AU2011244925B2 - Compositions and methods of treating disease with FGFR fusion proteins - Google Patents

Abstract

The invention provides FGFR fusion proteins, methods of making them, and methods of using them to treat proliferative disorders, including cancers and disorders of angiogenesis. The FGFR fusion molecules can be made in CHO cells and may comprise deletion mutations in the extracellular domains of the FGFRs which improve their stability. These fusion proteins inhibit the growth and viability of cancer cells in vitro and in vivo. The combination of the relatively high affinity of these receptors for their ligand FGFs and the demonstrated ability of these decoy receptors to inhibit tumor growth is an indication of the clinical value of the compositions and methods provided herein.

Description

AUSTRALIA Regulation 3.2 Patents Act 1990 Complete Specification Standard Patent APPLICANT: Five Prime Therapeutics, Inc. Invention Title: COMPOSITIONS AND METHODS OF TREATING DISEASE WITH FGFR FUSION PROTEINS The following statement is a full description of this invention, including the best method of performing it known to me: COMPOSITIONS AND METHODS OF TREATING DISEASE WIfH FGFR FUSION PROTEINS RELATED APPLICATIONS [001] The invention relates to applications 60/701,479, filed July 22, 2005; 60/729,401, filed October 21, 2005; 60/757,398, filed January 10, 2006; and 60/800,005, filed May 15, 2006, all of which are incorporated by reference in their entireties. TECHNICAL FIELD [0021 The present invention relates to fusion molecules comprising an extracellular domain of a fibroblast growth factor receptor (FGFR). It relates to polypeptide and polynucleotide sequences, vectors, host cells, compositions, kits, and animals comprising FGFR fusion proteins. The invention also relates to methods of making and using FGFR fusion molecules and variants and fragments thereof to diagnose, prevent, determine the prognosis for, and treat proliferative diseases, including cancer and disorders of angiogenesis. BACKGROUND ART [003] Fibroblast growth factors (FGFs) and their receptors (FGFR) are a highly conserved group of proteins with instrumental roles in angiogenesis, vasculogenesis, and wound healing, as well as tissue patterning and limb formation in embryonic development. FGFs and FGFRs affect cell migration, proliferation, and survival, providing wide-ranging impacts on health and disease. [004] The FGFR family comprises four major types of receptors, FGFR1, FGFR2, FGFR3, and FGFR4. These receptors are transmembraue proteins having an extracellular domain, a transmembrane domain, and an intracytoplasmic domain. Each of the extracellular domains contains either two or three immunoglobulin (Ig) domains. Some FGFRs exist in different isoforms which differ in specific segments of the molecule, such as FGFRI-ilb and FGFR1-IIc, which differ in the C-terminal region of the third Ig domain. Transmembrane FGFRs are monomeric tyrosine kinase receptors, activated by dimerization, which occurs at the cell surface in a complex of FGFR diners, FGF ligands, and heparin glycans or proteoglycans. Extracellular FGFR activation by FGF ligand binding to an FGFR initiates a cascade of signaling events inside the cell, beginning with the receptor tyrosine kinase activity. [005] To date, there are 23 known FGFs, each with the capacity to bind one or more FGFRs (Zhang et al., J.BioI. Chem. 281:15, 694-15,700 (2006)). Several FGFs can bind to and activate each of one or more FGFRs, often with large differences, for 2 order of magnitude differences in their affinities for the different FGFRs. Many FGFs bind their respective FGFRs with very high affinities, some in the picomolar range. Heparin is required for the binding of FGFs to FGFRs under some circumstances (Ornitz et al., Mol. Cell Biol. 12:240 (1992)). For example, the mitogenic response to FGF-2 (also known as basic FGF (bFGF)) mediated by FGFR1 has been shown to depend on the presence of heparin (Omitz et al., Mol. Cell Biol. 12:240 (1992)). (006] Previously proposed therapeutic approaches using specific antibodies to block FGF function do not address the issue of redundancy within the FGF family in activating multiple FGFRs, since cancers or other proliferative cells may express upregulated levels of more than one FGF or FGFR. Antisense oligonucleotide or related siRNA therapies have potential problems with specificity, serun half-life, and intracellular delivery. Gene transfer therapies, including those using adenovirus, have raised issues of patient safety and a number of clinical gene therapy studies have been halted due to patient death. Small molecule tyrosine kinase inhibitor therapies suffer from issues of target specificity, toxicity, and manifestations of drug resistance. To date, no drug which targets an FGFR signaling pathway has been approved for treating any human disease. SUMMARY [007] The invention provides an FGFR fusion protein comprising a first polypeptide that comprises an extracellular domain of an FGFR polypeptide and a fusion partner, wherein the extracellular domain comprises a C-terminus, wherein the C-terminus comprises a variant of a wildtype FGFR extracellular domain C-terminus, wherein the variant comprises a deletion of 1 - 22 amino acid residues present in a wildtype FGFRI, FGFR2, FGFR3, or FGFR4 extracellular domain C-terminus, and wherein the FGFR fusion protein binds at least one FGF ligand or a biologically active fragment thereof. In an embodiment, the deletion is C-terminal to a valine residue situated at the C-terminus of the IgM domain and commonly aligned among the wildtype FGFRi, FGFR2, FGFR3, and FGFR4 extracellular domain C-termini. In an embodiment, the FGFR fusion protein is less susceptible to cleavage. [008] The invention also provides an FGFR fusion protein comprising a first polypeptide that comprises an extracellular domain of an FGFR polypeptide and a fusion partner; wherein the extracellular domain comprises a C-terminus, wherein the C-terminus comprises a variant of a wildtype FGFR extracellular domain C-terminus, wherein the variant comprises at least one point mutation compared to a wildtype FGFR1, FGFR2, 3 FGFR3, or FGFR4 extracellular domain C-terminus; and wherein the point mutation renders the FGFR fusion protein less susceptible to cleavage. [009] Any of these FGFR fusion proteins may comprise an FGFR1 polypeptide, an FGFR2 polypeptide, an FGFR3 polypeptide, and/or an FGFR4 polypeptide. Any of these FGFR fusion proteins may comprise an Fc polypeptide. [010] In an embodiment, the extracellular domain of the FGFR fusion protein comprises an amino acid sequence of any of SEQ ID NO: 100, SEQ ID NO: 97 to SEQ ID NO.: 99, SEQ ID NO.: 101 to SEQ ID NO: 122, SEQ ID NO: 127 to SEQ ID NO: 132, SEQ ID NO: 137 to SEQ ID NO: 141, SEQ ID NO: 146 to SEQ ID NO: 150, SEQ ID NO: 162 to SEQ ID NO: 166, SEQ. ID. NOS.:178 to SEQ ID NO: 182, SEQ ID NO: 199 to SEQ ID NO: 203, SEQ ID NO: 206 to SEQ ID NO: 210, SEQ ID NO: 230 to SEQ ID NO: 234, and SEQ ID NO: 238 to SEQ ID NO: 242. These FGFR fusion proteins may lack a native leader sequence. In an embodiment, the Fc polypeptide comprises an amino acid sequence of any of SEQ ID NO: 171 to SEQ ID NO: 173. [011] The invention further provides an FGFR fusion protein produced in a CHO cell or a 293 cell comprising a first polypeptide comprising an extracellular domain of an FGFR polypeptide or a variant thereof and a fusion partner, wherein the FGFR fusion protein can bind to one or more FGF ligand. In an embodiment, this FGFR fusion protein comprises an amino acid sequence of any of SEQ ID NO.: 100, SEQ ID NO.: 95 to SEQ ID NO.: 99, SEQ ID NO: 102 to SEQ ID NO.: 126, SEQ ID NO: 156 to SEQ ID NO: 157, SEQ ID NO: 162 to SEQ ID NO: 166, SEQ ID NO: 176 to SEQ ID NO: 182, SEQ ID NO: 198 to SEQ ID NO: 202, SEQ ID NO: 205 to SEQ ID NO: 210, SEQ ID'NO: 228 to SEQ ID NO: 234, and SEQ ID NO: 236 to SEQ ID NO: 242. In an embodiment, this FGFR fusion protein lacks a native leader sequence. In an embodiment, it is produced using a CHEF expression system. [012] The invention yet further provides the use of any of the above-described FGFR fusion proteins as a medicament. It provides a composition comprising an effective amount of any of the above-described FGFR fusion proteins and a phannaceutically acceptable carrier. The invention provides a kit comprising this composition in a container and instructions for its administration into a subject in need of such a composition. In an embodiment the kit comprises either a single dose or multiple doses of the FGFR fusion protein. [013] The invention provides a nucleic acid molecule comprising a polynucleotide that encodes any of the above-described FGFR fusion proteins. In an 4 embodiment, a vector comprises this nucleic acid molecule and a promoter which regulates the expression of the nucleic acid molecule. The invention also provides a recombinant host cell comprising any of the above-described FGFR fusion proteins, this nucleic acid molecule, and/or this vector. In an embodiment, this recombinant host cell is a prokaryotic cell. In an embodiment, this recombinant host cell is a eukaryotic cell, for example, one of CHO or 293 lineage. In an embodiment, the invention provides a polypeptide expressed from such a recombinant host cell. [014] In another aspect, the invention provides a method of producing an FGFR fusion protein comprising providing the recombinant host cell described above and culturing it to express the FGFR fusion protein. In an embodiment, the method further comprises isolating the FGFR fusion protein from the cell culture. In an embodiment, the isolation procedure comprises contacting the expressed FGFR fusion protein with an affinity matrix, for example, Protein A, Protein G, Protein A/G, anti-Fc antibody, and anti FGFR antibody. In an embodiment, the isolation further comprises contacting the FGFR fusion protein with a hydrophobic matrix. [015] In a further aspect, the invention provides a method of detecting the level of FGFR expression in a subject comprising providing a ligand to FGFR, providing a tissue sample from the subject, allowing the ligand and the sample to interact under conditions that permit specific FGFR binding, measuring the specific binding; and comparing the amount of specific binding to that of a control sample, wherein the ligand binds to at least one of FGFR1, FGFR2, FGFR3, FGFR4, and a fragment of any of these. The tissue sample may comprise a blood, sern, or plasma sample. In an embodiment, the tissue sample comprises a sample of a diseased tissue. In an embodiment, the tissue sample comprises a sample of a tumor tissue. In an embodiment, the testing comprises a protein antibody binding or competition assay, or a nucleic acid hybridization assay. In an embodiment, the ligand comprises an FGF ligand or an antibody ligand. [016] The invention also provides a method of detecting the level of FGF expression in a subject comprising providing an FGFR or fragment thereof; providing a tissue sample from the subject; allowing the ligand and the sample to interact under conditions that permit specific binding; measuring the specific binding; and comparing the amount of specific binding to that of a control sample, wherein the FGFR or fragment thereof binds to the FGF. In an embodiment, the FGF is at least one of FGF-1, FGF-2, FGF-3, FGF-4, FGF-5, FGF-6, FGF-7, FGF-8, FGF-9, FGF-10, FGF-16, FGF-17, FGF 18, FGF-19 and FGF-20.

5 (017] The invention further provides a method of inhibiting the viability or proliferation of a proliferative cell in vitro, in vivo, or ex vivo comprising providing a composition comprising an effective amount of an FGFR fusion protein, as described above, and a pharmaceutically acceptable carrier, and contacting the proliferative cell with an amount of the composition effective to inhibit the viability or proliferation of the proliferative cell. In an embodiment, the proliferative cell is present in a subject and the subject expresses a higher level of one or more FGF ligand than normal, In an embodiment, a tissue of this subject expresses a higher level of the FGF ligand than normal. In an embodiment, the FGF ligand binds to at least one of FGFRI- Fc, FGFR2 Fc, FGFR3-Fc, or FGFR4-Fc, or a variant of any of these, as determined by measuring binding interactions in real time. Foi- example, this FGF may be selected from FGF-1, FGF-2, FGF-3, FGF-4, FGF-5, FGF-6, FGF-8, FGF-9, FGF-16, FGF-17, FGF-18, FGF-19 and FGF-20. In an embodiment, this method is performed wherein the proliferative cell is present in a subject and the subject expresses a higher level of an FGFR polypeptide than normal. In an embodiment, a tissue in the subject also expresses a higher level of the FGFR polypeptide, for example, FGFR1, FGFR2, FGFR3, or FGFR4, than normal. In an embodiment, this method inhibits the viability and/or proliferation of a proliferative cancer cell, a proliferative dysplastic cell, or a proliferative endothelial cell. In an embodiment, the proliferative cell comprises a breast cell, a pancreatic cell, a prostate cell, a lung cell, an ovarian cell, a kidney cell, a brain cell, a colorectal cell, a retinal cell, or another cell selected from any of Table 7 - Table 13. In an embodiment, the subject expresses a higher level of an FGFR polypeptide than normal and a higher level of an FGF ligand than normal. [018] In yet a further aspect, the invention provides a method of treating cancer in a subject, comprising providing any of the above-described FGFR fusion proteins and administering an effective amount of the FGFR fusion protein to the subject. In an embodiment, the cancer comprises at least one subpopulation of cells that is dependent on or responsive to growth stimulation by an FGF ligand. In an embodiment, the cancer comprises at least one subpopulation of cells that is dependent on or responsive to an angiogenic factor for production of blood vessels for growth. In an embodiment, the cancer is resistant to VEGF signaling pathway inhibition. In an embodiment, the cancer comprises metastasizing cancer, for example bone metastasis. In an embodiment, the cancer comprises a hematologic cancer, for example, chronic myelogenous leukemia, chronic lymphocytic leukemia, acute myelocytic leukemia, or hairy cell leukemia. In an 6 embodiment, the cancer comprises a solid tumor. In an embodiment, the cancer comprises breast cancer, pancreatic cancer, pituitary cancer, prostate cancer, lung cancer, ovarian cancer, renal cell cancer, oral squamous cell cancer, colorectal cancer, bladder cancer, retinal cancer, brain cancer, or another cancer listed in Table 7 - Table 13. [019) In an embodiment, this method further comprises administering a second anti-cancer therapeutic to the subject, foi. example, one comprising a cytostatic agent, a cytotoxic agent, an anti-angiogenic agent, a second FGFR fusion protein, an inhibitor of PDGF signaling, an inhibitor of VEGF signaling, or an inhibitor of EGF signaling. The second anti-cancer therapeutic may be administered before, after, or contemporaneously with the administration of the FGFR fusion protein. [020] The invention also provides a method of inhibiting angiogenesis in a subject comprising providing any of the above-described FGFR fusion proteins and administering an amount of the FGFR fusion protein to the subject effective to inhibit angiogenesis. In an embodiment, this method further comprises administering a second therapeutic agent to the subject, for example, a cytostatic agent, a cytotoxic agent, a second anti-angiogenic agent, or a second anti-angiogenic agent. [021] In an embodiment of this method, the subject is treated for macular degeneration. In an embodiment, the subject is treated for cancer. In an embodiment, the second therapeutic agent comprises an anti-cancer therapeutic agent, for example, a second FGFR fusion protein, an inhibitor of PDGF signaling, an inhibitor of VEGF signaling, an inhibitor of EGF signaling, an antibody, or an siRNA. In an embodiment, the invention provides a method of treating angiogenesis in a subject that has been or is being treated with Avastin". [022] The invention provides a method of inhibiting the viability or proliferation of a proliferative cell in viiho, in vivo, or ex vivo; a method of treating cancer, and an method of treating angiogenesis by administering a composition comprising an effective amount of any of the above-described FGFR fusion proteins intravenously, intramuscularly, subcutaneously, topically, orally, intraperitoneally, intraorbitally, by implantation, by inhalation, intrathecally, intraventricularly, and/or intranasally. In an embodiment, the method further comprises administering a second, anti-cancer therapeutic agent to the subject, wherein the second agent comprises surgery, chemotherapy, radiation therapy, and/or the administration of another biologic. [023] The invention also provides the use of any of the above-described FGFR fusion proteins for the manufacture of a medicament for treatment of a proliferative 7 disease, for example, cancer or macular degeneration. In an embodiment, the cancer comprises a hematologic cancer or a solid cancer. In an embodiment, the cancer comprises breast cancer, pancreatic cancer, pituitary cancer, prostate cancer, hmg cancer, ovarian cancer, renal cancer, oral cancer, colorectal cancer, bladder cancer, retinal cancer, brain cancer, or another cancer identified in Table 7 - Table 13. [0241 The invention provides a product comprising an FGFR fusion protein as described above and at least one anti-cancer therapeutic as a combined preparation for simultaneous, separate, or sequential use to treat cancer. BRIEF DESCRIPTION OF TIE DRAWINGS AND TABLES Brief Description of the Drawings [025) The accompanying drawings, which are incorporated into and constitute a part of this specification, illustrate several embodiments consistent with the invention. Together with the description, they serve to explain the principles of the invention, but do not limit the invention. [026] FIG. 1A shows an amino acid sequence alignment of a portion of the extracellular and transmenbrane domains of the seven FGFR isoforms, FGFR3-Blb, FGFR3-IIlc, FGFR1-Ib, FGFRI -Mc, FGFR2-lb, FGFR2-Iftc, and FGFR4, denoting the immunoglobulin (Ig) III domain, the truncation locations of the FGFR1 mutants R1Mut1, R1Mut2, R1Mut3, RlMut4, and R1Mut5 variants, and the position of the Fc portion of the fusion protein. [027] FIG. 1B shows the same amino acid sequence alignment as FIG 1A, denoting the locations of the FGFR1 mutants RlMut6, RlMut7, RlMut8, R1Mut9 and RiMut1o. [028] FIG. 2 shows the same amino acid sequence alignment as FIG. 1, denoting the truncation locations of the FGFR4 mutants R4Mutl, R4Mut2, R4Mut3, R4Mut4, R4Mut5, and R4Mut6. (0291 FIG. 3A shows a Western blot demonstrating that RiMut1, RlMut2, RlMut3, RlMut4, and RlMut5 were more resistant to proteolytic cleavage by MMP-2 than the parental FGFR1-lec-Fc. [030] FIG. 3B shows a quantitative Western blot demonstrating that RlMut4 is more resistant to MMP-2 cleavage, compared to parental FGFRl-mIc-Fc. Quantitative standards of FGFR1-flhc-Fc are shown on the left and cell culture medium from FGFRI Mc-Fc and RlMut4 are shown on the right.

8 [031] FIG. 4 shows a Western blot demonstrating that R4Mutl, R4Mut2, R4Mut3, R4Mut4, and R4MutS were more resistant to proteolytic cleavage by MMP-2 than the full-length parental FGFR4-Fc. [032] FIG. 5A shows a competition ELISA assay measuring the binding of increasing concentrations (ug/ml) of FGFR4 mutants to FGF-1 by measuring changes in

OD

4 5 0 . R4Mutl, R4Mut2, R4Mut3, R4Mut4, R4Mut5, and R4Mut6 each has a higher affinity for FGF-I than the parental FGFR4-Fc, whereas human IgG did not bind to FGF 1. [033] FIG. 5B shows a competition ELISA assay measuring the binding of increasing concentrations (ug/ml) of FGFR4 mutants to FGF-2 by measuring changes in

OD

450 . R4Mutl, R4Mut2, R4Mut3, R4Mut4, and R4Mut5 each has a higher affinity for FGF-2 than the parental FGFR4-Fc, whereas human TgG did not bind to FGF-2. [034] FIG. 5C shows a competition ELISA assay measuring the binding of increasing concentrations (ug/ml) of FGFR4 mutants to FGF-8b by measuring changes in OD450. R4Mutl, R4Mut2, R4Mut3, R4Mut4, R4Mut5, and R4Mut6 each has a higher affinity for FGF-8b than parental FGFR4-Fc, whereas human IgG did not bind to FGF-1. [035] FIG. 6 shows a direct ELISA assay measuring the binding of increasing concentrations (ug/mi) of FGFR4 mutants to FGF-2 by measuring changes in OD 4 50 . RiMuti, RIMut2, RlMut3, and RlMut4, but not RIMut5, were able to bind FGF-2 as well as or better than parental FGFR1-lIc-Fc. . [036] FIG. 7 shows a competition ELISA assay measuring the binding of increasing concentrations (ug/ml) of FGFR4 fusion proteins produced in either 293-6E cells or CHO cells to FGF-2. FGFR-Ilc-Fc produced by either 293-6E or by CHO host cells were able to bind and sequester FGF-2 to approximately the same extent, whereas human IgG was not able to bind FGF-2. [037] FIG. 8A shows a competition ELISA assay measuring the binding of increasing concentrations (ug/ml) of parental FGFR-Illc-Fc and R 1Mut4, both produced by DG44 host cells, were able to bind and sequester FGF-1 to approximately the same extent. Human IgG was not able to bind FGF-1. [038] FIG. 813 shows a competition ELISA assay measuring the binding of increasing concentrations (ug/ml) of parental FGFR-IIlc-Fc and RlMut4, both produced by DG44 host cells, were able to bind and sequester FGF-2 to approximately the same extent. Human IgG was not able to bind FGF-2.

9 [039] FIG. 9 shows a competition ELISA assay measuring the binding of increasing concentrations (ug/ml) of parental FGFR-BIc-Fc and R1Mut4, both produced by DG44 host cells, were able to bind and sequester FGF-8b to approximately the same extent. Human IgG was not able to bind FGF-8b. [040) FIG. 10 shows a competition ELISA assay measuring the ability of increasing concentrations (ug/ml) of parental FGFRl-IIc-Fc, FGFR3-fIIc-Fc, and FGFR4-Fc to inhibit FGF-1 binding to FGFR1 -IIIc-Fc. [041] FIG. 11 shows a competition ELISA assay, as described in FIG. 10, measuring the ability of parental FGFR1-IIlc-Fc, FGFR3-IIc-Fc, and FGFR4-Fc to inhibit FGF-2 binding to FGFR1-Mc-Fc. [042] FIG. 12 shows a competition ELISA assay, as described in FIG. 11, measuring the ability of FGFRI-UIc-Fc, FGFR3-Ic-Fc, and FGFR4-Fc to inhibit FGF-8 binding to FGFR1-IImc-Fc. [043] FIG. 13 shows a competition phospho-Erk ELISA assay, measuring changes in OD 45 0 induced by increasing concentrations (ug/mi) of parental FGFRI -Mc-Fc made by 293 cells or by CHO cells. Both parental constructs inhibited FGF-2 activated Erk phosphorylation, while human IgG was unable to do so. [044] FIG. 14 shows a competition phospho-Erk ELISA assay, measuring changes in OD 4 5 0 induced by increasing concentrations (ug/ml) of parental FGFR1-Ic-Fc, RlMut4, and RIMut5. Parental FGFRI-Mc-Fc and R.1Mut4, but not RIMut5 or human IgG, inhibited FGF-2 activated Erk phosphorylation. [045] FIG. 15 shows a CellTiterGloTM viability assay demonstrating the dose dependent inhibitory effect of FGFR1-Illc-Fc at concentrations of 20 ug/ml, 4.0 ug/mil, and 0.8 ug/ml on the viability and proliferation of U251 malignant glioblastoma cells plated at a concentration of 1000 cells per well in 10% FCS. Human IgG had no effect. The positive control, TRAIL, induced maximum inhibition. [046) FIG. 16 shows a CellTiterGloTM viability assay demonstrating the dose dependent inhibitory effect of FGFR1-IIc-Fc at concentrations of 20 ug/ml, 4.0 ug/ml, and 0.8 ug/nil on the viability and proliferation of U251 malignant glioblastoma cells plated at a concentration of 5000 cells per well in 1.0% FCS. Human IgG had no effect. The positive control, TRAIL, induced maximum inhibition. [047] FIG. 17 shows a CellTiterGloTm viability assay demonstrating the dose dependent inhibitory effect of FGFR1-IIlc-Fc at concentrations of 20 ug/ml, 4.0 ug/ml, and 0.8 ug/ml on the viability and proliferation of U251 malignant glioblastoma cells 10 plated at a concentration of 10,000 cells per well in 0.1% FCS. Human IgG had no effect. The positive control, TRAIL, induced maximum inhibition. [048] FIG. 18 lists cancer cell lines which were tested for their sensitivity to inhibition of their viability and proliferation by FGFR1-IlIc-Fc. Their malignancy origins and their sensitivity to FGFRI-IIc-Fc is shown. [049] FIG. 19 shows a CellTiterGloTM viability assay demonstrating the dose dependent inhibitory effect of both FGFR1-Ifc-Fc and FGFR4-Fc on the viability and proliferation of A549 lung carcinoma cells. [050] FIG. 20 compares the inhibitory effect of FGFR1-IIlb-Fc, FGFR1-IIEc-Fc, FGFR2-IlIb-Fc, FGFR2-IIlc-Fc, FGFR3-IIMb-Fc, FGFR3-lIIc-Fc, and FGFR4-Fc on the viability and proliferation of tunor cells from tumor cell lines, showing increasing inhibition (% inhibition) with increased concentration of the fusion proteins (treatment protein (ug/ml)). Data is shown for A549 cells (FIG. 20A), U118 cells (FIG. 20B), U251 cells (FIG. 20C), SF268 cells (FIG. 20D), T47D cells (FIG. 20E), and Caki-1 cells (FIG. 20F). FIG. 20G shows the concentration-dependent inhibitory effect of parental FGFRI IIMe and RlMut4, but not RlMut5, on the viability and proliferation of A549 cells and U251 cells. [051] FIG. 21 summarizes the results shown in FIG. 20 by listing the percent inhibition of the viability and proliferation of A549, U118, U251, SF268, T47D, and Caki 1 tumor cells induced by FGFRI-Ifb-Fc, FGFR1-Ilc-Fc, FGFR2-Mb-Fc, FGFR2-IIc Fc, FGFR3-IIIb-Fc, FGFR3-Ic-Fc, and FGFR4-Fc, respectively. [052] FIG. 22 shows the concentration of FGFR1-IIIc-Fc (ug/ml) in the sera of three mice injected with "mini-circle" vector cDNA encoding FGFRI-llc-Fc using the hydrodynamic tail vein injection method, as measured by direct ELISA assay and monitored for about 45 days post injection. [053] FIG. 23 shows the presence of functional, circulating FGFR1-lIc-Fc in sera of mice injected with "mini-circle" vector cDNA encoding FGFR1-IlIc-Fc using the hydrodynamic tail vein injection method, compared to sera from control mice, as measured by a quantitative competition ELISA assay. [054] FIG. 24 shows of the serum concentration of functional, circulating R1Mut4 in the sera of each of four mice by mice injected with "mini-circle" vector cDNA encoding R1Mut4 using the hydrodynamic tail vein injection method, measured by a quantitative competition ELISA assay on days 2 and 7 post-injection.

11 {055] FIG. 25 shows the inhibitory effect of FGFR1-EElc-Fc on Caki-1 tumor growth in an in vivo xenograft mouse model of tumor growth in which mice were injected with "mini-circle" vector cDNA encoding FGFRI-IIfc-Fc using the hydrodynamic tail vein injection method. The tumor volume increased more slowly in the animals treated with FGFRI-Illc-Fc than with saline. [056] FIG. 26 shows a Western blot of sera taken from mice 5 min, 30 min, 24 hr, 48 hr, and 72 hr after injection with purified FGFRI-IIlc-Fc produced from 293 cells or from CHO cells, compared to control mouse serum. The blot shows immunoreactivity with anti-human Fc antibody and demonstrates that FGFRI-IIlc-Fc from CHO cells persisted in the circulation longer (more than 72 hours after injection) and thus was more stable in vivo than FGFRl-Mc-Fc from 293-6E cells, which was undetectable after 24 hours. [057] FIG. 27 shows a chromatographic analysis of the N-labeled glycans of FGFRl-mc-Fc expressed from 293-6E cells (uppbr panel) and CHO cells (lower panel). [058] FIG. 28 shows the amount of FGFRI-HIc-Fc in mouse serum measured for 25 days following injection of FGFR1-Ulc-Fc protein, as detected by quantitative direct ELISA. [059] FIG. 29 shows the amount of functional FGFRI-Mc-Fc present in the sera of mice 30 min, 5 hours, 1 day, 2 days, 3 days, 4 days, 5 days, 7 days, 14 days, and 25 days following injection with FGFR1-Hlc-Fc protein, as measured by a quantitative FGF 2 competition ELISA assay. The decrease in functional FGFR1 -Inc-Fc is measured as a decrease in OD 450 resulting from an increasing volume of serum (ul). Functional FGFR1 IIc-Fc remains in mouse serum more than 14 days following injection with FGFRI-IIMc Fc. (060] FIG. 30A shows the inhibitory effect of three concentrations of FGFR1 IIc-Fc on Caki-1 tumor growth in an in vivo xenograft mouse model of tumor growth in which mice were injected intravenously with FGFR1-IIc-Fe fusion protein. [061] FIG. 30B shows the concentration-dependent inhibitory effect ofFGFR1 IIc-Fe and the inhibitory effect of R IMut4 on Caki- 1 tumor growth, as described in FIG. 30A. [062] FIG. 31 shows a quantitative Western blot of sera from mice injected with either FGFR1-Ic-Fc or R1Mut4 proteins at 4 hr, 3 days, and 7 days after injection, blotted with an anti-human Fc antibody, and compared to a set of FGFR1-Mc-Fc standards. FGFRI-Me-Fc and RlMut4 were both stable in vivo in mice up to at least 12 about 7 days following injection with "mini-circle" vector cDNA encoding FGFR1--IIc-Fc or RlMut4 by hydrodynamic tail vein injection. [063] FIG. 32 shows the real-time ligand binding profiles of FGFR1-IIfc-Fc and R1Mut4 to FGF-1, FGF-2, FGF-4, and FGF-5, measured by surface plasmon resonance. Definitions [064) The terms used herein have their ordinary meanings, and, more specifically, as sct forth below, and as can be further understood in the context of the specification. [065] The terms nucleicc acid molecule" and "polynucleotide" may be used interchangeably to refer to a polymer of nucleotides, such as DNA; RNA; RNAi; siRNA, whether genomic or cDNA or cRNA or anti-sense RNA; and may contain natural or non natural nucleic acids or polynucleotides or active fragments thereof. [066] The terms "polypeptide" and "protein" are used interchangeably to refer to a polymer of amino acid residues, comprising natural or non-natural amino acid residues, and are not limited to a minimum length. Thus, peptides, oligopeptides, dimers, multimers, and the like are included within the definition. Both full-length proteins and fragments thereof are encompassed by the definition. The terms also include post translational modifications of the polypeptide, including, for example, glycosylation, sialylation, acetylation, and phosphorylation. Furthermore, a "polypeptide" herein also refers to a modified protein such as single or multiple amino acid residue deletions, additions, and substitutions to the native sequence, as long as the protein maintains a desired activity. For example, a seine residue may be substituted to eliminate a single reactive cysteine or to remove disulfide bonding or a conservative amino acid substitution may be made to eliminate a cleavage site. These modifications may be deliberate, as through site-directed mutagenesis, or may be accidental, such as through mutations of hosts which produce the proteins or errors due to polymerase chain reaction (PCR) amplification. [067] An "extracellular domain" ("ECD") is the portion of a polypeptide that extends beyond the transmembrane domain into the extracellular space. The term "extracellular domain," as used herein, may comprise a complete extracellular domain or may comprise a truncated extracellular domain missing one or more amino acids. The extracellular domains of FGFRs (defined below) bind to one or more FGFs. The composition of the extracellular domain may depend on the algorithm used to determine which amino acids are in the membrane. Different algorithms may predict, and different systems may express, different extracellular domains for a given FGFR. For example, a 13 tyrosine (Y) residue may be considered as the first amino acid residue in the transmembrane domain or the last amino acid residue of the extracellular domain, depending on the method used to determine the extracellular domain. [068] A "fibroblast growth factor receptor" (FGFR) polypeptide, as used herein, is a polypeptide comprising the entirety or a portion of FGFR1, FGFR2, FGFR3, or FGFR4 including all its naturally occurring isofonns or allelic variants. An "FGFR1 polypeptide," for example, refers to a polypeptide having the amino acid sequence of any one of the known FGFRI polypeptides, such as FGFR1-Hlb and FGFRI-llc, and any fragment thereof, including those described in U.S. Patent Nos. 6,656,728; 6,384,191; 5,229,501; 6,255,454; 6,344,546; 5,474,914; and 5,288,855. FGFR1-Mb and FGFR1-fc differ from each other in their IgIIm domains (defined below). An FGFR2 polypeptide, for example, refers to a polypeptide having the amino acid sequence of any one of the known FGFR2 polypeptides, for example, FGFR2-Mb and FGFR2-Ic, and any fragments thereof. FGFR2-Illb and FGFR2-Mc differ from each other also in the IgIMI domains. An "FGFR3 polypeptide," for example, refers to a polypeptide having the amino acid sequence of any one of the known FGFR3 polypeptides, for example, FGFR3-Ilb and FGFR3-IIIc and any fragments thereof. FGFR3-lib and FGFR3-IIlc also differ from each other in their Iglfl domains. An "FGFR4 polypeptide," for example, refers to a polypeptide having the amino acid sequence of any one of the known FGFR4 polypeptides, and any fragments thereof. [069] An "FGFR fusion protein" is a protein as defined in the Tables and Sequence Listing and typically comprises a sequence of amino acids corresponding to the extracellular domain of an FGFR polypeptide or a biologically active fragment thereof, and a fusion partner. The fusion partner may be joined to either the N-terminus or the C terminus of the FGFR polypeptide and the FGFR may be joined to either the N-terminus or the C-terminus of the fusion partner. An FGFR fusion protein can be a product resulting from splicing strands of recombinant DNA and expressing the hybrid gene. It can be made by genetic engineering, for example, by removing the stop codon from the DNA sequence of the first protein, then appending the DNA sequence of the second protein in frame, so that the DNA sequence is expressed as a single protein. Typically, this is accomplished by cloning a cDNA into an expression vector in frame with an existing gene. An FGFR fusion protein may comprise a fusion partner comprising amino acid residues that represent all of, or more than one fragment of, more than one gene. An 14 FGFR fusion protein may also comprise a fusion partner which is not a polypeptide, but which is chemically attached. (070] The "valine residue that is situated at the C-terminus of the IgI1 domain and commonly aligned among the wildtype FGFR1, FGFR2, FGFR3, and FGFR4 ECD C termini" is the valine (V) residue shown in bold below. Ri-IIlb 348 ANQSAWLTVTRPVAKALEERPAVMTSPLYLE R1-XIIc 348 SHHSAWLTVL ---- EALEERPAVMTSPLYLE R2-IIb 349 ANQSAWLTVLPK-QQAPGRRKEITASPDYLE R2-IIIC 351 SFHSAWLTVLP--- -APORBKEITASPDYLE R3- IIb 347 ARKAFWLSVHGPRAAREELVEADEAGSVYAG R3-IIIc 348 SHHSAWLVVLP---ABEELVEADEAGSVYAG R4 342 SYQSAWLTVLP --- EEDPTWTAAAPEARYTD [071] A "fusion partner" is any component of a fusion molecule in addition to the extracellular domain of an FGFR or fragment thereof. A fusion partner may comprise a polypeptide, such as a fragment of an immunoglobulin molecule, or a non-polypeptide moiety, for example, polyethylene glycol. The fusion partner may comprise an oligomerization domain such as an Fc domain of a heavy chain immunoglobulin. (072] An "FGF ligand" is a fibroblast growth factor, or variant or fragment thereof, which binds to an FGFR. Currently, the known FGF ligands include FGF-1, FGF-2, FGF-3, FGF-4, FGF-5, FGF-6, FGF-7, FGF-8, FGF-9, FGF-10, FGF-1 1, FGF-12, FGF-13, FGF-14, FGF-15, FGF-16, FGF-17, FGF-18, FGF-19, FGF-20, FGF-21, FGF 22, and FGF-23. Each FGF may bind to one or more FGFR. An FGF ligand is "over expressed" when it is expressed at a higher level than normal for the cell, tissue, or organism expressing the ligand. (073] A "fragment crystallizable (Fc) polypeptide" is the portion of an antibody molecule that interacts with effector molecules and cells. It comprises the C-terminal portions of the iumunoglobulin heavy chains. As used herein, an Fc polypeptide comprises a fragment of the Fe domain with one or more biological activity of an entire Fc polypeptide. An "effector function" of the Fc polypeptide is an action or activity performed in whole or in part by an antibody in response to a stimulus and may include complement fixation or ADCC (antibody-dependent cellular cytotoxicity) induction. [074) "Wildtype" refers to a non-mutated version of a gene, allele, genotype, polypeptide, or phenotype, or a fragment of any of these. It may occur in nature or be produced recombinantly. A "wildtype FGFR ECD" refers to a protein or a nucleic acid molecule that contains an amino acid sequence or nucleic acid sequence that is identical to 15 that of a wildtype extracellular domain of an FGFR, in whole or in part, including all isoforms of FGFRI, FGFR2, FGFR3, and FGFR4. [075] A "variant" is a nucleic acid molecule or polypeptide that differs from a referent nucleic acid molecule or polypeptide by single or multiple amino acid substitutions, deletions, and/or additions and substantially retains at least one biological activity of the referent nucleic acid molecule or polypeptide. [076] A "point mutation" is a mutation that involves a single nucleotide or amino acid residue. The mutation may be the loss of a nucleotide or amino acid, substitution of one nucleotide or amino acid residue for another, or the insertion of an additional nucleotide or amino acid residue. [077] "Leader sequence" refers to a sequence of amino acid residues or polynucleotides encoding such, which facilitates secretion of a polypeptide of interest and is typically cleaved upon export of the polypeptide to the outside of the cell surface membrane. [0781 A "vector" is a plasmid that can be used to transfer DNA sequences from one organism to another or to express a gene of interest. A vector typically includes an origin of replication and regulatory sequences which regulate the expression of the gene of interest, and may or may not carry a selective marker gene, such as an antibiotic resistance gene. A vector is suitable for the host cell in which it is to be expressed. A vector may be termed a "recombinant vector" when the gene of interest is present in the vector. [079] A "host cell" is an individual cell or cell culture which can be or has been a recipient of any recombinant vector or isolated polynucleotide. Host cells include the progeny of a single host cell, which may not necessarily be completely identical, for example, in morphology or in total DNA complement, to the original parent cell due to natural, accidental, or deliberate mutation and/or change. A host cell includes a cell transfected or infected in vivo or in vitro with a recombinant vector or a polynucleotide of the invention. A host cell which comprises a recombinant molecule may be called a "recombinant host cell." Host cells may be prokaryotic cells or eukaryotic cells. Eukaryotic cells suitable for use as host cells include manunalian cells, such as primate or non-primate animal cells; fungal cells; plant cells; and insect cells. For example, host cells may be derived from 293 or CHO cells. [080] A "promoter" is a DNA regulatory region capable of binding RNA polymerase in a cell and initiating transcription of a downstream (5' to 3' direction) coding sequence operably linked thereto. Promoters include those that are naturally contiguous to 16 a nucleic acid molecule and those that are not naturally contiguous to a nucleic acid molecule. Additionally, the term "promoter" includes inducible promoters, conditionally active promoters such as a cre-lox promoter, tet inducible promoters, constitutive promoters, and tissue specific promoters. An "exogenous promoter" is one that is not operatively linked to a gene of interest in the naturally-occurring state. [081] "CHEF expression system" refers to an expression system utilizing regulatory DNA sequences derived from the hamster elongation factor-I (EF-1) alpha gene, as described in US Patent No. 5, 888,809 as the regulatory sequences. CHEF expression systems may use Chinese hamster ovary (CHO) cells as host cells. [082] The CHEF expression system comprises regulatory DNA sequences 5' to translated regions of the Chinese hamster ovary EF-lalpha gene and includes approximately 3.7 kb DNA extending from a Spel restriction site to the initiating methionine (ATG) codon of the EF-lalpha protein. Polynucleotides of less than 3.7 kb are also included in as much as the smaller fragment polynucleotides are capable of increasing transcription of an operatively linked gene. Examples of plasmids containing this regulatory system include pDEF2 and pDEF10. Plasmid pDEF2 in B. coli strain XL-1 Blue was deposited with from American Type Tissue Collection, 10801 University Boulevard, Manassas, VA 20110 and assigned Accession Number 98343. (083] "Isolating" a protein fiom cell culture means separating the protein from the remainder of the materials in the cell culture. "Isolating" can mean achieving a partial or a complete separation of the protein from the culture. "Isolating" and "purifying" are used interchangeably, as are "isolated" and "purified." [084] An "affinity matrix" refers to a composition that shows preferential affinity to a polypeptide or polynucleotide of interest and is used for purification or isolation of such fiom other materials naturally present in its environment, for example, in a cell culture. Materials suitable for use as an affinity matrix include, but are not limited to, Protein A, Protein G, a combination of Protein A and G, and an antibody, such as that attached to a solid substrate. [085) A "biologically active" entity, or an entity having "biological activity," is an entity having any function related to or associated with a metabolic or physiological process, and/or having structural, regulatory, or biochemical functions of a naturally occurring molecule. Biologically active polynucleotide fragments are those exhibiting activity similar, but not necessarily identical, to an activity of a polynucleotide of the present invention. A biologically active polypeptide or fragment thereof includes one that 17 can participate in a biological reaction, including, but not limited to, a ligand-receptor interaction or antigen-antibody binding. The biological activity can include an improved desired activity, or a decreased undesirable activity. An entity may demonstrate biological activity when it participates in a molecular interaction with another molecule, such as hybridization, when it has therapeutic value in alleviating a disease condition, when it has prophylactic value in inducing an immune response, when it has diagnostic and/or prognostic value in determining the presence of a molecule, such as a biologically active fragment of a polynucleotide that may be detected as unique for the polynucleotide molecule, and when it can be used as a primer in a polymerase chain reaction (PCR). [086] "Subject," "individual," "host," "animal," and "patient" are used interchangeably herein to refer to mammals, including, but not limited to, rodents, simians, humans, felines, canines, equines, bovines, porcines, ovines, caprines, mammalian laboratory animals, mammalian farm animals, mammalian sport animals, and mammalian pets. (087] A "tissue sample" is any biological specimen derived from a patient. The term includes, but is not limited to, biological fluids such as blood, serum, plasma, urine, cerebrospinal fluid, tears, saliva, lymph, dialysis fluid, lavage fluid, semen, and other liquid samples, as well as cell and tissues of biological origin. The term also includes cells or cells derived therefrom and the progeny thereof, including cells in culture, cell supernatants, and cell lysates. It further includes organ or tissue culture-derived fluids, tissue biopsy samples, tumor biopsy samples, stool samples, and fluids extracted from physiological tissues, as well as cells dissociated from solid tissues, tissue sections, and cell lysates. This definition encompasses samples that have been manipulated in any way after their procurement, such as by treatment with reagents, solubilization, or enrichment for certain components, such as polynucleotides or polypeptides. Also included in the term are derivatives and fractions of patient samples. A patient sample may be used in a diagnostic, prognostic, or other monitoring assay. [088] A "growth factor receptor signaling inhibitor," such as a "PDGF signaling inhibitor," a "VEGF signaling inhibitor," or an "EGF signaling inhibitor" is an agent that diminishes the effects of one or more of a series of events, such as in a signaling transduction pathway, beginning with the binding of the growth factor to its receptor and ending with a biological response, such as a proliferative response. A growth factor receptor signaling inhibitor may diminish one or more of many events, occurring both in 18 series and in parallel, including signal transduction, kinase activation, gene activation, and cell cycle modulation. [089] The terms "antibody" and "immunoglobulin" refer to a protein, generated by the immune system, made synthetically, or made recombinantly, that is capable of recognizing and binding to a specific antigen; antibodies are commonly known in the art. They can be polyclonal antibodies, monoclonal antibodies, single chain antibodies or antigen binding fragments thereof. [090] "Angiogenesis" is the development of new blood vessels, including capillary vessels. It can take place in health or disease, including cancer. The term includes neovascularization, revascularization, angiopoiesis, and vasculogenesis. New blood vessel growth typically results from stimulation of endothelial cells by angiogenic factors which may be active in proliferative conditions, such as in cancer or macular degeneration. An "angiogenic factor" is one that promotes angiogenesis. [091] "Cancer" and "tumor" are interchangeable terms that refer to any abnormal cell or tissue growth or proliferation in an animal. As used herein, the terms "cancer" and "tumor" encompass solid and hematologica/lymphatic cancers and also encompass malignant, pre-malignant, and benign growth, such as dysplasia. [092] "Metastasis" is the spread or dissemination of a disease-process, for example cancer, from one part of the body to another. It includes the spread or dissemination from an initial or primary site of disease to another site. "Metastasis" also refers to the process by which such spreading or dissemination occurs. The term is not limited to the mechanism of spread or dissemination. "Metastasis" includes the spread or dissemination of cancer cells by the lymphatics or blood vessels or by direct extension through serous cavities or subarachnoid or other spaces. [093] "Macular degeneration" is any condition in which the cells of the macula lutea degenerate, eventually resulting in blurred vision and possibly in blindness. [094) "Treatment," as used herein, covers any administration or application of a therapeutic for disease in a mammal, including a human, and includes inhibiting the disease, arresting its development, or relieving the disease, for example, by causing regression, or restoring or repairing a lost, missing, or defective function; or stimulating an inefficient process. Treatment may achieved with surgery, radiation, chemotherapy, and/or with a biologic. [095] A "pharmaceutically acceptable carrier" refers to a non-toxic solid, semisolid, or liquid filler, diluent, encapsulating material, formulation auxiliary, or carrier 19 conventional in the art for use with a therapeutic agent for administration to a subject. A pharmaceutically acceptable carrier is non-toxic to recipients at the dosages and concentrations employed and is compatible with other ingredients of the formulation. The pharmaceutically acceptable carrier is appropriate for the formulation employed. For example, if the therapeutic agent is to be administered orally, the carrier may be a gel capsule. If the therapeutic agent is to. be administered subcutaneously, the carrier ideally is not irritable to the skin and does not cause injection site reaction. [096) A "biologic" is a product which is naturally produced in some form by living organisms, whether modified or unmodified, whether in whole or a fragment thereof. A biologic may be prepared from a living source, such as animal tissue. The term includes, but is not limited to, a polynucleotide, polypeptide, antibody, cell, virus, toxin, vaccine, blood component or derivative, and fusion protein. A "biologic" may be used to treat an animal, including a human. [097] "Surface plasmon resonance" is a reduction in reflected light intensity which occurs when light is reflected off a thin metal film and a fraction of the incident light energy can interact with delocalised electrons in the metal film (plasmon). FGFR Extracellular Domain Fusion Molecules (098] The FGFR fusion molecules of the invention comprise a first polypeptide that comprises an extracellular domain (ECD) of an FGFR polypeptide and a fusion partner. The FGFR polypeptide can be any of FGFRL, FGFR2, FGFR3, and FGFR4, including all their variants and isoforms. Hence, the family of FGFR polypeptides suitable for use in the invention includes FGFR-IUIb, FGFR1-IIc, FGFR2-IIb, FGFR2-IIlc, FGFR3-Ib, FGFR3-UlIc, and FGFR4, for example. The extracellular domain of the FGFR can be the entire BCD or a portion thereof. The FGFR ECD is modified, as compared to the wildtype FGFR ECD, and possesses ligand binding activity. The modifications may be single or multiple amino acid deletions, additions, or substitutions. FGFR extracellular domains can be attached to fusion partners that provide desired pharmacokinetic properties, for example, increasing their half-life in vivo. The fusion partner of the FGFR fusion molecules of the invention can be any fusion partner conventional in the art, including those having oligomerization domains, such as dimerization domains, for example, an Fe fragment. Fusion partners of the invention also include those made by chemical modifications, such as pegylation. [099] FGFRs bind their cognate FGFs via their extracellular domains, thus the extracellular domain determines the ligand binding specificity. The FGFR extracellular 20 domain can comprise up to three immunoglobulin-like (Ig-like) domains, Igi, IglI, and IgIfI domains. Alternative mRNA splicing produces several forms of the extracellular domains. One splicing event eliminates the amino-terminal Ig-like domain (domain I) resulting in a short form of the receptor with only two Ig-like domains. Another mRNA splicing event takes place in FGFR1, FGFR2, and FGFR3, which results in three alternative versions of Ig-like domain III; namely, 1I1a, EIb, and IIc. So far, FGFR4 has not been reported to be alternatively spliced in this region. The third immunoglobulin-like domain can produce receptor splice variants with different ligand binding properties. [0100) The invention provides compositions comprising and methods of using such FGFR fusion molecules. FGFR fusion molecules of the invention can include the ECDs of FGFRI, for example those described in U.S. Patent Nos. 6,384,191; 6,656,728; 5,229,501; 6,344,546; and 5,474,914; including those annotated as NP_075594, NP_056934, or NP_000595, as described by the National Center for Bloinformatics Information (NCBI). FGFR fusion molecules of the invention can also include the ECDs of FGFR2, for example those annotated as 15281415 and NP_000132. FGFR fusion molecules of the invention can further include the ECDs of FGFR3, for example, those annotated as NP_056934, 17939658, P22607, NP_000133, orNP_075254. FGFR fusion molecules of the invention can yet further include the ECDs of FGFR4, for example those annotated NP_002002, 13991618, 2832350, 31372, and 182571. [0101) The fusion proteins of the invention can comprise an entire ECD or a portion of the ECD of wildtype or variant FGFRs. For example, the fusion proteins of the invention can comprise the entire FGFR1 ECD, including that of wildtype FGFR1-HIb or wildtype FGFR1-IIc ECD. The invention can also comprise a variant of the wildtype FGFR1 ECD, such as one having deletion of one or more and up to 22 amino acid residues, counting from the C-terminus of the wildtype FGFR1 ECD, provided that the variant ECD retains at least one of its FGF ligand binding activities. In an embodiment, the FGFR1 ECD has the 22 amino acids at the C-terminus deleted. In an embodiment, the deletion does not extend to or include the valine residue at amino acid residue 356 of the wildtype full length FGFR1-IIlb or FGFR1-IIIc. Examples of such variants include those having amino acid residues LYLE deleted (SEQ ID NO: 243), those having amino acid residues PLYLE (SEQ ID NO: 244) deleted, those having amino acid residues MTSPLYLE deleted (SEQ ID NO: 245), those having amino acid residues AVMTSPLYLE deleted (SEQ ID NO: 246), those having amino acid residues VMTSPLYLE (SEQ ID NO: 247) deleted, those having amino acid residues 21 EERPAVMTSPLYLE deleted (SEQ ID NO: 248), those having amino acid residues LEERPAVMTSPLYLE deleted (SEQ ID NO: 249), those having amino acid residues KALEERPAVMTSPLYLE deleted (SFQ ID NO: 250), those having amino acid residues EALEERPAVMTSPLYLE deleted (SEQ lID NO: 251), and those having amino acid residues RPVAKALEERPAVMTSPLYLE deleted (SEQ ID NO: 252), all as compared to wildtype FGFRL-Ib or FGFR1-ec. [0102] In an embodiment, the fusion proteins of the invention can comprise an entire FGFR2 ECD, including that of wildtype FGFR2-mb or wildtype FGFR2-lIc ECD. The invention can also comprise a variant of the wildtype FGFR2 ECD, such as one having deletion of one or more and up to 22 amino acid residues, counting from the C terminus of the wildtype FGFR2 ECD, provided that the variant ECD retains at least one of its FGF ligand binding activities. In an embodiment, the FGFR2 ECD has the 22 amino acids at the C-terminus deleted. In an embodiment, the deletion does not extend to or include the valine residue at amino acid residue 357 of the wildtype full length FGFR2 Mb or amino acid residue 359 of the wildtype full length FGFR2-HIc. Examples of such variants indlude those having amino acid residues DYLE deleted (SEQ ID NO: 253), those having amino acid residues PDYLE deleted (SEQ ID NO: 254), those having amino acid residues TASPDYLE deleted (SEQ ID NO: 255), those having amino acid residues ITASPDYLE deleted (SEQ ID NO: 256), those having amino acid residues EITASPDYLE deleted (SEQ ID NO: 257), those having amino acid residues GREKEI TASPDYLE deleted (SEQ ID NO: 258), those having amino acid residues PGREKEIT ASPDYLE deleted (SEQ ID NO: 259), those having amino acid residues APGREKEIT ASPDYLE deleted (SEQ ID NO: 260), those having amino acid residues PAPGREKE ITASPDYLE deleted (SEQ ID NO: 261), those having amino acid residues QAPGRE KBITASPDYLE deleted (SEQ ID NO: 262), and those having amino acid residues PKQQAPGREKEITASPDYLE deleted (SEQ ID NO: 263), all as compared to wildtype FGFR2-IIlb or FGFR2-Ilc. [0103] In an embodiment, the fusion proteins of the invention can comprise an entire FGFR3 ECD, including that of wildtype FGFR3-lflb or wildtype FGFR3-mc ECD. The invention can also comprise a variant of the wildtype FGFR3 ECD, such as one having deletion of one or more and up to 22 amino acid residues, counting from the C terminus of the wildtype FGFR3 ECD, provided that the variant ECD retains at least one of its FGF ligand binding activities. In an embodiment, the FGFR3 ECD has the 22 amino acids at the C-terminus deleted. In an embodiment, the deletion does not extend to or 22 include the valine residue at amino acid residue 355 of the wildtype full length FGFR3 IlIb or amino acid residue 356 of the wildtype full length FGFR3-IIc. Examples of such variants include those having amino acid residues VYAG deleted (SEQ ID NO: 264), those having amino acid residues SVYAG deleted (SEQ ID NO: 265), those having amino acid residues EAGSVYAG deleted (SEQ ID NO: 266), those having amino acid residues DEAGSVYAG deleted (SEQ ID NO: 267), those having amino acid residues ADEAGSVYAG deleted (SEQ ID NO: 268), those having amino acid residues ELVEADEAGSVYAG deleted (SEQ ID NO: 269), those having amino acid residues EELVEADEAGSVYAG deleted (SEQ ID NO: 270), those having amino acid residues ABEELVEADEAGSVYAG deleted (SEQ ID NO: 271), those having amino acid residues PAEEELVEADEAGSVYAG deleted (SEQ ID NO: 272), and those having amino acid residues GPRAAEEE LVEADEAGSVYAG deleted (SEQ ID NO: 273), all as compared to wildtype FGFR3-Ilb or FGFR3-IIc. [0104] In an embodiment, the fusion proteins of the invention can comprise an entire FGFR3 ECD, including that of wildtype FGFR4 BCD. The invention can also comprise a variant of the wildtype FGFR4 ECD, such as one having deletion of one or more and up to 22 amino acid residues, counting from the C-terminus of the wildtype FGFR4 ECD, provided that the variant ECD retains at least one of its FGF ligand binding activities. In an embodiment, the FGFR4 ECD has the 22 amino acids at the C-terminus deleted. In an embodiment, the deletion does not extend to or include the valine residue at amino acid residue 351 of the wildtype full length FGFR4. Examples of such variants include those having amino acid residues RYTD deleted (SEQ ID NO: 274), those having amino acid residues ARYTD deleted (SEQ ID NO: 275), those having amino acid residues APEARYTD deleted (SEQ ID NO: 276), those having amino acid residues AAPEARYTD deleted (SEQ ID NO: 277), those having amino acid residues AAAPEARYTD deleted (SEQ ID NO: 278), those having amino acid residues PTWTAAAPEARYTD deleted (SEQ ID NO: 279), those having amino acid residues DPTWTAAAPEARYTD deleted (SEQ ID NO: 280), those having amino acid residues EEDPTWTAAAPE ARYTD deleted (SEQ ID NO: 281), and those having amino acid residues PEEDPTWTAAAPEARYfD deleted (SEQ ID NO: 282), all as compared to wildtype FGFR4. [0105] In an embodiment, the fusion proteins of the invention can comprise variants that are point mutants, provided that they retain at least one FGF ligand binding activity. The point mutants can include any one or more of the amino acid residue 23 additions, deletions, or substitutions in the same regions of the C-terminus mentioned previously, that is, up to the valine residue at position 356 of FGFRI-Ilb or FGFRI-fc; position 357 of FGFR2-IIIb; position 359 of FGFR2-IHIc; position 355 of FGFR3-Ib; position 356 of FGFR3-fIc; and position 350 of FGFR4. For example, any one or more of the amino acid residues PAVM at positions 364 - 367 of the full length FGFR1-Mb or FGFRI-llc may be added, deleted, or substituted. [0106] The C-terminus of the extracellular and transmembrane domains of the FGFR polypeptides may differ, depending on the method used to identify the extracellular domain. Different algorithms predict different start residues for the transmembrane domain, thus different end residues for the extracellular domain. For example, the extracellular domain of FGFR1-Ic, is shown in the sequence listing (SEQ ID NO: 92) as ending with the amino acid sequence "YLE." Table 2 indicates that the transmembrane regions of FGFRl-Mc NP_056934, NP_075594, and NP_000595 begin with amino acid residues 373, which corresponds to the "L" in the "YLE" sequence. Thus, the "LE" residues may be considered as belonging either to the extracellular domain or the transmembrane domain of FGFRI-IJIc depending on the algorithm used for the prediction. This concept applies to all of the FGFRs described herein. Thus, with respect to FIG. 1, the extracellular domain of FGFR3-UIb may end in "VYAG" or "VY," the ECD of FGFR3-Me may end in "VYAG" or "VY," the ECD of FGFR1-Mb may end in "LYLE" or "LE," the ECD of FGFR1-IIc may end in "LYLE" or "LY," the ECD of FGFR2-Mb may end in "DYLE" or "DY," the ECD of FGFR2-UIc may end in "DYLE" or "DY," and the ECD of FGFR4 may end in "RYTD" or "RY." [0107) The invention provides FGFR fusion proteins comprising a fusion partner. The fusion partner can be a molecule having a dimerization domain, such as an Fc fragment of an immunoglobulin heavy chain. The Fc fragment may be a wildtype Fc . found in a naturally occurring antibody, a variant thereof, or a fragment thereof. In an embodiment, the Fc fragment belongs to the IgG1, IgG2, or IgG4 class. In an embodiment, the invention provides fusion molecules including, but not limited to, an Fc fragment of an immunoglobulin molecule belonging to the IgG1 class and/or having a C237S mutation. [0108] The invention provides FGFR fusion proteins comprising a linker which connects the first polypeptide and fusion partner. The fusion proteins of the invention may be used interchangeably with or without such a linker. In an embodiment, the linker comprises the amino acids GS or any nucleotide sequence encoding GS. The linker may 24 be convenient for constructing at least the first DNA construct in attaching the nucleic acid encoding the fusion protein to the nucleic acid encoding the FGFR BCD. Any linker conventional in the art may be used for this purpose to the extent that it does not diminish the desired properties of the fusion protein. [0109] The invention also provides multimeric FGFR fusion proteins that comprise more than one FGFR extracellular domain. For example, the invention provides FGFR fusion proteins, where the fusion partner comprises a second FGFR extracellular domain which is the same as the first polypeptide, fonning a homodimer; or an a second FGFR extracellular domain which is different from the first polypeptide, forming a heterodimer, or a biologically active fragment of either of these. Such a fusion protein may increase the affinity of the fusion protein to FGF ligand binding or expand the range of FGF ligands that can bind to the fusion protein. Its components may include two FGFR1 extracellular domains, two FGFR2 extracellular domains, two FGFR3 extracellular domains, two FGFR4 extracellular domains, or biologically active fragments of any of these. It may also include heterologous combinations of FGFR1, FGFR2, FGFR3, and FGFR4 extracellular domains, or biologically active fragments of any of these. [0110] The sequences of the FGFR fusion molecules of the invention are provided in the Sequence Listing and are further described in the Tables. Their sequence designations include both published sequences and the novel fusion proteins of the invention. The types of sequences include extracellular domains, linkers, fusion partners, deletion mutants, and other mutants. [0111] Table 1 shows the FGFR sequences of the Sequence Listing. Column 1 shows the internally designated identification number (Patent ID). Column 2 shows the nucleotide sequence ID number for the nucleic acids encoding the open reading frames of some of the polypeptides listed in column 3 (SEQ. ID. NO. (NI)). Column 3 shows the amino acid sequence ID number for the polypeptide sequences (SEQ. ID. NO. (P1)). Colunm 4 provides a description of the polypeptides, including the NCBI accession numbers when applicable (Protein ID). Column 5 provides a brief description of the encoded protein, designating the FGFR family (Protein). Column 6 provides an annotation of the protein, including a description of the variant FGFR, when applicable (Annotation). Column 7 provides a description of the variant or parental construct, including amino acid residues deleted or changed, when applicable (Description).

25 Table 1: SEQ ID NOS and Protein Identification Patent SEQ. ID. SEQ. ID. ID NO. (N1) NO. (P1) Protein ID Protein Annotation Descdiption HG1020 SEQ ID SEQ ID 122 NO: 1 NO: 92 NP 056934 1-374 FGFRIIIc ECD only HGo1020 SEQ ID SEQ ID 123 NO: 2 NO: 93 NP 075594 1-285 FOFR[IIUc BCD only HG1020 SEQ ID SEQ ID 124 NO: 3 NO: 94 NP 000595 1-376 FOPRIDIc ECD only NP_056934_1 HG1021 SEQ ID SEQ ID 374_GS_1793965 FGPRl + linker 602 NO: 4 NO: 95 8 233-464 C237S FGFRIllIc + Fe parental construct NP_056934_1 HG1020 SEQ ID SEQ ID 374_17939658_23 125 NO: 5 NO: 96 3-464 C237S FGFRIIlc R1Mut6 del GS NP_056934_I HG1020 SEQ ID SEQ ID 370_17939658_23 127 NO: 6 NO: 97 3-464 C237S FGFRIDIc RIMuti del LYLEGS NP_056934_I HG1020 SEQ ID SEQ ID 366_17939658_23 del 126 NO: 7 NO: 98 3-464 C237S PGFRIII~c RlMut2 MTSPLYLEGS NP_056934_1 HG1020 SEQ H) SEQ ID 3651793965823 del 128 NO: 8 NO: 99 3-464 C237S FGFRII~c RlMut3 VMTSPLYLEGS NP_056934_I- del H01020 SEQ ID SBQ ID 360_17939658_23 EBRPAVMTSPL 129 NO: 9 NO: 100 3-464 C237S FGFR1IlIc R1Mut4 YLEGS NP_056934_1- del HG1020 SEQ ID SEQ ID 355_17939658_23 VLEALEERPAV 130 NO: 10 NO: 101 3-464 C237S PGFRIIIIo R1Mut5 MTSPLYLBGS NP_056934_1 374_D364 HG1020 SEQ ID SEQ ID D365_7939658_2 131 NO: 11 NO: 102 33-464 C237S FOFRIIIc R1Mut7 del PA 26 Patent SEQ. ID. SEQ. ID. ID NO. (NI) NO. (PI) Protein ID Protein Annotation Description NP_056934_1 374_P364G_1793 HG1020 SEQ ID SBQ ID 9658233 132 NO: 12 NO: 103 464 C237S FGPRlIc RIMut8 P364G NP_056934_1 374_P364M_1793 HG1020 SEQ ID SEQ ID 9658233 133 NO: 13 NO: L04 464 C237S FOPRII1Ic RlMut9 P364M NP_056934_1 374_M367N_1793 HG1020 SEQ ID SEQ ID 9658_233 134 NO: 14 NO: 10S 464 C237S FGFRIIe RiMutl0 M367N NP_056934_1 374_P364MM36 HOG1020 SEQ ID SEQ ID 7N_17939658_233 135 NO: 15 NO: 106 -464 C237S FOPRIII1c RiMutil P364M M367N NP_075594_ 1G1020 SEQ ID SEQ ID 2851793965823 136 NO: 16 NO: 107 3-464 C237S FGFRIUIc RIMut6 del GS NP_075594_1 HG1020 SEQ ID SEQ ID 281_17939658_23 138 NO: 17 NO: 108 3-464 C237S FGFR IlHc RIMutI del LYLEGS NP_075594_1 HG1020 SEQ ID SEQ ID 277_17939658_23 del 137 NO: 18 NO: 109 3-464 C237S PGFRII[o RIMut2 MTSPLYLEGS NP_075594_1 HG1020 SEQ ID SBQ ID 276_17939658_23 del 139 NO: 19 NO: 110 3-464 C237S FGFRiJHc RIMut3 VMTSPLYLEGS NP_0755941 - del H01020 SEQ ID SEQ ID 271_17939658_23 EERPAVMTSPL 140 NO: 20 NO: I11 3-464 C237S FGFRIIIIc RlMut4 YLEGS HG1020 SEQ ID SEQ ID NP_075594_1 141 NO: 21 NO: 112 285 D275- FGFRlIc RIMut7 del PA 27 Patent SEQ. ID. SEQ. ID. ID NO. (N) NO. (P1) Protein ID Protein Annotation Descrptilon D276_17939658_2 33-464 C237S ____.- -- NP 075594 1 285_P275G_1793 1101020 SEQ ID SEQ ID 9658233 142 - NO: 22 NO: 113 464 C237S FGFRIIlJc RIMut8 P275G NP_075594 1 285_P275M_1793 HG1020 SEQ ID SEQ ID 9658233 143 NO: 23 NO: 114 464 C237S FGFRIILIc RlMut9 P275M NP_075594_1 285_M278N_1793 HG1o2o SEQ ID SEQ ID 9658_233 144 NO: 24 NO: 115 464 C237S FOFRIlIc RiMutl0 M278N NP_075594_1 285_P275MM27 HG1020 SEQ ID SEQ ID 8N_17939658_233 145 NO: 25 NO: 116 -464 C237S PGPRIIIc RIMutII P275MM278N NP_000595_1 HG1020 SEQ ID SEQ ID 376 17939658.23 146 NO: 26 NO: 117 3-464 C237S FGFRIIIc RIMuI6 del.GS NP_000595_I HG1020 SEQ [D SEQ ID 372_17939658_23 148 NO: 27 NO: 118 3-464 C237S FGPRlHifc RiMuti del LYLEOS NP_000595_1 HO1020 SEQ ID SEQ ID 368_17939658_23 del 147 NO: 28 NO: 119 3-464 C237S FGFRIIIc RlMut2 MTSPLYLEGS NP_000595_1 H01020 SEQ 1D SEQ ID 367_1793965823 del 149 NO: 29 NO: 120 3-464 C237S FGFRIIc R1Mut3 VMTSPLYLEGS NP_000595.1- del Ho1020 SEQ ID SEQ ID 362_17939658_23 BERPAVMTSPL ISO NO: 30 NO: 121 3-464 C237S FGFRIIUc RIMut4 YLEGS 28 Patent SEQ. ID. SEQ. ID. ID NO. (NI) NO. (PI) Protein ID Protein Annotation Description NP 000595)1 376_D366 HG1020 SEQ ID SEQ ID D367_179396582 151 NO: 31 NO: 122 33-464 C2378 FGPRIIHc RlMut7 del PA NP0005951 376 P366G_1793 M01020 SBQ ID SEQ ID 9658_233 152 NO: 32 NO: 123 464 C237S PGFRIIc RIMu8 P366G NP_000595_1 376_P366M_1793 HG1020 SEQ ID SEQ ID 9658233 153 NO: 33 NO: 124 464 C2378 FGFRIfJlc RlMu(9 P366M NP 000595_1 376_M369N_1793 HG1020 SEQ ID SEQ ID 9658_233 154 NO: 34 NO: 125 464 C237S FGFRIICo RiMutlo M369N NP_000595_ 376 P366MM36 HG 1020 SEQ ID SBQ ID 9N_17939658 233 155 NO: 35 NO: 126 -464 C237S FGFPRIlUc RiMutiJ P366M M369N 14G1020 SEQ ID SEQ ID 157 NO: 36 NO: 127 NP 056934 1-370 FGFRI)lc RiMuti delLYLEGS HG1020 SEQ ID SEQ ID del 156 NO: 37 NO: 128 NP 056934 1-366 FGFRIIlIc RlMut2 MTSPLYLEGS 1101020 SEQ ID SEQ ID del 158 NO: 38 NO: 129 NP 056934 1-365 FGFRIHIc RIMut3 VMTSPLYLBGS del HG1020 SEQ ID SEQ ID BERPAVMTSPL 159 NO: 39 NO: 130 NP 056934 1-360 FGFRIITe RlMu14 YLEGS del HG1020 SEQ ID SEQ ID VLBALBERPAV 160 NO: 40 NO: 131 NP 056934 1-355 FGPR1IUe RIMut5 MTSPLYLBOS HG1020 SEQ ID SEQ ID NP 056934 1- FOFRIIMe RIMuI7 delPA 29 Patent SEQ. ID. SEQ. ID. ID NO.'(Ni) NO. (P1) Protein ID Protein Annotatlon Description 161 NO: 41 NO: 132 374 D364-D365 H01020 SEQ ID SEQ ID NP 056934_ 162 NO: 42 NO: 133 374 P3640 FGFRIIc RIMuI8 P3640 HO1020 SEQ ID SEQ ID NP_056934 1 163 NO: 43 NO: 134 374 P364M FGFRIlIUc RIMu19 P364M H01020 SEQ ID SBQ ID NP_0569341 164 NO: 44 NO: 135 374 M367N FGFRIllc RiMutI0 M367N NP_056934_1 H01020 SEQ JD SEQ ID 374_P364M_M36 165 NO:45 NO: 136 7N FGFRilc RiMutil P364MM367N HG1020 SEQ ID SEQ ID 167 NO: 46 NO: 137 NP 075594 1-281 FGFRI1I1c RlMuti del LYLBGS 1G1020 SEQ ID SEQ ID dcl 166 NO: 47 NO: 138 NP 075594 1-277 FGFR1IIc R1Mut2 MTSPLYLEGS HG1020 SEQ ID SEQ ID del 168 NO: 48 NO: 139 NP 075594 1-276 FGFRIl~c RIMut3 VMTSPLYLEGS del HG1020 SEQ 0D SEQ ID EERPAVMTSPL 169 NO: 49 NO: 140 NP 075594 1-271 PGFRIIJc RIMut4 YLEGS HG1020 SEQ 1D SEQ ID NP_075594_1 170 NO: 50 NO 141 285 D275-D276 PGFRIIlTc RiMut7 dcl PA HG1020 SEQ ID SEQ ID NP._075594_l 171 NO: 51 NO: 142 285 P2750 FGFRIIc RIMutS P2750 HG 1020 SEQ ID SEQ ID NP_075594_1 172 NO: 52 NO: 143 285 P275M FGFRIIflc KIMut9 P275M HG1020 SEQ ID SEQ ID NP_075594_1 173 NO: 53 NO: 144 285 M278N FGFRIIlc RiMutl0 M278N NP.0755941 HG1020 SEQ ID SEQ ID 285_P275MM27 174 NO: 54 NO: 145 8N FGFRIIIlc RiMutll P275MM278N H01020 SEQ ID SEQ iD 176 NO: 55 NO: 146 NP 000595 1-372 GFRIIlc RIlMuti del LYLEGS 30 Patent SEQ. ID. SEQ. ID. ID NO. (NI) NO. (PI) Protein ID Prote_ Annotation t HG1020 SEQ ID SEQ ID del 175 NO: 56 NO: 147 NP 000595 1-368 GFRIIc RiMut2 MTSPEGS 1G1020 SEQ ID SEQ ID del 177 NO: 57 NO: 148 NP 000595 1-367 FGFRlIIIc RlMut3 VMTSPLYLBGS del HGt02O SEQ ID SEQ ID BERPAVMTSPL 178 NO: 58 NO: 149 NP 000595 1-362 FGFR]llc RMut4 YLEGS HG1020 SEQ ID SEQ ID NP_000595_1 179 NO: 59 NO: 150 376 D366-D367 FGPRIll_ RiMut7 del PA H(1020 SEQ ID SEQ ID NP_000595_1 ISO NO: 60 NO: 151 376 P366G FGFIFRIlUc_ RMuI8 P366G H01020 SEQ ID SEQ ID NP_000595_I 181 NO: 61 NO: 152 376 P366M FGFR1IIIc R1Mut9 P366M HO1020 SEQ ID SEQ ID NP_000595_1 182 NO: 62 NO: 153 376 M369N FGFRIIc R1MuI0 M369N NP_000595_1 Ho 1020 SEQ ID SEQ ID 376_P366M.M36 183 NO: 63 NO: 154 9N FGFRIIUc RIMutIl P366M M369N HG1020 SEQ ID SEQ ID 184 NO: 64 NO: 155 182571 1-369 FGFR4 FGPR4 ECD 182571_1 HG1020 SEQ ID SEQ ID 369_17939658_23 FGFR4 ECD + 185 NO: 65 NO: 156 3-464 C237S FGFR4 tinker+ Fc parental construct 182571_1 369_nolinker_179 HG1021 SEQ ID SEQ ID 39658233- FGFR4 ECD + 610 NO: 66 NO: 157 464 C237S FGFR4 Fc no linker 1131020 SEQ ID SEQ ID 13991618_1-159 other FGFR4 186 NO: 67 NO: 158 FGFR4 ECD H01020 SEQ ID SEQ ID NP_002002_1-369 other FPGFR4 187 NO: 68 NO: 159 FGFR4 BCD HG020 SEQ ID SEQ ID 313721-369 other PGFR4 188 NO: 69 NO: 160 _ FGFR4 ECD 31 Patent SEQ. ID. SEQ. ID. ID NO. (NI) NO. (P1) Protein ID Protein Annotation Description H01020 SBQ ID SEQ ID 2832350_1-369 other FGFR4 189 NO: 70 NO: 161 FGFR4 BCD 182571_1 HG1021 SEQ ID SEQ ID 364 1793965823 616 NO:71 NO: 162 3-464 C237S FGFR4 R4Muti del ARYTD 182571_1 HG1021 SEQ ID SEQ ID 359_17939658_23 del 617 NO: 72 NO: 163 3-464 C237S FGFR4 R4Mut2 AAAPBARYTD 1825711- del HG1021 SBQ ID SEQ ID 354_17939658_23

-

DPTWTAAAPBA 618 NO: 73 NO: 164 3-464 C237S FGFR4 R4Mut3 RYTD 1825711- del HG1021 SEQ ID SEQ ID 352_17939658_23 BDPTWTAAAP 619 NO: 74 NO: 165 3-464 C237S FGFR4 R4Mut4 BARYTD 182571_1-

.

dcl HG1021 SEQ ID SEQ ID 351_17939658_23 PBBDPTWTAAA 620 NO: 75 NO: 166 3-464 C237S FGFR4 R4Mut5 PEARYTD leader seq for HG1020 SEQ ID SEQ ID FGFRI (all 190 NO: 76 NO: 167 NP 056934 1-19 FGFRI _three variants) HG1020 SEQ ID SEQ ID 182571_1-19 leader seq for 191 NO: 77 NO: 168 FGFR4 FGFR4 H01020 SEQ ED SEQ ID 2832350_1-21 leader seq for 192 NO: 78 NO: 169 FGFR4 FGFR4 HO1020 SEQ ID SEQ ID 118 NO: 79 NO: 170 tinker sequence linker HG1020 SEQ ID SEQ ID 17939658233 119 NO.80 NO: 171 464 C237S FG Hl01020 SEQ ID SEQ ID 34528298_241 120 NO: 81 NO: 172 468 Fc HG1020 SEQ ID SEQ ID 19684073245 121 NO: 82 NO: 173 473 FC H(1020 SEQ ID SQ D NP_000133 1-375, FGFR3IIc BCD only FGFR311[e 32 Patent SEQ. ID. SEQ. ID. DNO_._(NI) NO.(P1) ProteinID Protein Annotation Description 374 NO: 83 NO: 174

-

H01020 SEQ ID SEQ ID 375 NO: 84 NO: 175 NP 075254 1-310 FGPR3IlIc BCD only NP 000133_1 HG1021 SEQ ID SEQ ID 375GS_1793965 FGFR31IIc + 603 NO: 85 NO: 176 8 233-464 C237S FGFR3IUc GS+Fc_ _ with liner NP_000133_1 HG1021 SEQ ID SEQ ID 375_17939658_23 604 NO: 86 NO: 177 3-464 C237S FGFR3IIIc FOFR31IIc+ Fc no linker . - NP_000133-1 HG1021 SEQ ID SEQ ID 371_17939658_23 605 NO: 87 NO: 178 3-464 C237S FGFR3Ic R3Mutl del VYAGGS NP_000133_1 HG1021 SEQ ID SEQ ID 367_17939658_23 del 606 NO: 88 NO: 179 3-464 C237S FGFR3IIc R3Mut2 EAGSVYAGGS NP_000133_1 HG1021 SEQ ID SEQ ID 366_17939658_23 del 607 NO: 89 NO: 180 3-464 C237S FGFR3DIc R3Mut3 DEAGSVYAGGS NP_000133_1- del H1G1021 SEQ ID SEQ ID 361_17939658_23 ELVEADBAGSV 608 NO: 90 NO: 181 3-464 C237S FGFR3IUc R3Mut4 YAGGS NP_000133_1- del HG1021 SEQ ID SEQ ID 355_17939658_23 VLPAEEBLVBA 609 NO: 91 NO: 182 3-464 C237S PGFR3Ic R3Mut5 DEAGSVYAGGS HG1021 SEQ ID SEQ ID 621 'NO: 183 NO: 197 NP 056934 1-374 FGFRIIIb ECD only FGFRIIfb_l HG1021 SEQ ID SEQ ID 374_17939658_23 622 NO: 184 NO: 198 3-464 C237S FGFRIIIlb ECD+Pc no linker FGFRIMb_1 HG1021 SEQ ID SEQ ID 3701793965823 623 NO: 185 NO: 199 3-464 C237S FGFRIIIb RIMutI del LYLE HG1021 SEQ ID SEQ ID FGFRIUlb 1- FGFRIUlb RlMut2 del MTSPLYLE 33 Patent SEQ. ID. SEQ. ID. ID NO. (NI) NO. (P1) Proteln ID Protein Aunotation Dcscription 624 NO: 186 NO: 200 3661793965823 S3-464 C237S PGFRIUb-1 HG1021 SEQ ID SEQ ID 365_17939658_23 625 NO: 187 NO: 201 3-464 C237S FGFRIIb RIMut3 del VMTSPLYLE FGFRIIIIb_1- del HG1021 SEQ ID SEQ ID 361_1793965823 ERPAVMTSPLY 626 NO: 188 NO: 202 3-464 C237S FGFRIII~b RIMut4 LE FGFRIIIthl- del HG1021 SEQ ID SEQ ID 355_17939658_23 VLEALEERPAV 627 NO: 189 NO: 203 3-464 C237S FGFRIIIb RIMut5 MTSPLYLE HG1021 SEQ ID SEQ ID 628 NO: 190 NO: 204 P22607 1-375 FGFR31Ib ECD only P22607_1 HG1021 SEQ ID SEQ ID 375_17939658_23 629 NO: 191 NO: 205 3-464 C237S FGFR3lUb BCD + Pc no linker P22607_1 HG1021 SEQ ID SEQ ID 371_17939658_23 630 NO: 192 NO: 206 3-464 C237S FGFR31Ub R3MutI del VYAG P22607_ HG1021 SEQ ID SEQ ID 367_17939658_23 631 NO: 193 NO: 207 3-464 C237S FGFR31U1b R3Mut2 del BAGSVYAG P226071 HG1021 SEQ ID SEQ ID 366_17939658_23 632 NO: 194 NO: 208 3-464 C237S FGFR3IIb R3Mut3 del DBAGSVYAG P22607_1- del HG1021 SEQ ID SEQ ID 362_17939658_23 LVEADEAGSVY 633 NO: 195 NO: 209 3-464 C237S FGFR31llb R3Mut4 AG P226071- del HG1021 SEQ ID SBQ ID 355_1793965823 VLPAEEELVEA 634 NO: 196 NO: 210 3-464 C237S FGFR3Hlb R3Mut5 DEAGSVYAG HG1021 SEQ ID SEQ ID 635 NO. 211 NO: 227 15281415 1-378 FGFR2b BCD only _ 34 Patent SEQ. ID. SEQ. ID. ID NO. (NI) NO. (P1) Protein ID Protein Annotation Description 15281415_1 HG1021 SEQ ID SEQ ID 378_17939658_23 636 NO: 212 NO: 228 3-464 C237S FGFR2b BCD + FC 152814151 RG1021 SEQ ID SEQ ID 378_GS_1793965 637 NO: 213 NO: 229 8 233-464 C237S FGFR2b ECD + GS+ Fc 152814151 HG1021 SEQ ID SEQ ID 374_17939658_23 638 NO: 214 NO: 230 3-464 C237S FGFR2b R2Mul del_ DYLB 15281415_1 HG1021 SEQ ID SEQ ID 370_17939658_23 639 NO: 215 NO: 231 3-464 C237S FGFR2b R2Mut2 del TASPDYLB 15281415_1 HG1021 SEQ ID SEQ ID 369_17939658_23 640 NO: 216 NO: 232 3-464 C237S FGFR2b R2Mut3 del ITASPDYLB 15281415_1- del HG1021 SEQ ID SEQ ID 365_17939658_23 REKEITASPDYL 641 NO: 217 NO: 233 3-464 C237S FOFR2b R2Mut4 B 152814151- del HO 1021 SEQ , ID SEQ ID 356_17939658_23 VLPKQQAPORB 642 NO: 218 NO: 234 3-464 C237S FGFR2b R2Mu5 KB1TASPDYLE HG1021 SEQ ID SEQ ID 643 NO: 219 NO: 235 NP 000132 1-377 FGFR2c BCD only NP_000132_1 HG1021 SEQ ID SEQ ID 377_17939658_23 644 NO: 220 NO: 236 3-464 C237S FGFR2c BCD + Fc NP_000132_1 HG1021 SEQ ID SEQ ID 377_GS_1793965 645 NO: 221 NO: 237 8 233-464 C237S PGFR2c ECD + GS+ Fc NP_000132_1 HG1021 SEQ ID SEQ ED 373_17939658_23 646 NO: 222 NO: 238 3-464 C237S FGFR2c R2MutI del DYLE HG1021 SBQ ID SBQ ID NP 000132 1- FGFR2c R2Mut2 del TASPDYLB 35 Patent SEQ. ID, SEQ. ED. ID NO. (NI) NO. () Protein ED Protein Annotation Description 647 NO: 223 NO: 239 369_17939658_23 3-464 C237S NPI000132_1 HG1021 SEQ ID SEQ ID 368_17939658_23 648 NO: 224 NO: 240 3-464 C237S FGFR2c R2Mut3 del ITASPDYLE NP_000132_1- del HG1021 SEQ ID SEQ ID 36417939658;23 RBKEITASPDYL 649 NO: 225 NO: 241 3-464 C237S PGFR2c R2Mut4 B NP_00013291- del HG1021 SEQ ID SEQ ID 358_1793965823 - VLPAPGRBKEIT 650 NO: 226 NO: 242 3-464 C237S FGFR2c R2Mut5 ASPDYLB HG1021 SEQ ID deleted 651 NO: 243 RI delfapnent I FGFRI fragment LYLE HG1O2I SEQ ID deleted 652 NO: 244 RI deltagment 2 FGFRI fragment PLYLB HG1021 SEQ ID deleted 653 NO: 245 Ri delfragment 3 FGFRI fragment MTSPLYLE HG1021 SEQ ID deleted 654 NO: 246 RI delframent 4 FGFRI fragment AVMTSPLYLB HG1021 SEQ ID deleted 655 NO: 247 RI delfragment 5 FGFRI fragment VMTSPLYLB G101021 SBQ ID deleted EERPAVMTSPL 656 NO: 248 RI delfragment 6 FGFRI frgment YLB HGI 021 SEQ ID deleted LEBRPAVMTSP 657 NO: 249 RI delfragment 7 FGFRI fiagment LYLE HG1021 SEQ ID deleted KALEBRPAVMT 658 NO: 250 RI delfragment 8 FGFRI fragment SPLYLB HG1021 SEQ ID deleted BALBERPAVMT 659 NO: 251 RI delfragment 9 FGFRI fragment SPLYLE HG1021 SEQ ID RI.delfragment_1 I deleted RPVAKALEBRP 660 NO: 252 0 FGFRI fragment AVMTSPLYLE HG1021 SEQ ID deleted 661 NO: 253 1R2 delfragment I FGFR2 fragment DYLB 36 Patent SEQ. ED. SEQ. ED. IID NO. (NM) NO. (P1)T Protcln ID _ Protein Annotation Description H01021 SEQ ED deleted 662 ____ NO: 254 R12 delfragment 2- FC3FR fragment -PDYLB HGJ1021 SEQ ID) deleted 663 ___ NO: 255 R2 FOFR2met fragment TASPDYLB H01 021 SEQ ID deleted 664 NO: 256 R.2 dclfiagment 4 GR2 fragment ITASPDYLE HG01021 SEQ ID deleted 665 NO: 257 P.2 deifragment. 5 FGFR2 -fragment EITASPDYLB 1101021 SEQ ID deleted GRPEKBITASPDY 666 _____NO: 258 R12 deiagment 6 FGFR2. fragment LE 1101021 SEQ ID deleted PGREKBITASPD 667 NO:259 11R2 delfragment 7 PGFR2 fragment YU3 HG11021 SEQ ID deleted APGREKEITASP 668 NO: 260 R2 delframent -8 FCJFR2 fragment DYLE BG01021 SEQ 1.1 deleted PAPUREKEITAS 669 ______NO: 261 R.2 delfragment 9 FGFR2 fragment PDYLE EGI2IS.EQ ID R dlrg~et deleted QAPGREKEITAS 670 ____ NO: 262 0 FGFR2 Ifragment 1 PDYLB IIG1021 SEQ ID 112 delfragment-I deleted PKQQAPGREKEI 671 ___ _ NO: 263 1 FGFR2 fragment TASPDYLE HG1021 SEQ JD deleted 672 ____ NO: 264 R13 delfiragment I FGFR3 fragment VYAG HG01021 SEQ ID deleted 673 ___ _ NO: 265 R.3 delfagment.2 FGFR3 fragment SVYAG HG01021 SEQ ID deleted 674 NO: 266 R3dlrant_3 FGFR3 fragment HAGSVYAG HG01021 SEQ ID deleted 675 NO: 267 R3 defragment 4 PGFR3 fragment DEAGSVYAG HG1021 SEQ ID deleted 676 ___ _ NO: 268 R3 delragment 5 TGF113 fragment ADEAGSVYAG_ HG11021 SEQ ID deleted ELVEADEAGSV 1677 ____ NO: 269 - R3 delftagment 6 FGFR3 fragment YAG HG1021 SEQO ID P3 defragment 7 FFR3 deleted BLVEADEAGS 37 Patent SEQ. ID. SEQ. ID. ID NO. (NI) NO. (P1) Protein ED Protein Annotation Description 678 -NO: 270 fragment VYAG HG1021 SEQ ID deleted ABBELVEADBA 679 NO: 271 R3 delfragment 8 FGFR3 fragment GSVYAG HG1021 SEQ ID deleted PABBELVEADE 680 NO: 272 R3 delfragment 9 FGFR3 fragment AGSGYA _ HG1021 SEQ ID R3_delfragment_1 deleted GPRAAEBBLVB 681i _ _ NO: 273 0 FGFR3 fragment ADBAGSVYAG HG1021 SEQ ID deleted 682 NO: 274 R4 delfragment 1 FGFR4 fragment RYTD HG1021 SEQ ID deleted 683 NO: 275 R4 delfragnent 2 FGFR4 fragment ARYTI) HG02t SEQ ID deleted 684 NO: 276 R4 delfragment 3 FGPR4 fragment APEARYTD HG1021 SEQ ID deleted 685 NO: 277 R4 delfragment 4 FGFR4 fragment AAPEARYTD HG1021 SEQ ID deleted 686 NO: 278 R4 delfragment 5 FGFR4 fragment AAAPEARYTD HG1021 SEQ ID deleted PTWTAAAPBAR 687 NO: 279 R4 delfragment 6 FGFR4 fragment YTD HG1021 SEQ ID deleted DPTWTAAAPEA 688 NO: 280 R4 detfragment 7 FGFR4 fragment RYTD HG1021 SEQ ID deleted BEDPTWTAAAP 689 NO: 281 R4 delfragment 8 FGFR4 fragment EARYTD HG1021 SEQ ID deleted PEBDPTWTAAA 690 NO: 282 R4 delfragment 9 FGFR4 fragment PI3ARYTD [0112] Table 2 shows information characterizing sequences relating to full-length FGFRI, FGFR3, and FGFR4 proteins. Column 1 shows the NCBI accession number (Protein ID). Column 2 designates whether the sequence relates to FGFRI, FGFR3, or FGFR4. Column 3 shows the predicted length of the polypeptide encoded by each protein (Protein Length). Column 4 (Treevote) shows the result of an algorithm that predicts whether the predicted amino acid sequence is secreted. A Treevote at or near 0 indicates a low probability that the protein is secreted while a Treevote at or near 1.00 38 indicates a high probability that the protein is secreted. Colunn 5 shows the predicted signal peptide coordinates (Signal Peptide Coords). Column 6 shows the mature protein coordinates, which refer to the coordinates of the amino acid residues of the mature polypeptide after cleavage of the secretory leader or signal peptide sequence (Mature Protein Coords). Column 7 shows alternate predictions of the signal peptide coordinates (Altern Signal Peptide Coords). Cohunn 8 specifies the coordinates of an alternative form of the mature protein (Altem Mature Protein Coords). The alternate coordinates result from alternative predictions of the signal peptide cleavage site; their presence may, for example, depend on the host used to express the polypeptides. Column 9 specifies the number of transmembrane domains (TM). Columns 10 and 11 provide the coordinates of the transmembrane and non-transmembrane sequences of the polypeptides. The transmembrane coordinates (TM Coords) designate the transmembrane domains of the molecule. The non-transmembrane coordinates (non-TM Coords) refer to the protein segments not located in the membrane; these can include extracellular, cytoplasmic, and luminal sequences. Coordinates are listed in terns of the amino acid residues beginning with "1" for the first amino acid residue at the N-terminus of the full-length polypeptide.

39 00 r'0 C 0n Cl opCl Cl C S- 0o o ? 00 me eT a Co a Cl rl 0 00 OR ~ ~ o' o n o o o o Nl C. C> co 0 m 00 00 ; I 4a 'ee '-' e a a ' a' I in o54 o o o o o o o - o -4 en S - - - - - - U 4' 90 8 Cl3' 0 Cl C l C .~Cl~ C 0% 0 0 0 0 0 40 Nucleic Acid Molecules Encoding FGFR Fusion Molecules [0113) The present invention provides nucleic acid molecules that comprise polynucleotide sequences that encode the FGFR fusion proteins of the invention. These nuclei acid molecules can be constructed with recombinant DNA techniques conventional in the art. The nucleic acid molecules include molecules relevant to the FGFR-IUIb ECD, such as those provided in SEQ ID NOS: 183 - 189; those relevant to FGFR--Illc ECD, such as those provided in SEQ ID NOS: 1 - 63; those relevant to FGFR2-IIb ECD, such as those provided in SEQ ID NOS: 211 -218; those relevant to FGFR2-IIc ECD, such as those provided in SEQ ID NOS: 219 - 226; those relevant to FGFR3--IIb ECD, such as those provided in SEQ ID NOS: 190 - 196; those relevant to FGFR3-IImc ECD, such as those provided in SEQ ID NOS: 83 - 91; and those relevant to FGFR4 ECD, such as those provided in SEQ ID NOS: 64 - 78. [0114] The nucleic acid molecules of the invention can include polynucleotide sequences that encode all or part of the ECD of an FGFR polypeptide, with or without its native homologous secretary leader sequence. If a homologous secretoiy leader sequence is not used in the construction of the nucleic acid molecule, then another secretory leader sequence may be used, for example, any one of the leader sequences described in PCT US06/02951. [0115] Typically, the nucleic acid molecule encoding the gene of interest, the FGFR ECD, is inserted into an expression vector, suitable for expression in a selected host cell, at a linker site and the nucleic acid molecule encoding the fusion partner is inserted at the site following the FGFR ECD such that they are in-frame when the nucleic acid molecule is transcribed and translated. FGFR Fusion Molecule Expression and Production Vectors [0116] The invention provides genetically engineered recombinant vectors comprising nucleic acid molecules encoding the fusion proteins of the invention, recombinant host cells comprising the recombinant vectors, the nucleic acid molecules encoding the fusion proteins of the invention, and the production of FGFR fusion proteins and fragments thereof. Vectors of the invention include those that are suitable for expression in a selected host, whether prokaryotic or eukaryotic, for example, phage, plasmid, and viral vectors. Viral vectors may be either replication competent or replication defective retroviral vectors. Viral propagation generally will occur only in complementing host cells comprising replication defective vectors. Vectors of the 41 invention may comprise Kozak sequences (Lodish et al., Molecular Cell Biology, 4'e ed., 1999) and may also contain the ATG start codon of an FGFR extracellular domain. Vectors of the invention include "minicircle" vectors, which are described in greater detail below. [0117] Copy number and positional effects are considered in designing transiently and stably expressed vectors. Copy number can be increased by, for example, dihydrofolate reductase amplification. Positional effects can be optimized by, for example, Chinese hamster elongation factor-I vector pDEF38 (CHEF1), ubiquitous chromatin opening elements (UCOE), scaffold/matrix-attached region of human (S/MAR), and artificial chromosome expression (ACE) vectors, as well as by using site-specific integration methods known in the art. The expression constructs containing the vector and gene of interest will further contain sites for transcription initiation, termination, and, in the transcribed region, a ribosome binding site for translation. The coding portion of the transcripts expressed by the constructs can include a translation initiating codon at the beginning and a termination codon (UAA, UGA, or UAG) appropriately positioned at the end of the polypeptide to be translated. [0118] Considering the above-mentioned factors, vectors suitable for expressing FGFR fusion molecules in bacteria include pTT vectors, available from Biotechnology Research Institute (Montreal, Canada), pQE70, pQE60, and pQE-9, available from Qiagen (Mississauga, Ontario, Canada); vectors derived from pcDNA3, available from Invitrogen (Carlsbad, CA); pBS vectors, Phagescript vectors, Bluescript vectors, pNI-8A, pNH6a, pNH18A, pNH46A, available from Stratagene. (La Jolla, CA); and ptrc99a, pKK223-3, pKK 2 3 3

-

3 , pDR540, pRIT5 available from Pharmacia (Peapack, NJ). Among suitable eukaryotic vectors are pWLNEO, pSV2CAT, pOG44, pXT1, and pSG available from Stratagene (La Jolla, CA); and pSVK3, pBPV, pMSG and pSVL, available from Pharmacia (Peapack, NJ). [0119] Vectors for expressing FGFR fusion molecules include those comprising a pTT vector backbone (Durocher et al., Nucl. Acids Res. 30:E9 (2002)). -Briefly, the backbone of a pTT vector may be prepared by obtaining pIRESpuro/EGFP (pEGFP) and pSEAP basic vector(s), for example from Clontech (Palo Alto, CA), and pcDNA3. 1, pCDNA3. 1/Myc-(His) 6 and pCEP4 vectors can be obtained from, for example, Invitrogen (Carlsbad, CA). As used herein, the pTT5 backbone vector can generate a pTT5-Gateway vector and be used to transiently express proteins in mammalian cells. The pTT5 vector can be derivatized to pTT5-A, pTT5-B, pTT5-D, pTTS-E, pTT5-H, and pTT5-I, for 42 example. As used herein, the pTT2 vector can generate constructs for stable expression in mammalian cell lines. [0120] A pTT vector can be prepared by deleting the hygromycin (BsnI and SalI excision followed by fill-in and ligation) and EBNA1 (ClaI and NsiI excision followed by fill-in and ligation) expression cassettes. The ColEI origin (Fspl-SalI fragment, including the 3' end of the 0-lactamase open reading frame (ORF) can be replaced with a FspI-Sal fragment from pcDNA3.1 containing the pMBI origin (and the same 3' end of p-lactaniase ORF). A Myc-(His) 6 C-terminal fusion tag can be added to SEAP (flindII-Hpal fragment from pSEAP-basic) following in-frame ligation in pcDNA3. 1/Myc-His digested with HindII and EcoRV. Plasmids can subsequently be amplified in E. coli (DH56) grown in LB medium and purified using MAXI prep columns (Qiagen, Mississauga, Ontario, Canada). To quantify, plasmids can be subsequently diluted in, for example, 50 mM Tris HCI pH 7.4 and absorbencies can be measured at 260 nm and 280 am. Plasmid preparations with A 260

/A

2 go ratios between about 1.75 and about 2.00 are suitable for producing the FGFR constructs. [0121] The expression vector pTTS allows for extrachromosomal replication of the eDNA driven by a cytomegalovirus (CMV) promoter. The plasmid vector pCDNA pDEST40 is a Gateway-adapted vector which can utilize a CMV promoter for high-level expression. SuperGlo GFP variant (sgGFP) can be obtained from Q-Biogene (Carlsbad, CA). Preparing a pCEP5 vector can be accomplished by removing the CMV promoter and polyadenylation signal of pCEP4 by sequential digestion and self-ligation using Sall and XbaI enzymes resulting in plasmid pCEP4A. A GblfI fragment from pAdCMV5 (Massie et al., J. Virol. 72:2289-2296 (1998)), encoding the CMV5-poly(A) expression cassette ligated in Bgfl-linearized pCEP4A, resulting in the pCEP5 vector. [0122] Vectors for expressing FGFR fusion molecules can include those comprising vectors optimized for use in CHO-S or CHO-S-derived cells, such as pDEF38 (CHEFI) and similar vectors (Running Deer et al., Biotechnol. Prog. 20:880-889 (2004). The CHEF vectors contain DNA elements that lead to high and sustained expression in CHO cells and derivatives thereof. They may include, but are not limited to, elements that prevent the transcriptional silencing of transgenes. [0123) FGFR polynucleotide vectors may be joined to a selectable marker for propagation in a host. Generally, a selectable marker allows the selection of transformed cells based on their ability to thrive in the presence or absence of a chemical or other agent that inhibits an essential cell function. The selectable markers confer a phenotype on a 43 cell expressing the marker, so that the cell can be identified under appropriate conditions. Suitable markers, therefore, include genes coding for proteins which confer drug resistance or sensitivity thereto, impart color to, or change the antigenic characteristics of those cells transfected with a molecule encoding the selectable marker, when the cells are grown in an appropriate selective medium. [0124] Suitable selectable markers include dihydrofolate reductase or (3418 for neomycin resistance in eukaryotic cell culture; and tetracycline, kanamycin, or ampicillin resistance genes for culturing in E. coli and other bacteria. Suitable selectable markers also include cytotoxic markers and drug resistance markers, whereby cells are selected by their ability to grow on media containing one or more of the cytotoxins or drugs; auxotrophic markers, by which cells are selected for their ability to grow ondefined media with or without particular nutrients or supplements, such as thymidine and hypoxanthine; metabolic markers for which cells are selected, for example, for ability to grow on defined media containing a defined substance, for example, an appropriate sugar as the sole carbon source; and markers which confer the ability of cells to form colored colonies on chromogenic substrates or cause cells to fluoresce. [0125] As mentioned above, vectors for the expression of FGFR fusion proteins can also be constructed in proteins retroviral vectors. One such vector, the ROSAPgeo retroviral vector, which maps to mouse chromosome six, was constructed with the reporter gene in reverse orientation with respect to retroviral transcription, downstream of a splice acceptor sequence (U.S. Patent No. 6,461,864; Zambrowicz et al., Proc. NaIl. Acad. Sci. 94:3789-3794 (1997)). Infecting embryonic stem (ES) cells with ROSAgeo retroviral vector resulted in the ROSAgeo26 (ROSA26) mouse strain by random retroviral gene trapping in the ES cells. [0126] A DNA insert comprising an FGFR fusion molecule can be operatively linked to an appropriate promoter, such as the phage lambda PL promoter; the E. coli lac, trp, phoA, and tac promoters; the SV40 early and late promoters; and promoters of retroviral LTRs. Suitable promoters also include the pCMV vector with an enhancer, pcDNA3.1; the pCMV vector with an enhancer and an intron, pCIneo; the pCMV vector with an enhancer, an intron, and a tripartate leader, pTT2, and CHEF1. Other suitable promoters will be known to the skilled artisan. The promoter sequences include the minimum number of bases or elements necessary to initiate transcription of a gene of interest at levels detectable above background. Within the promoter sequence may be a transcription initiation site, as well as protein binding domains (consensus sequences) 44 responsible for the binding of RNA polymerase. Eukaryotic promoters of the invention will often, but not always, contain "TATA" boxes and "CAT" boxes. [0127] The invention provides vectors for the in vivo expression of FGFR fusion molecules in animals, including humans, under the control of a promoter that fimctions in a tissue-specific manner. For example, promoters that drive the expression of FGFR fusion proteins of the invention may be liver-specific, as described in PCTUS06/00668. [0128] A region of additional amino acids, particularly charged amino acids, may be added to the N-terminus of the polypeptide to improve stability and persistence in the host cell purification throughout and subsequent handling and storage. Also, amino acid moieties may be added to the polypeptide to facilitate purification. Such amino acids may or may not be removed prior to the final preparation of the polypeptide. The FGFR fusion proteins of the invention can be fused to marker sequences, such as a peptide, that facilitates purification of the fused polypeptide. The marker amino acid sequence may be a hexa-histidine peptide such as the tag provided in a pQE vector (Qiagen, Mississauga, Ontario, Canada), among others, many of which are commercially available. As described in Gentz et al., Proc. Natl. Acad. Sci. 86:821-824 (1989), for instance, hexa-histidine provides for convenient purification of the fusion protein. Another peptide tag useful for purification, the hemagglutinin HA tag, corresponds to an epitope derived from the influenza hemagglutinin protein (Wilson et al., Cell 37:767-778 (1984)). Any of the above markers can be engineered using the polynucleotides or the polypeptides of the present invention. [0129) The expression constructs of the invention will further contain sites for transcription initiation, termination, and, in the transcribed region, a ribosome binding site for translation. The coding portion of the transcripts expressed by the constructs can include a translation initiating codon at the beginning and a termination codon (UAA, UGA, or UAG) appropriately positioned at the end of the polypeptide to be translated. Host Cells [0130] FGFR fusion proteins of the invention can be expressed by and produced from prokaryotic cells, such as bacterial cells and eukaryotic cells, such as fungal cells, plant cells, insect cells, and mammalian cells, according to procedures known in the art, for example, as shown in the examples and figures that follow. FGFR fusion proteins can be expressed by and produced from bacterial E. coli cells; Cos 7 cells; mammalian kidney epithelial 293 cells; and Chinese Hamster Ovary (CHO) cells, including CHO-S and DG44 cells, which are derived from CHO cells, They can also be produced in vivo in animals, 45 engineered or transfected with the nucleic acid molecules encoding the fusion proteins. For example, mice injected with DNA encoding FGFR fusion molecules can express FGFR fusion molecules following tail vein transfection (TVT). [0131] Introduction of the FGFR fusion proteins into the host cell can be accomplished by calcium phosphate transfection, DEAE-dextran mediated transfection, cationic lipid-mediated transfection, electroporation, transduction, infection, or other known methods. Such methods are described in many standard laboratory manuals, such as Sambrook et al., Molecular Cloning, A Laboratoy Manual. 3" ed. Cold Spring Harbor Laboratory Press (2001). FGFR fusion proteins of the invention can be transiently or stably transfected into the host cells, as described in greater detail below. FGFR fusion proteins of the invention can be purified from.host cells grown either in adherent culture or in suspension, and, as shown in greater detail below, can retain the biological properties of FGFR. [0132] Host cells of the invention can express proteins and polypeptides in accordance with conventional methods, the method depending on the purpose for expression. For large scale production of the protein, a unicellular organism, such as E. coli, B. subtils, S. cerevisiae; insect cells in combination with baculovirus vectors; or cells of a higher organism such as vertebrates, for example mammalian 293 (including 293-6E), CHO (including DG44), or COS 7 cells, can be used as the expression host cells. In some situations, it is desirable to express eukaryotic genes in eukaryotic cells, where the encoded protein will benefit from native folding and post-translational modifications, such as glycosylation. [0133] Accordingly, the invention provides a recombinant host cell that comprises nucleic acid molecules encoding the FGFR fusion proteins, vectors comprising such nucleic acid molecules, or FGFR fusion proteins and cultures containing such. These host cells may produce FGFR1, FGFR2, FGFR3, and FGFR4 fusion proteins of the invention. For example, they may produce FGFR-Fc fusion proteins and variants and fragments thereof. The host cells may be suitable for transient transfection and for stable transfection. FGFR fusion proteins expressed by any of the methods described.herein may be detected by methods known in the art. [0134] The post-translational glycosylation of the fusion proteins of the invention may vary according to their production source, as described in greater detail below. The glycosylation profile of a protein can affect its properties and/or function. Accordingly, the invention provides recombinant FGFR fusion proteins with or without altered 46 glycosylation profiles compared to the naturally-occurring forms. They may be produced in different host cells. For example, unglycosylated FGFR fusion proteins can be produced from E. coli. A form of glycosylated FGFR fusion protein can be produced in yeast cells, such as Saccharomyces cerevisae or Pichia pastoris, or fungal cells such as Aspergillus. A form of glycosylated FGFR fusion protein can be produced in plants, such as rice, wheat, oats, etc. form of glycosylated FGFR protein can be produced in mammalian cells, such as 293 cells or CHO cells or derivatives thereof. For example, the invention provides mutant constructs with additional arginine residues for the attachment of N-linked sugars. These constructs may comprise point mutations with arginine residues or may have larger substitutions with regions that include arginine residues. The invention also provides mutant constructs with omitted arginine residues. Glycosylation mutants of the invention can be made by altering the naturally-occurring sequences using methods known to those of skill in the art. [0135] Recombinant host cells of the invention are cultured under conditions conventional the art, including both inducible and non-inducible conditions. The FGFR fusion proteins may be made inside the cells, such as in inclusion bodies, for example, when the host cell is an E. coli cell, or they may be secreted into the cell culture, such as when the host cells are mammalian cells and the proteins are expressed using mammalian expression systems, for example, using a secretory leader sequence. The invention provides cell cultures comprising the FGFR fusion protein whether the FGFR fusion protein is present in the culture medium or residing inside the cells. Purification of FGFR Fusion Proteins {0136] The invention provides methods of purifying FGFR fusion proteins using a combination of techniques, each of which is conventional in the art. These techniques include, but are not limited to, the use of affinity matrices and hydrophobic interaction chromatography, for example, affinity chromatography. Suitable affinity ligands include any ligands of the FGFR extracellular domain or of the fusion partner, or antibodies thereto. For example, a Protein A, Protein G, Protein A/G, or an antibody affinity column may be used to bind to an Fc fusion partner to purify the FGFR fusion proteins. Antibodies to the FGFR portion of the fusion protein or to the fusion partner may also be used to purify the fusion protein. Hydrophobic interactive chromatography is also suitable for purifying FGFR fusion proteins of the invention. For example, a butyl or phenyl column may be used. Other methods of purification known to those skilled in the art may also be suitable for purifying the FGFR fusion molecules of the invention.

47 [0137] Protein A affinity chromatography may be used to purify FGFR fusion proteins of the invention comprising an Fe domain. Protein A is a cell wall component produced by several strains of Staphylococcus aureus and can be made in a recombinant fashion. It consists of a single polypeptide chain weighing approximately 42,000 daltons and contains little or no carbohydrate. Protein A binds specifically to the Fc region of most immunoglobulin molecules, including IgG (Sjoquist et al., Eur. J. Biochem. 29:572 578 (1972); Hjelm et al., Eur. J. Biochem. 57:395-403 (1975)). [01381 Protein G affinity chromatography may also be used to purify FGFR fusion proteins of the invention comprising an Fc domain. Protein G is a bacterial cell wall protein produced by group G streptococci and can also be made in a recombinant fashion. Like Protein A, Protein G binds to most mammalian immunoglobulins, primarily through their Fc regions (Bjorck et al., J ihnmunol. 133:969-974 (1984); Guss et al., EMBO J. 5:1567-1575 (1986) Akerstr6m et al., J. Biol. Chen. 261:10,240-10,247 (1986)). Affinity chromatography using chimeric Fc binding molecules may further be used to purify FGFR fusion proteins of the invention comprising an Fc domain. For example, Protein A/G is a genetically engineered protein that combines the IgG binding profiles of both Protein A and Protein G. Protein A/G is a gene fusion product, which can be secreted from, inter alia, nonpathogenic Bacillus. Protein A/G typically weighs approximately 50,000 daltons and was designed to contain four Fc binding domains from Protein A and two from Protein G (Sikkema, Aner. Biotech. Lab.7:42 (1989); Eliasson et al., J. Biol. Chent. 263:4323-4327 (1988). Hydrodynamic Tail Vein Transfection (TVT) [0139] The invention provides expression of FGFR fusion proteins in animals, following a hydrodynamics-based procedure of tail vein injection (Liu, F. et al., Gene There. 6:1258-1266 (1999); U.S. Patent No. 6,627,616; and Zhang et al., Hun. Gene Ther 10:1735 (1999). This technique provides for production of the FGFR fusion protein in vivo after administering the nucleic acid molecule encoding the fusion protein produced in a mini-circle vector construct. Serum from such injected animals containing the fusion . protein may be used to further characterize the protein, without first having to produce and purify the fusion protein from cell culture expression systems. [0140] In an embodiment, the invention provides vectors comprising nucleic acid molecules encoding an FGFR fusion protein for administration to animals and FGFR fusion proteins made thereby, following hydrodynamic injection of minicircle DNA comprising such nucleic acid molecules. Vectors for injection can be constructed, for 48 example, by the system described in Chen et al., Mol. There. 8:495-500 (2003) and U.S. Pat. Apple. No. 2004/0214329 Al. In brief, an expression cassette for an FGFR gene is flanked by attachment sites for a recombinase, which is expressed in an inducible fashion in a portion of the vector sequence outside of the expression cassette. Following recombination, the E. coli produce a minicircle vector comprising an expression cassette with an FGFR fusion protein gene. The vector as described in Chen et al. can be modified by inserting the nucleic acid molecule encoding the FGFR fusion protein following the intron present in the vector, instead of in the midst of the introns. [0141] Minicircle DNA vectors can be prepared with plasmids similar to pBAD.#C3l.hFIX and pBAD.4C31.RHB and used to transform E. coli. Recombinases known in the art, for example, lambda and cre, are useful in the minicircle vectors. The expression cassettes may contain sites for transcription initiation, termination, and, in the transcribed region, a ribosome binding site for translation. The coding portion of the transcripts expressed by the constructs can include a translation initiating codon at the beginning and a termination codon (UAA, UGA, or UAG) appropriately positioned at the end of the polypeptide to be translated. The plasmids may include at least one selectable marker, for example, dihydrofolate reductase, G418, or a marker of neomycin resistance for eukaryotic cell culture; and tetracycline, kanamycin, or ampicillin resistance genes for culturing in E. coli and other prokaryotic cell culture. The minicircle producing plasmids may include at least one origin of replication to allow for the multiplication of the vector in a suitable eukaryotic or a prokaryotic host cell. Origins of replication are known in the art, as described, for example, in Genes H, Lewin, B., ed., John Wiley & Sons, New York (1985). The FGFR fusion proteins produced from minicircles can also be fused to marker sequences, as described above. Fusion Partners and Conjugates [0142] Gene manipulation techniques have enabled the development and use of recombinant therapeutic proteins with fusion partners that impart desirable pharmacokinetic properties. FGFR polypeptides, including their immunogenic epitopes and other fragments, can be combined with heterologous molecules, resulting in therapeutically useful fusion molecules. The invention provides fusion molecules comprising the extracellular domain of any of FGFR1, FGFR2, FGFR3, and FGFR4. It provides fusion partners capable of imparting favorable pharmacokinetics and/or pharmacodynamics to the FGFR. In an embodiment, the invention provides a fusion molecule comprising all or a part of the extracellular domain of FGFR1-IlIb, FGFR -Mc, 49 FGFR2-IIrb, FRFR2-lIIc, FGFR3-IIIb, FGFR3-fIc, FGFR4, or fragments thereof and a fusion partner, such as an antibody Fc domain. [0143] FGFRs are expressed in many normal tissues and many cell types express more than one FGFR. In view of this, it is not obvious how a therapeutic which targets FGFR can be devised to last long enough in the circulation of a treated subject without causing harm to the normal tissues. [0144] Fusion molecules of the invention have an increased half-life in vivo, as compared to FGFR extracelular domains. The prolonged half-life of the FGFR fusion molecules described herein can require lower doses and a less-frequent dosing regimen than FGFRs alone. The resulting decreased fluctuation of FGFR serum levels can improve the safety and tolerability of FGFR therapeutics. [0145] The fusion partner can be linked to the C- terminus of the FGFR, or, alternatively, the FGFR can be linked to the C- terminus of the fusion partner. The fusion partner may comprise a linker, for example, a peptide linker, which may or may not comprise an enzyme cleavage site. The fusion partner may also comprise a molecule that extends the in vivo half-life by imparting improved receptor binding to FGFR within an acidic intracellular compartment, for example, an acid endosome or a lysosome. [0146] Fusion partners of the invention include polymers, polypeptides, lipophilic moieties, and succinyl groups. Examples of polypeptide fusion partners include serum albumin and the antibody Fc domain. Polymer fusion partners may comprise one or more polyethylene glycol moieties, branched or linear chains. Lipophilic fusion partners may increase the skin permeability of the fusion molecule. Oligomerization Domain Fusion Proteins [0147] Oligomerization offers functional advantages to a fusion protein, including multivalency, increased binding strength, and the combined function of different domains. These features are seen in natural proteins and may also be introduced by protein engineering. Accordingly, the invention provides an FGFR fusion molecule, wherein the fusion partner comprises an oligomerization domain, for example, a dimerization domain. Suitable oligomerization domains include coiled-coil domains, including alpha-helical coiled-coil domains; collagen domains; collagen-like domains, and dimeric immunoglobulin domains. Suitable coiled-coil polypeptide fusion partners of the invention include tetranectin coiled-coil domain, the coiled-coil domain of cartilage oligomeric matrix protein; angiopoietin coiled-coil domains; and leucine zipper domains. FGFR fusion molecules with collagen or collagen-like oligomerization domains as fusion 50 partner may comprise, for example, those found in collagens, mannose binding lectin, lung surfactant proteins A and D, adiponectin, ficolin, conglutinin, macrophage scavenger receptor, and emilin. Antibody Fc Domain Fusion Proteins (01481 In an embodiment, the invention provides fusion molecules having an Fc immunoglobulin domain. The FGFR fusion proteins of the invention can comprise Fc, various domains of the constant regions of the heavy or light chains of mammalian immunoglobulins and/or the first two domains of the human CD4 polypeptide. [0149] In an embodiment, the human Fc domain fusion partner comprises the entire Fe domain. In an embodiment it comprises one or more fragments of the Fc domain. For example, it may comprise a hinge and the CH2 and CH3 constant domains of a human IgG, for example, human IgGI, IgG2, or IgG4. The invention provides an FGFR fusion protein wherein the fusion partner is a variant Fe polypeptide or a fragment of a variant Fc polypeptide. The variant Fc may comprise a hinge, CH2, and CH3 domains of human IgG2 with a P33 IS mutation, as described in U.S. Patent No. 6,900,292. [0150] In an embodiment, a fusion protein of the invention may be a homodimeric protein linked through cysteine residues in the hinge region of IgG Fc, resulting in a molecule similar to an IgG molecule, but without CHI domains and light chains. [0151] Methods of making fusion proteins are well-known to the skilled artisan. In an embodiment, the Fe fusion partner of the invention comprises an amino acid sequence of any of SEQ ID NO: 171, SEQ ID NO: 172, and SEQ ID NO: 173. Albunin Fusion Proteins [0152] The invention provides an FGFR fusion molecule with an albumin fusion partner comprising albumin from human serum (human serum albumin or "HSA"), one or more fragments of albumin, a peptide that binds albumin, and/or a molecule that conjugates with a lipid or other molecule that binds albumin. In an embodiment, an FGFR-HSA fusion molecule may be prepared as described herein and as further described in U.S. Patent No. 6,686,179 with respect to an interferon alpha-HSA fusion molecule. Dimeric FGFR Fusion Proteins [0153] The invention provides FGFR fusion proteins, wherein the fusion partner comprises an FGFR extracellular domain or active fragment thereof. For example, the fusion protein may comprise two FGFRI, FGFR2, FGFR3, or FGFR4 extracellular domains or biologically active fragments thereof. The fusion molecule may also comprise 51 heterologous combinations of two different FGFR extracellular domains or biologically active fragments thereof, as described in greater detail above. [0154] In an embodiment, the FGFR fusion protein comprises an extracellular domain of an FGFR and/or one of its active fragments and further comprises a fusion partner comprising a dimerization domain as well as an FGFR extracellular domain. When the fusion partner comprises a dimerization domain, such as an Fc domain or an active fragment thereof, the FGFR fusion protein expressed in a mammalian cell expression system may naturally form a dimer during the production process. Fusion Proteins with Pegylated Moieties [0155] In addition to the recombinant molecules described above, the invention provides an FGFR fusion molecule, wherein the fusion partner comprises a polymer, such as a polyethylene glycol (PEG) moiety. PEG moieties of the invention may be branched or linear chain polymers. In an embodiment, the present invention contemplates a chemically derivatized polypeptide which includes mono- or poly- (e.g., 2-4) PEG moieties. Pegylation may be carried out by any of the pegylation reactions known in the art. Methods for preparing a pegylated protein product are generally known in the art. Optimal reaction conditions will be determined on a case by case basis, depending on known parameters and the desired result. [0156] There are a number of PEG attachment methods available to those skilled in the art, for example, EP 0 401 384; Malik et al., Exp. Hematol., 20:1028-1035 (1992); Francis, Focus on Growth Factors, 3:4-10 (1992); EP 0 154 316; EP 0 401 384; WO 92/16221; WO 95/34326; and the other publications cited herein that relate to pegylation. [0157] Pegylation may be performed via an acylation reaction or an alkylation reaction with a reactive polyethylene glycol molecule. Thus, protein products of the present invention include pegylated proteins wherein the PEG groups are attached via acyl or alkyl groups. Such products may be mono-pegylated or poly-pegylated (for example, those containing 2-6 or 2-5 PEG groups). The PEG groups are generally attached to the protein at the a- or &-amino groups of amino acids, but it is also contemplated that the PEG groups could be attached to any amino group attached to the protein that is sufficiently reactive to become attached to a PEG group under suitable reaction conditions. [0158] Pegylation by acylation generally involves reacting an active ester derivative of PEG with a polypeptide of the invention. For acylation reactions, the polymer(s) selected typically have a single reactive ester group. Any known or subsequently discovered reactive PEG molecule may be used to carry out the pegylation 52 reaction. An example of a suitable activated PEG ester is PEG esterified to N hydroxysuccinimide (NHS). As used herein, acylation is contemplated to include, without limitation, the following types of linkages between the therapeutic protein and a polymer such as PEG: amide, carbanate, urethane, and the like, see for example, Chamow, Bioconjugate Chem., 5:133-140 (1994). Reaction conditions maybe selected fiom any of those known in the pegylation art or those subsequently developed, but should avoid conditions such as temperature, solvent, and pH that would inactivate the polypeptide to be modified. [0159] Pegylation by acylation will generally result in a poly-pegylated protein. The connecting linkage may be an amide. The resulting product may be substantially only (e.g., >95%) mono, di- or tri-pegylated. However, some species with higher degrees of pegylation may be formed in amounts which depend on the specific reaction conditions used. If desired, more purified pegylated species may be separated from the mixture (particularly unreacted species) by standard purification techniques, including among others, dialysis, salting-out, ultrafiltration, ion-exchange chromatography, g'cl filtration chromatography, and electrophoresis. ' [0160] Pegylation by alkylation generally involves reacting a terminal aldehyde derivative of PEG with a polypeptide in the presence of a reducing agent. For the reductive alkylation reaction, the polymer(s) selected should have a single reactive aldchyde group. An exemplary reactive PEG aldehyde is polyethylene glycol propionaldchyde, which is water stable, or mono Cl-C1O alkoxy or aryloxy derivatives thereof, see for example, U.S. Patent No. 5,252,714. Variant and Mutant Polypeptides [0161] The FGFR fusion proteins of the invention can be made by protein engineering to improve or alter the native characteristics of FGFR fusion proteins. Recombinant DNA technology known to those skilled in the art can be used to create novel mutant proteins or "muteins," including single or multiple amino acid substitutions, deletions, and additions. Such modified polypeptides can possess properties desirable in a therapeutic agent, such as enhanced activity or increased stability. In addition, they may be purified in higher yields and be more water-soluble than the corresponding natural polypeptide, at least under certain purification and storage conditions. FGFR ECD Mutants [0162] As mentioned above, the invention provides polypeptide fusion molecules having one or more residues deleted from the amino and/or carboxyl terminus of the 53 amino acid sequences of the FGFR extracellular domains. For example, the invention provides deletion mutations of FGFR extracellular domains with deletions in the C terminal region. The invention provides deletion mutants missing one or more amino acids in the region N-terminally adjacent to the transmembrane domain. [0163] In an embodiment, the invention provides FGFR fusion proteins comprising variants of the ECD of wildtype FGFR polypeptides, where the variants have deletions or point mutations in the C-terminus of the ECD, for example, in the region of the MMP-2 cleavage site. These variants may be more resistant to cleavage by MMP-2 than wildtype FGFR extracellular domains. In an embodiment, the invention provides FGFR extracellular domains which are deletion mutants that have the MMP-2 cleavage site removed and which are more resistant to cleavage by MMP-2- than wildtype FGFR extracellular domains. For example, the invention provides deletion mutants of the FGFR extracellular domains in that region at the C-terminus of the IgIlI domain and N-terminal to the Fe domain. Any one or more amino acids of any of the seven FGFR extracellular domains in this region may be deleted or, otherwise mutated. By way of example, the invention provides deletion mutants of FGFRI, FGFR2, FGFR3, and FGFR4 corresponding to R1Mut1, RlMut2, RlMut3, and RlMut4, shown in FIG. 1A; RlMut7, as shown in FIG. 1B; and R4Mutl, R4Mut2, R4Mut3, R4Mut4, and R4Mut5, as shown in FIG. 2. These variants, as well as their parental, unmutated polypeptides, all bind at least one FGF ligand. The ligand-binding characteristics of these mutants, as well as their parent FGFRs, can be determined by binding assays known in the art. [0164] In an embodiment, the invention provides an FGFR fusion molecule comprising a first molecule that comprises one or more soluble extracellular domain of an FGFR and/or a biologically active fragment thereof and a second molecule, wherein the second molecule confers an extended half-life to the first molecule in an animal, wherein the second molecule is other than a naturally occurring Fc polypeptide, and wherein the fusion molecule is a variant Fc polypeptide. [0165] When FGFR1-IIlc-Fc fusion protein was produced in host cells, injected into animals, and examined by gel electrophoresis, it was observed that FGFRI-mc-Fc was partially cleaved both in vivo and in vitro. Degraded fragments in the cell culture media and serum samples were consistent with the size of fragments predicted following cleavage of the Fc fusion partner from the fusion protein. MMP-2 added exogenously to FGFRl-IIIc-Fc in vitro reproduced the degraded fragments observed in the serum and culture medium.

54 [0166] Accordingly, the invention provides recombinant FGFR fusion proteins resistant to cleavage and with improved pharmacokinetic profiles compared to the naturally-occurring forms. For example, the invention provides FGFR fusion proteins that are more resistant to degradation both in vitro and in vivo. These FGFR fusion protein constructs may have single amino acid changes in the cleavage sites of serum proteases. They may also have global deletions of the cleavage sites. These constructs can be made by altering the naturally-occurring sequences using methods known to those of skill in the art. The invention also provides FGFR constructs wherein the junction between the extracellular domains and transmembrane domains of the receptor are modified to remove proteolytic degradation sites by methods known to those of skill in the art. [0167] In addition to the deletion mutants described above, the invention provides single point mutants, such as substitution mutants resistant to MMiP-2 cleavage. Natural substrates of MPI-2 have a preponderance of proline at the third residue N-terminal to the MMP-2 cleavage site, but have not been described to have either methionine or glycine residues at this site. Accordingly, the invention provides for point mutants corresponding to RIMut8 (P364M), RlMut9 (M367N), and RIMut1O (P364G), such as shown in FIG. 1B, for example. P364M is expected to have a similar hydrophobicity profile and P364G is expected to have greater flexibility than wildtype. Also by way of example, the invention provides the M367N substitution mutation of the methionine of the MMP-2 cleavage site. Asparagine (N) has not been described in any natural or synthetic MMP-2 substrate, thus its substitution can be expected to attenuate or prevent cleavage by MMP-2. The M367N mutation also introduces a potential glycosylation site, namely, NTS into the FGFRI BCD. If this glycosylation site is utilized by the host cell in vitro. or in vivo, the N linked sugars could further shield the FGFRI-IIb or FGFR1-IIfc extracellular domain from proteolysis. The invention further provides the double variant of FGFRl-nib or FGFRI-Ill, P364G/M367N. Any of the mutants of the invention may, optionally, comprise a linker sequence. [0168) In addition to the deletion and substitution mutants of the FGFR fusion molecules described above, the invention provides insertion, inversion, and repeat mutants. Some amino acid sequences of FGFR fusion molecules can be varied without significant effect on the structure or function of the protein while others are critical for determining activity. Accordingly, the invention includes variations of the FGFR fusion proteins which show substantial FGFR polypeptide activity and/or which include regions of the FGFR extracellular domains. Mutants of the invention may have the receptor 55 activity of wild-type FGFR or may have activities enhanced or reduced, or broadened with respect to FGF ligand binding capability, as compared to wildtype. Methods of making these mutants are generally known in the art. [01691 Variations of the FGFR fusion proteins of the invention can be made and are included herein. Guidance concerning how to make phenotypically silent amino acid substitutions is provided by Bowie et al., Science 247:1306-1310 (1990). Genetic engineering techniques can be used to introduce amino acid changes at specific positions of the FGFR fusion protein and selections, or screens, can be used to identify sequences that maintain functionality. [0170] For example, the art cited herein can be followed to produce proteins tolerant of amino acid substitutions. This art indicates which amino acid changes are likely to be permissive at a certain position of a protein. For example, most buried amino acid residues require nonpolar side chains, whereas few features of surface side chains are generally conserved. Typically conservative substitutions of the FGFR fusion proteins are tolerated, such as replacements, one for another, among the aliphatic amino acids Ala, Val, Leu, and le; interchange of the hydroxyl residues Ser and Thr, exchange of the acidic residues Asp and Glu, substitution between the amide residues Asn and Gin, exchange of the basic residues Lys and Arg, and replacements between the aromatic residues Phe and Tyr. Substitutions of charged amino acids with other charged or neutral amino acids may produce proteins with desirable improved characteristics, such as less aggregation. Aggregation may not only reduce activity but also be problematic when preparing pharmaceutical formulations, because, for example, aggregates can be immunogenic (Pinckard et al., Clin. Exp. Imnmunol. 2:331-340 (1967); Robbins et al., Diabetes 36:838 845 (1987); Cleland ct al., Crit. Rev. Therapeutic Drug Carier Systems, 10:307-377 (1993)). As described above, the binding of FGF ligands to FGFRs is both selective and overlapping. Selected ligands bind to a particular FGFR, but more than one ligand can bind to a receptor and an FGF ligand may bind to multiple FGFR. Mutating amino acids in the FGFR fusion proteins of the invention can change the selectivity of FGF ligand binding to FGFRs. Transgenic, Knockout, and Other Animals [0171) The invention provides transgenic and knockout animals, respectively expressing exogenous FGFR and lacking endogenous FGFR, as well as animals injected with the FGFR fusion proteins of the invention or the nucleic acid molecules which encode them. Transgenic animals of the invention are generally made by expressing an 56 FGFR fusion molecule as described herein with a vector comprising an exogenous promoter and an FGFR extracellular domain and targeting the vector to a predetermined locus, wherein the expression pattern of the FGFR fusion transgene is determined by the expression pattern of the exogenous promoter. Knockout mice are generally made by selectively inactivating endogenous FGFR and replacing it with a mutant allele. [0172] Nonhuman animals of any species, including, but not limited to, mice, rats, rabbits, hamsters, guinea pigs, pigs, micro-pigs, goats, sheep, cows, and non-human primates, for example, baboons, monkeys, and chimpanzees, may be used to generate transgenic and knockout animals. In a specific embodiment, techniques described herein or otherwise known in the art, are used to express FGFR fusion molecules of the invention in humans, as part of a gene therapy protocol. [0173] Any technique known in the art may be used to introduce an FGFR transgene into animals to produce a founder lines of transgenic animals. Known techniques may also be used to "knock out' endogenous FGFR genes. Techniques for producing transgenic and knockout mice include, but are not limited to, pronuclear microinjection (Paterson et al., Appl. Microbiol. Biotechnol. 40:691-698 (1994); Carver et al., Biotechnology (NY) 11:1263-1270 (1993); Wright et al., Biotechnology (NY) 9:830 834 (1991); and Hoppe et al., U.S. Patent No. 4,873,191 (1989)); retrovirus mediated gene transfer into germ lines, blastocysts, or embryos (Van der Putten et al., Proc. Natt. Acad. Sci. 82:6148-6152 (1985)); gene targeting in embryonic stem cells (Thompson et al., Cell 56:313-321 (1989)); electroporation of cells or embryos (Lo, MoL. Cell. Biol. 3:1803-1814 (1983)); introduction of the polynucleotides of the invention using a gene gun (Ulmer et al., Science 259:1745 (1993)); introducing nucleic acid constructs into embryonic pluripotent stem cells and transferring the stem cells back into the blastocyst; and sperm mediated gene transfer (Lavitrano et al., Cell 57:717-723 (1989). For a review of such techniques, see Gordon, IntL. Rev. Cytol. 115:171-229 (1989); U.S. Patent No. 5,464,764 (Capecchi et al.); U.S. Patent No. 5,631,153 (Capecchi ct al.); U.S. Patent No. 4,736,866 (Leder et al.); and U.S. Patent No. 4,873,191 (Wagner et al.). Any technique known in the art may be used to produce transgenic clones containing polynucleotides of the invention, for example, nuclear transfer into enucleated oocytes of nuclei from cultured embryonic, fetal, or adult cells induced to quiescence (Campbell et al., Nature 380:64-66 (1996); Wilmut et al., Nature 385:810-813 (1997)). [0174] Gene targeting can be used to integrate the FGFR transgene into the chromosomal site of an endogenous gene. Briefly, vectors comprising nucleotide 57 sequences homologous to the endogenous gene are designed for the purpose of integrating, via homologous recombination with chromosomal sequences, into and disrupting the function of the nucleotide sequence of the endogenous gene. The FGFR transgene may also be selectively introduced into a particular cell type, thus inactivating the endogenous FGFR gene in only that cell type, by following, for example, the teaching of Gu et al., Science 265:103-106 (1994). The regulatory sequences required for such a cell-type specific inactivation will depend upon the particular cell type of interest, and will be apparent to those of skill in the art. [0175] The expression of the recombinant FGFR transgene or knockout allele may be assayed in the animals of the invention using standard techniques. Initial screening may be accomplished by Southern blot analysis or PCR techniques to verify that integration of the transgene or null allele has taken place. The level of mRNA expression of the transgene or null allele in animal tissues may be assessed using techniques which include, but are not limited to, Northern blot analysis, in situ hybridization analysis, and reverse transcriptase-PCR (RT-PCR). Samples of FGFR transgene-expressing or FGFR null tissue may also be evaluated immunocytochemically or immunohistochemically using specific antibodies. [0176] The invention provides transgenic animals which carry an FGFR transgene in all their cells, as well as animals which carry the transgene in some of their cells, such as mosaic or chirmeric animals. The transgene may be integrated as a single transgene or as multiple copies, such as in concatamers, for example, head-to-bead tandems or head-to tail tandems. An FGFR transgene may also be selectively introduced into and activated in a particular cell type by following, for example, the teaching of Lasko et al., Proc. Nati. Acad Sci. 89:6232-6236 (1992)). The regulatory sequences required for such a cell-type specific activation will depend upon the particular cell type of interest, and will be apparent to those of skill in the art. [0177] Founder transgenic animals may be bred, inbred, outbred, or crossbred to produce colonies of a particular animal. Examples of such breeding strategies include, but are not limited to, outbreeding of founder animals with more than one integration site in order to establish separate lines; inbreeding of separate lines in order to produce compound transgenics that express the transgene at higher levels because of the effects of additive expression of each transgene; crossing heterozygous transgenic animals to produce animals homozygous for a given integration site in order to both augment expression and eliminate the need for screening of animals by DNA analysis; crossing 58 separate homozygous lines to produce compound heterozygous or homozygous lines; and breeding to place the transgene on a distinct background appropriate to an experimental model of interest. [0178] In an embodiment, the FGFR fusion proteins of the invention are expressed in transgenic non-human animals produced by the method described in WO 03/020743. In this method, a cassette including a transgene of interest is targeted to one or more predetermined loci, including loci expressed in most or all cell types. The cassette can function as an autonomous unit, directing the expression of the transgene and-optional regulatory or accessory gcnes in the cassette. The transgene is expressed under the control of the exogenous promoter within the cassette. Thus, the expression pattern of the transgene is detennined by the nature of the exogenous promoter. [0179] Transgenic and knock-out animals of the invention have uses which include, but are not limited to, animal model systems useful in elaborating the biological function of FGFRs, studying conditions and/or disorders associated with aberrant expression of FGFRs, and in screening for compounds effective in modifyring or ameliorating these conditions and/or disorders. [0180] Animals comprising the nucleic acid molecules or the FGFR fusion proteins of the invention are those that have been injected with either the FGFR fusion protein or the nucleic acid construct, such as by hydrodynamic tail vein transfection method, and such as using the mini-circle vector construct previously described. FGFR Fusion Proteins as Decoy Receptor Traps [018.1] The FGFR fusion proteins of the invention can function as decoy receptors for trapping FGF ligands and inhibiting their interaction with FGFR on cell surfaces. Decoy receptors, such as those of the invention, recognize their ligands with high affinity and specificity but are structurally incapable, of signaling. They compete with wild-type receptors for ligand binding and participate in ligand/receptor interactions, thus modulating the activity of or the number of functioning receptors and/or the cellular activity downstream from the receptors. Decoy receptors can act as molecular traps for agonist ligands and thereby inhibit ligand-induced receptor activation. [0182] Prior to the teachings of the present invention, it was not known whether tumors or proliferative cells are dependent on FGF growth factors in vivo or which FGF ligand may be blocked in order to inhibit tumor progression or cell proliferation. Furthermore, it has not been previously reported that the ability to block FGF-induced proliferation correlates with lowered levels of FGFR activation.

59 [0183] The FGFR fusion proteins of the invention can be used in combination with other decoy receptor traps. Etanercept (Enbrel*) is an example of a genetically engineered decoy receptor trap comprising a fusion protein of the extracellular ligand-binding domain of the human TNF-a receptor and the Fe region of human IgG1. It acts as a decoy by competitively inhibiting TNF-a binding to naturally-occurring TNF-a receptors on the cell-surface, thus inhibiting TNF-t induced proinflammatory activity. Etanercept acts as a cytokine "sponge" and TNF-ct antagonist, rendering TNF-ca biologically inactive (Goldenberg et al., Clin. Ther. 21:75-87 (1999)). It is used to treat rheumatoid arthritis, juvenile rheumatoid arthritis, psoriatic arthritis, and ankylosing spondylitis. [0184] The FGFR fusion proteins of the invention can also be used with a VEGF Trap, which is another example of a soluble recombinant decoy fusion protein, and is currently in clinical trials. A genetically engineered fusion protein of one or more extracellular ligand-binding domain of the human vascular endothelial growth factor (VEGF) receptor and the Fe region of human IgG1, the VEGF Trap inhibits angiogenesis by acting as a decoy for naturally-occurring cell surface VEGF receptors. By inhibiting angiogenesis, VEGF decoys can shrink tumors which rely on angiogenesis for their viability. The biological activity of the decoy trap depends on the portion of the receptor used in the trap. For example, a fusion protein of the first three Ig domains of the VEGFR1 receptor isoform and the Fc region of human IgGI binds to VEGF with an affinity in the picomolar range and has potent anti-tumorigenic activity but a short in vivo half life and significant toxicity (Holash et al., Proc. Natl. A cad. Sci. 99:11,393-11,398 (2002)). VEGF decoy fusion proteins can be engineered to prolong the in vivo pharmacokinetic and pharmacodynamic profiles, minimize toxicities, and potently inhibit growth and vascularization. Removing a highly basic ten amino acid sequence from the third VEGF1 Ig domain, removing the entire first VEGF1 Ig domain, and fusing the second Ig domain ofVEGFRI with the third Ig domain of VEGFR2 have been reported to improve the clinical parameters (Holash et al., Proc. NatL. Acad. Sci. 99:11,393-11,398 (2002)). The combination of an FGFR fusion protein and the VEGF Trap can be more potent than either alone in inhibiting angiogenesis. [0185] The invention provides FGFR decoy receptor trap fusion proteins and demonstrates that they inhibit the binding of ligands to FGFRs, as shown in greater detail below. The decoy fusion proteins sequester the ligands, preventing ligand-receptor binding. The ability of the FGFR receptor trap fusion proteins to inhibit receptor-ligand 60 binding can be demonstrated using assays known in the art, for example, competition ELISA assays, as described in more detail below. [0186] The invention provides FGFR decoy receptor traps as therapeutic agents. The FGFR decoy receptor traps of the invention bind to various FGFs, described in more detail herein, which have been demonstrated to be over-expressed in proliferative disease states, compared to normal. These traps can bind FGF ligand with very high affinity, for example, they may bind FGF-2 with a Kd of approximately 15 picomolar. Furthermore, these traps can interfere with FGF signaling in abnormal tissues. The FGFRl-lIc-Fc and FGFR4-Fc traps of the invention, for example, can dampen the signaling of FGFR1-Hmc and FGFR4, and perhaps other members of the FGFR family (Zhang et al., J Biol. Cien. 281:15, 694-15,700 (2006)). [0187] Introducing gene trap vectors into embryonic stem cells has produced transgenic animal lines that reflect the gene expression patterns of receptor domains of interest (Coffin et al., Retroviruses Cold Spring Harbor Lab. Press (1997)). Accordingly, the invention provides the use of FGFR gene trap vectors to identify discrete expression patterns of FGFR genes during signal transduction events associated with normal and disease states. Constructs with a reporter gene but lacking a promoter are designed so that activation of the reporter gene depends on its insertion within an active transcription unit. Integration results in an expression pattern that reflects the pattern of the endogenous transcription unit. The reporter gene provides a molecular tag for cloning the "trapped" gene of the transcription unit. Reporter systems which can be used with gene trap vectors are known in the art. Following insertion, the tagged gene can be detected in space and time by assaying for the reporter gene product. Real-Time Detection of Ligand-Receptor Binding [0188) As described herein, FGFs are over-expressed in certain disease states. By acting as decoy receptor traps, FGFR fusion proteins attenuate the biological activities of the over-expressed FGFs. The profile of FGFs which bind to a particular FGFR fusion protein in vitro can predict the therapeutic profile of that fusion protein in vivo. Accordingly, the invention provides ligand-binding profiles for FGFR fusion proteins of the invention and methods of using FGFR fusion molecules of the invention to treat diseases that over-express FGF ligands. [0189] The invention provides direct ligand-receptor binding measurements by surface plasmon resonance (SPR) using Biacore technology (Biacore; Piscataway, NJ), which utilizes biosensor chips to measure binding interactions in real time (Dawson et al., 61 Molec. Cell. Biol. 25:7734-7742 (2005)). The technology is based on SPR optical phenomena and detects changes in refractive index that occur close to a sensor chip's surface. One of the interacting components is immobilized on a flexible dextran layer linked to the sensor chip surface, and an interacting partner flows in solution across the surface. Interaction between the two components immobilizes the interacting partner, increasing the mass at the sensor chip surface. The increased mass results in an optical signal, which is recorded in resonance units (RU). One RU represents approximately one picogram of protein bound to the surface. Biacore technology has been described in U.S. Patent Nos. 6,999,175 B2; 6,808,938 B2; and 5,641,640. [0190] As described in more detail below, FGF ligand binding to fusion proteins of the invention was measured using the Biacore* X system to measure surface plasmon resonance (SPR). This method provided a ranking of the relative affinities of FGF ligands for the FGFR fusion proteins of the invention, for example, FGFR1-IIlc-Fc, R1Mut4, FGFR3-MIc-Fc, and FGFR4-Fc. Accordingly, the invention provides a method of treating a disease characterized by one or more FGFs which are expressed at a higher level than normal by administering a binding FGFR fusion protein of the invention. Biomarkers of FGFR Fusion Molecule Treatment (0191] Biomarkers can be used to monitor the results of treating subjects with FGFR fusion proteins, including demonstrating efficacy as an end point in clinical trials. Suitable biomarkers will indicate that an FGF-FGFR signaling pathway is affected by the FGFR fusion protein. For example, FGF-2 is a suitable biomarker for FGFR1 because a decrease in FGF-2 levels in a subject treated with an FGFRI fusion protein would indicate that the fusion protein bound FGF-2, sequestered it from active FGFRs, and thus demonstrated treatment efficacy. [0192] Components of the FGFR signaling pathway may also serve as biomarkers to demonstrate treatment efficacy. For example, FGFRs produce intracellular responses to extracellular ligand binding by intracellular signaling. FGF binding to the extracellular domain of the intact transmembrane receptor activates-the catalytic tyrosine kinase domain present on the cytoplasmic portion of the receptor. The ligand induces the FGFR to autophosphorylate a cytoplasmic tyrosine residue, which then serves as part of a high affinity binding site for intracellular signaling proteins. One group of these signaling proteins, the extracellular signal regulated kinases (Brks), also known as mitogen-activated protein (MAP) kinases, become activated when FGFR phosphorylates a threonine and a nearby tyrosine on the Erk protein. Reportedly, ligand binding to cell surface FGFRs 62 initiate a signal transduction cascade that includes the phosphorylation of Erk to phospho Erk (pErk). Erk activation therefore provides a biomarker for FGFR fusion molecule treatment by providing a measurement of FGFR intracellular signaling activity, which can be quantified by measuring the phosphorylation of the threonine and tyrosine residues. Erk activation can be determined by methods known in the art and demonstrated in more detail below. Commercial reagents are available that detect Erk immunologically from cell lysates. ELISA and/or Western blot analyses can be performed using these reagents to identify and measure phosphorylated Erk by methods well-known in the art. [0193] Other biomarkers may also be used to monitor FGFR fusion protein treatment by providing a measurement of the FGFR signaling pathway. For example, a reduction in phosphorylated FGFR (pFGFR), or a reduction of the basal phosphorylation level of fibroblast growth factor receptor substrate 2 (pFRS2) and/or dual specificity phosphatase (DSP) would indicate efficacious treatment with FGFR fusion proteins. FGFR Fusion Proteins Inhibit the Viability and/or Proliferation of Proliferative Cells [0194] Proliferative cells often depend on extracellular signaling by growth factors for their survival and growth. The FGFR fusion proteins of the invention can inhibit the viability and/or the proliferation of cancer cells and other proliferative cells both in vitro and in vivo. Accordingly, the invention provides methods of inhibiting viability and/or proliferation of proliferative cells, methods of inhibiting angiogenesis, and methods of treating cancer in a subject by providing an FGFR fusion protein, as described herein, and administering the fusion protein to the subject. The effect of FGFR fusion proteins on cell viability and/or proliferation in vitro was examined on cultured tumor cells and their effects on tumor cell viability and proliferation in vivo was examined in an animal tumor model. [0195] The CellTiter-GloTM Luminescent Cell Viability Assay (Promega; Madison, WI), which was designed to measure the number of viable cells in culture (Sussman et at., Drug Disc. Dev. 5:71-71 (2002), was used herein to determine cell viability and proliferation. Cellular adenosine triphosphate (ATP) levels indicate cell viability; ATP levels drop rapidly when the cell loses viability. The assay uses a stable form of firefly luciferase to measure ATP as an indicator of metabolically active, i.e., viable, cells. The luciferase converts beetle luciferin to-luciferin oxide in the presence of ATP, magnesium, and oxygen. The resulting luminescent signal is proportional to the number of viable cells present in the culture and can be detected with a luminometer or CCD camera. Because the signal is proportional to cell number, it measures both viability 63 and proliferation. Under stated media and serum conditions, the assay is linear over a wide range of cell numbers. [0196] The invention provides methods of using the FGFR fusion proteins of the invention to inhibit the viability and/or proliferation of multiple proliferative cell types, whether dysplastic cells, premalignant cells or malignant tumor cells; methods to inhibit the viability and/or proliferation of other cell types, such as endothelial cells; and methods to inhibit angiogenesis, in vitro, ex vivo, or in vivo. As described in more detail below, FGFR fusion proteins of the invention can be used to inhibit a wide variety of cancer cell types, including lung, kidney, brain, breast, liver, ovarian, prostate, and/or colorectal cancer cells, for example. The FGFR fusion proteins of the invention each have different specificities to different cancer cell types, as shown in greater detail below, and can be used to treat different tumor types depending on such specificities. [0197] For example, the FGFR fusion proteins of the invention, such as FGFR1 'HIc-Fc fusion proteins, can be used to inhibit the viability and/or proliferation of malignant human glioma cells (for example, U251 cells); malignant human brain cancer cells (for example, SF268 cells); human lung cancer cells (for example, A549 cells); malignant lung non-squamous carcinoma cells (for example, NCI-H522 and NCI-H226 cells); malignant glioblastoma cells (for example, UI 18 and WTI Il cells); and malignant kidney cells (for example, Caki-1 cells). [0198] The invention also provides methods for using FGFR fusion proteins of the invention to inhibit the proliferation of tumor cells and other proliferative cells, such as endothelial cells, in vivo. This in vivo activity can be demonstrated by administering the FGFR fusion protein of the invention to inhibit in vivo formation of tumors in animal xenograft models. As shown herein, FGFRl-lIc-Fc effectively inhibited tumor growth in this model. Accordingly, the invention provides a method of inhibiting tumor growth and tumor cell proliferation in a subject by providing a composition comprising an FGFR fusion protein of the invention and administering the composition to the subject. [0199] The invention provides methods of inhibiting viability and/or proliferation of proliferative cells, such as endothelial cells, under conditions in which the proliferation of such cells is not desirable. For example, macular degeneration and tumor angiogenesis are conditions under which excess growth of blood vessels, thus endothelial cells, is undesirable. Accordingly, the invention further provides methods of inhibiting angiogenesis by administering the FGFR fusion proteins of the invention to a subject in need of such treatment. As described in greater detail below, dosing schedules and dosing 64 routes are generally known in the art; the latter may include intravenous, subcutaneous, intraperitoneal, and oral administration. FGF and FGFR Expression in Human Cancers [0200] Gene amplification is among the mechanisms of oncogene activation that can lead to specific types of cancers. The invention provides an analysis of the gene expression profile of breast cancer tissues residing in the proprietary GeneLogic database, and the finding that the FGFRI gene was amplified in 10-15% of breast cancer patients. As described herein, such gene amplification has implications for tumor cell growth and/or survival, thus interrupting signaling between an FGFR and an FGF ligand is a useful approach to inhibiting tumor growth. [0201] The invention also provides further analysis of the expression profllcs of different tumor types resident in the proprietary GeneLogic database for FGF and FGFR expression, and the finding that certain tumors expressed a higher level of FGFRI, FGFR3, and/or FGFR4. The invention further provides that certain tumors expressed a higher level of certain FGF ligands, implicating active FGF/FGFR signaling pathways in maintaining the viability and/or proliferative capacity of the cancer cells or the endothelial cells feeding the cancer cells. Information relating to the tumor types that over-express an FGFR and/or an FGF is provided in the tables below. [0202] Accordingly, the invention provides methods and compositions for FGFR fusion proteins which are suitable for use in treating proliferative diseases characterized by over-expression of an FGFR, FGF, or both. The analysis performed herein and described in greater detail in the Examples, provides the FGFR and FGF gene expression profiles of various hyperproliferative tissues. Thus, the over-expression of the FGFRs and their ligands are correlated with particular disease states. The FGFR fusion proteins of the invention are efficacious in treating the diseases in which the FGFR component, or its ligand, is over-expressed. Therapeutic Compositions and Formulations Routes of Administration and Carriers [0203] The FGFR fusion molecules of the invention can be administered in vivo by a variety of routes, including intravenous, intra-arterial, subcutaneous, parenteral, intranasal, intramuscular, intracardiac, intraventricular, intratracheal, buccal, rectal, intraperitoneal, intradermal, topical, transdermal, and intrathecal, or otherwise by implantation or inhalation. They may be administered in formulations, as described in more detail below. They may be administered in powder form intranasally or by 65 inhalation. They may be administered as suppositories, for example, as formulated by mixing with a variety of bases, such as emulsifying bases, water-soluble bases, cocoa butter, carbowaxes, and polyethylene glycols; which melt at body temperature, yet are solidified at room temperature. Jet injection can be used for intramuscular or intradermal administration (Furth et al., Anal. Biochem. 205:365-368 (1992)). The DNA can be coated onto gold microparticles and delivered intradermally by a particle bombardment device, or "gene gun" as described in the literature (Tang et al., Nature 356:152-154 (1992)), where gold microprojectiles are coated with the DNA, then bombarded into skin cells. These methods of in viva administration are known in the art. [0204] In some embodiments, fusion molecule compositions are provided in formulation with pharmaceutically acceptable carriers, a wide variety of which are known in the art (Gennaro, Remington: The Science and Practice of Pharnacy with Facts and Comparisons: Drug/acts Plus, 20th ed. (2003); Ansel et al., Pharmaceutical Dosage Forms and Drug Deliveiy Systems, 7th ed., Lippencott Williams and Wilkins (2004); Kibbe et al., Handbook of Pharmaceutical Excipients, 3d ed., Pharmaceutical Press (2000)). Pharmaceutically acceptable carriers, such as vehicles, adjuvants, carriers, or diluents, are available to the public. Moreover, pharmaceutically acceptable auxiliary substances, such as pH adjusting and buffering agents, tonicity adjusting agents, stabilizers, wetting agents and the like, are available to the public. [0205] The fusion molecules of the invention may be employed in combination with a suitable pharmaceutical carrier to comprise a pharmaceutical composition for parenteral administration. Accordingly, the invention provides a composition comprising an FGFR fusion molecule of the invention and a pharmaceutically acceptable carrier. Such compositions comprise a therapeutically effective amount of the polypeptide, agonist, or antagonist and a pharmaceutically acceptable carrier. Such a carrier includes, but is not limited to, saline, buffered saline, dextrose, water, glycerol, ethanol, and combinations thereof. The formulation should suit the mode of administration. [0206] In pharmaceutical dosage, the FGFR fusion molecule compositions can be administered in the form of their pharmaceutically acceptable salts, either alone or in appropriate association or combination with other pharmaceutically active compounds. The FGFR fusion molecule compositions are formulated in accordance with the mode of administration. Thus, the subject compositions can be formulated into preparations in solid, semi-solid, liquid, or gaseous forms, such as tablets, capsules, powders, granules, 66 ointments, solutions, suppositories, enemas, injections, inhalants, and aerosols. The methods and excipients cited herein are merely exemplary and are in no way limiting. [0207] The agents, polynucleotides, and polypeptides can be formulated into preparations for injection by dissolving, suspending, or emulsifying them in an aqueous or nonaqueous solvent, such as vegetable or other similar oils, synthetic aliphatic acid glycerides, esters of higher aliphatic acids or propylene glycol; and if desired, with conventional additives such as solubilizers, isotonic agents, suspending agents, emulsifying agents, stabilizers and preservatives. They may be formulated into preparations for administration via inhalation, for example as formulated into pressurized acceptable propellants such as dichlorodifluoromethane, propane, nitrogen, and the like. The FGFR fusion proteins of the invention can be formulated into a sustained release microcapsules, such as with biodegradable or non-biodegradable polymers, using techniques known in the art. An example of a biodegradable formulation suitable for use herein includes poly lactic acid-glycolic acid polymer. An example of a non biodegradable formulation suitable for use herein includes a polyglycerin fatty acid ester. A method of making these formulations is described in, for example, EP 1 125 584 Al. Other formulations for parenteral delivery can also be used, as conventional in the art. [0208] The FGFR fusion molecule compositions will be formulated and dosed in a fashion consistent with good medical practice, taking into account the clinical condition of the individual subject, the site of delivery of the fusion molecule composition, the method of administration, the scheduling of administration, and other factors known to practitioners. The effective amount of FGFR fusion molecule for purposes herein is thus determined by such considerations. [0209] The invention also provides a pharmaceutical pack or kit comprising one or more containers filled with one oi more of effective doses of the pharmaceutical FGFR fusion protein compositions of the invention. Associated with such container(s) can be a notice in the form prescribed by a governmental agency regulating the manufacture, use, or sale of pharmaceuticals or biological products. Such a notice reflects the agency's approval for manufacture, use, or sale for human administration. In addition, the FGFR fusion molecules of the invention may be employed in conjunction with other therapeutic agents. [0210] Unit dosage forms can be provided wherein each dosage unit contains a predetermined amount of the composition containing one or more agents. In an embodiment, an FGFR fusion molecule composition is supplied in single-use prefilled 67 syringes for injection. The composition may comprise saline, sucrose, or the like; a buffer, such as phosphate, or the like; and be formulated within a stable and effective pH range. In an embodiment, an FGFR fusion molecule composition is provided as a lyophilized powder in a multiple-use vial, which can be reconstituted upon addition of an appropriate liquid, for example, sterile bacteriostatic water. In an embodiment, an FGFR fusion molecule composition comprises one or more substances that inhibit protein aggregation, including, but not limited to, sucrose or arginine. In an embodiment, a composition of the invention comprises heparin and/or a proteoglycan. [0211] These pharmaceutical compositions are administered in an amount effective for treatment and/or prophylaxis of the specific indication. The effective amount is typically dependent on the weight of the subject being treated, his or her physical or health condition, the extensiveness of the condition to be treated, and/or the age of the subject being treated. In general, the FGFR fusion proteins of the invention are to be administered in an amount in the range of about 5 ug/kg body weight to about 10 mg/kg body weight per dose. Optionally, the FGFR fusion proteins of the invention can be administered in an amount in the range of about 10 ug/kg body weight to about 9 mg/kg body weight per dose. Further optionally, the FGFR fusion proteins of the invention can be administered in an amount in the range of about 100 ug/kg body weight to about 8 mg/kg body weight per dose. Still optionally, the FGFR fusion proteins of the invention can be administered in an amount in the range of about 1 mg/kg body weight to about 7 mg/kg body weight per dose. [0212] The FGFR fusion protein compositions can be administered as needed to subjects in need of inhibition of FGF ligand/FGFR signaling pathway. Determination of the frequency of administration can be made by persons skilled in the art, such as an attending physician based on considerations of the condition being treated, age of the subject being treated, severity of the condition being treated, general state of health of the subject being treated and the like. In one embodiment, an effective dose of the FGFR fusion protein is administered to a subject one or more times. In one embodiment, the FGFR fusion protein of the invention is administered to the subject at least twice a week for at least a week. In another embodiment, the FGFR fusion protein is administered at least three times a week for at least one week. In a further embodiment, the FGFR fusion protein is administered to the subject for at least two weeks. In yet another embodiment, the FGFR fusion protein of the invention is administered to the subject for at least three weeks. Administration of the FGFR fusion protein can be continuously for at least two or 68 three weeks or can be non-continuous, such as taking a one or two week break from treatment and resuming treatment after such break. Combination Therapy [0213) FGFR fusion molecules of the invention may be administered alone or with other rnodes of treatment. They may be provided before, substantially contemporaneous with, or after other modes of treatment, for example, surgery, chemotherapy, radiation therapy, or the administration of a biologic, such as a therapeutic antibody. Accordingly, the invention provides a method of combining treatment which blocks the signaling pathways utilized by fibroblast growth factors and receptors with treatment which blocks the signaling pathways utilized by other growth factors, which can be expected to be more effective in patients with tumors that express FGFs and/or FGFRs, as well as other growth factors and/or their receptors. This therapeutic approach can be applied to rapidly growing tumors and highly vascularized tumors, for example, glioblastomas. FGFR fusion molecules of the invention can be used in combination with fusion molecules of other growth factor receptors. For example, the FGFR fusion protein of the invention can be combined with a soluble VEGFR to inhibit tumor growth and/or to inhibit angiogenesis in tumors. [0214] Further, the invention provides combination therapy that blocks the FGF and other signaling pathways such as PDGF, VEGF, and/or EGF signaling pathways. The FGFR fusion molecules of the invention can be used in such combination therapy. One or more of the agents that inhibit PDGFR-alpha, PDGFR-beta, VEGFR, and/or EGF receptors can be combined with FGFR fusion proteins of the invention for therapeutic use. The compositions that block the FGF signaling pathways may be provided simultaneously or may be provided sequentially in any order with compositions that block the PDGF, VEGF, and/or EGF signaling pathways. Combination therapy may include the use of fusion proteins comprising the extracellular domains of PDGFR-alpha,

PDGFR

beta, VEGFR, EGFR, and the FGFR fusion proteins of the invention. Uses of FGFR Fusion Molecules [0215) FGFR fusion molecules of the invention, and fragments and variants thereof, may be used to diagnose, provide a prognosis for, prevent, treat, and develop treatments for disorders mediated, either directly or indirectly, by hyperactive or excess FGF ligand or FGFR. FGFR fusion molecules of the invention, and fragments and variants thereof, may be administered to a patient at risk for or suffering from such a disorder.

69 [0216] Accordingly, the invention provides a method of diagnosing a disease characterized by the over-expression of one or more FGF and/or FGFR, or a fragment or variant thereof, by measuring the real-time receptor ligand binding of one or more FGF to one or more FGFR. The method can be performed, for example, by providing a biological specimen from a subject, and measuring the binding of an FGF ligand or FGFR in the specimen to one or more cognate FGFR or FGF ligands. The results of the binding measurements can be used to diagnose the presence or absence of a disease characterized by the over-expression of the FGF(s) or FGFR(s). 10217] The invention also provides a method of treating a condition in a subject comprising providing a composition comprising an FGFR fusion molecule of the invention and administering the composition to the subject, wherein the condition comprises a proliferative disease, including cancers and disorders of angiogenesis. The FGFR fusion molecules of the invention are useful for inhibiting cancer cell proliferation and/or viability. The FGFR fusion molecules of the invention can be used accordingly in a variety of settings for the treatment of animal, including human, cancers. [0218) The FGFR fusion molecules of the invention can be used to treat, modulate, or prevent malignant, pre-malignant, and benign tumors. For example, they can treat metastasizing or non-metastasizing malignant tumors, which are typically at an advanced stage of tumor development, and may be life threatening. They can also be used to treat pre-malignant tumors, which are typically at a more advanced stage of tumor development than benign tumors, but have not progressed to malignancy. They can further be used to treat benign tumors, which typically show some abnormal cell characteristics and are at an early stage in tumor development. The benign tumor may or may not progress to a pre malignant or malignant tumor. The FGFR fusion molecules of the invention can be used to treat solid tumors formed by a collection of cells typically localized in a tissue or organ, for example, sarcomas and carcinomas such as, but not limited to fibrosarconas, myxosarcomas, liposarcomas, chondrosarcomas, osteogenic sarcomas, chordomas, angiosarcomas, endotheliosarcomas, lymphangiosarcomas, lymphangioendotheliosarcomas, synoviomas, mesotheliomas, Ewing's tumors, leiomyosarcomas, rhabdomyosarcomas, colon carcinomas, colorectal cancers, gastic cancers, pancreatic cancers, breast cancers, ovarian cancers, prostate cancers, squamous cell carcinomas, basal cell carcinomas, adenocarcinomas, sweat gland carcinomas, sebaceous gland carcinomas, papillary carcinomas, papillary adenocarcinomas, oystadenocarcinomas, medullary carcinomas, bronchogeaic carcinomas, renal cell 70 carcinomas, hepatomas, liver metastases, bile duct carcinomas, choriocarcinomas, seminomas, embryonal carcinomas, thyroid carcinomas such as anaplastic thyroid cancers, Wilms' tumors, cervical cancers, testicular tumors, lung carcinomas such as small cell lung carcinomas and non-small cell lung carcinomas, bladder carcinomas, epithelial carcinomas, gliomas, astrocytomas, mcdulloblastomas, craniopharyngiomas, ependymomas, pinealomas, hemangioblastomas, acoustic neuromas, oligodendroglionias, meningiomas, melanomas, neuroblastornas, and retinoblastomas. Also among the cancers within the scope of the invention are hematologic malignancies, breast cancer, such as infiltrating ductal carcinoma and adenocarcinoma; lung cancer, such as squamous cell carcinoma, non-small cell lung cancer, and lung adenocarcinoma; prostate cancer; bladder cancer, pancreatic cancer; ovarian cancer, salivary cancer; pituitary cancer; renal cell carcinoma; melanoma; glioblastoma; retinoblastoma; and/or cancer metastases in bone, including bone metastasis from prostate cancer. [0219] Tumors comprising dysproliferative changes, such as hyperplasias, metaplasias, and dysplasias, can be heated, modulated, or prevented with the present invention as well, such as those found in epithelial tissues, including the cervix, esophagus, and lung, for example. -yperplasia is a form of controlled cell proliferation involving an increase in cell number in a tissue or organ, without significant alteration in structure or function. By way of example, endometrial hyperplasia often precedes endometrial cancer. Metaplasia is a form of controlled cell growth in which one type of adult or fully differentiated cell substitutes for another type of adult cell. Metaplasia can occur in epithelial or connective tissue cells. Atypical metaplasia involves a somewhat disorderly metaplastic epithelium. Dysplasia is frequently a forerunner of cancer, and is found mainly in the epithelia; it is a disorderly form of non-neoplastic cell growth, involving losses in individual cell uniformity and in the cell's architectural orientation. Dysplasia characteristically occurs where there exists chronic irritation or inflammation and is often found in the cervix, respiratory passages, oral cavity, and gall bladder. Other examples of benign tumors which can be treated, modulated or prevented in accordance with the present invention include arteriovenous (AV) malformations, particularly in intracranial sites and myoleomas. [0220] The FGFR fusion molecules of the invention, or variants or fragments thereof, can be used to treat cancer patients sensitive to the effects of FGFR signaling. They are useful in a subset of patients that over-express FGFRI and/or FGF-2, for example, subsets of patients with breast cancer, lung cancer, kidney cancer, prostate 71 cancer, and glioblastoma. Treatment effectiveness can be assessed, for example, by measuring the patient's level of a biomarker, for example, FGF-2, pFGFR, DSP, and/or pFRS2 (Guddo et al., Hum. Patrol. 30:788-794 (1999)), as described above. [0221] The invention also provides compositions and methods for treating glioblastoma, a rapidly growing and highly vascularized tumor. Platelet-derived growth factor (PDGF) is expressed at high levels in many human glioblastomas. Further, in addition to their role in promoting tumor cell growth and survival, FGFs are potent angiogenic factors which may be expected to promote the growth of highly vascular tumors, such as glioblastomas. Blocking the effects of PDGF on cell growth or survival together with blocking the FGF signaling pathway with the FGFR fusion protein of the invention may thus retard the progression of glioblastoma development. [0222] FGF-2 and FGFR1 are expressed in the tumor cells and the tumor associated stromal cells and vessels of patients with non-small cell lung cancer. The FGFR fusion proteins of the invention, such as FGFRI-EIlc-Fc or RlMut4, for example, can be administered to such patients to block growth stimulation by FGF-2 binding to FGFRI and inhibit tumor growth. [0223] Stromal-epithelial interactions are important determinants of malignant vs. benign prostatic growth (Conte et al., Int. J. Cancer 107:1-10 (2003)). Prostate cancer, and also breast cancer, kidney cancer, and multiple myeloma, tend to metastasize to the bone. While breast cancer bone metastases tend to form lytic bone lesions, prostate cancer metastases tend to form blastic lesions characterized by an excess of abnormally dense bone. Further interaction of prostate cancer metastasis with the local bone environment may then alter normal bone homeostasis, shifting it toward an osteoblastic phenotype. Kidney metastases may exhibit both lytic and blastic bone lesions. [0224] Since FGFs contribute to normal bone formation and are expressed locally in the bone stromal environment, they may play a role in seeding, growth, and survival of prostate cancer bone metastases. FGFs have been implicated in bone formation, affecting osteoprogenitor cell replication, osteoblast differentiation, and apoptosis. Thus, agents which block FGF/FGFR interactions, including FGFR fusion molecules of the invention, or variants or fragments thereof, can be used to treat bone metastases in prostate cancer. Such agents will not only inhibit local osteoblastic conversion events, but also inhibit initial seeding, growth, and survival of prostate cancer bone metastases. [0225] The invention also provides methods of using FGFR fusion proteins of the invention, or variants or fragments thereof, to inhibit angiogenesis, for example, in 72 tumorigenesis and macular degeneration. By way of example, fusion proteins comprising FGFRI-llc and/or FGFR4 maybe used to bind undesirable proangiogenic FGFs and decrease angiogenesis. Useful compositions include those comprising fusion molecules comprising the fusion proteins of the invention as described herein, including those with Fc fusion partners. [0226] The invention provides methods of treating cancers resistant to other cancer therapeutic agents. For example, the FGFR fusion proteins of the invention can be used to treat cancers resistant to ErbB oncogene inhibitors, such as Herceptin*. They are also useful in treating cancers resistant to inhibitors of VEGF, such as Avastin*. [0227] The FGFR fusion proteins and the polynucleotide molecules that encode them are useful in treating proliferative diseases and diseases involving angiogenesis, including cancer. They can be used to diagnose, prevent, and treat these diseases. [0228] With respect to ranges of values, the invention encompasses each intervening value between the upper and lower limits of the range to at least a tenth of the lower limit's unit, unless the context clearly indicates otherwise. Further, the invention encompasses any other stated intervening values. Moreover, the invention also encompasses ranges including either or both of the upper and lower limits of the range, unless specifically excluded from the stated range. [0229] Unless defined otherwise, the meanings of all technical and scientific terms used herein are those commonly understood by one of skill in the art to which this invention belongs. One of skill in the art will also appreciate that any methods and materials similar or equivalent to those described herein can also be used to practice or test the invention. [0230] It must be noted that, as used herein and in the appended claims, the singular forms "a," "or," and "the" include plural referents unless the context clearly dictates otherwise. Thus; for example, reference to "a subject polypeptide" includes a plurality of such polypeptides and reference to "the agent" includes reference to one or more agents and equivalents thereof known to those skilled in the art, and so forth. [0231] Further, all numbers expressing quantities of ingredients, reaction conditions, % purity, polypeptide and polynucleotide lengths, and so forth, used in the specification and claims, are modified by the term "about," unless otherwise indicated. Accordingly, the numerical parameters set forth in the specification and claims are approximations that may vary depending upon the desired properties of the present invention. At the very least, and not as an attempt to limit the application of the doctrine 73 of equivalents to the scope of the claims, each numerical parameter should at least be construed in light of the number of reported significant digits, applying ordinary rounding techniques. Nonetheless, the numerical values set forth in the specific examples are reported as precisely as possible. Any numerical value, however, inherently contains certain errors from the standard deviation of its experimental measurement. [0232) The specification is most thoroughly understood in light of the references cited herein. Each of these references is hereby incorporated by the reference in its entirety. Exemplary Modes for Carrying out the Invention [0233] The examples, which are intended to be purely exemplary of the invention and should therefore not be considered to limit the invention in any way, also describe and detail aspects and embodiments of the invention discussed above. The examples are not intended to represent that the experiments below are all or the only experiments performed. Efforts have been made to ensure accuracy with respect to numbers used (for example, amounts, temperature, etc.) but some experimental errors and deviations should be accounted for. Unless indicated otherwise, parts are parts by weight, molecular weight is weight average molecular weight, temperature is in degrees Centigrade, and pressure is at or near atmospheric. Example 1: Sequence Alignment of Partial IgMI Domains and the Membrane Proximal Regions of FGFRs and FGFR Variants [0234] FIG. IA shows an alignment of the amino acid sequences of a part of the IgIm domain of each of the seven FGFR family members, FGFRI-lIb, FGFRI-mc, FGFR2-Mb, FGFR2.-II1c, FGFR3-Iflb, FGFR3-IIlc, and FGFR4, using the Clustal W version 1.8 program of the European Molecular Biology Laboratory (EMBL) bioinformatics search site. [0235] FIG. IA also illustrates the organization of the parental FGFR1-lIIc-Fc fusion protein (nucleotide SEQ ID NO: 4, protein SEQ ID NO: 95) and corresponding mutants. The IgM domain is followed by the C-terminal portion of the ECD, which is followed by the Fe portion of an antibody. The alignment marks the truncation locations of the R1Muti (nucleotide SEQ ID NO: 6, protein SEQ ID NO: 97), R1Mut2 (nucleotide SEQ ID NO: 7, protein SEQ ID NO: 98), RlMut3 (nucleotide SEQ ID NO: 8, protein SEQ ID NO: 99), RlMut4 (nucleotide SEQ ID NO: 9, protein SEQ ID NO: 100), and RlMut5 (nucleotide SEQ ID NO: 10, protein SEQ ID NO: 101) variants of FGFR1-IIIc 74 and the position of an Fe portion of an IgG antibody. R1Muti, RlMut2, RlMut3, RlMut4, and R1Mut5 also had the linker encoding amino acids "GS" removed from between the FGFR1-IIc and Fc domains. [0236] FIG. lB illustrates additional variants from the parental FGFRI-IIc-Fc, which were also made and used in the invention. The variants shown in FIG. 1B all had the linker GS removed. The variant RlMut6 (nucleotide SEQ ID NO: 5, protein SEQ ID NO: 96), had only the GS linker deleted. Another variant, R1Mut7 (nucleotide SEQ ID NO: 11, protein SEQ ID NO: 102), had a deletion of amino acid residues PA. The variant RlMut8 (nucleotide SEQ ID NO: 12, protein SEQ ID NO: 103), had an amino acid substitution P364M, in which the proline residue was substituted with methionine. The variant R1Mut9 (nucleotide SEQ ID NO: 13, protein SEQ ID NO: 104) had an amino acid substitution M367N, in which the methionine residue was substituted with asparagine. The variant RiMut10 (nucleotide SEQ ID NO: 44, protein SEQ 11) NO: 135) had an amino acid substitution P364G, in which the proline residue was substituted with glycine. [0237] FIG. 2 shows an alignment of the amino acid sequences of a part of the IgIII domain of each of the seven FGFR family members, using the Clustal W version 1.8 program of EMBL (European Molecular Biology Laboratory) bioinformatics search site. FIG. 2 illustrates the organization of the FGFR4-Fc fusion protein (nucleotide SEQ ID NO: 65, protein SEQ ID NO: 156) and corresponding mutants. The alignment marks the truncation locations of the R4Mut1 (nucleotide SEQ ID NO: 71, protein SEQ ID NO: 162), R4Mut2 (nucleotide SEQ ID NO: 72, protein SEQ ID NO: 163), R4Mut3 (nucleotide SEQ ID NO: 73, protein SEQ ID NO: 164), R4Mut4 (nucleotide SEQ ID NO: 74, protein SEQ ID NO: 165), R4Mut5 (nucleotide SEQ I) NO: 75, protein SEQ ID NO: 166) and R4Mut6 (nucleotide SEQ ID NO: 66, protein SEQ ID NO: 157) variants of FGFR4. FIG. 2 also indicates the position of an Fe portion of an IgG antibody. R4Mutl, R4Mut2, R4Mut3, R4Mut4, R4Mut5, and R4Mut6 also had the linker encoding amino acids 'GS' removed from between the FGFR4 and Fc domains. Example 2: Expression of FGFR Fusion Proteins [0238] The fusion proteins of the invention were expressed in 293-6E host cells using the pTT5 vector (Biotechnology Research Institute; Montreal, Canada) transfected into 293-6E cells (Biotechnology Research Institute; Montreal, Canada), which were then cultured to produce the fusion proteins. An expression vector that comprised the cDNA of FGFR1-.Tc-Fc (SEQ ID NO: 4), encoding the extracellular domain of human FGFR1-IIIc 75 (SEQ ID NO: 1) was constructed from an open-reading frame cDNA library prepared internally. This cDNA was linked at its C-terminus through a linker encoding the amino acids GS to cDNA encoding an Fc fragment of human IgGI heavy chain (SEQ ID NO: 80) to produce a fusion construct hereafter referenced as "FGFR1-Mc-Fc cDNA" and the expression product thereof as "FGFRI-Mc-Fc protein." The Fc fragment was also obtained from an open-reading frame cDNA library prepared internally. This cDNA fusion construct was inserted into a pTTS vector by conventional techniques to produce the FGFR1-IfIc-Fc/pTT5 expression vector, [0239] Expression constructs for expressing the FGFR3-mIe-Fc (nucleotide SEQ ID NO: 85, protein SEQ ID NO: 176) and FGFR4-Fc fusion proteins in 293-6E host cells using the pTT5 vector were made in a manner similar to that described above using cDNAs prepared internally and conventional techniques. Similar expression constructs for expressing FGFR variants, such as RIMut1, RlMut2, R1Mut3, RlMut4, RIMut5, R1Mut6 (GS deletion), RlMut7 (PA deletion), R1Mut8 (P364M), RlMut9 (M367N), R4Mutl, R4Mut2, R4Mut3, R4Mut4, R4Mut5, and R4Mut6 (GS deletion) using the pTTS vector were each made from the FGFRA-ffc-Fc cDNA using PCR and conventional mutagenesis techniques. [0240] The variants RiMutl, RIMut2, RlMut3, RlMut4, and R1Mut5, produced in this manner, each contained the same amino acid sequence as the parent FGFRI-Mc-Fc fusion protein (protein SEQ ID NO: 95) except for the deletion of the linker amino acids GS and also certain amino acid residues at the C-terminus of the wildtype FGFRI -Ine extracellular domain, as described in Example 1. RiMutI comprised an amino acid sequence ending with amino acid residues MTSP immediately preceding the Fc fragment. R1Mut2 comprised an amino acid sequence ending with amino acid residues RPAV immediately preceding the Fe fragment. RlMut3 comprised an amino acid sequence ending with amino acid residues ERPA immediately preceding the Fc fragment. RlMut4 comprised an amino acid sequence ending with amino acid residues LEAL immediately preceding the Fc fragment. R1Mut5 comprised an amino acid sequence ending with amino acid residues AWLT immediately preceding the Fc fragment. The variants R1Mut6, RlMut7, RIMut8, and RlMut9, also produced in this manner, each contained the same amino acid sequence as the parent FGFRI-Ulc-Fc the linker GS removed. [0241] The variants R4MutI, R4Mut2, R4Mut3, R4Mut4, R4Mut5, and R4Mut6, which were described in Example 1, were also produced from 293-6E host cells using the pTT5 vector in the manner described for FGFRI-MIc-Fc and the RI mutants. R4Mutl 76 comprised an amino acid sequence ending with amino acid residues AAPE immediately preceding the Fc fragment. R4Mut2 comprised an amino acid sequence ending with amino acid residues PTWT immediately preceding the Fe fragment. R4Mut3 comprised an amino acid sequence ending with amino acid residues LPEE immediately preceding the Fe fragment. R4Mut4 comprised an amino acid sequence ending with amino acid residues TVLP immediately preceding the Fc fragment. R4Mut5 comprised an amino acid sequence ending with amino acid residues LTVL immediately preceding the Fe fragment. R4Mut6 comprised an amino acid sequence ending with'amino acid residues RYTD immediately preceding the Fc fragment. [0242] The host cell line CHO-S can, in certain embodiments, produce recombinant proteins with higher yields and/or different glycosylation patterns than the 293-6E host cell line. Fusion proteins of the invention were expressed in CHO-S host cells with the vector pcDNA3.1 (Invitrogen; Carlsbad, CA). The expression vectors were transfected into the CHO-S host cells (Invitrogen; Carlsbad, CA), which were then cultured to produce the fusion proteins. The FGFR1-flc-Fc cDNA was subcloned into the pcDNA3.1 vector using PCR and conventional subcloning techniques. Expression constructs for expressing FGFR3-Mc-Fc and FGFR4-Fc fusion proteins in CHO-S host cells using the pcDNA3.1 vector were made in a manner similar to that described above, using PCR and conventional subcloning techniques. [0243) Similar expression constructs for expressing the FGFR4-Fc variants R4Mutl,'R4Mut2, R4Mut3, R4Mut4, R4Mut5, and R4Mut6 in CHO-S host cells using the ,pcDNA3.1 vector was made from the FGFR4-Fc cDNA using PCR and conventional subeloning techniques. Expression vectors for other FGFR-Fc fusion proteins and variants can also be made in a similar manner and the fusion proteins expressed as discussed herein using methods known in the art. [0244] DG44 is a cell line derivative of the CHO-S cell line and can, in some embodiments, produce higher yields of recombinant proteins than CHO-S cells. The fusion proteins of the invention were expressed in DG44 host cells (Invitrogen; Carlsbad, CA) using the pDEF38 vector (ICOS Corporation; Bothell, WA) transfected into DG44 cells as host cells, which were then cultured to produce the fusion proteins. For example, 'FGFRI -me-Fc cDNA was subcloned into the pDEF38 vector using PCR and conventional suboloning techniques. [0245] A similar expression construct for expressing the FGFR1 -lc-Fc variant RlMut4 in DG44 host cells using the pDEF38 vector was made using PCR and 77 conventional subeloning techniques. Expression vectors for other FGFR-Fc fusion proteins and variants, for example, FGFR3-Ic-Fc and FGFR4-Fc, can also be made in a similar manner and fusion proteins expressed as discussed herein. [0246] Long-term expression of fusion proteins in mice using a mini-circle vector with the fusion constructs, was performed as described in Chen el al., Hum. Gene Ther. 16:126 - 131 (2005); Rui, E. et al., Hum. Gene Ther. 16:558 - 570 (2005); and WO 04/020605, using a parent vector obtained from Dr. Mark Kay at Stanford University (Stanford, CA). This parent vector contained an alpha1-antitrypsin promoter, an apoE enhancer, a human factor IX intron, and a bovine polyA sequence. The parent vector was modified by inserting the FGFRI-IUc-Fc, FGFR4-Fc, or RlMut4 cDNA as the gcne of interest, placing such cDNA immediately after the human Factor DC intron in the parent vector using PCR and conventional subcloning techniques. Similar expression vectors for other FGFR-Fc fusion proteins, including FGFR3-Ilc-Fc and the variants described herein, can also be made, and fusion proteins expressed, as discussed herein using methods known in the art. Example 3: Transient Expression of Fusion Proteins in 293-6E Cells and CHO-S Host Cells [0247) The FGFRI-lUc-Fc/pTT5 expression vector was designed to provide transient expression in 293-6E host cells. The 293-6E cells were previously adapted to serum-free suspension culture in Free-Style medium (Invitrogen; Carlsbad, CA). The cells were transfected with the expression vector while in logarithmic growth phase (log phase growth) at a cell density of between 9 x 10 5 /ml and 1.2 x 10 6 /ml. [0248] In order to transfect 500 ml of cell suspension, a transfection mixture was first made by mixing 500 micrograms (ug) of the expression vector DNA in 25 milliliters (ml) of sterile phosphate buffered saline (PBS) with 1 milligram (mg) of polyethyleniinine (at a concentration of about 1 mg/ml solution in sterile water) in 25 ml of sterile PBS. This transfection mixture was incubated for 15 min at room temperature. Following incubation, the transfection mixture was added to the 293-6E cells in log phase growth to transfect the cells. The cells and the transfection mixture were then incubated at 370 C in 5% CO 2 . After 24 hr of incubation, Trypton-NI (Organotechnie S.A.; La Courneuve, France; 20% solution in sterile FreeStyle medium) was added to a final concentration of 0.5% (v/v). The mixture was maintained at 370 C and 5% CO 2 for about 6 -8 days until the cells reached a density of about 3 - 4 x 106 cells/ml and demonstrated greater than 78 about 80% viability. To harvest the fusion protein from the cell culture medium, cells were pelleted at 400 x g for 15 min at 4* C and the supernatant decanted then cleared of cell debris by centrifugation at 3,315 x g for 15 min at 4* C. The cleared supernatant containing the fusion protein was then purified, as described in more detail below. [0249] The FGFR fusion proteins FGFR3-IlIc-Fc, FGFR4-Fc, and the FGFR fusion variants RiMuti, R1Mut2, R1Mut3, RIMut4, RlMut5, R1Mut6, R1Mut7, R1Mut8, and RlMut9 were similarly produced by transient expression in 293-6E cells in pTT5 vectors constructed as described in Example 2. Other FGFR-Fc fusion proteins and variants can also be similarly made and expressed in 293-6E host cells using the methods discussed herein. [0250] Small batches (approximately 1 - 2 mg) of RMut4 protein, for example, for use in in vivo studies, were rapidly produced from CHQ-S cells grown in suspension and transiently transfected with the plasmid construct RIMut4/pcDNA3.l. Briefly, suspension CHO-S cells (Invitrogen; Carlsbad, CA) were cultured in CD-CHO scrum free medium supplemented with I- glutamine, and lx hypoxanthine/thymidine (HT) (Invitrogen; Carlsbad, CA). The day before transfection, CHO-S cells were seeded into a shaker flask at a density of about 5 x 10 5 /mi and reached a density of about 1 x 10 6 /m on the day of transfection. The cells were harvested and about 1 x 1 cells per transfection reaction were pelleted by centrifugation. Each cell pellet was resuspended in 0.1 ml Nucleofector V solution and transferred to an Amaxa Nucleofector cuvette (Amaxa; Cologne, Germany). About 5 ug of the RlMut4/pcDNA3.1 plasmid DNA was added and mixed with the suspended CHO-S cells in the cuvette. Cells were then electroporated with an Amaxa Nucleofector using program U-024. [02511 Larger batches can be produced, as well. For example, tW produce 200 ml, 12 transfcction reactions were carried out and the electroporated cells were cultured in CD-CHO medium (supplemented with L-glutamine, 1x hypoxanthine/thymidine (HT) at density of 0.5 x 10 6 /ml. After six days, the cell density reached about 6 -7 x 10 6 /ml with a viability of about 95%. The supernatant from the culture was harvested by centrifugation and was suitable for purification. Using this method, one mg of RlMut4 protein can be produced in about one week from 200 ml of transiently transfected cultured cells. [0252] The FGFR fusion proteins FGFR1-IIlc-Fc, FGFR3-IIc-Fc, and FGFR4-Fc, and the variants R4Mutl, R4Mut2, R4Mut3, R4Mut4, R4Mut5, and R4Mut6 were similarly produced by transient expression in CHO-S cells in pcDNA3.1 vectors 79 constructed as described in Bxample 2. Other FGFR-Fc fusion proteins and variants can also be made and expressed in CHO-S host cells using the methods discussed herein. Example 4: Cell Line Development for Stable Production of Fusion Proteins in CHO-S Host Cells [0253] The FGFRI-Mc-Fc/pcDNA3.1 expression vector was designed to provide stable expression in appropriate mammalian cells, such as CHO-S. This vector was transfected into CHO-S cells containing the dihydrofolate reductase (DHFR) gene, which were derived from adherent CHO-Ki cells by adaptation to senin-free suspension culture in CD-CHO medium (Invitrogen; Carlsbad, CA). [02541 Transfection was carried out using an Amaxa Nucleofector II (Amaxa; Cologne, Germany) according to manufacturer's recommendations. In this process, about 1 x 106 CHO-S cells were resuspended in 300 ul of Amaxa's Solution V (Amaxa; Cologne, Germany) and transferred into an electroporation cuvette. About 5 ug of plasmid DNA containing the FGFR1 -Itlc-Fc/pcDNA3.1 expression vector was added to the cells in the cuvette and DNA transfer was initiated using Amaxa program U-024 in Amaxa's Nucleofector Device IL After the DNA transfer to the CHO-S cells, the cell suspension was immediately transferred into 1 ml of pre-warmed CD-CHO medium and then incubated at 370 C for 10 min. The cell suspension was then transferred into 10 nil of pre warmed CD-CHO medium and cultured for 48 hr in a T-75 flask at 37* C and 5% CO 2 . The vector pcDNA3.1 carried the G418-selection gene (Invitrogen; Carlsbad, CA). About 48 hr after the DNA transfer with FGFR1-Il~c-Fc/pcDNA3. 1, G418 (Invitrogen; Carlsbad, CA) selection reagent was added to a final concentration of 400 ug/ml. About 2-3 weeks after introducing selective pressure, and when the cells reached confluency, they were expanded into T-225 flasks with fresh selection medium. The cells were then cryo preserved until use. [0255] The cryo-preservation medium contained 46.25% CD-CHO medium, 46.25% conditioned CD-CHO medium (usually supernatant from the culture being cryo preserved) and 7.5% dimethyl sulfoxide (DMSO). About 5 - 10 x 106 cells/vial were resuspended in I ml of cryo-preservation medium and slowly frozen (about 1* C/min) to about -80* C. The following day, the frozen cells were transferred into liquid N 2 (about 190* C). Upon use, the cells were thawed quickly, by transferring the cryo-vial into a 370 C water bath and resuspending the thawed cell suspension in at least 10 ml of fresh CD- 80 CHO medium. Usually about 60% of the cells would recover and start proliferating about 24 - 48 hr after thawing. [0256] Following cryo-preservation and recovery, the cells were plated on 96-well plates at a density of 2 cells/well/200 ul and cultured in CD-CHO medium at 37* C and 5% COz for three weeks. G418 selective pressure (400 ug/mi) was added after cell proliferation resumed. In order to identify transfected cell clones expressing FGFRI-II1c Fc fusion protein, the cell culture supernatant of each well was screened by Western blot. FGFR1-IIIc-Fc was detected using a polyclonal goat anti-human IgG Fc gamma-specific antibody (Jackson Immuno Research; West Grove, PA) conjugated to horseradish peroxidase (HIRP). (0257] FGFRl-Ifc-Fc produced from transfected CHO-S cells had a higher molecular weight than that produced from transfected 293-6E cells, indicating increased glycosylation in the CHO-S cell product. Thirty-one cell clones from wells which produced a distinct Fc-imnunoreactive band in Western blot were transferred to T-75 flasks with 10 nil of CD-CHO medium with 400 ug/mI G418. After two weeks, supernatants from each of these cultures was tested by SDS-PAGE and only those transfected cell clones producing a strongly visible band were continued for further analysis. The 14 highest expressing clones were tested for cell specific productivity and ranked accordingly. The two highest producing clones were adapted to suspension culture over a period of one month. A total of ten different culture media were tested regarding volumetric protein productivity and protein integrity. [0258] Stable CHO-S host cell lines producing FGFR4-Fc fusion protein from the pcDNA3.1 expression vector were also created in a manner similar to that described above for FGFRl-IIfc-Fc. Stable CHO-S host cell lines producing other FGFR-Fc fusion proteins and variants can also be created in a manner similar to that described herein, using the pcDNA3.1 expression vector described in Example 2. Example 5: Cell Line Development for Stable Production of Fusion Proteins in DG44 Cells [02591 The expression vector comprising RlMut4/pDEF38 described in Example 2 was used to transfect DG44 cells for the stable production of RlMut4 fusion protein. In this process, the untransfected DHFR-negative CHO cell line, DG44, was cultured in CHO-CD serum free medium (rvine Scientific; Irvine, CA) supplemented with 8 mM L glutamine, 1x hypoxanthine/thymidine (HT; Invitrogen; Carlsbad, CA), and 18 ml/L of 81 Pluronic-68 (Invitrogen; Carlsbad, CA). About 50 ug of plasmid DNA containing RlMut4/pDEF38 was first linearized by digestion with the restriction enzyme PvuI, then precipitated by addition of ethanol, briefly air-dried, and subsequently resuspended in 400 ul of sterile, distilled water. Cultured DG44 cells, as host cells, were seeded into a shaker flask at a cell density of about 4 x 105/ml the day before transfection, and reached a density of about 0.8 x 10 6 /ml on the day of transfection. The cells were harvested and about I x 107 cells per transfection unit were pelleted by centrifugation [0260] The cells were transfected by resuspending each cell pellet in 0.1 ml of Nucleofector V solution and transferred the suspension to an Amaxa Nucleofector cuvette (Amaxa; Cologne, Germany). About 5 ug of the resuspended linearized plasmid DNA was added and mixed with the suspended DG44 cells in the cuvette. Cells were then electroporated with an Amaxa Nucleofector Device II using program U-024. Blectroporated cells were cultured in CHO-CD medium for two days and then transferred into a selective medium comprising CHO-CD serum free medium supplemented with 8 mM L-glutamine, 18 ml/L Pluronic-68, and 10% dialyzed fetal calf serum (FCS; Invitrogen; Carlsbad, CA; without HT). This selective medium was changed once every week. After about 12 days, I ug/m R3 Long IGF I growth factor (Sigma; St. Louis, MO) was added to the medium and the culture was continued for another four days until confluent. The supernatants from pools of stably transfected cell lines were assayed by a sandwich FGFR1-IlIc-Fc ELISA to determine the product titer (for details of this sandwich ELISA, see Example 15). This transfection method generated an expression level of about 21 ug/mil for RlMut4 from the pools of stably transfected cells. [0261) Stable DG44 host cell lines producing FGFRI-Ic-Fc fusion protein from the pDEF38 expression vector were also created in a manner similar to that described herein for R1Mut4. Stable DG44 host cell lines producing other FGFR-Fc fusion proteins and variants can also be created in a similar manner described herein, using the pDEF38 expression vector described in Example 2. Example 6: Analysis of In Vitro Cleavage of FGFRi1-Ic-Fe, R1Mut4 and other FGFR1-mIc-Fc Variants During Cell Culture Production [0262] The resistance of FGFRI -IfIc-Fo to in vitro cleavage during transient protein expression was compared to the FGFR 1-Ic-Fe variants R1Mut1, RlMut2, R1Mut3, RlMut4, R1Mut5, RlMut6, RlMut7, RlMut8 and RIMut9. Fusion proteins were each expressed in 293-6E cells via transient transfection using the pTT5 vector as 82 described in Example 2. The supernatants of each transfectant were collected on day four post-transfection and about 5 ul of each was separated with SDS-PAGE in a 4 - 12% acrylamide gel under reducing conditions. The supernatants were from cultures that were matched for cell number, viability, and transfection conditions. The separated proteins were then probed with horseradish-peroxidase conjugated anti-human Fc antibody (anti human Fc HRP; Jackson InmunoResearch Laboratories, Inc.; West Grove, PA). The results are shown in FIG. 3A, which shows the Fc fragment cleaved from the parental FGFR1-Illc-Fc migrating between approximately 28 and 39 kD. Much less Fc product was cleared when the fusion protein was made with the truncation variants RIMuti, RiMut2, RlMut3, R1Mut4, and RIMut5. Similar experiments demonstrated that the variants of FGFR1-IIc-Fo RlMut6, RIMut7, also had less in vito cleavage during transient protein expression than the parental FGFR1-IIc-Fc fusion protein. [0263] To compare the in vitro cleavage of FGFR1-IIc-Fc and RlMut4 produced from stably transfected DG44 host cells, pooled supernatants from cultures of cells producing FGFR1-IIc-Fc and RlMut4 having similar cell viability (82.9% for FGFR1 IIfc-Fc and 79.2% for RlMut4), the same cultivation period (four days), and having similar cell densities (0.95 x 10 6 /ml for FGFRI-UIc-Fc and 0.65 x 106 /ml for RlMut4). The expressed recombinant proteins were separated on a 4 - 12% Bis-Tris PAGE gel (Bio-Rad; Hercules, CA) and subsequently transferred to a nitrocellulose membrane. The intact molecule, as well as the cleaved product of free human Fc fragment, was visualized by Western blot using goat anti-human Fe antibody (Jackson ImmunoResearch Laboratories, Inc.; West Grove, PA). The results are shown in FIG. 3B. The left panel of the Western blot shows 50 ag, 100 ng, and 300 ng of purified FGFRi-IIfc-Fc protein produced from CHO-S cells. The right panel shows a comparison of FGFRl-IIc-Fc and RlMut4 supernatants produced from DG44 cells, revealing the presence of an Fc cleavage product from the FGFR1-IIIc-Fc but little or no Fe cleavage product from'the RIMut4. These results indicate that Rl Mut4 was more resistant to proteolysis and had less product cleaved during production than its parent molecule FGFR1 -DIc-Fc. [0264] The resistance of parental FGFR4-Fc to in vitro cleavage during transient protein expression was compared to the FGFR4-Fe variants R4Mutl, R4Mut2, R4Mut3, R4Mut4, R4Mut5, and R4Mut6 (delta GS). Fusion proteins were each expressed in 293 6E cells via transient transfection using the pTT5 vector and the techniques described in Example 2. The supematants of each transfectant were collected on days seven and eight post-transfection and about 5 ul of each was separated by SDS-PAGE in a 4 - 1.2% 83 acrylamide gel under reducing conditions. The supernatants were obtained from cultures matched for cell number, viability, and transfection conditions. The separated proteins were then probed with horseradish-peroxidase conjugated anti-human Fc antibody (anti human Fc HRP; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA). The results are shown in FIG. 4. The Pc fragment was cleaved from the parental FGFR4-Fc and migrated between approximately 30 and 43 kD. Much less Fc product cleaved when the fusion protein was made with the truncation variants R4Mutt, R4Mut3, and R4Mut4 compared to the parental construct, R4Mut2, R4Mut5, and R4Mut6. Example 7: Purification of FGFRI-fflc-Fe [0265] FGFRI-mc-Fc expressed from recombinant host cells was purified from cell culture supernatant using a combination of Protein-A affinity chromatography and butyl hydrophobic interaction chromatography. The components of the media were separated first on a Protein-A Sepharose column, then on a butyl Sepharose column using a GE Healthcare Akta Purifier 100 (GE Healthcare Bio-Sciences; Piscataway, NJ). The Protein-A Sepharose 4 Fast Flow (GE Healthcare Bio-Sciences; Piscataway, NJ) was used as an affinity matrix to bind to the Fc region of the fusion molecule. The column was equilibrated with ten column volumes of a sterile buffer of 10 mM potassium phosphate, 500 mM NaCl, pH 7.0, then the cell culture supernatant fluid was applied to the column, The column was washed with eight column volumes of sterile 10 mM potassium phosphate, 500 mM NaCl buffer, pH 7.0; then the bound material, including FGFR1-mc Fc, was eluted at a rate of 10 ml/min with a step gradient of the elution buffer (100 mM glycine, 500 mM NaCI, pH 2.7) using sequential steps of two column volumes each of 15%, 30%, 45%, 60%, 75%, and 90% elution buffer, followed by five column volumes of 100% elution buffer. Ten-mi fractions were collected in tubes containing one ml I M Tris pH 7.0 (Ambion; Austin, TX) to neutralize the eluates. Fractions comprising FGFRI-Hlc Fe were identified by gel electrophoresis and pooled. FGFR1-Ilc-Fc was eluted at approximately 30-45% gradient-strength elution buffer. [0266] Pooled Protein-A column eluates comprising the bulk of FGFRI-IIfc-Fc were then subjected to further purification by butyl Sepharose hydrophobic interaction chromatography. Following the addition of an equal volume of 2.4 M ammonium sulfate to the eluate from the Protein-A column, the eluate was applied to a Butyl Sepharose 4 Fast Flow column (GE Healthcare Bio-Sciences; Piscataway, NJ) that had been equilibrated with five column volumes of sterile 10 mM potassium phosphate, 1.2 M 84 ammonium sulfate, pH 7.0. The column was washed with four column volumes of the equilibration buffer and the bound material was eluted at a rate of five ml/min with a linear gradient starting at 100% equilibration buffer and ending at 100% of the elution buffer (10 mM potassium phosphate, 30 mM NaCl, pH 7.0) over a total volume of 13 column volumes followed by an additional five column volumes 100% elution buffer. Fractions (14 ml) were collected and the fractions containing the bulk of FGFR1 -Ic-Fc were identified by gel electrophoresis and pooled, FGFRI-Mc-Fc was eluted with approximately 20-50% elution buffer. [0267] After purification, endotoxin levels were checked by the limulus amoebocyte lysate (LAL) assay (Cambrex; Walkersville, MD). When the values were higher than I endotoxin unit (BU)/mg of FGFRI-IJI-Fc protein, father purification was performed by CellufineTM ETClean chromatography (Chisso Corporation; Tokyo, Japan) following the manufacturer's instructions. FGFR--IJIc-Fc was dialyzed with PBS and applied to a CellufineTM ET Clean column (10 x 0.9 cm (I.D.); 9.6 ml) previously equilibrated with PBS, and the protein was collected in the flow through at a flow rate of 0.5 ml/min. The final FGFRl-I~c-Fc solution (in PBS without Ca 2 /Mg) was then re tested to confinn a value less than or equal to 1 EU/mg of protein as assessed by the LAL assay. [0268] These purification protocols were used to purify other FGFR-Fc fusion proteins and variants, such as FGFR3-Ic-Fc and RlMut4. These purification protocols may also be used to purify other FGFR-Fc fusion proteins and variants, and may be adjusted using methods known in the art to substantially purify FGFR-Fc fusion proteins, for example, other FGFR1-Fc fusion protein variants, FGFR2-Fc fusion proteins and variants, FGFR3-Ic-Fc fusion protein variants, and FGFR4-Fc proteins and variants. For example, components of the cell culture supernatant media may be separated by hydrophobic chromatography either prior to or subsequent to the Protein-A step. Both the Protein-A and hydrophobic chromatography can take place in a column, a slurry, or other similar embodiments. The column size may depend on the amount of FGFR-Fc estimated to be present in the cell culture supernatant, for example, 25 liters of CHO cell supernatant media transfected with FGFRL-Ic-Fc produced about 8 mg/L, or 200 mg of substantially pure FGFR1-UlIc-Fc, using the protocol described above.

85 Example 8: Specificity and Affinity of Ligand Binding to FGFR1-Ifc-Fc, R1Mut4, FGFR3-HIc-Fc, and FGFR4 Measured by Biacore Analysis [0269] The specificity of FGF ligand binding to FGFRl-Ic-Fc, RlMut4, FGFR3 Ilc-Fc, and FGFR4-Fc was assessed using Biacore* Ti00 surface plasmon resonance (SPR) technology (Biacore; Piscataway, NJ). FGFR1-Mc-Fc, R1Mut4, and FGFR4-Fc fusion proteins were produced from CHO-S host cells as described in Examples 2, 4, and 5. FGFR3-IIc-Fc fusion protein was produced from 293-6E host cells as described in Examples 2 and 3. Protein-A was covalently linked to a CMS chip, according to manufacturer's instructions and then a FGFR fusion protein was bound to the chip by the interaction of the Fc domain with the Protein-A. The FGF ligands were placed in contact with the FGFR fusion protein, also according to manufacturer's instructions, in the presence of HBS-P buffer (Biacore; Piscataway, NJ) supplemented with 50 ug/ml heparin (Sigma; St. Louis, MO). [0270] All the recombinant FGF ligands were from R&D Systems (Minneapolis, MN) except for FGF-1 8 which was from Wako Chemicals (Richmond, VA). FGF ligands were each tested at six to eight concentrations ranging from 4.5 ng/mI to 10 ug/ml. The FGF ligands were recombinant and of human origin, except for FGF8b and FGF-18, which were of recombinant mouse origin. [0271) The binding of FGFR1-HIc-Fc, RlMut4, FGFR3-IIlc-Fc, and FGFR4-Fc to various FGF ligands was measured in real time. FIG. 32 shows several representative binding traces from the experiments with FGFR1-IIc-Fc and RlMut4 and Table 3 below shows the resulting association constants (k,), dissociation constants (kd) and equilibrium dissociation constants (KD) that were determined from these studies. [0272] As summarized in Table 8-1, the relative rank of FGF binding affinity to FGFRI-IIc-Fc was FGF-I > FGF-18 > FGF-2, FGF-4 > FGF-9, FGF-20> FGF-5 > FGF 19. The relative rank of FGF binding affinity to R1Mut4 was FGF-1 > FGF-4, FGF-18 > FGF-2> FGF20> FGF-9 > FGF-5 >FGF-19. The relative rank of FGF binding affinity to FGFR3-IIIc-Fc was FGF-18 > FGF-1> FGF-9 > FGF-2, FGF-4 > FGF-20 > FGF-5 > FGF-7 > FGF-19. The relative rank of FGF binding affinity to FGFR4-Fc was FGF-1 > FGF-2. [0273] In another binding study between FGFR4-Fc and the various FGFs, conducted in a similar fashion as described above, the resulting equilibrium dissociation constants (K 0 ) and the relative rank of FGF binding affinity for FGFR4-Fc were: FGF-1 8 (K of 0.4 x 10-9 M)>FGF-17 (KD of 1.0 x 10-M) FGF-20 (KD of 1.2 x 10 9 M)> FGF- 86 8 (K,)of 3.9 x 10-'M) = FGF-4 (KDof 4.6 x 10-'M) > FGF-9 (KDof 9.8 x 10'9M) =FGF 16 (KDof 9.7 x 109M) > FGF-19 (Kof 12.3 x 109M) > FGF-l (KDof 16.3 x. 10 9 m) > FGF-6 (K(Dof 26.2 x 10-9M) > FGF-2 (Koof 44.2 x i0-9M) > FGF-3 (KID of 51.8 x 10- 9 M). FGF-5 showed no binding in this experiment. [0274] The affinity of RlMut4 for Al the ligands tested except FGF-19 was greater than that of the parental FGFRl-Lllc-Fc molecule. In addition, the relative rankings of ligand affinities were also different between RlIWO and the parental FGFRl-Mfc-Fc molecule. Table 3: IRcnl-Tinme Ligand Binding to FG1F1s FGPRI-HkE,- R!Mqt4 Ligand . dK, k 3 dK FGF-I 2.00x IOe M 1.99XIO04M 9.95x10-" M 3.46xW0M* 1.6]xO1 M* 5.07xd0" M* FF-2 3.75x]'M 2.3 IX10 4 M 6.17x 10'*M 4.120'M 1.80xIOA M UW8~O 0 M FGF-4 7.15x10 5 M 4.77x 10 M 6.67x 10-' 0 M I.06XI0 6 M 2.26x 10- M 2. 14d.0w OM FGF-S 1.71xIO 5 M 7.8500-OAM 4.58x 101M 3.7IxlOl M 9.65x 104M 2.95X 10 9 M FGP-7 n.d n.d n.d n.d n.d n.d FGP-9 4.'741O 5 M 5.29xiOr'M 1.I12x IT"M 5.66Xl10'M 5.20x 10 M 9.iqX1O- 1 *M FOP-18 1.1 lxIO M* 4A4IxO Ms 4.18xWO' M* 1.06x10' M* 2.77x 10- M* 2.75x10' 0 M* FGF-19 5.63x10 4 M 4.43x1-' M 7.87x IV M n.M. n.m. n.m. FGF-20 I .62x10 5 M 2.5SX104 M LI .5x10 M 2.42X1 O M I .9OxIO4 M 7.85x 10O" M PGFRO-lIff-Fe FGERA-F Ligand k. d Kle. d Kb FF-I 2.83x10 6 1v 3.31x1' 4 M* I.26x1& 0 OM' 1.68x I0 5 M 5.78x 10- M 3.45x10 9 M FGF-2 3.36x10 5 M 1.37x 10-3 M I .06xI 04 M 4.16x10' M 6.73xIO4 M 1.62x 10- M FGF-4 8.08XI 1 'M 1.55x 1 M 1.9IxIO- 9 M n.m. n.m n.m. FF-S 2.31xI'M 1.69X I0aM 9.7OxlO- 9 M n.d n.d n.d FOF-7 3.)9xIO'M 4.71 x 10 2 M 1.48x]0 7 M n.d n.d n.d FOF-9 7.76xIO'M 4.17x 104 M 53XWO 0 M n.d a-d n.d F-I8 5.l6xIO 6 M* I8OXIO-4M* 3.50XIO 1 M* n.d a.d n.d FGF-19 5.63A O M 4.43xlO-' M 7.87x104M nAd n.d n.d FGF-20 I.85xIO 5 M 4.04x IV M 2.17x10'M n.d n.d n.d * average of two independent measurements 87 n.d.= not determined n.m. =not measurable Example 9: Specificity and Affinity of Ligand Binding to FGFR4-Fc and FGFR4-Fc Deletion Mutants Measured by Competition ELISA [0275] FGFR4-Fc fusion protein and deletion variants, made as described in Examples 1, 2, and 3, were tested for their ability to sequester the soluble FGF ligands FGF-1, FGF-2, and FGF-8b, and to inhibit ligand binding to FGFR4- Fe fusion protein coated on a plate. [0276] Briefly, I BIND half-wells were coated with FGFR4-Fc of CHO-S-origin at a concentration of 5 ug/mI in PBS in a volume of 25 ul per well for 1 hr at room temperature. The wells were blocked by adding 150 ul BLOTTO per well and incubating for 2 hr at room temperature. The coated half-well plates were then washed six times with PBS and 0.05% Tween-20 to remove unbound FGFR1-IIc-Fc and BLOTTO. [0277] Varying amounts of FGFR4-Fc fusion protein and the deletion variants, produced from CHO-S cells, or 10 ug/ml of the negative control human IgG (Caltag; Burlingame, CA) were each first pre-incubated in 96-well U-bottom plates with 60 ng/ml recombinant human FGF-1 (from R&D Systems; Minneapolis, MN) in 50 ul for 30 min at 370 C on a shaker in the presence of 20 ug/ml heparin in 0.1X BLOTTO in PBS. About 40 ul of the above fusion proteins pre-incubated with FGF-1 were then added to the washed half-well plates coated with FGFR4-Fc and incubated for 30 min at 37* C with shaking. After incubation, the plates were washed as before six times with PBS and 0.05% Twcen-20 to remove any unbound FGF-1. After washing, about 2 ug/mI of anti human FGF-1 polyclonal biotinylated antibody (R&D Systems; Minneapolis, MN) in 1x BLOTTO was added to each well of the plate, which was then incubated for 30 min at 37* C with shaking, followed by washing as before to remove any unbound anti-FGF- I antibody. The bound anti-FGF-1 antibody was detected using a streptavidin-HRP linker provided in the ABC kit (Vector Laboratories; Burlingame, CA) according to manufacturer's protocol. After washing as before, reconstituted (OPD) solution (Sigma; St. Louis, MO) was added. The detection reaction proceeded for 10 to 20 min at room temperature and was followed by a reading of the absorbance at 450 nm. The binding curves from the competition ELISA and the resulting EC 5 0 values are shown in FIG. 5A. [0278] FIG. 5A showed that the FGFR4 deletion variants R4Mutl, R4Mut2, R4Mut3, R4Mut4, R4Mut5, and R4Mut6 had higher affinities for FGF-1 than did the 88 parental FGFR4-Fc. The FGFR4-Fc deletion variants had EC50 values of about 0.033 ug/m1 to about 0.057 ug/ml. In contrast, the parental FGFR4-Fc had an EC 5 0 value of about 0.123 ug/ml. The human IgGI negative control did not inhibit FGF-1 binding to the FGFR4-Fo coated on the plate. [0279] Similar competition ELISA experiments were conducted comparing the ability of the FGFR4- Fc fusion protein and the FGFR4- Fe deletion variants, produced in CHO-S host cells, to inhibit the binding of recombinant human FGF-2 (used at 200 ug/ml) and recombinant mouse FGF-8b (used at 200 ng/ml) (all from R&D Systems; Minneapolis, MN) to FGFR4-Fe derived from CHO-S cells and innobilized on an assay plate. [0280] FIG. 5B showed that the FGFR4 deletion variants R4Mut1, R4Mut2, R4Mut3, R4Mut4, and R4Mut5 had higher affinities for FGF-2 than did the parental FGFR4-Fc. The FGFR4-Fc deletion variants R4Mutl, R4Mut2, R4Mut3, R4Mut4, and R4Mut5 had EC 50 values from about 0.091 ug/mI to about 0.457 ug/ml, with R4Mut3 having the highest affinity for FGF-2; the ECSo was about 0.091 ug/ml. In contrast, the parental FGFR4-Fc had an EC 50 value of greater than 10 ug/ml, as did the deletion-variant R4Mut6. The human IgGI negative control did not inhibit FGF-1 binding to plate immobilized PGFR4-Fc. [0281) FIG. 5C showed that the FGFR4 deletion variants R4Mutl, R4Mut2, R4Mut3, R4Mut4, R4MutS, and R4Mut6 had higher affimities for FGF-8b than did the parental FGFR4-Fc. The FGFR4-Fc deletion variants had EC 5 0 values of about 0.137 ug/mi to about 0.209 ug/ml. In contrast, the parental FGFR4-Fo had an EC 5 0 value of about 0.631 ug/m. The human IgG1 negative control did not inhibit FGF-1 binding to the FGFR4-Fc coated on the plate. [0282] These experiments demonstrated that the FGFR4 deletion mutants R4Mutl, R4Mut2, R4Mut3, R4Mut4 and R4Mut5 had a higher affinity than the parental FGFR4-Fe in their ability to inhibit FGF-1 (as shown in FIG. 5A), FGF-2 (as shown in FIG. SB), and FGF-8b (as shown in FIG. 5C) binding to plate-immobilized FGFR4-Fc. Example 10: Affinity of Ligand Binding to FGFRI-Ic-Fe Deletion Mutants Measured by Direct ELISA [0283] The R1Mut1, R1Mut2, R1Mut3, RlMut4, and R1Mut5 fusion proteins were compared to parental FGFR1-IlIc-Fc fusion protein (all produced from 293-6E host cells as described in Example 3) for their ability to bind FGF-2 by a direct FGF-2 binding 89 ELISA assay. Briefly, FGF-2 (R&D Systems; Minneapolis, MN) was used to coat half well HI BIND wells (Becton Dickinson; Franklin Lakes, NJ) by diluting FGF-2 in PBS at a concentration of 5 ug/ml in 25 ul volume per well and incubating for 1 hr at room temperature while shaking. The wells were then blocked by adding 150 ul of BLOTTO (Pierce Biotechnology; Rockford, IL) to each well and incubating for 1 hr at room temperature. The plates were then washed six times with PBS comprising 0.05% Tween 20 to remove the FGF-2 and BLOTTO and the wells were then incubated overnight at 4* C with varying concentrations of FGFRI-LEc-Fc, RlMutl, R1Mut2, RlMut3, RlMut4, RIMut5, or human IgG (as a negative control), in the presence of 10 ug/mnl heparin diluted in 0. IX BLOTTO in PBS. The plates were washed as before, and then incubated with 25 ul of anti-human Fc antibody conjugated to HRP at 2.5 ug/ml in BLOTTO for 1 br at room temperature on a plate shaker. The BLOTTO was then removed and the plates were again washed as before. After washing, the wells were incubated with reconstituted OPD solution (Sigma; Saint Louis, MO) for 10 to 20 min at room temperature and the absorbance at 450 nn was measured. [0284] As shown in FIG. 6, RiMutI, RlMut2, RlMut3, and R1Mut4, but not R1Mut5, were each able to bind to FGF-2 as well as or better than the parental FGFR1 inc-Fc, with R1Mut4 having the highest apparent affinity of all the fusion proteins tested in this experiment. The MMP-2 cleavage site mutants R1Mut7, RlMut8, and Rl Mut9 (all produced in 293-6E host cells using the expression vector pTT5 as described in Example 2) also bound to FGF-2 with an affinity similar to that of parental FGFR1-IHc-Fc. Example 11: FGFR Fusion Proteins Inhibit FGFR1-Ic-Fc Ligand Binding [0285) FGFR1-IIIc-Fc fusion proteins produced from 293-61 and CHO cells, made as described in Examples 1, 2, and 3, were tested in a competition ELISA assay for their ability to sequester the soluble FGF ligands FGF-1, FGF-2, and FGF-8b, and to inhibit ligand binding to FGFRI-fIc-Fc fusion protein coated on a plate. [0286] Briefly, HI BIND half-wells were coated with FGFR1-IIIc-Fc of 293-6E origin at a concentration of 5 ug/ml in PBS in a volume of 25 ul per well for 1 hr at room temperature. The wells were blocked by adding 150 ul BLOTTO per well and incubating for 2 hr at room temperature. The coated half-well plates were then washed six times with PBS and 0.05% Tween-20 to remove unbound FGFR1-BIc-Fc and BLOTTO. [0287] Varying amounts of FGFRI -IIc-Fe fusion proteins produced from 293-6E cells or CHO cells, or 10 ug/ml of the negative control human IgG (Caltag; Burlingame, 90 CA) were each first pre-incubated in 96-well U-bottom plates with 200 ng/ml recombinant human FGF-2 (from R&D Systems) in 50 ul for 30 min at 370 C on a shaker in the presence of 20 ug/ml heparin in 0.1X BLOFO in PBS. About 40 ul of the above fusion proteins pre-incubated with FGF-2 were then added to the washed half-well plates coated with FGFR1-II1c-Fc and incubated for 30 min at 370 C with shaking. After incubation, the plates were washed as before six times with P3S and 0.05% Tween-20 to remove any unbound FGF-2. After washing, about 2 ug/ml of anti-human FGF-2 polyclonal biotinylated antibody (from R&D Systems) in Ix BLOTTO was added to each well of the plate, which was then incubated for 30 min at 370 C with shaking, followed by washing as before to remove any unbound anti-FGF-2 antibody. The bound anti-FGF-2 antibody was detected using a streptavidin--HRP linker provided in the ABC kit (Vector Laboratories, Burlingame, CA) according to the manufacturer's protocol. After washing as before, reconstituted OPD solution (Sigma; St. Louis, MO) was added. The detection reaction was developed for 10 to 20 min at room temperature followed by a reading of absorbance at 450 rn. Results are shown in FIG. 7. [0288] FIG. 7 showed that FGFR1-mc-Fc produced from 293-6E and from CHO cells had approximately equivalent binding potencies in their ability to sequester FGF-2 in an FGF-2 competition assay. Each of the two fusion proteins exhibited an EC 50 value of about 0.24 ug/ml. In contrast, the human IgGI negative control did not inhibit FGF-2 binding to the FGFR1 -eIc-Fo coated on the plate. [0289] Similar competition ELISA experiments were conducted comparing the ability of the FGFRl-IIIc-Fc fusion protein and RIMut4 fusion protein, both produced in DG44 host cells, to inhibit the binding of recombinant human FGF- 1 (at a concentration of 60 ng/ml), recombinant human FGF-2 (a concentration of at 200 ng/ml), and recombinant mouse FGF-8b (a concentration of at 200 ng/ml) (all from R&D Systems; Minneapolis, MN) to FGFR1 -IIc-Fc derived from 293 cells and immobilized on an assay plate. These experiments all demonstrated the equivalency of R1Mut4 and the parental FGFR1-IlIc-Fc fusion proteins in their ability to inhibit FGF-1 (as shown in FIG. 8A), FGF-2 (as shown in FIG. 8B), and FGF-8b (as shown in FIG. 9) binding to plate-immobilized FGFR1-Ilc-Fc. [0290] FGFR 1 -IIc-Fc produced in DG44 host cells and FGFR3-IIc-Fc and FGFR4-Fc produced in 293-6E host cells also inhibited ligand binding to FGFR1-IIlc-Fc produced in 293 cells and immobilized on an assay plate. A competition ELISA assay conducted as described above used recombinant human FGF-1, recombinant human FGF 2, and recombinant mouse FGF-8b (all from R&D Systems; Minneapolis, MN). Human 91 IgG was used as a negative control. The results are shown in FIG. 10, FIG. 11, and FIG. 12, which demonstrate both the effectiveness of the decoy fusion proteins in blocking ligand-receptor binding and the specificity of the fusion proteins for their respective ligands. [0291] FIG. 10 showed that FGFRI-IfIc-Fc, FGFR3-IIlc-Fc, and FGFR4-Fc all inhibited FGF- 1 binding to FGFRI-IIlc-Fc immobilized on an assay plate. FIG. 11 showed that FGFR3-Iflc-Fc and FGFR4-Fc were much less effective than FGFRI -MIc-Fc in inhibiting FGF-2 binding to FGFR1-IIfc-Fc immobilized on an assay plate. FIG. 12 showed that FGFR1-lIIc-Fc, FGFR3-Ic-Fc, and FGFR4-Fc were all similarly effective in inhibiting FGF-8 binding to FGFR1-IIlc-Fc immobilized on a plate. Example 12: FGFR1-IIIc-Fc Inhibited Phospho-Erk Signaling by FGF-2 [0292] FGFR1-IIlc-Fc fusion protein derived from 293-6E or CHO cells were approximately equally potent in inhibiting biological signaling by FGF-2. L6 cells transfected with FGFRI -IUc (ETH; Zurich, Switzerland) growing in a T-175 flask were trypsinized, washed and seeded at a concentration of 10,000 cells/well in a volume of 100 ul in Dulbecco's modified Eagle's medium (DMEM) supplemented with 0.5% FCS and 0.1% bovine serum albumin (BSA) in 96-well flat-bottom plates for 16 hr. Activation medium containing FGFRI-Iflc-Fc fusion proteins or human IgG1, at concentrations from 0.005 to 10 ug/ml, was prepared (in 0.1% BSA DMEM containing 100 ng/ml FGF-2, 10 ug/ml heparin) and incubated for 30 min at 37*C on a plate shaker. The L6 cells were then exposed to 25 ul of activation medium/well for 5 min at 37*C. The cells were then washed once with 200 ul of ice-cold PBS and lysed with 100 ul of ice-cold 1X lysis buffer for 30 min on ice following the manufacture's recommendations for the PathScan Phospho-p44/42 MAPK (T202/Y204) Sandwich ELISA Kit (Cell Signaling; Danvers, MA). At the end of the lysis period, the lysates were pipeted up and down approximately five times while minimizing foaming. About 80 ul of Sample Diluent (from Pathscan Sandwich ELISA kit) was added to each well of the phospho-ERK ELISA plate, then topped by 80 ul of cell lysate and mixed. The plate was sealed with plastic adhesive, incubated for 2 hr at 37 0 C, and washed six times with PBS containing 0.05% Tween-20. Then 100 ul of phospho-ERK Detection Antibody (from Pathscan Sandwich ELISA kit) was added to each well. The plate was sealed with adhesive cover, incubated for 1 hr at 37*C, washed as before and 100 ul of HRP-linked secondary antibody (from Pathscan Sandwich ELISA kit) was added to each well. The plate was again sealed, incubated for 92 30 min at 37 0 C, and then washed as before. About 100 ul of TMB substrate (3, 3', 5, 5' tetramethylbenzidine, from Pathscan Sandwich ELISA kit) was then added to each well and the plate was incubated for 30 min at 25*C. The color development was completed by adding 100 ul of STOP Solution (from Pathscan Sandwich ELISA kit) to each well and mixing. The absorbance at 450 am was recorded and plotted as shown in FIG. 13. [0293] FIG. 13 shows that Erk phosphorylation, as determined by ELISA, was similarly inhibited by FGFR1-IlIc-Fc made from either 293-6E or CHO cells. At the lower doses, FGFR1-IlIc-Fc blunted Erk activation and at higher doses, prevented activation by FGF-2. In the presence of 100 ng/ml FGF-2, the ECso values of the 293 cells-derived and the CHO cells-derived FGFR1-IIlc-Fc were 0.23 ug/ml and 0.29 ug/ml, respectively. Human IgG1 did not inhibit Erk phosphorylation activated with FGF-2. [0294] FGFRI-IIc-Fc and RIMut4 fusion proteins produced by 293-6E cells inhibited Erk phosphorylation with approximately the same potency, in contrast to RlMut5, which did not inhibit Erk phosphorylation. In the presence of 0.10 ug/mI FGF-2, the ECso of FGFR1-IIIc-Fc was 0.18 ug/ml and the ECso of RlMut4 was 0.29 ug/ml. At low doses, both FGFR1-IIc-Fc and RlMut4 blunted Erk activation; and at high doses, both prevented activation. Results are shown in FIG. 14. Example 13: FGFR1-IIIc-Fc Decreased Cancer Cell Viability and Proliferation [0295] FGFR-Fc fusion proteins of the invention decreased the viability and/or proliferation of cancer cells in culture, as measured with a Ce1lTiter-GloTM Luminescent Cell Viability Assay according to the manufacturer's instructions (Promega; Madison, WI). In this assay, luminescence quantitatively correlates with the number of viable cells. [02961 The effects of FGFR1-IIc-Fc on U251 malignant glioblastoma brain cancer cells obtained from the American Type Culture Collection (ATCC) (Manassas, VA) are shown in FIGS. 15-17. The effect of FGFR1-IIIc-Fc on cell viability and proliferation was dependent on the concentration of FGFR1-lIIc-Fc and on the growth conditions of the cells. The negative control human IgG (20 ug/ml) had no effect. The positive control, TRAIL, decreased viability and proliferation. FGF-2 (100 ug/rl) stimulated proliferation and differentiation. At the lower cell concentrations tested, FGF-2 did not blunt the inhibition induced by 20 pg/ml FGFRl-IIIc-Fc. [0297] U251 cells were grown in DMEM with 4 mM L-glutamine adjusted to contain 1.5 grams per liter (g/L) sodium bicarbonate and 4.5 g/L glucose, 10% heat- 93 inactivated FCS with 100 units/ml penicillin and 100 ug/lbn streptomycin (pen-strep, Invitrogen; Carlsbad, CA) in T-150 flasks until they reached 70% to 90% confluency. The cells were treated with 10 ml per flask of 0.25% trypsin solution in Hanks' Balanced Salt solution (Invitrogen; Carlsbad, CA) at room temperature for 3 min at 37* C and the trypsin-cell suspension mixed with 40 nl of ice-cold 0.1% FCS in DMEM. The cells were pelleted at 900 x g for 5 min at room temperature. This wash step was repeated with 50 ml of ice-cold 0.1% FCS in DMEM and the cells resuspended in 5 ml of ice-cold 0.1% FCS in DMEM. [02981 The resuspended U251 cells were plated in a volume of 150 ul per well in 96-well flat-bottom tissue-culture grade plastic plates (Nunc; Rochester, NY) in the presence of 20 ug/ml of porcine intestinal mucosa heparin (Sigma; St. Louis, MO). The cells were plated at three culture conditions (1) a concentration of 1000 cells per well in the presence of 10% FCS in DMEM with pen-strep; (2) a concentration of 5000 cells per well in the presence of 1.0% FCS in DMEM with pen-strep; (3) a concentration of 10,000 cclls per well in the presence of 0.1% FCS in DMEM with pen-strep. The cells were . treated with FGFR1-lIc-Fc protein or a control protein in four replicate wells per protein and incubated for five days in a humidified incubator at 37*C with 5% CO 2 . The FGFRI fIc-Fc protein was made from 293-6E cells and substantially purified as described in Examples 2 and 7. The cells were treated with FGFRI -lIlc-Fc at a concentration of 20 ug/ml, 4 ug/mI, or 0.8 ug/ml. Control proteins included 20 ug/ml purified human IgG dialyzed against PBS to remove preservative and then filter-sterilized (Caltag; Burlingame, CA) as a negative control, 100 ng/ml FGF-2 (R&D Systems; Minneapolis, MN), used either alone or in combination with 20 ug/ml FGFR1-IIc-Fc as a positive control; and 10 ng/nl TRAIL (APO2 ligand / tumor necrosis factor-related apoptosis inducing ligand) (R&D Systems; Minneapolis, MN), used as a positive control. [0299] Cell viability was then determined using the CellTiter-Glot" Luminescent Cell Viability Assay Kit (Promega; Madison, WI), according to the manufacturers instructions. Briefly, 100 ul/well of reconstituted CellTiter-GloTM reagent was added to the cells and incubated for 10 min in the dark. The well contents were mixed by pipeting and 100 ul from each well were transferred to opaque, white 96-well plates (Corning; Acton, MA). The luminescent output from each well was read using a 0.6 second per well recording time and the average relative huminescence units (RLU) of the four replicates were plotted along with their standard deviations.

94 [0300] As shown in FIG. 15, FGFR1-IIc-Fc at each of the three concentrations tested reduced the viability and proliferation of cultured U251 malignant glioblastoma cells plated at a concentration of 1000 cells per well in 10% FCS. Untreated cells showed an RLU of about 480. Cells treated with 20 ug/nl, 4 ug/ml, and 0.8 ug/ml FGFR1-Ic-Fc showed an RLU of about 380,400, and 400, respectively. Human IgG did not inhibit viability and proliferation, showing an RLU of about 500. FGF-2 alone enhanced viability and proliferation, showing an RLU of about 600. The combination of FGF-2 and 20 ug/mI FGFRI -Ic-Fc reduced cell viability and proliferation, showing an RLU of about 400. Treatment with TRAIL resulted in almost complete inhibition, with an RLU of about 10. [0301] As shown in FIG. 16, FGFRl-IIc-Fc at each of the three concentrations tested reduced the viability and proliferation of cultured J251 malignant glioblastoma cells plated at a concentration of 5000 cells per well in 1.0% FCS. Untreated cells showed a RLU of about 360. Cells treated with 20 ug/ml, 4 ug/ml, and 0.8 ug/ml FGFRl-IlIc-Fc showed an RLU of about 60, 140, and 200, respectively. Human IgG did not inhibit viability and proliferation, showing an RLU of about 400. FGF-2 alone showed an RLU of about 320. The combination of FGF-2 and 20 ug/ml FGFR1-IIIc-Fc reduced cell viability and proliferation, showing an RLU of about 90. Treatment with TRAIL resulted in almost complete inhibition. [0302] As shown in FIG. 17, FGFR1-Ilc-Fc reduced the viability and proliferation of cultured U251 malignant glioblastoma cells plated at a concentration of 10,000 cells per well in 0.1% FCS. These growth conditions generated more variability between wells, but FIG. 17 demonstrates that FGFR1-Mc-Fc inhibited the viability and proliferation of U251 cells in a fashion similar to that described in FIG. 15 and FIG. 16. [0303] The effect of FGFR1 -mc-Fc on the viability and proliferation of cancer cells from cancer cell lines representing various solid tumor types obtained from ATCC (Manassas, VA) or NCI (Bethesda, MD) was tested using the Cel lTiter-GloTMI Luminescent Cell Viability Assay. The cells were grown to about 70% to 90% confluency in T-150 flasks using the recommended growth media for each cell line. The cells were harvested and treated in a manner similar to that described above for U251 cells. The cells were plated at a density of 1000 cells per well in the presence of 10% FCS in DMBM, 5000 cells per well in the presence of 1.0% FCS in DMEM, or 10,000 cells per well in the presence of 0.1% FCS in DMEM; with 20 ug/ml FGFR1-IIc-Fc as the test agent or 20 ug/ml human IgG as the negative control.

95 [0304] The cancer cell lines tested included MDA-MB-435 (breast), MCF7 (breast), MDA-MB-231 (breast), T47D (breast), A549 (lung), NCI-H522 (lung), NCI H460 (lung), NCI-H23 (lung), NCI-H226 (lung), Ul 18 (brain), U87114 (brain), U251 (brain), SF268 (brain), WT11 (brain), DU145 (prostate), PC-3 (prostate), COLO 205 (colon), Caki-1 (kidney), SK-M[EL-2 (skin) and SK-OV-3 (ovary). The results are shown in FIG. 18. Of the 20 cancer cell lines tested, eight were susceptible to inhibition by FGFR1-LIc-Fc under one or more of the growth conditions tested. The eight susceptible cell lines were A549 (lung), NCI-H522 (lmg), NCI-H226 (lung), UI 18 (brain), U251 (brain), SF268 (brain), WTI I (brain), and Caki- I (kidney). [0305) Both FGFR1-IIlc-Fc and FGFR4-Fc inhibited the viability and proliferation of A549 lung carcinoma cells, as measured by the CellTiter-GloTM Luminescent Cell Viability Assay. The FGFR1 -lIc-Fc and FGFR4-Fc proteins tested were produced in 293-6B cells via transient transfection and purified as described in Examples 2 and 7. The effect of FGFR1-HIlc-FC and FGFR4-Fc on A549 cells was tested using a protocol similar to that described above for U251 cells. The A549 cells were seeded in four replicate wells at a density of 25,000 cells per well in a volume of 150 ul in flat-bottom 96-well plates in the presence of 20 ug/ml porcine intestinal mucosa heparin (Sigma; Saint Louis, MO) in 0.1% FCS DMEM with pen-strep. The A549 cells were cultured in the presence of concentrations of FGFRI-IIlc-Fc protein or FGFR4-Fc protein ranging from about 0.0000095 ug/ml to about 10 ug/ml in four-fold serial dilutions for five days at 37*C in 5% CO 2 . Human IgG (10 ug/mI) was used as a negative control. As shown in FIG. 19, cell viability and proliferation were expressed as percent inhibition (average RLU for untreated - average RLU for sample)/(average RLU for untreated) times 100. The error bars show the % Error (standard deviation of sample/average sample RLU) times 100. [0306] At the higher concentrations tested, FGFR1-lIc-Fc inhibited the viability and proliferation of A549 cells up to about 40%, as compared to the IgG control. The IC 50 of FGFR1-Illc-Fc was about 9.4 ng/ml, equivalent to 0.11 nanomolar (nM). At the highest concentration tested, FGFR4-Fc also inhibited the viability and proliferation of A549 cells up to about 40%, as compared to the IgG control. The ICso of FGFR4-Fc was about 100 ng/ml, equivalent to 1.2 nM; this is higher than that observed for FGFR1-IIc Fc, reflecting a greater potency of FGFR1-IIlc-Fc than FGFR4-Fc in inhibiting A549 cell viability and proliferation.

96 Example 14: FGFR-Fc Fusion Proteins Decreased Cancer Cell Viability and Proliferation (0307] FGFRI-IIb-Fc, FGFR1-UEc-Fc, FGFR2-Ufb-Fc, FGFR2-ec-Fc, FGFR3 IlIb-Fec, FGFR3-lIc-Fe, and FGFR4-Fc were tested for the ability to inhibit the viability and/or proliferation of six different cancer cell lines. The seven fusion proteins tested in this assay were obtained from a commercial source (R&D Systems). The cancer cell lines included A549 (lung), UI 18 (brain), U251 (brain), SF268 (brain), T47D (breast) and Caki I (kidney). (0308] The results obtained from the assays described in Example 13 provided the basis for determining the number of cancer cells plated per well and the concentration of FCS in their media. For instance, the A549 lung carcinoma cells were plated at 25,000 cells per well in 150 ul of DMEM with 0.1% FCS and pen-strep in a 96-well format. They were treated with two-fold serial dilutions of FGFR1 -me-Fc, FGFR2-IUc-Fe, FGFR3 IIc-Fc, or FGFR4-Fc ranging from about 0.078125 ug/ml to about 5.0 ug/ml. Human IgG was used as a positive control, as described in Example 11. Each fusion protein treatment was performed in triplicate wells and each data point represents an average of three wells. After five days of treatment, the cell viability of A549 cells was assayed with the CellTiter-GloTM Luminescent Cell Viability Assay as described above. The results are shown in FIG. 20 and indicated that the FGFR-Fc fusion protein inhibition of A549 cells was dose-dependent, reaching up to about 42% at the highest concentrations tested (5 ag/ml). The potencies of the fusion proteins were ranked as FGFR1-Mc-Fc = FGFR2 Me-Fe > FGFR3-Mc-Fc = FGFR1-Ib-Fc > FGFR2-JIb-Fc = FGFR4-Fc > FGFR3-Mllb Fe > human IgG. [0309] Similar experiments testing the effects of FGFR-Fc fusion proteins on viability and proliferation were performed with on the cancer cell lines.U1 18 (FIG. 2013), U251 (FIG. 20C), SF268 (FIG. 20D), T47D (FIG. 20B) and Caki- 1 (FIG. 20F). The protocols were similar to those described above, including the use of human Igo as a negative control. All of the cancer cell lines tested were inhibited by one or more of the seven FGFR-Fc fusion proteins. The results are summarized in FIG. 21. [0310) The FGFRI-flc-Fc mutant RlMut4, but not RIMut5, inhibited the viability and proliferation of the cancer cell lines A549 and U251 (FIG. 20G). The assay protocol was similar to that described above. The cells were treated with three-fold serial dilutions of FGFRI -IIc-Fe, R1Mut4, R1Mut5, or human IgG at concentrations ranging from about 0.00457 ug/mi to about 10 ug/ml. FGFRI-mc-Fe, R1Mut4, and R1Mut5 were 97 expressed in 293-6E cells using the pTTS vector as described in Example 2 and purified as described in Example 7. Each fusion protein treatment was performed in triplicate wells. After five days of treatment, the viability of the A549 and U251 cells was assayed with CeIlTiter-GloTM Luminescent Cell Viability Assay. Each data point represents an average of three wells. The results are shown in FIG. 20G. The A549 cells were inhibited to a similar extent (from approximately 40 RLU to 20 RLU) by both FGFR1 -mc-Fc and R1Mut4 at all the doses tested. The U251 cells were also inhibited to a similar extent by FGFRI-Ilc-Fc and byRlMut4. They displayed dose dependence at the concentrations tested, with no inhibition observed at the lowest dose and maximal inhibition (from approximately 100 RLU to 30 RLU) at the highest dose. Example 15: Sustained Expression of FGFR1-IIc -- c in Mice In Vivo [0311] The effect of sustained expression of human FGFR1-IlIc-Fc fusion protein in animal models, for example, animal tumor models, was tested using the hydrodynamic tail vein injection method to express FGFR1-IIlc-Fc in mice. Naked "mini-circle" vector cDNA encoding the FGFR1-IIIc-Fc fusion protein was injected into three-month old C57/B16 female mice (Charles River Laboratory; Hollister, CA). This "mini-circle" vector contained FGFRI Mc-Fc cDNA, and was generated as described in Examplc 2. The animals were injected via their tail veins using the hydrodynamic tail vein injection method as reported in Liu, F. et al., Gene Therapy 6:1258 - 1266 (1999) and U.S. Patent No. 6,627,616, at a DNA concentration of about 15 ug/ml in saline. About 2 ml of the DNA composition was injected in 5 - 8 seconds into each mouse. Three mice were injected with mini-circle DNA containing the FGFRI-IlIc-Fc cDNA and three mice were injected with saline as controls. Serum samples with a volume of about 50 ul were obtained from tail vein nicks on days 2, 9, 16, 24, 30, 37, and 44 post-injection. The concentration of the FGFR1-]Ic-Fc protein in the serum samples was analyzed by direct ELISA and the ligand binding activity of the FGFR1-Hic-Fc in the mouse sera was analyzed by FGF-2 competition ELISA. Both ELISA methods are described in further detail below. [0312] A direct sandwich ELISA to detect FGFR1-IIlc-Fc was developed and used to detect FGFRI-IIIc-Fc in the serum of the injected mice. Briefly, HI-BIND half-well plates (Corning; Acton, MA) were coated with anti-human FGFRI antibody (QED Bioscience, San Diego, CA) diluted in PBS to a concentration of 3 ug/mI for 1 hr at room temperature or overnight at 40C; the wells were then blocked with blocking buffer 98 (BLOTTO diluted to 3% in PBS) for 2 - 5 hr at room temperature. The plates were washed with PBS containing 0.05% Tween-20, and 50 ul of mouse serum from each of the test animals diluted in 0.6X BLOTTO was added, respectively, to each well and incubated for 2 hr at room temperature. The plates were then washed as before and incubated with 50 ul/well peroxidase-conjugated AffiPure goat anti-human Fe antibody (Jackson Immuno-Research Laboratories; West Grove, PA) diluted 1:3000 in Blocking Buffer for 60 min at room temperature. The plates were washed as before, and the wells were incubated for 10 to 20 min with reconstituted OPD solution (Sigma; Saint Louis, MO) at room temperature and the absorbance at 450 nm determined. [0313] The results are shown in FIG. 22, and demonstrate the amount of FGFRI IlIc-Fc fusion protein in the sera of each of the three injected mice in the days following injection of the cDNA encoding the fusion protein. The saline-injected mice showed no detectable levels of the fusion protein. Expression of FGFRI-Ic-Fc in the serum remained high for at least 44 days post-transfection in the highest expresser mouse. FGFR1-1Ilc-Fc was detected at about 10 ug/ml on day 2, about 100 ug/ml on day 9, about 80 ug/ml on day 16, about 50 ug/ml on day 24, and about 35 ug/ml on days 30 to 44 in this mouse. The other two mice injected with FGFR1-ilic-Fc cDNA showed lower but detectable levels of the fusion protein. This study demonstrated that high and sustained expression of a human FGFRI-IIc-Fc fusion protein could be achieved in mice after hydrodynamic tail vein injection of the cDNA and that these animals could be used to monitor treatment with this fusion protein. [0314] An FGF-2 competition BLISA demonstrated that the FGFRI-Ic-Fc fusion proteins expressed in these animals was functional. FGFR1-1Uc-Fc in the sera of the above-described eDNA-injected animals was capable of binding and sequestering a known ligand (for example, FGF-2). Briefly, serum from mice injected with FGFRI -IIc-Fe cDNA was pre-treated with FGF-2 and the amount of free FGF-2 remaining in the pre treated serum measured the ability of the expressed FGFR1-TIc-Fc fusion protein to bind its ligand. The amount of free FGF-2 was measured by the ability of the pre-treated serum to bind to FGFR1 -IIc immobilized on an assay plate. [0315] The serum from the injected mice was pre-treated with FGF-2. Briefly, the serum was diluted 1/500, 1/100, and 1/20 with 0.lx BLOTTO (diluted 1:10 in PBS) and added to 96-well U-bottom plates (Nune; Rochester, NY) with 200 ng/ml recombinant human FGF-2 (R&D Systems; Minneapolis, MN) in a volume of 50 ul for 30 min at 37* C on a shaker, in the presence of 20 ug/ml heparin. The pre-treatment of the serum with 99 FGF-2 sequestered the FGF-2 to the extent that the FGFRI-IIIc-Fc in the serum was able to bind its ligand FGF-2. [0316] The pretreated serum was then incubated with assay plates coated with FGFR1-Ifc-Fc and the binding of free FGF-2 in the serum measured. A high level of free FGF-2 binding indicates that the circulating FGFRI-flc-Fc was not able to bind its ligand FGF-2 and a low level of free FGF-2 binding indicates that the FGFRI-Mc-Fc expressed in the serum of the injected mice functioned to bind its ligand FGF-2. [03171 111-BIND half-wells (Coming; Acton, MA) were coated with FGFR1-IIIc Fe of 293-6E host cell origin, at a concentration of 5 ug/ml in PBS in a volume of 25 ul per we1l for 1 hr at room temperature. The wells were blocked by adding 150 ul BLOTTO per well for 2 hr at room temperature, The coated half-well plates were then washed with PBS containing 0.05% Tween-20. The washed, coated, half-well plates were then incubated with 40 ul of the pre-treated serum for 30 min at 370 C with shaking. The plates were washed as before with PBS containing 0.05% Tween-20. Then 2 ug/ml of biotinylated anti-FGF-2 polyclonal antibody (R&D Systems; Minneapolis, MN) in BLOTTO was added to each well and incubated for 30 min at 37"C with shaking. The plates were washed again as before and bound antibody was detected with the ABC kit according to manufacturer's protocol. After the final wash with PBS containing 0.05% Tween-20, reconstituted OPD solution (Sigma; Saint Louis, MO) was added to each well and incubated for 10 -20 min at room temperature and the absorbance of the wells at 450 nm determined. [0318] Results from two control mice injected with saline and two experimental mice injected with FGFRI-IIIc-Fc cDNA are shown in FIG. 23. Pretreated sera from control mouse 1 and mouse 2 showed little or no inhibition of FGF-2 binding to the FGFRl-Fc coated plates. Pretreatcd sera from mouse 3 and mouse 4, which expressed FGFR1-lc-Fc, sequestered FGF-2 in a dose-dependent manner, with the highest level of inhibition observed with the most concentrated sera (1/20 dilution). FIG. 23 also shows the standard curve of puified FGFRI-IIc-Fc used to calculate the amount of circulating FGFR1-IIIc-Fc in the injected mice. The serum of experimental mouse 4 had the functional equivalent of 64 ug/ml of FGFR1-c-Fo and the serum of experimental mouse 3 had the functional equivalent of 45 ug/mI serumn FGFR1-Mc-Fc.

100 Example 16: In Vivo Expression of RlMut4 Via Hydrodynamic Transfection [0319] An experiment similar to that described in Example 15 was performed by hydrodynamic transfection of RlMut4 cDNA into mice using the minicircle vector described in Example 2. Naked "mini-circle" vector cDNA encoding RlMut4 was injected into four month old CB 17 SCID mice (Charles River Laboratory; Hollister, CA). About 2 ml of the DNA was injected at a concentration of 20 tg/ml in 5 to 8 seconds into each of four control mice and four RlMut4 experimental mice. Serum samples were collected at day 2 and day 7. The concentration of RlMut4 in the serum samples was analyzed by direct ELISA (see Example 15), FGF-2 competition ELISA (see Example 15) and Western blot probing. [0320) Results from the direct ELISA test are shown in FIG. 24. M1 - M4 represent sera from the four experimental R1Mut4-injected mice. Mouse I expressed about 14,000 ug/ml of RlMut4 on day 2 and about 22,000 ug/ml of RlMut4 on day 7; Mouse 2 expressed about 23,000 ug/ml of RlMut4 on day 2 and about 17,000 ug/ml of RlMut4 on day 7; Mouse 3 expressed about 14,000 ug/ml of RlMut4 on day 2 and about 15,000 ug/ml of RIMut4 on day 7; and Mouse 4 expressed about 5,000 ug/ml of RlMut4 on day 2 and about 5,000 ug/ml of RlMut4 on day 7. Thus, the concentration of RlMut4 fusion protein in the mouse sera ranged from about 5 mg/ml to about 22.5 mg/nl, as measured by direct ELISA. Similar results were found using the FGF-2 competition ELISA and Western blot probing techniques. These results demonstrated a high and sustained systemic expression of RIMut4 by animals injected using- the hydrodynamic method, similar to that observed with FGFRI-Ulc-Fc, Example 17: In Vivo Comparison of FGFRI-HlIc-Fe Fusion Protein Produced by 293-6E and CHO-S Host Cells [0321] As shown in Example 5, FGFR1-I Ic-Fe expressed by CHO-S host cells showed superior in vitro stability compared to FGFRl-Ec-Fc expressed by 293-6E host cells. To determine whether FGFRI-IlIc-Fc expressed by CHO-S host cells also showed a superior in vivo stability profile compared to FGFR1 -Ic-Fc expressed by 293-6E host cells, FGFR1-IIIc-Fc protein from both sources was injected into mice and compared over a 72 hour time course by Western blot. [0322] C57BL6 mice were injected via tail vein with a dose of 3 mg/kilogram (kg) FGFRI -rn-Fe protein purified either from CHO-S or 293-6E host cells, as described in Examples 7 and 15. Blood was obtained retro-orbitally at 5 min, 30 min, 24 lir, 48 hr, and 101 72 hr post-injection and heparinized. Serum (100 pl) from each injected mouse and from an uninjected control mouse was separated on reducing 4-12% polyacrylamide gels via SDS-PAGE, transferred onto nitrocellulose membranes, and probed with IRP-conjugated anti-human Fc antibody (anti-human Fc HiRP) (Jackson ImmunoResearch Laboratories, Inc.; West Grove, PA). [0323] As shown in FIG. 26, FGFRI-Uc-Fc purified from 293-6B cells and injected into mice was quickly degraded in vivo and was undetectable via Western blot by 24 hr post-injection. FGFRI-lHIc-Fc expressed from CHO-S cells was more stable in vivo and was readily detectable in the serum, even at 72 hr post-injection. This study showed that the CHO-S host cell-derived FGFR1-Ic-Fc had a longer senun half life than the 293 6E host cell-derived material. [0324] As also shown in FIG. 26, FGFR1-IIc-Fc demonstrated different electrophoretic properties when expressed by 293-6E cells compared to CI-O-S cells. FGFRI -mc-Fc produced by CHO-S cells had an apparent molecular weight of approximately 90 kDa on reducing SDS-PAGE gels and migrated to a position 3-4 kD higher than FGFR1-Eflc-Fc produced by 293-6E cells. Also, the appearance of the CHO S-derived FGFR1-Illc-Fc in the gel was more compact in comparison to the more diffuse gel band of the FGFRl-llc-Fc derived from 293-6E host cells. Example 18: In Vivo Inhibition of Tumor Growth By FGFR1-Ic-Fc and RIMut4 [0325] Xenograft models of tumor growth can be used to assess the in vivo inhibitory properties of cancer therapeutic agents. Caki-1 human kidney tumor cells (ATCC; Manassas, VA) form tumors when injected into severe combined immunodeficient CB17 scid/scid (CB17SCID) mice. Treatment with FGFRI-IIc-Fc or RlMut4 following the injection of the tumor cells decreases the size of the tumors which form in the mice. [0326] Treating mice with FGFR1-IIc-Fc via hydrodynamic tail vein transfection following the injection of Caki-1 tumor cells reduced the volume of the tumors which formed in the mice, as compared to animals mock-transfected with saline. Nine-week old female CB17SCID mice (Charles River Laboratory) were subcutaneously implanted with 1.5 x 10 7 Caki-1 cells in a volume of 200 J. On day 5 post-tumor implantation, "mini circle" FGFRI-IIc-Fc cDNA was delivered at a concentration of 7.5 pg/ml by hydrodynamic tail vein transfection to 13 animals, as described in Example 15. Two ml of 102 the FGFRI-IIIc-Fc cDNA composition comprising 15 pg FGFR1 -Ic-Fc cDNA was injected into the tail vein within 5-8 seconds. Saline was injected into 13 control mice. The resulting tumors were measured by caliper at days 20, 25, 29, 36, 46, 57, and 64. The tumor volume (mm 3 ) was calculated by the formula (n/6)*L 2 *W, in which L (mm) designated the length and W (mm) designated the width of the tumor. The results, shown in FIG. 25; demonstrated that FGFR1-IIlc-Fc expression inhibited Caki-1 tumor growth by about 25% to 50% at all measured time points, as compared to the saline-treated controls. [0327] Recombinant FGFR1-IIlc-Fc also reduced the volume of Caki-1 tumors in mice. FGFR1-Umc-Fo was expressed by CHO-S host cells and purified as described in Example 7. Forty eight CB17SCID mice were injected subcutaneously in the flank with 1.5 x 107 human tumor Caki-1 cells in a PBS vehicle with an injection volume of 200 Pl and assigned to one of four treatment groups. Group 1 (n=12) received saline only; Group 2 (n1 1) received 1 mg/kg FGFR1-IUc-Fc; Group 3 (n=12) received 5 mg/kg FGFR1-ffc Fc; and Group 4 (n=13) received 15 mg/kg FGFRI-Ic-Fc. Saline or FGFR1-UIc-Fc treatment began one day after the tumors were injected and was given twice a week by injecting the appropriate dose of FGFRI-IIlc-Fc into the tail vein in a volume of 200 pl with saline vehicle. Group 1 control mice received only PBS vehicle injections. [0328] The length and width of each tumor resulting from the injected Cali-1 cells were measured with a caliper and the tumor volume was calculated using the equation volume = (n/6)*If *W, as described above. Measurements were made at seven time points between day 14 and day 57 following injection of the Caki-1 cells. The results are shown in FIG. 30A. FGFR1-Ic-Fc inhibited the growth of the Caki-1 cell-induced tumors by about 50% on day 57. All three doses of FGFRI-Ilc-Fc inhibited growth to approximately the same extent. [0329] The above experiment was also repeated using different a lower dose range of FGFRI-Iflc-Fc and a single dose of RlMut4. The FGFR1-Illc-Fc protein was produced in CHO-S host cells and the RI Mut4 protein was produced in DG44 host cells and the proteins purified as described in Example 7. [0330] Ninety CB17SCID mice were injected subcutaneously in the flank with 1.5 x I0 7 human tumor Caki-1 cells in a PBS vehicle with an injection volume of 200 pl and assigned to one of six treatment groups. Group 1 (n=15) received saline only; Group 2 (n=15) received 5 mg/kg FGFR1-IIlc-Fc; Group 3 (n=14) received 1 mg/kg FGFR1-Iltc Fc; Group 4 (n=14) received 0.3 mg/kg FGFRI-lIc-Fc; Group 5 (n=15) received 0.1 mg/kg FGFR1-IIlc-Fc and Group 6 (n=17) received 5 mg/kg RlMut4. Saline; FGFR1- 103 mc-Fc or RlMut4 treatment began one day after the tumors were injected and was given twice a week by injecting the appropriate dose of FGFRI-ll1c-c into the tail vein in a volume of 200 il with saline vehicle. Group I control mice received only PBS vehicle injections. [0331] The length and width of each tumor resulting from the injected Caki-1 cells were measured with a caliper and the tumor volume was calculated using the equation volume =(n/6)*L *W, as described above. Measurements were made at three time points between day 14 and day 27 following injection of the Caki-1 cells. The results are shown in FIG. 30B. Twice-weekly doses of 5 mg/ml of FGFR1-IUc-Fc or RlMut4 inhibited the growth of the Caki-1 cell-induced tumors by about 50% on day 27. This demonstrated that the RlMut4 fusion protein was similarly potent in preventing tumor cell growth in vivo as the parental FGFRi-lEc-Fe molecule. This experiment also demonstrated that doses of FGFRI-mle-Fe as low as 0.3 ing/mI, but not 0.1 mg/ml, inhibited tumor cell growth in a xenograft animal model. Example 19: FGFR1.-c-Fc Slalylation and Glycosylation by 293-6E and CHO-S Cells [0332] To investigate the factors responsible for the differences in stability of FGFRl-mc-Fc expressed by 293-6E cells and CHO-S cells, the sialic acid content of the two fusion proteins was compared. FGFR1-mc-Fc produced by 293-6E cells had a different sialylation pattern than FGFR1-le-Fc produced from CHO-S cells. [0333] FGFR1-llIc-Fc expressed by 293-6B host cells, FGFR1-Ic-Fc expressed by CHO-S host cells, and RlMut4 expressed by CHO-S host cells were analyzed for sialic acid content by high pH anion-exchange chromatography with pulsed amperometric detection (HPAEC-PAD) at the University of California at San Diego. Briefly, protein was treated with 2 M HOAc at 80 C for 3 hr. Sialic acids from the samples were collected by ultra-filtration through a 3,000 NMWCO membrane and eluted from a Dionex CarboPac PA-i HPAEC-PAD column (Dionex; Sunnyvale, CA) with a sodium acetate gradient separating the two common forms of mammalian sialic acids, N acetylneuraminic acid (Neu5Ac) and N-glycolylneuraminic acid (Neu5Gc). [0334] As shown in Table 4 below, FGFR1-Mc-Fc produced from 293-6E host cells, and FGFR1-flc-Fc and RlMut4 from CHO-S host cells are differentially sialated, with higher levels of both the Neu5Acand Neu5Gc types of sialic acid present in CHO-S cell-derived proteins. The increased sialic acid content of CHO-S-derived FGFRI-Ifc-Fc 104 may account, in whole or in part, for the observed differences in molecular weight and in, vivo stability. Table 4: Sialic Acid Analysis of FGFR1-Me-Fe and R1Mut4 Sample nMoles/mg Protein Moles/Mole 1111 Ne 5Go Nou5A FGFR1-]Uc-Fc (293 origin) 06.05 6.8 4.3_7 FGFRI1Mc-Fc (CHO origin) 0-6.73 1T90.g-2 12.47 R1Mut4 (CHO origin) 2.52 117.96 7.71 [0335) FIG. 27 (upper panel) shows a chromatographic. analysis of the N-labeled glycans of FGFR1-me-Fc expressed from 293-6E cells, following their separation on a Glycosep C column. FIG. 27 (lower panel) shows the same separation performed on FGFR1-IIIc-Fc expressed from CHO-S cells. The chromatographic analysis was perfonned by Prozyme (San Leandro, CA). The N-labeled glycans were identified by comparison to a 2-aminobenzamide (2-AB) labeled bovine fetuin N-linked glycan library. The positions of the neutral glycans, free 2-AB labeled glycans, and mono-, di-, tri-, and tetra-sialylated glycans are indicated. [0336] FGFR1-c-Fc expressed from 293-6B cells predominantly expressed neutral glycans (asialo-glycans). In contrast, FGFR1-IImc-Fc expressed from CHO cells predominantly expressed negatively charged sialylated glycans (about 86% of their carbohydrates). The negatively charged glycans of FGFRI-Ic-Fc expressed from 293-6B host cells were mostly monosialylated, whereas the negatively charged glycans of FGFR1 II1c-Fc expressed from CHO-S host cells comprised mono-, di-, tri-, and tetra-sialylated glycans, with the tetra-sialylated glycans comprising the major component. These results suggest that the differences in the levels of sialylation between FGFR1-Ic-Fc of 293 cell origin and that of CHO-S cell origin could be responsible for the in vivo stability differences between the two proteins, as shown in Example 18. Example 20: Pharmacodynanic Studies of FGFRI-IIIc-Fc in Mice [03371 Pharmacodynamic studies of FGFRI-IIlc-Fc in C57B16 mice show that the fusion protein is present in the serum for approximately 25 days post-injection and retains FGF-2 binding activity for approximately 14 days. The studies were performed by injecting mice with a fixed dose of recombinant FGFR1-Mc-FC expressed by CHO-S 105 cells. Nine-week old female C57/B16 mice (Charles River Laboratory) were injected with 200 ul of FGFRI-Mc-Fo protein solution at a final dose of 15 mg/kg. Serum samples (100 ul) were collected retroorbitally at 30 min, 5 hr, and days 1, 2, 3, 4, 5, 7, 14, and 25 post injection; four mice were examined at each time point. The concentration of FGFRI-IIc Fc in the serum samples was analyzed both by direct ELISA and FGF-2 competition ELISA, as described in Example 15. [0338] The results of the FGFR1 -Ic-Fc direct ELISA are shown in FIG. 28. The concentration of FGFR1-IIMc-Fc was about 62 ug/ni at 30 min, 27 ug/mnl at 5 hr, 21 ug/ml on day 1, 19 ug/ml on day 2, 18 ug/ml on day 3, 17 ug/ml on day 4, 18 ug/mI on day 5, 18 ug/ml on day 7, 10 ug/mi on day 14, and about 2 ug/mI on day 25. This result showed that the recombinant FGFRI-IIlc-Fc protein was stable in mice and was detectable at least until day 25 post-injection. [0339] The results of the FGF-2 competition ELISA measuring the FGF-2 binding capacity of the serum FGFR1-IIc-Fc are shown in FIG. 29. The sera were serially diluted and the amount of binding of the FGFR1-fllc-Fc in the scrum to FGF-2 (0.2 pg/ml ) was measured. As shown in FIG. 29, FGF-2 binding by FGFR1-IlIc-Fc in the serum of the injected mice diminished with time, as seen by the rightward shift in the binding curves. FGF-2 binding ability roughly paralleled the amounts of FGFRI -lEc-Fc measured by direct ELISA. FGF-2 binding activity remained detectable by the FGFR1-Mc-Fc in the mouse sera at day 14 but by day 25, the competition ELISA was not able to detect FGF-2 binding activity in the amounts of mouse serum which were tested. Example 21: Pharniacodynamic Studies of R1Mut4 in Mice [0340] Pharmacodynamic studies of R1Mut4 in C57 mice show that RlMut4 has approximately the same in vivo stability as FGFR1-IIc-Fc. C57 juice (Charles River Laboratory) at 2 - 3 months of age were injected subcutaneously with a dose of 10 mg/kg FGFRI-lc-Fc or RlMut4 in 200 ul of saline vehicle. The FGFR-IIc-Fc and RlMut4 were both prepared by expression in a pcDNA3.1 vector in CHO-S cells, as described in Example 2. Serum samples (200 ul) were collected at 4 hr, 3 days, and 7 days post injection. The concentrations of the FGFR1-Iic-Fc and RlMut4 proteins in the serum samples were analyzed by Western blot, as described in Example 5 and the results shown in FIG. 31. Four hours after injection, mice treated with FGFR1-IImc-Fcand RlMut4 showed about the same amount of reactivity with the anti-Fe antibody, about 6.3 ng or more, determined by comparison with the standards of known quantities of FGFR1-ffIc-Fc 106 prepared from CHO-S cells and shown in the right panel. By day 3, the amount of protein present in the serum of all the animals had decreased to about 3.1 ng or more determined by comparison with the standards. By day 7, the amount of protein present in the serum of all the animals had decreased to about 1.6 ng or more. These results demonstrated that recombinant FGFR1-IIc-Fc and RlMut4 proteins had similar stability in vivo. Example 22: Over-expression of FGFR1, FGFR3, and FGFR4 in Cancerous Tissues Relative to Normal Tissues [0341] The analysis and sorting of the expression data residing in the proprietary oncology database from GeneLogic (Gaithersburg, MD) herein identified cancers that over-expressed FGFR1, FGFR3, and FGFR4 compared to corresponding normal tissues. These cancers are therapeutic targets for the FGFR fusion proteins of the invention. The GeneLogic database was generated by hybridizing Affymetrix U133 (Santa Clara, CA) microarray chips with cRNAs derived from over 3000 malignant tissue samples and with cRNAs derived from over 4500 normal tissue samples. The Affymetrix U133 microarray chip contains probes corresponding to FGFRI, probes corresponding to FGFR3, and probes corresponding to FGFR4. [0342] Data derived from all malignant tissue samples and from all normal tissue samples were segregated into datasets corresponding to individual cancer types and to their corresponding nonnal tissues. Over 75 distinct cancer types are represented in the database. Cancer types with datasets containing samples expressing greater than the median expression value of FGFRi in the corresponding normal tissue dataset, greater than the median expression value of FGFR3 in the corresponding normal tissue dataset, and greater than the median expression value of FGFR4 in the corresponding normal tissue dataset were considered to over-express FGFR1, to over-express FGFR3, or to over express FGFR4, respectively. The proportion of samples in the dataset for any given cancer type over-expressing FGFRI, FGFR3, or FGFR4 was calculated as a percentage of the total number of samples in that dataset, as shown in Table 5.

107 Table 5: FGFR1, FGFR2, and FGFR4 Over-expression in Malignant Tissues Cancer Type Percent of Malignant Tissue Samples Over-expressing FGFR1 FGFR3, or FGFR4 FGFRI FGFR3 FGFR4 Leukemia . B-cell acute lymphoblastic leukemia 100 Chronic myelomonocytic leukemia 100 Chronic lymphocytic leukemia 100 Chronic myeloid leukemia 75 Lymqphomta Biurldtt"s tumor of extranodal site 7 Hodgkin's disease of lymph nodes 6 Malignant lymphoma of extranodal site 12 Malignant Iyphoma, non-Hodgkin's type 16 1 Myeloma Plasmacytoma 50_ Sarcoma -- Malignant neoplasm of bone 83 35 3 Malignant neoplasm of heart 100 Malignant neoplasm of soft tissues 37 15 7 Neurologic Malignant neoplasm of brain 89 21 Breast Malignant neoplasm of female breast 16 24 13 Malignant neoplasm of male breast 25 Digestive tract/Gasfrointestinal Malignant neoplasm of ampulla of Vater 50 25 Malignant neoplasm of appendix 50 Malignant neoplasm of colon 30 17 45 Malignant neoplasm of duodenum 63 18 36 Malignant neoplasm of esophagus 22 18 55 Malignant neoplasm of gallbladder 66 33 Malignant neoplasm of liver 32 55 52 Malignant neoplasm of pancreas 11 28 14 Malignant neoplasm of the eritoneum 35 Malignant neoplasm of rectum 26 1 14 59 108 Cancer Type Percent of Malignant Tissue Samples Over-expressing FGFR1I FGFR3, or FGFR4 .Malignant neoplasm of small intestine 69 30 7 Malignant neoplasm of stomach 30 14 28 Endocrine Cancers Malignant neoplasm of adrenal gland 50 100 Malignant neoplasm of islets of Langerhans 9 45 45 Malignant neoplasm of thyroid gland 43 3 Malignant neoplasm of eye 50 Gentitourinay Malignant neoplasm of bladder 8 56 Malignant neoplasm of kidney 77 5 27 Malignant neoplasm of prostate 11 23 Malignant neoplasm of testis 80 100 38 Malignant neoplasm of ureter 100 Gynecologic Malignant neoplasm of uterine cervix 26 52 Malignant neoplasm of myometrium 100 Malignant neoplasm of ovary 16 13 6 Malignant neoplasm of uterus 60 20 Malignant neoplasm of endometrium 69 16 7 Malignant neoplasm of placenta 100 Malignant neoplasm of vulva 5 5 Head & Neck Malignant neoplasm of larynx 36 26 Malignant neoplasnm of maior salivary gland 66 Malignant neoplasm of nasal cavity 100 Malignant neoplasm of oral cavity 20 60 Malignant neoplasm of parotid gland 27 18 9 Malignant neoplasm of tongue 71 14 Malignant neoplasm of tonsil . 100 knjratogy_ iracic Malignantneoplasmoflung 28 35 4 Malignant neolasm of thymus 50_ Malignant neoplasm of trachea 100 109 Cancer Type Percent of Malignant Tissue Samples Over-expressing FGFR1, FGFR3, or FGFR4 Skin I I Maligdant neoplasm of skin 22 16 1 [0343] FGFRI was over-expressed in leukemia, including B-cell acute lymphoblastic leukemia, chronic myelomonocytic leukemia, chronic lymphocytic leukemia, and chronic myeloid leukemia; in lymphoma, including Hodgkin's lymphoma, non-Hodgkin's lymphoma, and extranodal lymphoma; in myeloma, including plasmacytoma; in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the ampulla of Vater, appendix, colon, duodenum, esophagus, liver, pancreas, peritoneum, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the adrenal gland, islets of Langerhans, and thyroid gland; in eye cancer, including malignant neoplasms of the eye; in genitourinary cancer, including malignant neoplasms of the bladder, kidney, prostate, and testis; in gynecologic cancer, including malignant neoplasms of the uterine cervix, myometrium, ovary, uterus, endometrium, placenta, and vulva; in head and neck cancer, including malignant neoplasms of the larynx, salivary gland, nasal cavity, oral cavity, parotid gland, and tongue; in respiratory/thoracic cancer, including malignant neoplasms of the lung, thymus, and trachea; and in skin cancer, including malignant neoplasms of the skin (Table 5). [0344] FGFR3 was over-expressed in lymphoma, including Burkitt's lymphoma; in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in breast cancer, including malignant neoplasms of the female breast and male breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the ampulla of Vater, colon, duodenum, esophagus, gallbladder, liver, pancreas, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the islets of Langerhans and thyroid gland; in genitourinary cancer, including malignant neoplasms of the bladder, kidney, prostate, testis, and ureter; in gynecologic cancer, including malignant neoplasms of the uterine cervix, ovary, uterus, endoinetrium, and vulva; in head and neck cancer, including malignant neoplasms of the larynx, oral cavity, parotid gland, tongue, and tonsil; in 110 respiratoiy/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 5). [0345] FGFR4 was over-expressed in lymphoma, including non-Hodgkin's lymphoma; in sarcoma, including malignant neoplasms of the bone, heart, and soft tissues; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon, duodenum, esophagus, gallbladder, liver, pancreas, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the adrenal gland and islets of Langerhans; in genitourinary cancer, including malignant neoplasms of the kidney'and testis; in gynecologic cancer, including malignant neoplasms of the ovary and endometrium; in head and neck cancer, including malignant neoplasms of the parotid gland; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 5). [0346] Table 4 identifies tumors which over-expressed more than one FGFR, for example, malignant hyphoma, non-Hodgkin's type over-expressed FGFR1 and FGFR4; malignant neoplasm of bone, soft tissues, female breast, colon, duodenum, esophagus, liver, rectum, small intestine, stomach, islets of Langerhans, kidney, testis, ovary, endometrium, parotid gland, lung, and skin over-expressed FGFR1, FGFR3, and FGFR4; malignant neoplasm of brain, ampulla of Vater, thyroid gland, bladder, prostate, uterine cervix, uterus, vulva, larynx, oral cavity, and tongue over-expressed FGFRI and FGFR3; and malignant neoplasm of the gall bladder over-expressed FGFR3 and FGFR4. [0347] Our analysis indicated that FGFR1, and FGFR3, and/or FGFR4 were often over-expressed in cancer. This over-expression implicates active FGF signaling pathways in maintaining the viability and proliferative capability of the affected tumors. We concluded that blocking these signaling pathways in the affected tumors, such as with decoy receptors like FGFR1-Fe, FGFR3-Fc and FGFR4-Fc fusion proteins, or their variants will reduce the viability and proliferative capacity of these tumors. Example 23: Over-expression of FGF-1, FGF-2, FGF-4, and FGF-5 in Cancerous Tissues Relative to Normal Tissues [0348] An analysis of the GeneLogic (Gaithersburg, MD) database for the expression of FGF- 1, FGF-2, FGF-4, and FGF-5 in cancer tissue types and in corresponding normal tissue types was performed essentially as described in Example 22. The Affymetrix U1 33 microarray chip contains probes corresponding to FGF-1, FGF-2, FGF-4, and FGF-5. The proportion of samples in the dataset of any given cancer type III over-expressing FGF-1, FGF-2, FGF-4, or FGF-5 was calculated as a percentage of the total number of samples in that dataset and is shown in Table 6. Cancers which over expressed FGF-1, FGF-2, FGF-4, and FGF-5 compared to corresponding normal tissues are therapeutic targets for the FGFR fusion proteins of the invention. Table 6: FGF-1, FGF-2, FGF-4, and FGF-5 Over-expression in Malignant Tissues Cancer Type Percent of Malignant Tissue Samples Over-expressing FGF-1, FGF-2, FGF-4, or F F-5 FGF-1 FGF-2 FGF-4 FGF-5 Leukemia Acute monocytic/monoblastic leukemia 100_ Chronic lymphocytic leukemia Prolympocytic leukemia100 Chronic myloid leukemia 25 Lymiphomna Burkitt"s tumor of extranodal site 100 Hodgkin's disease of lymph nodes 20 6 Malignant lymphoma of extranodal site 12 3 3 Malignant lymphoma, non-Ho d& tye 10 11 6 Myeloma Plasmacytoma 50 Sarcoma Malignant neoplasm of bone 29 29 3 6 Malignant neoplasm of heart 100 100 Malignant neoplasm of soft tissues _25 42 4 18 Neurologic-- Malignant neoplasm of brain 3 53 3 Malignant neoplsm of female breast 19 5 7 Malignant neoplasm of male breast 50 25 Digesrtive trac/Gasroites5tinlt- Mali ant neoplasm of ampulla of Vater 50 25 Malignant neoplasm of appendix 50 Malignant neoplasm of colon 11 12 7 Mali ant neo lasm of duodenum 8 1 Malignant neoplasm of esophagus 7 22 3 112 Cancer Type Percent of Malignant Tissue Samples Over-expressing FGF, FGF-2, FGF-4, or FGF-5 Malignant neoplasm of gallbladder 33 Malignant neoplasm of liver ..

20 23 11 8 Malignant neoplasm of pancreas 30 53_ __ __53 Malignant neoplasm of the peritoneum 7 7 7 Malignantneoplasmofrectum 5 8 3 Malignant neoplasm of small intestine 30 53 Malignant neoplasm of stomach 19 28 9 Endocrine Cancers _ Malignant neoplasm of adrenal gland 50 _ Malignant neoplasm of islets of Langerhans _1 63 9 _ Malignant neoplasm of thyroid gland 22 30 7 1 E ye Malignant neoplasm of eye Genitourinary Maligat neoplasm of bladder 8 8 4 Malignant neoplasm of kidney 2 48 6 Malignant neoplasm of prostate 8 5 4 1 Malignant neoplasm of testis 28 52 42 Malignant neoplasm of ureter 33 Gynecologic Malignant neoplasm of fallopian tube 33 33 Malignant neoplasm of uterine cervix 17 8 13 Malignant neoplasm of myonetrium i1 22 11 Malignant ooplasm of ovary 12 9 6 1 Malignant neoplasm of uterus Malignant neoplasm of endonetrium 19 12 23 2 Malignant neoplasm of placenta 100 Malignant neoplasm of vulva 40 25 5 Head & Neck - Malignant neoplasm of larynx 21 10 is Malignant neoplasm of major salivary gland 33 66 Malignant neo'plasm of parotid gland 18 54 9 Malignant neoplasn of tongue 42 28 28 Malignant neoplasm of tonsil 113 Cancer Type Percent of Malignant Tissue Samples Over-expressing FGF-I, FGF-2, FGF-4, or FGF-5 Respiratory/Thoracc

-

Malignant neoplasm of lung 20 13 8 1 Malignant neoplasm of thymus 50 _ Malignant neoplasm of trachea 100 Skin Malignant neoplasm of skin 27 19 3 6 [0349] FGF-1 was over-expressed in leukemia, including acute monocytic/monoblastic leukemia and chronic myeloid leukemia; in lymphoma, including Burkitt's lymphoina, Hodgkin's lymphoma and non-Hodgkin's lymphoma, and extranodal lymphoma; in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in breast cancer, including malignant neoplasms of the female breast and male breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon, esophagus, gallbladder, liver, pancreas, peritoneum, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the islets of Langerhans and thyroid gland; in genitourinary cancer, including malignant neoplasms of the bladder, kidney, prostate, testis, and ureter; in gynecologic cancer, including malignant neoplasms of the fallopian tube, uterine cervix, myometrium, ovary, endometrium, and vulva; in head and neck cancer, including malignant neoplasms of the larynx, salivary gland, parotid gland, and tongue; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 6). [0350] FGF-2 was over-expressed in lymphoma, including Hodgkin's lymphoma, non-Hodgkin's lymphoma, and extranodal lymphoma; in myeloma, including plasmacytoma; in sarcoma, including malignant neoplasms of the bone, heart, and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the ampulla of Vater, appendix, colon, duodenum, esophagus, liver, pancreas, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the adrenal gland, islets of Langerhans and thyroid gland; in genitourinary cancer, including malignant neoplasms of the bladder, kidney, prostate, and testis; in gynecologic cancer, including malignant neoplasms of the 114 uterine cervix, myometrium, ovary, endometrium, placenta, and vulva; in head and neck cancer, including malignant neoplasms of the larynx, salivary gland, parotid gland, and tongue; in respiratory/thoracic cancer, including malignant neoplasms of the lung, thymus, and trachea; and in skin cancer, including malignant neoplasms of the skin (Table 6). [0351) FGF-4 was over-expressed in leukemia, including pro-lymphocytic leukemia; in lymphoma, including non-Hodgkin's lymphoma and extranodal lymphoma; in sarcoma, including malignant neoplasms of the bone, heart, and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the ampulla of Vater, colon, esophagus, liver, pancreas, peritoneum, rectum, and stomach; in endocrine cancer, including malignant neoplasms of the islets of Langerhans and thyroid gland; in genitourinary cancer, including malignant neoplasms of the bladder, kidney, prostate, and testis; in gynecologic cancer, including malignant neoplasms of the fallopian tube, uterine cervix, myometrium, ovary, endometrium, and vulva; in head and neck cancer, including malignant neoplasms of the larynx and tongue; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 6). [0352) FGF-5 was over-expressed in sarcoma, including malignant neoplasms of the bone and soft tissues; in breast cancer, including malignant neoplasms of the male breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the liver and peritoneum; in endocrine cancer, including malignant neoplasms of the thyroid gland; in gynecologic cancer, including malignant neoplasms of the ovary and endometrium; in head and neck cancer, including malignant neoplasms of the parotid gland; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in sidn cancer, including malignant neoplasms of the skin (Table 6). [0353) Table 6 demonstrates that FGF-1, FGF-2, FGF-4 and FGF-5, were often over-expressed in cancer. This over-expression implicates active FGF signaling pathways in maintaining viability or proliferative capability of the affected tumors. Blocking these signaling pathways in the affected tumors, such as by blocking the interactions between FGF-l, FGF-2, FGF-4, and FGF-5 and their respective receptors with decoy receptors, such as FGFR-1, FGFR3, FGFR4-Fc fusion proteins, or their variants will reduce the viability and proliferative capacity of these tumors.

115 Example 24: Over-expression of FGF--8, FGF-17, FGF-18, FGF-9, and FGF-20 in Cancerous Tissues Relative to Normal Tissues [0354] An analysis of the GeneLogic (Gaithersburg, MD) database for the expression of FGF-8, FGF-17, FGF-18, FGF-9, or FGF-20 in cancer tissue types and in corresponding normal tissue types was performed essentially as described in Example 22. The Affymetrix U133 microarray chip contains probes corresponding to FGF-8, FGF-17, FGF-18, FGF-9, and FGF-20. The proportion of samples in the dataset of any given cancer type over-expressing FGF-8, FGF- 17, FGF-18, FGF-9, or FGF-20 was calculated as a percentage of the total number of samples in that dataset and is shown in Table 6. Cancers which over-expressed FGF-8, FGF-17, FGF-18, FGF-9, and FGF-20 compared to corresponding normal tissues are therapeutic targets for the FGFR fusion proteins of the invention. Table 7: FGF-8, FGF-17, FGF-18, FGF-9, and FGF-20 Over-expression in Malignant Tissues Cancer Type Percent of Malignant Tissue Samples Over-expressing FGF-8, FGF-17, FGF-18, FGF-9, or FGF-20 FGF-8 FGF-17 FGF-18 FGF-9 FGF-.20 Leukemia B-cell acute lymphoblastic leukemia 100 Lymphoma Burkitt"s tumor of extranodal site __ Malignant lymphoma of extranodal site . _3 Malignant lyphoma, non-Hodgkin's type Sarcoma Malignant neoplasm of bone 6 32 Malignant neoplasin of soft tissues 2 32 16 Neurologic Malignant neoplasm of brain 32 Breast Malignant neoplasm of female breast 9 2 Malignant neoplasm of male breast 25 Digestive tract/Gastrointestinal __ Malignant neoplasm of ampulla of Vater 25 116 Cancer Type Percent of Malignant Tissue Saruples Over-expressing lvGF-8, FGF-17, FGILIS, FGF-9, or FGY-20 Malignant neoplasin of appendix so___5_ Mali ant neoplasm of colon 12__7_1 Malignant nooplasni of esophagus 22___3 Malignant neoplasm of gallbladder 66___33 Malignant neoplasm of liver 2____ Malignant nooplasni of pancecas 2 Malignant neoplasm of tho peritoneum a 3 Malignant 21pls o5rcu Malignant neoplasm of small intestine 23____ Malignant neoplasmi of stomach 3___6_ Malignant neoplasm of arena) gland so___ Maigan nopas o Idny7 9 Maigan nopas o tsts28 9 Malignant neoplasm of uteri cervix 4 17 21 Malignant neoplasm of ovary____ 1 66 30 3 Malignant neoplasm of uterus 60 20 ____ Malignant neoplasmn of endometrium 66 so 12 Head & Neck ___ Maligant neoplacm of major salivar gad ____ 33 Malignant neoplasm of parotid gland 50 27 Mali cant ae plam ftouc___ 14 ____ Respirnforyihoracic_____ Maligat neaoplasm oflung ____ __ _61 9 _ __ Malignant neoplasm of tmacheas___ so ____ Sil ____ Malignant neoplasm of skin _____4 4 117 [0355] FGF-8 was over-expressed in endocrine cancer, including malignant neoplasms of the islets of Langerhans (Table 7). [0356] FGF-17 was over-expressed in sarcoma, including malignant neoplasms of the soft tissues; and in gynecologic cancer, including malignant neoplasms of the uterine cervix and ovary (Table 7). [0357] FGF-18 was over-expressed in lymphoma, including Hodgkin's lymphoma and extranodal lymphoma; in sarcoma, including malignant neoplasms of the bone and soft tissues; in breast cancer, including malignant neoplasms of the female breast and malc breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the ampulla of Vater, appendix, colon, esophagus, gallbladder, liver, pancreas, peritoneum, rectum, and stomach; in endocrine cancer, including malignant neoplasms of the thyroid gland; in genitourinary cancer, including malignant neoplasms of the kidney and testis; in gynecologic cancer, including malignant neoplasms of the fallopian tube, uterine cervix, ovary, uterus, and endometrium; in head and neck cancer, including malignant neoplasms of the parotid gland; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 7). [0358] FGF-9 was over-expressed in leukemia, including B-cell acute lymphoblastic leukemia; in lymphoma, including Burkitt's lymphoma and non-Hodgkin's lymphoma; in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the appendix, colon, esophagus, gallbladder, pancreas, peritoneum, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the adrenal gland and islets of Langerhans; in genitourinary cancer, including malignant neoplasms of the bladder, kidney, and testis; in gynecologic cancer, including malignant neoplasms of the uterine cervix, myometrium, ovary, uterus, and endometrium; in head and neck cancer, including malignant neoplasms of the salivary gland, parotid gland, and tongue; in respiratory/thoracic cancer, including malignant neoplasms of the lung and trachea; and in skin cancer, including malignant neoplasms of the skin (Table 7). [0359] Table 6 demonstrates that FGF-8, FGF-17, FGF-18, FGF-9 and FGF-20 were often over-expressed in cancer. this over-expression implicates active FGF signaling pathways in maintaining viability or proliferative capability of the affected tumors. Blocking these signaling pathways in the affected tumors, such as by blocking the 118 interactions between FGF-8, FGF-17, FGF-9 and FGF-20 and their respective receptors, using decoy receptors, such as FGFR1-Fc, FGFR3-Fc and FGFR4-Fc fusion proteins, or any of their variants will reduce the viability and proliferative capacity of these tumors. [0360] FGF-20 was over-expressed in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon; in endocrine cancer, including malignant neoplasms of the islets of Langerhans and thyroid gland; and in gynecologic cancer, including malignant neoplasms of the ovary and endometrium (Table 7). Example 25: Over-expression of FGF-19, FGF-21, and FGF-23 in Cancerous Tissues Relative to Normal Tissues [0361] An analysis of the GeneLogic (Gaithersburg, MD) database for the expression of FGF-19, FGF-21, or FGF-23 in cancer tissue types and in corresponding normal tissue types was performed essentially as described in Example 22. The Affymetrix U133 microarray chip contains probes corresponding to FGF-19, FGF-21, and FGF-23 . The proportion of samples in the dataset of any given cancer type over expressing FGF-19, FGF-21, or FGF-23 was calculated as a percentage of the total number of samples in that dataset and is shown in Table 9. Cancers which over-expressed FGF-19, FGF-21, or FGF-23 compared to corresponding normal tissues are therapeutic targets for the FGFR fusion proteins of the invention. Table 8: FGF-19, FGF-21, and FGF-23 Over-expression in Malignant Tissues Cancer Type Percent of Malignant Tissue Samples Over-expressing

FGF

19, FGF-21, or FGF-23 FGF-19 FGF-21 FGF-23 Sarcoma -.. _---- Malignant neoplasm of bone 3 Malignant neoplasm of soft tissues I Neurologic Malignant neoplasm of brain 3 Digestive tract/Gastrointestinal Malignant neoplasm of colon 5 Malignant neoplasm of esophagus 3 Malignant neoplasm of gallbladder 33 Maliant neoplasm of liver 11 32 Malignant neoplasm of pancreas 22 119 Caner Type Percent of Malignant Tissue Samples Over-expressing FGF 19, FGF-21 or FGF-23 Malignant neoplasm of rectum 7 1 Malignant neoplasmn of small intestine 7 Malignant neoplasm of stomach I Endocrine Cancers Malignant neoplasm of thyroid gland 3_ Genitomtinaty. Malignant neoplasm of testis 19 Gynecologic- Malignant neoplasm of uterine cervix 8 Malignant neoplasm of myometrium Malignant neoplasm of ovary 5 Malignant neoplasm of endometrium 12 Malignant neoplasm of vulva 5 Head & Neck . Malignant neoplasm of larynx 5 RespimtoryThoracic Malignant neoplasm of lung 5 Skin Malignant neoplasm of skin 9 1 [0362] FGF-19 was over-expressed in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon, gallbladder, liver, pancreas, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the thyroid gland; in genitouriDary cancer, including malignant-neoplasms of the testis; in gynecologic cancer, including malignant neoplasms of the uterine cervix, ovary, endometrium, and vulva; in head and neck cancer, including malignant neoplasms of the larynx; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 8). [0363] FGF-21 was over-expressed in digestive tract/gastrointestinal cancer, including malignant neoplasms of the liver and rectum (Table 8).

120 [0364] FGF-23 was over-expressed in sarcoma, including malignant neoplasms of the soft tissues; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the esophagus; in gynecologic cancer, including malignant neoplasms of the myometrium; and in skin cancer, including malignant neoplasms of the skin (Table 8). [0365] Table 8 demonstrates that FGF-19, FGF-21 and FGF-23 were often over expressed in cancer. This over-expression implicates active FGF signaling pathways in maintaining viability or proliferative capability of the affected tumors. Blocking these signaling pathways in the affected tumors such as by blocking the interactions between FGF- 19, FGF-21 and FGF-23 and their respective receptors, using decoy receptors such as FGFRI-Fc, FGFR3-Fc and FGFR4-Fd fusion proteins or their variants will reduce the viability or proliferative capacity of these tumors. Example 26: Over-expression of FGFRI and Over-expression of FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-19, FGF-20, and FGF-21 in Cancerous Tissues Relative to Normal Tissues 103661 FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-19, FGF-20, and FGF-21 can induce proliferation in cancerous cells expressing FGFR1. An analysis of the GeneLogic (Gaithersburg, MD) database for the expression of FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-19, FGF-20, and FGF-21 in cancer tissue types and in corresponding normal tissue types was performed essentially as described in Example 22. The Affymetrix U1133 microarray chip contains probes corresponding to FGFR1, FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-19, FGF-20, and FGF-21. The proportion of samples in the dataset of any given cancer type over-expressing FGFR1, FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-19, FGF-20, or FGF-21 was calculated as a percentage of the total number of samples in that dataset and is shown in Table 10. Cancers which over-expressed FGFR1, FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-19, FGF-20, and FGF-21 compared to correspohding normal tissues are therapeutic targets for the FGFR fusion proteins of the invention.

121 Table 9: FGFR-1, FGF-1, FGF-2, FGF-4, FGF-8, FGF-9, FGF-17, FGF-19, FGF-20, and FGF-21 Over-expression In Malignant Tissues Cancer Type Percent of Malignant Tissue Samples Over-expressing FGFR1, FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-19, GF-20 or FGF-21 FOF FGF FGP FGF FOP FGF FGP FGF FGF FGF FGF RI -1 -2 -4 -5 -8 -9 -17 -19 -20 -21 Leukemia---- - - - - - B-cell acute lymphoblastic 100 100 leukemia-- - - - - - - - Chronic mycloid leukemia 75 25 __ Lymhona Hodgkin's disease of lymph nodes 6 20 6 Malignant lymphoma of extranodal 12 12 3 3 site Malignant lymphoma, non- 16 10 11 6 1 Hodi's tp Mylonia- - -- - - Plasmacytoma 50 50 Sarcoma --- - - - - - - Maligant neoplasmofbone 83 29 29 3 6 32 3 Malignant neoplasm ofoft tissues 37 25 42 4 18 ___ 16 2 1 Neurologic - - - --- - Malignant neoplasm of brain 89 3 53 3 __ 32 3 Breast-... - - --- - - - - - -- - - -- ~ Malignant neoplasm of female 16 19 5 7 2 breast--

-

Digestive tractGastrointestinal - - - Malignant neoplasm of ampulla of 50 50 25 Vater----- ----- ----- Malignant ne oplasm ofappendix 50 50 50 Malignantneoplasmofcolon 30 11 12 7 -7 - 5 Maligantneoplasmofduodenum 63 _ 8 -1-8- - -- - Malignant neoplasm of esophagus 22 7 22 3 Malignant neoplasm of liver 32 20 23 11 8 -11 32 Malignant neoplasm of pancreas 11 30 53 8 2 22 122 Cancer Type Percent of Malignant Tissue Samples Over-expressing FGFR1, FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF--9, FGF-17, FGF-19, GF-2, or FGF-21 Malignant neoplasm of the 35 7 7 7 35 peritofleum---------------------- Malignant neoplasm of rectum 26 5 8 3 5 7 7 Malignant neoplasm of small 69 30 53 23 7 intestine Malignant neoplasm of stomach 30 19 28 9 6 Endocrine Cancers - - Malignant neoplasm of adrenal 50 50 50 gland- - Malignant neoplasm of islets of 9 18 63 9 9 27 9 Langerhans - Malignant neoplasm of thyroid 43 22 30 7 1 3 3 gland Genit ourinary -- - Malignant neoplasm of bladder 8 8 8 4 21 4 Malignant ne oplasm ofkdne 77 2 48 6 19 Malignant neoplasm of rostate 11 8 5 4 __ __ Malignant neoIasm of testis 80 28 52 42 9 19 Gynecalogic-- Malignant neoplasm of uterine 26 17 8 13 21 4 8 cervix----------------------------- - Malignant neoplasm of 100 11 22 11 11 endometrium -------- -------- Malignant neoplasm of ovar 106 12 9 6 1 30 1 5 3 Malignant neoplasm of uterus 60 40

-

Malignant neoplasm of 69 19 12 23 2 50 12 12 Malignant neoplasm of placenta 100 __ too Mali antneo, lasmofvulva 5 40 25 -5 - - - _ 5__ Head & Neck- - -- - Malignant neolasm ofl x 36 021 Malignant neoplasm of major 66 33 66 33 salivary gland - n-p---a---8- Malignant neoplasm of parotid 27 18 54 9 27 gland

II-

123 Cancer Type Percent or Malignant Tissue Samples Over-expressing FGFR1, FGF-1, FGF-2, FGF-4, FGF-S, FGF-8, FGF-9, FGF-17, FGF-19, GF-or FF-21 Malignant neoplasms of tongue 71 42 28 28 ____ 14 RespiratorylhoracIC Malignant neoplasm of lung 28 20 1 8 1 9 5 Malignant neoplasm of thymus 50 Malignant neoplasm of trachea 100 -_ 100_ 50 Skin _ Malignant neoplasm of skin 22 27 19 3 6 4 9 [0367] FGFRI and FGF-1 were both over-expressed in leukemia, including chronic myeloid leukemia; in lymphoma, including Hodgkin's lymphoma, non-Hodgkin's lymphoma, and extranodal lymphoma; in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon, esophagus, liver, pancreas, peritoneum, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the islets of Langerhans, and thyroid gland; in genitourinary cancer, including malignant neoplasms of the bladder, kidney, prostate, and testis; in gynecologic cancer, including malignant neoplasms of the uterine cervix, myometrimn, ovary, endometrium, and vulva; in head and neck cancer, including malignant neoplasms of the larynx, salivary gland, parotid gland, and tongue; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 9). [0368] FGFRI and FGF-2 were both over-expressed in lymphoma, including Hodgkin's lymphoma, non-Hodgkin's lymphoma, and extranodal lymphoma; in mycloina, including plasmacytoma; in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the ampulla of Vater, appendix, colon, duodenum, esophagus, liver, pancreas, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the adrenal gland, islets of Langerhans, and thyroid gland; in genitourinary cancer, including malignant neoplasms of the bladder, kidney, prostate, and testis; in gynecologic cancer, including malignant neoplasms of the 124 uterine cervix, myometriun, ovary, endometrium, placenta, and vulva; in head and neck cancer, including malignant neoplasms of the larynx, salivary gland, parotid gland, and tongue; in respiratory/thoracic cancer, including malignant neoplasms of the lung, thymus, and trachea; and in skin cancer, including malignant neoplasms of the skin (Table 9). [0369] FGFRI and FGF-4 were both over-expressed in lymphoma, including non Hodgkin's lymphoma, and extranodal lymphoma; in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the ampulla of Vater, colon, esophagus, liver, pancreas, peritoneum, rectum, and stomach; in endocrine cancer, including malignant neoplasms of the islets of Langerhans, and thyroid gland; in genitourinary cancer, including malignant neoplasms of the bladder, kidney, prostate, and testis; in gynecologic cancer, including malignant neoplasms of the uterine cervix, myometrium, ovary, endometrium, and vulva; in head and neck cancer, including malignant neoplasms of the larynx and tongue; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 9). {0370] FGFRl and FGF-5 were both over-expressed in sarcoma, including malignant neoplasms of the bone and soft'tissues; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the liver and peritoneum; in endocrine cancer, including malignant neoplasms of the thyroid gland; in gynecologic cancer, including malignant neoplasms of the ovary and endometrium; in head and neck cancer, including malignant neoplasms of the parotid gland; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 9). [0371] FGFR1 and FGF-8 were both over-expressed in endocrine cancer, including malignant neoplasms of the islets of Langerhans (Table 9). [0372] FGFRI and FGF-9 were both over-expressed in leukemia, including B-cell acute lymphoblastic leukemia; in lymphoma, including non-Hodgkin's lymphoma; in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the appendix, colon, esophagus, pancreas, peritoneum, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the 125 adrenal gland and islets of Langerhans; in genitourinary cancer, including malignant neoplasms of the bladder, kidney, and testis; in gynecologic cancer, including malignant neoplasms of the uterine cervix, myometrium, ovary, uterus, and endometrium; in head and neck cancer, including malignant neoplasms of the salivary gland, parotid gland, and tongue; in respiratory/thoracic cancer, including malignant neoplasms of the lung and trachea; and in skin cancer, including malignant neoplasms of the skin (Table 9). [0373] FGFRI and FGF-17 were both over-expressed in sarcoma, including malignant neoplasms of the soft tissues; and in gynecologic cancer, including malignant neoplasms of the uterine cervix and ovary (Table 9). [0374] FGFR1 and FGF-19 were both over-expressed in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon, liver, pancreas, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the thyroid gland; in genitourinary cancer, including malignant neoplasms of the testis; in gynecologic cancer, including malignant neoplasms of the uterine cervix, ovary, endometrium, and vulva; in head and neck cancer, including malignant neoplasms of the larynx; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 9). [0375] FGFR1 and FGF-20 were both over-expressed in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon; in endocrine cancer, including malignant neoplasms of the islets of Langerhans, and thyroid gland; and in gynecologic cancer, including malignant neoplasms of the ovary and endometrium (Table 9). [0376] FGFRI and FGF-21 were both over-expressed in digestive tract/gastrointestinal cancer, including malignant neoplasms of the liver and rectum (Table 9). [0377] Table 9 demonstrates that FGFRI and any one or more of FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-19, FGF-20 and FGF-21 were often over expressed in cancer. This over-expression implicates active FGF signaling pathways in maintaining viability or proliferative capability of the affected tumors. Blocking these signaling pathways in the affected tumors, such as by blocking the interaction between FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-0, FGF-17, FGF-19 and/or FGF-20 and their respective receptors and between FGFR1 and its binding ligands, using decoy receptors 126 such as FGFR1-Fc, FGFR3-Fc, and FGFR4-Fc fusion proteins, or their variants will reduce the viability or proliferative capacity of these tumors. Example 27: Over-expression of FGFR3 and Over-expression of FGF-1, FGF-2, FGF-4, FGF-5, FGIF-8, FGF-9, FGF-17, FGF-18, FGF-19, and FGF-20 in Cancerous Tissues Relative to Normal Tissues [0378] FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19, and FGF-20 can induce proliferation in cancerous cells expressing FGFR3. An analysis of the GeneLogic (Gaithersburg, MD) database for the expression of FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19, and FGF-20 in cancer tissue types and in corresponding normal tissue types was performed essentially as described in Example 22. The Affymetrix U133 microan-ay chip contains probes corresponding to FGFR3, FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19, and FGF-20. The proportion of samples in the dataset of any given cancer type over-expressing FGFR3, FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19, and FGF-20 was calculated as a percentage of the total number of samples in that dataset and is shown in Table 11. Cancers which over-expressed FGFR3, FGF-1, FGF-2, FGF-4, FGF-5, FGF 8, FGF-9, FGF-17, FGF-18, FGF-19, and FGF-20 compared to corresponding normal tissues are therapeutic targets for the FGFR fusion proteins of the invention. Table 10: FGFR-3, FGF-1, FGF-2, FGF-4, FGF-S, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19, and FGF-20 Over-expression in Malignant Tissues Cancer Type Percent of Malignant Tissue Samples Over-expressing FGFR3, FGF-1, FGF-2, FGF-4, FGF-S, FGF-8, FGF-9, FGF-17, FGF-18, FG-19 or FGF-20 POP FF POP FOGP F OP F0P FOF FGF FGP FGF 3 -1 -2 -4 -5 -9 -9 -17 -18 -19 -20 Lynphoma- Burkitt"s tumor of extranodal site 7 lao 1 Sarcoma-- - - - - - - - - Malignant neoplasm ofbone 35 29 29 3 6 32 6 3 Malignant neoplasm ofsoft tssues 15 25 42 4 18 16 2 32 1 Neurolo I Malignantneoplasmofbrain 21 3 53 3 32 3 1 Breast 127 Cancer Type Percent of Malignant Tissue Samples Over-expressing FGFR3, FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-18, FGF--19 or FGF-20 Malignant neoplasm of female 24 19 5 7 2 9 breast--- -------------- ----- Malignant neoplasmof mltbreast 25 50 25 25 Digestive fract/GostroilestinIal

-

Malignant neoplasm of ampulla of 25 50 25 25 Vater

-

Malignant neoplasm of colon 17 11 12 7 7 12 5 1 Malignant neolasm of duodenum 18 18 - - Malignant neoplasm of esophagus 18 7 22 3 3 22 Malignant nolasn of gallbladder 66 33 - 33 66 33 Malignant neoplasm ofliver 55 20 23 11 8 2 11 Malignant neoplasm of pancreas 28 30 53 8 __ 2 22 22 Malignant neoplasm of rectum 14 5 8 3 __ __ 5 ___ 21 7 Malignant neoplasm of small 30 30 53 23 7 intestine--- - - - Malignant neoplasm of stomach 14 19 28 9 _6_ ___3 1 Endocrine Cancers -- -- - - Malignant neoplasm of islets of 45 18 63 9 9 27 9 Langerhans -.- - -- I Malignant neoplasm of thyroid 3 22 30 7 1 9 3 3 Genitourinary-- Malignant neoplasm of bladder 56 8 8 4 _- Malignant neoplasm of kidney 5 2 48 6 _ _ 9 100 28 52 42 9 28 191 Malipant neoplasm of ureter 100 33 __ GynecologIc ----.-.-. Malignant neoplasm of uterine 52 17 8 13 21 4 17 8 cervix I-

---

Malignant neoplasmofov 13 12 9 6 1 __ 30 1 66 5 3 Malignant neolasm of uterus 20 - 20 -0 Malignant neoplasm of 16 19 12 23 2 50 66 12 12 endometrium .- - - - - - - -- Malignant neoplasm of vulva 5 ,40 25 5 5 128 Cancer Type Percent of Malignant Tissue samples Over-expressing FGFR3, FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-18, GF-19, or FGF-20 Head & Neck __ Malignantneoplasmoflarynx 26 21 10 15 5 Malignant neoplasm of parotid 18 18 54 9 27 so glandI Malignant noplasmoftongue 14 42 28 28 14 R e s p r a t o r y l l r a c f c -" t

-

- -- 4 _Malignant neoplasmi of lung 35 20 13 8 1 __ 9 ___ 6 5 Malignantneoplasmofskin 16 27 19 3 6 4 1 4 9 [0379} FGFR3 and FGF-1 were both over-expressed in lymphoma, including Burkitt's lymphoma; in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in breast cancer, including malignant neoplasms of the female breast and male breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon, esophagus, gallbladder, liver, pancreas, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the islets of Langerhans and thyroid gland; in genitourinary cancer, including malignant neoplasms of the bladder, kidney, prostate, testis, and ureter; in gynecologic cancer, including malignant neoplasms of the uterine cervix, ovary, endometrium, and vulva; in head and neck cancer, including malignant neoplasms of the larynx, parotid gland, and tongue; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 10). [0380] FGFR3 and FGF-2 were both over-expressed in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the ampulla of Vater, colon, duodenum, esophagus, liver, pancreas, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the islets of Langerhans and thyroid gland; in genitourinary cancer, including malignant neoplasms of the bladder, kidney, prostate, and testis; in gynecologic cancer, including malignant neoplasms of the uterine cervix, ovary, endometrium, and vulva; in head and neck cancer, 129 including malignant neoplasms of the larynx, parotid gland, and tongue; in respiratory/tboracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 10). (0381] FGFR3 and FGF-4 were both over-expressed in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the ampulla of Vater, colon, esophagus, liver, pancreas, rectum, and stomach; in endocrine cancer, including malignant neoplasms of the islets of Langerhans and thyroid gland; in genitourinary cancer, including malignant neoplasms of the bladder, kidney, prostate, and testis; in gynecologic cancer, including malignant neoplasms of the uterine.cervix, ovary, endometrium, and vulva; in head and neck cancer, including malignant neoplasms of the larynx and tongue; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 10). [0382] FGFR3 and FGF-5 were both over-expressed in sarcoma, including malignant neoplasms of the bone and soft tissues; in breast cancer, including malignant neoplasms of the male breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the liver; in endocrine cancer, including malignant neoplasms of the thyroid gland; in gynecologic cancer, including malignant neoplasms of the ovary and endometrium; in head and neck cancer, including malignant neoplasms of the parotid gland; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 10). [0383] FGFR3 and FGF-8 were both over-expressed in endocrine cancer, including malignant neoplasms of the islets of Langerbans (Table 10). [0384] FGFR3 and FGF-9 were both over-expressed in lymphoma, including Burkitt's lymphoma; in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon, esophagus, gallbladder, pancreas, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the islets of Langerhans; in genitourinary cancer, including malignant neoplasms of the bladder, kidney, and testis; in gynecologic cancer, including malignant neoplasms of the uterine cervix, ovary, uterus, and endometrium; in head and neck cancer, including malignant neoplasms of the parotid gland and tongue; in respiratory/thoracic cancer, 130 including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 10). [0385] FGFR3 and FGF-17 were both over-expressed in sarcoma, including malignant neoplasms of the soft tissues, and in gynecologic cancer, including malignant neoplasms of the uterine cervix and ovary (Table 10). [0386] FGFR3 and FGF.-18 were both over-expressed in sarcoma, including malignant neoplasms of the bone and soft tissues; in breast cancer, including malignant neoplasms of the female breast and male breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the ampulla of Vater, colon, esophagus, gallbladder, liver, pancreas, rectum, and stomach; in endocrine cancer, including malignant neoplasms of the thyroid gland; in genitourinary cancer, including malignant neoplasms of the kidney and testis; in gynecologic cancer, including malignant neoplasms of the uterine cervix, ovary, uterus, and endometrium; in head and neck cancer, including malignant neoplasms of the parotid gland; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 10). [0387] FGFR3 and FGF-19 were both over-expressed in sarcoma, including malignant neoplasms of the bone and soft tissues; in neurologic cancer, including malignant neoplasms of the brain; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon; gallbladder, liver, pancreas, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the thyroid gland; in genitourinary cancer, including malignant neoplasms of the testis; in gynecologic cancer, including malignant neoplasms of the uterine cervix, ovary, endometrium, and vulva; in head and neck cancer, including malignant neoplasms of the larynx; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 10). [0388] FGFR3 and FGF-20 were both over-expressed in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon; in endocrine cancer, including malignant neoplasms of the islets of Langerhans and thyroid gland; and in gynecologic cancer, including malignant neoplasms of the ovary and endometrium (Table 10). [0389] Table 11 demonstrates that FGFR3 and any one or more of FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19 and FGF-20 were often over expressed in cancer. This ovdr-expression implicates active FGF signaling pathways in maintaining viability or proliferative capacity of the affected tumors. Blocking these 131 signaling pathways in the affected tumors, such as by blocking the interactions between FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19 and/or FGF-20 and their respective receptors and between FGFR3 and its binding ligands with decoy receptors such as FGFRl-Fc, FGFR3-Fo and FGFR4-Fc fusion proteins, or their variants will reduce the viability or proliferative capacity of these tumors. Example 28: Over-expression of FGFR4 and Over-expression of FGF-1, FGF-2, FGF-4, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19, FGF-20, FGF-21, and FGF-23 in Cancerous Tissues Relative to Normal Tissues [0390] FGF-1, FGF-2, FGF-4, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19, FGF-20, FGF-21, and FGF-23 can induce proliferation in cancerous cells expressing FGFR4. An analysis of the GeneLogic (Gaithersburg, MD) database for the expression of FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-1 8, FGF-19, and FGF-20 in cancer tissue types and in corresponding normal tissue types was performed essentially as described in Example 22. The Affymetrix Ul 33 microarray chip contains probes corresponding to FGFR3, FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF 18, FGF-19, and FGF-20. The proportion of samples in the dataset of any given cancer type over-expressing FGFR3, FGF-l, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19, and FGF-20 was calculated as a percentage of the total number of samples in that dataset and is shown in Table 11. Cancers which over-expressed FGFR3, FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19, and FGF-20 compared to corresponding normal tissues are therapeutic targets for the FGFR fusion proteins of the invention. Table 11: FGFR-4, FGF-1, FGF-2, FGF-4, FGF-8, FGF-9, FGF-17, FGF-18, FGF 19, FGF-20, FGF-21, and FGF-23 Over-expression in Malignant Tissues Cancer Type Percent of Malignant Tissue Samples Over-expressing FGFR4, FGF 1, FGF-2, FGF-4, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19, FGF-20, FGF-21, or FGF-23 ___ FOP PGF PGP FPO FG FGOP POP PO FOP POP POP POPF R4 -1 -2 -4 -8 -9 -17 -18 -19 -20 -21 -23 Lynwlwmna Malignant lymphoma, non- 1 10 11 6 1 1 odkins Srra 132 Cancer Type Percent of Malignant Tissue Samples Over-expressing FGFR4, FGF 1, FGF-2, FGF-4, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19, FGF-20, FG -Z1, or FGF-23 Malignant neoplasm of bone 3 29 29 _3 321 6 3 1 Malignant neoplasm of head 10 10 10 0 0 0 Malignant neoplasm of soft 7 25 42 4 16 2 32 I tissuesI II A Breast Malignant neoplasm of female 13 19 5 7 2 9 breast Digestive tract/GastrontfestIital___-- MalignantLneoplasmofcolon 12 7 72 5 1 Malignant neoplasm of 36 18 duodenum _ - Malignant neoplasm of 55 7 22 3 3 22 3 esophagus Malignant neoplasm of 33 33 33 66 33 gallbladder - Malignant neoplasm of liver 52 20 23 1L _ __ Maligant neoplasm of pancreas 14 30 53 2 1 2 22 22 Malignant neoplasm of rectum 59 5 8 3 S 1 21 7 1 Malignant neoplasm of small 7 30 53 23 7 intestine __ I- - --- Malignant neoplasm of stomach 28 19 28 9 6 3 1 Endactifte Cancers Malignant neoplasm of adrenal 10 50 50 gland- 01 Malignant neoplasm of islets of 45 18 63 9 9 27 9 Langerhans Geo muIy - - Malignant neoplasm of kidney 27 2 48 6 9 7 Malignant neoplasm of tests 38 28 52 42 9 28 19 Gynecologic - MalIgnant neoplasm of ovary 6 12 9 6 _ 30 1 66 5 3_ Malignant neoplasm of 7 19 12 23 50 66 12 12 endometrium Head & Neck 133 Cancer Type Percent of Malignant Tissue Samples Over-expressing FGFR4, FGF 1, FGF-2, FGF-4, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19, FGF-20, FGF-21, or FGF-23 Malignant neoplasm of parotid 9 18 54 27 50 gland Resporylorac------- Malignant neoplasm of lung 4 120 13 8 ___ 9__ 6 5 _ Skini Malignant neoplasm of skin I 27 19 3 4 4 9. 1 [0391] FGFR4 and FGF- 1 were both over-expressed in lymphoma, including non Hodgkin's lymphoma; in sarcoma, including malignant neoplasms of the bone and soft tissues; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon, esophagus, gallbladder, liver, pancreas, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the islets of Langerhans; in genitourinary cancer, including malignant neoplasms of the kidney and testis; in gynecologic cancer, including malignant neoplasms of the ovary and endometrium; in head and neck cancer, including malignant neoplasms of the parotid gland; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 11). [0392] FGFR4 and FGF-2 were both over-expressed in lymphoma, including non Hodgkin's lymphoma; in sarcoma, including malignant neoplasms of the bone, heart, and soft tissues; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon, duodenum, esophagus, liver, pancreas, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the adrenal gland and islets of Langerhans; in genitourinary cancer, including malignant neoplasms of the kidney and testis,; in gynecologic cancer, including malignant neoplasms of the ovary and endometrium; in head and neck cancer, including malignant neoplasms of the parotid gland; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 11). [0393] FGFR4 and FGF-4 were both over-expressed in lymphoma, including non Hodgkin's lymphoma; in sarcoma, including malignant neoplasms of the bone, heart, and soft tissues; in breast cancer, including malignant neoplasms of the female breast; in 134 digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon; esophagus, liver, pancreas, rectum, and stomach; in endocrine cancer, including malignant neoplasms of the islets of Langerhans; in genitourinary cancer, including malignant neoplasms of the kidney and testis; in gynecologic cancer, including malignant neoplasms of the ovary and endometrium; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 11). [0394] FGFR4 and FGF-8 were both over-expressed in endocrine cancer, including malignant neoplasms of the islets of Langerhans (Table 12). [03953 FGFR4 and FGF-9 were both over-expressed in lymphoma, including non Hodgkin's lymphoma; in sarcoma, including malignant neoplasms of the bone and soft tissues; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon, esophagus, gallbladder, pancreas, rectum, small intestine, and stomach; in endocrine cancer, including malignant neoplasms of the adrenal gland and islets of Langerhans; in genitourinary cancer, including malignant neoplasms of the kidney and testis; in gynecologic cancer, including malignant neoplasms of the ovary and endometrium; in head and neck cancer, including malignant neoplasms of the parotid gland; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 11). [0396] FGFR4 and FGF-17 were both over-expressed in sarcoma, including malignant neoplasms of the soft tissues; and in gynecologic cancer, including malignant neoplasms of the ovary (Table 11). [0397) FGFR4 and FGF-18 were both over-expressed in lymphoma, including non-Hodgkin's lymphoma; in sarcoma, including malignant neoplasms of the bone and soft tissues; in breast cancer, including malignant neoplasms of the female breast; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon, esophagus, gallbladder, liver, pancreas, rectum, and stomach; in genitourinary cancer, including malignant neoplasms of the kidney and testis; in gynecologic cancer, including malignant neoplasms of the ovary and endometrium; in head and neck cancer, including malignant neoplasms of the parotid gland; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 11).

135 [0398] FGFR4 and FGF-19 were both over-expressed in sarcoma, including malignant neoplasms of the bone and soft tissues; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon, gallbladder, liver, pancreas, rectum, small intestine, and stomach; in genitourinary cancer, including malignant neoplasms of the testis; in gynecologic cancer, including malignant neoplasms of the ovary and endometrium; in respiratory/thoracic cancer, including malignant neoplasms of the lung; and in skin cancer, including malignant neoplasms of the skin (Table 11). [0399] FGFR4 and FGF-20 were both over-expressed in digestive tract/gastrointestinal cancer, including malignant neoplasms of the colon; in endocrine cancer, including malignant neoplasms of the islets of Langerhans; and in gynecologic cancer,.including malignant neoplasms of the ovary and endometrium (Table 11). [0400] FGFR4 and FGF-21 were both over-expressed in digestive tract/gastrointestinal cancer, including malignant neoplasms of the liver and rectum (Table 11). [0401] FGFR4 and FGF-23 were both over-expressed in sarcoma, including malignant neoplasms of the soft tissues; in digestive tract/gastrointestinal cancer, including malignant neoplasms of the esophagus; and in skin cancer, including malignant neoplasms of the skin (Table 11). [0402] This analysis demonstrated that FGFR4, FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF-17, FGF-18, FGF- 19, FGF-20, FGF-21 and FGF-23 are commonly over-expressed in cancer. This over.expression implicates active FGF signaling pathways in maintaining viability and/or proliferative capability of the affected tumors. Blocking these FGFR signaling pathways in the affected tumors, such as with FGFRl-Fc, FGFR3 Fc, or FGFR4-Fc fusion proteins, will reduce the viability and/or proliferative capacity of these tumors. [0403] Table I1 demonstrates that FGFR4 and any of FGF-1, FGF-2, FGF-4, FGF-5, FGF-8, FGF-9, FGF, FGF-17, FGF-18, FGF-19, FGF-20, FGF-21 and FGF-23 were often over-expressed in cancer. This over-expression implicates active FGF signaling pathways in maintaining viability or proliferative capability of the affected tumors. blocking these signaling pathways in the affected tumors, such as by blocking the interactions between FGF-1, FGF-2, FGF-4, FGF-8, FGF-9, FGF-17, FGF-18, FGF-19, FGF-20 and FGF-23; and between FGFR4 and its binding ligands with decoy receptors such as FGFR1-Fc, FGFR3-Fc and FGFR4-Fc fusion proteins, or their variants, will reduce the viability or proliferative capacity of these tumors.

136 INDUSTRIAL APPLICABILITY [0404] The FGFR fusion proteins and the polynucleotide molecules that encode them are useful in treating proliferative diseases and diseases involving angiogenesis, including cancer. They can be used to diagnose, prevent, and treat these diseases.

Editorial Note: Patent Application No. 2011244925 The description (pages 1-136) is followed by the sequence listing (pages 137-211) before the claims (pages 212-221).

Claims (73)

1. 2. The FGFR fusion protein of claim I, wherein the I - 22 C-terminal amino acid residues are C-terminal to a valine residue situated at the C-terminus of the IgIll domain and commonly aligned among the wildtype FGFR 1, FGFR2, FGFR3, and FGFR4 extracellular domain C termini.
2. 3. The FGFR fusion protein of any of claims 1-2, wherein the FGFR polypeptide comprises an FGFRI polypeptide. 213
3. 4. The FGFR fusion protein of any of claims 1-2, wherein the FGFR polypeptide comprises an FGFR2 polypeptide.
4. 5. The FGFR fusion protein of any of claims 1-2, wherein the FGFR polypeptide comprises an FGFR3 polypeptide.
5. 6. The FGFR fusion protein of any ofclaims 1-2, wherein the FGFR polypeptide comprises an FGFR4 polypeptide.
6. 7. The FGFR fusion protein of any of claims 1-2, wherein the fusion partner comprises an Fc polypeptide.
7. 8. The FGFR fusion protein of any of claims 1-2, wherein the fusion protein is less susceptible to cleavage.
8. 9. The FGFR fusion protein of claim I, wherein the fusion protein comprises an amino acid sequence of any of SEQ ID NO: 100, SEQ ID NO: 97 to SEQ ID NO: 99, SEQ ID NO: 101, SEQ ID NO: 108 to SEQ ID NO: 11, SEQ ID NO: 118 to SEQ ID NO: 121, SEQ ID NO: 127 to SEQ ID NO: 131, SEQ ID NO: 137 to SEQ ID NO: 140, SEQ ID NO: 146 to SEQ ID NO: 149, SEQ ID NO: 162 to SEQ ID NO: 166, SEQ ID NOS:178 to SEQ ID NO: 182, SEQ ID NO: 199 to SEQ ID NO: 203, SEQ ID NO: 206 to SEQ ID NO: 210, SEQ ID NO: 230 to SEQ ID NO: 234, and SEQ ID NO: 238 to SEQ ID NO: 242.
9. 10. The FGFR fusion protein of claim 1, wherein the fusion protein comprises an amino acid sequence of any of SEQ ID NO: 100, SEQ ID NO: 97 to SEQ ID NO: 99, SEQ ID NO: 101, SEQ ID NO: 108 to SEQ ID NO: I l1, SEQ ID NO: I18 to SEQ ID NO: 121, SEQ ID NO: 127 to SEQ ID NO: 131, SEQ ID NO: I37 to SEQ ID NO: 140, SEQ ID NO: 146 to SEQ ID NO: 149, SEQ ID NO: 162 to SEQ ID NO: 166, SEQ ID NO: 178 to SEQ ID NO: 182, SEQ ID NO: 199 to SEQ ID NO: 203, SEQ ID NO: 206 to SEQ ID NO: 210, SEQ ID NO: 230 to SEQ ID NO: 234, and SEQ ID NO: 238 to SEQ ID NO: 242, wherein said amino acid sequence lacks a native leader sequence. 214
10. 11. The FGFR fusion protein of claim 10, wherein the Fc polypeptide comprises an amino acid sequence of any ofSEQ ID NO: 171 to SEQ ID NO: 173.
11. 12. The FGFR fusion protein of claim 1, produced in a CHO cell or a 293 cell.
12. 13. The FGFR fusion protein of claim 12, wherein the fusion protein comprises an amino acid sequence of any of SEQ ID NO: 100, SEQ ID NO: 97 to SEQ ID NO: 99, SEQ ID NO: 108 to SEQ ID NO: I 11, SEQ ID NO: 118 to SEQ ID NO: 121, SEQ ID NO: 162 to SEQ ID NO: 166, SEQ ID NO: 178 to SEQ ID NO: 182, SEQ ID NO: 199 to SEQ ID NO: 202, SEQ ID NO: 206 to SEQ ID NO: 210, SEQ ID NO: 230 to SEQ ID NO: 234, and SEQ ID NO: 238 to SEQ ID NO: 242, wherein said amino acid sequence optionally lacks a native leader sequence.
13. 14. The FGFR fusion protein of claim 12, wherein the fusion protein lacks a native leader sequence.
14. 15. The FGFR fusion protein of claim 12, wherein the fusion protein is produced using a CHEF expression system.
15. 16. An FGFR fusion protein according to any one of claims 1-15 for use as a medicament.
16. 17. A composition comprising an effective amount of the FGFR fusion protein of any of claims 1-15 and a pharmaceutically acceptable carrier.
17. 18. A kit comprising the composition of claim 17 in a container and instructions for administration into a subject in need of such a composition.
18. 19. The kit of claim 18, wherein the composition comprises a single or multiple doses of the FGFR fusion protein. 215
19. 20. A nucleic acid molecule comprising a polynucleotide that encodes the FGFR fusion protein of any of claims 1-16.
20. 21. A vector comprising the nucleic acid molecule of claim 20 and a promoter that regulates the expression of the nucleic acid molecule.
21. 22. A recombinant host cell comprising a host cell and the FGFR fusion protein of any of I 15, the nucleic acid molecule of claim 20, or the vector of claim 21.
22. 23. The recombinant host cell of claim 22, wherein the host cell is a eukaryotic cell.
23. 24. The recombinant host cell of claim 23, wherein the eukaryotic cell is of CHO or 293 lineage.
24. 25. A polypeptide expressed from the recombinant host cell of claim 23.
25. 26. A method of producing an FGFR fusion protein comprising: (a) providing the recombinant host cell of any of claims 22-24; and (b) culturing the recombinant host cell to express the FGFR fusion protein.
26. 27. The method of claim 26, wherein the method further comprises isolating the FGFR fusion protein from the cell culture.
27. 28. The method of claim 27, wherein isolating the FGFR fusion protein comprises contacting the expressed FGFR fusion protein with an affinity matrix.
28. 29. The method of claim 28, wherein the affinity matrix comprises at least one of Protein A, Protein G, Protein A/G, anti-Fc antibody, and anti-FGFR antibody.
29. 30. The method of any of claims 27-29, wherein isolating the FGFR fusion protein further comprises contacting the FGFR fusion protein with a hydrophobic matrix. 216
30. 31. A method of detecting the level of FGF expression in a subject comprising: (a) providing an FGFR fusion protein according to any one of claims 1-15; (b) providing a tissue sample from the subject; (c) allowing the FGFR fusion protein and the sample to interact under conditions that permit specific FGF binding; (d) measuring the specific binding; and (e) comparing the amount of specific binding to that of a control sample, wherein the FGFR fusion protein binds to the FGF.
31. 32. The method of claim 31, wherein the FGF is at least one of FGF- 1, FGF-2, FGF-3, FGF 4, FGF-5, FGF-6, FGF-7, FGF-8, FGF-9, FGF-10, FGF-16, FGF-17, FGF-18, FGF-19, or FGF 20.
32. 33. A method of inhibiting the viability or proliferation of a proliferative cell in vitro, in vivo, or ex vivo comprising: (a) providing a composition of claim I 7; and (b) contacting the proliferative cell with an amount of the composition effective to inhibit the viability or proliferation of the proliferative cell.
33. 34. The method of claim 33, wherein the proliferative cell is present in a subject and the subject expresses a higher level of one or more FGF ligand than normal.
34. 35. The method of claim 34, wherein a tissue of the subject expresses a higher level of the FGF ligand than normal.
35. 36. The method of claim 35, wherein the FGF ligand binds to at least one of FGFR I -Fc, FGFR2-Fc, FGFR3-Fc, or FGFR4-Fc, or a variant of any of these, as determined by measuring binding interactions in real time. 217
36. 37. The method of claim 36, wherein the FGF ligand is selected from FGF-I, FGF-2, FGF-3, FGF-4, FGF-5, FGF-6, FGF-8, FGF-9, FGF-16, FGF-17, FGF-18, FGF-19 and FGF-20.
37. 38. The method of claim 33, wherein the proliferative cell is present in a subject and the subject expresses a higher level of an FGFR polypeptide than normal.
38. 39. The method of claim 38, wherein a tissue in the subject expresses a higher level of the FGFR polypeptide than normal.
39. 40. The method of claim 38, wherein the FGFR polypeptide comprises FGFR1, FGFR2, FGFR3, or FGFR4.
40. 41. The method of any of claims 33-40, wherein the proliferative cell is a cancer cell, a dysplastic cell, and/or an endothelial cell.
41. 42. The method of claim 41, wherein the proliferative cell comprises a breast cell, a pancreatic cell, a prostate cell, a lung cell, an ovarian cell, a kidney cell, a brain cell, a colorectal cell, a retinal cell, or one selected from Table 5 - Table I .
42. 43. The method of claim 38, wherein the subject expresses a higher level of an FGF ligand than normal.
43. 44. A method of treating cancer in a subject, comprising: (a) providing an FGFR fusion protein of any of claims I - 16; and (b) administering an effective amount of the FGFR fusion protein to the subject to treat cancer.
44. 45. The method of claim 44, wherein the cancer comprises at least one subpopulation of cells that is dependent on or responsive to growth stimulation by an FGF ligand. 218
45. 46. The method of claim 44, wherein the cancer comprises at least one subpopulation of cells that is dependent on or responsive to an angiogenic factor for production of blood vessels for growth.
46. 47. The method of claim 44, wherein the cancer is resistant to VEGF signaling pathway inhibition.
47. 48. The method of claim 44, wherein the cancer comprises metastasizing cancer.
48. 49. The method of claim 44, wherein the cancer comprises a hematologic cancer.
49. 50. The method of claim 49, wherein the hematologic cancer comprises chronic myelogenous leukemia, chronic lymphocytic leukemia, acute myelocytic leukemia, or hairy cell leukemia.
50. 51. The method of claim 49, wherein the cancer comprises a solid tumor.
51. 52. The method of claim 51, wherein the cancer comprises breast cancer, pancreatic cancer, pituitary cancer, prostate cancer, lung cancer, ovarian cancer, renal cell cancer, oral squamous cell cancer, colorectal cancer, bladder cancer, retinal cancer, brain cancer, or a cancer listed in Table 5 - Table 11.
52. 53. The method of claim 48, wherein the cancer comprises bone metastasis.
53. 54. The method of claim 44, further comprising administering a second anticancer therapeutic to the subject.
54. 55. The method of claim 54, wherein the second anti-cancer therapeutic comprises a cytostatic agent, a cytotoxic agent, or an antiangiogenic agent. 219
55. 56. The method of claim 54, wherein the second anti-cancer therapeutic comprises a second FGFR fusion protein, an inhibitor of PDGF signaling, an inhibitor of VEGF signaling, or an inhibitor of EGF signaling.
56. 57. The method of claim 54, wherein the second anti-cancer therapeutic is administered before, after, or contemporaneously with the administration of the FGFR fusion protein.
57. 58. A method of inhibiting angiogenesis in a subject comprising: (a) providing a FGFR fusion protein of any of claims I - 16; and (b) administering an amount of the FGFR fusion protein to the subject effective to inhibit angiogenesis.
58. 59. The method of claim 58, further comprising administering a second therapeutic agent to the subject.
59. 60. The method of claim 59, wherein the second therapeutic agent comprises a cytostatic agent, a cytotoxic agent, or a second anti-angiogenic agent.
60. 61. The method of claim 58 or 59, wherein the subject is treated for macular degeneration.
61. 62. The method of claim 58 or 59, wherein the subject is treated for cancer.
62. 63. The method of claim 59, wherein the second therapeutic agent comprises an anti-cancer agent.
63. 64. The method of claim 63, wherein the anti-cancer therapeutic agent comprises a second FGFR fusion protein, an inhibitor of PDGF signaling, an inhibitor of VEGF signaling, an inhibitor of EGF signaling, an antibody, or an siRNA. 220
64. 65. The method ofany of claims 33, 44, or 58, wherein the composition is administered intravenously, intramuscularly, subcutaneously, topically, orally, intraperitoneally, intraorbitally, by implantation, by inhalation, intrathecally, intraventricularly, and/or intranasally.
65. 66. The method of claim 63, wherein the anti-cancer therapeutic agent comprises surgery, chemotherapy, radiation therapy, and/or the administration of another biologic.
66. 67. Use of an FGFR fusion protein according to any one of claims I - 16 for the manufacture of a medicament for treatment of a proliferative disease.
67. 68. Use of an FGFR fusion protein as in claim 67, wherein the disease comprises cancer or macular degeneration.
68. 69. Use of an FGFR fusion protein as in claim 68, wherein the cancer comprises a hematologic cancer or a solid cancer.
69. 70. Use of an FGFR fusion protein as in claim 69, wherein the cancer comprises breast cancer, pancreatic cancer, pituitary cancer, prostate cancer, lung cancer, ovarian cancer, renal cancer, oral cancer, colorectal cancer, bladder cancer, retinal cancer, brain cancer or a cancer identified in Table 5 - Table 11.
70. 71. A product comprising an FGFR fusion protein according to any one of claims I to 16 and at least one anti-cancer therapeutic as a combined preparation for simultaneous, separate, or sequential use to treat cancer.
71. 72. The FGFR fusion protein according to any one of claims I, 11, 12, 15, or 16, wherein the FGFRI extracellular domain sequence consists of SEQ ID NO: 130, optionally lacking a native leader sequence.
72. 73. The FGFR fusion protein according to claim 72, wherein the FGFRI extracellular domain sequence consists of amino acids 22-360 of SEQ ID NO: 130. 221
73. 74. The FGFR fusion protein according to any one of claims I, 11, 12, 15, or 16, wherein the fusion protein sequence comprises amino acids 22-592 of SEQ ID NO: 100.