin comprising a primate MAdCAM can inhibit the interaction a primate MAdCAM with an ao4p7 integrin.

100. The method of claim 98, wherein said hybrid immunoglobulin contains a fusion protein, comprising a first moiety and a second moiety, wherein said first moiety is a primate MAdCAM and said second moiety is at least a portion of an immunoglobulin chain. "09

101. An isolated nucleic acid which encodes a primate MAdCAM, substantially as hereinbefore described with reference to any one of the Examples.

102. An isolated primate MAdCAM, substantially as hereinbefore described with reference to any one of the Examples.

103. A method for producing a primate MAdCAM, substantially as hereinbefore described with reference to any one of the Examples.

104. An antibody or functional portion thereof which binds primate MAdCAM, S substantially as hereinbefore described with reference to any one of the Examples.

105. A method of detecting a selected primate MAdCAM in a sample, substantially as hereinbefore described with reference to any one of the Examples.

106. A method of detecting or identifying a ligand of or an agent which binds a primate MAdCAM, substantially as hereinbefore described with reference to any one of the Examples.

107. A method of detecting an inhibitor of binding of primate MAdCAM to a ligand thereof, substantially as hereinbefore described with reference to any one of the Examples.

108. A method of detecting an inhibitor of cellular adhesion mediated by MAdCAM, substantially as hereinbefore described with reference to any one of the Examples. Dated 2 September, 1999 Leukosite Inc. I Patent Attorneys for the Applicant/Nominated Person SPRUSON FERGUSON IR \LIBFF]08399.doc gcc 1