WO1991006668A1

WO1991006668A1 - Monoclonal antibodies to basic fibroblast growth factor that inhibit its biological activity

Info

Publication number: WO1991006668A1
Application number: PCT/US1990/005658
Authority: WO
Inventors: Janet Lynn Dzubow; William Fitts Herblin; Thomas Michael Reilly
Original assignee: The Du Pont Merck Pharmaceutical Company
Priority date: 1989-10-27
Filing date: 1990-10-11
Publication date: 1991-05-16
Also published as: AU6605690A

Abstract

Monoclonal antibodies to human bFGF and fragments of such antibodies that inhibit the biological activity of bFGF, including in vitro mitogenic activity and in vivo angiogenic activity. Hybridomas which express such antibodies. Use of such antibodies in studying bFGF and in treatment of angiogenic diseases.

Description

TITLE Monoclonal Antibodies to Basic Fibroblast Growth Factor That Inhibit Its Biological Activity BACKGROUND OF THE INVENTION

Basic fibroblast growth factor (bFGF) is a heparln-binding polypeptide mitogen that has been isolated from a variety of cells and tissues (Esch et al. (1985) Biochem. Blophvs. Res. Co mun. 133, 554-562; Lobb et al. (1985) Biochemistry 24, 4969- 4973). bFGF is a potent mitogen for a wide range of cell types including endothelial cells, fibroblasts, chondrocytes, and glial cells (Esch et al . (1985) Biochem. Biophvs. Res. Commun. 133. 554-562). Basic FGF has been identified 1n a wide variety of normal and malignant tissues (Klagsbrun et al . (1986) Proc. Nat! . Acad. Sci. USA 83. 2448-2452; Moscatelli et al. (1986) Proc. Natl. Acad. Sci . USA 83, 2091-2095; Neufeld et al. (1987) Endocrinology 121. 597-603; Schweigerer et al . (1987) Biochem. Biophvs. Res. Commun. 143. 934-940; Schweigerer et al. (1987) Nature 325. 257-259; Schweigerer et al . (1987) Endocrinology 120. -796-800; Schweigerer et al. (1987) Proc. Natl. Acad. Sci. USA 84. 842-846; Vlodavsky et al . (1987) Proc. Natl. Acad. Sci. USA 84. 2292-2296) and many cell types reported to synthesize bFGF respond to it as well (Schweigerer et al . (1987) Nature 325. 257- 259; Schweigerer et al. (1987) Endocrinology 120. 796-800; Halaban et al. (1988) Oncogene Res. 3, 177-186; Sato et al .

(1988) J. Cell. Biol. 107. 1199-1205). Thus, bFGF may act as an autocrine regulator of specific cellular functions. As an autocrine stimulator of cell proliferation, bFGF could play a pivotal role in the development of many tumors. In addition, cells transfected with bFGF exhibit amplified autocrine growth in monolayer culture and soft agar (Neufeld et al . (1988) J. Cell. Biol. 106. 1385-1394; Rogelj et al . (1988) Nature 331. 173-175). Cells transfected with a coding sequence for addle FGF also exhibit a loss of contact inhibited growth and produce small, non-progressive tumors in nude mice (Jaye et al. (1988) EMBO 7, 963-969). These data indicate that the expression of FGF-like molecules could play a role in cellular transformation and in the continued proliferation of neoplastic cells.

Basic FGF is also a potent inducer of new blood vessel formation, or neovascularization, a phenomenon essential to the growth of solid tumors (Esch et al . (1985) Proc. Natl. Acad. Sci. USA 82, 6507-6511; Lobb et al . (1985) Biochemistry 24. 4969- 4973). The process of neovascularization has been associated with malignant progression in solid tumors (Folkman et al . (1987) Science 235. 442-447). Therefore, bFGF could influence tumor development by directly stimulating tumor cell growth or by promoting tumor vascularization.

The role of bFGF may, for example, be of significance in the growth and propagation of human gliomas, since bFGF is relatively abundant in neural tissue and is known to exert actions on several normal cell types of neural origin (Morrison et al. (1981) Proc. Natl. Acad. Sci. USA 78. 7205-7209; Gospodarowicz et al. (1984) Proc. Natl. Acad. Sci. USA 81, 6963- 6967; Ann Eccleston et al . (1985) Develop. Brain Res. 21, 315- 318; Pettmann et al . (1985) FEBS 189. 102-108; Morrison et al . (1986) Proc. Natl. Acad. Sci. USA 83, 7537-7541; Walicke et al . (1986) Proc. Natl. Acad. Sci. USA 83. 3012-3016). Moreover, the natural history of gliomas, which kill patients by invasion and local spread rather than by metastasis, may be significantly influenced by peptides that stimulate the growth, invasion, and vascularization of the primary tumor 1n the cranium. It has recently been reported that a malignant human glioma cell line expresses and responds to bFGF (Morrison et al . submitted for publication) . Basic FGF has a wide spectrum of biological activities including stimulating proliferation in a variety of cell types (Baird et a . (1988) Proc. Natl. Acad. Sci. USA 85, 2324), modifying neurite outgrowth (Schubert et al . (1987) J. Cell. Biol . 101. 635-643), and promoting neovascularization (Esch et al. (1985) Proc. Natl. Acad. Sci. USA 85. 6507-6511), promoting certain cell-to-cell interactions, and promoting adhesion (Gospodarowicz et al . (1986) Mol . Cell. Endocrin. 46. 187-204; Gospodarowicz et al. (1985) Endocrinology 118. 82-90; MoscatelH et al. (1986) J. Cell. Physiol. 129. 273-276; Baird et al . (1986) Rec. Prog. Horm. Res. 42, 143-205; Gospodarowicz et al . (1985) Endocrinology 117. 2383-2391; Gospodarowicz et al . (1984) Proc. Natl. Acad. Sci. USA 81, 6963-6967). These potent activities of bFGF suggest a.^' critical role for this mitogen, not only in physiological functions associated with reproduction, growth and development, but also in pathophyslological processes Including growth of tumors, diabetic proliferative retinopathies, and wound healing responses.

Recently, polyclonal antibodies capable of neutralizing the biological activity of bFGF, as defined in various 1_n vitro assays, have been described (Kurokawa et al. (1989) J. B1ol. Chem. 264. 7686-7691; Sakaguchi et al . (1988) FEBS Lett. 233. 163-166; Joseph-Silverstein et al . (1988) J. Immunol. Methods 110. 183-192). However, the effects of these antibodies on the In vivo biological activity of bFGF were not evaluated. Such antibodies are subject to the usual limitations of polyclonal antiserum in physiological studies, Including heterogeneity and batch to batch variability (Zalcberg (1985) Pharmacol . Ther. 28, 273).

Monoclonal antibodies directed against bFGF, particularly those with bFGF-neutralizing or bFGF-inhibiting properties, would represent powerful tools for providing new insights into the biological actions of bFGF. While monoclonal antibodies to bFGF have been described and utilized for isolation, detection, and measurement of bFGF (Seno et al . (1989) Hvbridoma 8, 209-221; European Patent Application 0-288-687; Massoglia et al . (1987) JL Cell. Phys. 132. 531-537), none of these have been shown to inhibit cellular binding or to inhibit biological activity of bFGF in vitro or i_n vivo.

Seno et al. (Hvbridoma (1989) 8:209-221; EPA 0-288-687) disclose the following monoclonal antibodies:

Epitopes were mapped by peptide competition. MAb52 and MAb98 did not block ³H-thymidine uptake into quiescent BALB/c3T3 cells treated with bFGF. MAbl2 and MAb78 were apparently not neutralizing since aa 1-15 are not required for the mitogenic activity or receptor binding of bFGF.

Massoglia et al . (J. Cell. Phvslol. (1987) 132: 531-537) reported the following monoclonal antibodies to bFGF: Designation Epitope Recognized

McAB 6 aa 1-10

McAB 8 aa 1-10

McAb 38 aa 1-10 McAb 42 "core"

Epitopes were mapped using a peptide and a form of bFGF lacking aa 1-10. McAB 42 was found to be ineffective in blocking biological activity (mitogenic or receptor binding) of bFGF.

SUMMARY OF THE INVENTION This invention provides monoclonal antibodies to human bFGF, and antigen-binding fragments of such antibodies, which inhibit 1_n vitro bFGF-1nduced ³H-thymidine incorporation into cultured fibroblasts and gliomal cells, inhibit n vivo the angiogenic response to bFGF in a rat kidney capsule model of angiogenesis, and/or inhibit 1_n vivo tumor growth 1n a mouse B16 melanoma model .

Monoclonal antibodies and fragments within this invention also inhibit i_n vitro binding of bFGF to bFGF receptors on cultured baby hamster kidney (BHK) cells. Preferred monoclonal antibodies and fragments of the invention inhibit 1_n vitro binding of bFGF to low affinity bFGF receptors on BHK cells to a greater extent than they inhibit binding of bFGF to high affinity bFGF receptors on cultured BHK cells. Other monoclonal antibodies and fragments of the invention inhibit 1_n vitro binding of bFGF to high affinity bFGF receptors on cultured BHK cells to a gre^'ater extent than they inhibit binding of bFGF to low affinity bFGF receptors on cultured BHK cells. The invention Includes hybrldoma cell lines which produce the monoclonal antibodies of this invention.

The invention also Includes therapeutic use of monoclonal antibodies and fragments of the invention to Inhibit bFGF- ediated angiogenesis and solid tumor growth in mammals. In addition, the invention includes pharmaceutical compositions comprising a therapeutically effective amount of a monoclonal antibody or fragment of the invention in a pharmaceutically acceptable vehicle.

DETAILED DESCRIPTION OF THE INVENTION

A panel of five murine monoclonal IgGi antibodies (mAbs) to a recombinant form of basic fibroblast growth factor (bFGF) was produced using somatic cell fusion techniques (see Example 1). These mAbs are designated as: DE6, AF11, FE8, AE4, and DG2. Two of the mAbs, DG2 and DE6 were identified as a bFGF- neutralizing antibody on the basis of their ability to inhibit, in vitro, the binding of [¹²⁵I]-bFGF to high and low affinity bFGF sites on cultured baby hamster kidney (BHK) cells (Example 2) and bFGF-induced [³H]-thymid1ne incorporation in cultured 3T3 cells and human glioma SNB-56 cells (Example 3). In addition DG2 was found to inhibit in vivo the angiogenic response to bFGF in a rat kidney capsule model of angiogenesis (Example 4). Moreover, DG2 was found to inhibit the growth of the B16 melanoma and DLD-2 colon carcinoma in mice (Example 5). Hybridoma cells lines, designated DE6 and DG2, which produce antibodies DE6 and DG2, were deposited in the American Type Culture Collection (ATCC), Rockville, MD, in accordance with the provisions of the Budapest Treaty. ATCC accession numbers for these hybridoma cell lines are HB10203 and HB10202, respectively. Neutralizing antibodies such as DG2 and DE6, and antigen- binding fragments of such antibodies, are useful for a number of applications in bFGF research, including immunoassays, immunohistochemical studies, isolation and purification of bFGF, and as functional antagonists of bFGF for examining its role in physiological processes such as reproduction, growth, and development. Antibodies such as DG2 which exhibit j_n vivo inhibition of bFGF, and antigen-binding fragments thereof, are useful for the therapeutic treatment of diseases mediated or contributed to by bFGF. Additional studies, including defining the epitopes recognized by these particular antibodies, will be important to provide additional insights into the modulation of bFGF-mediated responses. The potent neutralizing activity of DG2 identifies it as an extremely useful reagent for evaluating biological activities of bFGF in various physiological and pathophysiological processes.

The potential problem of immunogenicity when introducing murine antibodies into humans may be reduced or eliminated by: 1) utilizing smaller, less immunogenic antigen-binding fragments of murine antibodies such as Fab, F(ab')2 and Fab' fragments (Garvey et al. (1977) Methods in Immunology. W. A. Benjamin,

Reading, MA); 2) utilizing human monoclonal antibodies (Schlom et al. (1980) Proc. Natl. Acad. Sci. USA 77. 6841-6845; Satoh et al . (1983) N. Engl. J. Med. 309. 217-220); or 3) utilizing human- mouse chimeric antibodies produced by genetic engineering techniques (Oi et al. (1986) Biotechnioues 4. 214-221).

For therapeutic application purified bFGF-neutralizing monoclonal antibodies and fragments of the invention can be administered as a sterile, nonpyrogenic parenteral solution. Concentration of the antibody or fragment in the solution will ordinarily be about 0.5 to 10 mg/ml. The aqueous parenteral vehicle could be, for example, Sterile Water for Injection USP, 0.9% Sodium Chloride for Injection USP or 5% Dextrose Injection USP. Pharmaceutical compositions of the present invention comprise an effective amount of purified soluble bFGF- neutralizing monoclonal antibody or fragment and a pharmaceutically acceptable carrier such as listed above, and optionally additional excipients, such as preservatives and buffers as known to those skilled in the art of pharmaceutical formulations. See, for example, Remington's Pharmaceutical Sciences, a standard reference 1n the field. bFGF-neutralizing monoclonal antibody or fragment may be prepared in a stable formulation ready for administration, or for dilution in an appropriate intravenous solution. Preferably, to increase product shelf life, the bFGF-neutralizing monoclonal antibody or fragment may be formulated, for example, as a sterile lyophilized powder to be reconstituted aseptlcally or as a buffered solution.

As is known to those skilled in the art, Individual patient dosage will vary depending upon known factors such as the pharmacodynamic characteristics of the particular antibody or fragment; the age, health, and weight of the recipient; the nature and extent of symptoms; the kind of concurrent treatment and frequency of treatment; and the effect desired. Generally, daily dosage of antibody or fragment 1n the range of about 0.1 to 100 milligrams per kilogram of body weight can be administered in a single dose or in 2-6 equally divided doses.

Diseases which are mediated or contributed to by bFGF and can be treated by administration of preferred antibodies and fragments of this invention are known as "angiogenic diseases". (Folkman, J. and M. Klagsbrun, "Angiogenic Factors", Science 235:442-447, 1987, and Folkman, J., "Successful Treatment of an

Angiogenic Disease", New England J. Med.. 320:1211-1212, 1989. A current list of such disorders includes primary and metastatic malignancies, diabetic retinopathy, retrolental flbroplasia, neovascular glaucoma, rheumatoid arthritis, hemangiomas, angiofibro as, psoriasis, capillary proliferation within atherosclerotic plaques, and pulmonary capillary hemangio atosis. Solid malignant tumors which can be treated with the preferred antibodies and fragments of the invention to Inhibit tumor growth include melanoma, colon carcinoma and Kaposi's sarcoma. Monoclonal antibodies of this invention can be generated using purified bFGF or a fragment thereof as immunogen. Kohler and Milstein (Nature (1975) 256, 495) first described methods of making monoclonal antibodies by fusing spleen cells from an immunized mouse to a drug-resistant plasmacytoma cell line and isolating the hybrid clones by growth on selective medium. Monoclonal antibodies can overcome many of the problems associated with the use of polyclonal antisera; namely purity, specificity, homogeneity and availability.

Monoclonal antibodies can be prepared by methods known to one skilled in the art, such as, by the method set forth herein. Briefly, an animal is immunized with bFGF or a fragment thereof, an antibody-producing organ is subsequently harvested, a cellular homogenate is prepared, and cells in the homogenate are fused with tumor cells to produce hybrido a cells. Hybridoma cells which produce antibody specific for bFGF are selected by assaying for binding of hybridoma-produced antibody to bFGF using standard radio-labeled bFGF and radioimmunoassay (RIA) procedures, or bFGF immobilized on a solid support and standard enzyme-linked immunoassay (ELISA) procedures. Hybridomas producing monoclonal antibody specific for bFGF may be expanded and the monoclonal antibody produced may be harvested.

Typically, an animal is immunized with the antigen of interest emulsified in an adjuvant and boosted at regular intervals. The serum is assayed for the presence of the desired antibody by any convenient method, frequently in ELISA or a RIA. When an acceptable titer of antibody is detectable in the serum, the animal is sacrificed and the spleen is removed aseptlcally for fusion.

Several different murine (mouse) myeloma cell lines deficient in hypoxanthine guanlne phosphoribosyl transferase y

(HGPRT) are known to be suitable as fusion partners. The features of some of these cell lines are described in Current Topics in Microbiology and Immunology, jBJ., F. Melchers, M. Potter, and N. Warner, ed., Springer-Verlag, 1878. Fusion is carried out most commonly by using polyethylene glycol as a fusion promoter. After fusion, the cells are diluted, and cultured in a selective medium containing hypoxanthine, aminopterin and thymidine (HAT). Cells may be supplemented with insulin to enhance the formation and growth of hybridomas. When sufficient cell growth has occurred, the culture supernatant is sampled and tested by any convenient means, frequently ELISA or RIA. Those cultures which contain antibody of interest are then cloned by limiting dilution, re-tested, and expanded. Large volumes of antibody can then be obtained by growing the hybridoma i_n vitro and harvesting the culture supernatant. Antibodies may also be harvested from the ascites fluid of syngeneic mice which have been Injected intraperitoneally with the hybridoma cells. The antibody is purified using techniques well-known in the art. Chromatography on staphylococcal protein A 1s one such method. The monoclonal antibody thus generated can be characterized by its immunog!obulin class and subclass.

Reagents used in the following examples: The bFGF, provided by Synergen, Inc. (Boulder, CO), was a human recombinant form cloned and expressed in Escherichia coli and previously shown to be indistinguishable from human placental bFGF in various assays (Squires et al . (1988) J. Biol. Chem. 263. 16297- 16302). [¹²⁵I]-bFGF (40 cpm/pg) was produced by the lactoperoxidase method as described for nerve growth factor

(Sonnenfeld et al. (1985) J. Neurosd. 5. 1717-1728) and purified over a heparin-agarose affinity column (Gospodarowicz et al . (1984) Proc. Natl. Acad. Sci. USA 81. 6963-6967). Bovine acidic and basic FGF's were purchased form R & D Systems, Inc. (Minneapolis, MN), [¹²⁵I]Na from Du Pont-New England Nuclear (Billerica, MA), heparin-agarose and MOPC 21 IgG_j from Sigma Chemical Company (St. Louis, MO), monophosphoryl lipid A (MPL) , and trehalose dimycolate (TDM) from RIBI Immunochem Research, Inc. (Hamilton, MT) . Balb/c mice and Sprague-Dawley rats were from Charles River Laboratories (Charles River, MA). Swiss 3T3 cells (clone DI) were kindly provided by Elaine Raines of Synergen, Inc. B16 melanoma cells were from ATCC, Rockville, MD.

Example 1

Generation of bFGF Monoclonal Antibodies Female balb/c mice were primed with 20 μq of bFGF in the emulsion adjuvant MPL-TDM by intraperitoneal (IP) injection and boosted (IP) at 15 weeks and 19 weeks with 20 μς- and 10 g bFGF, respectively, in TDM emulsion. At 23 weeks 10 μg of bFGF in 50 μλ PBS, was equally divided and administered directly into the spleen and the intraperitoneal cavity via aseptic abdominal surgery. Three days later, the mouse was sacrificed and the spleen cells were fused with P3X63-Ag8.653 cells in PEG 1500 (Reilly et al . (1987) Biochem. Biophvs. Res. Commun. 143. 133- 139). After selection in HAT medium, the supernatants were screened for specific antibody production by an ELISA using microtiter wells coated with 500 ng bFGF per well. Positive hybridomas were cloned by limiting dilution. Monoclonal antibodies were purified from culture supernatant by affinity chromatography over a protein A-Sepharose CL-4B column (Ey et al . (1978) Immunochemistry 15. 429-437). The dissociation constant (KQ) for each antibody was determined by non-linear regression analysis (Scatchard (1949) Ann. N.Y. Acad. Sci. 51. 660-668) of data obtained in a competitive radioimmunoprecipitation assay (Massoglia et al . (1987) J. Cell. Phvs. 132. 531-537). Cross reactivities of antibodies for the acidic and basic forms of native, bovine FGF were calculated as previously described (Abraham (1969) J. Clin. Endocrinol. Metab. 29. 886-895). Nineteen murine hybridomas secreting antlboαπes LO υrι_r were Identified by ELISA from a fusion of P3X63-Ag8.653 myeloma cells with spleen cells of a bFGF-immunized mouse. Five of the antibodies, designated AE4, AFll, DE6, DG2 and FE8, were selected for further studies based on their values in the bFGF ELISA. All antibodies were identified as IgGχ in a typing ELISA. The dissociation constants (KQ) of the antibodies for bFGF, calculated from analysis of radioimmunoassay data, ranged from 0.822 to 20.1 nM (Table 1). Cross reactivities for each of the antibodies with the basic (bFGF) and acidic (aFGF) form of native, bovine FGF were also determined by radioimmunoassay. Human bFGF differs from its bovine counterpart by only two conservative amlno add substitutions (Abraham et al . (1986) Science 233. 545-548), while a 55% homology exists between aFGF and bFGF (Esch et al . (1985)

Biochem. Biophvs. Res. Commun. 133. 554-562). All the antibodies recognized bovine bFGF, while only AE4 and DE6 displayed any significant cross reactivity towards the acidic form of FGF (Table 1). Table 1

KQ'S were calculated from analyses of data obtained from radioi munoprecipitation assays. Crossreactlvities of antibodies with bovine acidic and basic FGFs were also determined by the radioimmunoprecipitation assay. Example 2 Effect of Antibodies on Binding of bFGF to Low and High Affinity Sites on BHK Cells Iodination Procedure: ¹²⁵l was obtained as Nal from New England Nuclear at a concentration of 100 millicuries/mL 1n dilute NaOH (NEZ 033A) . Male Sprague-Dawley rats were from Charles River Farms and all other reagents were purchased from standard commercial sources.

One miHicurie of Na¹²⁵I was centrifuged at 600 rpm for 5 minutes in its original container. To the Na¹²⁵I we added 10 μl 0.2 N HCl in 0.1 M KH₂P0 (pH 7.4), 16.7 μl of 0.1 M KH₂P0 (pH 7.4), 3.3 μl of 3 mg/mL bFGF, 5 μl of 50 g/mL lactoperoxidase in 0.1 M KH₂P0_j, and 5 μl of 0.004% H₂0₂ in 0.1 M KH2PO4. The mixture was allowed to stand at room temperature for 15 minutes, and then received an additional 5 μl of 0.004% H2O2 in buffer. After an additional 15 minutes, 350 μl of buffer A (50 mM Tris, pH 7.4, 0.15% gelatin) containing 0.5 mg/mL phenol red was added. Triplicate 1 μl samples of the reaction mixture were taken for TCA precipitation analysis and the remaining reaction mixture was loaded onto a 20 cm x 1 cm Sephadex-G15 column that had been equilibrated with buffer A. The column was eluted with buffer A and 100.5 L fractions were collected. Triplicate 1 μl samples of the peak fraction were taken for TCA precipitation analysis.

Purification Procedure: Immediately following the iodination, the fractions containing the peak from the Sephadex- G15 column were loaded onto a 5 cm x 1 cm heparin agarose column that had previously been equilibrated with TRIS buffer containing 0.5 M NaCl. The column was washed with at least 50 mL of the 0.5 M NaCl buffer or until the eluate was free of counts. Next, the ¹²⁵I-bFGF was eluted by washing the column with Tris buffer containing 2 M NaCl. Fractions of approximately 200 μl (4 drops) were collected and counted to determine the bFGF content. TCA Precipitation: Each of the triplicate 1 μl samples for analysis is added to 1 mL of buffer A containing 1 mg/mL BSA. For each sample, a 50 μl aliquot 1s taken for counting and 100 μl is added to 100 μl 20% trichloroacetlc acid (TCA) 1n a 1.5 L icrofuge tube, mixed well, refrigerated for at least one hour and centrifuged at 5,000 x g for two minutes. A 100 μl aliquot of the supernatant 1s taken for counting. The total counts incorporated into protein and the cpm/pg protein are calculated from the samples of the reaction mixture before and after TCA precipitation. The samples of the peak fraction after the Sephadex column are used to determine the recovery and the percent precipitability of the recovered protein.

Binding Assay: Baby hamster kidney (BHK) cells were grown in 24-well plates in Dulbecco Modified Eagle medium (DMEM) with 20% fetal calf serum. bFGF binding was determined using the procedure described by Moscatelli (J. Cell. Biol. (1988) 107. 753-759). The medium was removed from the cells and replaced with DMEM containing 0.15% gelatin and 20 mM HEPES buffer at pH 7.5. This solution contained 40 pM radlolodinated bFGF and various concentrations of unlabeled bFGF. After incubation on ice for 2 hours, the incubation medium was removed and the cells were washed with cold PBS. The medium and wash were used to determine the unbound bFGF. The cells were washed twice more with PBS and those washes discarded. The cells were then washed twice with 2 M NaCl in 20 M HEPES at pH 7.5 to remove bFGF bound to low affinity receptors and twice with 2 M NaCl in 25 M acetate buffer at pH 4.0 to remove bFGF bound to high affinity receptors. The washes were counted in a Packard Gamma counter for ¹²⁵I. Non-specific binding was determined by the addition of 3 / unlabeled bFGF.

Table 2 shows the binding of bFGF to high and low affinity receptor sites on BHK cells. Table 2 High and Low Affinity Binding of bFGF to Receptors on BHK Cells S ecific Bindin

Table 3 shows the effect of various concentrations of various mAbs on the binding of the [¹²⁵I]-bFGF to the low and high affinity binding sites on BHK cells. For this evaluation, [¹²⁵I]-bFGF in binding buffer (40 pM initial concentration) was pretreated with various concentrations of the appropriate antibody for 4 hr at 25°C, and then the binding assay was performed as described above.

The low and high affinity components have been suggested to reflect [¹²⁵I]-bFGF binding to glycosaminoglycans and specific membrane receptors respectively (MoscatelH (1987) J. Cell. Physiol. 131. 123-130; Moscatelli (1988) J. Cell. Biol. 107. 753-759) BHK cells contain Urge numbers of both low and high affinity binding sites for bFGF (Moscatelli (1987) J. Cell. Physiol. 131. 123-130; Neufeld, et al . (1985) J. Biol. Chem. 260. 13860-13868).

Table 3 Inhibition of bFGF Binding by Monoclonal Antibodies

% of Control Bindin

As shown in Table 3, the mAb-dependent extent of inhibition of binding of bFGF to the high affinity site 1s greater than that observed for the binding of bFGF to the low affinity site, for all of the mAbs tested, except DG2. These results indicate that, with the exception of DG2, the bFGF- specific mAbs inhibit the binding of bFGF to the high affinity bFGF-binding site on BHK cells to a greater extent than they inhibit the binding to the low affinity site on BHK cells. In Table 4, the monoclonal antibodies at the listed concentrations were combined with bFGF and incubated for various times at room temperature in binding buffer. The mixtures were added to the wells containing BHK cells on ice and the binding assay performed. As shown in Table 4, only DG2 exhibited greater inhibition of the binding of bFGF to the low affinity binding site than that for the high affinity receptor site.

Tabl e 4

Effect of Preincubation of mAb and bFGF on the Binding of bFGF to BHK Receptors

% of Control Binding

All the antibodies had pronounced effects in inhibiting bFGF binding to both sites with the order of decreasing potency for high affinity sites: DE6 > FE8 > AFll > AE8 > DG2, and for low affinity sites: DG2 > DE6 > AFll > FE8 > AE8 (Table 4). Values are expressed relative to the binding observed 1n the presence of a control mouse IgGj antibody, MOPC 21, which had negligible effects on the binding assay of bFGF to receptors on BHK cells.

Example 3 Effects of Monoclonal Antibodies on bFGF-induced Thymidine Incorporation 3T3 Cell Mitogenesis Assays: bFGF, pretreated in the absence or presence of appropriate antibody 1n DMEM containing 0.5% calf serum for 1 hr at 25^βC, was added to duplicate cultures of confluent Swiss 3T3 fibroblasts in 24 well plates (2 ng bFGF per well) for a 20 hr incubation at 37^βC. Cultures were washed and received fresh medium containing [³H]-thymidine (0.5 /iCi/well) for an additional 2 hr incubation at 37^βC. Cultures were then washed in PBS, fixed in cold 5% TCA for 30 min, solubilized at room temperature in 0.25 N NaOH, and the NaOH- soluble material processed for scintillation counting. Basic FGF is a potent stimulator of thymidine incorporation in 3T3 cells (Baird et al. (1988) Proc. Natl . Acad. Sci. USA 85, 2324-2328). To determine the effect of the antibodies on the mitogenic activity of bFGF, we evaluated their ability to inhibit bFGF-induced thymidine incorporation in these cells. For these studies, bFGF was preincubated for 1 hr with various concentrations of the antibodies prior to its incubation with cells. As shown in Table 5, a dose-dependent Inhibition in thymidine incorporation was observed with antibodies DE6 and DG2. 85% of bFGF-induced activity was inhibited with 1.3 μg of DG2, the most potent antibody in this assay. The remaining 3 anti- bFGF antibodies had much smaller Inhibitory effects on thymidine incorporation. The MOPC 21 control antibody had no effect on the mitogenic activity of bFGF, even at the highest concentration of 100 /ig/ml .

Solutions of bFGF (2 ng), pre-incubated for 1 hour with 5 μg, 1.3 μg, or 0.5 μg of the appropriate antibody, were added to duplicate wells of 3T3 cells and incubated for 20 hrs. [³H]-thymidine incorporation was then measured as described in the text, and expressed as the percentage observed in the absence of any antibody. Glioma Cell Mitoqenesis Assays: Human glioma SNB-56 cells were maintained in a humidified 5% CO2: 95% air atmosphere at 37°C in complete medium, defined as Ham's F12/Dulbecco's Modified Eagle Medium (DMEM) (1:1) with 20% (v/v) heat inactivated fetal calf serum (FCS). Cells were trypsinized and suspended at 1 x 10^ cells per 35 mm culture dish. Duplicate cultures were prepared for each antibody tested. Twenty-four hours post-plating, cells were washed twice with sterile phosphate-buffered saline (PBS) and incubated for an additional 24 hrs in serum-free F12/DMEM (1:1). Cultures were then incubated (1.5 L/dish) in chemically defined media (CDM) which consisted of F12/DMEM (1:1) with 30 nM selenium, 50 /.g/ml transferrin, 10 nM hydrocortisone, 5 /g/ml insulin, and 1.0 /.Ci/ml [³H] thymidine. Purified IgG (20 .l per 1.5 mL CDM) was added in duplicate for each antibody tested. After 1 hr, 1 ng/mL basic fibroblast growth factor (bFGF) (Stock in PBS with 0.1% CHAPS) was added to half of the cultures. Cultures were incubated for 17 hrs at 37°C in a humidifies atmosphere. [³H]thymidine incorporation was measured as follows. Labeled medium was removed from cultures, and cultures were washed twice with cold PBS. Cells were incubated for 20 minutes at 37°C with 1 mL of 0.2 M NaOH/dish. The solubilized cell extracts were placed in glass tubes, and incubated on ice for 30 minutes with equal vol (1 mL) of 20% (w/v) of trichloracetic add (TCA). The extracts were filtered onto GF/B filters and filters were washed three times with cold 10% TCA. Radioactivity on filters was measured in a Packard Tricarb scintillation counter.

Table 6

Inhibition of bFGF-induced [³H]-thymidine Incorporation by Monoclonal Antibodies to bFGF

[³H]-thymidine Incorporation (DPM) av . of 2 dishes

The numbers in parentheses indicate the concentration of antibody in each dish.

As shown in Table 6, mAbs DG2 and DE6 appeared to inhibit the bFGF-dependent uptake of [³H]-thymidine into human glioma SNB-56 cells.

Example 4 Effects of Monoclonal Antibodies on bFGF-induced Angiogenesis Rat Angiogenesis Model: The inhibitory activities of antibodies DG2 and DE6 in the in vitro mitogenic assay prompted us to evaluate their ability to neutralize the biological activity of bFGF in vivo. For this purpose, we utilized a recently described in vivo model (Hayek et al . (1987) Biochem. Biophvs. Res. Commun. 147. 876-880), where the angiogenic response to bFGF is evaluated after its implantation 1n slow release beads under the kidney capsules of rats.

Briefly, this involved adsorption of Aff1-Gel blue agarose beads complexed with bFGF (1 μg) and mAb where appropriate, onto sterile Gelfoa ® pads and surgical Implantation of such pads beneath the left kidney capsule of anesthetized, male Sprague-Dawley rats.

Male Sprague-Dawley rats (200-300 grams) were anesthetized with a mixture of chloral hydrate (42.5 mg/mL) and pentobarbital (8.9 mg/mL) given at 0.4 mL/100 g. The left side of the rat's abdomen 1s shaved and disinfected with benzoin tincture (Cliniswab, CHnipad Corp.). A left side laparotomy 1s performed to expose the left kidney. Using microforceps, a 3x3 mm piece of Gelfoam® is implanted underneath the kidney capsule through an incision at the lower pole and gently advanced to the upper pole. The abdominal muscle and skin layers are closed with 2-0 Dexon sutures and the rats allowed to recover. Seven days later, the rats are reanesthetized and a m1dl1ne laparotomy performed. The angiogenic response is viewed under 2-fold stereoscopic magnification and scored on a scale of 0 to 4, where 0 is no response; 1 is a very weak response; 2 a weak response; 3 is a moderate response and 4 indicates a strong response. To prepare the Gelfoam® patches, 30 μg of basic fibroblast growth factor from Synergen, Inc. (Boulder, CO) 1s combined with 300 μl Affi-Gel Blue beads (150-300 μ, Bio-Rad, Richmond, CA) and saline with or without antibody 1n a total volume of 3.6 mL. The mixture is incubated at 37 degrees for 30 minutes and centrifuged at 1000 x g for 10 minutes at room temperature. 3 L of the supernatant 1s discarded and the beads are resuspended in the remaining 600 μl. Each patch received 20 μl of the bead suspension which contained 1 μg bFGF, which previous work had shown to be the optimum dose.

An intense angiogenic response was visualized seven days after the subcapsular Implantation of beads containing 1 μg of bFGF, with a mean value of 3.2 (using 23 rats) (Table 7). The beads alone did not elicit a substantial neovascular response US90/05658

24

(1.9). Pretreatment of the bFGF with various concentrations of DG2 resulted in a concentration-dependent reduction in the bFGF- induced angiogenic re.sponse (Table 7), with 10 μg of the antibody giving a mean value of 1.8. This was similar to the control (i.e., beads alone). The addition of 40 μg of either DE6 or the control IgGj monoclonal antibody MOPC 21 had a very small effect on bFGF-induced angiogenesis (Table 7).

To our knowledge, these results provide the first demonstration of antibody-mediated neutralization of the biological activity of bFGF. It is interesting to note that this bFGF-neutralizing antibody, DG2, was the most potent inhibitor of bFGF binding to low affinity sites on BHK cells (Tables 3 and 4), which probably represent heparin-like molecules. Our results therefore support the notion proposed by others (Moscatelli et al. (1987) J. Cell. Physiol. 131. 123-130; Moscatelli et al.

(1988) J. Cell. Biol. 107. 753-759; Baird et al . (1987) J. Cell. Phvs. Supplmt. 5, 101-106) that the binding of bFGF to heparin or heparin-like molecules may be important in mediating its biological activity. Table 7

Effect of Monoclonal Antibodies on bFGF-induced Angiogenesis in the Rat Kidney Capsule Model Pretreatment

Data are expressed as mean + standard deviation, "n" represents the number of rats analysed in each treatment group. Example 5 Effect of bFGF-spedfic mAbs on Tumor Growth in Mice B16 Melanoma Model : The effect of bFGF-spedfic mAbs was also tested 1n the mouse B16 melanoma model. C57BL16 mice (N=5) were injected with 1 x 10^ B16 melanoma cells on day 0. On day 10, when the tumors were approximately 300 mg, DG2 monoclonal antibody to bFGF (5 /ig/injection), MOPC-21, a control murine monoclonal antibody (5 /ig/injection), or vehicle (PBS) were injected intratumorally for 5 consecutive days. Tumors were measured on day 13 and 17. The results are shown 1n Table 8. As shown in Table 8, DG2 inhibited the growth of the B16 melanoma tumor in the mouse.

Table 8 Inhibition of B16 Melanoma 1n Mice by bFGF-specific Monoclonal Antibody Days Foilowing Treatment

0 3

Treatment Tumor Weight (mg) PBS 293 + 98 446 + 356 1052 + 268

MOPC-21 304 + 93 501 + 321 1011 + 104

DG2 302 + 89(0) 300 + 259(40) 568 + 211(44)

Mean for 5 mice +

Claims

WHAT IS CLAIMED IS:

1. A monoclonal antibody to human basic fibroblast growth factor (bFGF), or an antigen-binding fragment of such an antibody, which inhibits in vitro bFGF-1nduced ³H-thymid1ne incorporation into cultured fibroblasts and gliomal cells.

2. A monoclonal antibody to human bFGF, or an antigen- binding fragment of such an antibody, which inhibits in vivo the angiogenic response to bFGF in a rat kidney capsule model of angiogenesis.

3. A monoclonal antibody or fragment of claim 2 which inhibits i_n vitro binding of bFGF to low affinity bFGF receptors on cultured baby hamster kidney (BHK) cells to a greater extent than it inhibits binding of bFGF to high affinity bFGF receptors on cultured BHK cells.

4. A monoclonal antibody to human bFGF, or an antigen- binding fragment of such an antibody, which inhibits tumor growth in a mouse B16 melanoma model.

5. A monoclonal antibody or fragment of claim 4 which inhibits in vitro binding of bFGF to low affinity bFGF receptors on cultured baby hamster kidney (BHK) cells to a greater extent than it inhibits binding of bFGF to high affinity bFGF receptors on cultured BHK cells.

6. The monoclonal antibody of claim 5 designated DG2 which is produced by hybridoma cell line designated DG2 which is on deposit in ATCC under accession number HB10202.

7. A monoclonal antibody or fragment of claim 1 which inhibits in vitro binding of bFGF to high affinity bFGF receptors on cultured BHK cells to a greater extent than it inhibits binding of bFGF to low affinity bFGF receptors on cultured BHK cells.

8. The monoclonal antibody of claim 7 designated DE6 which is produced by hybridoma cell line designated DE6 which is on deposit 1n ATCC under accession number HB102 3.

9. A hybridoma which produces a monoclonal antibody of any one of claims 1-8.

10. A pharmaceutical composition comprising a therapeutically effective amount of a monoclonal antibody or fragment of any of claims 2-61n a pharmaceutically acceptable vehicle.

11. A method of inhibiting bFGF-mediated angiogenesis 1n a mammal comprising administering to the mammal a therapeutically effective amount of a monoclonal antibody or fragment of any of claims 2-6.

12. A method of inhibiting growth of a solid tumor in a mammal comprising administering to the mammal a therapeutically effective amount of a monoclonal antibody or fragment of any of claims 2-6.

13. Method of claim 12 wherein the tumor is a melanoma.

14. Method of claim 12 wherein the tumor is a colon carcinoma.