AU2001234721A1 - Human Cyr61 - Google Patents

Description

CYR61 COMPOSITIONS AND METHODS

FIELD OF THE INVENTION

The present invention is directed to materials and methods involving extracellular matrix signaling molecules in the form of polypeptides involved in cellular responses to growth factors. More particularly, the invention is directed to Cyrβl-, Fispl2-, and CTGF-related polynucleotides, polypeptides, compositions thereof, methods of purifying these polypeptides, and methods of using these polypeptides.

BACKGROUND OF THE INVENTION The growth of mammalian cells is tightly regulated by polypeptide growth factors. In the adult animal, most cells are metabolically active but are quiescent with regard to cell division. Under certain conditions, these cells can be stimulated to reenter the cell cycle and divide. As quiescent cells reenter the active growth and division phases of the cell cycle, a number of specific genes, the immediate early genes, are rapidly activated. Reentry to the active cell cycle is by necessity tightly regulated, since a breakdown of this control can result in uncontrolled growth, frequently recognized as cancer. Controlled reentry of particular cells into the growth phase is essential for such biological processes as angiogenesis (e.g., blood vessel growth and repair), chondrogenesis (e.g., skeletal development and prosthesis integration), oncogenesis (e.g., cancer cell metastasis and tumor neovascularization), and other growth-requiring processes.

Angiogenesis, the formation of new blood vessels from the endothelial cells of preexisting blood vessels, is a complex process which involves a changing profile of endothelial cell gene expression, associated with cell migration, proliferation, and differentiation. Angiogenesis begins with localized breakdown of the basement membrane of the parent vessel. In vivo, basement membranes (primarily composed of laminin, collagen type IV, nidogen/entactin, and proteoglycan) support the endothelial cells and provide a barrier separating these cells from the underlying stroma. The basement membrane also affects a variety of biological activities including cell adhesion, migration, and growth during development and differentiation.

Following breakdown of the basement membrane, endothelial cells migrate away from the parent vessel into the interstitial extracellular matrix (ECM), at least partially due to chemoattractant gradients. The migrating endothelial cells form a capillary sprout, which elongates. This elongation is the result of migration and proliferation of cells in the sprout. Cells located in the leading capillary tip migrate toward the angiogenic stimulus, but neither synthesize DNA nor divide. Meanwhile, behind these leading tip cells, other endothelial cells undergo rapid proliferation to ensure an adequate supply of endothelial cells for formation of the new vessel. Capillary sprouts then branch at their tips, the branches anastomose or join with one another to form a lumen, the basement membrane is reconstituted, and a vascular connection i s established leading to blood flow. Alterations in at least three endothelial cell functions occur during angiogenesis: 1) modulations of interactions with the ECM, which require alterations of cell-matrix contacts and the production of matrix-degrading proteolytic enzymes; 2) an initial increase and subsequent decrease in endothelial cell migration, effecting cell translocation towards an angiogenic stimulus; and 3) a transient increase in cell proliferation, providing cells for the growing and elongating vessel, with a subsequent return to the quiescent cell state once the vessel is formed. These three functions are realized by adhesive, chemotactic, and mitogenic interactions or responses, respectively. Therefore, control of angiogenesis requires intervention in three distinct cellular activities: 1 ) cell adhesion, 2) cell migration, and 3) cell proliferation. Another biological process involving a similar complex array of cellular activities is chondrogenesis.

Chondrogenesis is the cellular process responsible for skeletal organization, including the development of bone and cartilage. Chondrogenesis, like angiogenesis, involves the controlled reentry of quiescent cells into the growth phase of the cell cycle. The growth phase transition is associated with altered cell adhesion characteristics, changed patterns of cell migration, and transiently increased cell proliferation. Chondrogenesis involves the initial development of chondrogenic capacity (i.e., the proto-differentiated state) by primitive undifferentiated mesenchyme cells. This stage involves the production of chondrocyte-specific markers without the ability to produce a typical cartilage ECM. Subsequently, the cells develop the capacity to produce a cartilage-specific ECM as they differentiate into chondrocytes. Langille, Microscop. Res. & Tech. 25:455-469 (1994). Chondrocyte migration, adhesion, and proliferation then contribute to the development of bony, and cartilaginous, skeleton. Abnormal elaboration of the programmed development of cells participating in the process of chondrogenesis results in skeletal defects presenting problems that range from cosmetic concerns to life-threatening disorders.

Like angiogenesis and chondrogenesis, oncogenesis is characterized by changes in cell adhesion, migration, and proliferation. Metastasizing cancer cells exhibit altered adhesion and migration properties. Establishment of tumorous masses requires increased cell proliferation and the elaboration of the cellular properties characteristic of angiogenesis during the neovascularization of tumors.

Abnormal progression of angiogenesis or chondrogenesis, as well as mere progression of oncogenesis, substantially impairs the quality of life for afflicted individuals and adds to modern health care costs. The features common to these complex biological processes, comprising altered cell adhesion, migration, and proliferation, suggest that agents capable of influencing all three of these cellular activities would be effective in screening for, and modulating, the aforementioned complex biological processes. Although the art is aware of agents that influence individual cellular activities, e.g., integrins and selectins (cell adhesion), chemokines (cell migration), and a variety of growth factors or cytokines (cell proliferation), until recently no agent has been identified that exerts an influence over all three cellular activities in humans.

Murine Cyr61 (CYsteine-Rich protein) is a protein expressed in actively growing and dividing cells that may influence each of these three cellular activities. RNase protection analyses have shown that the gene encoding murine Cyrόl , murine cyrόl, is transcribed in the developing mouse embryo. O'Brien et al, Cell Growth & Diff. 5:645-654 (1992). In situ hybridization analysis showed that expression of cyrόl during mouse embryogenesis is closely correlated with the differentiation of mesenchymal cells, derived from ectoderm and mesoderm, into chondrocytes. In addition, cyrόl is expressed in the vessel walls of the developing circulatory system. These observations indicate that murine cyrόl is expressed during cell proliferation and differentiation, which are characteristics of expression of genes involved in regulatory cascades that control the cell growth cycle.

Further characterization of the Cyrόl polypeptide has been hampered by an inability to purify useful quantities of the protein. Efforts to purify Cyrόl in quantity by overexpression from either eukaryotic or prokaryotic cells typically fail. Yang, University of Illinois at Chicago, Ph.D. Thesis (1993). One problem associated with attempting to obtain useful quantities of Cyrόl is the reduction in mammalian growth rates induced by overexpression of Cyrόl . Another problem with Cyrόl purification is that the cysteine-rich polypeptide, when expressed in bacterial cells using recombinant DNA techniques, is often found in insoluble protein masses. Nevertheless, Cyrόl has been characterized as a polypeptide of 349 amino acids, containing 39 cysteine residues, a hydrophobic putative N-terminal signal sequence, and potential N-linked glycosylation sites (Asn₂₈ and Asn₂₂₅). U.S. Patent No. 5,408,040 at column 3, lines 41-54, Grotendorst et al., incoφorated herein by reference (the '040 Patent). Recently, proteins related to Cyrό 1 have been characterized. For example, a human protein, Connective Tissue Growth Factor (CTGF), has been identified. (See '040 Patent). CTGF is expressed in actively growing cells such as fibroblasts and endothelial cells ('040 Patent, at column 5, lines 62-64), an expression pattern shared by Cyrό 1. In terms of function, CTGF has been described as a protein growth factor because its primary biological activity has been alleged to be its mitogenicity ('040 Patent, at column 2, lines 25-27 and 53-55). In addition, CTGF reportedly exhibits chemotactic activity. '040 Patent, at column 2, lines 56-59. In terms of structure, the polynucleotide sequence encoding CTGF, and the amino acid sequence of CTGF, have been published. '040 Patent, SEQ ID NO:7 and SEQ ID NO:8, respectively. Another apparently related protein is the mouse protein Fisp 12 (Fibroblast

Secreted Protein). Fisp 12 has been subjected to amino acid sequence analysis, revealing a primary structure that is rich in cysteines. Ryseck et al, Cell Growth & Diff. 2:225-233 (1991), incoφorated herein by reference. The protein also possesses a hydrophobic N-terminal sequence suggestive of the signal sequence characteristic of secreted proteins.

Sequence analyses involving Cyrόl, Fispl 2, CTGF, and other proteins, have contributed to the identification of a family of cysteine-rich secreted proteins. Members of the family share similar primary structures encoded by genes exhibiting similar sequences. Each of the proteins in this emerging family is further characterized by the presence of a hydrophobic N-terminal signal sequence and 38 cysteine residues in the secreted forms of the proteins. Members of the family identified to date include the aforementioned Cyrόl (human and mouse), Fisp 12 (mouse), and CTGF (the human ortholog of Fisp 12), as well as CEF10 (chicken), and Nov (avian). One of several applications for a purified protein able to affect cell adhesion, migration, and proliferation properties involves the development of stable, long term ex vivo hematopoietic stem cell cultures. Patients subjected to high-dose chemotherapy have suppressed hematopoiesis; expansion of stem cells, their maturation into various hematopoietic lineages, and mobilization of mature cells into circulating blood routinely take many weeks to complete. For such patients, and others who need hematopoietic cell transplantation, introduction into those patients of autologous stem cells that have been manipulated and expanded in culture is advantageous. Such hematopoietic stem cells (HSC) express the CD34 stem cell antigen, but do not express lineage commitment antigens. These cells can eventually give rise to all blood cell lineages (e.g., erythrocytes, lymphocytes, and myelocytes). Hematopoietic progenitor cells that can initiate and sustain long term cultures (i.e., long term culture system-initiating cells or LTC-IC) represent a primitive population of stem cells. The frequency of LTC-IC has been estimated at only 1-2 per 10⁴ cells in normal human marrow and only about 1 per 50-100 cells in a highly purified CD34⁺ subpopulation. Thus, it would be useful to have methods and systems for long term cell culture that maintain and expand primitive, pluripotent human HSC to be used for repopulation of the hematopoietic system in vivo.

Cell culture models of hematopoiesis have revealed a multitude of cytokines that appear to play a role in the hematopoietic process, including various colony stimulating factors, interleukins, stem cell factor, and the c-kit ligand. However, in ex vivo cultures, different combinations of these cytokines favor expansion of different sets of committed progenitors. For example, a factor in cord blood plasma enhanced expansion of granulocyte-erythroid-macrophage- megakaryocyte colony forming unit (CFU-GEMM) progenitors, but expansion in these cultures favored the more mature subsets of cells. Therefore, it has been difficult to establish a culture system that mimics in vivo hematopoiesis.

An HSC culture system should maintain and expand a large number of multi- or pluripotent stem cells capable of both long term repopulation and eventual lineage commitment under appropriate induction. However, in most ex vivo culture systems, the fraction of the cell population comprised of LTC-IC decreases steadily with continued culturing, often declining to 20% of their initial level after several weeks, as the culture becomes populated by more mature subsets of hematopoietic progenitor cells that are no longer pluripotent. Moreover, the proliferative capacity exhibited by individual LTC-IC may vary extensively. Thus, a need exists in the art for HSC culture systems comprising biological agents that maintain or promote the pluripotent potential of cells such as LTC-IC cells. In addition to a role in developing ex vivo HSC cultures, biological agents affecting cell adhesion, migration, and proliferation are useful in a variety of other contexts. Proteins that potentiate the activity of mitogens but have no mitogenic activity themselves may play important roles as signaling molecules in such processes as hematopoiesis. Moreover, these signaling proteins could also serve as probes in the search for additional mitogens, many of which have not been identified or characterized. Several biological factors have been shown to potentiate the mitogenic activity of other factors, without being mitogenic themselves. Some of these potentiators are associated with the cell surface and/or extracellular matrix. Included in this group are a secreted basic Fibroblast Growth Factor-binding protein (bFGF-binding protein), the basal lamina protein perlecan, and the Human Immunodeficiency Virus- 1 TAT protein, each protein being able to promote bFGF-induced cell proliferation and angiogenesis. Also included in this group of mitogen potentiators are thrombospondin, capable of activating a latent form of Transforming Growth Factor-β, and an unidentified secreted growth-potentiating factor from vascular smooth muscle cells (Nakano et al, J. Biol. Chem. 270:5702-5705 [ 1995]), the latter factor being required for efficient activation of Epidermal Growth Factor- or thrombin-induced DNA synthesis. Further, the B cell stimulatory factor- l/interleukin-4, a T cell product with no demonstrable mitogenic activity, is able to 1) enhance the proliferative response of granulocyte-macrophage progenitors to granulocyte-colony stimulating factor, 2) enhance the proliferative response of erythroid progenitors to erythropoietin, and 3) together with erythropoietin, induce colony formation by multipotent progenitor cells. Similarly, interleukin-7 enhanced stem cell factor-induced colony formation by primitive murine bone marrow progenitors, although interleukin-7 had no proliferative effect by itself. In addition, lymphocyte growth enhancing factor (LGEF) was found to enhance mitogen- stimulated human peripheral blood lymphocyte (PBL) or purified T cell proliferation in a dose-dependent fashion. LGEF alone did not stimulate PBL or T cell proliferation.

Therefore, a need continues to exist for biological agents capable of exerting a concerted and coordinated influence on one or more of the particularized functions (e.g., cell adhesion, cell migration and cell proliferation) collectively characterizing such complex biological processes as angiogenesis, chondrogenesis, and oncogenesis. In addition, a need persists in the art for agents contributing to the reproduction of these in vivo processes in an ex vivo environment, e.g., the development of HSC cultures. Further, there continues to be a need for tools to search for the remaining biological components of these complex processes, e.g., mitogen probes, the absence of which impedes efforts to advantageously modulate and thereby control such processes.

SUMMARY OF THE INVENTION

The present invention provides extracellular matrix (ECM) signaling molecule-related materials and methods. In particular, the present invention is directed to polynucleotides encoding ECM signaling molecules and fragments or analogs thereof, ECM signaling molecule-related polypeptides and fragments, analogs, and derivatives thereof, methods of producing ECM signaling molecules, methods of using ECM signaling molecules, methods of screening for modulators of an ECM signaling molecule activity, those modulators, and methods of using those modulators to treat diseases or conditions related to angiogenesis, chondrogenesis, oncogenesis, cell migration, cell adhesion and cell proliferation.

One aspect of the present invention relates to a purified and isolated polypeptide comprising an ECM signaling molecule. The polypeptides according to the invention retain at least one biological activity of an ECM signaling molecule, such as the ability to stimulate cell adhesion, cell migration, or cell proliferation; the ability to modulate angiogenesis, chondrogenesis, or oncogenesis; immunogenicity or the ability to elicit an immune response; and the ability to bind to polypeptides having specific binding sites for ECM signaling molecules, including antibodies and integrins. The polypeptides may be native or recombinant molecules. Further, the invention comprehends full-length ECM signaling molecules, and fragments thereof, such as an isolated human Cyrόl fragment comprising a sequence selected from the group consisting of residues 280-290 of SEQ ID NO:33 and residues 305-310 of SEQ ID NO:33, wherein said Cyrόl fragment retains at least one biological function of human Cyrόl . Of course, the human Cyrόl fragment may comprise the sequence set forth in SEQ ID NO:33, provided such a fragment, like all of the fragments of the invention, retains at least one biological function of human Cyrόl . More generally, the human Cyrόl fragment may comprise a sequence that is at least 95% similar to the sequence set forth in SEQ ID NO:33, wherein the human Cyrόl fragment again retains at least one biological function of human Cyrό l . Human Cyrόl fragments contemplated by the invention may be encoded by a polynucleotide comprising a sequence that is at least 95% similar (using BLAST software as described with default settings) to a polynucleotide encoding a polypeptide having the sequence set forth at SEQ ID NO:33, wherein said polypeptide retains at least one biological function of human Cyrόl . Tn addition, the polypeptides of the invention may be underivatized, or derivatized in conformity with a native or non-native derivatization pattern. The invention further extends to polypeptides having a native or naturally occurring amino acid sequence, and variants (i.e., polypeptides having different amino acid sequences), analogs (i.e., polypeptides having a non-standard amino acid or other structural variation from the conventional set of amino acids) and homologs (i.e., polypeptides sharing a common evolutionary ancestor with another polypeptide) thereof. Polypeptides that are covalently linked to other compounds, such as polyethylene glycol, or other proteins or peptides, i.e., fusion proteins, are contemplated by the invention.

Exemplary ECM signaling molecules include mammal ian Cyrό 1 , Fisp 12, and CTGF polypeptides. Beyond ECM signaling molecules, the invention includes polypeptides that specifically bind an ECM signaling molecule of the invention, such as the aforementioned antibody products. A wide variety of antibody products fall within the scope of the invention, including polyclonal and monoclonal antibodies, antibody fragments, chimeric antibodies, CDR-grafted antibodies, "humanized" antibodies, and other antibody forms known in the art. Other molecules such as peptides, carbohydrates or lipids designed to bind to an active site of the ECM molecules thereby inhibiting their activities are also contemplated by the invention. However molecules such as peptides that enhance or potentiate the activities of ECM molecule are also within the scope of the invention. The invention further extends to a pharmaceutical composition comprising a biologically effective amount of a polypeptide and a pharmaceutically acceptable adjuvant, diluent or earner, according to the invention. A "biologically effective amount" of the biomaterial is an amount that is sufficient to result in a detectable response in the biological sample when compared to a control lacking the biomaterial. Another aspect of the invention relates to a purified and isolated polynucleotide comprising a sequence that encodes a polypeptide of the invention. A polynucleotide according to the invention may be DNA or RNA, single- or double-stranded, and may be may purified and isolated from a native source, or produced using synthetic or recombinant techniques known in the art. The invention also extends to polynucleotides encoding fragments, analogs (i.e., polynucleotides having a non-standard nucleotide), homologs (i.e., polynucleotides having a common evolutionary ancestor with another polynucleotide), variants (i.e., polynucleotides differing in nucleotide sequence), and derivatives (/^'. e., polynucleotides differing in a structural manner that does not involve the primary nucleotide sequence) of ECM molecules. Vectors comprising a polynucleotide according to the invention are also contemplated. In addition, the invention comprehends host cells transformed or transfected with a polynucleotide or vector of the invention. In a related aspect, the invention contemplates a mammalian cell comprising a cyrόl mutation selected from the group consisting of an insertional inactivation of a cyrόl allele and a deletion of a portion of a cyrόl allele. The mammalian cell is preferably a human cell and the mutation is either heterozygous or homozygous. The mutation, resulting from insertional inactivation or deletion, is either in the coding region or a flanking region essential for expression such as a 5' promoter region. Cells are also found associated with non-human animals.

Other aspects of the invention relate to methods for making or using the polypeptides and/or polynucleotides of the invention. A method for making a polypeptide according to the invention comprises expressing a polynucleotide encoding a polypeptide according to the present invention in a suitable host cell and purifying the polypeptide. Other methods for making a polypeptide of the invention use techniques that are known in the art, such as the isolation and purification of native polypeptides or the use of synthetic techniques for polypeptide production. In particular, a method of purifying an ECM signaling molecule such as human Cyrόl comprises the steps of identifying a source containing human Cyrόl , exposing the source to a human Cyrόl -specific biomolecule that binds Cyrόl such as an anti-human Cyrόl antibody, and eluting the human Cyrόl from the antibody or other biomolecule, thereby purifying the human Cyrό 1.

Another aspect of the invention is a method of screening for a modulator of angiogenesis comprising the steps of: (a) contacting a first biological sample capable of undergoing angiogenesis with a biologically effective (i.e., angiogenically effective) amount of an ECM signaling molecule-related biomaterial and a suspected modulator (inhibitor or potentiator); (b) separately contacting a second biological sample with a biologically effective amount of an ECM signaling molecule-related biomaterial, thereby providing a control; (c) measuring the level of angiogenesis resulting from step (a) and from step (b); and (d) comparing the levels of angiogenesis measured in step (c), whereby a modulator of angiogenesis is identified by its ability to alter the level of angiogenesis when compared to the control of step (b). The modulator may be either a potentiator or inhibitor of angiogenesis and the ECM signaling molecule- related biomaterial includes, but is not limited to, Cyrόl , and fragments, variants, homologs, analogs, derivatives, and antibodies thereof. The invention also extends to a method of screening for a modulator of angiogenesis comprising the steps of: (a) preparing a first implant comprising Cyrόl and a second implant comprising Cyrόl and a suspected modulator of Cyrόl angiogenesis; (b) implanting the first implant in a first comea of a test animal and the second implant in a second comea of the test animal; (c) measuring the development of blood vessels in the first and second corneas; and (d) comparing the levels of blood vessel development measured in step (c), whereby a modulator of angiogenesis is identified by its ability to alter the level of blood vessel development in the first cornea when compared to the blood vessel development in the second cornea. Another aspect of the invention is a method of screening for a modulator of angiogenesis comprising the steps of: (a) contacting a first endothelial cell comprising a cyrόl allele with a suspected modulator of angiogenesis; (b) measuring the Cyrό 1 activity of the first endothelial cell; (c) measuring the Cyrόl activity of a second endothelial cell comprising a cyr61 allele; and (d) comparing the levels of Cyrόl activity measured in steps (b) and (c), thereby identifying a modulator of angiogenesis. A related aspect of the invention is drawn to a method of screening for a modulator of angiogenesis comprising the steps of: (a) contacting a first endothelial cell with a polypeptide selected from the group consisting of a Cyrό 1 , a Fisp 12, a CTGF, a NO V, an ELM- 1 (WISP- 1 ), a WISP- 3, a COP-1 (WISP-2), and fragments, analogs, and derivatives of any of the aforementioned polypeptides, which are members of the CCN family of proteins; (b) further contacting the first endothelial cell with a suspected modulator of angiogenesis; (c) contacting a second endothelial cell with the polypeptide of step (a); (d) measuring the angiogenesis of the first endothelial cell; (e) measuring the angiogenesis of the second endothelial cell; and (f) comparing the levels of angiogenesis measured in steps (d) and (e), thereby identifying a modulator of angiogenesis.

Yet another related aspect of the invention is a method of screening for modulators of angiogenesis comprising the steps of: (a) constructing a transgenic animal comprising a mutant allele of a gene encoding a polypeptide selected from the group consisting of a Cyrόl , a Fisp 12, a CTGF, a NOV, an ELM-1 (WISP-1), a WISP-3, a COP-1 (WISP-2); (b) contacting the transgenic animal with a suspected modulator of angiogenesis; (c) further contacting a wild-type animal with the polypeptide, thereby providing a control; (d) measuring the levels of angiogenesis in the transgenic animal; (e) measuring the level of angiogenesis of the wild-type animal; and (f) comparing the levels of angiogenesis measured in steps (d) and (e), thereby identifying a modulator of angiogenesis.

Another aspect of the invention relates to a method of screening for a modulator of chondrogenesis comprising the steps of: (a) contacting a first biological sample capable of undergoing chondrogenesis with a biologically effective (e.g. chondrogenically effective) amount of an ECM signaling molecule- related biomaterial and a suspected modulator; (b) separately contacting a second biological sample capable of undergoing chondrogenesis with a biologically effective amount of an ECM signaling molecule-related biomaterial, thereby providing a control; (c) measuring the level of chondrogenesis resulting from step (a) and from step (b); and (d) comparing the levels of chondrogenesis measured in step (c), whereby a modulator of chondrogenesis is identified by its ability to alter the level of chondrogenesis when compared to the control of step (b). The modulator may be either a promoter or an inhibitor of chondrogenesis; the ECM signaling molecules include those defined above and compounds such as mannose-6-phosphate, heparin, and tenascin.

The invention also relates to an in vitro method of screening for a modulator of oncogenesis comprising the steps of: (a) inducing a first tumor and a second tumor; (b) administering a biologically effective amount of an ECM signaling molecule-related biomaterial and a suspected modulator to the first tumor; (c) separately administering a biologically effective amount of an ECM signaling molecule-related biomaterial to the second tumor, thereby providing a control; (d) measuring the level of oncogenesis resulting from step (b) and from step (c); and (e) comparing the levels of oncogenesis measured in step (d), whereby a modulator of oncogenesis is identified by its ability to alter the level of oncogenesis when compared to the control of step (c). Modulators of oncogenesis contemplated by the invention include inhibitors of oncogenesis. Tumors may be induced by a variety of techniques including, but not limited to, the administration of chemicals, e.g. , carcinogens, and the implantation of cancer cells. A related aspect of the invention is a method for treating a solid tumor comprising the step of delivering a therapeutically effective amount of a Cyrόl inhibitor to an individual, thereby inhibiting the neovascularization of the tumor. Inhibitors include, but are not limited to, inhibitor peptides such as peptides having the "RGD" motif, and cytotoxins, which may be free or attached to molecules such as Cyrόl .

Yet another aspect of the invention is directed to a method of screening for a modulator of cell adhesion comprising the steps of: (a) preparing a surface compatible with cell adherence; (b) separately placing first and second biological samples capable of undergoing cell adhesion on the surface; (c) contacting a first biological sample with a suspected modulator and a biologically effective amount of an ECM signaling molecule-related biomaterial selected from the group consisting of a human Cyrόl , a human Cyrόl fragment, a human Cyrό 1 analog, and a human Cyrόl derivative; (d) separately contacting a second biological sample with a biologically effective amount of an ECM signaling molecule- related biomaterial selected from the group consisting of a human Cyrό l , a human Cyrόl fragment, a human Cyrόl analog, and a human Cyrόl derivative, thereby providing a control; (e) measuring the level of cell adhesion resulting from step (c) and from step (d); and (f) comparing the levels of cell adhesion measured in step (e), whereby a modulator of cell adhesion is identified by its ability to alter the level of cell adhesion when compared to the control of step (d).

In a related aspect, the invention provides a method of screening for a modulator of cell adhesion comprising the steps of: (a) contacting a first fibroblast cell with a suspected modulator of cell adhesion and a biologically effective amount of an ECM signaling molecule-related biomaterial selected from the group consisting of a Cyrόl , a Fispl2, a CTGF, aNOV, an ELM-1 (WISP-1), a WISP-3, a COP- 1 (WISP-2), and fragments, analogs, and derivatives of any of the aforementioned members of the CCN family of proteins; (b) separately contacting a second fibroblast cell with a biologically effective amount of an ECM signaling molecule-related biomaterial described above, thereby providing a control; (c) measuring the level of cell adhesion resulting from step (a) and from step (b); and (d) comparing the levels of cell adhesion measured in step (c), whereby a modulator of cell adhesion is identified by its ability to alter the level of cell adhesion when compared to the control of step (b). In a preferred embodiment of the method, the fibroblast cells present the α₆β, integrin. Also preferred are fibroblast cells that present a sulfated proteoglycan, such as a heparan sulfate proteoglycan or a chondroitin sulfate proteoglycan.

Yet another aspect of the invention is a method of screening for modulators of macrophage adhesion comprising the steps of: (a) contacting a first macrophage with a polypeptide of the CCN family, such as Cyrόl , and a suspected modulator; (b) further contacting a second macrophage with the polypeptide of step (a); (c) measuring the binding of the first macrophage to the polypeptide; (d) measuring the binding of the second macrophage to the polypeptide; and (e) comparing the binding measurements of steps (d) and (e), thereby identifying a modulator of macrophage adhesion. Analogous methods of the invention are used to screen for modulators of an inflammatory response.

The invention also extends to a method of screening for a modulator of cell migration comprising the steps of: (a) forming a gel matrix comprising Cyrό 1 and a suspected modulator of cell migration; (b) preparing a control gel matrix comprising Cyrόl ; (c) seeding endothelial cells capable of undergoing cell migration onto the gel matrix of step (a) and the control gel matrix of step (b); (d) incubating the endothelial cells; (e) measuring the levels of cell migration by inspecting the interior of the gel matrix and the control gel matrix for cells; (0 comparing the levels of cell migration measured in step (e), whereby a modulator of cell migration is identified by its ability to alter the level of cell migration in the gel matrix when compared to the level of cell migration in the control gel matrix. The endothelial cells include, but are not limited to, human cells, e.g., human microvascular endothelial cells. The matrix may be formed from gelling materials such as Matrigel, collagen, or fibrin, or combinations thereof.

In a related aspect, the invention comprehends a method of screening for modulators of fibroblast cell migration comprising the steps of: (a) contacting a first fibroblast cell with a suspected modulator of cell migration and a biologically effective amount of an ECM signaling molecule-related biomaterial selected from the group consisting of a Cyrόl , a Fisp 12, a CTGF, a NOV, an ELM-1 (W1SP-1), a WISP-3, a COP-1 (WISP-2), and fragments, analogs, and derivatives of any of the aforementioned members of the CCN family of proteins; (b) separately contacting a second fibroblast cell with a biologically effective amount of an ECM signaling molecule-related biomaterial described above, thereby providing a control; (c) measuring the level of cell migration resulting from step (a) and from step (b); and (d) comparing the levels of cell migration measured in step (c), whereby a modulator of cell migration is identified by its ability to alter the level of cell migration when compared to the control of step (b). Preferred embodiments of the methods of screening for modulators of cell migration involve the use of fibroblasts presenting an α₆β, integrin and/or a sulfated proteoglycan.

Another aspect of the invention is directed to an in vitro method of screening for cell migration comprising the steps of: (a) forming a first gelatinized filter and a second gelatinized filter, each filter having two sides; (b) contacting a first side of each the filter with endothelial cells, thereby adhering the cells to each the filter; (c) applying an ECM signaling molecule and a suspected modulator of cell migration to a second side of the first gelatinized filter and an ECM signaling molecule to a second side of the second gelatinized filter; (d) incubating each the filter; (e) detecting cells on the second side of each the filter; and (f) comparing the presence of cells on the second side of the first gelatinized filter with the presence of cells on the second side of the second gelatinized filter, whereby a modulator of cell migration is identified by its ability to alter the level of cell migration measured on the first gelatinized filter when compared to the cell migration measured on the second gelatinized filter. The endothelial cells are defined above. The ECM signaling molecules extend to human Cyrό 1 and each of the filters may be placed in apparatus such as a Boyden chamber, including modified Boyden chambers.

The invention also embraces an /// vivo method of screening for a modulator of cell migration comprising the steps of: (a) removing a first central portion of a first biocompatible sponge and a second central portion of a second biocompatible sponge; (b) applying an ECM signaling molecule and a suspected modulator to the first central portion and an ECM signaling molecule to the second central portion; (c) reassociating the first central portion with said first biocompatible sponge and said second central portion with the second biocompatible sponge; (d) attaching a first filter to a first side of the first biocompatible sponge and a second filter to a second side of the first biocompatible sponge; (e) attaching a third filter to a first side of the second biocompatible sponge and a fourth filter to a second side of the second biocompatible sponge; (f) implanting each of the biocompatible sponges, each biocompatible sponge comprising the central portion and the filters, in a test animal; (e) removing each the sponge following a period of incubation; (f) measuring the cells found within each of the biocompatible sponges; and (g) comparing the presence of cells in the first biocompatible sponge with the presence of cells in the second biocompatible sponge, whereby a modulator of cell migration is identified by its ability to alter the level of cell migration measured using the first biocompatible sponge when compared to the cell migration measured using the second biocompatible sponge. ECM signaling molecules include, but are not limited to, human Cyrόl ; the ECM signaling molecule may also be associated with Hydron. In addition, the in vivo method of screening for a modulator of cell migration may include the step of providing a radiolabel to the test animal and detecting the radiolabel in one or more of the sponges.

Another aspect of the invention relates to a method for modulating hemostasis comprising the step of administering an ECM signaling molecule in a pharmaceutically acceptable adjuvant, diluent or carrier. Also, the invention extends to a method of inducing wound healing in a tissue comprising the step of contacting a wounded tissue with a biologically effective amount of an ECM signaling molecule, thereby promoting wound healing. The ECM signaling molecule maybe provided in the form of an ECM signaling molecule polypeptide or an ECM signaling molecule nucleic acid, e.g., using a gene therapy technique. For example, the nucleic acid may comprise an expression control sequence operably linked to an ECM signaling molecule which is then introduced into the cells of a wounded tissue. The expression of the coding sequence is controlled, e.g., by using a tissue-specific promoter such as the K 14 promoter operative in skin tissue to effect the controlled induction of wound healing. The nucleic acid may include a vector such as a Heφesvirus, an Adenovirus, an Adeno-associated Virus, a Cytomegalovirus, a Baculovirus, a retrovirus, and a Vaccinia Virus. Suitable wounded tissues for treatment by this method include, but are not limited to, skin tissue and lung epithelium. A related method comprises administering a biologically effective amount of an ECM signaling molecule, e.g. Cyrόl , to an animal to promote organ regeneration. The impaired organ may be the result of trauma, e.g. surgery, or disease. Another method of the invention relates to improving the vascularization of grafts, e.g., skin grafts. Another method of the invention is directed to a process for promoting bone implantation, including bone grafts. The method for promoting bone implantation comprises the step of contacting a bone implant or receptive site with a biologically effective (i.e., chondrogenically effective) amount of an ECM signaling molecule. The contacting step may be effected by applying the ECM signaling molecule to a biocompatible wrap such as a biodegradable gauze and contacting the wrap with a bone implant, thereby promoting bone implantation. The bone implants comprise natural bones and fragments thereof, as well as inanimate natural and synthetic materials that are biocompatible, such as prostheses. In addition to direct application of an ECM signaling molecule to a bone, prosthesis, or receptive site, the invention contemplates the use of matrix materials for controlled release of the ECM signaling molecule, in addition to such application materials as gauzes.

Still another related aspect of the invention is a method of screening for modulators of wound healing comprising the steps of: (a) contacting a first activated platelet with a polypeptide of the CCN family, such as Cyrόl , and a suspected modulator; (b) further contacting a second activated platelet with the polypeptide of step (a); (c) measuring the binding of the first activated platelet to the polypeptide; (d) measuring the binding of the second activated platelet to the polypeptide; and (e) comparing the binding measurements of steps (d) and (e), thereby identifying a modulator of wound healing. Preferably, the wound healing involves the participation of platelet binding in the process of blood clotting. Also preferred are platelets presenting the _1Ibβ₃ integrin.

Yet another aspect of the invention relates to a method of screening for a modulator of cell proliferation comprising the steps of: (a) contacting a first biological sample capable of undergoing cell proliferation with a suspected modulator and a biologically effective (i.e., mitogenically effective) amount of an ECM signaling molecule-related biomaterial selected from the group consisting of a human Cyrόl , a human Cyrόl fragment, a human Cyrόl analog, and a human Cyrόl derivative; (b) separately contacting a second biological sample capable of undergoing cell proliferation with a biologically effective amount of an ECM signaling molecule-related biomaterial selected from the group consisting of a human Cyrόl , a human Cyrόl fragment, a human Cyrόl analog, and a human Cyrό 1 derivative, thereby providing a control ; (c) incubating the first and second biological samples; (d) measuring the level of cell proliferation resulting from step (c); and (e) comparing the levels of cell proliferation measured in step (d), whereby a modulator of cell proliferation is identified by its ability to alter the level of cell adhesion when compared to the control of step (b).

In a related aspect, the invention contemplates a method of screening for modulators of fibroblast cell proliferation comprising the steps of: (a) contacting a first fibroblast cell with a suspected modulator of cell proliferation and a biologically effective amount of an ECM signaling molecule-related biomaterial selected from the group consisting of a Cyrόl, a Fisp 12, a CTGF, a NOV, an ELM-1 (W1SP-1 ), a WISP-3, a COP-1 (WISP-2), and fragments, analogs, and derivatives of any of the aforementioned members of the CCN family of proteins; (b) separately contacting a second fibroblast cell with a biologically effective amount of an ECM signaling molecule-related biomaterial described above, thereby providing a control; (c) measuring the level of cell proliferation resulting from step (a) and from step (b); and (d) comparing the levels of cell proliferation measured in step (c), whereby a modulator of cell proliferation is identified by its ability to alter the level of cell proliferation when compared to the control of step (b). Preferred embodiments of the methods of screening for modulators of cell proliferation involve the use of fibroblasts presenting an α₆β, integrin and/or a sulfated proteoglycan.

A related aspect of the invention is a modulator of an ECM signaling molecule identified using the screening methods of the invention. In one embodiment, Cyrόl is the ECM signaling molecule whose activity is modulated, induction of angiogenesis and/or cell adhesion is the activity that is modulated, and the modulator is a peptide. A preferred peptide modulator of Cyrόl -induced angiogenesis and/or cell adhesion is a peptide having a sequence derived from domain III of Cyrόl, including peptides having the sequence H₂N-G QKCIVQTTSWSQCSKS-CO₂H and peptides having amino acid sequences of 80%, 90%), 95% and preferably, 99% sequence similarity. The peptide modulators of the invention include ECM signaling molecule fragments of varying length, variants thereof preferably exhibiting conservative amino acid substitutions, additions or deletions, and derivatives thereof, including, but not limited to, peptides that are glycosylated, PEGylated, phosphorylated, and attached or fused to other peptides or non-peptide carrier molecules.

Also comprehended by the invention is a method for expanding a population of undifferentiated hematopoietic stem cells in culture, comprising the steps of: (a) obtaining hematopoietic stem cells from a donor; and (b) culturing said cells under suitable nutrient conditions in the presence of a biologically effective (i.e., hematopoietically effective) amount of a Cyrόl polypeptide.

Another method according to the invention is a method of screening for a mitogen comprising the steps of: (a) plating cells capable of undergoing cell proliferation; (b) contacting a first portion of the cells with a solution comprising Cyrόl and a suspected mitogen; (c) contacting a second portion of the cells with a solution comprising Cyrόl, thereby providing a control; (c) incubating the cells; (d) detecting the growth of the first portion of cells and the second portion of the cells; and (e) comparing growth of the first and second portions of cells, whereby a mitogen is identified by its ability to induce greater growth in the first portion of cells when compared to the growth of the second portion of cells. The cells include, but are not limited to, endothelial cells and fibroblast cells. Further, the method may involve contacting the cells with a nucleic acid label, e.g., [ H]- thymidine, and detecting the presence of the label in the cells. Another method relates to improving tissue grafting, comprising administering to an animal a quantity of Cyrόl effective in improving the rate of neovascularization of a graft.

In other aspects, the invention is drawn to a method of screening for a modulator of binding between a Cyrό 1 polypeptide and a Cyrό 1 receptor integrin selected from the group consisting of α_vβ₃, _vβ₅, α₆β,, α„β₃ and α_Mβ₂, comprising the steps of: a) contacting a Cyrό l polypeptide with an integrin receptor composition in the presence and in the absence of a potential modulator compound; b) detecting binding between the polypeptide and the integrin receptor; and c) identifying a modulator compound in view of decreased or increased binding between the Cyrόl polypeptide and the integrin receptor in the presence of the modulator, as compared to binding in the absence of the modulator. Suitable Cyrόl polypeptides include human Cyrόl, and fragments, analogs and derivative thereof. For example, a preferred Cyrόl fragment would have at least 95% amino acid similarity to residues 1-281 of SEQ ID NO:4, thereby containing domains I-III, but not domain IV, of Cyrόl . Purified modulators identified by the above-described method of screening for a modulator of binding between a Cyrόl polypeptide and a particular integrin receptor are also embraced by the invention. Further, the invention contemplates the use of a modulator of a Cyrόl -integrin α_vβ interaction for preparation of a medicament for the treatment of a condition selected from the group consisting of atherosclerosis, heart disease, tumor growth, tumor metastasis, fibrosis, disorders associated with inadequate angiogenesis, disorders associated with aberrant granulation tissue development, aberrant fibroblast growth and wounds. In still another aspect, the invention provides a method for treating a condition characterized by defective connective tissue in a mammalian subject, comprising the steps of: (a) identifying a mammalian subject in need of treatment for the condition, and (b) administering to the mammalian subject a composition comprising a modulator of an interaction between Cyrό 1 and an integrin receptor selected from the group consisting of α_vβ₃, _vβ₅, α₆β,, α„β₃ and α_Mβ₂, in an amount effective to mitigate the symptoms of the condition in the mammalian subject. Suitable modulators include, but are not limited to, a modulator selected from the group consisting of heparin, heparan sulfate and a polypeptide that binds an integrin receptor. The modulator may be administered in a wide variety of ways, including the situation where administering is accomplished by expressing an exogenous polypeptide coding region in cells of the type affected by said condition. An exemplary polypeptide is a fragment of a Cyrόl polypeptide, for example a fragment comprising a sequence selected from the group consisting of residues 280-290 of SEQ ID NO:4 and residues 306-312 of SEQ ID NO:4. Further, a variety of conditions or disorders (e.g., diseases) and integrin receptors are contemplated, such as a condition that involves a defect in fibroblast adhesion and the integrin receptor is α₆β_t, a condition that involves a defect in fibroblast chemotaxis and the integrin receptor is α β₅, and a condition that involves a defect in fibroblast proliferation and the integrin receptor is α_vβ -

Another aspect of the invention is drawn to methods of treating conditions or disorders, such as diseases, associated with gene under- or over-expression by delivering a therapeutically effective amount of an ECM Signaling Molecule (e.g.. a Cyrόl polypeptide, Fispl2, CTGF), or a biologically active fragment thereof, or modulator of a Cyrόl -integrin receptor interaction, using delivery means known in the art. A related aspect of the invention is drawn to a method for modulating gene expression comprising the step of administering a biologically effective amount of a human Cyrόl fragment to a cell capable of Cyrόl -modulated gene expression.

In yet another aspect, the invention is drawn to a method for treating a condition characterized by a defect in smooth muscle tissue in a mammalian subject, comprising the steps of: (a) identifying a mammalian subject in need of treatment for the condition, and (b) administering to the mammalian subject a composition comprising a modulator of an interaction between Cyrόl and an α₆β, integrin receptor in an amount effective to mitigate the symptoms of the condition in the mammalian subject. An exemplary modulator is selected from the group consisting of heparin, heparan sulfate and a polypeptide that binds an ₆β, integrin receptor. One of many suitable means of administering the composition is accomplished by expressing an exogenous polypeptide coding region in cells of the type affected by said condition. Again, exemplary Cyrόl polypeptides include fragments of Cyrόl , such as a fragment that comprises a sequence selected from the group consisting of residues 280-290 of SEQ ID NO:4 and residues 306-312 of SEQ ID NO:4. Conditions that may be treated by the method include a condition selected from the group consisting of atherosclerosis, heart disease, tumor growth, tumor metastasis, fibrosis, disorders associated with inadequate angiogenesis, disorders associated with aberrant granulation tissue development, aberrant fibroblast growth and wounds.

Another aspect of the invention is directed to a kit for assaying for Cyrό 1 - integrin receptor interactions, the kit comprising a Cyrόl polypeptide and a composition comprising an integrin receptor selected from the group consisting ofα_vβ₃, α_vβ„ ₆β,, α_Mβ and α_Mβ₂. An exemplary kit contains an integrin receptor composition that comprises an α_vβ₃ integrin localized in a mammalian fibroblast cell membrane.

Numerous additional aspects and advantages of thepresent invention will be apparent upon consideration of the following drawing and detailed description. BRIEF DESCRIPTION OF THE DRAWING

FIGURE 1 presents the comparative amino acid sequences of members of the cysteine-rich protein family of growth-regulating proteins.

DETAILED DESCRIPTION OF THE INVENTION In the mouse, the Cyrό 1 protein has been found to influence cell adhesion, migration, and proliferation. The cyrόl gene, which encodes Cyrόl , is an immediate-early gene that is transcriptional ly activated by serum growth factors in mouse fibroblasts. Lau et al, EMBO J. 4:3145-3151 (1985), incoφorated herein by reference; Lau et al, Proc. Natl. Acad. Sci. (USA) 54:1 182-1 186 (1987), incoφorated herein by reference. The murine cyr67 cDNA coding sequence is set forth in SEQ ID NO:l . (The human cyrόl cDNA coding sequence is provided in SEQ ID NO:3). The amino acid sequence of murine Cyrόl is set out in SEQ ID NO:2. (The human Cyrόl amino acid sequence is presented in SEQ ID NO:4). Cyrόl is a 41 kDa polypeptide exhibiting 39 cysteine residues, approximately 10% of the 379 amino acids constituting the unprocessed protein. Yang et al, Cell Growth & Diff. 2:351-357 (1991), incoφorated herein by reference. Investigations have revealed that murine Cyrό 1 binds heparin and is secreted. Yang et al. Consistent with the observed secretion of Cyrόl is the identification of an N-terminal signal sequence in nascent Cyrόl , deduced from inspection of the murine cyrόl cDNA sequence. Yang et al. Additionally, Cyrόl is not found in the conditioned medium of cultured cells expressing cyrόl, but is found associated with the extracellular matrix (ECM) and the cell surface. Yang et al. Structurally similar cysteine-rich mammalian proteins have been characterized. Fisp 12. a cysteine-rich murine protein, exhibits structural similarity to

Cyrόl . The cDNA sequence encoding Fispl 2 is set forth in SEQ ID NO:5; the amino acid sequence of Fisp 12 is presented in SEQ ID NO:6. Murine Fispl 2, like Cyrόl, influences cell adhesion, proliferation and migration. The human ortholog of Fisp 12 is Connective Tissue Growth Factor (CTGF), a protein similar in structure and function to Cyrό 1. Fisp 12, and CTGF, are distinguishable from Cyrόl, however. For example, a greater proportion of secreted Fispl2 is found in the culture medium than is the case with Cyrόl ; a correspondingly lower proportion of Fispl 2 is localized in the area of expressing cells (cell surface and nearby extracellular matrix) than is found with Cyrόl . Additional similarities and distinctions among the proteins comprising the ECM signaling molecules of the invention will become apparent in the recitations below.

The present invention has multiple aspects, illustrated by the following examples. Example 1 describes the cloning of polynucleotides encoding members of the cysteine-rich protein family of ECM signaling molecules; Example 2 describes sequence analyses; Example 3 describes RNA analyses; Example 4 describes the production of transgenic animals; Example 5 describes the expression of Cyrόl polypeptides; Example 6 describes the expression of Fisp 12 polypeptides; Example 7 sets out methods of polypeptide purification; Example 8 provides a characterization of the polypeptides of the invention; Example 9 discloses a heparin binding assay for the polypeptide members of the cysteine-rich protein family; Example 10 is directed to receptors for the polypeptides; Example 11 describes anti-ECM signaling molecule antibodies; Example 12 is directed to inhibitory peptides; Example 13 describes cell adhesion and polypeptide-based methods for influencing the process of cell adhesion; Example 14 describes polypeptide-influenced migration of fibroblasts; Example 15 describes the migration of endothelial cells and in vitro assays for migration; Example 16 describes an in vitro assay for inhibitors of endothelial cell migration; Example 17 describes an in vivo assay for endothelial cell migration; Example 18 describes mitogen potentiation by the polypeptides of the invention; Example 19 describes an in vivo cornea assay for angiogenic factors and modulators; Example 20 is directed to methods for influencing blood clotting using the polypeptides of the invention; Example 21 discloses the use of the polypeptides for ex vivo hematopoietic stem cell cultures; Example 22 addresses organ regeneration; Example 23 describes chondrogenesis and the expression of extracellular matrix signaling molecules in mesenchyme cells; Example 24 describes the promotion of cell adhesion in the process of chondrogenesis using the polypeptides of the invention; Example 25 describes chondrogenesis and the influence of the polypeptides of the invention on cell aggregation; Example 26 describes the promotion of cell proliferation by polypeptides of the invention in the process of chondrogenesis; Example 27 addresses methods for using the polypeptides of the invention to affect chondrogenesis; Example 28 provides genetic approaches to the use of polynucleotides of the invention; Example 29 describes Fibroblast adhesion, Example 30 addresses angiogenesis, Example 31 relates to insertional inactivation or knock-out genetic constructs, Example 32 describes adhesion to platelets and macrophages, and Example 33 describes peptide modulators of Cyrόl activity. These examples are intended to be illustrative of the present invention and should not be construed to limit the scope of the invention.

Example 1

Polynucleotide Cloning

Initially, an attempt was made to isolate a human cyrόl cDNA from a human placental cDNA library by probing with the murine cyrόl cDNA sequence using techniques that are standard in the art. See Sambrook et al, incoφorated herein by reference. Isolation of the complete murine cyrόl cDNA from a BALB/c 3T3 (ATCC CRL-1658) cDNA library has been described. O'Brien et al. Moi. Cell. Biol. 70:3569-3577 (1990), incoφorated herein by reference. The nucleotide and deduced amino acid sequences of murine cyrόl are available from the Genbank database under accession number M32490. The nucleotide sequence of murine cyrόl is presented in SEQ ID NO:l ; the murine Cyrόl amino acid sequence is presented in SEQ ID NO:2.

The human cDNA library was constructed using λgtl 1 (Promega Coip., Madison, WI) as a vector which was transfected into E. coli and plated on LB agar. A murine cDNA expression construct cloned in pGEM-2 (O'Brien et al, [ 1990]), containing the entire murine cyrόl coding sequence [nucleotides 56- 1560, using the numbering of O'Brien et al, (1990); see SEQ ID NO: 1 ] was used as a probe. The mouse cDNA probe was radiolabeled by techniques standard in the art. Sambrook et al. Plaque screenings using the mouse probe were performed using standard techniques. Sambrook et al

More particularly, agar plates containing the human cDNA library described above were exposed to nitrocellulose filters (BA85, 82 mm, Schleicher & Schuell, Keene, NH) were placed on each plate. After plaque adsoφtion (approximately 20 minutes), the filters were removed and air dried for approximately 30 minutes. Subsequently, each filter was sequentially submerged for 30-60 seconds in 0.2 M NaOH, 1.5 M NaCl (100 ml); 2X SSC, 0.4 M Tris- HCl, pH 7.4 (100 ml); and 0.2X SSC (100 ml). Filters were then dried at room temperature for approximately 1 hour and subjected to 80°C under vacuum for 2 hours. Filters were probed with radiolabeled murine cyrόl cDNA. The excessive number of signals, indicative of a high level of false positive signals from related sequences, prevented identification of cyrόl cDNA.

Turning to a complicated, cyrόl -specific cloning strategy, human cyrόl cDNA clones were identified with probes generated by RT-PCR. In particular, the probe for screening the human placental cDNA library was a PCR fragment generated with degenerate primers by RT-PCR of total RNA from logarithmically growing WI38 cells. The primers were derived from the sequences corresponding to the most conserved region of the open reading frame of the mouse cyrόl cDNA. One primer, designated H61 -5 [5 '- GGGAATTCTG(TC)GG(GATC)TG(TC)TG(TC)AA(GA)GT(GC)TG-3 '], contains a degenerate sequence which, with the exception of the "GGGAATTC" sequence at the 5 ' end which was used to introduce an EcoRI site, is derived from nucleotides 327-346 (sense strand) of the mouse cyrόl sequence set forth in SEQ ID NO: l . The degeneracies appear in positions corresponding to the third position of codons in SEQ ID NO: l . The second primer used for PCR amplification of a human cyrόl sequence was designated H61 -3 [5'- CCGGATCC(GA)CA(GA)TT(GA)TA(GA)TT(GA)CA-3'], which, with the exception of the 5' sequence "CCGGATCC" used to introduce a BamHI site, corresponds to the anti-sense strand complementary to nucleotides 1236-1250 of the mouse cyrόl sequence set forth in SEQ ID NOT . The degeneracies occur in positions complementary to the third positions of codons in mouse cyrόl as set forth in SEQ ID NOT . The amplified cyrόl cDNA was cloned into the pBlueScript SK+ vector (Stratagene, La Jolla, CA) and sequenced with a Sequenase II kit (U.S. Biochemicals, Cleveland, OH).

Serial screenings of the human placental cDNA library led to the isolation of a clone containing a human cyrόl cDNA. The human cyrόl cDNA is approximately 1,500 bp in length. The human cDNA is contained on an EcoRI fragment cloned into the EcoRI site in pGΕM-2. As shown in SEQ ID NO:3, the human cDNA sequence includes the entire coding region for human Cyrό 1 , along with 120 bp of 5' flanking sequence, and about 150 bp of 3' flanking sequence. The polynucleotides of the invention maybe wholly or partially synthetic,

DNA or RNA, and single- or double-stranded. Because polynucleotides of the invention encode ECM signaling molecule polypeptides which may be fragments of an ECM signaling molecule protein, the polynucleotides may encode a partial sequence of an ECM signaling molecule. Polynucleotide sequences of the invention are useful for the production of ECM signaling molecules by recombinant methods and as hybridization probes for polynucleotides encoding ECM signaling molecules.

DNA polynucleotides according to the invention include genomic DNAs, cDNAs, and oligonucleotides comprising a coding sequence of an ECM signaling molecule, or a fragment or analog of an ECM signaling molecule, as described above, that retains at least one of the biological activities of an ECM signaling molecule such as the ability to promote cell adhesion, cell migration, or cell proliferation in such biological processes as angiogenesis, chondrogenesis, and oncogenesis, or the ability to elicit an antibody recognizing an ECM signaling molecule. Other polynucleotides according to the invention differ in sequence from sequences contained within native ECM signaling molecule polynucleotides (i.e., by the addition, deletion, insertion, or substitution of nucleotides) provided the polynucleotides encode a protein that retains at least one of the biological activities of an ECM signaling molecule. A polynucleotide sequence of the invention may differ from a native ECM signaling molecule polynucleotide sequence by silent mutations that do not alter the sequence of amino acids encoded therein. Additionally, polynucleotides of the invention may specify an ECM signaling molecule that differs in amino acid sequence from native ECM signaling molecule sequences or subsequences, as described above. For example, polynucleotides encoding polypeptides that differ in amino acid sequence from native ECM signaling molecules by conservative replacement of one or more amino acid residues, are contemplated by the invention. The invention also extends to polynucleotides that hybridize under standard stringent conditions to polynucleotides encoding an ECM signaling molecule of the invention, or that would hybridize but for the degeneracy of the genetic code. Exemplary stringent hybridization conditions involve hybridization at 42°C in 50% formamide, 5X SSC, 20 mM Na^»PO₄, pH 6.8 and washing in 0.2X SSC at 55°C. It is understood by those of skill in the art that variation in these conditions occurs based on the length and GC nucleotide content of the sequences to be hybridized. Formulas standard in the art are appropriate for determining exact hybridization conditions. See Sambrook et al. , Molecular Cloning: A Laboratory Manual (Second ed., Cold Spring Harbor Laboratory Press 1989) §§ 9.47-9.51.

ECM signaling molecule polynucleotides comprising RNA are also within the scope of the present invention. A preferred RNA polynucleotide according to the invention is an mRNA of human cyrόl . Other RNA polynucleotides of the invention include RNAs that differ from a native ECM signaling molecule mRNA by the insertion, deletion, addition, or substitution of nucleotides (see above), with the proviso that they encode a polypeptide retaining a biological activity associated with an ECM signaling molecule. Still other RNAs of the invention include anti-sense RNAs (i.e., RNAs comprising an RNA sequence that is complementary to an ECM signaling molecule mRNA).

Accordingly, in another embodiment a set of DNA fragments collectively spanning the human cyrόl cDNA were cloned in pGEM-2 and M13 derivatives using methods well known in the art to facilitate nucleotide sequence analyses. The pGEM-2 clones provided substrates for the enzymatic generation of serial deletions using techniques known in the art. This collection of clones, collectively containing a series of DNA fragments spanning various parts of the cyrόl cDNA coding region, are useful in the methods of the invention. The resulting series of nested pGEM-2 clones, in turn, provided substrates for nucleotide sequence analyses using the enzymatic chain terminating technique. The fragments are also useful as nucleic acid probes and for preparing Cyrόl deletion or truncation analogs. For example, the cyrόl cDN A clones may be used to isolate cyrόl clones from human genomic libraries that are commercially available. (Clontech Laboratories, Inc., Palo Alto, CA). Genomic clones, in turn, may be used to map the cyrόl locus in the human genome, a locus that may be associated with a known disease locus.

Other embodiments involve the polynucleotides of the invention contained in a variety of vectors, including plasmid, viral (e.g., prokaryotic and eukaryotic viral vectors derived from Lambda phage, Heipesviruses, Adenovirus, Adeno-associated viruses, Cytomegalovirus, Vaccinia Virus, the M13-fl-fd family of viruses, retroviruses, Baculovirus, and others), phagemid, cosmid, and YAC (i.e., Yeast Artificial Chromosome) vectors.

Yet other embodiments involve the polynucleotides of the invention contained within heterologous polynucleotide environments. Polynucleotides of the invention have been inserted into heterologous genomes, thereby creating transgcnes, and transgenic animals, according to the invention. In particular, two types of gene fusions containing partial murine cyrόl gene sequences have been used to generate transgenic mice. (See below). One type of fused gene recombined the coding sequence of cyrόl with one of three different promoters: 1 ) the K14 keratin promoter, 2) the β-actin promoter, or 3) the phosphoglycerokinase promoter. Adra et al, Gene 60:65-74 ( 1987). These fusion constructs were generated using standard techniques, as described below in the context of a phosphoglycerokinase promoter (pgk-l)-cyrόl fusion. An Xhol-Scal genomic DNA fragment containing the entire cyrόl coding region and all introns, but lacking the transcription initiation site and polyadenylation signal, was cloned into plasmid pgk/β-gal, replacing the lacZ coding sequence. The resulting construct placed cvroJ under the control of the strong pgk-1 promoter which is active in all cells. The second type of gene fusion recombined the cyrόl expression control sequences (i.e., promoter) with the E. coli β-galactosidase coding sequence. The cyrόl -lacZ fusion gene was constructed using the following approach. A DNA fragment spanning nucleotides -2065 to +65 relative to the transcription initiation nucleotide was used to replace the pgk-1 promoter (Adra et al. [ 1987]) in plasmid pgk/β-gal by blunt-end cloning. In addition, the polyadenylation signal from the bovine growth hormone gene was cloned into the plasmid containing the fusion gene. The resulting construct, plasmid 2/lacZ, has the E. coli lacZ gene under the transcriptional control of a 2 kb DNA fragment containing the cyrόl promoter. The related plasmid 1 Λ/lacZ was derived from plasmid llaeZby removing about 600 bp of cyrόl DNA found upstream of an Afffl site. Also, plasmid 2M/lacZ resembles plasmid 2/lacZ, except for a C-to-T transition in the CArG Box, created by PCR. These constructs were excised from the vectors by Notl digestion, purified using GeneClean (BiolOl, Inc., La Jolla, CA), and used to generate transgenic mice (see below). A cDNA fragment encoding mouse fisp 12 has also been cloned using standard techniques. Ryseck et al, Cell Growth & Diff. 2.225-233 ( 1991 ), incoφorated herein by reference. The cloning was accomplished by ligating an XlioW fragment containing the fispl 2 cDNA coding region into -S HI-cleaved pBlueBacIlI, a baculovirus expression vector (Invitrogen Corp., San Diego, CA). Recombinant baculovirus clones were obtained as described in Summers et al, TXAg. Exp. Sta., Bulletin 1555 (1987).

The human ortholog of fispl 2, the gene encoding CTGF, was cloned by screening a fusion cDNA library with anti-Platelet-Derived Growth Factor (anti- PDGF) antibodies, as described in U.S. Patent No. 5,408,040, column 12, line 16, to column 13, line 29, incoφorated herein by reference. The screening strategy exploited the immunological cross-reactivity of CTGF and PDGF.

The cloned copies of the cyrόl, fispl 2, and ctg/XDNAs provide a ready source for polynucleotide probes to facilitate the isolation of genomic coding regions, as well as allelic variants of the genomic DNAs or cDNAs. In addition, the existing cDNA clones, or clones isolated by probing as described above, may be used to generate transgenic organisms. For example, transgenic mice harboring cyr67 have been generated using standard techniques, as described in the next Example. A clone, hCyrόl cDNA, containing the human cyrόl cDNA sequence set forth in SEQ ID NO:3, and a bacterial strain transformed with that clone, Escherichia coli DH5α (hCyrόlcDNA), were deposited with the American Type Culture Collection, 10801 University Blvd., Manassas, VA 201 10-2209 USA (formerly at 12301 Parklawn Drive, Rockville, MD 20852 USA), on March 14, 1997.

Example 2 Sequence Analyses

The nucleotide sequence of murine cyrόl has been described, O'Brien et al. (1990); Latinkic et al. Nucl. Acids Res. /P:3261 -3267 (1991 ), and is set out herein as SEQ ID NOT .

The deduced amino acid sequence of murine Cyrό l has been reported, O'Brien et al ( 1990), and is set forth in SEQ ID NO:2.

The nucleotide sequence of the human cyr67 cDNA was determined using the method of Sanger, as described in Sambrook et al Sequencing templates were generated by constructing a series of nested deletions from a pGEM-2 human cyrόl cDNA clone, as described in Example 1 above. The human cyroJ cDNA sequence is set forth in SEQ ID NO:3. The amino acid sequence of human Cyrόl was deduced from the human cyrόl cDNA sequence and is set forth in SEQ ID NO:4.

A comparison of the mouse and human Cyrόl sequences, presented in SEQ ID NO:2 and SEQ ID NO:4, respectively, reveals 91% similarity. Both sequences exhibit an N-terminal signal sequence indicative of a processed and secreted protein; both proteins also contain 38 cysteine residues, distributed throughout both proteins but notably absent from the central regions of both murine and human Cyrόl . Notably, the region of greatest sequence divergence between the mouse and human Cyrόl coding regions is this central region free of cysteine residues. However, the 5 ' untranslated regions of the mouse and human cyr67 cDNAs are even more divergent (67% similarity). In contrast, the 3' untranslated regions are the most similar regions (91% similarity). In overall length, the encoded murine Cyrόl has 379 amino acids; human Cyrόl has 381 amino acids.

Afispl2 cDNA sequence has also been determined and is set out in SEQ ID NO:5. The amino acid sequence of Fispl2 has been deduced from the fisp!2 cDNA sequence and is set forth in SEQ ID NO:6. A comparison of the amino acid sequences of murine Cyrό 1 and Fispl 2 reveals that the two proteins are 65% identical. The structural similarity of Cyrόl and Fispl2 is consistent with the similar functional properties of the two proteins, described below.

A partial cDNA sequence of CTGF, containing the complete CTGF coding region, has also been determined. The CTGF cDNA sequence was obtained using Ml 3 clones as templates for enzymatic sequencing reactions, as described. '040 Patent, at column 12, line 68 to column 13, line 14. Additional cloning coupled with double-stranded enzymatic sequencing reactions, elucidated the entire sequence of the cDNA encoding CTGF. U.S. Patent No. 5,408,040, column 14, line 44, to column 15, line 8, incoφorated herein by reference. The nucleotide sequence of the cDNA encoding CTGF is presented herein in SEQ ID NO:7. The deduced amino acid sequence of the cDNA encoding CTGF is presented in SEQ ID NO:8.

Example 3 RNA Analyses

Polynucleotide probes are useful diagnostic tools for angiogenic, and other, disorders correlated with Cyrόl expression because properly designed probes can reveal the location, and level, of cyroJ gene expression at the transcriptional level. The expression of cyrόl, in turn, indicates whether or not genes controlling the process of angiogenesis are being expressed at typical, or expected, levels.

Using these tools, the mouse cyrόl mRNA expression pattern was determined using an RNase protection technique. O'Brien et al, (1992). In particular, a 289 nucleotide antisense riboprobe was used that would protect 246 nucleotides of the murine cyrόl mRNA (nucleotides 67 to 313 using the numbering of O'Brien et al.) The assays showed levels of cyrόl mRNA in PSA-1 cells (10 μg of total RNA) from either the undifferentiated state or stages 1, 2, and 3 of differentiation (PSA-1 cells undergo three stages of cellular differentiation corresponding to mouse embryonic cells of the following gestational ages, in days: 4.5-6.5 [PSA-1 stage 1]; 6.5-8.5 [PSA-1 stage 2]; 8.5- 10.5 [PSA-1 stage 3]). A comparison of the protection of whole embryonic and placental total RNAs (20 μg each) showed that cyrόl is expressed in embryonic tissues at times that are coincident with the processes of cell differentiation and proliferation. Expression characteristics of human cyrόl were determined by Northern analyses, using techniques that are standard in the art. Sambrook et al RNA was isolated from the human diploid fibroblastic cell line WI38 (ATCC CCL-75). In addition, RNA was isolated from rat cells (REF52), hamster cells (CHO), and monkey cells (BSC40). Each of the cell lines was grown to confluence in MEM- 10 (Eagle's Minimal Essential Medium with Earle's salts [GIBCO-BRL, Inc.], 2 mM glutamine, and 10%> fetal calf serum [fcs]) and maintained in MEM-0.5 (a 0.5% serum medium) for two days. Cultures were then stimulated with 20% fcs, in the presence or absence of cycloheximide, by techniques known in the art. Lau et al. (1985; 1987). Ten microgram aliquots of RNA isolated from these cell lines were then fractionated by formaldehyde-agarose gel electrophoresis, transferred and immobilized on nitrocellulose filters, and exposed to a full-length [ ²P]-radiolabeled murine cyrόl cDNA probe under hybridization conditions of high stringency. Human cyrόl RNA expression was similar to murine cyrόl expression. Both mouse and human cyroJ expression yielded approximately 2 kilobase RNAs. Additionally, both mouse and human expression of Cyrόl were stimulated by serum and were resistant to cycloheximide.

The distribution of human cyrόl mRNA was also examined using multiple tissue Northern blots (Clontech). The blots were hybridized in an ExpressHyb Solution (Clontech) according to the manufacturer's instructions. The results showed that cyrόl mRNA is abundant in the human heart, lung, pancreas, and placenta; is present at low levels in skeletal muscle, kidney and brain; and is not detectable in liver. These results are consistent with the expression of cyro7 in mouse tissues. In addition, total cellular RNA was isolated from human skin fibroblasts

(HSFs) that were either quiescent, growing exponentially, stimulated by serum, or exposed to cycloheximide. HUVE cells (ATCC CRL 1730) were maintained in Ham's F 12 medium supplemented with 10% fbs (Intergene), 100 μg/ml heparin (Gibco BRL) and 30 μg/ml endothelial cell growth supplement (Collaborative Biomedical Products). Human skin fibroblasts (HSF, ATCC CRL-1475) and WI38 fibroblasts (ATCC CCL-75) were grown in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% fbs. Quiescent HSFs were prepared by growth in DMEM supplemented with 10% fbs to confluence followed by changing the medium to DMEM containing 0.1% fbs, for 2 days. Serum stimulation was carried out by changing the medium to 20% fbs for 1 hour. Where indicated, cycloheximide was added to 10 μg/ml simultaneously with serum for 3 hours.

RNAs from the aforementioned cells were isolated using a guanidinium isothiocyanate protocol. Chomczynski et al, Anal. Biochem. 7(52. 156-159 (1987). RNA samples were analyzed by electrophoretic separation in formaldehyde- agarose gels followed by transfer to nylon filters. Blots were hybridized with random-primed probes generated using either cyrόl or GAPDH as a template. Adams et al, Nature 355:632-634 (1992). The results indicated that human cyrόl mRNA is not detectably present in quiescent human skin fibroblasts, is abundant in logarithmically growing and seaim stimulated HSFs, and is superinduced by cycloheximide.

The analysis of RNA encoding CTGF also involved techniques that are standard in the art. In particular, investigation of RNA encoding CTGF involved the isolation of total cellular RNA and Northern analyses, performed as described in U.S. Patent No. 5,408,040, column 1 1, line 59, to column 12, line 14, and column 13, lines 10-29, incoφorated herein by reference. A 2.4 kb RNA was identified. The expression of CTGF was high in the placenta, lung, heart, kidney, skeletal muscle and pancreas. However, CTGF expression was low in the liver and brain.

Example 4

Transgenic Animals

The construction of transgenic mice bearing integrated copies of recombinant cyr<57 sequences was accomplished using linear DNA fragments containing a fusion gene. The cyrόl coding sequence was independently fused to the β-actin, K14, and pgk promoters, described above. Expression of cyrόl was driven by these promoters in the transgenic animals. The fusion gene was produced by appropriate restriction endonuclease digestions, using standard techniques. The fusion gene fragments were injected into single-cell zygotes of

Swiss Webster mice. The injected zygotes were then implanted into pseudopregnant females. Several litters of mice were produced in this manner. Newboms exhibiting unusual phenotypes were subjected to additional analyses. For example, neonatal transgenic mice expressing cyr< under the pgk promoter exhibited skeletal deformities, including curly tails, immobile joints, and twisted limbs, resulting in locomotive difficulties. These mice typically were runted and died within seven days of birth. Transgenic mice expressing cyr<57 under the β- actin promoter showed no obvious phenotype except that the mice were smaller. When mice bearing the transgene were back-crossed to the in-bred strain C57BL/6, the progeny mice became progressively more runted with continued back-crossing. After three to four such back-crosses, essentially no progeny survive to reproduce. Transgenic mice expressing cyro^~7 under the K 14 promoter exhibited a form of fibrotic dermatitis. The pathology involved excessive surface scratching, sometimes resulting in bleeding. Transgenic organisms having knockout mutations of cyrόl can also be created using these standard techniques, Hogan et al, Manipulating the Mouse Embryo: A Laboratory Manual (Cold Spring Harbor Laboratory Press 1994), and are useful as models of disease states.

Example 5 Cyrόl Expression

Native Cyrό 1 is expressed in embryonic tissues and is induced in a variety of wounded tissues. See below; see also, O'Brien et al. (1992). The tissue distribution of Cyrόl was examined with rabbit anti-Cyrόl polyclonal antibodies elicited using a conventional immunological technique (Harlow et al., 1987) and affinity-purified. Using affinity-purified anti-Cyrόl polyclonal antibodies according to the invention, cyrόl expression was found in a variety of tissues, including smooth muscle, cardiomyocytes, and endothelia of the cardiovascular system; brain, spinal cord, ganglia and neurons, and retina of the nervous system; cartilage and bone of the skeletal system; epidermis, hair, oral epithelia, and cornea of the skin; bronchioles and blood vessels of the lung; and placenta] tissues. In addition to expression studies directed towards native cyrόl (mRNA and protein), studies using cyr<57 transgenes, as described above, have contributed to our understanding of Cyrόl expression. The use of transgene fusions comprising the expression control sequences of cyrόl and the coding sequence of lacZ (encoding β-galactosidase) has provided a convenient colorimetric assay for protein expression.

The colorimetric assay involves the use of 5-Bromo-4-Chloro-3-Indolyl-β- D-Galactopyranoside (i.e., X-Gal) as a substrate for β-galactosidase, the gene product of lacZ. Enzymatic cleavage of X-Gal by β-galactosidase produces an intensely colored indigo dye useful in histochemical staining. In practice, embryonic and adult tissues subjected to analysis were dissected and fixed in 2% formaldehyde, 0.2% glutaraldehyde, 0.02% Nonidet P-40, and 0.01 sodium deoxycholate, in standard phosphate-buffered saline (PBS). Fixation times varied from 15-120 minutes, depending on the size and density of organ or embryo samples being subjected to analysis. Subsequently, samples were rinsed in PBS and stained overnight at 37°C in a PBS solution containing 5 mM potassium ferrocyanide, 5 mM potassium ferricyanide, 2 mM MgCl₂, 0.02% Nonidet P-40, 0.01% sodium deoxycholate and 1 mg/ml of X-Gal (40 mg/ml in dimethylsulfoxide [DMSO]). Samples were then rinsed in PBS, post-fixed in 4%> paraformaldehyde for 1 -2 hours, and stored in 70% ethanol at 4°C until subjected to microscopic examination. Mice containing the cyrόl -lacZ transgene were used to map the expression profile of cyrόl . The results are presented in Table I for embryonic tissues at day 12.5.

Table I

¹ Transgenic lines, S- stable (established) transgenic lines; T- transient lines +/- Expression pattern only partially reproduced.

The results indicate that Cyrόl is expressed in a variety of embryonic cell types. Additional infor ation has been gleaned from the ectopic expression of Cyrόl resulting from another type of transgene fusion comprising a heterologous expression control sequence coupled to the coding sequence of cyrόl . The control sequences, the K14 keratin promoter, the β-actin promoter, and the phosphoglycerokinase promoter, directed the expression of Cyrόl in a pattern that differed from its native expression.

Transgenic mice ectopically expressing Cyrό 1 were routinely smaller than wild type mice and exhibited a reduction in average life span. Moreover, these transgenic mice had abnormal hearts (i.e., thickened chamber walls with a corresponding reduction in internal capacity) and abnormal skeletons characterized by curved spines, joints swollen to the point of immobility, and curly tails. Therefore, ectopic expression of Cyrόl interferes with angiogenesis (blood vessel development and heart development) and chondrogenesis (skeletal development). In addition, transgenic mice carrying knockout mutations of cyrόl may be developed and tested as models of disease states associated with a lack of Cyrόl activity.

A strategy for the expression of recombinant cyrόl was designed using a Baculovirus expression vector in Sf9 cells. Expression systems involving Baculovirus expression vectors and Sf9 cells are described in Current Protocols in Molecular Biology §§ 16.9.1 - 16.12.6 (Ausubel et al., eds., 1987). One embodiment of the present invention implemented the expression strategy by cloning the murine cyrόl cDNA into pBlueBac2, a transfer vector. The recombinant clone, along with target AcMNPV (i.e., Autographa califomica nuclear polyhedrosis virus, or Baculovirus) DNA, were delivered into Sf9 cells by liposome-mediated transfection, using the MaxBac Kit (Invitrogen, Inc., San Diego, CA) according to the manufacturer's instructions. Recombinant virus was plaque-purified and amplified by 3 passages through Sf9 cells via infection.

Conditioned medium of Sf9 insect cells infected with a baculovirus construct driving the synthesis of murine Cyrόl was used as a source for purification of Cyrόl (see below). The purified recombinant Cyrόl retains certain characteristics of the endogenous protein, e.g., the heparin-binding activity of Cyrόl (described below) from 3T3 fibroblast cells and had a structure similar to the endogenous protein as revealed by independent peptide profiles produced by partial proteolysis using either chymotrypsin or trypsin (sequencing grade; Boehringer-Mannheim, Inc., Indianapolis, IN).

Human cyr<57 was also expressed using the baculovirus system. A Smal- Hindm fragment (corresponding to nucleotides 100-1649 of SEQ ID NO:3) of cyrόl cDNA spanning the entire human cyrόl open reading frame was subcloned into a pBlueBac3 baculovirus expression vector (Invitrogen). Recombinant baculovirus clones were obtained, plaque purified and amplified through three passages of Sf9 infection, using conventional techniques. Infection of Sf9 cells and human Cyrόl (hCyrόl ) purification was performed using standard techniques, with some modifications. Sf9 cells were maintained in serum-free Sf900-II medium (Sigma). Sf9 cells were seeded, at 2-3 x 10⁶ cells per 150 mm dish, in monolayer cultures and were infected with 5 plaque forming units (PFU) of recombinant virus per cell. The conditioned medium was collected at 8 and 96 hours post-infection, cleared by centrifugation (5000 .Y g, 5 minutes) and adjusted to 50 mM MES [2-(N-Moφholino)ethanesulfonic acid], pH 6.0, 1 mM PMSF (phenylmethylsulfonyl fluoride), and 1 mM EDTA. The medium was mixed with Sepharose S beads equilibrated with loading buffer (50 mM MES, pH 6.0, 1 mM PMSF, 1 mM EDTA, 150 mM NaCl) at a ratio of 5 ml Sepharose S beads per 500 ml of conditioned medium and the proteins were allowed to bind to the Sepharose S at 4°C (o/n) with gentle stirring. Sepharose S beads were collected by sedimentation without stirring for 20 minutes and applied to the column. The column was washed with 6 volumes of 0.3 M NaCl in loading buffer and recombinant human Cyrόl was eluted from the column with a step gradient of NaCl (0.4-0.8 M) in loading buffer. This procedure resulted in 3-4 milligrams of purified Cyrόl protein from 500 ml of conditioned medium, and the purified Cyrόl was over 90% pure as judged by Coomassie Blue staining of SDS-gels.

In another embodiment, the complete human cyrόl cDNA is cloned into a cytomegalovirus vector such as pBK-CMV (Stratagene, LaJolla, CA) using the Polymerase Chain Reaction (Hayashi, in PCR: The Polymerase Chain Reaction 3- 13 [Mullis et al. eds., Birkhauser 1994]) and Taq Polymerase with editing function, followed by conventional cloning techniques to insert the PCR fragment into a vector. The expression vector is then introduced into HUVE cells by liposome-mediated transfection. Recipient clones containing the vector-borne neo gene are selected using G418. Selected clones are expanded and Cyrόl expression is identified by Reverse Transcription-Polymerase Chain Reaction (i.e., RT-PCR; Chelly et al, in PCR: The Polymerase Chain Reaction 97-109 [Mullis et al. eds., Birkhauser 1994]) or Enzyme- Linked Immunosorbent Assays (i.e., ELISA; Stites et al, in Basic and Clinical Immunology 243 [Stites et al. eds., Appleton & Lange 1991]) assays.

In other embodiments of the invention, Cyrόl protein is expressed in bacterial cells or other expression systems (e.g., yeast) using the cyrόl cDNA coding region linked to promoters that are operative in the cell type being used. Using one of these approaches, Cyrόl protein may be obtained in a form that can be administered directly to patients, e.g., by intravenous routes, to treat angiogenic, chondrogenic, or oncogenic disorders. One of skill in the art would recognize that other administration routes are also available, e.g., topical or local application, liposome-mediated delivery techniques, or subcutaneous, intradermal, intraperitoneal, or intramuscular injection.

Example 6 Fispl 2 Expression The expression of Fispl 2, and a comparison of the expression characteristics of Cyrόl and Fispl 2, were investigated using immunohistochemical techniques. For these immunohistochemical analyses, tissue samples (see below) were initially subjected to methyl-Carnoy's fixative

(60% methanol, 30% chloroform and 10% glacial acetic acid) for 2-4 hours. They were then dehydrated, cleared and infiltrated in Paraplast X-tra wax at 55- 56°C for minimal duration. 7 μm thick sections were collected on poly-L-lysine- coated slides (Sigma), mounted and dewaxed. They were then treated with 0.03% solution of H₂O₂ in methanol for 30 minutes to inactivate endogenous peroxidase activity. After rehydration, sections were put in Tris-buffered saline (TBS: 10 mM Tris, pH 7.6 and 140 mM NaCl) for 15 minutes. At that point, sections were blotted to remove excess TBS with paper towels and blocked with 3% normal goat serum in TBS for 10 minutes in a humid chamber. Excess buffer was then drained and primary antibodies applied. Affinity purified anti-Cyrόl antibodies were diluted 1 :50 in 3% normal goat serum-TBS solution. Dilution for affinity-purified anti-Fispl2 antibody was 1 :25. Routine control was 3% normal goat serum-TBS, or irrelevant antibody (for example, monoclonal anti- smooth muscle cell α-actin). Specificity of staining was confiπned by incubation of anti-Cyrόl or anti-Fispl 2 antibodies with an excess of the corresponding antigen on ice for at least two hours prior to applying to sections. Complete competition was observed. By contrast, cross-competition (incubation of anti- Cyrόl antibodies with Fisp 12 antigen and vice versa) did not occur.

Primary antibodies were left on sections overnight at 4°C. They were then washed with TBS twice, and subjected to 30 minutes incubation with secondary antibodies at room temperature. Secondary antibodies used were goat anti-rabbit horseradish peroxidase conjugates from Boehringer-Mannheim, Inc., Indianapolis, IN (used at 1 :400 dilution). Sections were washed twice in TBS and chromogenic horseradish peroxidase substrate was applied for 5 minutes (1 mg/ml of diaminobenzidine in 50 mM Tris-HCl, pH 7.2 and 0.03% H₂O₂). Sections were then counterstained in Ehrlich's haematoxylin or in Alcian blue, dehydrated and mounted in Permount.

Mouse embryos between the neural fold (E8.5, embryo day 8.5) and late organogenesis (El 8.5) stages of development were sectioned and subjected to immunostaining with antigen-affinity-purified rabbit anti-Cyrόl and anti-Fispl 2 antibodies. As various organs developed during embryogenesis, the presence of Cyrόl and Fispl 2 was determined. Cyrόl and Fisp 12 were co-localized in a number of tissues and organs. A notable example is the placenta, where both proteins were readily detectable. In particular, both Cyrό 1 and Fisp 12 were found in and around the trophoblastic giant cells, corroborating the previous detection of cyrόl mRNA in these cells by in situ hybridization (O'Brien and Lau, 1992). Both Cyrόl and Fispl2 signals in immunohistochemical staining were blocked by either the corresponding Cyrόl or Fisp 12 antigen but not by each other, nor by irrelevant proteins, demonstrating specificity. In general, Cyrόl and Fisp 12 proteins could be detected both intracellularly and extracellularly. In addition to the placenta, both Cyrόl and Fisp 12 were detected in the cardiovascular system, including the smooth muscle, the cardiomyocytes, and the endothelia. Both proteins were also found in the bronchioles and the blood vessels in the lung. Low levels of anti-Cyrόl and anti-Fispl 2 staining could be detected transiently in the skeletal muscle. This staining is associated with connective tissue sheets, rather than myocytes; in this instance the staining pattern was clearly extracellular.

A more complex pattern of distribution was found in the epidermis and the epithelia. Both Cyrόl and Fispl 2 staining could be detected in the early, single-cell layer of embryonic epidermis, as well as in later, multilayered differentiating epidermis. Fisp 12 in epidermis declined to an undetectable level by the end of gestation and remained as such through adulthood, whereas Cyrόl was readily detectable in the epidermis. In the neonate, a strong staining for Fispl2 was seen in the oral epithelia where Cyrόl staining was much weaker, while Cyrόl was found in the upper jawbone where Fispl 2 was not observed. The anti-Fispl 2 signal in the oral epithelia gradually increased and remained intense into adulthood. In the tongue, both Cyrόl and Fisp 12 were seen in the keratinized epithelia, although the Fispl2 staining pattern, but not that of Cyrόl, excludes the filiform papillae.

Aside from the aforementioned sites of localization, Cyrόl and Fispl2 were also uniquely localized in several organ systems. For example, Cyrόl , but not Fisp 12, was present in skeletal and nervous systems. As expected from in situ hybridization results (O'Brien and Lau, 1992), Cyrόl protein was readily detected in the sclerotomal masses of the somites, and in cartilage and bone at later stages of development. In contrast, Fisp 12 was not detectable in the skeletal system. Since correlation with chondrocytic differentiation is one of the most striking features of cyrόl expression (O'Brien and Lau, 1992), the absence of Fispl 2 in the skeletal system may underscore an important difference in the biological roles of Cyrόl and Fispl2. In the E14.5 embryo, Cyrόl could be detected in the ventral spinal cord, dorsal ganglia, axial muscle and sclerotome- derived cartilaginous vertebrae. Fisp 12, however, was not detected in these tissues.

By contrast, Fisp 12 was uniquely present in various secretory tissues. Beginning at El 6.5, Fisp 12 could be detected in the pancreas, kidneys, and salivary glands. In the pancreas, Fisp 12 was strictly localized to the periphery of the islets of Langerhans. In the kidney, strong Fispl 2 staining was seen in the collecting tubules and Henle's loops, regions where Cyrόl was not found. In the mucous-type submandibular salivary gland only collecting ducts stained for Fispl 2, whereas in the mixed mucous-serous submandibular gland, both serous acini and collecting ducts stained. The signal in acini was peripheral, raising the possibility that Fisp 12 is capsule-associated. In simple holocrine sebaceous glands a strong acellular Fisp 12 signal was detected.

In summary, Cyrόl and Fisp 12 have been co-localized in the placenta, the cardiovascular system, the lung and the skin. Neither protein was detected in the digestive system or the endocrine glands. Unique localization of Cyrόl can be detected in the skeletal and central nervous system, and Fisp 12 is found in secretory tissues where Cyrόl is not.

An issue closely related to protein expression concerns the metabolic fate of the expressed proteins. Members of the cysteine-rich protein family have been localized. As discussed above, secreted Cyrόl is found in the ECM and on the cell surface but not in the culture medium (Yang et al. , 1991), yet secreted Fispl 2 was readily detected in the culture medium (Ryseck et al. , 1991 ). To address the question of whether Fisp 12 is also ECM-associated, the fate of both Cyrόl and Fisp 12 was followed using pulse-chase experiments. Serum-stimulated, sub- confluent NIH 3T3 fibroblasts were metabolically pulse-labeled for 1 hour and chased in cold medium for various times. Samples were fractionated into cellular, ECM, and medium fractions followed by immunoprecipitation to detect Cyrόl and Fispl 2. Both proteins have a similar short half-life of approximately 30 minutes in the cellular fraction, which includes both newly synthesized intracellular proteins as well as secreted proteins associated with the cell surface

(Yang and Lau, 1991). It should be noted that since Cyrόl is quantitatively secreted after synthesis and only a minor fraction is stably associated with the

ECM, the bulk of secreted Cyrό 1 is cell-surface associated (Yang and Lau, 1991 ).

A fraction of Cyrόl was chased into the ECM where it remained stable for several hours. Newly synthesized Fisp 12 was also chased into the ECM, where its half-life was only about 1 hour. A larger fraction of Fispl 2 was chased to the conditioned medium, where no Cyrόl was detectable. Fispl 2 in the conditioned medium also had a short half-life of about 2 hours. Thus, whereas Cyrόl is strongly associated with the ECM, Fisp 12 is associated with the ECM more transiently. This result suggests that Fisp 12 might be able to act at a site distant from its site of synthesis and secretion, whereas Cyrόl may act more locally.

Since many ECM proteins associate with the matrix via interaction with heparan sulfate proteoglycans, the affinity with which a protein binds heparin might be a factor in its interaction with the ECM. The results of heparin binding assays, described below, are consistent with this hypothesis.

Example 7 Protein Purification

Serum-stimulated NIH 3T3 fibroblast cells were lysed to provide a source of native murine Cyrόl . Yang et al. Similarly, human fibroblasts are a source of native human Cyrόl .

Recombinant murine Cyrόl was purified from Sf9 cells harboring the recombinant Baculovirus vector, described above, containing the complete cyrόl coding sequence. Although murine Cyrόl in Sf9 cell lysates formed insoluble aggregates as was the case with bacterial cell extracts, approximately 10% of the Cyrόl synthesized was secreted into the medium in a soluble form. The soluble, secreted form of Cyrόl was therefore subjected to purification.

Initially, subconfluent Sf9 cells in monolayer cultures were generated in supplemented Grace's medium (GIBCO-BRL, Inc., Grand Island, NY). Grace, Nature 795:788 (1962). The Sf9 cells were then infected with 10 plaque- forming-units/cell of the recombinant Baculovirus vector, incubated for 16 hours, and fed with serum-free Grace's medium. These cells were expanded in serum- free Grace's Medium. The conditioned medium was collected 48 hours post- infection, although Cyrόl expression could be detected in the medium 24 hours after infection. Subsequently, the conditioned medium was cleared by centrifugation at 5000 x g for 5 minutes, chilled to 4°C, adjusted to 50 mM MES, pH 6.0, 2 mM EDTA (Ethylenediamine tetraacetic acid), 1 mM PMSF (Phenylmethylsulfonyl fluoride) and applied to a Sepharose S column (Sigma Chemical Co., St. Louis, MO) at 4°C (5 ml void volume per 500 ml medium). The column was washed with a buffer (50 mM MES, pH 6.0, 2 mM EDTA, 0.5 mM PMSF) containing 150 NaCl, and bound proteins were eluted with a linear gradient of NaCl (0.2-1.0 M) in the same buffer. The pooled fractions of Cyrόl eluted at 0.6-0.7 M NaCl as a distinct broad peak. The column fractions were 90%) pure, as determined by 10% SDS-PAGE followed by Coomassie Blue staining or Western analysis, using techniques that are standard in the art. Yang et al. ; see also, Sambrook et al, supra. For Western analysis, blots were probed with affinity-purified anti-Cyrόl antibodies as described in Yang et al, supra. After antibody probing, Western blots were stained with ECL™ (i.e., Enhanced ChemiLuminescence) detection reagents (Amersham Coφ., Arlington Heights, IL). Fractions containing Cyrόl were pooled, adjusted to pH 7.5 with Tris-HCl, pH 7.5, and glycerol was added to 10% prior to storage of the aliquots at -70°C. Protein concentration was determined by the modified Lowry method using the BioRad protein assay kit (BioRad Laboratories, Inc., Hercules, CA). This purification procedure was repeated at least five times with similar results. The typical yield was 3-4 mg of 90% pure Cyrόl protein from 500 ml of conditioned medium.

Fisp 12 was purified using a modification of the Cyrόl purification scheme (Kireeva et al, Exp. Cell Res. 233:63-11 [1997]). Serum-free conditioned media (500 ml) of Sf9 cells infected at 10 pfu per cell were collected 48 hours post- infection and loaded onto a 5-ml Sepharose S (Sigma Chemical Co., St. Louis, MO) column. A fter extensive washing at 0.2 M and 0.4 M NaCl, bound proteins were recovered by step elution with 50 mM MES (pH 6.0) containing 0.5 M NaCl. Fractions containing Fispl 2 of greater than 80% purity were pooled, NaCl adjusted to 0.15 M and the protein was concentrated 3-5 fold on a 0.5 ml Sepharose S column with elution of the protein at 0.6 M NaCl.

This purification scheme allowed the isolation of 1.5 mg of recombinant Fisp 12 protein of at least 80% purity from 500 ml of serum-free conditioned media. CTGF was purified by affinity chromatography using anti-PDGF cross- reactivity between CTGF and PDGF, as described in U.S. Patent No. 5,408,040, column 7, line 15, to column 9, line 63, incoφorated herein by reference.

Example 8 Polypeptide Characterization

The murine Cyrόl protein has a M_r of 41,000 and is 379 amino acids long including the N-terminal secretory signal. There is 91% amino acid sequence identity with the 381 amino acid sequence of the human protein. Those regions of the mouse and human proteins contributing to the simi larity of the two proteins would be expected to participate in the biological activities shared by the two polypeptides and disclosed herein. However, the mouse and human proteins do diverge significantly in the central portion of the proteins, where each protein is devoid of cysteines. See, O'Brien et al, Cell Growth & Diffi 3:645-654 ( 1992). A cysteine-free region in the murine Cyrόl amino acid sequence is found between amino acid residues 164 to 226 (SEQ ID NO:2). A corresponding cysteine-free region is found in the human Cyrόl amino acid sequence between amino acid residues 163 to 229 (SEQ ID NO:4). More particularly, the mouse and human Cyrόl proteins are most divergent between Cyrόl amino acids 170-185 and 210- 225. Other members of the ECM signaling molecule family of cysteine-rich proteins, e.g., Fispl2 (SEQ ID NO:6) and CTGF (SEQ ID NO:8), exhibit similar structures suggestive of secreted proteins having sequences dominated by cysteine residues.

Because murine Cyrόl contains 38 cysteines in the 355 amino acid secreted portion, the contribution of disulfide bond formation to Cyrόl tertiary structure was investigated. Exposure of Cyrό 1 to 10 mM dithiothreitol (DTT) for 16 hours did not affect the ability of Cyrό l to mediate cell attachment (see below). However, Cyrόl was inactivated by heating at 75°C for 5 minutes, by incubation in 100 mM HCl, or upon extensive digestion with chymotrypsin. These results indicate that murine Cyrόl is a heat- and acid-labile protein whose active conformation is not sensitive to reducing agents. The aforementioned structural similarities of murine and human Cyrόl polypeptides suggests that human Cyrόl may also be sensitive to heat or acid, but insensitive to reducing agents. In addition, Cyrόl is neither phosphorylated nor glycosylated. To determine if the purified recombinant murine Cyrόl described above was the same as native murine Cyrόl , two additional characteristics of mouse Cyrόl were determined. First, two independent protein fingeφrints of recombinant and native murine Cyrόl were obtained. Purified recombinant murine Cyrόl and a lysate of serum-stimulated 3T3 cells, known to contain native murine Cyrόl, were subjected to limited proteolysis with either trypsin or chymotrypsin, and their digestion products were compared. Partial tryptic digests of both the recombinant protein and cell lysate resulted in two Cyrόl fragments of approximately 21 and 19 kDa. Similarly, fingeφrinting of both preparations by partial chymotrypsin digestion produced stable 23 kDa fragments from recombinant murine Cyrόl and native murine Cyrόl .

Another criterion used to assess the properties of recombinant Cyrόl was its ability to bind heparin, described below. Purified recombinant murine Cyrόl bound quantitatively to heparin-sepharose at 0.15 M NaCl and was eluted at 0.8-1.0 M NaCl. This heparin binding capacity is similar to native murine Cyrόl obtained from serum-stimulated mouse fibroblasts. Because of the similarities of the murine and human Cyrόl proteins, recombinant human Cyrόl should exhibit properties similar to the native human Cyrόl , as was the case for the murine polypeptides.

The polypeptides of the invention also extend to fragments, analogs, and derivatives of the aforementioned full-length ECM signaling molecules such as human and mouse Cyrόl . The invention contemplates peptide fragments of ECM signaling molecules that retain at least one biological activity of an ECM signaling molecule, as described above. Candidate fragments for retaining at least one biological activity of an ECM signaling molecule include fragments that have an amino acid sequence corresponding to a conserved region of the known ECM signaling molecules. For example, fragments retaining one or more of the conserved cysteine residues of ECM signaling molecules would be likely candidates for ECM signaling molecule fragments that retain at least one biological activity. Beyond the naturally occurring amino acid sequences of ECM signaling molecule fragments, the polypeptides of the invention include analogs of the amino acid sequences or subsequences of native ECM signaling molecules. ECM signaling molecule analogs are polypeptides that differ in amino acid sequence from native ECM signaling molecules but retain at least one biological activity of a native ECM signaling molecule, as described above. These analogs may differ in amino acid sequence from native ECM signaling molecules, e.g., by the insertion, deletion, or conservative substitution of amino acids. A conservative substitution of an amino acid, i.e., replacing an amino acid with a different amino acid of similar properties (e.g., hydrophilicity, degree and distribution of charged regions) is recognized in the art as typically involving a minor change. These minor changes can be identified, in part, by considering the hydropathic index of amino acids, as understood in the art. Kyte et al, J. Moi Biol. /57.T05- 132 (1982). The hydropathic index of an amino acid is based on a consideration of its hydrophobicity and charge, and include the following values: alanine (+1.8), arginine (-4.5), asparagine (-3.5), aspartate (-3.5), cysteine/cystine (+2.5), glycine (-0.4), glutamate (-3.5), glutamine (-3.5), histidine (-3.2), isoleucine (+4.5), leucine (+3.8), lysine (-3.9), methionine (+1.9), phenylalanine (+2.8), proline (-1.6), serine (-0.8), threonine (-0.7), tryptophan (- 0.9), tyrosine (- 1.3), and valine (+4.2). It is known in the art that amino acids of similar hydropathic indexes can be substituted and still retain protein function. Preferably, amino acids having hydropathic indexes of ± 2 are substituted.

The hydrophilicity of amino acids can also be used to reveal substitutions that would result in proteins retaining biological function. A consideration of the hydrophilicity of amino acids in the context of a polypeptide permits calculation of the greatest local average hydrophilicity of that polypeptide, a useful measure that has been reported to correlate well with antigenicity and immunogenicity. U.S. Patent No. 4,554, 101 , incorporated herein by reference. Hydrophilicity values for each of the common amino acids, as reported in U.S. Patent No. 4,554,101 , are: alanine (-0.5), arginine (+3.0), asparagine (+0.2), aspartate (+3.0 ± 1 ), cysteine (- 1.0), glycine (0), glutamate (+3.0 ± 1 ), glutamine (+0.2), histidine (-0.5), isoleucine (-1.8), leucine (-1.8), lysine (+3.0), methionine (-1.3), phenylalanine (-2.5), proline (-0.5 ± 1), serine (+0.3), threonine (-0.4), tryptophan (-3.4). tyrosine (-2.3), and valine (-1.5). Substitution of amino acids having similar hydrophilicity values can result in proteins retaining biological activity, for example immunogenicity, as is understood in the art. Preferably, substitutions are performed with amino acids having hydrophilicity values within ± 2 of each other. Both the hydrophobicity index and the hydrophilicity value of amino acids are influenced by the particular side chain of that amino acid. Consistent with that observation, amino acid substitutions that are compatible with biological function are understood to depend on the relative similarity of the amino acids. and particularly the side chains of those amino acids, as revealed by the hydrophobicity, hydrophilicity, charge, size, and other properties.

Additionally, computerized algorithms are available to assist in predicting amino acid sequence domains likely to be accessible to an aqueous solvent. These domains are known in the art to frequently be disposed towards the exterior of a protein, thereby potentially contributing to binding determinants, including antigenic determinants. Having the DNA sequence in hand, the preparation of such analogs is accomplished by methods well known in the art (e.g., site- directed) mutagenesis and other techniques.

Derivatives of ECM signaling molecules are also contemplated by the invention. ECM signaling molecule derivatives are proteins or peptides that differ from native ECM signaling molecules in ways other than primary structure (i.e., amino acid sequence). By way of illustration, ECM signaling molecule derivatives may differ from native ECM signaling molecules by being glycosylated, one form of post-translational modification. For example, polypeptides may exhibit glycosylation patterns due to expression in heterologous systems. If these polypeptides retain at least one biological activity of a native ECM signaling molecule, then these polypeptides are ECM signaling molecule derivatives according to the invention. Other ECM signaling molecule derivatives include, but are not limited to, fusion proteins having a covalently modified - or C-terminus, PEGylated polypeptides, polypeptides associated with lipid moieties, alkylated polypeptides, polypeptides linked via an amino acid side- chain functional group to other polypeptides or chemicals, and additional modifications as would be understood in the art. In addition, the invention contemplates ECM signaling molecule-related polypeptides that bind to an ECM signaling molecule receptor, as described below.

The various polypeptides of the present invention, as described above, may be provided as discrete polypeptides or be linked, e.g., by covalent bonds, to other compounds. For example, immunogenic carriers such as Keyhole Limpet Hemocyanin may be bound to a ECM signaling molecule of the invention.

Example 9

Heparin Binding Assay

The heparin binding assay for native murine Cyrόl , described in Yang et al, was modified for the purified recombinant murine protein. Initially, recombinant pun lied Cyrό 1 was suspended in RIPA (Radioimmunoprecipitation assay) buffer (150 mM NaCl, 1.0% NP-40, 0.5% deoxycholate, 0.1% SDS, 50 mM Tris-HCl, pH 8.0, 1 M phenylmethylsulfonyl fluoride). Next, 200 μl of a 50% (v/v) slurry of heparin-Sepharose CL 6B beads (Pharmacia-LKB Biotechnology, Inc., Piscataway, NJ) was added to 100 μl of the recombinant Cyrόl solution and incubated for 1 hour. Under these conditions, human Cyrόl was quantitatively bound to heparin-agarose. Application of a salt concentration gradient in RIPA buffer resulted in the elution of recombinant murine Cyrόl at 0.8-1.0M NaCl. The elution profile of the recombinant protein was similar to the elution profile for native murine Cyrόl . One might expect that Fispl 2 would bind heparin with lower affinity than Cyrόl, as it does not interact with the ECM as strongly as Cyrόl . To examine this possibility, metabolically labeled ([MSJ-cysteine; 100 μCi per 100 mm dish; ICN) cell lysates were incubated with heparin agarose beads which were subsequently washed to remove unbound proteins. Bound proteins were eluted in increasing salt concentrations. Fispl 2 from cell lysates was retained on heparin agarose but was eluted by 0.2 to 0.6 M NaCl with peak elution at 0.4 M NaCl. This is in contrast to Cyrόl , which was eluted at significantly higher concentrations of NaCl. This difference in heparin binding is consistent with the differing affinities of Cyrόl and Fisp 12 for the ECM, suggesting that binding to heparan sulfate proteoglycans may be a primary mechanism by which both proteins associate with the ECM.

Example 10 Receptors Human Cyrόl, like murine Cyrόl, was localized to the cell surface and

ECM. The localization of Cyrόl to the cell surface implicated a cell surface receptor binding Cyrόl . Consistent with that implication, the biological effects of Cyrόl are mediated by the α_vβ₃ integrin, or vitronectin receptor. The α_vβ₃ integrin, in association with other integrins, forms protein clusters providing focal points for cytoskeletal attachment. Cyrόl induces the formation of protein clusters, including the protein clusters containing the α_vβ₃ integrin. In addition, using an in vitro assay, the biological effects of Cyrόl , including Cyrόl -induced cell adhesion and mitogenesis, were abolished by the addition of either one of two monoclonal antibodies- LM609 (Cheresh, Proc. Natl. Acad. Sci. [USA] 84:6411 - 6475 [1987]) or anti-VnR l (Chen et al, Blood 5(5:2606-2615 [1995])- directed to the α β₃ integrin. This data led to the identification of the α,,β₃ integrin as the Cyrόl receptor.

Cyrόl induction of HUVE cell adhesion, described in Example 13 below, led to an investigation of the divalent cation-sensitive cell surface receptors expressed by HUVE cells. The cell adhesion properties of Cyrόl were used to identify the receptor, which is a divalent cation-sensitive cell surface receptor. The ability of Cyrόl to mediate cell adhesion, coupled with the strict requirement for divalent cations in the process, indicated that Cyrόl interacts with one of the divalent cation-dependent cell adhesion molecules from the integrin, selectin, or cadherin families. Ruoslahti et al, Exp. CellRes. 227: 1-11 (1996). Using well- characterized approaches to receptor identification, a series of inhibition studies were conducted. Inhibitors, or blocking agents, of various degrees of specificity (EDTA, similar to the EGTA described above; inhibitory peptides bearing variants of the RGD (single letter amino acid code) integrin recognition motif, such as RGDS, SGDR, and RGDSPK (Ruoslahti, et al, Science 235:491 -497 [] 9S7\, Ruoslahti. E., Ann. Rev. of Cell and Dev. Biol 72:698-715 [1996]); and known, specific anti-receptor antibodies) were used to identify a Cyrόl receptor. That receptor was the α_vβ integrin, also known to function as the vitronectin receptor. Confirmation of that identification was obtained by showing that antibody LM609, a specific anti-α_vβ₃ integrin antibody, could block the effect of Cyrό 1 on cell adhesion. Integrins form a large family of heterodimeric adhesion receptors, with a broad ligand specificity range, involved in cell-cell and cell-matrix interactions. Beyond their requirement for divalent cations and their involvement in cell-matrix adhesion events [Hynes, R.O., Cell (59: 1 1-25 (1992)], integrins also are involved in cell migration [Damsky etal, Curr. Opin. Cell Biol. 4:772-781 (1992); Doerr et al, J. Biol Chem. 271 :2443-447 (1996)] and proliferation [Juliano et al, J. Cell Biol, 720:577-585 (1993); Plopper et al, Moi Biol. Cell 6:1349-1365 (1995); and Clark et al, Science 268:233-239 (1995)], two additional processes associated with Cyrόl activity. The c^β₃ integrin was found to be essential for Cyrόl -mediated cell adhesion.

Characterization of CTGF binding to cells has been reported to occur through a cell surface receptor that also interacts with PDGF-BB (the BB isoform of PDGF), as recited in U.S. Patent No. 5,408,040, column 1 1 , line 10, to column 12, line 14, incoφorated herein by reference. The identification of the foregoing receptors permits the design and production of molecules and which bind to the respective receptors to inhibit the activities of ECM molecules.

Example 11 A nti-ECM Signaling Molecule A ntibodies

Antibodies, optionally attached to a label or to a toxin as described below, are also contemplated by the present invention. The availability of the human cyrόl cDNA sequence and the Cyrόl deduced protein sequence facilitate the implementation of methods designed to elicit anti-Cyrόl antibodies using a number of techniques that are standard in the art. Harlow et al.

In one embodiment, polyclonal antibodies directed against Cyrόl are generated. The generation of anti-Cyrόl antibodies specific for human Cyrόl , for example, is optimized by designing appropriate antigens. The human Cyrόl protein is 381 amino acids long, including the N-terminal secretory signal. As described above, human Cyrό 1 exhibits a 91 % amino acid sequence identity with the 379 amino acid sequence of the mouse protein. However, the mouse and human proteins diverge most significantly in the central portion of the proteins. where they are devoid of cysteines (see above). These sequence differences are exploited to elicit antibodies specific to the human Cyrόl by using as an antigen a peptide having a sequence derived from one of the divergent regions in the human protein, although antibodies directed to a conserved region are also contemplated by the invention. In another embodiment of the present invention, monoclonal antibodies are elicited using intact recombinant human Cyrόl although a fragment may be used. Female BALB/c mice are inoculated intraperitoneally with a mixture of 0.25 ml recombinant human Cyrόl (5-50 micrograms), bacterially produced or produced in eukaryotic cells, and 0.25 ml complete Freund's adjuvant. Fourteen days later the injections are repeated with the substitution of incomplete Freund's adjuvant for complete Freund's adjuvant. After an additional two weeks, another injection of human Cyrόl in incomplete Freund's adjuvant is administered. About two weeks after the third injection, tail bleeds are performed and serum samples are screened for human anti-Cyrόl antibodies by immunoprecipitation with radiolabeled recombinant human Cyrόl . About two months after the initial injectioπ, mice whose sera yield the highest antibody titers are given booster injections of Cyrόl (5-50 micrograms in incomplete Freund's adjuvant, 0.1 ml intravenously and 0.1 ml intraperitoneally). Three days after the booster injection, the mice are sacrificed. Splenocytes are then isolated from each mouse using standard techniques, and the cells are washed and individually fused with a myeloma cell line, e.g., the X63Ag8.653 cell line (Harlow et al), using polyethylene glycol, by techniques that are known in the art. Other suitable cell lines for fusion with splenocytes are described in Harlow et al, at page 144, Table 6.2, incoφorated herein by reference. Fused cells are removed from the PEG solution, diluted into a counter-selective medium (e.g., Hypoxanthine- Aminopterin-Thymidine or HAT medium) to kill unfused myeloma cells, and inoculated into multi-well tissue culture dishes.

About 1 -2 weeks later, samples of the tissue culture supematants are removed from wells containing growing hybridomas, and tested for the presence of anti-Cyrόl antibodies by binding to recombinant human Cyrόl bound to nitrocellulose and screening with labeled anti-immunoglobulin antibody in a standard antibody-capture assay. Cells from positive wells are grown and single cells are cloned on feeder layers of splenocytes. The cloned cell lines are stored frozen. Monoclonal antibodies are collected and purified using standard techniques, e.g., hydroxylapatite chromatography. In an alternative, Cyrό l peptides used as antigens, may be attached to immunogenic carriers such as keyhole limpet hemocyanin carrier protein, to elicit monoclonal anti-Cyrόl antibodies.

Another embodiment involves the generation of antibody products against a fusion protein containing part, or all, of human Cyrόl , including enough of the protein sequence to exhibit a useful epitope in a fusion protein. The fusion of the large subunit of anthranilate synthase (i.e., TφE) to murine Cyrόl , and the fusion of glutathione S-transferase (i.e., GST) to murine Cyrόl, have been used to successfully raise antibodies against murine Cyrόl . Yang et al. In addition, a wide variety of polypeptides, well known to those of skill in the art, may be used in the formation of Cyrόl fusion polypeptides according to the invention.

More particularly, Yang reported a TφE-Cyr61 fusion polypeptide that was expressed from a recombinant clone constructed by cloning a fragment of the murine cyrόl cDNA containing nucleotide 456 through nucleotide 951 (encoding Cyrόl amino acids 93-379) into the Sacl site of the pATHl vector. Dieckman et al, J. Biol. Chem. 2(50: 1513-1520 (1985). The recombinant construct was transformed into a bacterial host, e.g., E. coli KI2, and expression of the fusion protein was induced by addition of 25 μg/ml indoleacrylic acid to growing cultures. Subsequently, cells were lysed and total cell lysate was fractionated by electrophoresis on a 7.5% polyacrylamide gel. The fusion protein of predicted size was the only band induced by indoleacrylic acid; that band was eluted from the gel and used as an antigen to immunize New Zealand White rabbits (Langshaw Farms) using techniques that are standard in the art. Harlow et al. In addition to polyclonal antibodies, the invention comprehends monoclonal antibodies directed to such fusion proteins.

In other embodiments of the invention, recombinant antibody products are used. For example, chimeric antibody products, "humanized" antibody products, and CDR-grafted antibody products are within the scope of the invention. Kashmiri etal, Hybridoma 14:461-413 (1995), incoiporated herein by reference. Also contemplated by the invention are antibody fragments. The antibody products include the aforementioned types of antibody products used as isolated antibodies or as antibodies attached to labels. Labels can be signal-generating enzymes, antigens, other antibodies, lectins, carbohydrates, biotin, avidin, radioisotopes, toxins, heavy metals, and other compositions known in the art; attachment techniques are also well known in the art.

Anti-Cyrόl antibodies are useful in diagnosing the risk of thrombosis, as explained more fully in Example 20 below. In addition, anti-Cyrόl antibodies are used in therapies designed to prevent or relieve undesirable clotting attributable to abnormal levels of Cyrόl . Further, antibodies according to the invention can be attached to toxins such as ricin using techniques well known in the art. These antibody products according to the invention are useful in delivering specifically- targeted cytotoxins to cells expressing Cyrόl, e.g., cells participating in the neovascularization of solid tumors. These antibodies are delivered by a variety of administrative routes, in pharmaceutical compositions comprising carriers or diluents, as would be understood by one of skill in the art.

Antibodies specifically recognizing Fisp 12 have also been elicited using a fusion protein. The antigen used to raise anti-Fispl 2 antibodies linked glutathione-S-transferase (GST) to the central portion of Fisp 12 (GST-Fispl2), where there is no sequence similarity to Cyrόl (O'Brien and Lau, 1992). A construct containing cDNA encoding amino acids 165 to 200 of Fisp 12 was fused to the glutathione-S-transferase (GST) coding sequence. This was done by using polymerase chain reaction (PCR) to direct synthesis of a fragment of DNA encompassing that fragment of fispl 2 flanked by a 5' BamHI restriction site and a 3' EcoRI restriction site. The 5' primer has the sequence 5'-GGGGATCTGTGACGAGCCCAAGGAC-3' (SΕQ ID NO:9) and the 3 ^' primer has the sequence 5 '-GGGAATTCGACCAGGCAGTTGGCTCG-3XSΕQ ID NO: 10). For Cyrό 1 -specific antiserum, a construct fusing the central portion of Cyrόl (amino acids 163 to 229), which contains no sequence similarity to Fisp 12, to GST was made in the same manner using the 5' primer 5 '-GGGGATCCTGTGATGAAGACAGCATT-3' (SEQ ID NOT 1) and the 3' primer 5'-GGGAATTCAACGATGCATTTCTGGCC-3' (SEQ ID NO: 12). These were directionally cloned into pGEX2T vector (Pharmacia-LKB, Inc.) and the clones confirmed by sequence analysis. The GST-fusion protein was isolated on glutathione sepharose 4B (Pharmacia-LKB, Inc.) according to manufacturer's instructions, and used to immunize New Zealand white rabbits. For affinity purifications, antisera were first passed through a GST-protein affinity column to remove antibodies raised against GST, then through a GST-Fispl 2 or GST- Cyrόl protein affinity column to isolate anti-Fispl 2 or anti-Cyrό l antibodies (Harlow et al. , [ \ 988]). These antibodies immunoprecipitated the correct size Fisp 12 protein product synthesized in vitro directed by fispl 2 mRNA. The antibodies are specific for the Fisp 12 polypeptide and show no cross-reactivity with Cyrόl .

Polyclonal antibodies recognizing CTGF are also known. U.S. Patent No. 5,408,040, column 7, line 41 , to column 9, line 63, incoφorated by reference hereinabove, reveals an immunological cross-reactivity between PDGF and CTGF, as described above.

Example 12 Inhibitory peptides Another embodiment of the present invention involves the use of inhibitory peptides in therapeutic strategies designed to inhibit the activity of the Cyrόl protein. One approach is to synthesize an inhibitory peptide based on the protein sequence of Cyrόl . For example, a peptide comprising an amino acid sequence that is conserved between murine Cyrόl (SEQ ID NO:2) and human Cyrόl (SEQ ID NO:4) competes with native Cyrόl for its binding sites. This competition thereby inhibits the action of native Cyrόl . For example, administration of an inhibitory peptide by well-known routes inhibits the capacity of Cyrόl to influence the cascade of events resulting in blood clots, the vascularization of tumors, or the abnormal vascularization of the eye (e.g., eye disorders characterized by vascularization of the retina or the vitreous humor), etc. In particular, an inhibitory peptide prevents Cyrό 1 from inhibiting the action of Tissue Factor Pathway Inhibitor, or TFPI, as described below.

In an embodiment of the invention, inhibitory peptides were designed to compete with Cyrόl . These inhibitory peptides, like the antibodies of the preceding Example, exemplify modulators of Cyrόl activity, as described in the context of a variety of assays for Cyrόl activity that are disclosed herein. The peptide design was guided by sequence comparisons among murine Cyrόl , Fispl 2, and Nov (an avian proto-oncogene). The amino acid sequences of several members of this family are compared in FIGURE 1. These types of sequence comparisons provide a basis for a rational design for a variety of inhibitory peptides. Some of these designed peptides, for example peptides spanning amino acids 48-68 (SEQ ID NO: 13), 1 15-135 (SEQ ID NO:14), 227-250 (SEQ ID NO: 15), 245-270 (SEQ ID NO: 16), and 310-330 (SEQ ID NO:17) of SEQ ID NO:2, have been synthesized. A comparison of the murine Cyrόl amino acid sequence and the human Cyrόl amino acid sequence reveals that similar domains from the human protein may be used in the design of peptides inhibiting human Cyrόl . In addition, sequence comparisons may involve the human Cyrόl amino acid sequence; comparisons may also include the human homolog of Fispl 2, Connective Tissue Growth Factor, also identified as a member of this protein family. O'Brien et al. (1992).

Inhibitory peptides may also be designed to compete with other ECM signaling molecules, e.g., Fisp 12 or CTGF, for binding to their respective receptors. The design of inhibiting peptides is facilitated by the similarity in amino acid sequences among the ECM signaling molecules. In addition, inhibitory peptide design may be guided by one or more of the methods known in the art for identifying amino acid sequences likely to comprise functional domains (e.g., hydrophilic amino acid sequences as external/surface protein domains; sequences compatible with -helical formation as membrane-spanning domains). These methods have been implemented in the form of commercially available software, known to those of ordinary skill in the art. See e.g., the Intelligenetics Suite of Analytical Programs for Biomolecules. Intelligenetics, Inc., Mountain View, CA. Using these approaches, inhibitory peptides interfering with the biological activity of an ECM signaling molecule such as Cyrόl, Fispl 2 or CTGF, may be designed. With the design of the amino acid sequence of an inhibitory peptide, production of that peptide may be realized by a variety of wel 1- known techniques including, but not limited to, recombinant production and chemical synthesis. Exemplary peptides that have been shown to specifically inhibit at least one biological activity of Cyrό l include peptides exhibiting the "RGD" motif, or motif variants such as "RGDS," "RGDSPK," "GDR," or "SGDR," (Ruoslahti, et al, Science 238:491 -497 [ 1987], Ruoslahti, E., Ann. Rev. of Cell and Dev. Biol. 72:698-715 [1996]) as described in Example 10 above.

Example 13 Cell Adhesion Another embodiment of the invention is directed to the use of Cyrόl to mediate cellular attachment to the extracellular matrix. Induction of cellular adhesion was investigated using murine Cyrόl , fibronectin, and bovine serum albumin (BSA). Immunological 96-well plates (Falcon brand) were coated with

50 μl of 0.1% BSA in PBS at 4°C in the presence of 0-30 μg/ml concentrations of murine Cyrόl or fibronectin. After two hours exposure to the coating solution, non-diluted immune or pre-immune antisera (30 μl/well), or affinity-purified anti-Cyrόl antibodies were added. For some wells, the coating mixture was adjusted to 10 mM DTT or 100 mM HCl. After 16 hours incubation, the coating solution was removed and the well surface was blocked with 1% BSA in phosphate-buffered saline (PBS) for 1 hour at room temperature. HUVE cells were plated in Ham's complete F12K medium [GIBCO-BRL, Inc.; Ham, Proc. Natl. Acad. Sci. (USA) 53:786 (1965)] at 5 x lθMθ⁴ cells/well. Cycloheximide was added to 100 μg/ml immediately before plating and monensin was added to 1 μM 14 hours before plating. After a 2-hour incubation at 37°C, the wells were washed with PBS and attached cells were fixed and stained with methylene blue. The attachment efficiency was determined by quantitative dye extraction and measurement of the extract absorbance at 650 nm. Oliver et al, J. Cell. Sci. 92:513-518 (1989).

HUVE cells attached poorly to dishes treated with BSA alone, but adhered well to dishes coated with fibronectin. Murine Cyrόl-coated surfaces also supported HUVE cell attachment in a dose-dependent manner, similar to fibronectin. For example, at 1 μg/ml, Cyrόl and fibronectin yielded A₆₅₀ values of 0.1 . An A₆₅₀ value of 0.5 corresponded to the attachment of 6 x 10³ cells. At the other end of the tested concentration range, 30 μg/ml, Cyrόl yielded an A₀₅₀ of 0.8: fibronectin yielded an A₆₅₀ of 0.9. Cyrόl also promoted the attachment of NIH 3T3 cells, though less effectively than fibronectin. Cyrόl -mediated cell attachment can be observed as early as 30 minutes after plating, as visualized by light microscopy. The adhesion of HUVE cells on murine Cyrόl -coated surfaces was specifically inhibited by anti-Cyrόl antiserum and by affinity-purified anti-Cyrόl antibodies, but not by pre-immune serum. In contrast, attachment of cells to fibronectin-coated dishes was not affected by either the anti-Cyrόl antiserum or affinity-purified anti-Cyrόl antibodies. These results show that enhancement of cell adhesion is a specific activity of the Cyrόl protein. Furthermore, the Cyrόl -mediated cell attachment was insensitive to cycloheximide or monensin treatment, indicating that Cyrόl does not act by inducing de novo synthesis of ECM components, stimulation of fibronectin, or collagen secretion. Rather, the data support the direct action of Cyrόl on cells in effecting adhesion. The Cyrόl -mediated attachment of HUVE cells was completely abolished by the presence of EGTA; however, attachment was restored by the addition of CaCl₂ or MgSO, to the medium. These results indicate that the interaction between Cyrόl and its cell surface receptor requires divalent cations, consistent with the observations leading to the identification of the α. β₃ integrin as the Cyrόl receptor described in Example 10, above.

The ability of Cyrό 1 to promote cell adhesion, and the ability of molecules such as anti-Cyrόl antibodies to inhibit that process is exploited in an assay for modulators of cell adhesion. The assay involves a comparison of cell adhesion to surfaces, e.g., plastic tissue culture wells, that are coated with Cyrόl and a suspected modulator of cell adhesion. As a control, a similar surface is coated with Cyrόl alone. Following contact with suitable cells, the cells adhering to the surfaces are measured. A relative increase in cell adhesion in the presence of the suspected modulator, relative to the level of cell adherence to a Cyrόl -coated surface, identi fies a promoter of cell adhesion. A relative decrease in cell adhesion in the presence of the suspected modulator identifies an inhibitor of cell adhesion.

The identification of a Cyrό l receptor led to the development of a rapid and specific ligand-receptor assay (i.e., integrin binding assay) for Cyrόl . Monoclonal antibody LM609 (anti-α β₃) has been described. Cheresh, 1987. Monoclonal antibody JBS5 (anti-fibronectin antibody) was purchased from Chemicon. Anti-human and anti-bovine vitronectin antisera were from Gibco BRL. HRP-conjugated goat anti-rabbit antibody was from KPL. RGDSPK peptide was from Gibco BRL; RGDS and SDGR peptides were from American Peptide Company. The peptides for functional assays were dissolved in PBS at 10 mg/ml and the pH was adjusted to 7.5-8.0 with NaOH. Human plasma vitronectin was from Collaborative Biomedical Products. _vβ₃ integrin purification from HUVE cell lysates was done as described in Pytela et al, Meth. Enzymol, 144:415-489 ( 1987). Briefly, 10⁸ cells were lysed in 1 ml of PBS containing 1 M CaCl₂ , 1 mM MgCl, , 0.5 mM PMSF and 100 mM octylglucoside. The lysate was passed four times through a 0.5 ml column containing RGDSPK Sepharose (prepared from the cyanogen bromide activated Sepharose CL 4B as described in Lam, S. C.-T, J. Biol. Chem., 2(57:5649-5655 ( 1992). The column was washed with 10 ml of the lysis buffer and the bound protein was eluted with 2 ml of the same buffer containing 1 mM RGDS peptide at room temperature. The α_vβ₃ integrin was dialyzed against PBS containing 1 mM CaCl₂, 1 mM MgCl₂, 5 mM octylglucoside and 0.1 mM PMSF with three changes of the dialysis buffer to remove the RGDS peptide. The protein was stored in aliquots at -70°C. The purity of the integrin was determined by SDS-PAGE under non-reducing conditions, followed by silver staining. Western blotting with anti-CD47 antibody showed that this cc β, integrin preparation does not contain any integrin-associated proteins.

The integrin binding assay was developed in accordance with the disclosures in Brooks et al, Cell 55:683-693 ( 1996), and Lam, S. C.-T. ( 1992). Approximately 50 ng of the integrin in a total volume of 50 μl were added per well of 96-well immunological Pro-Bind plates (Falcon) and incubated overnight at 4°C. Non-specific sites were blocked with 20 mg/ml BSA in the same buffer and washed four times in that buffer. Treated plates were incubated with 1 μg/ml Cyrόl or 0.1 μg/ml vitronectin for 3 hours at room temperature. EDTA (5 mM), RGDS peptide (0.5 mM) and blocking antibodies were either preincubated with the immobilized integrin for 1 hour before the addition of the protein ligand or added along with the ligand. The final dilution of the LM609 ascites fluid was 1 :200. Bound proteins were detected by specific polyclonal antisera (anti-Cyrόl antiserum was diluted 1 :500 and anti -vitronectin antiserum was diluted 1 : 1000 in PBS containing 1 mM CaCl₂, 1 mM MgCl₂, and 5 mg/ml BSA) followed by a secondary antibody-horseradish peroxidase conjugate (1 :20000 in the same buffer). Plates were rinsed four times with PBS containing 1 mM CaCl₂ and 1 mM MgCl₂ after each incubation. Horseradish peroxidase (HRP) was detected with an HRP immunoassay kit (Bio-Rad Laboratories). The colorimetric reaction was developed for 15-30 minutes at room temperature, stopped by the addition of H₂SO₄, and the absorbance at 450 nm was measured. Those of ordinary skill in the art will understand that a variety of detection techniques could be employed in place of the enzyme-linked immunological approach exemplified. For example, other labels such as radiolabels, fluorescent compounds and the like could be bound, e.g., covalently, to an antibody or other agent recognizing the peptide of interest such as Cyrόl .

The results of integrin binding assays showed that vitronectin and Cyrόl bound to the immobilized integrin. Further, both Cyrόl and vitronectin binding to α_vβ₃ were saturable. The concentration of Cyrόl at which saturation was reached was significantlyhigherthan the concentration of vitronectin required for saturation. This difference may reflect a lower affinity of α β₃ for Cyrόl compared to vitronectin, which is in agreement with the results of cell adhesion assays, which show that HUVE cells adhere to vitronectin and, more weakly, to Cyrόl , in a concentration-dependent manner (see below). The specificity of the interaction was addressed by blocking the ligand binding site of the integrin using any one of several techniques, including divalent cation deprivation, RGDS peptide competition, and LM609 antibody inhibition. The interaction of both proteins (Cyrόl and vitronectin) with α_vβ₃ was inhibited by EDTA, the RGDS peptide, and the LM609 antibody. These properties of the Cyrόl interaction with α_vβ₃ were also in agreement with the results of the cell adhesion assay and indicated that HUVE cell adhesion to Cyrόl was mediated by the direct interaction of Cyrόl with the α_vβ₃ integrin.

In addition, Cyrόl induces focal adhesion, i.e., cell surface foci for cytoskeletal attachments. Focal adhesion is effected by cell surface protein complexes or clusters. These protein clusters are complex, including a variety of receptors from the integrin family, and a variety of protein kinases. The induction of focal adhesion by Cyrό l is reflected in the capacity of Cyrόl to induce particular members of these cell surface protein clusters. For example, Cyrόl induces the phosphorylation of Focal Adhesion Kinase, a 125 kDa polypeptide, and Paxillin, another protein known to be involved in the focal adhesion cell surface protein complexes. Moreover, indirect immunofluorescence studies have shown that Cyrόl is bound to a receptor (see above) in focal adhesive plaques. The plaques, in turn, are characteristic of focal adhesion protein complexes. Focal Adhesion Kinase, Paxillin, and α_xβ₃ Integrin are co-localized to the focal adhesion plaques produced by focal adhesion complex formation induced by Cyrόl . These focal adhesion protein complexes bind Cyrόl at the cell surface; the complexes also attach internally to the cytoskeleton. Therefore, murine Cyrόl , and human Cyrόl (see below), are, in part, adhesion molecules, a characteristic distinguishing Cyrόl from conventional growth factors. Those of skill in the art will also recognize that the α_vβ₃ integrin can be used, in conjunction with Cyrόl , to screen for modulators of Cyrό I bindingto its receptor. In one embodiment, the integrin is immobilized and exposed to either (a) Cyrό l and a suspected modulator of receptorbinding; or (b) Cyrόl alone. Subsequently, bound Cyrό l is detected, e.g., by anti-Cyrό l antibody that is labeled using techniques known in the art, such as radiolabelling, fluorescent labeling, or the use of enzymes catalyzing colorimetric reactions. A promoter of Cyrόl binding to its receptor would increase binding of Cyrόl (and an inhibitor would decrease Cyrόl ), relative to the binding by Cyrόl alone.

In another embodiment of the invention, the effect of murine Cyrόl on cell moφhogenesis was assessed by a cell spreading assay. Polystyrene Petri dishes were coated with 2 ml of a 10 μg/ml solution of Cyrόl or fibronectin in PBS with 0.1 % BSA and treated as described above. A third plate was treated with BSA and served as a control. Each dish received 7 x 10⁶ cells and was incubated for 2 hours. Cell spreading was analyzed by microscopy at 100-fold magnification. The results indicate that murine Cyrόl induces HUVE cell spreading to approximately the same extent as fibronectin. The efficient attachment (see above) and spreading of cells on murine Cyrόl -coated substrates indicated that Cyrόl may interact with a signal-transducing cell surface receptor, leading to a cascade of cytoskeletal rearrangements and possible fonuation of focal contacts. Consequently, Cyrόl and Cyrόl -related polypeptides may prove useful in controlling cell adhesion, e.g., the cell adhesion events that accompany metastasizing cancer cells, organ repair and regeneration, or chondrocyte colonization of prosthetic implants, discussed below.

In contrast to mouse Cyrόl which mediated both HUVE cell attachment and migration, hCyrόl was found to mediate cell adhesion but not spreading of HUVE cells. Immunological plates (96-well ProBind assay plates, Falcon) were coated with 0.1 -30 μg/ml hCyrόl , fibronectin (Gibco BRL) or vitronectin (Gibco BRL) in phosphate-buffered saline (PBS) containing 0.1 %> protease-free BSA (Sigma) for 16 hrs at 4°C. The wells were blocked with 1 % BSA in PBS for 1 hr at room temperature and washed with PBS. HUVE cells were harvested with 0.02% EDTA in PBS, washed twice with serum-free F12 medium and resuspended in serum-free F12. In some experiments, fbs was added to 5-10%. Also, in experiments involving vitronectin-coated plates, endogenous vitronectin was removed from fbs by immunoaffinity chromatography using bovine polyclonal anti-vitronectin antibodies (Gibco). Norris et al, J. Cell Sci. 95:255- 262 ( 1990). Cells were plated at 10⁴ cells/well. After 2 hours, cells were fixed with 4% paraformaldehyde, stained with methylene blue and quantified as described. Oliver et al, J. Cell Sci. 92:513-518 (1989).

Under serum-free conditions, hCyrόl mediated cell attachment but not spreading of HUVE cells. Attachment of HUVE cells to hCyrόl -coated plates was enhanced by inclusion of serum in the culture medium. In the presence of serum, HUVE cells attached and spread on hCyrόl in a manner similar to that seen on fibronectin. Human Cyrόl supported HUVE cell adhesion in a dose- dependent manner both under high-serum (10%) and low-serum (0.5%) conditions. However, in the presence of 10% fbs, the maximal proportion of the cells attaching at a lower concentration of hCyrόl , and the proportion of the cells attached, was higher. Human Cyrόl was also found to cooperate with vitronectin in promoting HUVE cell adhesion and spreading. Two major cell-adhesive proteins found in mammalian sera are fibronectin and vitronectin, also known as "serum spreading factor." For review, see Felding-Habermann et al. , Curr. Opin. Cell Biol. 5:864-868 (1993). Cell attachment, spreading and growth on tissue- culture plastic depended upon vitronectin, rather than fibronectin, in serum for the following reasons: (1 ) considerable depletion of fibronectin in the batches of fbs due to "clotting" at 4°C; and (2) inability of fibronectin to efficientlv coat the plastic in the presence of an excess amount of other serum proteins. In contrast, vitronectin coated the plastic surfaces efficiently under the same conditions.

The ability of HUVE cells to adhere to hCyrόl -coated plates in the presence of mock-immunodepleted fbs and serum immunodepleted with anti- bovine vitronectin antibodies were compared. HUVE cells adhered to hCyrόl - coated surfaces significantly better in the presence of soluble vitronectin or mock- immunodepleted fbs than they did in the presence of serum-free medium or medium supplemented with vitronectin-immunodepleted fbs. The addition of vitronectin (30 μg/ml) to vitronectin-immunodepleted serum restored the ability of HUVE cells to adhere and spread on hCyrόl -coated plates to the same level observed when whole serum was used in the cell attachment assay. Furthermore, soluble vitronectin alone, at a concentration equal to its level in 10% serum (30 μg/ml), restored the level of cell adhesion and spreading to the level found in the presence of 10% serum. Thus, vitronectin is a necessary and sufficient serum component contributing to HUVE cell adhesion and spreading on hCyrόl -coated plastic surfaces. Control studies showed that the effect of vitronectin was not due to its preferential retention on the plastic dish surfaces in the presence of hCYRόl .

Additionally, HUVE cell attachment and spreading in the presence of an increasing quantity of vitronectin was examined. The solutions for coating the dishes contained increasing amounts of vitronectin (0-10 μg/ml) with a fixed amount of hCyrόl (10 μg/ml). The results indicated that more cells adhered to plates coated with the two proteins than would have been expected by adding the individual adhesive capacities of vitronectin and hCyrόl . This non-additive increase of adhesion in the presence of vitronectin and hCyrό l was not due to higher amounts of vitronectin absorbed on the plastic. ELISA assay with anti- human vitronectin antibodies showed that the amount of vitronectin adsorbed to plastic dishes exposed to the vitronectin/hCyrόl mixture did not exceed that of vitronectin alone by more than 20%. This difference is insufficient to explain the observed difference in cell adhesion (3-5 fold in different experiments). In addition, a higher proportion of HUVE cells also adhered to the mixture of proteins when the coating solution contained diluted vitronectin (2.5 μg/ml) than were found to adhere to dishes coated with higher concentrations of pure vitronectin ( 10 μg/ml) or pure hCyrό 1 (10 μg/ml). Thus, vitronectin and hCyrό 1 functionally cooperate and exert a synergistic effect on HUVE cell adhesion. The capacity of Fispl 2 to affect cell adhesion was also investigated.

Fisp 12 cell attachment assays were performed essentially as described (Oliver et al. , [ 1989]). 96-well immunological plates were coated for 16 hours at 4°C with 20 μg/ml Cyrόl , Fisp 12 or fibronectin (Gibco BRL) in PBS containing 0.1 mg/ml BSA and blocked with 1 mg/ml BSA for 1 hour at room temperature. HUVE cells were plated at 10⁴ cells/well in F 12K media with 10% FBS (Hyclone Laboratories, Inc., Logan, Utah); NIH 3T3 fibroblasts were plated at 3 x 10⁴ cells/well and MvlLu cells were plated at 5 x 10⁴ cells/well in minimal essential medium (MEM) with 10%> FBS. After 1 hour incubation cells were fixed, stained with methylene blue and quantified as described (Oliver et al, [1989]). Cell spreading was examined on cells plated on 100 mm polystyrene petri dishes coated with 2.5 ml of a 20 μg/ml solution of Cyrόl, Fispl 2 or fibronectin. 10⁷ cells were plated on each dish and cell spreading was analyzed 90 minutes after plating by microscopy at 100X magnification.

The results indicated that Fisp 12, as well as Cyrόl , when coated on plastic dishes, promoted the attachment of three different cell types: HUVE cells, NIH 3T3 fibroblasts, and mink lung epithelial (MvlLu) cells. These cells attached poorly to uncoated plastic dishes or plastic dishes coated with bovine serum albumin, but attached significantly better to dishes coated with either fibronectin, Cyrόl , or Fispl 2. The ability of either Cyrόl or Fispl2 to mediate cell attachment is comparable to that of fibronectin for all three cell types. While the ability of Cyrόl to mediate cell attachment was previously demonstrated for fibroblasts and endothelial cells (Kireeva et al. Moi Cell Biol. 7(5:1326-1334 [1996]), these studies show cell attachment activity for both Fispl 2 and Cyrόl in epithelial cells in addition to endothelial cells and fibroblasts. Like cell attachment to fibronectin and Cyrόl (Kireeva et al, [1996]),

Fispl2-mediated cell attachment was inhibited when EDTA was added to the culture medium. This inhibition was completely abolished by the addition of excess MgCl₂, indicating a requirement for divalent cations in Fispl 2-mediated cell attachment. In addition to cell attachment, Fisp 12 also promotes cell spreading. Similar cell spreading was found when NIH 3T3 cells were plated on dishes coated with either fibronectin, Cyr61 , or Fispl 2, but not BSA. Endothelial and epithelial cells also spread when plated on fibronectin, Cyrόl, or Fispl2. Example 14 Migration of Fibroblasts

Cyrόl also affects chondrocytes, e.g., fibroblasts involved in skeletal development. In particular, Cyrόl influences the development, and perhaps maintenance, of cartilage, in contrast to the variety of growth-related proteins that exclusively influence development and maintenance of the bony skeleton. The chemotactic response of NIH 3T3 cells to murine Cyrόl was examined using a modified Boyden chamber (Neuroprobe Inc., catalog no. AP48). Grotendorst,

Meth. Enzymol. 747: 144-152 (1987). Purified Cyrόl protein was serially diluted in DMEM containing bovine serum albumin (BSA; 0.2 mg/ml) and added to the lower well of the chamber. The lower well was then covered with a collagen-coated polycarbonate filter (8 μm pore diameter; Nucleopore Coφ., Pleasanton, CA). Cells (6 x 10⁴) were then loaded into the upper well. After 5 hours incubation (10%> CO₂, 37°C), the filter was removed and the cells were fixed and stained using Wright-Giemsa stain (Harleco formulation; EM Diagnostic Systems, Gibbstown, NJ). Cells from the upper surface of the filter were then removed by wiping with a tissue swab. The chemotactic response was determined by counting the total number of migrating cells detected in ten randomly selected high-power microscopic fields (400-fold magnification) on the lower surface of the filter. Duplicate trials were performed for each experiment and the experiment was repeated three times to ensure reproducibility of the data. NIH 3T3 cells responded to Cyrόl as a chemotactic factor in a dose- dependent manner in the Boyden chamber assay. Without Cyrόl , approximately 4.8 cells had migrated per high-power field. In the presence of 0.5 μg/ml murine Cyrόl , about 5.2 cells were found in each field. As the concentration of Cyrόl was raised to 1 , 5 and 10 μg/ml, the average number of migrating cells detected per field rose to 7.5, 18.5, and 18.7. Thus, murine Cyrό l acts as a chemoattractant for fibroblasts. The optimal concentration for the chemotactic activity of Cyrό 1 is 1 -5 μg/inl in this assay; this concentration range is consistent with the reported ranges at which other ECM molecules provide effective chemotactic stimulation. For example, Thrombospondin, at 5-50 μg/ml, has a chemotactic effect on endothelial cells (Taraboletti et al, J. Cell Biol. 117:765- 772 (1990); fibronectin also functions as a chemotactic agent at 1 -30 μg/ml (Car sons et al, Role of Fibronectin in Rheumatic Diseases, in Fibronectin [Mosher, ed., Academic Press 1989]; Carsons et al, Arthritis. Rheum. 25:601- 612 [1985]) as determined using similar Boyden chamber assays. The human Cyrό 1 polypeptide maybe used to cheiuoattract fibroblasts in a manner analogous to murine Cyrόl . In experiments analogous to the studies of NIH 3T3 cell migration in response to murine Cyrόl, the migration of 1064SK human skin fibroblasts in response to wild-type human Cyrό 1 , and human Cyrό 1 NT (a Cyrό 1 polypeptide fragment described in greater detail in Example 29), was determined. Cyrό 1 induced migration of the human fibroblast cells at comparable levels to the murine Cyrόl levels inducing NIH3T3 cell chemotaxis. The human Cyrόl NT induced 1064SK cell migration at levels comparable to the effective levels of wild-type human Cyrόl. Antibody studies showed that an antibody specific for the _v integrin subunit blocked fibroblast migration in the presence of Cyrόl polypeptides, while an antibody (i.e., GoH3) specific for the α₆ integrin subunit did not block migration. Further investigation revealed that a monoclonal antibody specifically recognizing the _vβ₅ integrin inhibited Cyrόl -induced human fibroblast migration, but a monoclonal antibody specifically recognizing the integrin α_vβ₃ did not inhibit that migration. Thus, integrin _vβ₅ mediates the fibroblast migration response to Cyrόl polypeptides. These results contrast with the results of antibody studies analyzing the chemotactic response of endothelial cells to Cyrόl polypeptides, which is dependent on integrin α_vβ₃. Human CTGF has also been reported to induce the migration of non-human mammalian cells such as NIH 3T3 eel Is (mouse fibroblasts) and B ASM cells (bovine aortic smooth muscle cells), as described in U.S. Patent No. 5,408,040, column 7, line 65 to column 1 1 , line 7, incoφorated herein by reference.

Thus, one aspect of the invention is a method of screening for modulators of cell migration, such as fibroblast cell migration. In general, the method involves exposing cells that preferably present at least one suitable integrin receptor to a Cyrόl polypeptide in the presence or absence of a potential, or suspected, modulator. Subsequent measurement of relative cell migration rates (in the presence and absence of the potential modulator) identifies modulators of Cyrόl -induced cell migration. A variety of small chemicals, both inorganic and organic, as well as various peptides, are expected to be potential modulators. Examples include mannose and its derivative, mannose-6-phosphate, with the latter expected to function as a modulator of Cyrόl -induced cell migration by inhibition. The potential modulators screened by methods of the invention can have any of a wide variety of structures and may be tested individually or as part of a more systematic approach using any of the chemical or peptide libraries known in the art.

In an alternative embodiment, an assay for modulators of cell migration, such as the migration of chondrocytes, involves a combination of a suspected modulator of cell migration and Cyrό l being added to the lower well of a Boyden chamber. As a control, Cyrόl is separately added to the lower well of another Boyden chamber. Relative cell migrations are then measured. An increase in cell migration in the presence of the suspected modulator relative to cell migration in response to Cyrό l alone identifies a promoter of chondrocyte cell migration, while a relative decrease in cell migration in the presence of the suspected modulator identifies an inhibitor.

Example 15 Migration of Endothelial Cells- In Vitro Assays

The end product of (// vitro angiogenesis is a well-defined network of capillary-like tubes. When cultured on gel matrices, e.g., collagen, fibrin, or

Matπgel gels, endothelial cells must first invade the matrix before forming mature vessels. (Matrigel is a complex mixture of basement membrane proteins including laminin, collagen type IV, nidogen/entactin, and heparan sulfate proteoglycan, with additional growth factors. Kleinman et al, Biochem. 25:312- 318 ( 1986). The invasive structures are cords which eventually anastomose to form the vessel-like structures. The angiogenic effect of human Cyrόl on confluent monolayers of human umbilical vein endothelial cells is assessed by seeding the cells onto three-dimensional collagen or fibrin gels, in the presence or absence of Cyrόl . HUVE cells do not spontaneously invade such gels but do so when induced by agents such as tumor promoters.

Collagen gels were prepared by first solubihzing type I collagen (Collaborative Research, Inc., Bedford, MA) in a sterile 1 : 1000 (v/v) dilution of glacial acetic acid (300 ml per gram of collagen). The resulting solution was filtered through sterile triple gauze and centrifuged at 16,000 x g for 1 hour at 4°C. The supernatant was dialyzed against 0.1X Eagle's Minimal Essential Medium (MEM; GIBCO-BRL, Inc.) and stored at 4°C. Gels of reconstituted collagen fibers were prepared by rapidly raising the pH and ionic strength of the collagen solution. The pH and ionic strength adjustments were accomplished by quickly mixing 7 volumes of cold collagen solution with one volume of 10X MEM and 2 volumes of sodium bicarbonate (1 1.76 mg/ml) in a sterile flask. The solution was kept on ice to prevent immediate gelation. The cold mixture was dispensed into 18 mm tissue culture wells and allowed to gel for 10 minutes at 37°C. Fibrin gels were prepared by dissolving fibrinogen (Sigma Chemical Co..

St. Louis, MO) immediately before use in calcium-free MEM to obtain a final concentration of 2.5 mg of protein/ml. Clotting was initiated by rapidly mixing 1.35 ml of fibrinogen solution with 15 μl of 10X MEM containing 25 U/ml thrombin (Sigma Chemical Co.) in a plastic tube. The mixture was transferred immediately into 18 mm tissue culture wells and allowed to gel for about 2 minutes at 37°C.

In some wells, Cyrόl was mixed into the gel matrix before gelation (final concentration 10 μg/ml), while in other wells, Cyrόl was not in the gel matrix but was added as part of the nutrient medium (similar range of concentrations as in the matrix) after the cells reached confluency. HUVE cells were seeded onto the gel matrix surface at 5 x 10⁴ cells per well in Ham's F12K medium [GIBCO-BRL, Inc.] containing 10% fetal bovine serum, 100 μg/ml heparin, and 30 μg/ml endothelial cell growth factor. When the cells reached confluency, the medium was removed, the cells were rinsed with PBS, and fresh medium without endothelial cell growth factor was supplied. Some cultures received purified recombinant Cyrόl, while others received Cyrόl and polyclonal anti-Cyrόl antibodies. Thus, the variety of cultures at confluency included: a) cultures with no Cyrόl ; b) cultures with Cyrόl within the matrix; c) cultures with Cyrόl supplementing the medium; and d) cultures with Cyrόl supplementing the medium along with polyclonal anti-Cyrόl antibodies.

Invasion of the gel matrix was quantified about 4-7 days after treatment of the confluent cultures. Randomly selected fields measuring 1.0 mm x 1.4 mm were photographed in each well under phase-contrast microscopy with a Zeiss Axiovert inverted photomicroscope. Photographs were taken at a single level beneath the surface monolayer. Invasion was quantified by measuring the total length of all cell cords that penetrated beneath the surface monolayer. Results were expressed as the mean length in microns per field for at least 3 randomly selected fields from each of at least 3 separate experiments.

In order to examine the network of cords within the matrix for capillary- like tube formation, cultures were fixed ... situ overnight with 2.5% glutaraldehyde and 1 % tannic acid in 100 mM sodium cacodylate buffer, pH 7.4. Cultures were then washed extensively in 100 mM sodium cacodylate buffer, pH 7.4. The gels were cut into 2 mm x 2 mm fragments, post-fixed in 1% osmium tetroxide in veronal acetate buffer (to minimize tissue swelling; see Hayat, in Principles and Techniques of Electron Microscopy: Biological Applications 1 :38 [Litton Educational Publishing, Inc. 1970]) for 45 minutes, stained en bloc with 0.5% uranyl acetate in veronal buffer for 45 minutes, dehydrated by exposure to a graded ethanol series, and embedded in Epon in flat molds. Semi-thin sections were cut peφendicular to the culture plane with an ultramicrotome, stained with 1% toluidine blue, and photographed under transmitted light using an Axiophot photomicroscope (Zeiss).

In an alternative embodiment, a suspected modulator of angiogenesis is combined with Cyrόl and the combination is added before, or after, formation of a gel. In this embodiment, a control is established by using Cyrόl alone. The migration of cells in response to the suspected modulator and Cyrόl is then compared to the migration of cells in response to Cyrόl alone. A promoter or positive effector will increase cell migration while an inhibitor or negative effector will decrease cell migration. In an alternative in vitro assay for angiogenic activity, an assay for endothelial cell migration was developed. This chemotaxis assay has been shown to detect the effects of Cyrόl concentrations on the order of nanograms per milliliter. Primary Human Microvascular Endothelial Cells (HMVEC PO51, Clonetics, San Diego, CA) were maintained in DME with 1 % donor calf serum (Flow Laboratories, McLean, VA) and 100 μg/ml endothelial celi mitogen (Biomedical Technologies Inc., Stoughton, MA). The cells were used between passages 10 and 15. To measure migration, cells were starved for 24 hours in DME containing 0.1% BSA, harvested, resuspended in DME with 0.1% BSA, and plated at 1.75 x 10⁴ cells/well on the lower surface of a gelatinized 0.5 μm filter (Nucleopore Coφoration, Pleasanton, CA) in an inverted modified Boyden chamber. After 1 -2 hours at 37°C, during which time the cells were allowed to adhere to the filter, the chamber was reverted to its normal position. To the top well of separate chambers, basic Fibroblast Growth Factor (a positive control), Cyrόl , or a negative control solution (conditioned medium known to lack chemoattractants or DME plus BSA, see below) was added at concentrations ranging from 10 ng/ml to 1 μg/ml. Chambers were then incubated for 3-4 hours at 37°C to allow migration. Chambers were disassembled, membranes fixed and stained, and the number of cells that had migrated to the top of the filter in 3 high-powered fields was determined. Tolsma et al. J. Cell. Biol 722:497-51 1 ( 1993) (incoiporated by reference), and references cited therein. DME with 0.1 % BSA was used as a negative control and either bFGF (10 ng/ml) or conditioned media from angiogenic hamster cell lines (20 μg/ml total protein) were used as positive controls. Rastinejad et al, Cell 5(5:345-355 (1989). Each sample was tested in quadruplicate (test compound such as Cyrόl, positive control, conditioned medium as a negative control, and DME plus BSA as a negative control) in a single experiment; experiments were repeated at least twice.

To allow comparison of experiments performed on different days, migration data is reported as the percent of maximum migration towards the positive control, calculated after subtraction of background migration observed in the presence of DME plus BSA. Test compounds that depressed the random movement of endothelial cells showed a negative value for the percent migration. Very high concentrations of thrombospondin (TSP) caused endothelial cells to detach from the membrane. Detachment was detected by counting cells on the lower face of the membrane. When cell loss exceeded 10%, the number of migrated cells was corrected for this loss. The results indicate that 0.01 -10 μg/ml bFGF induced the migration of a constant 92 cells per three high-powered microscope fields Migration in the presence of Cyrόl revealed a greater dependence on concentration. At 10 ng/ml, Cyrόl induced an average of 64 cells to migrate per three high-powered fields examined. At 100 ng/ml Cyrόl , approximately 72 cells were found in three fields; at 1 μg/ml Cyrόl , a peak of 87 cells had migrated; at approximately 7 μg/ml Cyrόl , about 61 cells were observed; and at 10 μg/ml Cyrόl, approximately 57 cells were found to have migrated. The negative control revealed a constant basal level of endothelial cell migration of 53 cells per three high-powered microscope fields. In addition to these results, there is a perfect correlation of the results from this in vitro assay and the results from the in vivo cornea assay, described below.

To monitor toxicity. endothelial cells were treated with each of the tested compounds at a range of concentrations, under conditions identical to those used in the migration assay. Cells were then stained with Trypan blue and cells excluding Trypan blue were counted. The results showed that cells remained viable and that the inhibition of migration could not be attributed to toxicity. Where relevant, endothelial cells were pretreated for 36-48 hours with peptides at 20 μM in DME with 0.1% BSA before use in the migration assays. Toxicity was also tested over these time frames and found to be negligible. The ability of Cyrό l to induce matrix invasion and tube formation by

HUVE cells, as well as the ability of Cyrόl to induce human microvascular endothelial cells to migrate, demonstrates the angiogenic properties of this protein. It is anticipated that other members of the ECM signaling molecule family of cysteine-rich proteins, such as Fisp 12 and CTGF, have similar properties that may be used in methods of the invention for screening for, and modulating, angiogenic conditions. In particular, one of ordinary skill in the art understands that an in vitro assay for angiogenic inhibitors involves the assay described above, including an effective amount of Cyrόl, with and without the candidate inhibitor.

Example 16

Migration of En doth elial Cells- A n In Vitro A ssay For A ngiogenesis In ft ibitors

The inclusion of an effective amount of an ECM signaling molecule, such as Cyrόl , in the in vitro migration (i.e., chemotaxis) assay described in the preceding Example, provides an assay designed to detect inhibitors of ECM signaling molecules and angiogenesis. Because of the crucial role of neovascularization in such processes as solid tumor growth and metastasis, the development of assays to detect compounds that might antagonize these processes would be useful.

The above-described in vitro migration assay was adapted to include an ECM signaling molecule, Cyrόl . Cyrόl was included at 1 μg/ml, which was found to be the optimal dosage in titration studies. As in the preceding Example, human microvascular endothelial cells (Clonetics) were used. In one series of assays, several carbohydrates and carbohydrate derivatives were analyzed. These compounds included 10 mM mannose, 10 mM mannose-6-phosphate, and 10 mM galactose. Results of these assays showed that Cyrόl plus mannose yielded approximately 73 cells per set of three high-powered microscope fields (see above). Cyrό 1 plus galactose induced the migration of approximately 74 cells per set of three high-powered fields. However, Cyrόl plus mannose-6-phosphate yielded approximately 2 migrating cells for each set of three high-powered fields examined. Control experiments demonstrate that the inhibition of Cyrόl activity by mannose-6-phosphate is specific.

The angiogenic activity of basic FGF ( 10 ng/ml) was also tested, as described above, with and without mannose-6-phosphate. In the presence of 10 mM mannose-6-phosphate, bFGF induced 51 cells per set of three high-powered fields to migrate; in its absence, bFGF induced the migration of approximately 52 cells. However, when either Cyrόl or Insulin Growth Factor II (IGF-II) were tested, mannose-6-phosphate reduced the number of migrating cells from approximately 48 or 47 cells, respectively, to approximately 12 or 11 cells, respectively. The effect of mannose-6-phosphate on IGF II activity was anticipated because mannose-6-phosphate is known to compete with IGF II for their common receptor (the IGF II receptor). Thus, mannose-6-phosphate specifically inhibits the chemotactic activity of Cyrό 1 on human endothelial cells. Moreover, because there is an essentially perfect correlation between the in vitro migration assay and the in vivo angiogenesis assay, described below, mannose-6- phosphate has been identified as an inhibitor of angiogenesis based on the results of the assay disclosed herein. Accordingly, the invention contemplates a method of inhibiting angiogenesis comprising the step of administering an inhibitor the angiogenic activity of Cyrόl such as mannose-6-phosphate. Assays such as that described above may also be used to screen for other inhibitors of angiogenesis which may be useful in the treatment of diseases associated with angiogenesis such as cancer, and diseases of the eye which arc accompanied by neovascularization.

In an embodiment of the invention, a method of screening for modulators of angiogenesis involves a comparative assay. One set of conditions involves exposure of cells to a combination of Cyrό l and a suspected modulator of cell migration. As a control, a parallel assay is performed that exposes cells to Cyrόl alone. A promoter of cell migration elevates the rate of in vitro cell migration relative to the rate of migration in the presence of Cyrόl alone; the converse is true for an inhibitor of the chemoattracting ability of Cyrό 1.

Example 17 Migration of Endothelial Cells- An In Vivo Assay

An in vivo assay for endothelial cell migration has also been developed.

In general, the assay protocol is consistent with the disclosure of Tolsma et al, 1993. To assess angiogenesis associated with the formation of granulation tissue (i.e., the newly formed, proliferative, fibroblastic dermal tissue around wounds during healing), sponge implants were used as pieviously described (Fajardo, et al, Lab. Invest. 55:718-724 [1988]). Polyvinyl-alcohol foam discs ( 10-mm diam x 1-mrn thick) were prepared by first removing a 2-mm diameter central core of sponge. PBS or an RGDS peptide (other possible test compounds include fragments of Cyrόl , RGDS peptide, small molecules such as mannose-6- phosphate) at 100 μM were added to the sponge core which was then coated with 5 μl of sterile Hydron (Interferon Sciences, New Brunswick, NJ). After solidifying, the coated core was returned to the center of the sponge which was then covered on both sides with 5 μm filters and secured in place with glue (Millipore Coφ., Bedford, MA). One control and one test disc were then implanted subcutaneously in the lower abdomen of anesthetized Balb/c female mice where granulation tissue could invade the free perimeter of the disc. Wounds were closed with autoclips and animals left undisturbed until sacrificed. Quantitative estimates of thymidine incoφoration in situ into endothelial cells in the discs were obtained as previously described (Polverini, et al, J. Immunol. I /5:529-532 [ 1977]). Sponge implants were evaluated at days 5, 7, 10, and 14 after implantation. Thirty minutes before sacrifice, mice were injected with a solution containing [³H]-thymidine in saline (specific activity 6.7 Ci/mM; New England Nuclear/Du Pont, Wilmington, DE) to a level of 1 μCi per gram of body weight. Sponges were removed and facially embedded to provide a uniform section of the entire circumference. Tissues were fixed in 10% neutral buffered formalin, dehydrated through a graded series of alcohols, and embedded in glycol methacrylate (Polysciences, Miles, IL). Autoradiograms were prepared by dipping sections mounted on acid-cleaned glass slides into NTB type 2 emulsion (Eastman Kodak). After exposure for 4 weeks at 4°C, autoradiographs were developed in half strength D-19 developer, fixed in Kodak Rapid Fixer, and stained with hematoxylin and eosin. Quantitation of endothelial cell labeling was performed by counting all endothelial cells that lined capillaries and venules extending from the periphery to the center of the sponge by rectilinear scanning under oil immersion (x 1 ,000). A total of 500-700 endothelial cells were counted in each of two sponges containing either PBS, TSP, or peptide fragments (i.e., thrombospondin fragments). Cells were considered labeled i f five or more grains were detected over the nucleus. The percentage of labeled cells was calculated and a chi-square analysis of data derived from control and experimental sponges was performed.

The results of the foregoing assay established that thrombospondin fragments could inhibit the process of angiogenesis. More generally, one of ordinary skill in the art would appreciate that the scope of the present invention extends to such in vivo assays for suspected modulators of ECM signaling molecule activities, such as the chemotactic ability of Cyrόl to induce cell migration. As with other assays of the invention, a comparative assay involves exposure of cells, in vivo, to a sponge laden with Cyrόl in the presence or absence of a suspected modulator of Cyrόl activity. Following implantation, incubation, and removal, the relative rates of cell migration are determined. A promoter of Cyrόl activity will increase the rate of cell migration relative to cell migration induced by Cyrόl alone; an inhibitor will decrease the rate of cell migration relative to the level ascribable to Cyrόl alone. Example 18 Mitogen Potentiation

In another aspect of the invention, murine Cyrόl enhanced the mitogenic effect of growth factors on fibroblasts and endothelial cells. When NIH 3T3 fibroblasts or HUVE cells were treated with a non-saturating dose of either basic

Fibroblast Growth Factor (bFGF) or Platelet-Derived Growth Factor (PDGF-BB), the addition of murine Cyrό I or human Cyrόl polypeptides (wild-type Cyrόl and

Cyrό 1 NT), as well as Fisp I 2/CTGF, significantly increased the incorporation of radiolabeled thymidine compared to cells treated with the growth factors alone. The thymidine incoφoration assay is a standard technique for determining whether cells are actively growing by assessing the extent to which the cells have entered the S phase and are synthesizing DNA. The Cyrόl enhancement of bFGF- or PDGF-BB-induced thymidine incoφoration was dose-dependent, requiring a minimum concentration of 0.5 - 1.0 μg/ml of recombinant protein for either cell type. Cyrόl polypeptides are expected to have analogous enhancing or potentiating effects on the ability of other growth factors (e.g.,PDGF-BB) to induce the mitogenesis of endothelial cells, fibroblasts, and other mammalian cell types as measured, e.g., by thymidine incoφoration (i.e., DNA synthesis). The enhancement of DNA synthesis by Cyrόl was inhibited by the addition of specific anti-Cyrόl antiserum.

More specifically, NIH 3T3 fibroblast cells were plated on 24- well plates at 3 x 10⁴ cells/well and grown in DMEM with 10% fetal bovine serum (Intergen Co., Purchase, NY) for 3-4 days and incubated with medium containing 0.2% FBS forthe following 48 hours. The following compounds, at the parenthetically noted final concentrations, were then added to the plated cells in fresh DMEM containing 0.2% fbs and [³H]-thymidine (1 μCi/ml final concentration; ICN Biomedicals, Inc., Costa Mesa, CA): bFGF ( 15 ng/ml), PDGF-BB (30 ng/ml), and murine Cyrόl (0.5-5 μg/ml). These compounds were added to individual plates according to the following pattern: 1) no supplementation; 2) murine Cyrόl ; 3) bFGF; 4) murine Cyrόl and bFGF; 5) PDGF-BB; and 6) murine Cyrόl and PDGF. After 18-20 hours of incubation, ceils were washed with PBS and fixed with 10% trichloroacetic acid. DNA was dissolved in 0.1 N NaOH and thymidine incoφoration was determined. The results indicated that murine Cyrόl , in the absence of a growth factor, did not stimulate DNA synthesis as measured by tritiated thymidine incoφoration. Without any supplements, 3T3 cells incoφorated approximately 1.8 x 10⁴ cpm of[³H]-thymidine, in the presence or absence of Cyrόl . Cells exposed to bFGF alone incoφorated about 1.2 x 10⁵ cpm; cells contacting bFGF and murine Cyrόl incoφorated 2 x 10⁵ cpm. Cells receiving PDGF-BB incoφorated about 1.2 x 10⁵ cpm; and cells exposed to PDGF-BB and murine Cyrόl incoφorated approximately 2.4 x 10⁵ cpm. Therefore, murine Cyrόl did not function as a mitogen itself, but did potentiate the mitogenic activity of bFGF and PDGF-BB, two known growth factors.

The ability of murine Cyrόl to potentiate the mitogenic effect of different levels of bFGF also revealed a threshold requirement for the growth factor. Human umbilical vein endothelial cells were plated essentially as described above for 3T3 cells and exposed to a constant amount of murine Cyrόl ; controls received no Cyrόl . Different plates were then exposed to different levels of bFGF, comprising a series of bFGF concentrations ranging from 0-10 ng/ml. Following culture growth in the presence of [³H]-thymidine for 72 hours, cells exposed to 0-0.1 ng/ml of bFGF exhibited a baseline level of thymidine incoφoration (approximately 4 x 10² cpm), in the presence or absence of Cyrόl . At 1 ng/ml bFGF, however, HUVE cells increased their thymidine incoφoration in the presence of bFGF to 6 x 10² cpm; in the presence of 1 ng/ml bFGF and murine Cyrόl, HUVE cells incorporated 1.3 x 10³ cpm. At 10 ng/ml bFGF, cells exposed to bFGF incoφorated about 1 .8 x 10³ cpm thymidine; cells receiving 10 ng/ml bFGF and Cyrόl incoφorated approximately 6.1 x 10³ cpm.

The capacity of murine Cyrόl to potentiate the mitogenic activity of bFGF was verified by a thymidine incoφoration assay involving HUVE cells and various combinations of bFGF, Cyrό l , and anti-Cyrόl antibodies. Cells were plated and grown as described above. The following combinations of supplements (final plate concentrations noted parenthetically) were then pre- incubated for 1 hour before addition to individual plates: 1) pre-immune antiserum (3%); 2) bFGF (15 ng/ml) and pre-immune antiserum (3%); 3) pre- immune antiserum (3%) and Cyrόl (4 μg/ml); 4) pre-immune antiserum (3%), Cyrόl (4 μg/ml), and bFGF (15 ng/ml); 5) anti-Cyrόl antiserum (3%); 6) anti- Cyrόl antiserum and bFGF (15 ng/ml); 7) anti-Cyrόl antiserum (3%) and Cyrόl (4 μg/ml); and 8) anti-Cyrόl antiserum (3%), Cyrόl (4 μg/ml), and bFGF (15 ng/ml).

Following incubation in the presence of [³H]-thymidine as described above, cells exposed to pre-immune antiserum incoφorated about 2 x 10² cpm thymidine; cells contacting pre-immune antiserum and bFGF incoφorated 1.3 x 10³ cpm; cells receiving pre-immune antiserum and Cyrόl incoφorated 1 x 10² cpm; cells receiving pre-immune antiserum, Cyrόl , and bFGF incoφorated 3.6 x 10³ cpm; cells exposed to anti-Cyrόl antiserum incoφorated 2 x 10² cpm; cells receiving anti-Cyrόl antiserum and bFGF incoφorated about 1.3 x 10³ cpm; cells contacting anti-Cyrόl antiserum and Cyr61 incoφorated about 1 x 10²; and cells receiving anti-Cyrόl antiserum, Cyrόl , and bFGF incoφorated 1 x 10³ cpm. These results indicate that pre-immune antiserum had no effect on Cyrό 1 -induced potentiation of bFGF mitogenic activity. Anti-Cyrόl antiserum, however, completely abolished the potentiation of bFGF by Cyrόl . Moreover, the effect of anti-Cyrόl antiserum was specific to Cyrόl -induced mitogenic potentiation because anti-Cyrό 1 antiserum had no effect on the mitogenic activity of bFGF per se. Therefore, Cyrόl can be used as a reagent to screen for useful mitogens.

Additional antibody studies using integrin-specific monoclonal antibodies showed that from an entire panel of specific anti-integrin antibodies, only an antibody (LM609) specific for integrin α_vβ₃ inhibited the induction of DNA synthesis by Cyrόl polypeptides, including wild-type Cyrόl and Cyrόl NT. Thus, integrin α_vβ₃ is required for Cyrό 1 -induced mitogenesis, although it is not involved in fibroblast adhesion (see Example 19). DNA synthesis for HUVE cells and NIH 3T3 fibroblasts was measured by thymidine incoφoration as described previously (Kireeva et al, [\ 996]) with minor modifications. HUVE cells were grown in 24- well plates to a subconfluent state, serum-starved for 24 hours and treated with F12K medium containing 10% fetal bovine serum (FBS), 1 μCi/ml [³H] -thymidine and 10 ng/ml basic Fibroblast Growth Factor (bFGF) (Gibco-BRL, Inc.) with various concentrations of Cyrόl and Fispl 2 as indicated. NIH 3T3 fibroblasts were grown to subconfluence, serum-starved for 48 hours, and treated with Minimal Essential Medium (MEM) containing 0.5% FBS, 1 μCi/ml [³H]-thymidine, bFGF and various concentrations of Cyrόl or Fispl2. Thymidine incoφoration into the trichloroacetic acid- insoluble fraction was determined after 24 hour incubation. Logarithmically grown mink lung epithelial cells (Mvllu, CCL64) were treated with various concentrations of TGF-βl (Gibco-BRL) and 2 μg/ml of Cyrόl or Fispl2 for 18 hours; [ H]-thymidine was then added to 1 μCi/ml for 2 hours. Thymidine incoφoration was determined as described above.

Purified recombinant Fisp 12 protein did not exhibit any mitogenic activity under any tested assay conditions. Rather, Fisp 12 was able to enhance DNA synthesis induced by fibroblast growth factor in either NIH 3T3 fibroblasts or HUVE-cells. This activity was nearly indistinguishable from that exhibited by Cyrό l .

Whereas in fibroblasts and endothelial cells, Cyrόl and Fisp 12 enhance growth factor-induced DNA synthesis, both proteins can also enhance growth factor-mediated actions in another way. It is known that TGF-β acts to inhibit DNA synthesis in epithelial cells (Satterwhite et al. , 1994). It was observed that both Cyrόl and Fisp 12 enhanced the ability of TGF-β to inhibit DNA synthesis in mink lung epithelial cells. The data demonstrate that both recombinant Cyrόl and Fisp 12, purified from serum-free sources, are not mitogenic by themselves, but have the ability to synergize with the actions of polypeptide growth factors. Cyrό 1 and Fisp 12 enhance DNA synthesis induction by FGF, and enhance DNA synthesis inhibition by TGF-β. Beyond the use of Cyrόl polypeptides and Fisp 12, the present invention comprehends the use of CTGF in methods to potentiate the mitogenic effect of true growth factors, orto screen for true growth factors. Those contemplated uses are in contrast to the reported use of CTGF as a mitogen or growth factor itself. U.S. Patent No. 5,408,040, column 7, line 65, to column 1 1 , line 7, incoφorated herein by reference hereinabove. It is expected that, in addition to murine and human Cyrόl polypeptides. Fisp 12, and CTGF, other ECM signaling molecules, such as members of the CCN family of proteins, will function to potentiate the mitogenic activity of true growth factors and will not function as true growth factors themselves.

Further, the invention comprehends methods of screening for modulators of mitogen potentiation. A comparative assay exposes subconfluent cells to an ECM signaling molecule such as Cyrόl , a growth factor, and a suspected modulator of an ECM signaling molecule. As a cυntrol, similar cells are exposed to the ECM signaling molecule and the growth factor. A further control exposes similar cells to the growth factor and the suspected modulator in the absence of the ECM signaling molecule. Based on the relative cell proliferation rates, as measured by, e.g.. [³H]-thymidine incoφoration, an identification of a suspected modulator as a promoter of mitogen potentiation (elevated cell proliferation in the presence of all three molecules) or an inhibitor of mitogen potentiation (decreased cell proliferation in the presence of the three molecules) can be made.

Additionally, the invention comprehends the use of ECM signaling molecules such as Cyrόl polypeptides in methods for treating conditions or disorders, such as diseases, associated with an under- or over-proliferation of mammalian cells. One of ordinary skill would readily comprehend that Cyrόl polypeptides having growth-factor-enhancing activity are suitable for treating conditions characterized by an undesirably low rate of cell proliferation. The use of Cyrόl polypeptides lacking such an activity could be used to treat conditions characterized by an over-proliferation of cells. Any means known in the art for delivering the therapeutic polypeptides or modulators is suitable, including direct injection into a mammal, such as a human, using any known route (e.g., subcutaneous, intramuscular, intravenous, intraperitoneal), optionally in the presence of a known adjuvant, excipient, carrier, or vehicle such as a liposome, or by gene therapy using any conventional nucleic acid delivery system. Any of these delivery mechanisms may be modified to target delivery of the therapeutic by mechanisms known in the art (e.g., association with a targeting molecule such as an antibody recognizing a desired cell type).

Example 19 Cornea Assay For Angiogenic Factors And Modulators Another assay for modulators of angiogenesis is an in vivo assay for assessing the effect of a suspected modulator in the presence of an ECM signaling molecule-related biomaterial, such as Cyrόl, on angiogenesis is the Cornea Assay. The Cornea Assay takes advantage of the absence of blood vessels in the cornea, which in the presence of an angiogenic factor, results in the detectable development of capillaries extending from the sclera into the cornea. Friedlander et al, Science 270:1500-1502 (1995). This ingrowth of new blood vessels from the sclera can be microscopically monitored. Further, the visually determined rate of migration can be used to assess changes in the rate of angiogenesis. These cornea assays may be performed using a wide variety of animal models. Preferably, the cornea assays are performed using rats. By way of example, an assay for suspected modulators of Cyrόl using this assay is disclosed. To perform this assay, Cyrόl is initially titrated using primary capillary endothelial cells to determine effective concentrations of Cyrόl . Subsequently, Cyrό 1 , in the presence or absence of a suspected modulator, is surgically implanted into the corneas of mammalian laboratory animals, e.g.. rabbits or rats, hi a preferred embodiment, Cyrόl (or Cyrό l and a suspected modulator) is embedded in a biocompatible matrix, using matrix materials and techniques that are standard in the art. Subsequently, eyes containing implants are visually observed for growth of the readily visible blood vessels within the eye. Control implantations may consist of physiologically balanced buffers embedded in the same type of matrix and implanted into eyes of the same type of laboratory animal receiving the Cyrόl -containing implants.

The development of an in vivo cornea assay for angiogenic factors has advantages over existing in vitro assays for these factors. The process of angiogenesis involves four distinct phases: induction of vascular discontinuity, endothelial cell movement, endothelial cell proliferation, and three-dimensional restructuring and sprouting. In vitro assays can evaluate only two of these steps: endothelial cell migration and mitogeuesis. Thus, to provide a comprehensive assay for angiogenic factors, an in vivo assay such as the cornea assay is preferred.

The cornea assay has been used to confirm the effect of angiogenic factors such as Cyrόl, Fisp 12, CTGF, and Nov, on the process of angiogenesis. Moreover, modifying the cornea assay by including any of these angiogenic factors and a suspected modulator of their activity results in a cornea assay for modulators of angiogenesis. For example, in one embodiment of the invention, dose of an angiogenic factor such as Cyrόl could be used in cornea assays for positive effectors of the angiogenic activity of Cyrόl . An appropriate dose of Cyrό 1 would initially be determined by titration of the dose response relationship of Cyrόl with angiogenic events. Inclusion of a control assay lacking Cyrόl would eliminate compounds having a direct effect on angiogenesis. In an alternative embodiment of the invention, an effective dose of an angiogenic factor such as Cyrόl could be used to assay for negative modulators of the activity of an angiogenic factor. In yet another alternative embodiment, a corneal implant comprises Cyrόl and another corneal implant comprises Cyrόl and a suspected modulator of angiogenesis. Measure ents of the development of blood vessels in the implanted corneas provides a basis for identifying a suspected modulator as a promoter of angiogenesis (elevated blood vessel development in the comea containing an implant comprising the suspected modulator. A relative decrease in blood vessel development identifies an inhibitor of angiogenesis. The rat is preferred as the animal model for the comea assay. Disclosures in the art have established the rat model as a well-characterized system for analyzing angiogenesis. Parameters such as implant size, protein release dynamics, and suitable surgical techniques, have been well characterized. Although any strain of rat can be used in the comea assay, preferred strains will be well-characterized laboratory strains such as the Sprague-Dawley strain.

Although rats of various sizes can be used in the co ea assay, a preferred size for the rats is 150-200 g'animal. Anesthesia is induced with Methoxyflurane and is maintained for 40-60 minutes with sodium pentobarbital (50 mg/kg, delivered intraperitoneally). The eyes are gently opened and secured in place by clamping the upper eyelid with a non-traumatic hemostat. Two drops of sterile proparacaine hydrochloride (0.5%) are then placed on each eye as to effect local anesthesia. Using a suitable surgical blade such as a No. 1 i Bard Parker blade, an approximately 1.5 mm incision is made approximately I mm from the center of the comea. The incision extends into the stroma but not through it. A curved iris spatula approximately 1.5 mm in width and approximately 5 mm in length is then inserted under the lip of the incision and gently blunt-dissected through the stroma toward the outer canthus of the eye. Slight finger pressure against the globe of the eye helps to steady the eye during dissection. The spatula penetrates the stroma no more than approximately 2.5 mm. Once the cornea pocket is made, the spatula is removed and the distance between the limbus and base of the pocket is measured to make sure the separation is at least about 1 mm.

To provide slow release of the protein after implantation in the co ea, protein is mixed with poly-2-hydroxyethylmethacrylate (Hydron⁵'), or an equivalent agent, to form a pellet of approximately 5 μl. Implants made in this way are rehydrated with a drop of sterile lactated Ringers solution and implanted as described above. After implantation, the corneal pocket is sealed with erythromycin ointment. After implantation, the protein-Hydron pellet should remain near the limbus of the co ea (cornea-sclera border) and vision should not be significantly impaired. Following surgery, animals are examined daily for seven days with the aid of a stereomicroscope to check for inflammation and responses. To facilitate examination, the animal is anesthetized with Methoxyflurane and the anesthetic is continuously administered by nose cone during examination. During this seven day period, animals are monitored for implant position and comeal exudate. Animals exhibiting comeal exudate are sacrificed. A preferred method of euthanasia is exsanguination. Animals are initially anesthetized with sodium pentobarbital (50 mg/kg) and then perfused, as described below.

After seven days, animals are perfused with colloidal carbon (e.g.. India Ink). Anesthesia is induced with Methoxyflurane, and is maintained with sodium pentobarbital (50 mg/kg, intraperitoneally). Each animal is perfused with 100-200 ml warm (37°C) lactated Ringers solution per 150 g of body mass via the abdominal aorta. Once the snout of the animal is completely blanched, 20-25 ml of colloidal carbon is injected in the same way as the Ringers solution, until the head and thoracic organs are completely black. Eyes are then enucleated and fixed. Corneas are excised, flattened, and photographed.

Each protein is typically tested in three doses, in accordance with the practice in the art. Those of ordinary skill in the art realize that six positive comeal responses per dose are required to support an identification of an angiogenic response. An exemplary comea assay includes three doses of the protein under study, with six rats being tested at each dose. Additionally, six animals are exposed to a buffer-Hydron implant and serve as negative controls. Exposure of at least three animals to a known angiogenic factor-Hydron implant serve as positive controls. Finally, to demonstrate the specificity of any observed response, six animals are exposed to implants containing a single dose of the protein under study, an excess of neutralizing antibody, and Hydron.

A comea assay as described above was performed to assess the ability of Cyrόl to induce angiogenesis. Four animals were given negative control implants containing a buffer-Hydron pellet (both eyes). Each of these animals failed to show any blood vessel development in either eye after seven days. Six animals received implants containing a biologically effective amount of Fibroblast Growth Factor (0.15 μM) in one eye and a control pellet in the other eye; all six showed angiogenic development in the eye receiving FGF, none showed neovascularization in the eye receiving the negative control. Seven animals received 1 μg/ml Cyrόl, in one eye and all seven of these eyes showed blood vessel growth; one of the seven eyes receiving a negative control showed angiogenic development. Finally, four animals received implants locally releasing 1 μg/ml Cyrόl (Hydron prepared with a 10 μg/ml Cyrόl solution) and a specific anti-Cyrόl antibody in three-fold excess over Cyrόl ; none of the eyes of this group showed any angiogenic development. Thus, the in vivo assay for angiogenesis identifies angiogenic factors such as FGF and Cyrόl . The assay also is able to reveal inhibition of angiogenic development induced ECM signaling molecules such as Cyrόl .

Example 20 Blood Clotting

ECM signaling molecules are also useful in correcting hemostasis, or abnormal blood clotting. A defect in blood clotting caused by, e.g., low level expression of cyr<57 which thereby allows Ti ssue Factor Pathway Inhibitor (TFPI) to act unchecked can be corrected by expression or use of recombinant Cyrόl protein.

Cyrόl can interact with TFPI, a protein that inhibits extrinsic blood coagulation. TFPI inhibits blood clotting in a two step process. First, TFPI binds to factor Xa and the TFPFXa complex then interacts with the Tissue Factor (TF):Factor Vila complex, thereby inhibiting the latter complex. The TF:Factor Vila complex is the complex that activates factors IX and X. By inhibiting TF:VIla, TFPI regulates coagulation by preventing the activation of Factors IX and X. required for blood clotting. The interaction of Cyrόl with TFPI inhibits the activity of TFPI, thus promoting blood coagulation. Cyrόl is, thus, a tissue factor agonist. Because of the interaction of Cyrόl and TFPI, Cyrόl can control the ability of TFPI to inhibit coagulation, thereby regulating hemostasis. A defect in Cyrόl may lead to the inability to inhibit TFPI at the appropriate time, resulting in excessive inhibition of tissue factor, thereby preventing clot formation. Deregulated expression of Cyrόl will conversely inhibit the activity of TFPI constitutively, and thus tissue factor is constantly active, resulting in excessive clotting. When the expression of cyrόl in endothelial cells is deregulated, one possible outcome is thrombosis.

In addition to Cyrόl, other ECM signaling molecules, such as Fisp 12 and CTGF, have been shown to exert effects on cells participating in angiogenesis.

Consequently, it is anticipated that a variety of ECM signaling molecule-related biomaterials, alone or in combination, may be used in the methods of the invention directed towards modulating hemostasis.

Example 21 Ex vivo Hematopoietic Stem Cell Cultures

To investigate the effect of Cyrό 1 on the growth of primitive multipotent stem cells, several assays that distinguish these cells from more mature progenitor cells in a hematopoietic culture are employed. These assays make use of physicochemical (fibronectin-binding) or growth and development-related (generation of progenitor blast colonies) differences between immature and mature subsets of cells.

Two cell lines which require conditioned media for growth are used as a source of hematopoietic stem cells (HSC). These cloned, factor-dependent murine lines are B6Sut (cloned from long term bone marrow culture and capable of growing in liquid medium without differentiation, but multipotent in agar, as described in Greenberger et al, Proc. Natl. Acad. Sci. /USA] 50:2931 [1983]), and FDCP-mix (cloned from long term bone marrow culture cells infected with the recombinant virus src-MoMuLV, and are multipotent in agar cultures, as described in Spooncer el al. Nature 3/0:2288 [ 1984]). BόSut cells are propagated in Kincaid's medium with 10% fetal calf serum (FCS) and 10% 6X- concentrated WEHI-conditioned medium. Greenberger et al. FDCP-mix cells are propagated in Fischer's medium with 20% horse serum and 10% 6X- concentrated WEHI-conditioned medium. The cell lines are propagated at 37°C, 5% CO₂.

Various ex vivo or in vitro cultures are assayed for population growth in the presence or absence of exogenously supplied murine Cyrόl or polyclonal anti-Cyrόl antibodies. Under limiting dilution conditions, the cobblestone area forming cell (CAFC) assay is used to identify cells with long term repopulating ability. Ploemacher et al, Blood 74:2155 (1989); Ploemacher et al, Blood 75:2527 (1991). Cells identified as having long term repopulating ability by the CAFC assay are then analyzed by measuring three parameters: Rate of population doubling, mitotic index, and rate of DNA synthesis.

Long term cultures, with or without supplementation with Cyrόl , are assayed for their levels of primitive HSC in the CAFC assay, van dev Sluijs et al, Exp. Hematol 22:1236 (1994). For example, M2-10B4 stromal cells provide a stromal cell layer for either B6Sut or FDCP-mix, which are each subjected to the CAFC assay in the following manner. Stromal cell layers are prepared by inoculating 5 x \⁽ M2-10B4 stromal cells (a cell line cloned from bone marrow stroma. Sutherland et al, Blood 75:666 [1991 ]) into each well of a 96-well culture plate in DMEM with 10% FCS. When the cells approach confluency, they are rinsed with PBS and irradiated (20 Gy of gamma-irradiation, 1 .02-1.04 Gy/minute) to prevent replication of any hematopoietic cells within the stroma, without affecting the stroma's ability to support hematopoiesis. B6Sut or FDCP-mix cells (sources of HSC) are added to the irradiated stromal cells in DMEM with 10% FCS, in the presence or absence of Cyrόl (10 μg/ml final concentration). After overlaying BόSut or FDCP-mix on the stromal cells, the cultures are incubated (e.g., 28-35 days for the murine system) and the number of cobblestone-forming areas (which identifies cells with long-term repopulating ability) are counted to determine the HSC frequency. ln a variation of the above-described CAFC assay, cells (e.g., BόSut or FDCP-mix) are initially maintained in parallel cultures in the presence or absence of a Cyrόl polypeptide. Subsequently. CAFC assays are performed, as generally described above; however, the assay is preferably performed in the complete absence of CAFC. It is expected that cells cultured in the presence of Cyrόl polypeptides will exhibit an increase in the frequency of HSC relative to cells initially cultured in the absence of Cyrόl polypeptides.

Following identification of cells with long-term repopulating ability, population doubling rates are determined, e.g., by microscopic examination of cell moφhology to determine the numbers of long term repopulating cells (and more mature short term progenitor cells) present in the various experimental long term cultures. Subsequent investigation of the expansion and differentiation capacities of the potential long term HSC cultures is used for confiπriation of suitable candidate cell lines. The mitotic index is determined according to procedures standard in the art. Keram et al. Cancer Genet. Cytogenet. 55:235 (1991). Harvested cells are fixed in methanokacetic acid (3: 1, v:v), counted, and resuspended at 10^h cells/ml in fixative. Ten microliters of this suspension is placed on a slide, dried, and treated with Giemsa stain. The cells in metaphase are counted under a light microscope, and the mitotic index is calculated by dividing the number of metaphase cells by the total number of cells on the slide. Statistical analysis of comparisons of mitotic indices is performed using the 2-sided paired t-test.

The rate of DNA synthesis is measured using a thymidine incoφoration assay. Various cultures are propagated in 1 μCi/ml [³H]-thymidine (ICN Biomedicals, Inc., Costa Mesa, CA) for 24-72 hours. Harvested cells are then rinsed with PBS and fixed with 10% trichloroacetic acid. DNA is dissolved in 0.1 N NaOH, and thymidine incorporation is determined, for example by liquid scintillation spectrophotometry.

The function of Cyrό 1 polypeptides in promoting the maintenance and/or expansion of long-term hematopoietic stem cell cultures was confirmed by antibody studies. Having determined that Cyrό l is a stroma-associated component important for the maintenance/expansion of long-term HSC cultures, stromal-contact bone marrow cell cultures were exposed to Cyrόl polypeptides in the presence or absence of anti-Cyrό 1 antibodies. Those cultures in which the Cyrόl activity had been neutralized by exposure to anti-Cyrόl antibodies showed a decrease in vitality of the culture, an increase in the number of visually detectable differentiated hematopoietic cells in the culture, and a decrease in the cell fraction comprising HSC , and a concomitant increase in the cell fraction comprised of committed cells, relative to HSC cultures maintained in the presence of Cyrό 1 , but the absence of anti-Cyrόl antibodies. Consequently, the activity of Cyrόl polypeptides is expected to be important for the stromal- ■ dependent expansion of undifferentiated hematopoietic stem cells.

The use of an ECM signaling molecule-related biomaterial, such as Cyrόl , can be used in the ex vivo expansion of hematopoietic stem cell cultures. In addition, more than one ECM signaling molecule-related biomaterial may be used to expand these cultures. For example, Cyrόl , with its expression targeted locally, may be combined with Fisp 12, which is secreted and, consequently, exhibits a more expansive targeting. As an alternative, CTGF maybe substituted for Fisp 12, its mouse ortholog. One of skill in the art would be able to devise other combinations of ECM signaling molecule-related biomolecules that are within the spirit of the invention.

Those of ordinary skill in the art will recognize that the successful expansion of hematopoietic stem cell cultures in the presence of ECM signaling molecules such as Cyrό 1 polypeptides provides a basis for a method of screening for suspected modulators of that expansion process. As in the other methods of the invention, a suspected modulator is combined with an ECM signaling molecule such as Cyrόl and exposed to primitive cells. In parallel, the ECM signaling molecule is exposed to similar cells. The relative rates of expansion may be used to identify a promoter, or inhibitor, of the ability of the ECM signaling molecule to expand pluripotent hematopoietic stem cell cultures. Cyr61 , alone or in combination with other hematopoietic growth factors, may also be used to expand stem cell populations taken from a patient and which may, after expansion, be returned to the patient or other suitable recipient patient after for example, chemotherapy or other treatment modalities that result in the depletion of blood cells in a patient. Stem cell populations expanded according to the present invention may also be used in bone marrow transplants in a patient in need thereof. Further, the invention contemplates the delivery of an ECM signaling molecule such as a Cyrόl polypeptide (including polypeptide fragments that inhibit endogenous Cyrόl activity) to a subject (e.g., patient) that would benefit from increased, or decreased, HSC production or hematopoiesis, using delivery means known in the art.

Example 22 Organ regeneration

The role of Cyrόl in the various cellular processes invoked by changes in the cellular growth state indicate that this protein would be effective in promoting organ regeneration. Towards that end, studies were conducted to determine the expression profile of murine cyr<57 in remaining liver tissue following a partial hepatectomy. (The response of remaining liver tissue following partial hepatectomy is a model for the liver's response to a variety of injuries, including chemical injuries, e.g., exposure to toxic levels of carbon tetrachloride.)

BALB/c 3T3 (Charles River) mice were subjected to partial hepatectomies removing approximately 67% of their liver tissue. Higgins et al,

Archs. Path. 72: 186-202 ( 1931 ). Twenty microgram aliquots of RNA were removed from the remaining liver tissue at varying times following the operation and liver RNA was isolated by tissue homogenization followed by guanidinium isothiocyanate, cesium chloride precipitation. Sambrook et al. RNAs were then immobilized on nitrocellulose filters and probed with radiolabeled clones containing various regions of murine cyrόl cDNA. Results were visualized by autoradiography and indicated that removal of liver tissue induced cyrόl mRNA expression, particularly in cells found near the injury site. Consequently, induction of cyrόl expression, e.g., by recombinant techniques, might promote the regeneration of organs such as liver. For example, cyrόl expression can be controlled, e.g., by introducing recombinant cyrόl constructs that have been engineered to provide the capacity to control expression of the gene, e.g., by the use of tissue-specific promoters, e.g., the K14 promoter for expression in skin. The recombinant cyrt57 may be introduced to cells of the relevant organ by gene therapy techniques using vectors that facilitate homologous recombination (e.g., vectors derived from Heφesviruses, Adenovirus, Adeno-associated Virus, Cytomegalovirus, Baculovirus, retroviruses, Vaccinia Vims, and others). Techniques for introducing heterologous genes into eukaryotic cells, and techniques for integrating heterologous genes into host chromosomes by homologous recombination, are well known in the art.

The development of skin, another organ, is also affected by Cyrόl . The expression of cyrόl is induced in cells in the vicinity of skin injuries. Also, as described above, Cyrόl has a chemotactic effect (i.e., Cyrόl induces cell migration) on endothelial cells and fibroblasts. Further, Cyrόl induces the proliferation of endothelial cells and fibroblasts. Both processes are involved in the healing of skin wounds. Accordingly, Cyrόl administration, e.g., by localized or topical delivery, should promote skin regeneration.

Cyrόl is also highly expressed in lung epithelium. These cells are frequently injured by exposure to environmental contaminants. In particular, lung epithelium is frequently damaged by air-bome oxidants. The administration of Cyrόl , e.g., in atomizers or inhalers, may contribute to the healing of lung epithelium damaged, e.g., by environmental contaminants. Example 23 Chondrogenesis- ECM Signaling Molecules Are Expressed in Mesenchyme

Some ECM signaling molecules are also expressed in cells, such as mesenchyme cells, that ultimately become a part of the skeletal system. In this Example, Cyrόl is identified as one of the ECM signaling molecules expressed in mesenchyme cells. Limb mesenchymal cells were grown in micromass culture as described above on glass coverslips (Fisher) for 3 days. Cultures were fixed in 4% paraform ldehyde in PBS, incubated for 30 minutes at room temperature with 1 mg/ml bovine testicular hyaluronidase (type IV, Sigma) in 0.1 N sodium acetate (pH 5.5) with protease inhibitors phenymethylsulfonyl fluoride (PMSF, 1 mM), pepstatin (1 μg/ml), leupeptin (1 μg/ml), aprotinin (1 μg/ml), aminocaproic acid (50 mM), benzamidine (5 mM), and EDTA (1 mM), blocked with 10% goat serum in PBS and incubated overnight at 4°C with primary antibodies against Cyrόl (Yang et al, [1991]), fibronectin fGibco) and tenascin (Gibco). Controls were incubated with anti-Cyrόl antibodies neutralized with 1 μg/ml purified Cyrόl . Cultures were subsequently incubated with FITC- conjugated goat-anti-rabbit secondary- antibody (Zymed), for 1 hour at room temperature.

For whole mount immunohistochemical staining, mouse embryos from gestational days 10.5 to 12.5 were fixed in 4% paraformaldehyde in PBS, dehydrated in methanol/PBS and stored at -20°C in absolute methanol. After rehydration, embryos were incubated with anti-Cyr61 antibodies as described in Hogan et al, Development 720:53-60 (1994), incoφorated herein by reference. Controls were incubated with anti-Cyrόl antibodies neutralized with 1 μg/ml purified Cyrό 1. Jmmunostained embryos were fixed, cleared and photographed.

Consistent with the transient expression of the cyrόl mRNA in somitic mesenchymal cells that are differentiating into chondrocytcs (O'Brien et al,

[1992]), the Cyrόl protein was found in the developing embryonic skeletal system. Cyrόl was localized by whole mount immunohistochemical staining to the proximal limb bud mesenchyme in gestational day 10.5 to 12.5 embryos. The Cyrό l protein was localized to the developing vertebrae, the calvarial frontal bone and the first brachial arch, as well as in the heart and umbilical vessels, forming an expression pattern that was consistent with the cyrόl mRNA expression pattern (O'Brien et al, [1992]). Cyrόl protein could be detected by immunoblot analysis in whole limb buds and in micromass cultures of limb bud mesenchymal cells. The level of Cyrόl protein remained at relatively constant levels throughout the 4 day culture period during which chondrogenesis occurred. Using quantitative immunoblot analysis, Cyrόl was estimated to represent approximately 0.03% of total cellular and extracellular proteins in the mesenchymal cell cultures. Cyrόl, tenascin (Gibco), and fibronectin were localized to the cartilage nodules by immunofluorescent staining in the mesenchymal cell cultures. Cyrόl and tenascin were primarily localized among the intranodular cells, while a fibrillar staining pattern was also observed around and bet een the cartilage nodules with anti-fibronectin antibodies. A similar immunofluorescent staining pattern was observed in transverse sections of the micromass cultures for all three antibodies. Together, these results show that endogenous Cyrόl is localized in the developing limb bud mesenchyme, both in vivo and in vivo.

Example 24 Chondrogenesis- ECM Signaling Molecules Promote Cell Adhesion

Cyrόl is a secreted protein that mediates the adhesion of fibroblasts and endothelial cells to non-tissue culture-treated plastic surfaces (Kireeva et al,

[1996]). The attachment of limb bud mesenchymal cells on non-tissue culture dishes coated with BSA, Cyrόl , tenascin, and fibronectin, were compared. Cyrόl , fibronectin (Gibco), or tenascin (Gibco) were diluted in 0.1 % protease-free bovine serum albumin (BSA) in PBS with 0.5 mM PMSF, to a final concentrations of 10 or 50 μg/ml. A 10 μl drop/well was placed in a non-tissue culture treated 24-well plate (Coming), and incubated at room temperature for 2 hours. The wells were blocked with 1 % BSA in PBS for 1 hour at room temperature, and rinsed with serum-free MEM (Modified Eagle's Medium). Limb mesenchymal cells, suspended at 5 x lO^ cell/ml in serum-free MEM, were added at a volume of 400 μl/well, and incubated at 37°C, 5% CO₂ for 1 or 3 hours. At each time point, the cell suspension was removed, the wells were rinsed with MEM and the remaining adherent cells were photographed.

Cells attached poorly to BSA-coated dishes, but adhered as clusters of rounded cells to Cyrόl- and tenascin-coated dishes within 1 hour of plating. In contrast, cells plated on fibronectin-coated dishes attached uniformly and started to spread. When cells were allowed to attach for 3 hours, many more adherent cells were observed. Furthermore, intercellular clustering and rounded cell moφhology were maintained in cells plated on Cyrόl and tenascin, while cells plated on fibronectin spread to form a monolayer. These observations show that Cyrόl mediates the adhesion and maintenance of a rounded cellular moφhology which is conducive for mesenchymal cell chondrogenesis (Zanetti et al., Dev. Biol. 739:383-395 [1990]; Solursh et al, Dev. Biol 94:259-264 [1982]), similar to that previously reported for tenascin (Mackie et al, J. Cell Biol. 705:2569- 2579 [1987]).

As mentioned previously, ECM signaling molecules such as Cyrόl may be used in methods of screening for modulators of cell adhesion, including, but not limited to, the adhesion of chondrocytes. The comparative assay, described above, measures the relative adhesion levels of cells exposed to a combination of an ECM signaling molecule and a suspected modulator of cell adhesion and cells exposed to the ECM signaling molecule alone, whereby the relative levels provide a basis for identifying either a promoter or an inhibitor of cell adhesion.

Example 25

Chondrogenesis- ECM Signaling Molecules Promote Cell Aggregation

Since aggregation is an essential step for chondrogenic differentiation

(Solursh, M., In The role of extracellular matrix in development, pp. 277-303

(Trelstad, R., ed.) (Alan R. Liss, New York 1984)), the ability of Cyrόl to mediate intercellular aggregation in suspension cultures of mesenchymal cells was assessed. The number of cells remaining at various times after isolation were counted. Untreated mesenchymal cells in suspension began to aggregate soon after isolation, as the number of single cells was decreased to 30% of the initial number within a 2 hour incubation period. Cell aggregation was significantly inhibited in cultures treated with affinity-purified anti-Cyrόl antibodies, indicating that endogenous Cyrό 1 is important for mesenchymal cell aggregation. To rule out the possibility that the affinity-purified anti-Cyrόl antibodies might contain undefined components that interfere with aggregation, anti-Cyrόl antibodies, described above, werepre-incubated with purified Cyrόl protein prior to addition to cells. These pre-incubated antibodies affected cell aggregation no more than the IgG and Cyrόl buffer controls, indicating that the anti-Cyrόl antibodies achieved their inhibition of cell aggregation by neutralizing the endogenous Cyrόl protein of mesenchymal cells. In addition to the ceil aggregation in suspension cultures described above, the effect of Cyrόl on mesenchymal cell aggregation in micromass cultures was also examined. When purified Cyrόl protein (0.3 μg/ml) was added to limb mesenchymal cells, precocious cellular aggregation was observed within 24 hours, unlike control cells which had not received Cyrόl . Neither Cyrόl -treated nor control cultures had differentiated into cartilage nodules at this time. By culture day 3, the development of internodular cellular condensations between the distinct cartilage nodules was more extensive in Cyrόl-treated cultures. These internodular condensations subsequently undergo chondrogenesis, observed as Alcian blue-staining cartilaginous matrix on culture day 4. Taken together, these results indicate that Cyrόl is able to promote cell-cell aggregation, a necessary step in chondrogenesis of mesenchymal cells in micromass culture. Example 26 Chondrogenesis- ECM Signaling Molecules Promote Cell Proliferation

Some ECM signaling molecules, such as Cyrόl , affect chondrogenesis, as revealed by effects on limb bud mesenchyme cells in micromass culture, as described above. Ahrens et al, Dev. Biol 60:69-82 (1977), has reported that these cells, in micromass culture, undergo chondrogenesis in a manner similar to the in vivo process. Mesenchyme cells were obtained from mouse embryonic limb buds by trypsin digestion (1 mg/ml, 1 :250 dilution of porcine pancreatic trypsin, Sigma Chemical Co.). Cells were explanted in plastic tissue culture wells and allowed to attach for 2 hours at 37°C, 5% CO₂. Cells were then incubated for 24 hours at 37°C, 5% CO₂ in MEM with 10% FBS, penicillin (50 U/ml), and streptomycin (50 μg/ml). At this point, the composition of the medium was changed by substituting 4% NuSerum (Collaborative Biomedical Products, Inc.) for 10% FBS. Individual cultures then received Cyrόl , fibronectin. heparin. (each at approximately 1 μg/ml) or buffer as a negative control. An additional control was provided by adding a 1 :100 dilution of affinity-purified anti-Cyrόl antibody (approximately 13 μg/ml stock solution), elicited and purified by standard techniques. Harlow et al.

Cell proliferation was assessed by the thymidine assay, described above, and by microscopic cell counts. Chondrogenesis was assessed by quantifying the incoφoration of [³⁵S]-sulfate (ICN Biomedicals, Inc.) into sulfated glycosaminoglycans, and by qualitatively deteπnining the extent of chondrogenesis by cell staining with Alcian Blue. Cultures, described above, were labeled with 2.5 μCi/ l [³⁵S]-sulfate for 18 hours, washed twice in PBS, fixed with Kahle's fixative (Pepper et al, J. Cell Sci. 709:73-83 [ 1995]) and stained for 18 hours in 0.5% Alcian Blue, pH 1.0. The extent of chondrogenesis is correlated with the intensity of Alcian Blue staining. San Antonio et al, Dev. Biol. 775:313-324 (1986). The results show that Cyrόl specifically increased limb bud mesenchyme cell proliferation and aggregation, leading to enhanced chondrogenesis. In addition to demonstrating that purified Cyrόl enhanced growth factor- induced DNA synthesis in fibroblasts and endothelial cells, the effects of Cyrόl on cell proliferation were directly examined. Cell proliferation during the 4 day culture period was determined by counting cell number and by incoφoration of [³H]-thymidine. To determine cell number, cells were harvested by trypsin/EDTA (Sigma) and counted with a Coulter counter. In parallel cultures, [³H]-thymidine ( 1 μCi/ml; ICN) was added to the media for 18 hours and incorporation in the TCA-insoluble layer was determined by liquid scintillation counting. Purified Cyrόl protein added to limb mesenchymal cells both increased cell number and enhanced DNA synthesis after 2 and 3 days in culture, although the total cell number in Cyrόl-treated and Cyrόl -untreated cultures leveled off at the same level after 4 days.

The role of Cyrόl in chondrogenesis may also improve the integration of prosthetic devices. For example, skeletal injuries and conditions frequently are treated by the introduction of a prosthesis e.g., hip prosthesis, knee prosthesis. Beyond questions of histocompatibihty, the successful implantation of a prosthetic device requires that the foreign element become integrated into the organism's skeletal structure. The capacity of Cyrόl polypeptides to affect cell adhesion, migration, and proliferation, and the ability of Cyrόl polypeptides to induce the differentiation of mesenchyme cells into chondrocytes, should prove valuable in the treatment of skeletal disorders by prosthesis implantation. For example, integration of a prosthetic device by chondrocyte colonization would be promoted by therapeutic treatments involving the administration of Cyrόl in a pharmaceutically acceptable adjuvant, carrier or diluent, using any of the administration routes known in the art or by coating the prosthesis device with Cyrόl polypeptides in a suitable caπier. The carrier may also be a slow-release type vehicle to allow sustained release of the polypeptides.

As noted in previously, the methods of the invention include a method of screening for modulators of cell proliferation, including chondrocytes. A comparison of the relative rates of cell proliferation in the presence of a control comprising an ECM signaling molecule alone (e.g., Cyrόl) and in the presence of a combination of an ECM signaling molecule and a suspected modulator of cell proliferation provides a basis for identifying a suspected modulator as a promoter, or inhibitor, of chondrocyte proliferation.

Example 27

Chondrogenesis- ECM Signaling Molecules Promote Chondrogenesis

Chondrogenic differentiation was quantitated by incoφoration of [ ^SS]- sulfate (ICN) into sulfated glycosaminoglycans and assessed qualitatively by

Alcian Blue staining. Cultures were radiolabeled with 2.5 μCi/ml [³⁵S]-sulfate for 18 hr, fixed with Kahle's fixative and stained with 0.5% Alcian Blue, pH 1.0 (Lev et al, 1964). The extent of chondrogenesis is correlated with the intensity of Alcian Blue staining (San Antonio et al, 1986). [³⁵S]-Sulfate incoφoration in the fixed cell layer was quantitated by liquid scintillation counting.

Exogenous purified Cyrόl protein promoted limb mesenchymal cell aggregation and resulted in greater Alcian blue-staining cartilaginous regions in micromass cultures, suggestive of a chondrogenesis-promoting effect. This effect was quantified by the incoφoration of [³⁵S]-sulfate into sulfated glycosaminoglycans (San Antonio et al, 1986) in Cyrόl-treated micromass cultures. Exogenous Cyrόl enhanced [^3:)S]-sulfate incoφoration in a dose- dependent manner, resulting in a 1.5-fold and 3.5-fold increase with 0.3 and 5 μg/ml Cyrό 1 , respectively, and was correlated qualitatively by increased Alcian Blue staining. The increase observed at the 5 μg/ml Cyrόl dose (120 nM) is an under-estimation of the actual extent of chondrogenesis, since some of the large cartilage nodules which were formed at this dose detached from the dish. Cultures treated with 10 μg/ml Cyrόl formed a more massive mound of cartilage.

A review of the literature indicated that chondrogenesis in limb mesenchymal cell micromass cultures was increased 2-fold with the addition of

10 μg/ml heparin (San Antonio et al. , [1987]; Resh et al. , [ 1985]) and 3-fold with

50 μg/ml tenascin (200 nM) (Mackie et al, [1987]). The results demonstrated that purified Cyrόl was effective at concentrations ( 10- 100 nM) similar to or less than those of other molecules known to promote chondrogenesis in this cell system.

Since a certain threshold cell density must be reached for initial aggregation to occur (Umansky, [1966]; Ahrens et al, [1977]), embryonic mesenchymal cells plated at low densities are normally unable to differentiate into chondrocytes, although the addition of exogenous factors such as heparin or poly-L-lysine (San Antonio et al, [1986]; San Antonio et al, [1987]) have been shown to support chondrogenesis in cells plated under these conditions. Therefore, the ability of Cyrόl to promote differentiation of mesenchymal cells plated at densities above and below the threshold for chondrogenesis was assessed. Cells plated at 2.5 x 10° cell/ml incoφorated little [³⁵S]-sulfate. However, when Cyrόl was added, these sub-threshold density cultures formed nodules and incoφorated sulfate to a level similar to that in cultures plated at 3 x 10⁶ cells/ml, which supports chondrogenesis. Therefore, Cyrόl can promote chondrogenesis in mesenchymal cells plated at non-chondrogenic, sub-threshold densities.

It is conceivable that when mesenchymal cells are plated in a high density micromass, the extent of chondrogenesis may be maximal and cannot be enhanced further by exogenous factors, which also may not be accessible to all cells. However, addition of exogenous Cyrόl resulted in a 2-fold enhancement in [³⁵S]-sulfate incoφoratiυn in cultures plated at densities ranging from 3 to 10 x 10⁶ cell/ml. Therefore, Cyrόl can further enhance chondrogenesis in high density micromass cultures, which have apparently not reached a maximal degree of differentiation.

It is possible that the increased [³³S]-sul fate incoφoration in Cyrόl -treated cultures is at least partly due to an increase in cell number, since Cyrόl also promotes cell proliferation. If this were true, then normalization of sulfate incoφoration with respect to cell number would eliminate any differences between control and Cyrό 1 -treated cultures. This was not found to be the case. Cyi'ό 1 -treated cultures still showed an approximately 2-fold increase in normalized sulfate incoφoration over control, indicating that Cyrόl promotes a net increase in chondrogenesis. On culture day 2, the sulfate/cell number ratio was lower in Cyrό 1 -treated cultures compared to controls and is reflective of a low level of [³⁵S]-sulfate incoφoration relative to cell number, since mesenchymal cells are mostly proliferating rather than differentiating in these early stage cultures (Ede, 1 83).

The presence of endogenous Cyrόl in these cells, both in vivo and in vitro, indicates that Cyrόl may indeed function biologically to regulate chondrogenic differentiation. The ability of exogenously added purified Cyrόl to promote intercellular aggregation and to increase [³⁵S]-sulfate incoφoration and Alcian-blue staining in limb mesenchymal cells demonstrates that Cyrόl can act as a chondrogenesis enhancing factor. As shown above in Example 11. anti- Cyrό 1 antibodies can neutralize both the cell adhesion and DNA-synthesis enhancement activities of Cyrόl . Anti-Cyrόl antibodies were added to the mesenchymal cell culture media or mixed the cell suspension prior to plating. Chondrogenesis was inhibited in the cultures treated with anti-Cyrόl antibodies, as demonstrated by decreases of [³⁵S]-sulfate incoφoration to 50% and 30% of controls, when antibodies were added to the media, and mixed with the cells, respectively. These observations were correlated with decreased Alcian Blue staining. However, mixing of the anti-Cyrό 1 antibodies with mesenchymal cells prior to plating resulted in complete detachment in some of the treated cultures within 24 hours.

To eliminate the possibility of an unidentified component in the antibody preparation as a cause of cell detachment, anti-Cyrό 1 antibody was preincubated with 1 μg/ml purified Cyrόl protein prior to mixing with cells. The inhibition of chondrogenesis in mesenchymal cells mixed with neutralized anti-Cyrόl antibodies was abolished.

Generally, the invention contemplates a method of screening for modulators of chondrogenesis. A comparative assay involves the exposure of chondrocytes to either (a) a combination of a suspected modulator of chondrogenesis and an ECM signaling molecule such as Cyrόl , or (b) the ECM signaling molecule alone. Measurements of the relative rates of chondrogenesis then provide a basis for identifying the suspected modulator of chondrogenesis as a promoter or inhibitor of that process.

The results described in this Example demonstrate that endogenous Cyrό 1 is present in mesenchymal cells and is important for their chondrogenesis. Accordingly, the use of an ECM signaling molecule, such as Cyrόl , to induce bone healing is contemplated by the invention. For example, a biologically effective amount of Cyrόl is introduced into a matrix such as a sponge, as described above, and this material is then applied to set bone fractures or used to gather together the fragments of a comminuted bone fracture. A biodegradable matrix may be employed, or the matrix may be removed at an appropriate later time. Alternatively, Cyrόl may be applied directly to bone In addition, Cyrό l may be applied to inanimate objects such as. biocompatible prosthesis, as described in Example 26.

Example 28 Genetics

Another way to control the effects of an ECM signaling molecule-related biomaterial is to inactivate it by creating dominant negative mutations in the relevant gene in actively growing and dividing cells. One approach involves the use of recombinant techniques, e.g., to create homozygous mutant genotypes in ex vivo cultures such as HSC cultures. Introduction of these cells into an organism, e.g., a human patient, would then provide an opportunity for the introduced mutant cells to expand and alter the expression of the ECM signaling molecule in vivo. Mutants homozygous for such a mutation could affect the expression of an endogenous wild type ECM signaling molecule in other cells.

Heterozygous mutants might produce altered ECM signaling molecules capable of interacting with the wild type ECM signaling molecule, also being expressed, in such a way that the ECM signaling molecule's activities are modulated or abolished.

Furthermore, because of the role played by ECM signaling molecules such as Cyrόl in regulating chondrogenesis (i.e., skeletal development), genetic manipulations that alter the expression of human Cyrόl may prove medically important for prenatal screening methods and gene therapy treatments related to skeletal conditions, in addition to angiogenic conditions. For example, the cyrόl gene is expressed when mesenchymal cells of both ectodeπnal and mesodermal origins differentiate to form chondrocytes. Thus, one of the roles that Cyrόl might play is to regulate the commitment of mesenchyme cells to chondrocyte cell lineages involved in skeletal development. Consistent with this view, transgenic mice overexpressing cyrόl ectopically are born with skeletal abnormalities. In all cases examined, the presence of the skeletal deformities correlates with expression of the transgene. These results suggest that the human form of Cyrόl may also regulate chondrogenesis and skeletal development. It is also possible that the human cyr(57 gene may correspond to a genetic locus already known to affect skeletal development or birth defects relating to bone moφhogenesis. Knowledge of the human Cyrόl protein sequence, presented in SEQ ID NO:4 herein, and the coding sequence of the cDNA, presented in SEQ ID NO:3 herein, provide the basis for the design of a variety of gene therapy approaches.

This information also provides a basis for the design of probes useful in genotypic analyses, e.g.. Restriction Fragment Length Polymoφhism analyses. Such analyses are useful in the fields of genetic counseling, e.g., in diagnosing diseases and conditions and the likelihood of their occurrence, as well as in forensic analyses.

By way of example, the materials of the present invention are useful in the prenatal screening for a variety of conditions or disorders, including blood disorders, skeletal abnormalities, and cancerous conditions. Well known techniques for obtaining fetal cells, e.g., amniocentesis, provide the materials needed for diagnosis, hi one embodiment of the invention, the fetal cells are expanded and DNA is isolated. In another embodiment, fetal cells are lysed and polymerase chain reactions are performed using oligonucleotide primers according to the invention. Using either approach, the DNA is then subjected to analysis. One analytical approach involves nucleotide sequence determination of particular regions of cyrόl or of the entire gene. The available human cyrόl coding sequence, presented in SEQ ID NO:3 herein, facilitates the design of sequencing primers that brings nucleotide sequence analysis into the realm of practical reality. An alternative to nucleotide sequence analysis is an investigation of the expression characteristics of the fetal nucleic acid. The capacity of the fetal nucleic acid to be expressed might be dispositive in the diagnosis of Cyrόl -related angiogenic, chondrogenic, or oncogenic disorders.

The invention also comprehends a kit comprising Cyrόl . The kits according to the invention provide Cyrόl in a form that is useful for performing the aforementioned methods of the invention. Kits according to the invention contain isolated and purified recombinant human Cyrόl in a suitable buffer, optionally stabilized by the addition of glycerol for storage at -20°C. In addition to the Cyrόl provided in the kit, the invention also contemplates the inclusion of any one of a variety of buffering agents, salts of various types and concentrations, and additional protein stabilizing agents such as DTT, all of which are well known in the art. Other kits according to the invention incoφorate isolated and purified murine Cyrόl . Kits incoφorating a Cyrόl polypeptide and an inhibitory peptide or an anti-Cyrόl antibody, as described above, are also contemplated.

Example 29 Fibroblast Adhesion

Primary human skin fibroblasts adhere to proteins of the CCN family, notably Cyrόl and CTGF, through integrin ₆β,, rather than either α_vβ₃ or α_IIbβ₃.

Fibroblast adhesion to either Cyrόl or CTGF requires interaction of the protein with cell surface heparan sulfate proteoglycan, which serves as coreceptor with - I l l - α,,β, for both Cyrόl and CTGF. In addition to its involvement in fibroblast adhesion, Cyrόl induces angiogenic factors such as vascular endothelial growth factor (i.e. ,VEGF) in these cells. It is expected that CTGF will also induce VEGF expression. Both Cyrόl and CTGF serve as bona fide signaling molecules acting through adhesion receptors. Fibroblast adhesion to Cyrόl and CTGF results in focal adhesion plaques, which can be visualized by staining with either anti-β , antibodies, or anti-focal adhesion kinase (FAK), anti-paxillin, or anti-vinculin antibodies. Moφhologically, fibroblasts adhered to Cyrόl form filipodia, consistent with a migratory response. Further, adhesion results in the rapid tyrosine phosphorylation of FAK and paxillin, as well as activation of MAP kinase and JNK kinase through phosphorylation. Adhesion to Cyrό l results in altered gene expression in fibroblasts. such as induction of MMP1 (collagenase; matrix metalloproteinase 1 ). The ₆β, integrin is the adhesion receptor for Cyr61 and CTGF in fibroblasts. Moreover, Cyrόl mediated adhesion through ₆β, involves concurrent binding to heparan sulfate proteoglycan. By contrast, Cyrό 1 adhesion through _vβ₃ in endothelial cells does not require binding to proteoglycans. Thus, the adhesion of endothelial cells consists of two components: one mediated through integrin α_vβ₃ and the other mediated through α₆β,. A small amount of Cyrόl -mediated adhesion in endothelial cells that could not be blocked by LM609 has been observed. This residual adhesion was completely blocked by anti- ₆β, antibodies. Thus α₆β, is the primary receptor for Cyrόl in fibroblasts, and the secondary receptor in endothelial cells. The functional-blocking monoclonal antibodies against integrins were purchased from Chemicon Inc: JB1 A (anti-β,); FB I 2 (anti-rx,); P 1 E6 (anti-α₂): P1 B5 (antι-α₃); P1 H4 (anti-α₄); P1 D6 (anti-α_s); NKl-GoH3 (anti-α₆); P3G8 (anti-α_v); LM609 (anti-α₄β₃). Polyclonal anti-Cyrόl antibody was raised in rabbits and affinity purified as described (Kireeva et al, [ 1997]). Synthetic peptides GRGDSP (SEQ TD NO:31) and GRGESP (SEQ ID NO:32) were purchased from Gibco-BRL. Heparin, Chondroitin sulfate A, Chondroitin sul fate C, and Decorin were from Sigma. Chondroitin sulfate B (Dermatan sulfate) and low molecular weight (3K) heparin were from Fluka.

To further characterize Cyrόl -mediated fibroblast adhesion, Cyrόl variants or mutants were constructed that harbored alanine substitutions within the putative heparin-binding sites between, approximately, amino acids 280-290 and amino acids 306-312. Constructions involved site-directed mutagenesis using a two-step polymerase chain reaction (PCR) procedure. Briefly, two overlapping internal oligonucleotide primers containing the altered sequences in opposite orientation along with outside primers were used in two separate PCR reactions. Mouse cvrόl cDNA was used as a template in the first PCR reaction. Resulting products were gel purified, combined, and used as a template for the second PCR reaction. The final mutant PCR product was digested with BsrCΛ, which cuts at sites flanking the mutated sequences. The BsrG\ fragment containing the wild-type cyrόl cDNA in the pSG5 vector was substituted with the mutated PCR product. The orientation and mutations were confirmed by DNA sequencing. Finally, the mutant cyrόl cDNA was released from pSG5 by digestion with EcoRI from pSG5 and cloned into the Baculovirus expression vector pBlueBac 4.5 (Invitrogen). For the HI mutant, the internal primers were 5'-

GCGGC ATGC AGCGCGACCGCGAAATCCCCAGA ACCAGTC-3 ' (primer fH l ; SΕQ ID NO: 18) and 5'-TCGCGCTGCATGCCGCGCCCGC- TTTTAGGCTGCTGTACACTG-3' (primer rHl ; SΕQ ID NO: 19). For the H2 mutant, the internal primers were 5'-GTCGCGGCATACGCGCCCAAATAC- TGCGGCTCM (primer fH2; SΕQ ID NO:20) and 5'-GCGCGTATGCCGC- GACACTGGAGCATCCTGC-3' (primer rH2; SΕQ ID NO:21 ). The outside primers used in the second PCR reaction for each mutant were 5'- CAGACCACGTCTTGGTCC-3' (upstream PCR primer; SΕQ ID NO:22) and 5'- GAATAGGCTGTACAGTCGG-3' (downstream PCR primer; SΕQ ID NO:23). To construct a double mutant (dmcyrό l), the H2 mutant cvrόl -containing amplified polynucleotide was used as a PCR template when introducing the mutation found in H 1.

Recombinant human Cyrόl and mutant Cyrόl, synthesized in a Baculovirus expression system using Sf9 insect cells, were purified from serum- free conditioned media by chromatography on Sepharose S columns. The purity and yield of the proteins were determined by SDS-PAGE followed by Coomassie blue staining and immuno- blotting. Human fibronectin, human vitronectin, rat tail Type-I collagen and mouse laminin were obtained from Collaborative Research, MA. Primary human foreskin fibroblast cell line 1064SK (ATCC CRL-2076,

Starting passage 2) was cultured in DMEM (4.5 g/liter glucose, Gibco) with 10% fetal bovine serum (Intergene). Human umbilical vein endothelial cells (HUVEC) were from Cascade Biologies Inc. and were grown in the medium provided by Cascade Biologies. Cell adhesion assays were performed under serum-free conditions. Briefly, a protein under study was diluted to the desired concentration (0.1-10 μg/ml) in PBS, applied to 96-well microtiter plates (50 μl per well) and incubated at 4°C for 16 hours. Unsaturated protein binding capacity was blocked with 1% BSA at room temperature for 1 hour. Cells were washed twice with PBS and harvested by incubation in PBS containing 2.5 mM EDTA at room temperature for 10 minutes. Detached cells were washed with serum-free basal culture medium containing 1% BSA and resuspended at 2.5 x 10³ cells/ml in the same medium. Where indicated, reagents (EDTA, heparin, peptides, etc.) were mixed with cells prior to plating, and antibodies were incubated with cells at room temperature for 30 minutes before plating. To each well, 50 μl of cell suspension were plated and, after incubation at 37°C for 30 minutes, wells were washed twice with PBS. Adherent cells were fixed with 10% formalin, stained with methylene blue, and quantified by dye extraction and measurement of absorbance at 620 nm. Inhibition of glycosaminoglycan sulfation was achieved by growing cells in medium containing the indicated amount of sodium chlorate for 24 hours. Involvement of cell surface proteoglycan was examined by pretreating cells with heparinase I (2U/ml, Sigma) or chondroitinase ABC (2U/ml, Sigma) at 37°C for 30 minutes.

The ability of Cyrόl to mediate cell adhesion in normal fibroblasts was investigated. Microtiter wells were coated with purified recombinant Cyrόl protein, and 1064SK primary human foreskin fibroblasts were allowed to adhere under serum-free conditions. In particular, washed 1064SK fibroblasts were detached with 2.5 mM EDTA and resuspended in serum-free DMEM medium at 2.5X10³ cells/ml. 50 μl cell suspensions were plated on microtiter wells coated with varying concentrations (0.5-5.0 μg/ml) of Cyrόl protein. After incubation at 37°C for 30 minutes, adherent cells were fixed and stained with methylene blue. Extracted dye was quantified by absorbance at 620 nm and means from three trials showed an absorbance of about 0.3 A₆₂₀ (0.5 μg/ml Cyrόl), with the absorbance ranging from 0.45-0.55 A₆₂₀ for Cyrόl solutions of 1-5 μg/ml, respectively. Adhesion of 1064SK cells to Cyrόl was dose-dependent and saturable.

In another experiment, microtiter wells were coated with either BSA, Cyrόl (2 μg/ml), or vitronectin (VN; 0.5 μg/ml) and blocked with affinity- purified anti-Cyrόl antibodies for one hour at 37°C before 1064SK cells were plated. The A₆₂₀ for BSA-coated wells was about 0.05 with orwithout anti-Cyrόl blocking antibody. With Cyrό 1 -coated wells, the A₆₂₀ was 0.45 (without blocking antibody) or 0.15 (with blocking antibody). With VN-coated wells, the A₆₂₀ was about 0.55, with or without blocking antibody. These data are means of triplicate determinations. The results showed that affinity-purified anti-Cyrόl antibodies inhibited 1064SK cell adhesion to Cyrόl but not to vitronectin, indicating that the ability to mediate fibroblast adhesion was an intrinsic property of the Cyrόl protein.

The effects of divalent cations in cell adhesion to Cyrόl was also investigated. Cells were added to microtiter wells coated with Cyrόl (2 μg/ml), Type I collagen (Col. 1 , 2 μg/ml), Vitronectin (VN, 0.5 μg/ml) or BSA (control); EDTA (2.5 m) or Mg²' (5.0 mM) was added to some cells. Cells were plated on microtiter wells coated with Cyrόl (2 μg/ml); and one of the following components was added: Nothing, Ca²⁺, Mg²¹ Mn²", Ca²⁺/Mg²', orCa²7Mn²' (5.0 mM each); the A₆₂₀ values were 0.07, 0.46. 0.07, 0.46, 0.50, 0.08, and 0.48. respectively. Data are means from triplicate determinations. Fibroblast adhesion to Cyrόl was completely blocked by 2.5 mM EDTA (about 0.12 A₆₂₀), and was restored by the addition of 5.0 mM Mg²⁺ (0.50 A_b2Q). As expected, similar effects were observed for fibroblasts plated on collagen (0.10 A₆₂₀ with EDTA, 0.58 A₆₂₀ with Mg²⁺) or vitronectin (0.05 A₆₂₀ with EDTA, 0.55 A₆₂₀ with Mg²X The presence of Ca²⁺ abolished cell adhesion to Cyrόl completely, whereas the addition of Mg² or Mn²⁺ had no effect. Inhibition by Ca²¹ is characteristic of adhesion through some members of the β, family of integrins, including receptors that bind type I collagen, laminin, and vitronectin (integrins α₂β,, ₆β,, and α_vβ,) in fibroblasts, as well as the lymphocyte-specific integrin α_Lβ₂. This observation also helped to exclude some integrins whose adhesive properties are supported by CaX The presence of Mn² ' , but not Mg ' , was able to overcome the inhibitory effect of Ca²⁺on cell adhesion to Cyrό 1 , suggesting that MiM can bind the Cyrό 1 adhesion receptor with higher affinity than Ca^{2 *}

A number of integrins expressed in fibroblasts, notably the α_v integrins (cc β, α_vβ₃ and α_vβ₅ vitronectin receptors) and integrin ₅β, (fibronectin receptor), are sensitive to inhibition by RGD-containing peptides. The effects of RGD peptides on Cyrόl -mediated fibroblast binding was tested by plating cells in wells coated with Cyrόl (2 μg/ml). Type I collagen (2 μg/ml) or vitronectin (0.5 μg/ml); 2 mM GRGDSP or GRGESP peptide was then added. For Cyrόl- coated wells, no addition, GRGDSP, or GRGESP gave A₆₂₀ values of about 0.50, 0.48, and 0.46, respectively. For Type I collagen, the A₆₂₀ values were 0.50, 0.47 and 0.49, respectively, and for VN, the values were 0.45, 0.05, and 0.41 , respectively.

Cells were also pre-incubated with 40 μg/ml monoclonal antibodies against the integrin α₅ or the integrin α₁ subunits at room temperature for 30 minutes, then plated on wells coated with Cyrόl (2 μg/ml), vitronectin (0.5 μg/ml) or fibronectin (FN, 1 μg/ml). For Cyrόl -coated wells, the A₆₂₀ values were 0.75 (no addition), 0.73 (anti-α_v), and 0.77 (anti- ₅); the corresponding values for VN were 0.78, 0.32, and 0.72; for fibronectin, the corresponding values were 0.72, 0.65, and 0.36. The data are means, representative of three experiments. The peptide RGDSP, but not the control peptide RGESP, completely abolished 1064SK cell adhesion to vitronectin. By contrast, RGDSP had no effect on fibroblast adhesion to Cyrόl or type I collagen. This result indicated that fibroblast adhesion to Cyrόl is not mediated through either the α_v integrins (α_vβ, _vβ₃, and α_vβ₅) or ₅β,. 1064SK cells were also challenged by pre-incubation with monoclonal antibodies specifically recognizing the _v and α₅ integrin subunits. Whereas 1064SK adhesion to vitronectin and fibronectin was blocked by monoclonal antibodies against α_v and ₅, respectively, these antibodies had no effect on adhesion to Cyrόl . Thus, 1064SK fibroblast adhesion to Cyrόl is mediated through one of the subset of β, integrins (e.g., α β,, α β,, or α₆β, ) known to be inhibited by Ca^2" and insensitive to inhibition by RGD-containing peptides.

The 1064 SK cells were also analyzed using the monoclonal antibody JB1A, which binds the β, integrin subunit. Cells were pre-incubated with the antibody (50 μg/ml) before plating on microtiter wells coated with Cyrόl (2 μg/ml), Type I collagen (2 μg/ml), or vitronectin (0.5 μg/ml). Anti-β, antibody inhibited cell adhesion to Cyrόl (A₆₂₀ of 0.53 without antibody and 0.14 with antibody) by 75%, confirming that 1064SK fibroblast adhesion to Cyrό 1 requires the involvement of a β, integrin. As expected, adhesion to type I collagen was inhibited by about 62% (A_f)20 of 0.56 without and 0.21 with antibody) by the anti- β, antibody, whereas adhesion to vitronectin was unaffected (A₆₂₀ of 0.48 without, and 0.44 with, antibody). Data are means, representative of three experiments.

To identify the specific β, integrin that mediates fibroblast adhesion to

Cyrόl , the inhibitory effects of function-blocking monoclonal antibodies were tested against various integrin subunits. Cells were pre-incubated with monoclonal antibodies against the integrin βl subunit (50 μg/ml, see results described above) or the integrin α₆ subunit (20 μg/ml) at room temperature for 30 minutes, then plated on wells coated with Cyrόl (2 μg/ml), laminin (5 μg/ml), or fibronectin (1 μg/ml). The data are means, representative of at least three experiments. Monoclonal anti-α₆ antibody blocked 1064SK fibroblast adhesion to Cyrόl (A₆₂₀ of 0.53 without, and 0.08 with, antibody) by more than 80%, while having no effect on adhesion to fibronectin. Adhesion to laminin, a ligand for integrin otgβ _] , was only partially blocked (A₆₂₀ of 0.62 without, and 0.48 with, antibody, or a block of about 22%). Cells were pre-incubated with 40 μg/ml monoclonal antibodies against integrin αl, α2, α3 or α4, or treated with a cocktail of αl , α2 and α3 antibodies ( 40 μg/ml each) at room temperature for 30 minutes, then plated on wells coated with Cyr61(2 μg/ml), vitronectin (0.5 μg/ml), or Type I collagen (2 μg/ml). Data are means, representative of three experiments, and are presented in Table II.

TABLE II

Function-blocking monoclonal antibodies against integrin α,, α₂, α₃ or α₄ subunits, or a combination of antibodies against ^ α^ and α- ,had little effect on Cyrόl -mediated cell adhesion (Table II). In contrast, a mixture of anti-α,, anti-α₂, and anti-cc_j antibodies almost completely inhibited fibroblast adhesion to collagen. Thus, 1064SK fibroblast adhesion to Cyrόl is mediated through integrin ctgβ_j .

The role of heparin in the binding of fibroblasts to Cyrόl was also examined. First, various amounts of soluble heparin were added to suspensions of 1064SK fibroblasts prior to plating on either Cyrόl- or fibronectin-coated wells. In particular, cells were plated on wells coated with Cyrόl (2 μg/ml) or fibronectin (1 μg/ml). Different amounts of heparin (0.001-1000 μg/ml) were included in the cell suspension before plating. Cells were plated on wells coated with Cyrόl (2 μg/ml) or vitronectin (0.5 μg/ml). Data in terms of means, representative of at least three experiments, showed that heparin levels as low as 1 ng/ml detectably inhibited binding, with 0.1 μg/ml or more completely inhibiting cell adhesion to Cyrόl . However, heparin levels up to 1000 μg/ml had no effect on cell adhesion to fibronectin. .

The influence of chondroitin sulfates on cell adhesion to Cyrό i was additionally examined. Cells were plated on wells coated with Cyrόl (2 μg/ml) or vitronectin (0.5 μg/ml). Chondroitin sulfate A (1 mg/ml), chondroitin sulfate B (100 μg/ml), chondroitin sulfate C (10 mg/ml), or decorin (100 μg/ml) was included in the cell suspension before plating. The data (three trials) showed that 1 mg/ml chondroitin A or 10 μg/ml of either chondroitin B or decorin inhibited cell adhesion to Cyrό l ; chondroitin C failed to inhibit adhesion to Cyrόl at all concentrations tested (0.01 -1000 μg/ml). However, the concentrations of chondroitin sulfates A, B, or decorin needed for this inhibition were orders of magnitude higher than the effective inhibitory concentration of heparan sulfate (e.g., 0.1 μg/ml heparan sulfate versus 1 mg/ml chondroitin sulfate A). The addition of sodium chlorate (an inhibitor of proteoglycan sulfation) to cell suspensions in Cyrόl -treated (2 μg/ml) wells resulted in a dose-dependent response, with 90% inhibition of adhesion at 40 mM sodium chlorate. In contrast, this concentration of chlorate resulted in only a 10-20% inhibition of cell adhesion to other substrates (fibronectin, type I collagen, vitronectin, and laminin). Data are means, representative of three experiments. The chlorate inhibition study was conducted by culturing cells in media containing 0-50 mM sodium chlorate for 24 hours, washing, and harvesting as described above, then plating on Cyrόl (2 μg/ml), fibronectin (1 μg/ml), Type I collagen (Coll. I, 2 μg/ml), vitronectin (VN, 0.5 μg/ml), or laminin (5 μg/ml). Chlorate-mediated adhesion inhibition was rescued by the addition of 10 mM MgSO₄. Cells were cultured in media containing 50 mM sodium chlorate, or 50 mM sodium chlorate plus 10 mM magnesium sulfate for 24 hours, washed and harvested, then plated on wells coated with Cyrόl (2 μg/ml), vitronectin (0.5 μg /ml), or fibronectin (1 μg/ml). The 1064SK fibroblasts were also treated with heparatinase, which rendered cells unable to adhere to Cyrόl (A₆₂₀ of 0.13 with, and 0.59 without, treatment), but had no effect on cell adhesion to vitronectin (A₆₂₀ of 0.63 with, and 0.70 without, treatment) or fibronectin (A₆₂₀ of 0.66 with, and 0.74 without, treatment). In contrast, chondroitinase A, B, C had no effect (A₆₂₀ values of 0.57, 0.69, and 0.65, for Cyrόl, vitronectin and fibronectin, respectively) on cell adhesion, indicating that chondroitin sulfates do not contribute significantly to human foreskin fibroblast adhesion to Cyrόl . Thus, cell surface proteoglycans, such as heparan sulfate proteoglycans, are involved in Cyrόl -mediated fibroblast cell adhesion. The adhesion of human fibroblasts to Cyrόl is mediated through integrin α^β,, and sulfated proteoglycans play a role in that adhesion.

Mutant Cyrόl proteins deficient in heparin binding were generated to examine the effect of such changes on fibroblast adhesion. Conventional site- directed mutagenesis techniques were used to produce mutant Cyrό 1 polypeptides having altered heparin-binding moti fs. Two putative heparin-binding motifs were found within the carboxyl-terminal domain in Cyrόl that conform to the consensus XBBXB sequence for heparin binding (where B denotes basic amino acid residues such as lysine or arginine). Site-directed mutagenesis was used to replace the lysine and arginine residues in the motifs with alanine, thus creating two Cyrόl variants (HI and H2) each having one of the two heparin binding motifs mutated. In addition, both motifs were mutated in a Cyrόl double mutant (DM) variant. A comparison of the mutated amino acid sequence H₂N- SLKAGAACSATAKSPEPVRFTYAGCSSVAAYAPKYCG-COTi (SEQ ID NO:30) with residues 278-314 of SEQ ID NO:2 (wild-type mouse Cyrόl), shows clusters of amino acid changes between residues 280-290 (HI, underscored above) and between residues 305-310 (H2, underscored above); both sets of clustered changes are found in DM. These mutations were created using the full- length cyrόl ; the mutant constructs were expressed in, and purified from, recombinant Baculovirus-transformed insect cells. Equal amounts of conditioned media of insect SF9 cells infected with Baculovirus expressing wild- type or mutant Cyrόl protein were loaded on CL-6B Heparin Sepharose columns. After washing with 20 bed volumes of RIPA buffer, bound protein was eluted with RIPA buffer containing increasing concentrations of sodium chloride. Equal amounts of eluate from each fraction were analyzed on SDS -PAGE gels followed by Western blotting to visualize Cyrόl protein. Antibodies used were rabbit polyclonal antibodies against bacterial GST-Cyrόl . The HI mutant Cyrόl polypeptide eluted over the range of 04-0.8 M NaCl; H2 eluted over the range 0.4-1 .0 (primarily between 0.6-0.8) M NaCl; DM eluted during the washing and up to 0.25 M NaCl; and wild-type Cyrόl eluted at 0.8- 1.0 M NaCl. These elution profiles indicate that HI and H2 exhibited somewhat decreased heparin-binding affinities, whereas DM was severely deficient in heparin binding

To examine the activities of these Cyrόl variants (i.e., mutant Cyrόl proteins), various concentrations of the variants were separately coated onto microtiter wells and 1064SK fibroblasts were added. Adhesion assays were perfoπried by optionally pre-incubating cells with 20 μg/ml monoclonal anti-α₆ antibody at room temperature for 30 minutes, then plating on wild-type Cyrόl, mutani Cyrόl HI , mutant Cyrόl H2. or vitronectin (0.5 μg/ml). The results (three trials) showed that both HI (e.g., 2.5 μg/ml) and H2 (e.g., 2.5 μg/ml) were able to support fibroblast adhesion with adhesion isotheπns that were comparable to that of wild-type Cyrόl (e.g., 2 μg/ml), although maximal adhesion was reached at a lower concentration of wild-type protein ( 1 μg/ml) compared to the mutant proteins (2.5-5.0 μg/ml). Moreover, adhesion to either HI or H2 was blocked by antibodies against the integrin subunit ^, indicating that cell adhesion to the mutant Cyrόl proteins is also mediated through integrin ot/A . In particular, the antibody inhibited Cyrόl -mediated binding by 78%, Hl - mediated binding by 69%, and H2-mediated binding by 70%; fibroblast binding to vitronectin was only inhibited by 9%. Thus, the integrin o.gβ _| binding sites of Cyrόl are distinct from the heparin-binding sites mutated in H I and H2. Mutant Cyrόl proteins that preserved either heparin-binding site exhibited sufficient residual heparin-binding activity to support fibroblast adhesion. In contrast, DM was unable to support 1064SK fibroblast adhesion at any concentration tested, indicating that the intrinsic heparin binding activity of Cyrό 1 is essential for mediating fibroblast adhesion.

Experiments designed to examine the effect of deleting the carboxy- terminal domain of Cyrόl on Cyrόl -mediated adhesion of 1064 SK fibroblasts yielded consistent results in establishing that the carboxy-terminal domain, containing the heparin binding site, were necessary for fibroblast adhesion to Cyrόl . The experiments followed the protocol described above, but the recombinant human cyrόl construct encoded Cyrόl NT, a Cyrόl polypeptide containing amino acids 1-281 ofSEQ IDNO:4 (i e., lacking the carboxy-terminal amino acid residues 282-381 ofSEQ IDNO:4). Thus, mature Cyrόl NT contains domains I, II and III, con-esponding to the IGFBP homology domain, the von Willebrand factor type C repeat domain, and the thrombospondin type 1 repeat domain, respectively, while lacking the heparan binding domain IV. Results of the experiments demonstrated that Cyrόl NT did not support fibroblast adhesion, although this Cyrόl polypeptide fragment did support fibroblast migration, as described above (see Example 14). Further, immunological analyses showed that an antibody (i.e., GoH3) specifically recognizing the integrin α blocked human fibroblast adhesion to Cyrό l polypeptides.

The following experiment showed that the requirement for Cyrόl heparin binding sites was distinct from Cyrόl -mediated adhesion through the α_vβ₃ integrin. Washed HUVE cells were detached by 2.5 mM EDTA and resuspended in serum-free F-12K medium at 5 x 10⁵ cells/ml. Cells were pre-incubated with 2.5 mM EDTA, 1 mM RGD peptide, or 40 μg/ml anti-α,.β₃ monoclonal antibody (LM609) at room temperature for 30 minutes, then plated on wild-type Cyrόl (5 μg/ml). double-mutant Cyrόl (DM Cyrόl, 10 μg/ml), or vitronectin (0.5 μg/ml). Immobilized cells were stained with methylene blue and absorbances (A₆₂₀) were recorded. Relative binding capacities (binding of cells not exposed to a pre- incubation compound defined as 100%) are presented in Table III.

Table III

The data were based on three independent trials. Thus, DM still mediated HUVEC binding, establishing that the failure of DM to bind to fibroblasts was specific to that cell type.

Cell adhesion assays were also perfonned on smooth muscle cells. Bovine aortic smooth muscle cells (BASM cells) were subjected to the cell adhesion assay described above in the context of assaying fibroblasts. Results showed that Cyrόl polypeptides induced adhesion of BASM cells. Heparin, but not "RGD" peptides (see, e.g., SEQ ID NO:31), inhibited Cyrόl -induced adhesion of the smooth muscle cells. To identify particular integrin receptors mediating the Cyrόl induction of BASM cell adhesion, antibody studies were conducted using anti-integri n antibodies recognizing speci fie integrins. Antibody GoH3, recognizing integrin _(l, completely abolished BASM cell adhesion. Similarly, antibody JB la, specifically recognizing integrin β,, eliminated Cyrόl - induced BASM cell adhesion. Thus, integrin α₍₎β, mediates Cyrόl-induced adhesion of smooth muscle cells. It is expected that Cyrόl polypeptides that include the heparan binding domain IV will induce adhesion of any mammalian smooth muscle cell.

Accordingly, another aspect of the invention is directed to a method of screening for a modulator of cell adhesion comprising the steps of: (a) contacting a first fibroblast cell with a suspected modulator of cell adhesion and a biologically effective amount of an ECM signaling molecule-related biomaterial selected from the group consisting of a Cyrόl , a Fisp 12, a CTGF, a NOV, an ELM-1 (WISP-1), a WISP-3, a COP-1 (WISP-2), and fragments, analogs, and derivatives of any of the aforementioned members of the CCN family of proteins; (b) separately contacting a second fibroblast cell with a biologically effective amount of an ECM signaling molecule-related biomaterial described above. thereby providing a control; (c) measuring the level of cell adhesion resulting from step (a) and from step (b); and (d) comparing the levels of cell adhesion measured in step (c), whereby a modulator of cell adhesion is identified by its ability to alter the level of cell adhesion when compared to the control of step (b). Preferably, the fibroblast cells present the α₆β, integrin. Also preferred are fibroblast cells that present a sulfated proteoglycan, such as hepaian sulfate proteoglycan. Any one of a number of CCN polypeptides may be used in the methods of the invention, such as Cyrό 1 (mouse- SEQ ID NO:2, human- SEQ ID NO:4, rat- Genbank Ace. No. AB015877), Fispl 2/CTGF (mouse- SEQ ID NO:6, human- SEQ ID NO:8, N vi idescens- Genbank Ace. No. AJ271 167, Susscrofa- Genbank Ace. No. U70060. X. laevis- Genbank Ace. No. U43524, 73. taurus- Genbank Ace. No. AF000137, and rat- Genbank Ace. No. AF120275), NOV (human- Genbank Ace. No. NM_002514, mouse- Genbank Ace. No. Y09257, and G. gallus- Genbank Ace. No. X59284), ELM-1 (Wisp- 1 ; human- Genbank Ace. No. NM_003882, mouse- Genbank Ace. No. AB004873), COP-1 (Wisp-2; human- Genbank Ace. No. NM_003881 , mouse- Genbank Ace. No. AF 100778), and Wisp-3 (human- Genbank Ace. No. NM 003880).

Further, the invention comprehends screens for modulators of interactions between ECM signaling molecules, such as human Cyrόl polypeptides, and α₆β,, a specific integrin receptor whose activities include, but are not limited to, mediating the adhesion of fibroblasts, smooth muscle cells and endothelial cells. The screening methods involve contacting a CyrC.1 polypeptide with α₆β,, preferably found in a composition including a mammalian cell membrane such as an endothelial, fibroblast or smooth muscle cell membrane, in the presence and absence of a potential or suspected modulator of the Cyrόl -integrin interaction. Detection of relative levels of interaction, preferably in the form of detecting relative levels of cell membrane (or whole cell) adhesion, leads to the identification of modulators. The invention further extends to methods of treating conditions or disorders, such as diseases, associated with excessive or inadequate cell adhesion, such as various forms of fibrosis, defective angiogenesis, tumor growth, tumor metastasis, granulation tissue disorders, inflammatory responses, and certain muscle disorders known in the art. Treatment involves delivery of a therapeutically effective amount of a Cyrόl polypeptide or a modulator of the Cyrόl -α₆β, integrin receptor interaction to a mammal such as a human by any means known in the art.

The invention also contemplates analogous methods of screening for modulators of fibroblast cell migration or proliferation. The method described below identifies modulators of cell migration; the described method applies to methods of screening for modulators of cell proliferation by substituting the parenthetically noted terms. A method of screening for modulators of fibroblast cell migration comprises the steps of: (a) contacting a first fibroblast cell with a suspected modulator of cell migration (proli feration) and a biologically effective amount of an ECM signaling molecule-related biomaterial selected from the group consisting of a Cyrόl, a Fisp 12, a CTGF, a NOV, an ELM-1 (W1SP-1 ), a WISP-3, a COP- 1 (WISP-2), and fragments, analogs, and derivatives of any of the aforementioned members of the CCN family of proteins; (b) separately contacting a second fibroblast cell with a biologically effective amount of an ECM signaling molecule-related biomaterial described above, thereby providing a control; (c) measuring the level of cell migration (proliferation) resulting from step (a) and from step (b); and (d) comparing the levels of cell migration (proliferation) measured in step (c), whereby a modulator of cell migration (proliferation) is identified by its ability to alter the level of cell migration (proliferation) when compared to the control of step (b). Preferred embodiments of the methods of screening for modulators of either cell migration or cell proliferation involve the use of fibroblasts presenting an α_fιβ, integrin and/or a sulfated proteoglycan.

Example 30 Cyrόl -Mediated Regulation of Gene Expression

Soluble Cyrόl protein added to primary human foreskin fibroblasts in culture elicits significant changes in gene expression within 24 hours. In contrast, immobilized Cyrόl is significantly less efficient in inducing such expression changes. Among the changes that occur is: (1) upregulation of matrix degrading enzymes, including MMP1. MMP3, and uPA; and (2) downregulation of matrix protein genes such as the type 1 collagen chain gene. Cyrόl also induces expression of the inflammation-related proteins IL-1 and lL-6. In addition, Cyrόl induces a substantial upregulation of the VEGF1 (vascular endothelial growth factor I or VEGF-A), a potent angiogenic factor, and VEGF3 (VEGF-C), an important angiogenic factor for the lymphatic system. These gene expression alterations indicate that Cyrόl is useful in screening assays designed to identify modulators of angiogenesis through detection of an effect on Cyrό l activity.

Purified, soluble, recombinant Cyrό l protein was added to cultures of primary human foreskin fibroblasts for 24 hours. From these cells, mRNA was isolated for preparation of probes to hybridize to a multi-gene blot (Clontech

Atlas Human Array 1.2 cat. no. 7850-1) containing about 650 genes. The levels of expression of these genes in Cyrό l-treated cells were compared to that of control cells. From this comparison, it was found that the expression of several groups of genes was altered: upregulation of matrix degrading enzymes including MMPl , MMP3, and uPA; downregulation of matrix protein genes such as the type 1 collagen chain gene; induction of the inflammation-related cytokines IL- 1 and IL-6; and induction of the angiogenic molecules VEGF1 and VEGF3. These findings have been confirmed by Northern blot analysis showing increases or decreases of the mRNAs in question in Cyrόl-treated cells. A time course of expression changes has been established. In particular, by 6 hours after its addition, Cyrόl has induced expression of VEGF -A mRNA and the induction is still evident at 12 and 24 hours post-addition. Also at 12 and 24 hours post- addition, Cyrόl has induced expression of VEGF-A polypeptide. With respect to VEGF-C, induction of mRNA is clearly evident 12 hours after addition of Cyrόl and the induction persists at least through 24 hours post-addition. Cyrό l is expected to induce the expression of VEGF-C polypeptide with similar kinetics.

Cyrόl appears to be the key mediator for the action of TGF-beta, which induces Cyrόl strongly and is known to regulate matrix protein synthesis. Using mouse embryo fibroblasts derived from Cyrόl knockout mice (see Example 31 ), it was shown that Cyrόl mediated TGF-beta function. Whereas TGF-beta can induce collagen expression in embryo fibroblasts derived from a cyrόl ^~ mouse (littermate of a cyrόl^'1' mouse), it cannot do so in cells that have an insertional inactivation of cyrόl (i.e., cvrόl knock-out cells). Also, TGF-beta can induce BEFG expression in cyril ¹' fibroblasts, but not in cyr6\^~'^~ fibroblasts. Stimulation of fibroblasts by serum is known to induce the expression of many genes. Whereas cyrόF^1' cells respond to serum stimulation with the induction of VEGF, cyrόl '^' cells do not exhibit serum induction of VEGF (although the background expression level is the same). This finding indicates that Cyrόl is the mediator of VEGF induction under stimulation by serum growth factors, and confirms the ability of Cyrόl to regulate VEGF expression. To prove that Cyrόl -mediated serum induction of VEGF occurs at the transcriptional level, collagen I and collagen II promoters linked to reporter genes were transfected into cyrόl knock-out cells and control cells. Consistent with the aforementioned results, transcription from the transfected gene constructs was induced by TGF-beta in control cells but not in knock-out cells.

Thus, when fibroblasts are stimulated, Cyrόl expression can lead to gene expression changes that produce proteins for matrix degradation and remodeling, cytokines that are chemotactic for macrophages and lymphocytes, and growth factors for angiogenesis. As noted above, Cyrόl polypeptides interact with the α₆β, integrin receptor of fibroblasts. Additional gene expression studies have shown that both Cyrόl NT (lacking the heparan binding domain IV) and dmCyrόl (containing substitutions in the two consensus XBBXB heparan binding motifs at amino acids 280-290 (H I) and 305-310 (H2) of SEQ ID NO:4) do not induce the expression of the above-referenced genes (e.g., VEGF, MMPl, and MMP3), in contrast to the effects of wild-type Cyrόl . Based on these results, the heparan binding domain of Cyrό 1 appears to be required for induction of gene expression in fibroblasts. It is expected that Cyrόl peptides such as TSP1 will inhibit gene expression resulting from Cyrόl induction mediated by the α₀β, integrin receptor. It is also expected that Cyrόl will induce the expression of these genes, and related genes, in other mammalian cell types, and that the heparan binding domain will be required for such induction regardless of cell type.

The identification of target genes regulated by Cyrόl, including genes involved in matrix remodeling (wound healing, metastasis, etc), inflammation, and angiogenesis, provides indications of suitable targets of therapy using Cyrόl .

Thus, in accordance with these findings, another aspect of the invention is a method of screening for a modulator of angiogenesis comprising the steps of:

(a) contacting a first endothelial cell comprising a cyrόl allele with a suspected modulator of angiogenesis; (b) measuring the Cyrόl activity of the first endothelial cell; (c) measuring the Cyrόl activity of a second endothelial cell comprising a cvrόl allele; and (d) comparing the levels of Cyrόl activity measured in steps (b) and (c), thereby identifying a modulator of angiogenesis. A related aspect of the invention is drawn to a method of screening for a modulator of angiogenesis comprising the steps of: (a) contacting a first endothelial cell with a polypeptide selected from the group consisting of a Cyrό 1 , a Fispl2, a CTGF, a NOV, an ELM-l (WISP-1), a WISP-3, a COP-1 (WISP-2), and fragments, analogs, and derivatives of any of the aforementioned members of the CCN family of proteins; (b) further contacting the first endothelial cell with a suspected modulator of angiogenesis; (c) contacting a second endothelial cell with the polypeptide of step (a); (d) measuring the angiogenesis of the first endothelial cell; (e) measuring the angiogenesis of the second endothelial cell; and (f) comparing the levels of angiogenesis measured in steps (d) and (e), thereby identifying a modulator of angiogenesis.

Another aspect of the invention is drawn to methods of treating conditions or disorders, such as diseases, associated with gene under- or over-expression by delivering a biologically or therapeutically effective amount of an ECM Signaling Molecule (e.g., a Cyrόl polypeptide, Fisp 12, CTGF), or modulator of a Cyrόl- integrin receptor interaction, using delivery means known in the art. A therapeutically effective amount of an ECM Signaling Molecule is that amount that results in mitigation of the gene under- or over-expression. Genes whose expression can be affected by an ECM Signaling Molecule include, but are not limited to, MMPl , MMP3, uPA, the type 1 collagen chain gene, the inflammation-related cytokines IL-1 and IL-6; and the genes encoding the angiogenic molecules VEGF 1 and VEGF3. Exemplary conditions, disorders and diseases include excessive or inadequate cell adhesion, such as various forms of fibrosis, defective angiogenesis, tumor growth, tumor metastasis, granulation tissue disorders, inflammatory responses, and certain muscle disorders known in the art.

The methods of identifying modulators of angiogenesis take advantage of the potential for modulators to influence angiogenesis by affecting the activity levels of a CCN protein such as Cyrόl by either influencing the level of expression of the protein or by influencing the specific activity of the expressed protein.

Example 31 Cyrόl Knock-out Mice

The mouse cyrόl gene was insertionally inactivated (i.e., knocked out) in vivo by targeted gene disruption and the phenotypes of heterozygous and homozygous knock-out mice were examined. Heterozygous mice (cyrόL'^") appeared to be normal, as these mice did not exhibit any apparent phenotype. The cyrόl^'1' homozygous mice, however, exhibited severe vascular defects and apparent neuronal defects as well. Most of the cyrόl^'1' mice died in utero, starting from E 10.5 through parturition, with most embryos dying around El 3.5. There is a spectrum of developmental defects and phenotypes at the time of embryonic death. The initial step in preparing knock-out mice was to construct a targeting vector that contained the mouse cyrόl gene insertionally inactivated by introducing the bacterial lacZ gene encoding β-galactosidase, which facilitated screening for knock-out mice. A commercially available 129 SvJ mouse genomic DNA library (Stratagene) was screened with a cyrόl probe and Clone 61 -9 was identi fied. Clone 61 -9 phage DNA was then prepared and digested with Stul and BamHI using conventional techniques. The 6 kb fragment containing the cyr<57 promoter and coding region was ligated to a blunt-ended Kpn\ linker, thereby attaching the linker to the Stul site. The fragment was then digested with BamHI and Kpnλ and inserted into BamHI, Kpnl digested pBluescript KS+. The recombinant pBluescript KS+ was cut with Smal and then ligated to an Xho linker. After linker ligation, the recombinant plasmid was cut with Xhol and the Xhol fragment bearing the lacZ coding region from pSAβgc./ (F iedrich et al , Genes Dev. 5: 1513-1532 ( 1991)) was inserted. The PGK-TK-blue plasmid containing a thymidine kinase gene driven by the PGK promoter (Mansour et al, Nature 33(5:348-352 ( 1988)) was cut with EcoRI and the ends were blunted with Klenow. The blunt-ended fragment was then ligated to Kpnl linkers. Finally, the cyrόl- Sgal-neo DNA and the modified PGK-TK DNA were each cut with Kpnl and ligated to generate pόl geo, the final targeting construct. Thus, pόlgeo contained functional βgal and neo coding regions flanked on the 5' side by a 1.7 kb fragment containing an intact cyrόl promoter and flanked on the 3' side by a 3.7 kb fragment containing the 3' end of the cyrόl coding region (exons 2-5 and 3' flanking sequence). Homologous recombination of this insert into the mouse chromosome would disrupt the cyrόl coding region and place the βgal and neo coding regions into the genome.

Cell culturing was performed according to Genome Systems instructions for mouse embryonic fibroblasts (MΕFs), or as described by Li et al, Cell (59:915-926 (1992), with modifications, for Jl ΕS cells. Briefly, MΕFs were cultured in 7.5% CO₂ in an incubator at 37°C with DMΕM (high glucose) medium (Gibco/BRL #1 1965-084) and 10% heat-inactivated Fetal Calf Serum (HyClone), 2 mM glutamine, 0.1 mM non-essential amino acids, and optionally with 100 U of Penicillin/Streptomycin. MΕFs were isolated from mouse embryos at Ε14.5 and supplied at passage 2.

For feeder cells, MEFs were mitotically inactivated by exposure to 10 μg/ml Mytomycin C(Sigma) in culture medium at 37°C (7.5% CO_?) for 2-5 hours. Cells were then washed 3 times with PBS. Mitotically inactivated MEFs were harvested with trypsin-EDTA(Gibco/BRL) and plated at about 1 x 107cm² with MEF medium. 1 embryonic stem(ES) cells were cultured in DMEM (no pyruvate, high glucose formulation; Gibco/BRL# 1 1965-084) supplemented with 15% heat inactivated FCS (Hyclone), 2 mM glutamine (Gibco/BRL), 0.1 mM non-essential amino acids(Gibco/BRL), 10 mM HEPES buffer (Gibco/BRL), 55 μM β- mercaptoethanol (Gibco/BRL), and 1 ,000 U/ml ESGRO (leukemia inhibitory factor, LlF)(Gibco/BRL). J l cells were routinely cultured in ES medium on a feeder layer of mitotically inactivated MEFs in a humidity saturated incubator at 37°C in 7.5%o CO₂. Normally, 1.5 x 10° Jl cells were seeded in a 25 cm² tissue culture flask and the medium was changed every day. Cell cultures were divided 2 days after seeding, usually when the flask was about 80% confluent. To dissociate ES cells, cells were washed twice with PBS (Ca- and Mg-free) and trypsinizedwith Trypsin/EDTA at 37°C for4 minutes. Cells were than detached, mixed with trypsin/EDTA thoroughly, and incubated for an additional 4 minutes. The cell suspension was then pipetted several (20-30) times to break up the cell clumps. A complete dissociation of cells was checked microscopically. ES cells were frozen with ES medium having 10% FCS and 10% DMSO(Sigma) at about 4-5 x 10⁶ cells/ml, with 0.5 ml/tube. Frozen cells were stored at -70°C overnight and transferred into liquid nitrogen the next day. Frozen cells were quickly thawed in a 37°C water bath, pelleted in 5 ml ES medium to remove DMSO, and plated in 25 cm² flasks with fresh MEF feeder cells.

To transfect mouse cells with a transgene, the pόlgeo targeting constmct was linearized by Notl digestion, suspended in PBS at 1 μg/ml, and introduced into Jl ES cells by electroporation. Rapidly growing (subconfluent, medium newly refreshed) Jl ES cells were trypsinized, counted, washed and resuspended in the electroporation buffer containing 20 mM HEPES, pH 7.0, 137 mM ΝaCl, 5 mM KC1, 6 M D-glucose, and 0.7 mM Νa₂HPO₄, at 1 x 10' cells/ml. Linearized DNA was added to the cell suspension at 45 μg/ml, mixed, and incubated at room temperature for 5 minutes. A 0.8 ml aliquot of cell-DNA mix was then transfeπed to a cuvette and subjected to electroporation with a BioRad Gene Pulser using a single pulse at 800 V, 3 μF. Cells were left in the buffer for 10 minutes at room temperature, and then plated at 4 x 10⁶ cells/100 mm plate with neomycin-rcsistant MEF feeder cells. Cells were then cultured under standard conditions without drug selection. After 24 hours, selection medium containing ES medium supplemented with 400 μg/ml (total) G418 (Gibco/BRL) and 2 μM Ganciclovir (Roche) was substituted. Selection medium was refreshed daily. Individual colonies were placed in microtiter wells and cells were dissociated with 25 μl 0.25% trypsin-EDTA /well on ice and subsequently incubated in a humidified incubator at 37°C with 7.5 CO₂, for 10 minutes. Cell suspensions were then mixed with 25 μl ES medium and pipetted up and down 10 times to break up clumps of cells. The entire contents of each well were then was transferred to a well in a 96-well flat-bottom dish with 150 μl of ES medium in each well and grown using conventional culturing techniques for 2 days.

Confluent ES cell clones were washed and treated with lysis buffer (10 mM Tris (pH 7.7), 10 mM NaCl, 0.5% (w/v) sarcosyl, and 1 mg/ml proteinase K) in a humid atmosphere at 60°C overnight. After lysis, a mixture of NaCl and ethanol (150 μl of 5 M NaCl in 10 ml of cold absolute ethanol) was added (100 μl/well) and genomic DNA was isolated. The genomic DNA of each ES cell clone was digested with EcoRI (30 μl/well) and subjected to Southern blot assay.

Southern blotting was preformed as described in "Current Protocols in

Molecular Biology" (Ausubel et al. [1999]). Briefly, EcoRI fragments of genomic DNA were fractionated by electrophoresis through 0.8% agarose gels and blotted onto nylon membranes (Bio-Rad) by downward capillary transfer with alkaline buffer (0.4 M NaOH). The probes, a BamHl-EcoR fragment 3' to the long arm of the targeting construct (pόl geo) or the neo coding region sequences, were prepared by random primer labeling (Prim-it II, Stratagene) using [a- ³²P] dCTP (NΕN). Membranes were prehybridi/εd in hybridization buffer (7% SDS, 0.5 M NaHPO₄ (pH 7.0), and 1 mM ΕDTA) at 65°C for 15 minutes in a rolling bottle. Fresh hybridization buffer was added with the probe and membranes were hybridized for 18 hours. Hybridized membranes were briefly rinsed in 5% SDS, 40 mM NaHPO₄ (pH 7.0), 1 mM ΕDTA and then washed for 45 minutes at 65°C with fresh wash solution. This wash solution was replaced with 1% SDS, 40 M NaHPO₄ (pH 7.0), 1 mM ΕDTA and washed twice for 45 minutes at 65°C with fresh solution. After washing, membranes were exposed to a screen, which was then scanned using a Phosphorlmager (Molecular Dynamics). Blots were routinely stripped and re-probed with the control neo probe to ensure that random integration had not occurred, using conventional techniques. Results of the Southern analysis showed that the genomic DNA of 14 colonies (231 colonies examined) contained a mutant cvrόl allele in a location consistent with integration via homologous recombination. The sizes of the detected fragments were 6.4 kb for the wild-type cyrόl allele and 7.4 kb for the mutant allele with the cyrόl probe; no band for the wild-type cyrόl allele and a 7.4 kb band for the mutant allele with the neo probe.

Genotyping was also done by PCR using a RoboCycler (Stratagene). Primers were designed to amplify a 2.1 kb DNA fragment from mutant alleles. The PCR product covers the 5'-flank of the short arm of the targeting construct through to the sequence oflacZ (β-gal) within the targeting constmct. The upper PCR primer sequence was 5'-CACAACAGAAGCCAGGAACC-3' (SEQ ID NO:24) and the lower PCR primer sequence was 5'-GAGGGGACGACGAC- AGTATC-3' (SEQ ID NO:25). PCR reaction conditions were 95°C for 40 seconds, 63°C for 40 seconds, and 68°C for one minute, for 35 cycles. For genotyping mouse tails or embryo tissues, two sets of primers were included in the same PCR reaction to amplify both wild-type and mutant alleles. A single upper PCR primer (b) was used, which had the sequence 5'- CAACGGAGCCAGGGGAGGTG-3' (SEQ ID NO:26). The lower PCR primer for amplifying the wild-type allele, lower wt primer, had the sequence 5'- CGGCGACACAGAACCAACAA-3' (SEQ ID NO:27) and would amplify a fragment of 388 bp. The lower PCR primer for amplifying the mutant allele was the lower mutant primer and had the sequence 5'-GAGGGGACG- ACGACAGTATC-3' (SEQ ID NO:28); a 600 bp fragment was amplified from mutant alleles. Reaction conditions were: 95°C for one minute, 63°C for one minute, and 72°C for one minute, for 30 cycles.

PCR amplification of mutant alleles of cyrό l using the mutant-specific primers produced a fragment of 2.1 kb and attempts to ampli fy the wild-type allele with those primers failed to produce a detectably amplified fragment, in agreement with expectations. Southern analyses identified a 7.4 kb band (mutant allele) and a 6.4 kb band (wild type) in heterozygous mutants; only the 6.4 kb band was detected when probing wild-type DNAs. Both the PCR data and the Southern data indicate that mutant cyrόl alleles were introduced into the mouse genome in a manner consistent with homologous recombination.

The selected ES cell clones were then expanded for micro-injection into E3.5 blastocysts from C57BL/6J mice. Embryo manipulations were carried out as described by Koblizek et al, Curr. Biol. 5:529-532 (1998) and Suri et al, Science 252:468-471 (1998), with modifications. Briefly, the Jl ES cell clones were harvested and dissociated with trypsin-EDTA. The cells were resuspended in CO₂-independent medium (Gibco-BRL) with 10% FBS and kept on ice. About 15-20 ES cells were injected into each blastocyst from C57BL/6J (Jackson Labs). Injected blastocysts were cultured for 1-2 hours prior to transfer into the uterine horns of pseudopregnant foster mothers (CD-I , Harlan). Chimeras were identified by coat color. Male chimeras with a high percentage of agouti coat color were caged with C57BL/6J females to test germ-line transmission of the ES-cell genotype. F, offspring carrying the targeted (i.e., mutant) allele were then back-crossed with C57BL/6J females for a few rounds to establish an inbred C57BL genetic background. In addition, a mutant mouse line having the inbred 129SvJ genetic background was obtained by mating germ-line chimera males with 129SvJ females. Five ES cell clones were injected and generated chimeric offspring with

ES cell contributions ranging from 30%o-100%, as judged by the proportion of agouti coat color. Four and two chimeric males derived from ES cell clones 4B7 and 2A1 1 , respectively, efficiently transmit the targeted allele through the germline. The cvrόl heterozygous mutant mice appeared healthy and fertile. The 4B7 chimeric line was either bred to 129SvJ mice to maintain the targeted allele in a Sv.1129 genetic background, or back-crossed with C57BL/6.I mice to transfer the mutation into the C57BL/6J background. The 2A1 1 targeted line was maintained in the 129SvJ genetic background. Similar phenotypes were exhibited by the 4B7,_2l), 4B7_C57B| , and 2A11 mouse lines. Among the offspring from intercrosses of cyrόl ^17" mice that were examined, 141 were cyrόl A 225 were cyrόl ^', and no homozygous cyrόl ^"'^"mice were observed at this age, except that 10 cyrόl^"'^" pups were bom alive and died within 24 hours of birth. Based on Mendelian ratios, the majority (>90%) of the cyr61 '^' animals should have died before birth. Thus, staged prenatal fetuses were examined by PCR, as described above. Starting from El 0.5, the numbers of homozygous mutant embryos were found to be less than expected based on a Mendelian ratio, which might have been due to resoφtion of homozygous mutant embryos. However, most (80%) of the E10.5 cyr6l '^' embryos appeared normal compared to littermates. At this stage (E10.5). the failure of chorioallantoic fusion was found in some embryos and this phenotype resulted in early embryonic lethality. The allantois of this type of embryo appeared ball-shaped and often was filled with blood. While no other defects were specifically identified, hemontiage began to appear in a few of the cyrόl -null embryos. At El 1.5, about 50% of cvrόl '^" embryos were indistinguishable from wild-type or heterozygous mutant littermates by appearance. By El 1.5, embryos lacking a chorioallantoic fusion were consistently deteriorated. Increasing numbers and severity of hemorrhage were also observed in cyrόl -null embryos. Hemorrhages occurred in different areas, including the placenta, intra-uterus, intra-amnion, embryo body trunks, body sides, and head. At this stage, placental defects were also found in some null mutant embryos. The placentae associated with these embryos showed a sub-standard vasculature network. Unlike the early lethality associated with the failure of chorioallantoic fusion, embryos with placental defects typically lived and developed normally. At El 2.5, cvrόl^"'^" embryos still presented three phenotypes: 1 ) unaffected,

2) alive with hemoπhage and/or placental defects, and 3) deteriorated, though with the proportion of categories changed from earlier stages. About 30% of the cyrόl -null embryos remained unaffected at this stage. About 50% of the null mutant embryos showed signs of hemoirhage and/or placental defects and 20% of this type of embryo did not survive the vascular or the placental defects. About 20% of cyrό 1^"Λ embryos did not have a chorioallantoic fusion and died at much earlier stages, as judged by the under-development of defective embryos. By El 3.5, none of the cyrόl^7" embryos that had shown hemorrhage, placental defects, or failure of chorioallantoic fusion were alive, although about 30% of the total Cyrόl -deficient embryos showed no apparent phenotype. Embryos examined at later stages (>E14.5) showed the same phenotypic pattern and the same proportion for each type of defect, but with increasing severity.

Additional investigation, at the cellular and sub-cellular levels, was perfoπned using the following techniques. MEF cells were harvested as described by Hogan et al, Manipulating the Mouse Embryo- A Laboratory Manual ( 1994). Briefly, E 11.5 embryos from crosses of two heterozygous cyr61 - targeted parents were dissected in DMEM without serum. The limbs, internal organs, and brain were removed. Embryo carcasses were then minced with a razor blade and dissociated with trypsin/EDTA at 37°C with rotation for 10 minutes. Half of the dissociation buffer was then added to an equal volume of DMEM plus 10% FBS. Dissociation and collection steps were repeated five times. Collected cells were expanded and split at a 1 : 10 ratio to select the proliferating fibroblast cells.

To prepare total cell lysates, a 100 mm plate of MEF cells was cultured to near confluency. Cells were activated with fresh medium containing 10% serum and incubated at 37°C for 1.5 hours before being harvested. Cells were then washed and centrifuged using conventional procedures. The cell pellets were resuspended in 100 μl RIPA buffer (0.5% sodium deoxycholate, 0.1 % SDS, 1 % Nonidet P-40, 50 mM Tris-Cl, pH 8.0, 150 mM NaCl, aprotinin 0.2 units/ml, and 1 mM PMSF) and put on ice for 10 minutes to lyse the cells. The cell suspension was centrifuged and the supernatant (cell lysate) was stored at -70°C for further analysis. One third of the supernatant was subjected to Western blot analysis using a TrpE-mCyrόl polyclonal anti-serum.

To confirm that homozygous cyr 1 ^''^' animals did not express Cyrό 1 , MEF cells were prepared from E l 1 .5 embryos resulting from intercrosses of two cyrόl^{+ '} parents. Cell lysates were collected from serum-stimulated MEFs of different genotypes and were subjected to Western blot analyses using anti-Cyrόl antiserum (trpE-mCyrόl ). The Western blot demonstrated that the Cyrόl protein level was not detectable in KO (knockout) MEF cells, while heterozygous cyrόl ^{+ "} cells expressed the Cyrόl protein at high levels under the same culture conditions and se m stimulation. The lack of expression of Cyrόl in cyr61^"'^" animals was further confirmed by Northern blot analyses, in which cyrόl mRNA was not detectable in serum-induced KO MEF cells. Thus, the null mutation of cyrόl^{" "} has been confirmed as eliminating Cyrόl expression at both the mRNA and protein levels.

Defects in placental development, a major cause of embryonic death in cyrόl^"'^" mice, were further analyzed. Histological analyses of mouse placentae generally followed Suri et al, (1998). Briefly, placentae from E 12.5 embryos were dissected in cold PBS and fixed with 4% parafoπnaldehyde in 0.1 M phosphate buffer (PB) at 4°C for overnight. Fixed placentae were then dehydrated through increasing concentrations of alcohol (50%>, 75%, 90%, 95%, and 100%) two times. Dehydrated tissue was then cleared with Hemo-De (a xylene alternative), 1 :1 ethanol/Hemo-De (Fisher), and 100% Hemo-De, and the clearing process was repeated. Cleared tissues were then equilibrated in a 1 :1 mixture of paraffin:Hemo-De at 60°C for one hour in a vacuum oven and the process was repeated. Tissues were embedded in paraffin with Histoembedder (Leica). The paraffin-embedded placentae were cut into 10 μm slices with a microtome (Leica). Finally, tissue sections were subjected to Harris' Hematoxylin and Eosin staining (Asahara et al, Circ. Res. 53:233-240 [1998]). Placentae for immunohistochemical staining were dissected in cold PBS and fixed in 4% paraformaldehyde at 4°C ovemight. Fixed tissue was transfeπed to 30% sucrose in PBS at 4°C overnight. Placentae were then embedded in O.C.T. (polyvinyl alcohol, carbowax solution) on dry ice. Frozen blocks were stored at -70°C or cut into 7 μm sections with a cryotome (Leica). Immuno- histochemical staining was done as recommended by the manufacturer (Zymed). Briefly. frozen sections were post-fixed with 100% acetone at 4°C for 10 minutes. Endogenous peroxidase was blocked with Peroxo-Block (Zymed). Sections were incubated with a 1 :250 dilution of biotinylated rat anti-mouse PECAM-1 (i.e., platelet endothelial cell adhesion molecule- 1) monoclonal antibody MEC 13.3 (Pharmingen) at 4°C overnight. A Histomouse-SP kit with Horse Radish Peroxidase (Zymed) was used to detect PECAM-1 signals.

The results of histological and immunohistochemical analyses showed that Cyrόl -null placentae contained a limited number of embryonic blood cells and were largely occupied by maternal blood sinuses. Abnormally compact trophoblastic regions were-also observed. PECAM-1 staining demonstrated the highly-vascularized labyrinthine zone in a heterozygous mutant placenta. Under higher magnification, flows of fetal blood cells within the PECAM- 1 -stained vessels were identified. Consistent with the variation in phenotypes among the Cyrόl -deficient embryos, the staining of placentae from numerous Cyrόl^"' embryos also reflected placental defects to various degrees. Nonetheless, the placental defects observed with PECAM-1 -staining can be classified into two groups, groups I and II. Group I of type II (type I- embryos with complete failure of chorioallantoic fusion not surviving E10.5; type II- embryos with partially defective chorioallantoic fusion surviving through about E 13.5) exhibits a set of placental defects that is characterized by the virtual absence of embryonic vessels, the presence of condensed trophoblasts, and the presence of a compressed labyrinthine zone. A higher magnification view confirms that no vessels developed in the labyrinthine with placental defects of this kind. Placentae with a group II defect showed fair amounts of PECAM- 1 -positive staining and condensed capillary structures. However, the PECAM-1 -stained vessel-like structures were degenerated and collapsed, with no fetal blood cells inside.

Thus, the lack of Cyrόl causes two types of placental defects. In type I, the failure of chorioallantoic fusion results in the loss of physical connection between the embryo and the placenta. In type II placental defects, the physical connection is established by successful chorioallantoic fusion. However, the embryonic vessels only reach to the surface of the placenta or, with successful penetration through the chorionic plate, develop an immature non- functional vascular structure in the labyrinthine zone.

X-gal staining was also used to assess embryonic development in various xyrό 1 backgrounds. (The targeting DNA, p61 geo, was designed to knock out the Cyrόl gene and also to "knock in" a -gal gene as a marker to reflect the expression of Cyrόl). X-gal (i.e., 5-Bromo-4-chloro-3-indolyl- β-D- galactopyranoside) staining for β-galactosidase expression was performed on heterozygous cyr61^+/" embryos staged from E9.5 to El 1.5. The staining was done as described (Suri et al, [1998]). Staged embryos were fixed in a 0.2% parafonnaldehyde solution at 4°C overnight. Fixed tissue was incubated in 30% sucrose in PBS plus 2 mM MgCl₂ at 4°C overnight. Tissue was then embedded in OCT on dry ice and cut with a cryotome into 7 μm sections. Frozen tissue sections were post-fixed in 0.2% paraformaldehyde and stained with X-gal (1 mg/ml) at 37°C for 3 hours in the dark. Slides were counter-stained with % Orange G. Stained slides were then serially dehydrated through increasing concentrations of methanol, cleared with Hemo-De, and slides were mounted.

X-gal staining of the E9.5 embryos, including the extra-embryonic tissues, showed β-galactosidase expression, driven by the cyr61 promoter, at the tip of the allantois adjacent to the chorion in the chorioallantoic placenta. The staining of more advanced El 0.5 embryos illustrated that large vessels branching from the allantoic vessels were developed in the chorionic plate and could easily be identified in the endothelial lining using X-gal. Further developed El 1.5 placenta showed the same expression pattern as El 0.5 embryos. While the staining was highly associated with the endothelium of the umbilical and chorionic vessels, no detectable staining in the labyrinthine zone, where a microvasculature network was developing, was seen at El 1.5. The presence of Cyrόl in the allantois at, and proximal to, the fusion surface with the chorion, and in the umbilical and chorionic vessels, further supports the important role of Cyrόl in angiogenesis. Cyrόl was involved in chorioallantoic fusion and was critical for proper angiogenic development as placentation progressed. Moreover, a staining of the El 1.5 embryo confirmed that Cyrό 1 was expressed in the paired dorsal aortae and the major arteries branching from the heart, which is consistent with the hemorrhaging seen in Cyrόl -null mutants. Apparent from the preceding description is another aspect of the invention, which is directed to a method of screening for modulators of angiogenesis comprising the steps of: (a) constructing a non-human transgenic animal comprising a mutant allele of a gene encoding a polypeptide selected from the group consisting of a Cyrόl, aFispl 2, a CTGF. aNOV, an ELM-1 ( WISP-1), a WISP-3, a COP-1 (WISP-2); (b) contacting the transgenic animal with a suspected modulator of angiogenesis; (c) further contacting a wild-type animal of the same species with the polypeptide, thereby providing a control; (d) measuring the levels of angiogenesis in the transgenic animal; (e) measuring the level of angiogenesis of the wild-type animal; and (f) comparing the levels of angiogenesis measured in steps (d) and (e), thereby identifying a modulator of angiogenesis.

Transgenic animals are characterized as described above and, based on such characterizations, a variety of genotypes may be usefully employed in the methods of the invention. For example, the transgenic animal may be either homozygous or heterozygous and the mutant allele may result in no expression (i.e., a null mutation) or altered activity levels. A preferred transgenic animal is a mouse, although any non-human vertebrate organism may be used, including other mammals (e.g., rat, rabbit, sheep, cow, pig, and horse, among others) or birds (e.g., chicken). A preferred transgene is an insertional inactivation, or knock-out, of a gene encoding a CCN protein (e.g., cyrόl); also preferred is an insertional inactivation resulting from the introduction, or "knocking in", of an identi fiable marker gene such as lαcZ encoding β-galactosidase. Of course, many transgene constructions are possible, including transgenes resulting from the replacement of wild-type sequence by related sequences that specify variant amino acid sequences. It should be understood from the preceding discussion that the invention comprehends gene therapy approaches involving the introduction of a transgene into a cell to treat any of a variety of conditions or disorders, such as diseases.

Also apparent from the preceding description is another aspect of the invention, which is drawn to a mammalian cell comprising a cyrόl mutation selected from the group consisting of an insertional inactivation of a cyrόl allele and a deletion of a portion of a cyrόl allele. The mammalian cell is preferably a human cell and the mutation is either heterozygous or homozygous. The mutation, resulting from insertional inactivation or deletion, is either in the coding region or a flanking region essential for expression such as a 5 ' promoter region. Cells are also found associated with non-human animals.

Example 32 Adhesion to Platelets and Macrophages

Platelet immobilization plays an important role in wound healing, for example by contributing to thrombosis in the process of stanching the flow of blood. Proteins of the CCN family, such as Cyrό l and Fisp l2/CTGF, promote platelet adhesion by interacting with the α_llbβ₃ integrin.

Recombinant Cyrόl and Fispl 2/mCTGF, synthesized in a Baculovirus expression system using Sf9 insect cells, were purified from serum-free conditioned media by chromatography on Sepharose S as described (Kireeva et al, [1997]; Kireeva, et al, [1996]). SDS-PAGE analysis of purified Cyrόl and

Fispl 2/mCTGF revealed the presence of single Coomassie Blue-stained bands of 40-kDa and 38-kDa, respectively. On immunoblots, the purified proteins reacted specifically with their cognate antibodies. Protein concentrations were determined using the BCA protein assay (Pierce) with bovine serum albumin

(BSA) as the standard.

Microtiter wells were coated with purified recombinant Fispl 2/mCTGF or Cyrόl , and the adhesion of isolated platelets to these proteins was detected with ^{l 2}T-mAbl 5, an anti-β₃ monoclonal antibody (Frelinger et cιl, J. Biol. Chem. 265:6346-6352 [1990]). Platelets were obtained from venous blood drawn from healthy donors and collected into acid-citrate-dextrose (ACD). Washed platelets were prepared by differential centrifugation as described (Kinlough-Rathbone et al. T romb. Haemostas. 2:291-308 [1997]), and finally resuspended in HEPES- Tyrode's buffer (5 mM HEPES, pH 7.35, 1 mM MgCl₂, 1 mM CaCl₂, 135 mM NaCl, 2.7 mM KC1, 11.9 mM NaHCO,, 1 mg/ml dextrose and 3.5 mg/ml BSA). The platelet concentration was adjusted to 3 x 10^s platelets/ml.

Microtiter wells (Immulon 2 removawell strips, Dynex Technologies, Inc.) were coated with Cyrόl, Fispl2/mCTGF, or fibrinogen (25 μg/ml, 50 μl/well) incubated overnight at 22°C, and then blocked with 3% BSA at 37°C for 2 hours. Washed platelets were added to the wells (100 μl/well) in the presence and absence of platelet agonists and incubated at 37°C for 30 minutes. The wells were washed with HEPES-Tyrode's buffer and adherent platelets were detected with ^{l !}XmAbl5, an anti-β, monoclonal antibody. Exposure to the labeled antibody (50 nM, 50 μl/well) proceeded for 1 hour at 22°C. After extensive washing with HEPES-Tyrode's buffer, bound radioactivity was determined by γ- counting. In inhibition studies, washed platelets were pre-incubated with blocking peptides or antibodies at 37°C for 15 minutes prior to addition to microtiter wells. In experiments to examine the effect of divalent cation chelation, EDTA (5 mM) was added to suspensions of washed platelets and pre- incubated at 37°C for 15 minutes.

The anti-β₃ antibody was radioiodinated with carrier-free Na^l25I (Amersham Corp.) using the IODO-BEADS iodination reagent (Pierce) to a specific activity of approximately 2 μCi/μg. This antibody binds equally to α_IIbβ₃ present on activated (see below) and unactivated platelets. As controls, BSA- and fibrinogen-coated (Kabi Vitrum Inc.) wells were also used. Initially, the adhesion of unactivated versus activated platelets to immobilized Fisp l2/mCTGF and Cyrό l was compared. To ensure that platelets were not activated during the washing procedures, PGI₂ (lOO nM), which inhibits activation by raising platelet cAMP levels, was added to the platelet suspensions. Unactivated platelets failed to adhere to either protein. However, activation of platelets with 0.1 U/ml thrombin, 500 nM U46619, or 10 μM ADP caused a dramatic increase in platelet adhesion to both Fispl2/mCTGF- and Cyrόl -coated wells. To confirm that the adhesion process was activation- dependent, PGI₂ (100 nM) was added with the agonists to prevent platelet activation. Under these conditions, platelet adhesion to both Fisp 12/mCTGF and Cyrόl was significantly inhibited.

For comparison, platelet adhesion to fibrinogen-coated wells was assessed. While unactivated platelets were capable of adhering to immobilized fibrinogen at a low level, platelet adhesion to Cyrόl and Fispl2/mCTGF appeared to be absolutely dependent on cellular activation. Following platelet activation with strong agonists such as thrombin and U46619, platelet adhesion to Cyrό 1 and Fisp 12/mCTGF was comparable to fibrinogen. The weaker agoni st, ADP, caused a lesser response. Because ADP does not induce secretion of α- granule proteins from washed human platelets and does not induce platelet aggregation in the absence of exogenous fibrinogen, ADP was used to induce platelet adhesion in subsequent experiments.

To further substantiate the activation-dependent adhesion of platelets to these proteins, an acid phosphatase assay designed to quantitate the relative numbers of adherent platelets was performed. This assay measured the acid phosphatase activity of adherent platelets. Following the adhesion and washing procedures described above, the substrate solution (0.1 mM sodium acetate, pH 5.0, 20 mM /7-nitrophenylphosphate, and 0.1% Triton X-100; 150 μl/well) was added and incubated for 2 hours at 37°C. The reaction was stopped by the addition of 20 μl 2N NaOH, and absorbance at 405 nm was measured. Both the ^l25I-mAbl 5 binding assay and the acid phosphatase assay for adhesion of ADP- stimulated platelets to fibrinogen, Fispl 2/mCTGF, and Cyrόl , produced similar results. Because the amounts of bound ^l25-mAbl 5 were directly proportional to the quantity of integrin α_πbβ₃ on the adherent platelets, the acid phosphatase assay was used in subsequent studies. The adhesion of ADP-activated platelets to Fisp l 2/mCTGF and Cyrόl was dose-dependent and saturable. In the presence of PGI₂, unactivated platelets adhered poorly to both proteins, even at high coating concentrations. The specificity of the adhesion process was characterized in inhibition studies using anti-peptide polyclonal antibodies raised against the central variable regions of Fispl2/mCTGF and Cyrόl . On immunoblots, rabbit polyclonal anti- Fispl 2/mCTGF and anti-Cyrόl, prepared as described in Example 29, reacted specifically with Fispl2/mCTGF and Cyrόl, respectively. No cross-reactivity was observed. In addition, anti-Fispl 2/mCTGF antibody inhibited platelet adhesion to Fisp 12/mCTGF, but not to Cyrό 1 , and anti-Cyrό 1 antibody inhibited Cyrόl -mediated platelet adhesion but not that mediated by Fispl2/mCTGF. No inhibition was observed with normal rabbit IgG. Also, neither anti- Fispl 2/mCTGF antibody nor anti-Cyrόl antibody inhibited platelet adhesion to fibrinogen-coated wells. Thus, the abilities of Fispl 2/mCTGF and Cyrόl to mediate platelet adhesion are intrinsic properties of these proteins.

Upon platelet activation, the ligand binding affinities of integrins α_llbβ₃ and _vβ₃ are upregulated (Shattil, et al, Blood 91:2645-2651 [1998 , Bennett, et al, J. Biol Chem. 272:8137-8140 [1997]). To determine whether these integrin receptors mediated platelet adhesion to Fisp 12/mCTGF and Cyrό 1 , the inhibitory potentials of peptide antagonists and the divalent cation chelator, EDTA, were tested. Preincubation of platelets with EDTA at 37°C completely abolished platelet adhesion to both proteins, indicating that the adhesion process was divalent cation dependent. Cation dependency of adhesion is consistent with the involvement of an integrin receptor. The major platelet integrin, α_πbβ₃, is sensitive to inhibition by RGD- containing peptides and the dodecapeptide H₁₂ (H₂N-HHLGGAKQAGDV- CO₂H, SEQ ID NO:29; Research Genetics Inc.) derived from the fibrinogen γ chain (Plowetal, Proc. Natl Acad. Sci. (USA) 52:8057-8061 [1985]; Lam, etal, J. Biol Chem. 262:941-950 [1987]). The adhesion of ADP-activated platelets to Cyrόl and Fispl2/mCTGF was specifically inhibited by GRGDSP (SEQ ID NO:31), but not by GRGESP (SEQ ID NO:32). (Peninsula Laboratories). Likewise, the RGD-containing snake venom peptide echistatin (Gan et al, J. Biol. Chem. 263: 19821- 9832 [1988]; Sigma Chemical Co.) also completely blocked platelet adhesion to both proteins. It has also been shown that the dodecapeptide H₁₂ preferentially interacts with integrin α_llbβ₃ as compared to integrin α_vβ₃ (Cheresh et al, Cell 55:945-953 [1989]; Lam et al, J. Biol. Chem. 264:3142-3149 [1989]). Thus, the finding that H₁₂ inhibited platelet adhesion to Cyrόl and Fispl2/mCTGF indicated that this process was mediated by α_πbβ₃ rather than α_vβ₃. While the complex-specific monoclonal antibody AP-2 (anti- a _nβf, Pidardetal., J. Biol. Chem. 255:12582-12587 [1983]) completely blocked the adhesion of ADP-activated platelets to Fispl2/mCTGF and Cyrόl, no inhibition was observed with LM609 (anti-α_vβ₃; Cheresh et al, J. Biol. Chem. 262:11103-111 1 [1987]) or with normal mouse IgG. In control samples, the adhesion of ADP-activated platelets to fibrinogen was also completely inhibited by EDTA, RGDS, echistatin, H₁₂ or AP-2, but not by RGES or LM609. These results indicate that platelet adhesion to these proteins is mediated through interaction with activated integrin α_IIbβ₃.

A solid-phase binding assay to detect the receptor-ligand interactions showed that α_IIbβ₃ binds directly to Fispl2/mCTGF and Cyrόl . In these experiments, activated and unactivated α_IIbβ₃ were purified from platelet lysates. Activated α_Ilbβ₃ was purified by RGD affinity chromatography, as described (Knezevic et al, J. Biol. Chem. 277:16416-16421 [1996]). Briefly, outdated human platelets were isolated by differential centrifugation and solubihzed in lysis buffer (10 mM HEPES, pH 7.4, 0.15 M NaCl, containing 1 mM CaCl₂, 1 mM MgCl₂ 100 μM leupeptin, 1 mM phenylmethylsulfonyl fluoride, 10 mM -ethylmaleimide, and 50 mM octylglucoside). The octylglucoside extract was incubated with 1 ml GRGDSPK-coupled Sepharose 4B overnight at 4°C. After washing with 15 ml column buffer (same as lysis buffer except it contained 25 mM octylglucoside), bound α_1Ibβ₃ was eluted with 1.7 mM H,₂ (2 ml) in column buffer. The H_I2 eluate was applied to a Sephacryl S-300 High Resolution column ( 1.5 x 95 cm), and α_llbβ₃ was eluted with 10 mM HEPES, pH 7.4, 0.15 M NaCl, 1 mM CaCl₂, 1 mM MgCl₂ and 25 M octylglucoside.

Unactivated α_Ilbβ₃ was isolated by the method of Fitzgerald et al, Anal. Biochem. 757: 169-177 (1985), with slight modifications. The flow-through fraction of the GRGDSPK-Sepharose column was applied to a concanavalin A- Sepharose 4B column (1 x 20 cm). Unbound proteins were washed with 50 ml column buffer, and bound α_IIbβ₃ was then eluted with 100 mM mannose dissolved in column buffer. Fractions containing α_Ilbβ₃ were further purified on a Sephacryl S-300 High Resolution column To perform the solid-phase binding assay, purified α_Ilbβ₃ was added to wells coated with either Cyrό 1 or Fi sp 12 (mCTGF) in the presence or absence o f inhibitors and binding was allowed to proceed for 3 hours at 37°C. Unbound receptor was removed and the wells wee washed twice with HEPES-Tyrode's Buffer. The binding of purified α _bβ₃ to Cyrόl or Fispl2/mCTGF immobilized onto microtiter wells was detected with ^l25I-mAbl5.

Both activated and unactivated α_IIbβ₃ were indistinguishable on SDS- PAGE analysis as detected by silver staining. However, activated α_1Ibβ₃, but not the unactivated receptor, was capable of binding to immobilized fibrinogen. Likewise, greater binding of activated versus unactivated α_IIbβ₃ to Fisp 12/mCTGF and Cyrόl was observed. In contrast, the background bindings of activated and unactivated α_Ilbβ₃ to control wells coated with BSA were similar. Thus, activated, but not unactivated, platelets adhered to Cyrόl and Fispl 2/mCTGF.

To further characterize the interaction of α_1Ibβ₃ with Fispl2/mCTGF and Cyrόl , binding isotherms were determined for varying concentrations of RGD- affinity purified α_IIbβ₃. These binding isotherms showed that the dose-dependent binding of activated α_IIbβ₃ to Fispl2/mCTGF and Cyrόl was saturable with half- saturation occurring at 15 nM and 25 nM α_ιn,β₃, respectively. Again, no significant binding of α_I[bβ₃ to control BSA-coated wells was observed. To demonstrate the specificity of the interaction, inhibition studies were performed. As expected, the binding of activated α_IIbβ₃ to Fispl2/mCTGF and Cyrόl was specifically blocked by RGDS but not by RGES. Furtheπnore, echistatin and the H₁₂ peptide also effectively inhibited α_Mbβ₃ binding to these proteins. These findings are consistent with results obtained in the platelet adhesion assay. Collectively, these functional and biochemical data demonstrate that activated integrin _|Ibβ₃ is the receptor mediating activation-dependent platelet adhesion to Cyrόl and Fispl2/mCTGF.

Thus, another aspect of the invention is a method of screening for modulators of wound healing comprising the steps of: (a) contacting a first activated platelet with a polypeptide of the CCN family, such as Cyrόl , and a suspected modulator; (b) further contacting a second activated platelet with the polypeptide of step (a); (c) measuring the binding of the first activated platelet to the polypeptide; (d) measuring the binding of the second activated platelet to the polypeptide; and (e) comparing the binding measurements of steps (d) and (e), thereby identifying a modulator of wound healing. Preferably, the wound healing involves the participation of platelet binding in the process of blood clotting. Also preferred are platelets presenting the α_llbβ₃ integrin.

In addition to the above-described binding properties of members of the CCN family of proteins, antibody inhibition studies with anti-α_M and anti-β₃ antibodies have shown that Cyrόl binds to macrophages via yet another integrin, the α_Mβ₂ integrin. Based on these results, it is expected that mammalian CCN proteins, such as human or mouse Cyrόl, will bind to the macrophages of mammals. It is also expected that Cyrόl will promote the migration of macrophages, thus serving a role in attracting and retaining macrophages at the site of a wound. Consequently, Cyrόl is expected to play a role in the inflammatory response of mammals, and modulation of Cyrόl activity is expected to influence the inflammatory response.

Yet another aspect of the invention is a method of screening for modulators of macrophage adhesion comprising the steps of: (a) contacting a first macrophage with a polypeptide of the CCN family, such as Cyrόl , and a suspected modulator; (b) further contacting a second macrophage with the polypeptide of step (a); (c) measuring the binding of the first macrophage to the polypeptide; (d) measuring the binding of the second macrophage to the polypeptide; and (e) comparing the binding measurements of steps (d) and (e), thereby identifying a modulator of macrophage adhesion.

Example 33

Peptide Modulators of ECM Signaling Molecule Activity

Screening assays for modulators of cell adhesion are designed to identify modulators (e.g., inhibitors or activators) of ECM signaling molecule activities, such as Cyrόl activities, involved in angiogenesis, chondrogenesis, oncogenesis, cell adhesion, cell migration or cell proliferation. In developing a screening assay for modulators of cell adhesion, candidate peptide modulators were designed as described in Example 12. In particular, Cyrόl peptide fragments were designed with the expectation that such peptides would modulate Cyrόl binding to an integrin receptor such as the α₆β, integrin. To facilitate success with the approach to peptide design described in

Example 12, experiments were conducted to refine the location of domains involved in one or more of the activities of Cyrόl, such as angiogenic activities. Initially, a deletion of mouse cyrόl that removed domain IV of the encoded Cyrόl (corresponding to amino acids 282-381 of SEQ ID NO:4) was generated. For a general description of Cyrόl domains 1-4, see Lau et al., Exp. Cell Research 248:44-57 (1999), incoφorated herein by reference in its entirety. The deletion constmct, cloned using a conventional Baculovirus system, was expressed and was subjected to a rat co eal implant assay as generally described in Example 19, and the results indicated that the truncated polypeptide induced angiogenesis. Further, the truncated polypeptide was subjected to an in vitro endothelial cell migration assay as described in Example 15, and the polypeptide containing Cyrόl domains I, II and III induced cell migration.

To further localize domains involved in angiogenesis, including endothelial cell migration, the coding regions for domains I, II and III of Cyrόl were separately fused to a coding sequence for glutathione S-transferase (GST), again using conventional technology. The individual fusion constructs were introduced into bacterial hosts, expressed, and purified with glutathione columns, using conventional techniques. Cell adhesion assays using endothelial cells and fibroblasts showed that, in isolation, only domain III (corresponding to amino acids 212-281 of SEQ ID NO:4) supported cell adhesion, and the fusion polypeptide containing this domain, when immobilized, supported the adhesion of both resting endothelial cells and fibroblasts.

To localize the relevant domain still further, several overlapping peptides were synthesized. The peptides, each having a sequence set forth in one of SEQ ID NOS: 33-38, show significant similarity to an ECM signaling molecule such as mouse Cyrόl (SEQ ID NO:2) or human Cyrόl (SEQ ID NO:4). These peptides were then separately tested for their capacities to inhibit Cyrόl adhesion to endothelial cells or fibroblasts. Peptides were separately added at various concentrations to Cyrόl - coated wells containing endothelial cells or fibroblasts, as described above. The peptide designated TSP1 (SEQ ED NO:33) inhibited Cyrόl adhesion to endothelial cells or fibroblasts at concentrations of 25 μM or higher. None of the other peptides tested (SEQ ID NO. 34-38) showed any inhibition of cell adhesion to Cyrόl. To determine if TSP1 was selectively inhibiting Cyrόl binding to α₅β₃or α₆β,, two integrins found on endothelial cells that bind Cyrόl, specific monoclonal antibody blocking studies were performed. These studies involved the separate addition of monoclonal antibody GoH3 (α₆-specific, including α₆β,; see Example 29) or LM609 (α₅β₃ specific; see Example 29) to cells prior to addition of the mixture to Cyrόl -coated wells in the presence or absence of the peptide under study. The results showed that TSP1 inhibited Cyrόl binding to the α₆β, integrin, a major Cyrόl receptor active in resting, or unactivated, endothelial cells and in fibroblasts. TSP1 did not inhibit Cyrόl binding to α_vβ₃, a major Cyrόl receptor in activated endothelial cells. Endothelial cells interacting with Cyrόl require integrin α_vβ₃ for cell migration and cell survival. However, to form tubules in vitro in a matrigel, integrin α₆β, is required. Thus, the ability of TSP1 to inhibit tubule formation in vitro using a matrigel assay as described in Example 15 (and as known in the art, Davis et al., Exp. Cell Research 216:1 13-123 (1995), incoφorated herein by reference in its entirety) was performed. As expected, TSP1 inhibited tubule formation in the matrigel assay.

To confirm that the TSP 1 peptide was inhibiting Cyrόl activity by directly interacting with the α₆β , , integrin, TSP 1 and the peptides col lectively having SEQ ID NOS: 34-38 were used in the aforementioned cell adhesion assay in the absence of immobilized Cyrόl . In these assays, wells were coated with one of the peptides and the immobilized peptides were then separately exposed to either endothelial cells (resting or activated) or fibroblasts. The results showed that TSP 1 , by itself, supported adhesion of endothelial cells and fibroblasts. None of the other tested peptides (having SEQ ID NOS: 34-38, collectively) supported cell adhesion in these assays. However, it is expected that peptides comprising sequences from domain II (von Willebrand Factor domain) of Cyrόl will inhibit the interaction between Cyrόl and the α_vβ₃ integrin receptor, thereby inhibiting the participation of endothelial cells in the process of angiogenesis. Such peptides are expected to show at least 95%>, and preferably 98%, similarity to a subsequence of the sequence set forth from amino acid 93 to amino acid 211 of SEQ ID NO:4 using the comparison algorithm of Altschul et al. used for BLAST searching in the GenBank nucleotide database (http://www.ncbi.nlm.nih.gov/) with default settings in place. Peptides exhibiting activity are expected to be at least seven amino acids in length, with no upper bound, although the most economical peptides will probably have no more than 20 amino acids.

This example establishes that peptide modulators of ECM signaling molecules, such as mouse and human Cyrόl can be identified using the in vitro angiogenesis and in vitro cell adhesion assays of the invention. The modulators themselves, showing promise as potential therapeutics for diseases and conditions related to angiogenesis, chondrogenesis, and cell adhesion, and for oncogenesis, cell migration and cell proliferation, constitute another aspect of the invention.

One of ordinary skill in the art, apprised of the information disclosed in this application, could readily identify other peptides having Cyrόl modulating activity by designing candidate peptides having sequence similarity to an ECM signaling molecule such as Cyrόl . Sequence variation is guided by a knowledge of conservative amino acids substituting as described above, peptide length may be varied between about 8-50 amino acid residues. Further, peptide derivatives

(e.g. , glycosylated, PEGylated, phosphorylated, may be used, as would be known in the art.

Numerous modifications and variations in the practice of the invention as illustrated in the above examples are expected to occur to those of ordinary skill in the art. Consequently, the illustrative examples are not intended to limit the scope of the invention as set out in the appended claims.

Claims (53)

CLA1MSWhat is claimed is:

1. A method of screening for a modulator of cell adhesion comprising the steps of: (a) contacting a first fibroblast cell with a suspected modulator of cell adhesion and a biologically effective amount of an ECM signaling molecule-related biomaterial selected from the group consisting of a human Cyrόl, a human Cyrόl fragment, a human Cyrόl analog, and a human Cyrόl derivative;

(b) separately contacting a second fibroblast cell with a biologically effective amount of an ECM signaling molecule-related biomaterial selected from the group consisting of a human Cyrόl, a human Cyrόl fragment, a human Cyrόl analog, and a human Cyrόl derivative, thereby providing a control;

(c) measuring the level of cell adhesion resulting from step (a) and from step (b); and

(d) comparing the levels of cell adhesion measured in step (c), whereby a modulator of cell adhesion is identified by its ability to alter the level of cell adhesion when compared to the control of step (b).

2. The method according to claim 1 wherein said first and second fibroblast cells comprise an α₆β, integrin.

3. The method according to claim 1 wherein said first and second fibroblast cells each comprise a heparan sulfate sulfated proteoglycan.

4. A method of screening for a modulator of cell migration comprising the steps of: (a) forming a gel matrix comprising Cyrόl and a suspected modulator of cell migration;

(b) preparing a control gel matrix comprising Cyrόl ;

(c) seeding fibroblast cells presenting an α₆β, integrin onto the gel matrix of step (a) and the control gel matrix of step (b);

(d) incubating said fibroblast cells;

(e) measuring the levels of cell migration by inspecting the interior of said gel matrix and said control gel matrix for cells;

(f) comparing the levels of cell migration measured in step (e), whereby a modulator of cell migration is identified by its ability to alter the level of cell migration in the gel matrix when compared to the level of cell migration in the control gel matrix.

5. The method according to claim 4 wherein said first and second fibroblast cells comprise an α₆β, integrin.

6. The method according to claim 4 wherein said first and second fibroblast cells comprise a heparan sulfate proteoglycan.

7. The method according to claim 4 wherein said fibroblast cells are human cells.

8. The method according to claim 4 wherein said matrix is selected from the group consisting of Matrigel, collagen, and fibrin.

9. A method of screening for a modulator of fibroblast cell proliferation comprising the steps of: (a) contacting a first fibroblast cell presenting an α₆β, integrin with a suspected modulator and a biologically effective amount of an ECM signaling molecule- related biomaterial selected from the group consisting of a human Cyrόl, a human Cyrόl fragment, a human

Cyrόl analog, and a human Cyrόl derivative;

(b) separately contacting a second fibroblast cell presenting an α₆β, integrin with a biologically effective amount of an ECM signaling molecule-related biomaterial selected from the group consisting of a human Cyrόl, a human Cyrόl fragment, a human Cyrόl analog, and a human Cyrόl derivative, thereby providing a control;

(c) incubating said first and second fibroblast cells; (d) measuring the level of cell proliferation resulting from step (c); and (e) comparing the levels of cell proliferation measured in step (d), whereby a modulator of cell proliferation is identified by its ability to alter the level of cell proliferation when compared to the control of step (b).

10. The method according to claim 9 wherein said first and second fibroblast cells comprise an α₆β, integrin.

1 1. The method according to claim 9 wherein said first and second fibroblast cells comprise a heparan sulfate proteoglycan.

12. A method of screening for a modulator of angiogenesis comprising the steps of:

(a) contacting a first endothelial cell comprising a cyrόl allele with a suspected modulator of angiogenesis; (b) measuring the Cyrόl activity of the first endothelial cell;

(c) measuring the Cyrόl activity of a second endothelial cell comprising a cyrόl allele; and

(d) comparing the levels of Cyrόl activity measured in steps (b) and (c), thereby identifying a modulator of angiogenesis.

13. A method of screening for a modulator of angiogenesis comprising the steps of:

(a) contacting a first endothelial cell with a polypeptide selected from the group consisting of a Cyrόl, a Fisp 12, a CTGF, a NOV, an ELM- 1 (WISP- 1 ), a WISP-3 , a COP- 1 ( WISP-

2), and fragments, analogs, and derivatives of any of the aforementioned members of the CCN family of proteins;

(b) further contacting the first endothelial cell with a suspected modulator of angiogenesis; (c) contacting a second endothelial cell with the polypeptide of step (a);

(d) measuring the angiogenesis of the first endothelial cell;

(e) measuring the angiogenesis of the second endothelial cell; and (f) comparing the levels of angiogenesis measured in steps (d) and (e), thereby identifying a modulator of angiogenesis.

14. A method of screening for modulators of angiogenesis comprising the steps of:

(a) constructing a transgenic animal comprising a mutant allele of a gene encoding a polypeptide selected from the group consisting of a Cyrόl, a Fispl2, a CTGF, a NOV, an ELM-1 (WISP-1), a WISP-3, a COP-1 (WISP-2); (b) contacting the transgenic animal with a suspected modulator of angiogenesis;

(c) further contacting a wild-type animal with said polypeptide, thereby providing a control; (d) measuring the levels of angiogenesis in the transgenic animal;

(e) measuring the level of angiogenesis of the wild-type animal; and

(f) comparing the levels of angiogenesis measured in steps (d) and (e), thereby identifying a modulator of angiogenesis.

15. A method of screening for modulators of wound healing comprising the steps of:

(a) contacting a first activated platelet with a polypeptide of the CCN family, such as Cyrό 1 , and a suspected modulator; (b) further contacting a second activated platelet with the polypeptide of step (a);

(c) measuring the binding of the first activated platelet to the polypeptide;

(d) measuring the binding of the second activated platelet to the polypeptide; and

(e) comparing the binding measurements of steps (d) and (e), thereby identifying a modulator of wound healing.

16. The method according to claim 15 wherein the platelets present the α_1Ibβ₃ integrin.

17. A method of screening for modulators of macrophage adhesion comprising the steps of: (a) contacting a first macrophage with a Cyrόl and a suspected modulator;

(b) further contacting a second macrophage with said polypeptide of step (a); (c) measuring the binding of the first macrophage to said polypeptide;

(d) measuring the binding of the second macrophage to said polypeptide; and

(e) comparing the binding measurements of steps (d) and (e), thereby identifying a modulator of macrophage adhesion.

18. A mammalian cell comprising a cyr61 mutation selected from the group consisting of an insertional inactivation of a cyr61 allele and a deletion of a portion of a cyrόl allele.

19. The cell according to claim 18 wherein said mutation is a homozygous mutation.

20. The cell according to claim 18 wherein said mutation is in the coding region of cyrόl .

21. The cell according to claim 18 wherein said cell is associated with an organism.

22. The cell according to claim 18 wherein said cell is human.

23. The cell according to claim 18 wherein said cell is a fibroblast.

24. An isolated human Cyrόl fragment comprising a sequence selected from the group consisting of residues 280-290 of SEQ ID NO:33 and residues 305-310 of SEQ ID NO:33, wherein said Cyrόl fragment retains at least one biological function of human Cyrόl .

25. The human Cyrόl fragment according to claim 24 comprising the sequence set forth in SEQ ID NO:33 and retaining at least one biological function of human Cyrό 1.

26. A human Cyrόl fragment comprising a sequence that is at least 95%> similar to the sequence set forth in SEQ ID NO:33, wherein said human Cyrόl fragment retains at least one biological function of human Cyrόl .

27. A human Cyrόl fragment encoded by a polynucleotide comprising a sequence that is at least 95%> similar to a polynucleotide encoding a polypeptide having the sequence set forth at SEQ ID NO:33, wherein said polypeptide retains at least one biological function of human Cyrόl .

28. A method for modulating gene expression comprising the step of administering a biologically effective amount of a human Cyrόl fragment according to claim 24 to a cell capable of Cyrόl -modulated gene expression.

29. A method of screening for a modulator of binding between a Cyrόl polypeptide and a Cyrόl receptor integrin selected from the group consisting of α_vβ₃, a_vβ₅, α₆β,, α_πβ₃ and α_Mβ₂, comprising the steps of: a) contacting a Cyrόl polypeptide with an integrin receptor composition in the presence and in the absence of a potential modulator compound; b) detecting binding between the polypeptide and the integrin receptor; and c) identifying a modulator compound in view of decreased or increased binding between the Cyrόl and the integrin receptor in the presence of the modulator, as compared to binding in the absence of the modulator.

30. The method according to claim 29 wherein said Cyrόl polypeptide lacks amino acid residues 282-381 of SEQ ID NO:4.

31. The method according to claim 29 wherein said Cyrόl receptor integrin is α_vβ₃.

32. The method according to claim 29 wherein said Cyrόl receptor integrin is α_vβ₅.

33 The method according to claim 29 wherein said Cyrόl receptor integrin is α₆β,

34. The method according to claim 29 wherein said Cyrόl receptor integrin is α„β₃

35. The method according to claim 29 wherein said Cyrόl receptor integrin is α_Mβ₂.

36. A purified compound that is a modulator of binding between a Cyrόl polypeptide and an integrin receptor, said compound identified by the method according to claim 29.

37. The use of a modulator of a Cyrόl -integrin α_vβ₃ interaction for preparation of a medicament for the treatment of a condition selected from the group consisting of atherosclerosis, heart disease, tumor growth, tumor metastasis, fibrosis, disorders associated with inadequate angiogenesis, disorders associated with aberrant granulation tissue development, abeπant fibroblast growth and wounds.

38. A method for treating a condition characterized by defective connective tissue in a mammalian subject, comprising the steps of:

(a) identifying a mammalian subject in need of treatment for said condition, and

(b) administering to said mammalian subject a composition comprising a modulator of an interaction between Cyrόl and an integrin receptor selected from the group consisting of α_vβ₃, α_vβ₅, α₆β,, α,,β₃ and α_Mβ₂, in an amount effective to mitigate the symptoms of the condition in said mammalian subject.

39. The method according to claim 38 wherein said modulator is selected from the group consisting of heparin, heparan sulfate and a polypeptide that binds an integrin receptor.

40. The method according to claim 39 wherein said administering is accomplished by expressing an exogenous polypeptide coding region in cells of the type affected by said condition.

41. The method according to claim 39 wherein said polypeptide is a fragment of a Cyrόl polypeptide.

42. The method according to claim 39 wherein said fragment of a Cyrόl polypeptide comprises a sequence selected from the group consisting of residues 280-290 of SEQ ID NO:4 and residues 306-312 of SEQ ID NO:4.

43. The method according to claim 38 wherein said condition involves a defect in fibroblast adhesion and said integrin receptor is α₆β,.

44. The method according to claim 38 wherein said condition involves a defect in fibroblast chemotaxis and said integrin receptor is α_vβ₅.

45. The method according to claim 38 wherein said condition involves a defect in fibroblast proliferation and said integrin receptor is α_vβ₃.

46. A method for treating a condition characterized by a defect in smooth muscle tissue in a mammalian subject, comprising the steps of:

(a) identifying a mammalian subject in need of treatment for said condition, and

(b) administering to said mammalian subject a composition comprising a modulator of an interaction between Cyrόl and an α₆β, integrin receptor in an amount effective to mitigate the symptoms of the condition in said mammalian subject.

47. The method according to claim 46 wherein said modulator is selected from the group consisting of heparin, heparan sulfate and a polypeptide that binds an α₆β, integrin receptor.

48. The method according to claim 47 wherein said administering is accomplished by expressing an exogenous polypeptide coding region in cells of the type affected by said condition.

49. The method according to claim 46 wherein said polypeptide is a fragment of Cyrόl.

50. The method according to claim 48 wherein said fragment comprises a sequence selected from the group consisting of residues 280-290 of SEQ ED NO:4 and residues 306-312 of SEQ ID NO:4.

51. The method according to claim 46 wherein said condition is selected from the group consisting of atherosclerosis, heart disease, tumor growth, tumor metastasis, fibrosis, disorders associated with inadequate angiogenesis, disorders associated with abeπant granulation tissue development, aberrant fibroblast growth and wounds.

52. A kit for assaying for Cyrόl -integrin receptor interactions, said kit comprising a Cyrόl polypeptide and a composition comprising an integrin receptor selected from the group consisting of α_vβ₃, α_vβ₅, α₆β,, α„β₃ and α_Mβ₂.

53. The kit according to claim 52 wherein said integrin receptor composition comprises an α_vβ₃ integrin localized in a mammalian fibroblast cell membrane.