MATERIALS AND METHODS FOR INHIBITING CANCER CELL INVASION
RELATED TO FGFR4
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application claims priority to U.S. Provisional Patent Application No.
61/093,925, filed September 3, 2008, and U.S. Provisional Patent Application No.
61/156,634, filed March 2, 2009.
TECHNICAL FIELD OF THE INVENTION
[0002] The invention generally relates to cancer therapy and also to antibodies and
antibody fragments that bind fibroblast growth factor receptor-4 (FGFR4) and uses thereof
and combination therapies containing them.
BACKGROUND OF THE INVENTION
[0003] Tumor cell invasion plays an important role in cancer pathogenesis. The invasion
of carcinoma cells through underlying basement membrane and metastasis to distant organs
are considered rate limiting steps in carcinogenesis and cancer spread. During these
processes, tumor cells rely on targeted proteolytic activities on the cell surface to traverse
basement membrane barriers, and invade and grow in collagen or fibrin-rich interstitial and
temporary matrixes. The formation of secondary tumors in distant regions of the body
complicates therapeutic options and often results poor clinical outcomes in cancer patients.
[0004] Current strategies in cancer treatment focus on blocking tumor growth through pro-
apoptotic, anti-proliferative, and anti-angiogenic therapies. For example, growth factor
receptors, e.g., fibroblast growth factor receptors, have been suggested as a possible target for
inhibiting proliferation. See, e.g., St. Bernard et al, Endocrinology, 146(3): 1145-1153
(2005). Possible regulators of growth factor receptor signaling include, e.g., small molecules,
inactivating ligands, and antibodies. Kwabi-Addo et al., Endocrine-Related Cancer, 11: 709-
724 (2004). Chen et al., Hybridoma, 24(3): 152-159 (2005), for instance, purportedly
identified antibodies that bind FGFR4 on human breast cancer tumor cell lines. However,
attempts to inhibit tumor invasion have not been as successful. Thus, the development of
novel interventions to target cell invasion is of key importance for more efficient cancer
treatments.

SUMMARY OF THE INVENTION
[0005] The invention generally relates to materials that are useful in the treatment of
neoplastic disorders such as cancer, including antibody substances, nucleic acids,
polypeptides, and compositions. The invention further relates to methods of using such
materials, including methods of treatment, medical uses, and uses for making pharmaceutical
compositions. The invention also relates to tools for screening for novel therapeutics and
new combination therapies.
[0006] The invention provides an isolated antibody or antibody fragment that (i) binds an
extracellular epitope of a fibroblast growth factor receptor-4 (FGFR4) that is expressed by
mammalian cells and (ii) inhibits cancer cell invasion. In addition, the invention provides
monoclonal antibody F90-10C5, as well as an isolated antibody, antibody fragment, or
polypeptide that comprises one or more, and preferably all complementarity determine
regions (CDR) of monoclonal antibody F90-10C5 and binds an extracellular epitope of
FGFR4 that is expressed in mammalian cells. In another aspect, the invention provides an
isolated antibody, such as a monoclonal antibody, or fragment thereof that binds an epitope of
FGFR4 that is bound by monoclonal antibody F90-10C5. The FGFR4 recognized by the
antibody can comprise the amino acid sequence of the FGFR4 G388 protein or the FGFR4
R388 variant. Compositions comprising the antibody, fragment thereof, or polypeptide,
optionally combined with other therapeutics and/or with pharmaceutically acceptable
carrier(s), excipient(s), adjuvants, or the like, also are included in the invention.
[0007] The invention also provides materials and methods for making the claimed
antibody or fragment thereof. For example, the invention provides an isolated polynucleotide
that encodes the inventive antibody or fragment thereof, a vector comprising the
polynucleotide, an isolated host cell comprising the polynucleotide or vector, and a
hybridoma. An isolated polynucleotide comprising a nucleotide sequence that encodes at
least one amino acid sequence selected from the group consisting of an antibody heavy chain
variable region and an antibody light chain variable region also is provided by the invention.
The heavy chain variable region and light chain variable region comprise complementarity
determine regions (CDR) identical to monoclonal antibody F90-10C5 CDRs. The invention
also provides a method of identifying an antibody or antibody fragment. The method
comprises obtaining one or more antibodies or antibody fragments that bind FGFR4;

screening the antibodies or antibody fragments in a tumor cell invasiveness assay; and
identifying an antibody that inhibits invasiveness in the assay by at least 50%.
[0008] The invention further includes methods of using the inventive antibody or fragment
thereof. For example, a method of modulating invasion, ingrowth, or metastasis of cancer
cells is provided. The method comprises contacting a population of cancer cells with a
composition comprising the inventive antibody or fragment thereof in an amount effective to
modulate cancer cell invasion, ingrowth, or metastasis. The method can be performed in
vivo, such that the cancer cells are in a mammalian subject, and the contacting step comprises
administering the composition to the mammalian subject.
[0009] In another aspect, the invention includes a method of treating a subject by
administering a composition comprising the inventive antibody, fragment thereof, or
polypeptide. For example, in one embodiment, the method comprises selecting for treatment
a mammalian subject diagnosed with or treated for cancer; and administering to the subject
the inventive composition in an amount effective to modulate cancer cell invasion, ingrowth,
or metastasis. A method of treating cancer also is provided. The method comprises
administering to the subject the composition comprising the inventive antibody or fragment
thereof in an amount effective to treat cancer. Optionally, (i) the antibody or fragment
thereof binds an epitope of FGFR4 that is bound by monoclonal antibody F90-10C5 and (ii)
the method further comprises contacting the population of cancer cells with (or administering
to the subject) an antibody or fragment thereof that binds an epitope of FGFR4 that is
different than the epitope recognized by mAb F90-10C5. Alternatively or in addition, the
method can comprise contacting the population of cancer cells with (or administering to the
subject) an MT1-MMP inhibitor.
[0010] In some variations of the invention, the subject has a cancer that includes cells that
contain at least one FGFR4 allele that is characterized by an arginine at amino acid position
388 (FGFR4 R388). This particular allele is linked to increased cancer cell invasion and poor
patient prognosis, and may obtain unexpected benefit from methods of the invention. The
cancer cells may have an FGFR4 R288 allele due to mutation localized to the cancer or due
to inheritance of the allele. Selecting for treatment a cancer patient with one or more FGFR4
R388 alleles in the cancer is specifically contemplated as an aspect of the invention.
[0011] The following numbered paragraphs each succinctly define one or more exemplary
variations of the invention:

[0012] 1. An isolated antibody or antibody fragment thereof that binds an extracellular
epitope of a fibroblast growth factor receptor-4 (FGFR4) that is expressed by mammalian
cells and inhibits cancer cell invasion.
[0013] 2. An isolated polypeptide that comprises a fragment of an antibody that binds an
extracellular epitope of a fibroblast growth factor receptor-4 (FGFR4) that is expressed by
mammalian cells, wherein the antibody and the polypeptide inhibit cancer cell invasion.
[0014] 3. An isolated antibody or antibody fragment thereof that bind an extracellular
epitope of a fibroblast growth factor receptor-4 (FGFR4) on mammalian cells that express
FGFR4 R388 (SEQ ID NO: 2) and inhibits fibroblast growth factor 2 (FGF2)-induced
phosphorylation of FGFR4 in the cells.
[0015] 4. An isolated polypeptide that comprises a fragment of an antibody that bind an
extracellular epitope of a fibroblast growth factor receptor-4 (FGFR4) on mammalian cells
that express FGFR4 R388 (SEQ ID NO: 2), wherein the antibody and the polypeptide inhibit
fibroblast growth factor 2 (FGF2)-induced phosphorylation of FGFR4 in the cells.
[0016] 5. An isolated antibody or antibody fragment thereof that bind an extracellular
epitope of a fibroblast growth factor receptor-4 (FGFR4) on mammalian cells that co-express
FGFR4 R388 (SEQ ID NO: 2) and fibroblast growth factor receptor-1 (FGFR1), wherein the
antibody or fragment enhances fibroblast growth factor 2 (FGF2)-induced FGFR1
degradation in the cells.
[0017] 6. An isolated polypeptide that comprises a fragment of an antibody that bind an
extracellular epitope of a fibroblast growth factor receptor-4 (FGFR4) on mammalian cells
that co-express FGFR4 R388 (SEQ ID NO: 2) and fibroblast growth factor receptor-1
(FGFR1), wherein the antibody and the polypeptide enhance fibroblast growth factor 2
(FGF2)-induced FGFR1 degradation in the cells.
[0018] 7. The antibody, antibody fragment, or polypeptide of paragraph 5 or 6 that also
inhibits FGF2-induced phosphorylation of FGFR4 in the cells.
[0019] 8. An isolated antibody or antibody fragment thereof that bind an extracellular
epitope of a fibroblast growth factor receptor-4 (FGFR4) on mammalian cells that co-express
FGFR4 R388 (SEQ ID NO: 2) and membrane type-1 metalloproteinase (MT1-MMP),
wherein the antibody or fragment inhibits complex formation between FGFR4 and MT1-
MMP in the cells.
[0020] 9. An isolated polypeptide that comprises a fragment of an antibody that binds an
extracellular epitope of a fibroblast growth factor receptor-4 (FGFR4) on mammalian cells
that co-express FGFR4 R388 (SEQ ID NO: 2) and membrane type-1 metalloproteinase
(MT1-MMP), wherein the antibody and the polypeptide inhibit complex formation between
FGFR4 and MTl-MMPin the cells.
[0021] 10. The antibody, antibody fragment, or polypeptide of any one of paragraphs 3-9
that inhibits cancer cell invasion.
[0022] 11. The antibody, antibody fragment, or polypeptide of any one of paragraphs 1 -
10, wherein the antibody binds an FGFR4 that comprises the amino acid sequence of SEQ ID
NO:l.
[0023] 12. The antibody, antibody fragment, or polypeptide of any one of paragraphs 1-
10, wherein the antibody binds an FGFR4 that comprises the amino acid sequence of SEQ ID
NO: 2.
[0024] 13. The antibody, antibody fragment, or polypeptide of any one of paragraphs 1-
12, wherein the antibody binds an FGFR4 peptide that consists of an amino acid sequence
selected from the group consisting of SEQ ID NOS: 5-9.
[0025] 14. The antibody, antibody fragment, or polypeptide of paragraph 13 that binds an
FGFR4 peptide consisting of SEQ ID NO: 7.
[0026] 15. The antibody, antibody fragment, or polypeptide of paragraph 13, wherein the
antibody or antibody fragment binds an epitope of SEQ ID NO: 1 or 2 that comprises amino
acid residues 79-81.
[0027] 16. An isolated antibody or fragment thereof that binds an epitope of FGFR4 that
is bound by monoclonal antibody F90-10C5.
[0028] 17. The polypeptide of any one of paragraphs 2, 4, 6, and 9, wherein the antibody
is monoclonal antibody F90-10C5.
[0029] 18. The antibody fragment or polypeptide of any one of paragraphs 1-17, wherein
the antibody fragment is an ScFv, Fv, Fab1, Fab, diabody, or F(abr)2 antigen-binding fragment
of an antibody.
[0030] 19. An isolated antibody, antibody fragment, or polypeptide that comprises all
complementarity determine regions (CDR) of monoclonal antibody F90-10C5, wherein the
antibody, antibody fragment, or polypeptide binds an extracellular epitope of FGFR4 that is
expressed by mammalian cells.
[0031] 20. The antibody, antibody fragment, or polypeptide of paragraph 19 that
comprises the variable regions of monoclonal antibody F90-10C5.
[0032] 21. The antibody, antibody fragment, or polypeptide of any one of paragraphs 1-
20, wherein the antibody or antibody fragment inhibits invasion of MDA-MB-231 cells
expressing FGFR4 R388 protein in a tumor cell invasiveness assay.
[0033] 22. The antibody, antibody fragment, or polypeptide of paragraph 21, wherein the
antibody or antibody fragment reduces cell invasion in a tumor cell invasiveness assay by at
least 25%.
[0034] 23. The antibody, antibody fragment, or polypeptide of paragraph 21, wherein the
antibody or antibody fragment reduces cell invasion in a tumor cell invasiveness assay by at
least 50%.
[0035] 24. The antibody of paragraph 20 that is monoclonal antibody F90-10C5.
[0036] 25. An isolated antibody, antibody fragment, or polypeptide that comprises all
complementarity determine regions (CDR) of monoclonal antibody F85-6C5, wherein the
antibody, antibody fragment, or polypeptide binds an extracellular epitope of FGFR4 that is
expressed by mammalian cells.
[0037] 26. The antibody, antibody fragment, or polypeptide of paragraph 25 that
comprises the variable regions of monoclonal antibody F85-6C5.
[0038] 27. The antibody of paragraph 26 that is monoclonal antibody F85-6C5.
[0039] 28. An isolated antibody, antibody fragment, or polypeptide that comprises all
complementarity determine regions (CDR) of monoclonal antibody F90-3B6, wherein the
antibody, antibody fragment, or polypeptide binds an extracellular epitope of FGFR4 that is
expressed by mammalian cells.
[0040] 29. The antibody, antibody fragment, or polypeptide of paragraph 28 that
comprises the variable regions of monoclonal antibody F90-3B6.
[0041] 30. The antibody of paragraph 29 that is monoclonal antibody F90-3B6.
[0042] 31. The antibody or antibody fragment of any one ofparagraphs 1, 3, 5, 7, 8, 10-
15, and 21-23 wherein the antibody is a monoclonal antibody.
[0043] 32. The antibody or antibody fragment of any one of paragraphs 1, 3, 5, 7, 8, 10-
15, and 20-31, wherein the antibody is a humanized antibody, a human antibody, or a
chimeric antibody.
[0044] 33. A humanized antibody that comprises the variable regions of mAb F90-10C5,
F85-6C5, or F90-3B6 or a fragment of any of the foregoing that binds FGFR4.
[0045] 34. The antibody, antibody fragment, or polypeptide of any one of paragraphs 1-
33, further comprising an anti-neoplastic or cytotoxic agent conjugated or attached thereto.
[0046] 35. The antibody, antibody fragment, or polypeptide of paragraph 34, wherein the
anti-neoplastic agent comprises a radionucleotide.
[0047] 36. A composition comprising the antibody, antibody fragment, or polypeptide of
any one of paragraphs 1-35 and a physiologically acceptable carrier.
[0048] 37. The composition of paragraph 36, further comprising a standard of care anti-
cancer therapeutic compound.
[0049] 38. The composition of paragraph 36 or 37, further comprising an agent that
inhibits VEGF-D or VEGF-C stimulation of VEGFR-3 or VEGFR-2.
[0050] 39. The composition of paragraph 38, wherein the agent comprises a member
selected from the group consisting of:
antibodies or antibody fragments that bind to VEGF-C, VEGF-D, or the extracellular
domain of VEGFR-3 or VEGFR-2;
a soluble protein comprising a VEGFR-3 extracellular domain or fragment thereof
effective to bind VEGF-C or VEGF-D; and
a soluble protein comprising a VEGFR-2 extracellular domain or fragment thereof
effective to bind VEGF-C or VEGF-D.
[0051] 40. The composition of any one of paragraphs 36-38, wherein the antibody,
antibody fragment, or polypeptide is a monoclonal antibody or fragment thereof ("the first
monoclonal antibody or fragment thereof).
[0052] 41. The composition of paragraph 40, further comprising a second monoclonal
antibody or fragment thereof that binds a second epitope of FGFR4 that is different than the
epitope recognized by the first monoclonal antibody or fragment thereof.
[0053] 42. The composition of paragraph 41, wherein the second monoclonal antibody or
fragment thereof is a human or humanized antibody.
[0054] 43. The composition of any one of paragraphs 36-42, further comprising a
membrane type-1 metalloproteinase (MT1-MMP) inhibitor.
[0055] 44. The composition of paragraph 43, wherein the MT1-MMP inhibitor is an
antibody or fragment thereof that binds MT1-MMP or a small molecule inhibitor of MT1-
MMP.
[0056] 45. The composition of paragraph 43, wherein the MT1-MMP inhibitor is an
inhibitor nucleic acid that hybridizes with MT1-MMP genomic DNA or mRNA and inhibits
MT1-MMP transcription or translation.
[0057] 46. An isolated polynucleotide that comprises a nucleotide sequence that encodes
the antibody, antibody fragment, or polypeptide of any one of paragraphs 1-33.
[0058] 47. A vector that comprises the polynucleotide of paragraph 46.
[0059] 48. The vector of paragraph 47 that is an expression vector.
[0060] 49. The vector of paragraph 48 that is a replication-deficient viral vector.
[0061] 50. A composition comprising the vector of paragraph 49 and a physiologically
acceptable carrier.
[0062] 51. An isolated cell transformed or transfected with the polynucleotide or vector
of any one of paragraphs 46-49.
[0063] 52. An isolated cell that produces the antibody, antibody fragment, or polypeptide
of any one of paragraphs 1-33.
[0064] 53. A hybridoma that produces the monoclonal antibody or antibody fragment of
any one of paragraphs 24, 27, and 30-32.
[0065] 54. A method of modulating invasion, ingrowth, or metastasis of cancer cells,
wherein the method comprises contacting a population of cancer cells with an antibody,
antibody fragment, polypeptide, polynucleotide, or composition of any one of paragraphs 1-
50, in an amount effective to modulate cancer cell invasion, ingrowth, or metastasis.
[0066] 55. The method of paragraph 54, wherein the cancer cells are in a mammalian
subject, and the contacting comprises administering the composition to the mammalian
subject.
[0067] 56. A method of treating a mammalian subject comprising:
selecting for treatment a mammalian subject diagnosed with or treated for cancer; and
administering to the subject the composition of any one of paragraphs 36-45 and 50,
in an amount effective to modulate cancer cell invasion, ingrowth, or metastasis.
[0068] 57. The method of paragraph 55 or 56, wherein (i) the composition comprises an
antibody, antibody fragment, or polypeptide according to any one of paragraphs 13-17, and
wherein the method further comprises administering to the mammalian subject an
antibody or fragment thereof that binds a second epitope of FGFR4 that is different than the
epitope recognized by the antibody, antibody fragment, or polypeptide of the composition.
[0069] 58. The method of paragraph 57, wherein the antibody or fragment thereof that
binds a second epitope is mAb F90-3B6 or a fragment thereof.
[0070] 59. The method of paragraph 55 or 56, wherein the composition is a composition
according to any one of paragraphs 36-42, and wherein the method further comprises
administering to the mammalian subject a composition comprising a membrane type-1
metalloproteinase (MT1-MMP) inhibitor.
[0071] 60. The method of paragraph 59, wherein the MT1-MMP inhibitor is an antibody
or fragment thereof that binds MT1-MMP or a small molecule inhibitor of MT1-MMP.
[0072] 61. The method of paragraph 55 or 56, wherein the composition is a composition
according to any one of paragraphs 36-37 and 40-42, and wherein the method further
comprises administering to the mammalian subject a composition that comprises an agent
that inhibits VEGF-D or VEGF-C stimulation of VEGR-3 or VEGFR-2.
[0073] 62. The method of any one of paragraphs 55-61, wherein the method further
comprises administering to the mammalian subject a standard of care anti-cancer therapy.
[0074] 63. A method of treating cancer in a subject, wherein the method comprises
administering to the subject the composition of any one of paragraphs 36-45 and 50, in an
amount effective to treat cancer.
[0075] 64. Use of an antibody, antibody fragment, polypeptide, polynucleotide, or
composition of any one of paragraphs 1-50 to inhibit invasion, ingrowth, or metastasis of
cancer cells in a mammalian subject.
[0076] 65. The use according to paragraph 64 in combination with a MTl-MMP inhibitor
or an inhibitor of VEGF-C or VEGF-D binding to VEGFR-3 or VEGFR-2 to inhibit invasion,
ingrowth, or metastasis of cancer cells in a mammalian subject.
[0077] 66. The use according to paragraph 64 or 65, wherein (i) the composition
comprises an antibody, antibody fragment, or polypeptide according to any one of paragraphs
13-17, used in combination with an antibody or fragment thereof that binds a second epitope
of FGFR4 that is different than the epitope recognized by the antibody, antibody fragment, or
polypeptide of the composition.
[0078] 67. The method or use of any one of paragraphs 54-66, wherein the subject is
human.
[0079] 68. The method or use of any one of paragraphs 54-67, wherein the cancer is
selected from the group consisting of breast cancer, bladder cancer, melanoma, prostate
cancer, mesothelioma, lung cancer, testicular cancer, thyroid cancer, squamous cell
carcinoma, glioblastoma, neuroblastoma, uterine cancer, colorectal cancer, and pancreatic
cancer.
[0080] 69. An isolated polynucleotide that comprises a nucleotide sequence that encodes
at least one amino acid sequence selected from the group consisting of an antibody heavy
chain variable region (Vh) and an antibody light chain variable region (Vl), wherein the Vh
and the Vl comprise complementarity determine regions (CDR) identical to monoclonal
antibody F90-10C5 CDRs.
[0081] 70. A vector that comprises a polynucleotide according to paragraph 69.
[0082] 71. A cell comprising a polynucleotide according to paragraph 69 or a vector
according to paragraph 70, wherein (a) the cell expresses an antibody or antibody fragment
containing the Vh and the Vl, and (b) the antibody or antibody fragment binds FGFR4.
[0083] 72. A method of selecting an antibody or antibody fragment, wherein the method
comprises:
(a) obtaining one or more antibodies or antibody fragments that bind FGFR4;
(b) screening the antibodies or antibody fragments in a tumor cell invasiveness assay;
and
(c) selecting an antibody that inhibits invasiveness in the assay by at least 50%.

[0084] 73. The method of paragraph 72, wherein (b) comprises detecting invasion of
tumor cells expressing FGFR4 in a three-dimensional collagen invasion assay using
fibroblast growth factor-2 as a chemoattractant.
[0085] 74. An isolated antibody or antibody fragment selected by the method of
paragraph 72 or paragraph 73.
[0086] 75. A hybridoma cell line deposited under Deutsche Sammlung von
Mikroorganismen und Zellkulturen GmbH (DSMZ) Deposit Accession Number DSM
ACC2967.
[0087] 76. A hybridoma cell line deposited under DSMZ Deposit Accession Number
DSM ACC2966.
[0088] 77. A hybridoma cell line deposited under DSMZ Deposit Accession Number
DSM ACC2965.
[0089] 78. An isolated cell capable of producing antibody mAb F90-3B6.
[0090] 79. The isolated cell of paragraph 78, wherein the cell is hybridoma F90-3B6
(DSMZ Deposit Accession Number DSM ACC2965).
[0091] 80. An isolated cell capable of producing antibody mAb F90-10C5.
[0092] 81. The isolated cell of paragraph 80, wherein the cell is hybridoma F90-10C5
(DSMZ Deposit Accession Number DSM ACC2967).
[0093] 82. An isolated cell capable of producing antibody mAb F85-6C5.
[0094] 83. The isolated cell of paragraph 82, wherein the cell is hybridoma F85-6C5
(DSMZ Deposit Accession Number DSM ACC2966).
[0095] 84. An isolated antigenic peptide consisting of 5-25 amino acids of the amino acid
sequence encoding FGFR4, wherein the peptide comprises the amino acid sequence set forth
in any one of SEQ ID NOs: 5-9 or a fragment thereof.
[0096] 85. An isolated polynucleotide encoding the antigenic peptide of paragraph 84.
[0097] 86. A vector comprising the polynucleotide of paragraph 85.
[0098] 87. An isolated cell comprising the vector of paragraph 86.
[0099] 88. A composition comprising the peptide of paragraph 84 and an adjuvant.

[0100] 89. The method of any one of paragraphs 56-63, comprising a step of determining
the presence or absence of an FGFR4 allele that encodes FGFR4 R388 in the cancer, wherein
the treatment is administered if the cancer has at least one FGFR4 allele that encodes FGFR4
R388.
[0101] 90. A method of treating a mammalian subject comprising: selecting for treatment
a mammalian subject diagnosed with or treated for cancer, wherein the cancer includes cells
that contain at least one FGFR4 allele that encodes FGFR4 R388; and administering to the
subject the composition of any one of claims 36-45 and 50, in an amount effective to
modulate cancer cell invasion, ingrowth, or metastasis.
[0102] 91. The method of paragraphs 89 or 90, wherein the presence or absence of an
FGFR4 R388 allele is determined by assaying FGFR4 protein with an antibody or antibody
fragment that differentially binds FGFR4 R388 and G388 alleles.
[0103] 92. The method of paragraphs 89 or 90, wherein the presence or absence of an
FGFR4 R388 allele is determined by assaying nucleic acid from the subject or from the
cancer.
[0104] 93. A method of treating a mammalian subject comprising: selecting for treatment
a mammalian subject diagnosed with or treated for cancer; and administering to the subject a
first anti-FGFR4 antibody or FGFR4-binding fragment thereof and a second anti-FGFR4
antibodies or FGFR4-binding fragment thereof, wherein the first anti-FGFR4 antibody or
fragment inhibits FGF2-induced phosphorylation of FGFR4 R388, and wherein the second
anti-FGFR4 antibody or fragment inhibits ligand-independent FGFR4 phosphorylation.
[0105] 94. The method of paragraph 93, wherein the first and second antibodies or
fragments thereof are separately administered, either simultaneously or sequentially, as a first
composition containing the first antibody or fragment and a second composition containing
the second antibody or fragment.
[0106] 95. The method according to any one of paragraphs 90-94, further comprising
administering to the subject a standard of care chemotherapy for the cancer.
[0107] The foregoing summary is not intended to define every aspect of the invention, and
additional aspects are described in other sections, such as the Detailed Description. The
entire document is intended to be related as a unified disclosure, and it should be understood
that all combinations of features described herein are contemplated, even if the combination
of features are not found together in the same sentence, or paragraph, or section of this
document. Where protein (e.g., antibody) therapy is described, embodiments involving
polynucleotide therapy (using polynucleotides/vectors that encode the protein) are
specifically contemplated, and the reverse also is true. Where embodiments of the invention
are described with respect to a specific antibody, such as an FGFR4 monoclonal antibody, it
should be appreciated that analogous embodiments involving antibody fragments, variants,
and the like are specifically contemplated.
[0108] In addition to the foregoing, the invention includes, as an additional aspect, all
embodiments of the invention narrower in scope in any way than the variations specifically
mentioned above. With respect to aspects of the invention described as a genus, all
individual species are individually considered separate aspects of the invention. With respect
to aspects of the invention described or claimed with "a" or "an," it should be understood that
these terms mean "one or more" unless context unambiguously requires a more restricted
meaning. With respect to elements described as one or more within a set, it should be
understood that all combinations within the set are contemplated.
[0109] Although the applicant(s) invented the full scope of the claims appended hereto, the
claims appended hereto are not intended to encompass within their scope the prior art work of
others. Therefore, in the event that statutory prior art within the scope of a claim is brought
to the attention of the applicants by a Patent Office or other entity or individual, the
applicant(s) reserve the right to exercise amendment rights under applicable patent laws to
redefine the subject matter of such a claim to specifically exclude such statutory prior art or
obvious variations of statutory prior art from the scope of such a claim. Variations of the
invention defined by such amended claims also are intended as aspects of the invention.
Additional features and variations of the invention will be apparent to those skilled in the art
from the entirety of this application, and all such features are intended as aspects of the
invention.
DESCRIPTION OF THE FIGURES
[0110] Figure 1 is a graph illustrating the relative level of MMP2 activation (active/latent)
(Y-axis) for various kinases (listed on X-axis) demonstrating greater than 2-fold induced
MMP2 activation relative to mock transfected control cells.
[0111] Figure 2 is a graph comparing absorbance (Y-axis) caused by ligand-receptor
binding and concentration (nM) of potential binding blockers (X-axis).
[0112] Figure 3A-C are illustrations of alternative views of the three dimensional structure
of dimerized FGFR4, depicting the location of the epitope region bound by antibody F90-
10C5 (SEQ ID NOs: 5-9) as beaded regions.
[0113] Figures 4A-4C are graphs comparing response units (Y-axis) and concentration
(nM) of mAb 10C5 (also referred to herein as Ab F90-10C5) (Figure 4A), mAb 6C5 (also
referred to herein as Ab F85-6C5) (Figure 4B), and mAb 3B6 (also referred to herein as Ab
F90-3B6) (Figure 4C) (X-axis).
[0114] Figure 5 is a graph summarizing the number of collagen invasion foci (Y-axis) with
treatment by control antibody, mAb 10C5, mAb 6C5, and mAb 3B6 (X-axis).
[0115] Figure 6 depicts an immunoblot prepared from MDA-MB-231 cells transfected
with expression vectors encoding V5-tagged FGFR4 R388 (FGFR4 R). The cells were
pretreated with mAb F90-3B6, mAb F90-10C5, or a combination of mAb F90-3B6 and mAb
F90-10C5 overnight, and left unstimulated (-) or incubated with FGF2 (+). Cell extracts were
immunoprecipitated with antibodies against FGFR4 (IP:FGFR4) and immunoblotted using
antibodies against the V5 tag (IB: V5) or phosphotyrosine residues (IB: pY).
[0116] Figure 7 depicts an immunoblot prepared from COS-1 cells transfected with
expression vectors encoding V5-tagged FGFR4 G388 ("FGFR4 G" or "FR4 G"), V5-tagged
FGFR4 R388 ("FGFR4 R" or "FR4 R"), and FGFR1, alone or in combination. The cells
were pretreated with mAb F85-6C5 or mAb F90-10C5, and incubated with FGF2 (+) or left
unstimulated (-). Cell extracts were immunoprecipitated using anti-FGFR4 antibodies
(IP:FGFR4) and immunoblotted using anti-phosphotyrosine antibodies (IB:pY), antibodies
against FGFR1 (B:FGFR1), or antibodies against the V5 tag (IB:V5). mAb F90-10C5
treatment inhibited FGF2-induced FGFR4 R388 phosphorylation and FGFR1
downregulation, whereas mAb F85-6C5 reduced ligand independent FGFR4 phosphorylation
and FGFR4/FGFR1 heterodimerization after FGF2 stimulation.
DETAILED DESCRIPTION OF THE INVENTION
[0117] The invention relates, at least in part, to the unexpected identification of certain
anti-fibroblast growth factor receptor 4 (FGFR4) antibodies that inhibit invasion of cancer
cells into surrounding tissue. While not wishing to be bound to a particular mechanism of
action, the inventors surprisingly determined that FGFR4 is functionally linked with human
membrane-type matrix metalloproteinase 1 (MT1-MMP), which is expressed in cancer and
reactive cells. MT1-MMP is largely sequestered in tissue microenvironments, enabling it to
escape inactivation by many known MMP inhibitors. FGFR4 represents a novel target for
MTl-MMP-mediated metastatic events. The invention provides an isolated antibody or
fragment thereof that binds an FGFR4 and inhibits or downregulates cancer cell invasion, as
well as methods of using the antibody or fragment thereof to modulate invasion, ingrowth, or
metastasis of cancer cells. In this manner, antibodies of the invention have therapeutic utility
to slow cancer progression.
FGFR4
[0118] FGFR4 is one of four transmembrane receptor tyrosine kinases activated by FGF
(Givol et al, FASEB J., 6: 3362-3369, 1992). The receptor is composed of three
immunoglobulin (Ig)-like extracellular domains, a transmembrane domain, a tyrosine kinase,
and a COOH-terminal tail (Givol et al., supra). At least two known human FGFR4 amino
acid sequences have been reported, differing as a result of a mutation affecting codon 388,
resulting in either a glycine (FGFR4 G388, SEQ ID NO: 1) or an arginine (FGFR4 R388,
SEQ ID NO: 2) at this position. The allele with the R388 mutation correlates with aggressive
tumor progression and is considered an indicator of poor clinical outcome (see, e.g., Bange et
al, Cancer Res., 62(3): 840-7, 2002). The mutation results in substitution of a hydrophobic
with a hydrophilic amino acid in the transmembrane domain of the protein. In this regard, a
wild-type FGFR4 transmembrane domain comprises approximately the amino acid sequence
RYTDIILYASGSLALAVLLLLAGLY (SEQ ID NO: 3), while the R388 mutant comprises a
transmembrane domain comprising approximately the amino acid sequence
RYTDIILYASGSLALAVLLLLARLY (SEQ ID NO: 4). The R388 FGFR4 mutant is further
described in, e.g., Bange et al., supra; and U.S. Patent 6,770,742, incorporated herein by
reference.
Antibodies and Fragments Thereof
[0119] Some embodiments or aspects of the invention relate to an antibody or fragment
thereof that binds an extracellular epitope of FGFR4 (comprising the amino acid sequence of
any FGFR4 polypeptide, including any naturally occurring isoforms or allelic variants of
FGFR4) and inhibits cancer cell invasion. For example, the antibody or fragment thereof
binds an FGFR4 expressed on the surface of a mammalian (e.g., human) cell. The antibody

or fragment thereof can bind to an FGFR4 comprising the amino acid sequence set forth in
SEQ ID NO: 1, which is commonly referred to as the FGFR4 G388 allele. Alternatively or in
addition, the antibody or fragment thereof can bind an FGFR4 wherein the glycine at position
388 of wild-type FGFR4 is substituted with an arginine (the FGFR4 R388 allele) (SEQ ID
NO: 2).
[0120] Preferably, the antibody or fragment thereof binds an extracellular epitope of a
FGFR4. The three immunoglobulin (Ig)-like domains of FGFR4's extracellular region
extend beyond the transmembrane domain into the extracellular space and, with reference to
SEQ ID NOs: 1 and 2, comprise approximately amino acid residues 25-369 of the FGFR4
amino acid sequence. In certain aspects of the invention, the antibody or fragment thereof
binds an extracellular epitope located in the region of the extracellular domain spanning
amino acid residues 25-366, such as the first immunoglobulin-like domain of the extracellular
region of FGFR4 (spanning approximately amino acid residues 50-107 of the FGFR4 amino
acid sequence). The extracellular domain of FGFR4 is further described in Loo et al., Int. J.
Biochem. Cell Biol., 32: 489-97, 2000; and Sorenson et al, J. Cell. Sci., 117:1807-1819,
2004, the disclosures of which pertaining to FGFR4 are hereby incorporated by reference.
[0121] Any type of antibody is suitable in the context of the invention, including
polyclonal, monoclonal, chimeric, humanized, or human versions having full length heavy
and/or light chains. The invention also includes antibody fragments (and/or polypeptides that
comprise antibody fragments) that retain FGFR4 binding characteristics of FGFR4 antibodies
of the invention. Antibody fragments include antigen-binding regions and/or effector regions
of the antibody, e.g., F(ab')2, Fab, Fab', Fd, Fc, and Fv fragments (fragments consisting of the
variable regions of the heavy and light chains that are non-covalently coupled), or single-
domain antibodies (nanobodies). In general terms, a variable (V) region domain may be any
suitable arrangement of immunoglobulin heavy (Vh) and/or light (Vl) chain variable
domains. Thus, for example, the V region domain may be dimeric and contain Vh-Vh, Vh-
Vl, or Vl-Vl dimers that bind FGFR4. If desired, the Vh and Vl chains may be covalently
coupled either directly or through a linker to form a single chain Fv (scFv). For ease of
reference, scFv proteins are referred to herein as included in the category "antibody
fragments." Similarly, antibody fragments may be incorporated into single domain
antibodies, maxibodies, minibodies, intrabodies, diabodies, triabodies, tetrabodies, variable
domains of new antigen receptors (v-NAR), and bis-single chain Fv regions (see, e.g.,
Hollinger and Hudson, Nature Biotechnology, 23(9): 1126-1136, 2005) to inhibit cancer cell
invasion. Additionally, the invention provides an isolated polypeptide that comprises a
fragment of an antibody that binds an extracellular epitope of a fibroblast growth factor
receptor-4 (FGFR4) that is expressed by mammalian cells. If desired, the antibody fragment
is fused to a moiety with effector function (e.g., cytotoxic activity, immune recruitment
activity, and the like), a moiety that facilitates isolation from a mixture (e.g., a tag), a
detection label, or the like. It will be appreciated that the features of the inventive antibody
or fragment thereof described herein extend also to a polypeptide comprising the antibody
fragment.
[0122] The antibody or antibody fragment can be isolated from an immunized animal,
synthetic, or genetically-engineered. Antibody fragments derived from an antibody can be
obtained, e.g., by proteolytic hydrolysis of the antibody. For example, papain or pepsin
digestion of whole antibodies yields a 5S fragment termed F(ab')2 or two monovalent Fab
fragments and an Fc fragment, respectively. F(ab)2 can be further cleaved using a thiol
reducing agent to produce 3.5S Fab monovalent fragments. Methods of generating antibody
fragments are further described in, for example, Edelman et aL, Methods in Enzymology, 1:
422 Academic Press (1967); Nisonoff et aL, Arch. Biochem. Biophys., 89: 230-244, 1960;
Porter, Biochem. J., 73: 119-127, 1959; U.S. Patent No. 4,331,647; and by Andrews, S.M.
and Titus, J.A. in Current Protocols in Immunology (Coligan et al., eds), John Wiley & Sons,
New York (2003), pages 2.8.1-2.8.10 and 2.10A.1-2.10A.5.
[0123] An antibody or fragment thereof also can be genetically engineered such that the
antibody or antibody fragment comprises, e.g., a variable region domain generated by
recombinant DNA engineering techniques. For example, a specific antibody variable region
can be modified by insertions, deletions, or changes in or to the amino acid sequences of the
antibody to produce an antibody of interest. In this regard, polynucleotides encoding
complementarity determining regions (CDRs) of interest are prepared, for example, by using
polymerase chain reaction to synthesize variable regions using mRNA of antibody-producing
cells as a template (see, for example, Courtenay-Luck, "Genetic Manipulation of Monoclonal
Antibodies," in Monoclonal Antibodies: Production, Engineering and Clinical Application,
Ritter et aL (eds.), page 166 (Cambridge University Press 1995); Ward et aL, "Genetic
Manipulation and Expression of Antibodies," in Monoclonal Antibodies: Principles and
Applications, Birch et aL, (eds.), page 137 (Wiley-Liss, Inc. 1995); and Larrick et aL,
Methods: A Companion to Methods in Enzymology, 2: 106-110, 1991). Current antibody
manipulation techniques allow construction of engineered variable region domains containing

at least one CDR and, optionally, one or more framework amino acids from a first antibody
and the remainder of the variable region domain from a second antibody. Such techniques
are used, e.g., to humanize an antibody or to improve its affinity for a binding target.
[0124] "Humanized antibodies" are recombinant proteins in which complementary
determining regions of monoclonal antibodies have been transferred from heavy and light
variable chains of non-human immunoglobulin into a human variable domain. Constant
regions need not be present, but if they are, they optionally are substantially identical to
human immunoglobulin constant regions, i.e., at least about 85-90%, about 95% or more
identical, in certain embodiments. Hence, in some instances, all parts of a humanized
immunoglobulin, except possibly the CDR's, are substantially identical to corresponding
parts of natural human immunoglobulin sequences. For example, in one aspect, humanized
antibodies are human immunoglobulins (host antibody) in which hypervariable region
residues of the host antibody are replaced by hypervariable region residues from a non-
human species (donor antibody) such as mouse, rat, rabbit, or a non-human primate having
the desired specificity, affinity, and capacity. Humanized antibodies such as those described
herein can be produced using techniques known to those skilled in the art (Zhang et al.,
Molecular Immunology, 42(12): 1445-1451, 2005; Hwang et al, Methods, 36(1): 35-42,
2005; Dall'Acqua et al., Methods, 56(1): 43-60, 2005; Clark, Immunology Today, 2/(8): 397-
402,2000, and U.S. Patent Nos. 6,180,370; 6,054,927; 5,869,619; 5,861,155; 5,712,120; and
4,816,567, all of which are all hereby expressly incorporated herein by reference).
[0125] In one embodiment, the antibody is a human antibody, such as, but not limited to,
an antibody having variable regions in which both the framework and CDR regions are
derived from human germline immunoglobulin sequences as described, for example, in Kabat
et al. (1991) Sequences of proteins of Immunological Interest, Fifth Edition, U.S. Department
of Health and Human Services, NIH Publication No. 91-3242. If the antibody contains a
constant region, the constant region also preferably is derived from human germline
immunoglobulin sequences. Human antibodies may comprise amino acid residues not
encoded by human germline immunoglobulin sequences to, e.g., enhance the activity of the
antibody, but do not comprise CDRs derived from other species (i.e., a mouse CDR placed
within a human variable framework region).
[0126] The antibody or fragment thereof binds any region of FGFR4 so long as cancer cell
invasion is inhibited (or reduced) and/or one or more of the other desired activity parameters
is retained. In one embodiment, the invention further provides an isolated antibody or
antibody fragment that binds an epitope of FGFR4 that is bound by monoclonal antibody
(mAb) F90-10C5 (also referred to herein as "10C5"), further described in the Examples
below. mAb F90-10C5's binding activity is localized to the first immunoglobulin-like
domain of the extracellular region of FGFR4 (see Figures 3A-3C). Surprisingly, mAb F90-
10C5 recognizes a linear epitope within approximately amino acids 67-93 of the FGFR4
amino acid sequence, as determined by an immunoblotting array composed of a series of 15
amino acid fragments (the sequences of which overlapped by three amino acids) that spanned
amino acids 67-93 of FGFR4's extracellular domain (excluding signal sequence). mAb F90-
10C5 binds the following FGFR4 fragments: YKEGSRLAPAGRVRG (SEQ ID NO: 5);
GSRLAPAGRVRGWRG (SEQ ID NO: 6); LAPAGRVRGWRGRLE (SEQ ID NO: 7);
AGRVRGWRGRLEIAS (SEQ ID NO: 8); and VRGWRGRLEIASFLP (SEQ ID NO: 9).
The isolated antibody or fragment thereof preferably binds a peptide comprising any one or
more of the amino acid sequences set forth in SEQ ID NOs: 5-9. More preferably, the
isolated antibody or fragment thereof binds a peptide comprising (or consisting of) the amino
acid sequence of SEQ ED NO: 7.
[0127] Other antibodies that bind FGFR4 and inhibit cancer cell invasion also are suitable
in the context of the invention. For example, in various embodiments, the invention includes
administering an antibody or fragment thereof that (i) competes for binding with mAb F90-
10C5, (ii) binds the region of FGFR4 recognized by mAb F90-10C5, or (iii) binds at or near
amino acid residues 67-93 (e.g., amino acids residues 73-87 or amino acid residues 79-81) of
the FGFR4 extracellular region, while inhibiting cancer cell invasion. If desired, the antibody
fragment comprises all or part of the antigen-binding elements of an antibody, such as mAb
F90-10C5, including the variable region of mAb F90-10C5 (or any other antibody of the
invention). The antibody fragment can comprise all or part of the antigen-binding elements
of an antibody while lacking all or part of the framework regions of an antibody. In this
regard, the isolated antibody or fragment thereof comprises one, two, three, four, five, or six
(i.e., all) complementary determining regions (CDRs) of an FGFR4-binding antibody that
inhibits cancer cell invasion, e.g., mAb F90-10C5. Methods of identifying complementarity
determining regions and specificity determining regions are known in the art and further
described in, for example, Tamura et al., J. Immunol, 164: 1432-1441, 2000. In one
embodiment, the antibody or fragment thereof is mAb F90-10C5 or an FGFR4-binding
fragment thereof.

[0128] In the context of the invention, antibody binding refers to immuno-reacting between
the variable regions of the antibody and an antigen as distinct from other protein-protein
interactions (such as Staphylococcus aureus protein A interactions with immunoglobulins, for
example). The antibody or fragment thereof preferably preferentially binds to FGFR4,
meaning that the antibody or fragment thereof binds FGFR4 with greater affinity than it binds
to an unrelated control protein. More preferably, the antibody or fragment thereof
specifically recognizes and binds FGFR4 (or a portion thereof). "Specific binding" means
that there is essentially no cross-reactivity with an unrelated control protein. In some
variations of the invention, the antibody binds FGFR4 substantially exclusively (i.e., is able
to distinguish FGFR4 from other known polypeptides (e.g., other FGFRs) by virtue of
measurable differences in binding affinity). Depending on the embodiment, the antibody or
fragment thereof binds to FGFR4 with an affinity that is at least 5, 10, 15, 20, 25, 50, 100,
250, 500, 1000, or 10,000 times greater than the affinity for an unrelated control protein. In
other variations the antibody cross-reacts with other FGFR sequences. Screening assays to
determine binding specificity/affinity of an antibody, as well as identify antibodies that
compete for binding sites (i.e., cross-block binding of, e.g., mAb F90-10C5, to FGFR4), are
well known and routinely practiced in the art. For example, binding affinity or cross-
blocking can be determined using the methods described in the Examples. Competition
binding assays employing a Biacore machine, which measures the extent of interactions using
surface plasmon resonance technology, also are appropriate. Another suitable assay uses an
ELISA-based approach to measure competition between antibodies in terms of their binding
to FGFR4. For a comprehensive discussion of binding assays, see Harlow et al. (Eds),
Antibodies A Laboratory Manual; Cold Spring Harbor Laboratory; Cold Spring Harbor, NY
(1988), Chapter 6. In general, an antibody that "competes" with or "cross-blocks" mAb F90-
10C5 prevents the binding of mAb F90-10C5 to FGFR4 by 50% to 100% (e.g., 60%, 65%,
70%, 75%, 80%, 85%, 90%, or 95%).
Materials and Methods for Producing Antibodies and Fragments Thereof
[0129] Antibodies according to the invention can be obtained by any suitable method, such
as by immunization and cell fusion procedures as described herein and known in the art
and/or screening libraries of antibodies or antibody fragments using FGFR4 extracellular
domain epitopes described herein. Monoclonal antibodies of the invention are generated
using a variety of known techniques (see, for example, Coligan et al. (eds.), Current
Protocols in Immunology, 1:2.5.12.6.7 (John Wiley & Sons 1991); Monoclonal Antibodies,
Hybridomas: A New Dimension in Biological Analyses, Plenum Press, Kennett, McKearn,
and Bechtol (eds.) (1980); Antibodies: A Laboratory Manual, Harlow and Lane (eds.), Cold
Spring Harbor Laboratory Press (1988); and Picksley et aL, "Production of monoclonal
antibodies against proteins expressed in E. coli," in DNA Cloning 2: Expression Systems,
2nd Edition, Glover et aL (eds.), page 93 (Oxford University Press 1995)). In one
embodiment, the invention provides an isolated cell capable of producing antibody mAb F90-
3B6, mAb F90-10C5, or mAb F85-6C5. Typically, monoclonal antibodies are produced by a
hybridoma, and the invention provides a hybridoma that produces the inventive monoclonal
antibody or antibody fragment. Hybridoma cell lines that produce antibodies F90-10C5, F85-
6C5, and F90-3B6 are provided by the invention and have been deposited with the Deutsche
Sammlung von Mikroorganismen und Zellkulruren GmbH (DSMZ), Mascheroder Wep lb.
D-38124, Germany, under the provisions of the Budapest Treaty for the International
Recognition of the Deposit of Microorganisms for the Purpose of Patent Procedure
("Budapest Treaty"), and assigned Deposit Accession Nos. DSM ACC2967, DSM ACC2966,
and DSM ACC2965, respectively.
[0130] Likewise, human antibodies are generated by any of a number of techniques
including, but not limited to, Epstein Barr Virus (EBV) transformation of human peripheral
blood cells (e.g., containing B lymphocytes), in vitro immunization of human B cells, fusion
of spleen ceils from immunized transgenic mice carrying inserted human immunoglobulin
genes, isolation from human immunoglobulin V region phage libraries, or other procedures as
known in the art and based on the disclosure herein. Methods for obtaining human antibodies
from transgenic animals are further described, for example, in Bruggemann et aL, Curr. Opin.
Biotechnol., 8: 455-58,1997; Jakobovits et al., Ann. N. Y. Acad. Sci., 764: 525-35,1995;
Green et al., Nature Genet., 7: 13-21, 1994; Lonberg et al., Nature, 368: 856-859,1994;
Taylor et al., Int. Immun. 6: 579-591, 1994; and U.S. Patent No. 5,877,397.
[0131] For example, human antibodies are obtained from transgenic animals that have
been engineered to produce specific human antibodies in response to antigenic challenge.
For example, International Patent Publication No. WO 98/24893 discloses transgenic animals
having a human Ig locus, wherein the animals do not produce functional endogenous
immunoglobulins due to the inactivation of endogenous heavy and light chain loci.
Transgenic non-primate mammalian hosts capable of mounting an immune response to an
immunogen, wherein the antibodies have primate constant and/or variable regions, and

wherein the endogenous immunoglobulin encoding loci are substituted or inactivated, also
have been described. International Patent Publication No. WO 96/30498 discloses the use of
the Cre/Lox system to modify the immunoglobulin locus in a mammal, such as to replace all
or a portion of the constant or variable region to form a modified antibody molecule.
International Patent Publication No. WO 94/02602 discloses non-human mammalian hosts
having inactivated endogenous Ig loci and functional human Ig loci. U.S. Patent No.
5,939,598 discloses methods of making transgenic mice in which the mice lack endogenous
heavy chains, and express an exogenous immunoglobulin locus comprising one or more
xenogeneic constant regions. Using a transgenic animal, such as a transgenic animal
described herein, an immune response can be produced to a selected antigenic molecule, and
antibody producing cells can be removed from the animal and used to produce hybridomas
that secrete human-derived monoclonal antibodies. Immunization protocols, adjuvants, and
the like are known in the art, and are used in immunization of, for example, a transgenic
mouse as described in International Patent Publication No. WO 96/33735. The monoclonal
antibodies can be tested for the ability to inhibit or neutralize the biological activity or
physiological effect of the corresponding protein.
[0132] The invention provides materials for generating the inventive anti-FGFR4
antibodies and fragments thereof. For example, the invention provides an isolated cell (e.g., a
hybridoma) that produces the inventive antibody or antibody fragment, such as hybridoma
cell lines F90-10C5, F85-6C5, and F90-3B6 further described herein. The invention further
relates to an isolated polynucleotide encoding the inventive antibody or antibody fragment.
In one aspect of the invention, the isolated polynucleotide comprises a nucleotide sequence
that encodes an antibody heavy chain variable region (Vh) and/or an antibody light chain
variable region (Vl), wherein the Vh and the Vl comprise complementarity determining
regions (CDRs) identical to monoclonal antibody F90-10C5 CDRs.
[0133] In a related embodiment, the invention provides a vector (e.g., an expression
vector) comprising a polynucleotide of the invention to direct expression of the
polynucleotide in a suitable host cell. Such vectors are useful, e.g., for amplifying the
polynucleotides in host cells to create useful quantities thereof, and for expressing peptides,
such as antibodies or antibody fragments, using recombinant techniques. In preferred
embodiments, the vector is an expression vector wherein the polynucleotide of the invention
is operatively linked to a polynucleotide comprising an expression control sequence.
Autonomously replicating recombinant expression constructs such as plasmid and viral DNA

vectors incorporating polynucleotides of the invention are specifically contemplated.
Expression control DNA sequences include promoters, enhancers, and operators, and are
generally selected based on the expression systems in which the expression construct is to be
utilized. Preferred promoter and enhancer sequences are generally selected for the ability to
increase gene expression, while operator sequences are generally selected for the ability to
regulate gene expression. Expression constructs of the invention may also include sequences
encoding one or more selectable markers that permit identification of host cells bearing the
construct. Expression constructs may also include sequences that facilitate, and preferably
promote, homologous recombination in a host cell. Preferred expression constructs of the
invention also include sequences necessary for replication in a host cell.
[0134] Exemplary expression control sequences include promoter/enhancer sequences,
e.g., cytomegalovirus promoter/enhancer (Lehner et aL, J. Clin. Microbiol., 29: 2494-2502,
1991; Boshart et aL, Cell, 41: 521-530,1985); Rous sarcoma virus promoter (Davis et aL,
Hum. Gene Ther., 4:151,1993); Tie promoter (Korhonen et aL, Blood, 86(5): 1828-1835,
1995); simian virus 40 promoter; DRA (downregulated in adenoma; Alrefai et aL, Am. J.
Physiol. Gastrointest. Liver Physiol, 293: G923-G934, 2007); MCT1 (monocarboxylate
transporter 1; Cuff et aL, Am. J. Physiol. Gastrointet. Liver Physiol, G977-G979. 2005); and
Mathl (mouse atonal homolog 1; Shroyer et al., Gastroenterology, 132: 2477-2478, 2007),
for expression in the target mammalian cells, the promoter being operatively linked upstream
(i.e., 5') of the polypeptide coding sequence (the disclosures of the cited references is
incorporated herein by reference in their entirety and particularly with respect to the
discussion of expression control sequences). In another variation, the promoter is an
epithelial-specific promoter or endothelial-specific promoter. The polynucleotides of the
invention may also optionally include a suitable polyadenylation sequence (e.g., the SV40 or
human growth hormone gene polyadenylation sequence) operably linked downstream (i.e.,
3') of the polypeptide coding sequence.
[0135] If desired, the polynucleotide of the invention also optionally comprises a
nucleotide sequence encoding a secretory signal peptide fused in frame with the polypeptide
sequence. The secretory signal peptide directs secretion of the polypeptide of the invention
by the cells that express the polynucleotide, and is cleaved by the cell from the secreted
polypeptide. The polynucleotide may further optionally comprise sequences whose only
intended function is to facilitate large scale production of the vector, e.g., in bacteria, such as
a bacterial origin of replication and a sequence encoding a selectable marker. However, if the

vector is administered to an animal, such extraneous sequences are preferably at least
partially cleaved. One can manufacture and administer polynucleotides for gene therapy
using procedures that have been described in the literature for other transgenes. See, e.g.,
Isner et al, Circulation, 91: 2687-2692, 1995; and Isner et al., Human Gene Therapy, 7: 989-
1011,1996; incorporated herein by reference.
[0136] In some embodiments, polynucleotides of the invention further comprise additional
sequences to facilitate uptake by host cells and expression of the antibody or fragment thereof
(and/or any other peptide). In one embodiment, a "naked" transgene encoding an antibody or
fragment thereof described herein (i.e., a transgene without a viral, liposomal, or other vector
to facilitate transfection) is employed.
[0137] Vectors also are useful for "gene therapy" treatment regimens, wherein, for
example, a polynucleotide encoding an antibody or fragment thereof is introduced into a
subject suffering from or at risk of suffering from invasive cancers in a form that causes cells
in the subject to express the antibody or fragment thereof in vivo. Any suitable vector may be
used to introduce a polynucleotide that encodes an antibody or fragment thereof into the host.
Exemplary vectors that have been described in the literature include replication deficient
retroviral vectors, including but not limited to lentivirus vectors (Kim et al., J. Virol., 72(1):
811-816, 1998; Kingsman & Johnson, Scrip Magazine, October, 1998, pp. 43-46); parvoviral
vectors, such as adeno-associated viral (AAV) vectors (U.S. Patent Nos. 5,474,9351;
5,139,941; 5,622,856; 5,658,776; 5,773,289; 5,789,390; 5,834,441; 5,863,541; 5,851,521;
5,252,479; Gnatenko et al, J. Invest. Med., 45: 87-98, 1997); adenoviral (AV) vectors (U.S.
Patent Nos. 5,792,453; 5,824,544; 5,707,618; 5,693,509; 5,670,488; 5,585,362; Quantin et
al., Proc. Natl. Acad. Sci. USA, 89: 2581-2584,1992; Stratford Perricaudet et al., J. Clin.
Invest., 90: 626-630, 1992; and Rosenfeld et al., Cell, 68: 143-155, 1992); an adenoviral
adeno-associated viral chimeric (U.S. Patent No. 5,856,152) or a vaccinia viral or a
herpesviral vector (U.S. Patent Nos. 5,879,934; 5,849,571; 5,830,727; 5,661,033; 5,328,688);
Lipofectin mediated gene transfer (BRL); liposomal vectors (U.S. Patent No. 5,631,237); and
combinations thereof. All of the foregoing documents are incorporated herein by reference in
their entirety and particularly with respect to their discussion of expression vectors. Any of
these expression vectors can be prepared using standard recombinant DNA techniques
described in, e.g., Sambrook et al., Molecular Cloning, a Laboratory Manual, 2d edition, Cold
Spring Harbor Press, Cold Spring Harbor, N.Y. (1989), and Ausubel et al., Current Protocols
in Molecular Biology, Greene Publishing Associates and John Wiley & Sons, New York,

N.Y. (1994). Optionally, the viral vector is rendered replication-deficient by, e.g., deleting or
disrupting select genes required for viral replication.
[0138] Other non-viral delivery mechanisms contemplated include calcium phosphate
precipitation (Graham and Van Der Eb, Virology, 52: 456-467, 1973; Chen and Okayama,
Mol. Cell Biol, 7: 2745-2752, 1987; Rippe et al., Mol. Cell Biol, 10: 689-695, 1990) DEAE-
dextran (Gopal, Mol. Cell Biol, 5: 1188-1190,1985), electroporation (Tur-Kaspa et al, Mol
Cell Biol, 6: 716-718,1986; Potter et al., Proc. Nat. Acad. Sci. USA, 81: 7161-7165,1984),
direct microinjection (Harland and Weintraub, J. Cell Biol, 101: 1094-1099, 1985, DNA-
loaded liposomes (Nicolau and Sene, Biochim. Biophys. Acta, 721: 185-190, 1982; Fraley et
al., Proc. Natl Acad. Sci. USA, 76: 3348-3352, 1979; Feigner, Sci Am., 276(6): 102-6, 1997;
Feigner, Hum Gene Ther., 7(15): 1791-3, 1996), cell sonication (Fechheimer et al., Proc.
Natl. Acad. Sci. USA, 84: 8463-8467, 1987), gene bombardment using high velocity
microprojectiles (Yang et al., Proc. Natl. Acad. Sci USA, 87: 9568-9572, 1990), and receptor-
mediated transfection (Wu and Wu, J. Biol Chem., 262: 4429-4432, 1987; Wu and Wu,
Biochemistry, 27: 887-892, 1988; Wu and Wu, Adv. Drug Delivery Rev., 12: 159-167, 1993).
[0139] The expression vector (or the antibody or fragment thereof discussed herein) may
be entrapped in a liposome. Liposomes are vesicular structures characterized by a
phospholipid bilayer membrane and an inner aqueous medium. Multilamellar liposomes
have multiple lipid layers separated by aqueous medium. They form spontaneously when
phospholipids are suspended in an excess of aqueous solution. The lipid components
undergo self-rearrangement before the formation of closed structures and entrap water and
dissolved solutes between the lipid bilayers (Ghosh and Bachhawat, In: Liver diseases,
targeted diagnosis and therapy using specific receptors and ligands, Wu G, Wu C ed., New
York: Marcel Dekker, pp. 87-104 (1991)). The addition of DNA to cationic liposomes
causes a topological transition from liposomes to optically birefringent liquid-crystalline
condensed globules (Radler et al., Science, 275(5301): 810-814, 1997). These DNA-lipid
complexes are potential non-viral vectors for use in gene therapy and delivery.
[0140] Liposome-mediated nucleic acid delivery and expression of foreign DNA in vitro
has been successful. Also contemplated in the invention are various commercial approaches
involving "lipofection" technology. In certain embodiments of the invention, the liposome
may be complexed with a hemagglutinating virus (HVJ). This has been shown to facilitate
fusion with the cell membrane and promote cell entry of liposome-encapsulated DNA
(Kaneda et al, Science, 243: 375-378, 1989). In other embodiments, the liposome is
complexed or employed in conjunction with nuclear nonhistone chromosomal proteins
(HMG-1) (Kato et al., J. Biol. Chem., 266: 3361-3364, 1991). In yet further embodiments,
the liposome are complexed or employed in conjunction with both HVJ and HMG-1. Such
expression constructs have been successfully employed in transfer and expression of nucleic
acid in vitro and in vivo. In some variations of the invention, an FGFR4 targeting moiety,
such as an FGFR4 antibody or fragment, is included in the liposome to target the liposome to
cells (such as cancer cells) expressing FGFR4 on their surface.
[0141] Transferring a naked DNA expression construct into cells can be accomplished
using particle bombardment, which depends on the ability to accelerate DNA coated
microprojectiles to a high velocity allowing them to pierce cell membranes and enter cells
without killing them (Klein et al., Nature, 327: 70-73, 1987). Several devices for
accelerating small particles have been developed. One such device relies on a high voltage
discharge to generate an electrical current, which in turn provides the motive force (Yang et
al, Proc. Natl. Acad. Sci USA, 87: 9568-9572, 1990). The microprojectiles used have
consisted of biologically inert substances such as tungsten or gold beads.
[0142] In embodiments employing a viral vector, preferred polynucleotides still include a
suitable promoter and polyadenylation sequence as described above. Moreover, it will be
readily apparent that, in these embodiments, the polynucleotide further includes vector
polynucleotide sequences (e.g., adenoviral polynucleotide sequences) operably connected to
the sequence encoding a polypeptide of the invention.
[0143] The invention further provides a cell that comprises the polynucleotide or the
vector, e.g., the cell is transformed or transfected with a polynucleotide encoding the
inventive antibody or fragment thereof or a vector comprising the polynucleotide. In certain
aspects of the invention, the cell expresses an anti-FGFR4 antibody or antibody fragment
containing the Vh and the Vl comprising CDRs identical to those of mAb F90-10C5. The
cell may be a prokaryotic cell, such as Escherichia coli (see, e.g., Pluckthun et al., Methods
Enzymol, 178: 497-515,1989), or a eukaryotic host cell, such as an animal cell (e.g., a
myeloma cell, Chinese Hamster Ovary cell, or hybridoma cell), yeast (e.g., Saccharomyces
cerevisiae), or a plant cell (e.g., a tobacco, corn, soybean, or rice cell). Use of mammalian
host cells is expected to provide for such translational modifications (e.g., glycosylation,
truncation, lipidation, and phosphorylation) that may be desirable to confer optimal biological
activity on recombinant expression products. Similarly, the invention embraces polypeptides
that are glycosylated or non-glycosylated and/or have been covalently modified to include
one or more water soluble polymer attachments such as polyethylene glycol, polyoxyethylene
glycol, or polypropylene glycol.
[0144] Polynucleotides of the invention may be introduced into the host cell as part of a
circular plasmid, or as linear DNA comprising an isolated protein coding region or a viral
vector. Methods for introducing DNA into the host cell, which are well known and routinely
practiced in the art, include transformation, transfection, electroporation, nuclear injection, or
fusion with carriers such as liposomes, micelles, ghost cells, and protoplasts. As stated
above, such host cells are useful for amplifying the polynucleotides and also for expressing
the polypeptides of the invention encoded by the polynucleotide. The host cell may be
isolated and/or purified. The host cell also may be a cell transformed in vivo to cause
transient or permanent expression of the polypeptide in vivo. The host cell may also be an
isolated cell transformed ex vivo and introduced post-transformation, e.g., to produce the
polypeptide in vivo for therapeutic purposes. The definition of host cell explicitly excludes a
transgenic human being.
[0145] Particular methods for producing antibodies from polynucleotides are generally
well known and routinely used. For example, basic molecular biology procedures are
described by Maniatis et al., Molecular Cloning, A Laboratory Manual, 2nd ed., Cold Spring
Harbor Laboratory, New York, 1989 (see also Maniatis et al, 3rd ed., Cold Spring Harbor
Laboratory, New York, 2001). Additionally, numerous publications describe techniques
suitable for the preparation of antibodies by manipulation of DNA, creation of expression
vectors, and transformation and culture of appropriate cells (see, e.g., Mountain and Adair,
Chapter 1 in Biotechnology and Genetic Engineering Reviews, Tombs ed., Intercept,
Andover, UK, 1992); and Current Protocols in Molecular Biology, Ausubel ed., Wiley
Interscience, New York, 1999).
[0146] The invention further provides a peptide comprising (or consisting of) amino acids
67-93 of FGFR4 or a subregion of amino acids 67-93 of FGFR4. In this regard, the invention
provides a peptide comprising (or consisting of) the amino acid sequence
YKEGSRLAPAGRVRG (SEQ ID NO: 5); GSRLAPAGRVRGWRG (SEQ ID NO: 6);
LAPAGRVRGWRGRLE (SEQ ID NO: 7); AGRVRGWRGRLEIAS (SEQ ID NO: 8); or
VRGWRGRLEIASFLP (SEQ ID NO: 9). The peptides of the invention may be used to, for
example, generate antibodies against FGFR4 and/or identify antibodies for anti-FGFR4
activity. In addition, the invention contemplates use of the peptides as immunogens for, e.g.,
stimulating the immune system against tumor cells displaying FGFR4. In one aspect, the
invention provides an isolated antigenic peptide consisting of 5-25 amino acids of an amino
acid sequence encoding FGFR4, wherein the peptide comprises the amino acid sequence set
forth in any one of SEQ ID NOs: 5-9 or a fragment thereof. The invention also provides a
composition comprising any of the foregoing peptides and one or more excipient(s),
adjuvant(s), chemotherapeutic agent(s), and the like. It will be appreciated the materials and
methods for generating an antibody or fragment thereof also apply to the peptides described
herein. For example, the invention provides a polynucleotide that encodes any of the
foregoing peptides, a vector comprising the polynucleotide, and an isolated cell comprising
the polynucleotide (optionally incorporated into a vector).
[0147] It will be understood that the polynucleotide, vector, and cell of the invention can
be used in methods of inhibiting cancer cell invasion in vitro and in vivo (e.g., in a method of
treating cancer in a subject).
Inhibiting Cancer Cell Invasion/Methods of Treatment
[0148] The antibody or antibody fragment binds FGFR4 and inhibits cancer cell invasion.
As used herein, "cancer cell invasion" refers to ingrowth of cancer cells into surrounding
tissue or collagen or fibrin-rich interstitial and temporary matrixes. In this regard, the
invention also provides a method of modulating invasion, ingrowth, or metastasis of cancer
cells, wherein the method comprises contacting a population of cancer cells with the
inventive antibody or fragment thereof (e.g., a composition comprising the antibody or
fragment thereof) in an amount effective to modulate cancer cell invasion, ingrowth, or
metastasis. When the cancer cells are present in a mammalian subject, the population of
cancer cells is contacted by administering the composition comprising the inventive antibody
or fragment thereof to the mammalian subject. In vivo, cancer cell invasion and metastasis is
a multifaceted process requiring degradation of extracellular matrix, including basement
membrane and collagen-rich interstitial matrixes, as well as active cell migration from a
primary tumor. The antibody or fragment thereof blocks one or more of the processes
associated with cancer cell invasion in vivo, such as extracellular matrix degradation.
Preferably, the antibody or fragment thereof inhibits (i.e., downregulates or slows) both
extracellular matrix (collagen and basement membrane) degradation and cell migration in a
three-dimensional tissue environment.
[0149] The efficacy of the antibody to inhibit cancer cell invasion is demonstrated using an
in vitro invasiveness assay, and preferably confirmed in an animal model for cancer. In this
regard, the invention provides a method of identifying an antibody or antibody fragment,
wherein the method comprises (a) obtaining one or more antibodies or antibody fragments
that bind FGFR4; (b) screening the antibodies or antibody fragments in a tumor cell
invasiveness assay; and (c) identifying an antibody that inhibits invasiveness in the assay by,
e.g., at least 50%. An exemplary in vitro dual-chamber tumor cell invasiveness assay is
described in the Examples. In some variations, the cells used in the invasiveness assay co-
express the FGFR4 with other receptors, such as FGFR1. In some variations, the receptor(s)
of interest are recombinantly expressed. In other variations, the cells are isolated from a
tumor (primary isolates) or are from a tumor cell line that expresses the receptor(s) of
interest.
[0150] In an exemplary assay, cancer cells (e.g., MDA-MB-231 cells) expressing FGFR4
(e.g., the FGFR4 R388 protein) are applied to a three-dimensional culture (e.g., collagen) gel.
Optionally, a chemoattractant, such as FGF2, is applied (to the bottom chamber if a dual-
chamber format is employed) to activate cell migration. Invasion can be determined by
measuring the number of cells migrated into the culture gel, or by measuring the length of the
cellular arms reaching into the gel. A decrease in cell invasion into the culture gel mediated
by a candidate antibody, compared to invasion in the absence of the antibody, is indicative of
an antibody or fragment thereof that inhibits cancer cell invasion. Tumor cell invasion assays
are further described in, e.g., Puiffe et al, Neoplasia, 9(10): 820-829,2007; Alonso-Escolano
et al, J. Pharm. Exper. Ther., 318: 373-380, 2006; and Keese et al., BioTechniques, 33: 842-
850, 2002. Antibodies or fragments thereof identified by the method are provided.
[0151] It will be appreciated that antibodies or fragments thereof can be characterized
using other assays known in the art, such as those described in the Examples. For example,
FGFR4 activation can be examined by detecting receptor phosphorylation using, e.g.,
immunoprecipitation and immunoblotting techniques. In certain aspects, the inventive
antibody or fragment thereof inhibits or reduces ligand independent or ligand dependent (e.g.,
fibroblast growth factor 2 (FGF2)-induced) phosphorylation of FGFR4. Similar techniques
can be employed to examine the effect of an antibody or fragment thereof on co-localization
of FGFR4 and MT1-MMP in cells.
[0152] Additionally, the ability of an antibody or fragment thereof to inhibit cancer cell
invasion in vivo can be determined using any suitable animal model, such as a bone invasion
model (see, e.g., Kang, Cancer Cell, 3: 537-549, 2003; and Pauli et al., Cancer Research, 40:
4571-4580, 1980), an animal model of bladder carcinoma invasion (Mohammed et al., Mol.
Cancer Ther., 2(2): 183-188, 2003; and Kameyama et al., Carcinogenesis., 14(S): 1531-1535,
1993), a mouse melanoma metastasis model (Lee et al., Cancer Chemother. Pharmacol.,
57(6): 761-71, 2006), or a screening assay employing modified chorioallantoic membrane
(see, e.g., Ossowski, J. Cell Biol, 107(6): 2437-2445, 1988). Methods of monitoring
metastasis in a human patient are well known and include, for instance, examination of tissue
biopsies to detect metastatic cells.
[0153] "Inhibiting," "blocking," or "impeding" cancer cell invasion does not require a
100% abolition of invasion or metastasis. Any decrease in cancer cell invasion constitutes a
beneficial biological effect in a subject. In this regard, the antibody or fragment thereof can
inhibit cell invasion (e.g., MDA-MB-231 cell invasion of a 3-D collagen tumor cell
invasiveness assay) by, e.g., at least about 5% (at least about 10%, at least about 20%, or at
least about 25%) compared to levels of cancer cell invasion observed in the absence of the
antibody or fragment thereof (e.g., in a biologically-matched control subject, specimen, or
cell culture that is not exposed to the antibody or fragment thereof). In some embodiments,
cell invasion is reduced by at least about 50%, e.g., by at least about 60%, at least about 70%,
at least about 80%, at least about 90%, or at least about 95%. In in vitro or animal models or
in clinical trials involving repetition, inhibition of invasiveness can be confirmed using
standard statistical analyses to confirm that results are statistically significant (e.g., to a level
of significance of p < 0.05).
[0154] In addition, the invention provides a method of treating cancer in a subject (e.g., a
mammal, such as a human). The method comprises administering to the subject the inventive
antibody or fragment thereof in an amount effective to treat cancer. 'Treating cancer"
encompasses inhibiting or arresting the development or progression of a disorder marked by
abnormal cell or tissue growth or proliferation, particularly those with metastatic potential.
"Treating cancer" also encompasses impeding the progress of invasive growth in surrounding
tissue, thereby slowing cancer cell spread to distant organs and the growth of secondary
tumors (metastases). 'Treating cancer" also encompasses alleviating, in whole or in part, a
disorder marked by hyperproliferation. Any cancer with metastatic potential or an invasive
aspect and that expresses FGFR4 is suitable for treatment in the context of the invention.
Indeed, in certain embodiments, the method comprises administering the inventive antibody
or fragment thereof to a patent diagnosed with metastatic cancer. Exemplary cancers include
cancers of the oral cavity and pharynx, the digestive system, the respiratory system, bones
and joints (e.g., bony metastases), soft tissue, the skin (e.g., melanoma), breast, the genital
system, the urinary system, the eye and orbit, the brain and nervous system (e.g., glioma), or
the endocrine system (e.g., thyroid). In some embodiments, the cancer of the inventive
method is breast cancer, bladder cancer, melanoma, prostate cancer, colorectal cancer,
thyroid cancer, glioma, mesothelioma, lung cancer, testicular cancer, or pancreatic cancer.
The R388 FGFR4 mutant has been detected in cell lines derived from breast tumors,
squamous cell carcinoma, glioblastomas, neuroblastomas and uterine cancer (see U.S. Patent
6,770,742); thus, the materials and methods of the invention are particularly suitable for the
treatment of those disorders.
[0155] The progress of the inventive method in treating cancer (e.g., impeding cancer cell
invasion in surrounding tissues) can be ascertained using any suitable method, such as those
methods described herein and currently used in the clinic to track cancer progress. If desired,
the efficacy of the inventive method is determined by detection of new tumors, detection of
tumor antigens or markers, biopsy, positron emission tomography (PET) scans, survival,
disease progression-free survival, time to disease progression, quality of life assessments
such as the Clinical Benefit Response Assessment, and the like, all of which can point to the
overall progression (or regression) of cancer in a human.
[0156] In some embodiments of the invention, the method of the invention comprises
determining the presence or absence of an FGFR4 allele that encodes FGFR4 R388 in the
cancer. Depending on the particular embodiment, the treatment is administered if the cancer
has at least one FGFR4 allele that encodes FGFR4 R388. Thus, in one aspect, the invention
provides a method of treating a mammalian subject, wherein the method comprises selecting
for treatment a mammalian subject diagnosed with or treated for cancer, wherein the cancer
includes cells that contain at least one FGFR4 allele that encodes FGFR4 R388; and
administering to the subject a composition comprising the antibody or fragment thereof (or
nucleic acid encoding the antibody or fragment thereof) in an amount effective to modulate
cancer cell invasion, ingrowth, or metastasis.
[0157] The presence or absence of an FGFR4 R388 allele in a biological sample can be
determined using a variety of techniques. Samples typically are isolated from blood, serum,
urine, or tissue biopsies from, e.g., muscle, connective tissue, nerve tissue, and the like. Once
obtained, cells from the sample are examined to detect the presence or absence of FGFR4
R388. One method for identifying FGFR4 R388 comprises assaying nucleic acid (e.g.,
obtaining nucleic acid sequence data) from a biological sample taken from a subject (e.g., a
cancer specimen or tissue biopsy). Genomic DNA, RNA, or cDNA is obtained from a
biological sample and, optionally, the nucleic acid encoding FGFR4 is amplified by
polymerase chain reaction (PCR). The DNA, RNA, or cDNA sample is then examined. The
presence of FGFR4 R388 can be determined by sequence-specific hybridization of a nucleic
acid probe specific for the R388 allele. One of skill in the art has the requisite knowledge
and skill to design a probe so that sequence-specific hybridization will occur only if the
biological sample contains an FGFR4 R388 coding sequence. Alternatively or in addition,
the presence or absence of FGFR4 R388 is determined by directly sequencing DNA or RNA
obtained from a subject.
[0158] In one aspect, the presence or absence of an FGFR4 R388 allele in a biological
sample (e.g., a cell) is determined by assaying FGFR4 protein with an antibody or antibody
fragment that differentially binds FGFR4 R388 and G388 alleles. The term "differentially
binds" refers to the antibody's ability to distinguish between the R388 and G388 FGFR4
proteins. For example, an antibody or fragment thereof that differentially binds FGFR4 R388
binds the protein with greater affinity (e.g., at least 10, 15, 20, 25, 50, or 100 times greater
affinity) than it binds to FGFR4 G388. Exemplary methods for detecting FGFR4 R388
protein include, but are not limited to, immunoassays, e.g., immunofluorescent
immunoassays, immunoprecipitations, radioimmunoasays, ELISA, Western blotting, and
fluorescence activated cell sorting (FACS). These methods comprise contacting a biological
sample with an antibody or fragment thereof that differentially binds FGFR4 R388, and
detecting antibody binding to FGFR4 R388.
Non-Antibody Based Inhibitors
[0159] Some variations of the invention include use of non-antibody-based inhibitors
against FGFR4 and, optionally, MT1-MMP. Matrix metalloproteinases (MMP) constitute a
family of enzymes responsible for extracellular matrix degradation in cancer tissue (Nagase
et al., J. Biol. Chem., 274: 21491-21494, 1999; Nelson et al, J. Clin. Oncol., 18: 1135-1149,
2000). MMPs are zinc-dependent multidomain endopeptidases that, with few exceptions,
share a basic structural organization comprising propeptide, catalytic, hinge, and C-terminal
(hemopexin-like) domains (Nagase et al., supra; Massova et al., FASEBJ., 12: 1075-1095,
1998). All MMPs are produced in a latent form (pro-MMP) requiring activation for catalytic
activity, a process that is usually accomplished by proteolytic removal of the propeptide
domain.
[0160] MT1-MMP (MMP-14) is a multifunctional enzyme that degrades a variety of
extracellular matrix components including fibrillar collagen and fibrin (Pei et al., J. Biol.
Chem., 271: 9135-9140, 1996; d'Ortho et al., FEBS Lett., 421: 159-164, 1998; Ohuchi et al.,
J. Biol Chem., 272: 2446-2451, 1997). In addition, both MMP-2 and MT1-MMP are
associated with metastatic potential in many human cancers, and enhance tumor cell invasion
in experimental systems. MT-MMP1 is frequently upregulated in both cancer cells and
reactive stromal cells in various forms of cancer, and cancer cells overexpressing MT1-MMP
invade, proliferate, and metastasize in nude mice at remarkably higher rates than control
cells. The amino acid sequence of MT1-MMP is provided in SEQ ID NO: 10.
[0161] FGFR4 or MT1-MMP inhibitors modulate activity by targeting FGFR4 or MT1-
MMP directly, i.e., at the protein level, targeting transcription or translation of FGFR4 or
MT1-MMP, or targeting a downstream molecule required for realization of FGFR4 or MT1-
MMP function. On the nucleic acid level, inhibitors inactivate or disrupt FGFR4 or MT1-
MMP coding sequence. Inhibition may also block transcription or translation by targeting
genomic DNA or FGFR4 mRNA, FGFR4 ligand mRNA, MT1-MMP mRNA and/or mRNA
of downstream targets. In this regard, antisense therapy is one method for inhibiting
expression, described below with particular reference to FGFR4 and MT1-MMP with the
understanding that the description is equally suitable for other gene targets.
[0162] Antisense oligonucleotides negatively regulate FGFR4 (or MT1-MMP) expression
via hybridization to messenger RNA (mRNA) encoding FGFR4 (or MT1-MMP). The
nucleic acid sequences encoding FGFR4 G388 protein and FGFR4 R388 protein are known,
e.g., as reported in GenBank Accession No. X57205 for the nucleic acid sequence of FGFR4
G388 (SEQ ID NO: 11). The nucleic acid sequence of FGFR4 R388 is provided in SEQ ID
NO: 12. Likewise, the nucleic acid sequence encoding MT1-MMP is known in the art, e.g.,
as reported in GenBank Accession No. X90925 (SEQ ID NO: 13). These sequences may be
used to prepare and optimize antisense molecules using any methods known in the art. In this
regard, the invention includes use of an inhibitor nucleic acid that hybridizes with MT1-MMP
(or FGFR4) genomic DNA or mRNA and inhibits MT1-MMP (or FGFR4) transcription or
translation. All classes of nucleic acid inhibitor described herein can be used alone or in
combination with other inhibitor substances described herein (see section below relating to
combination therapies).
[0163] In one aspect, antisense oligonucleotides at least 5 to about 50 nucleotides in
length, including all lengths (measured in integer number of nucleotides) in between, which
specifically hybridize to mRNA encoding FGFR4 or MT1-MMP and inhibit mRNA
expression, and as a result FGFR4 or MT1-MMP protein expression, are contemplated for
use in the inventive method. Antisense oligonucleotides include those comprising modified
internucleotide linkages and/or those comprising modified nucleotides which are known in
the art to improve stability of the oligonucleotide, i.e., make the oligonucleotide more
resistant to nuclease degradation, particularly in vivo. It is understood in the art that, while
antisense oligonucleotides that are perfectly complementary to a region in the target
polynucleotide possess the highest degree of specific inhibition, antisense oligonucleotides
which are not perfectly complementary, i.e., those which include a limited number of
mismatches with respect to a region in the target polynucleotide, also retain high degrees of
hybridization specificity and therefore inhibit expression of the target mRNA. Accordingly,
the invention includes methods using antisense oligonucleotides that are perfectly
complementary to a target region in a polynucleotide encoding FGFR4 or MT1-MMP, as
well as methods that utilize antisense oligonucleotides that are not perfectly complementary,
i.e., include mismatches, to a target region in the target polynucleotide to the extent that the
mismatches do not preclude specific hybridization to the target region in the target
polynucleotide. Preparation and use of antisense compounds is described in U.S. Patent No.
6,277,981, the disclosure of which is incorporated herein by reference in its entirety.
[0164] Another class of therapeutics for inhibiting expression of the target genes described
herein is ribozymes. Ribozyme inhibitors include a nucleotide region which specifically
hybridizes to a target polynucleotide and an enzymatic moiety that digests the target
polynucleotide. Specificity of ribozyme inhibition is related to the length the antisense region
and the degree of complementarity of the antisense region to the target region in the target
polynucleotide. The invention therefore includes use of ribozyme inhibitors of FGFR4 or
MT1-MMP comprising antisense regions from 5 to about 50 nucleotides in length, including
all nucleotide lengths in between, that are perfectly complementary, as well as antisense
regions that include mismatches to the extent that the mismatches do not preclude specific
hybridization to the target region in the target FGFR4- or MT1-MMP-encoding
polynucleotides. Ribozymes useful in methods of the invention include those comprising
modified internucleotide linkages and/or those comprising modified nucleotides which are
known in the art to improve stability of the oligonucleotide, i.e., make the oligonucleotide

more resistant to nuclease degradation, particularly in vivo, to the extent that the
modifications do not alter the ability of the ribozyme to specifically hybridize to the target
region or diminish enzymatic activity of the molecule. Because ribozymes are enzymatic, a
single molecule is able to direct digestion of multiple target molecules thereby offering the
advantage of being effective at lower concentrations than non-enzymatic antisense
oligonucleotides. Preparation and use of ribozyme technology is described in U.S. Patent
Nos. 6,696,250; 6,410,224; and 5,225,347, the disclosures of which are incorporated herein
by reference in their entireties.
[0165] Another class of therapeutics for inhibiting expression (and therefore activity) of
target genes/pathways described herein is interfering RNA technology, also known as RNA
interference (RNAi) or short interfering RNA (siRNA). Using the knowledge of the
sequence of target genes such as FGFR4 or MT1-MMP, siRNA molecules are formed that
interfere with the expression of the genes. siRNA describes a technique by which post-
transcriptional gene silencing (PTGS) is induced by the direct introduction of double stranded
RNA (dsRNA: a mixture of both sense and antisense strands) (Fire et al., Nature, 391: 806-
811, 1998). Current models of PTGS indicate that short stretches of interfering dsRNAs (21-
23 nucleotides; siRNA also known as "guide RNAs") mediate PTGS. The siRNAs are
apparently produced by cleavage of dsRNA introduced directly or via a transgene or virus.
These siRNAs may be amplified by an RNA-dependent RNA polymerase (RdRP) and are
incorporated into the RNA-induced silencing complex (RISC), guiding the complex to the
homologous endogenous mRNA, where the complex cleaves the transcript. It is
contemplated that RNAi may be used to disrupt the expression of a gene in a tissue-specific
manner. By placing a gene fragment encoding the desired dsRNA behind an inducible or
tissue-specific promoter, it should be possible to inactivate genes at a particular location
within an organism or during a particular stage of development.
[0166] Also contemplated is double-stranded RNA (dsRNA) wherein one strand is
complementary to a target region in a target FGFR4- or MT1-MMP-encoding
polynucleotides. In general, dsRNA molecules of this type less than 30 nucleotides in length
are referred to in the art as short interfering RNA (siRNA). The invention also includes,
however, use of dsRNA molecules longer than 30 nucleotides in length, and in certain aspects
of the invention, these longer dsRNA molecules can be about 30 nucleotides in length up to
200 nucleotides in length and longer, and including all length dsRNA molecules in between.
As with other RNA inhibitors, complementarity of one strand in the dsRNA molecule can be
a perfect match with the target region in the target polynucleotide, or may include
mismatches to the extent that the mismatches do not preclude specific hybridization to the
target region in the target FGFR4- or MT1-MMP-encoding polynucleotides. As with other
RNA inhibition technologies, dsRNA molecules include those comprising modified
internucleotide linkages and/or those comprising modified nucleotides which are known in
the art to improve stability of the oligonucleotide, i.e., make the oligonucleotide more
resistant to nuclease degradation, particularly in vivo. Exemplary lentiviral shRNA
constructs targeting MT1-MMP include TRCN0000050854 (GenBank Accession No.
NM_004995) and TRCN0000050585 (GenBank Accession No. NM006703) from Open
Biosystems (Huntsville, Alabama) (catalog nos. catalog RHS3979-9618053 and RHS3979-
9617784; described further in Tatti et al., Exp. Cell. Res., 3/4(13): 2501-14, 2008). HP
Validated siRNA SI03648841 (SEQ ID NO: 14) from Qiagen (Hilden, Germany) also
effectively downregulates MT1-MMP. Exemplary FGFR4-targeting siRNAs include HP
Validated siRNA SI02659979 (SEQ ID NO: 15), HP Validated siRNA SI02665306, and HP
GenomeWide siRNA SI00031360 (SEQ ID NO: 16) from Qiagen (Hilden, Germany).
Preparation and use of RNAi compounds is described in U.S. Patent Application No.
20040023390, the disclosure of which is incorporated herein by reference in its entirety.
[0167] The invention further contemplates methods wherein inhibition of FGFR4 or MT1-
MMP is effected using RNA lasso technology. Circular RNA lasso inhibitors are highly
structured molecules that are inherently more resistant to degradation and therefore do not, in
general, include or require modified internucleotide linkage or modified nucleotides. The
circular lasso structure includes a region that is capable of hybridizing to a target region in a
target polynucleotide, the hybridizing region in the lasso being of a length typical for other
RNA inhibiting technologies. As with other RNA inhibiting technologies, the hybridizing
region in the lasso may be a perfect match with the target region in the target polynucleotide,
or may include mismatches to the extent that the mismatches do not preclude specific
hybridization to the target region in the target FGFR4- or MT1-MMP-encoding
polynucleotides. Because RNA lassos are circular and form tight topological linkage with the
target region, inhibitors of this type are generally not displaced by helicase action unlike
typical antisense oligonucleotides, and therefore can be utilized as dosages lower than typical
antisense oligonucleotides. Preparation and use of RNA lassos is described in U.S. Patent
6,369,038, the disclosure of which is incorporated herein by reference in its entirety.
[0168] Anti-sense RNA and DNA molecules, ribozymes, RNAi and triple helix molecules
directed against FGFR4 or MT1-MMP can be prepared by any method known in the art for
the synthesis of DNA and RNA molecules. These include techniques for chemically
synthesizing oligodeoxyribonucleotides well known in the art including, but not limited to,
solid phase phosphoramidite chemical synthesis. Alternatively, RNA molecules may be
generated by in vitro and in vivo transcription of DNA sequences encoding the antisense
RNA molecule. Such DNA sequences may be incorporated into a wide variety of vectors
which incorporate suitable RNA polymerase promoters such as the T7 or SP6 polymerase
promoters. Alternatively, antisense cDNA constructs that synthesize antisense RNA
constitutively or inducibly, depending on the promoter used, can be introduced stably or
transiently into cells.
[0169] Aptamers are another nucleic acid based method for interfering with the interaction
of FGFR4 or MT1-MMP is the use of an aptamer. Aptamers are DNA or RNA molecules
that have been selected from random pools based on their ability to bind other molecules.
Aptamers have been selected which bind nucleic acid, proteins, small organic compounds,
and even entire organisms. Methods and compositions for identifying and making aptamers
are known to those of skill in the art and are described e.g., in U.S. Patent No. 5,840,867 and
U.S. Patent No. 5,582,981 each incorporated herein by reference. Aptamers that bind FGFR4
or MT1-MMP are specifically contemplated to be useful in the present therapeutic
embodiments.
[0170] Recent advances in the field of combinatorial sciences have identified short
polymer sequences with high affinity and specificity to a given target. For example, SELEX
technology has been used to identify DNA and RNA aptamers with binding properties that
rival mammalian antibodies, the field of immunology has generated and isolated antibodies
or antibody fragments which bind to a myriad of compounds and phage display has been
utilized to discover new peptide sequences with very favorable binding properties. Based on
the success of these molecular evolution techniques, it is certain that molecules can be
created which bind to any target molecule. A loop structure is often involved with providing
the desired binding attributes as in the case of: aptamers which often utilize hairpin loops
created from short regions without complimentary base pairing, naturally derived antibodies
that utilize combinatorial arrangement of looped hyper-variable regions, and new phage
display libraries utilizing cyclic peptides that have shown improved results when compared to
linear peptide phage display results. Thus, sufficient evidence has been generated to suggest
that high affinity ligands can be created and identified by combinatorial molecular evolution
techniques. For the present invention, molecular evolution techniques can be used to isolate
binding constructs specific for ligands described herein. For more on aptamers, see
generally, Gold et al., J. Biotechnol., 74: 5-13, 2000. Relevant techniques for generating
aptamers may be found in U.S. Patent No. 6,699,843, which is incorporated by reference in
its entirety.
[0171] In some embodiments, the aptamer may be generated by preparing a library of
nucleic acids; contacting the library of nucleic acids with a target, e.g., FGFR4 or MT1-
MMP, wherein nucleic acids having greater binding affinity for the target (relative to other
library nucleic acids) are selected and amplified to yield a mixture of nucleic acids enriched
for nucleic acids with relatively higher affinity and specificity for binding to the target. The
processes may be repeated, and the selected nucleic acids mutated and re-screened, whereby a
target aptamer is identified.
[0172] Other inhibitors target FGFR4 or MT1-MMP directly, i.e., at the protein level. In
this regard, chemical compound inhibitors of FGFR4 or MT1-MMP are contemplated. Small
molecule compounds (i.e., compounds having a molecular weight of less than 1000 Daltons,
typically between 300 and 700 Daltons) are generally preferred because the reduced size
renders the molecule more accessible for uptake by a target cell. Synthetic inhibitors of
FGFR4 include PD173074 (Pfizer; Ezzat et al, Clinical Cancer Res., 11: 1336-1341, 2005,
and Kwabi-Addo et al., Endocrine-Related Cancer, 11; 709-724, 2004). Synthetic inhibitors
capable of blocking MT1-MMP activity include Ro-28-2653, described in Maquoi et al.,
Clin. Cancer Res., 15: 4038-47 (2004). Synthetic inhibitors are further described in Nisato et
al, Cancer Res., 65(20): 9377-9387, 2005; and Galvez et al., J. Biol. Chem., 276: 37491-500,
2001.
[0173] Other inhibitors include FGFR4 binding agents that specifically bind to FGFR4 to
block or impair binding of human FGFR4 to one or more ligands, such as FGF2. While such
agents bind the receptor, they do not trigger the signaling cascade responsible for FGFR4
activity. Alternatively, soluble FGFR4 receptors may be used to sequester ligands away from
FGFR4. In this regard, the extracellular region of FGFR4 can be fused to another moiety to
increase serum half-life, e.g., an Fc antibody domain, to make a fusion protein, or to PEG
moieties, to generate a soluble FGFR4 receptor.
Administration Considerations
[0174] When treating cancer or modulating cancer cell invasion in vivo, the method is
preferably performed as soon as possible after it has been determined that a subject is at risk
for cancer (e.g., cancer markers are detected) or as soon as possible after cancer and/or
invasion of surrounding tissues is detected (e.g., following tumor resection). To this end, the
antibody or fragment thereof is administered before tumor invasion is detected to protect, in
whole or in part, against cancer cell invasion, ingrowth, or metastasis. The antibody or
fragment thereof also can be administered after tumor invasion has begun to prevent, in
whole or in part, further invasion or formation of secondary tumors. In this regard, the
invention provides a method of treating a mammalian subject comprising (i) selecting for
treatment a mammalian subject diagnosed with or treated for cancer; and (ii) administering to
the subject the inventive antibody or fragment thereof (e.g., the inventive antibody or
fragment thereof formulated in a composition) in an amount effective to modulate cancer cell
invasion, ingrowth, or metastasis.
[0175] In preferred embodiments, the antibody or antibody fragment (and/or any other
therapeutic agent described herein) is formulated in a composition, such as a physiologically-
acceptable composition, comprising a carrier (i.e., vehicle, adjuvant, or diluent). The
particular carrier employed is limited only by chemico-physical considerations, such as
solubility and lack of reactivity, and by the route of administration. Physiologically-
acceptable carriers are well known in the art. Illustrative pharmaceutical forms suitable for
injectable use include sterile aqueous solutions or dispersions and sterile powders for the
extemporaneous preparation of sterile injectable solutions or dispersions (for example, see
U.S. Patent No. 5,466,468). Injectable formulations are further described in, e.g.,
Pharmaceutics and Pharmacy Practice, J. B. Lippincott Co., Philadelphia. Pa., Banker and
Chalmers, eds. (1982), and ASHP Handbook on Injectable Drugs, Toissel, 4th ed. (1986)). A
pharmaceutical composition comprising any of the materials described herein may be placed
within containers, along with packaging material that provides instructions regarding the use
of such pharmaceutical compositions. Generally, such instructions include a tangible
expression describing the reagent concentration, as well as, in certain embodiments, relative
amounts of excipient ingredients or diluents (e.g., water, saline or PBS) that may be
necessary to reconstitute the pharmaceutical composition.
[0176] A particular administration regimen for a particular subject will depend, in part,
upon the particular antibody used, the presence of other therapeutics, the amount
administered, the route of administration, and the cause and extent of any side effects. The
amount administered to a subject (e.g., a mammal, such as a human) in accordance with the
invention should be sufficient to affect the desired response over a reasonable time frame.
The size of the dose also will be determined by the route, timing, and frequency of
administration. Accordingly, the clinician may titer the dosage and modify the route of
administration to obtain the optimal therapeutic effect, and conventional range-finding
techniques are known to those of ordinary skill in the art. Purely by way of illustration, the
inventive method can comprise administering, e.g., from about 0.1 µg/kg to up to about 100
mg/kg or more, depending on the factors mentioned above. In other embodiments, the
dosage may range from 1 µg/kg up to about 100 mg/kg; or 5 µg/kg up to about 100 mg/kg; or
10 µg/kg up to about 100 mg/kg. Due to the hyperproliferative nature of cancer, a single
dose of antibody or fragment thereof may not accomplish a complete anti-cancer (anti-
invasive) effect. Indeed, as with most chronic diseases, prolonged treatment involving
multiple doses of a therapeutic agent may be required. Accordingly, in one embodiment, the
inventive method comprises delivering multiple doses of pharmaceutical composition over a
period of time.
[0177] Suitable methods of administering a physiologically-acceptable composition, such
as a pharmaceutical composition comprising an anti-FGFR4 antibody or fragment thereof, are
well known in the art. Although more than one route can be used to administer an agent, a
particular route can provide a more immediate and more effective reaction than another route.
Depending on the circumstances, a pharmaceutical composition comprising the agent is
applied or instilled into body cavities, absorbed through the skin or mucous membranes,
ingested, inhaled, and/or introduced into circulation. For example, in certain circumstances,
it will be desirable to deliver a physiologically-acceptable (e.g., pharmaceutical) composition
through injection by intravenous, intraperitoneal, intracerebral (intra-parenchymal),
intracerebroventricular, intramuscular, intra-ocular, intraarterial, intraportal, intralesional,
intramedullary, intrathecal, intraventricular, transdermal, subcutaneous, intraperitoneal,
intranasal, enteral, topical, sublingual, urethral, vaginal, or rectal means, by sustained release
systems, or by implantation devices. If desired, the antibody or fragment thereof is
administered regionally via intraarterial or intravenous administration feeding the region of
interest, e.g., via the hepatic artery for delivery to the liver. Alternatively, the composition is
administered locally via implantation of a membrane, sponge, or another appropriate material
on to which the antibody has been absorbed or encapsulated. Where an implantation device

is used, the device may be implanted into any suitable tissue or organ, and delivery of the
antibody may be via diffusion, timed-release bolus, or continuous administration. In other
aspects, the agent is administered directly to exposed tissue during tumor resection or other
surgical procedures or by targeted injection (e.g., intratumoral injection). Therapeutic
delivery approaches are well known to the skilled artisan, some of which are further
described, for example, in U.S. Patent No. 5,399,363.
Combination Therapy
[0178] When appropriate, the agent is administered in combination with other substances
(e.g., therapeutics) and/or other therapeutic modalities to achieve an additional (or
augmented) biological effect. For example, in one embodiment, the inventive method
comprises administering two or more different anti-FGFR4 antibodies (or fragments thereof)
to a subject. In this regard, when the inventive method entails using an antibody that binds an
FGFR4 epitope recognized by mAb F90-10C5, the method may further comprise
administering to a subject (or contacting a population of cancer cells with) an antibody or
fragment thereof that binds an epitope of FGFR4 that is different than the epitope recognized
by mAb F90-10C5. Exemplary second antibodies and fragments thereof include (i) F85-6C5
and F90-3B6 (also referred to herein as "6C5" and "3B6," respectively), described in the
Examples, (ii) antibodies or fragments thereof that compete for binding of FGFR4 with F85-
6C5 and/or F90-3B6, and (iii) antibodies or fragments thereof that bind the region of FGFR4
recognized by F85-6C5 and/or F90-3B6. Surprisingly, exposing cancer cells to mAb F90-
10C5 in combination with F85-6C5 or F90-3B6 results in a greater reduction in total MT1-
MMP protein and activated MT1-MMP protein in the cells compared to treatment with mAb
F90-10C5 alone. Combination treatment with two or more anti-FGFR4 antibodies
recognizing different FGFR4 epitopes (especially different extracellular epitopes) can
enhance the inhibitory effect of mAb F90-10C5.
[0179] Alternatively (or in addition), multiple antibodies are delivered to a subject to
obtain multiple biological effects. In one aspect, the invention provides a method of treating
a mammalian subject comprising administering to a subject diagnosed with or treated for
cancer a first and a second anti-FGFR4 antibody or FGFR4-binding fragment thereof,
wherein the first anti-FGFR4 antibody or fragment inhibits FGF2-induced phosphorylation of
FGFR4 R388, and the second anti-FGFR4 antibody or fragment inhibits ligand-independent
FGFR4 phosphorylation. The antibodies or fragments thereof may be formulated in a single
composition, or administered in separate compositions (i.e., a first composition containing the
first antibody or fragment and a second composition containing the second antibody or
fragment) to be administered simultaneously or sequentially. The inventive method also may
entail administering the anti-FGFR4 antibody in combination with a non-antibody based
FGFR4 inhibitor, such as those described further herein. For example, the method can
comprise administering to the subject a standard of care chemotherapy for cancer.
[0180] Alternatively or in addition, the inventive method further comprises administering
to a subject (or contacting a population of cancer cells with) an MT1-MMP inhibitor. Any
inhibitor of MT1-MMP is suitable for use in the context of the invention, and non-antibody-
based (e.g., small molecule) MT1-MMP inhibitors are described above. In one aspect, the
inhibitor is an antibody or fragment thereof that binds MT1-MMP to inhibit the enzyme's
activity (e.g., inhibit extracellular matrix degradation). Several anti-MT1-MMP antibodies
are known in the art. For example, antibodies LEM-1 and LEM-2, further described in
Nisato et al., Cancer Res., 65(20): 9377-9387, 2005, inhibit cytokine-induced bovine
microvascular endothelial (BME) cell invasion of three-dimensional collagen gels in a dose-
dependent manner. Anti-MT1-MMP antibodies also are described in Galvez et al., J. Biol.
Chem., 276: 37491-500, 2001. The discussion of antibodies and fragments thereof provided
above with respect to FGFR4 antibodies also is relevant to anti-MT1-MMP antibodies.
[0181] Endogenous inhibitors of MT1-MMP have been identified and are contemplated for
use in the inventive method. For example, once activated, MMPs are specifically inhibited
by a group of endogenous tissue inhibitors of metalloproteinases (TIMPs) that bind to the
active site, inhibiting catalysis (Nagase et al, supra). MT1-MMP is inhibited by TIMP-2,
TIMP-3 and TIMP-4, but not by TIMP-1 (Will et al., J. Biol. Chem., 271:11119-17123,
1996; Bigg et al, Cancer Res., 61: 3610-3618, 2001). RECK (reversion inducing-cysteine
rich protein with Kazal motifs), a GPI anchored glycoprotein, is another inhibitor of MT1-
MMP (Oh et al., Cell, 107: 789-800, 2001). Mice containing mutated RECK are embryonic
lethal at E10.5 showing defects in collagen fibrils, the basal lamina, and vascular
development - a phenotype that may correlate with excessive MMP activity.
Chondroitin/heparan sulfate proteoglycans, testican 1, testican 3, and a splice variant of
testican 3, N-Tes, have also been shown to inhibit MT1-MMP (Nakada et al., Cancer Res.,
61: 8896-8902, 2001).

[0182] Inhibitors of vascular endothelial growth factor receptor-3 (VEGFR-3) or vascular
endothelial growth factor receptor-2 (VEGFR-2) also are contemplated for use with the
inventive antibody, fragment thereof, polypeptide, or polynucleotide. The inventive method
can comprise administering an agent that inhibits VEGF-D or VEGF-C stimulation of
VEGFR-3 or VEGFR-2, such as an antibody or antibody fragment that binds to VEGF-C,
VEGF-D, or the extracellular domain of VEGFR-3 or VEGFR-2; a soluble protein
comprising a VEGFR-3 extracellular domain or fragment thereof effective to bind VEGF-C
or VEGF-D; or a soluble protein comprising a VEGFR-2 extracellular domain or fragment
thereof effective to bind VEGF-C or VEGF-D. Exemplary agents and methods for
modulating VEGFR-3 and VEGFR-2 activity are described in, e.g., U.S. Patents 7,034,105
and 6,824,777, U.S. Patent Publication Nos. 2005/0282233 and 2006/0030000; and
International Patent Publication Nos. WO 2005/087812 (Application No.
PCT/US2005/007742), WO 2005/087808 (Application No. PCT/US2005/007741), WO
2002/060950 (Application No. PCT/US2002/001784), and WO 2000/021560 (Application
No. PCT/US1999/023525).
[0183] At any given time medical practitioners have one or more "standard of care"
therapies that are regarded as appropriate or preferred for a particular cancer, stage of
progression, and patient type, for example. The invention includes, as an additional variation,
administration/use of standard or care therapies in combination with therapies described
herein.
[0184] Other therapeutics/co-treatments suitable for use in conjunction with the inventive
method include, for example, radiation treatment, hyperthermia, surgical resection,
chemotherapy, anti-angiogenic factors (for instance, soluble growth factor receptors (e.g.,
sflt), growth factor antagonists (e.g., angiotensin), etc.), pain relievers, and the like. Each
therapeutic factor is administered according to a regimen suitable for that medicament. This
includes concurrent administration (i.e., substantially simultaneous administration) and non-
concurrent administration (i.e., administration at different times, in any order, whether
overlapping or not) of the inventive antibody or fragment thereof and one or more
additionally suitable agents(s). It will be appreciated that different components may be
administered in the same or in separate compositions, and by the same or different routes of
administration. In this regard, the inventive composition can comprise an antibody or
fragment thereof that binds an epitope of FGFR4 that is different than the epitope recognized
by mAb F90-10C5 and/or an MT1-MMP inhibitor. Alternatively or in addition, the inventive
antibody or fragment thereof can comprise an anti-neoplastic agent (e.g., a radionucleotide)
or cytotoxic agent conjugated or attached thereto. For further discussion of radionucleotide-
antibody conjugates, see, e.g., Appelbaum et al., Blood, 75(8): 2202, 1989; and U.S. Patent
No. 6,743,411.
[0185] Chemotherapy treatment for use in conjunction with the invention employ anti-
neoplastic agents including, but not limited to, alkylating agents including: nitrogen mustards,
such as mechlor-ethamine, cyclophosphamide, ifosfamide, melphalan and chlorambucil;
nitrosoureas, such as carmustine (BCNU), lomustine (CCNU), and semustine (methyl-
CCNU); ethylenimines/methylmelamine such as thriethylenemelamine (TEM), triethylene,
thiophosphoramide (thiotepa), and hexamethylmelamine (HMM, altretamine); alkyl
sulfonates such as busulfan; triazines such as dacarbazine (DTIC); antimetabolites including
folic acid analogs such as methotrexate and trimetrexate, pyrimidine analogs such as 5-
fluorouracil, fluorodeoxyuridine, gemcitabine, cytosine arabinoside (AraC, cytarabine), 5-
azacytidine, 2,2'-difluorodeoxycytidine, purine analogs such as 6-mercaptopurine, 6-
thioguanine, azathioprine, 2'-deoxycoformycin (pentostatin), erythrohydroxynonyladenine
(EHNA), fludarabine phosphate, and 2-chlorodeoxyadenosine (cladribine, 2-CdA); natural
products including antimitotic drugs such as paclitaxel, vinca alkaloids including vinblastine
(VLB), vincristine, and vinorelbine, taxotere, estramustine, and estramustine phosphate;
pipodophylotoxins such as etoposide and teniposide; antibiotics such as actimomycin D,
daunomycin (rubidomycin), doxorubicin, mitoxantrone, idarubicin, bleomycins, plicamycin
(mithramycin), mitomycinC, and actinomycin; enzymes such as L-asparaginase; biological
response modifiers such as interferon-alpha, IL-2, G-CSF and GM-CSF; miscellaneous
agents including platinium coordination complexes such as cisplatin and carboplatin,
anthracenediones such as mitoxantrone, substituted urea such as hydroxyurea,
methylhydrazine derivatives including N-methylhydrazine (MIH) and procarbazine,
adrenocortical suppressants such as mitotane (o,p'-DDD) and aminoglutethimide; hormones
and antagonists including adrenocorticosteroid antagonists such as prednisone and
equivalents, dexamethasone and aminoglutethimide; progestin such as hydroxyprogesterone
caproate, medroxyprogesterone acetate and megestrol acetate; estrogen such as
diethylstilbestrol and ethinyl estradiol equivalents; antiestrogen such as tamoxifen; androgens
including testosterone propionate and fluoxymesterone/equivalents; antiandrogens such as
flutamide, gonadotropin-releasing hormone analogs and leuprolide; and non-steroidal
antiandrogens such as flutamide.
[0186] Cytokines that are effective in inhibiting tumor metastasis are also contemplated for
use in the combination therapy. Such cytokines, lymphokines, or other hematopoietic factors
include, but are not limited to, M-CSF, GM-CSF, TNF, IL-1, IL-2, IL-3, IL-4, IL-5, IL-6, IL-
7, IL-8, IL-9, IL-10, IL-11, IL-12, IL-13, IL-14, IL-15, IL-16, IL-17, IL-18, IFN, TNFa,
TNF1, TNF2, G-CSF, Meg-CSF, GM-CSF, thrombopoietin, stem cell factor, and
erythropoietin.
EXAMPLES
[0187] The invention, thus generally described, will be understood more readily by
reference to the following examples, which are provided by way of illustration and are not
intended to limit the invention.
Example 1
[0188] This Example identifies FGFR4 and other kinase molecules as regulators of tumor
cell invasion using an unbiased gain of function kinome screen for MT1-MMP activity.
[0189] To identify kinase molecules that regulate tumor cell invasion, an unbiased gain of
function kinome screen for MT1-MMP activity using gelatin zymography was developed.
Since MT1-MMP is the most prominently expressed MT-MMP and the main MMP-2
activator in HT-1080 cells (Lehti et al., J. Biol. Chem., 27(10): 8440-48, 2002), MMP-2
activation served as an indirect measure of MT1-MMP activity. A cDNA library encoding
~93% of all human protein kinases (564 cDNAs encoding 480 separate kinases) (Varjosalo et
al, Cell, 133(3): 537-48, 2008) was transiently transfected with FuGENE6 (Roche) into
human HT-1080 fibrosarcoma cells plated in 96-well plates at a density of 1 xlO4 cells/well.
After transfection, the cells were incubated in complete medium for 24 hours and serum-free
medium for an additional 20 hours. Aliquots of conditioned medium were dissolved in non-
reducing Laemmli sample buffer and separated by electrophoresis using discontinuous
3.5:10% polyacrylamide gels containing 1 mg/ml gelatin. SDS was removed to permit MMP
refolding and autoactivation as described in Lohi et al., Eur. J. Biochem., 239(2): 239-47,
1996. The gels were then incubated at 37°C for 16 hours and stained with Coomassie Blue in
10% acetic acid, 5% methanol.

[0190] The gels were developed via gelatin zymography, and relative protein levels and
activation levels of proMMP-2 and proMMP-9 were quantified from the zymographic
images. A subset of kinases that induced the greatest relative level of proMMP-2 were
selected for a secondary screen. Since MT1-MMP and MMP-9 are both biologically
important MMPs commonly implicated in remodeling processes (including invasion), these
MMPs could share common regulatory pathways. However, the kinases that enhanced the
relative levels of MMP-9, which was detectable only as proenzyme, were mostly distinct
from those enhancing MT1-MMP activity and MMP-2 activation.
[0191] In the secondary MT1-MMP/MMP-2 cascade screen, the expression of 22 kinases
prompted a more than 2-fold increase in pro-MMP-2 activation relative to the control (Figure
1). These kinases included both novel MT1-MMP/MMP-2 cascade regulators and pathway
components downstream of known MT1-MMP-inducing stimuli. Among this later group are
receptors of TGF-P family members and kinases related to the inflammatory signaling
pathways activated by IL-1 or TNF-a. Unexpected hits which significantly increased MMP-2
activation included receptor tyrosine kinases FGFR4 and EphA2.
[0192] FGFR4 significantly increased MMP-2 activation, which is an indirect measure of
MT1-MMP activity. The results signal FGFR4's role as regulator of tumor cell invasion.
Example 2
[0193] This Example demonstrates that FGFR4's modulation of MT1-MMP activity
occurs post-transcriptionally.
[0194] Since MT1-MMP gene expression is frequently upregulated in malignant versus
normal tissues, the potential contribution of selected kinases on MT1-MMP expression was
examined. HT-1080 cells were transfected with expression vectors encoding FGFR4,
EphA2, or the most potent kinases in the TGFß, IL-1, and TNFa pathways. Levels of MT1-
MMP mRNA were determined by quantitative real-time PCR. IRAKI, MAP3K13,
ACVRIC, and EphA2 moderately but significantly increased the levels of MT1-MMP mRNA
(1.5 to 2.5 fold, n=3, p<0.005). Consistent with the in vitro data, correlation blots from the In
Silico Transcriptomics (IST) database containing normalized expression data from 8478
malignant samples revealed significant positive correlations between the expression of MT1-
MMP and IRAKI, EphA2, or ACVR1L in tissue samples from several types of cancers.
Interestingly, FGFR4, which was among the strongest hits in the MT1-MMP/MMP-2 screen,
had negligible effects on MT1-MMP mRNA levels in HT-1080 cells. Likewise, only weak
correlations between FGFR4 and MT1-MMP expression levels were observed in tissue
samples from different types of tumors. This is consistent with independent transcriptional
regulation of these two genes and raises the possibility of more direct mechanism of MT1-
MMP regulation by FGFR4.
[0195] To assess the posttranscriptional effects of kinases on the MT1-MMP levels and
subcellular distribution, MT1-MMP was co-expressed with FGFR4, EphA2, or IRAKI in
COS-1 cells that do not express detectable endogenous MT1-MMP. The cells were subjected
to immunofluorescence staining with anti-MT1-MMP antibodies. Upon viewing, MT1-MMP
appeared mainly localized in intracellular perinuclear compartments in cells transfected with
only MT1-MMP. Interestingly, co-expression of FGFR4 resulted in punctuate MT1-MMP
localization in both cytoplasmic and cell surface membrane structures. These alterations
coincided with increased cell spreading in FGFR4 expressing cells, which was also seen in
EphA2 transfected cells. IRAK increased the intensity of MT1-MMP staining on the cell
surface. The catalytic kinase activities were essential for MT1-MMP regulation, since MT1-
MMP remained mainly perinuclear in cells expressing mutant kinases with inactivating point
mutation in active-site motifs (Varjosalo et al, supra).
[0196] The amount of MT1-MMP protein in transfected cells also was studied. MDA-
MB-231 human breast cancer cells were transiently transfected with expression vectors
encoding various kinases, and the MT1-MMP protein expression assessed by
immunoblotting. The cell surface levels of MT1-MMP protein were assessed by Sulfo-NHS-
biotin labeling and immunoprecipitation with antibodies against MT1-MMP. FGFR4,
IRAKI, and EphA2 all markedly increased the total and cell surface levels of MT1-MMP,
although only IRAKI slightly increased MT1-MMP mRNA expression as assessed by real-
time PCR. Both activated MT1-MMP and the autocatalytically processed 43 kDa form,
which frequently correlates with high cell surface MT1-MMP activity (Lehti et al., Biochem.
J., 334: 345-53, 1998; Lehti et al, J. Biol. Chem., 275: 15006-13, 2000), were detected in
IRAKI and FGFR4 expressing cells. In contrast, MT1-MMP levels were not markedly
affected by most nonfunctional kinases.
[0197] These results suggest that active FGFR4 post-transcriptionally increases MT1-
MMP protein levels and alters its distribution.

Example 3
[0198] This Example demonstrates that FGFR4 and MT1-MMP physically interact and
highlights potential mechanisms of MT1-MMP regulation by FGFR4.
[0199] To examine possible mechanisms of MT1-MMP regulation by FGFR4, double
immunofluorescence staining was performed for FGFR4 or IRAKI and MT1-MMP in MDA-
MB-231 cells transfected with expression vectors encoding the respective kinases. FGFR4-
transfected cells were incubated with or without FGF2 (25 ng/ml) for 30 minutes, and then
stained with immunofluorescent antibodies against MT1-MMP and FGFR4. IRAKI
transfected cells were incubated with or without IL-1ß (5 ng/ml) for 30 minutes, and
immunofluorescent staining was carried out for MT1-MMP and IRAKI. Interestingly,
FGFR4 largely co-localized with MT1-MMP in vesicular membrane structures of untreated
cells, and FGF2 stimulation further enhanced this co-localization. In IRAKI expressing
cells, MT1-MMP localized prominently on the cell surface with and without IL-ip
stimulation. Unlike FGFR4, cytoplasmic IRAK staining did not specifically co-localize with
MT1-MMP, although both proteins tended to accumulate at the same regions of the cell.
[0200] To define whether MT1-MMP and FGFR4, which were located in the same
membrane vesicles, would physically interact in the same membrane receptor complexes, the
cells were co-transfected to produce HA-tagged MT1-MMP and V5-tagged FGFR4. This
was followed by immunoprecipitation and immunoblotting analysis. For immunoblotting
experiments, confluent cell cultures were washed and incubated for 24 hours in serum free
DMEM. Transiently transfected cells were incubated for 16 hours after transfection before
transferring to serum free medium. The conditioned media was then harvested and cell
lysates prepared as described (Lehti et al. 1998, supra). SDS-PAGE was carried out using 4-
20% gradient Laemmli polyacrylamide gels (Bio-Rad, Hercules, CA). The proteins were
transferred to nitrocellulose membranes, and their immunodetection was performed as
described (Lohi et al., Eur. J. Biochem., 239: 239-47, 1996). The immunoblotting analysis
was performed using antibodies against the protein tags, HA and V5.
[0201] FGFR4 was clearly detectable by immunoblotting with anti-V5 antibodies in MT1-
MMP complexes that had been immunoprecipitated with anti-HA antibodies. Likewise, HA-
tagged MT1-MMP was detected in the FGFR4 complexes immunoprecipitated with anti-V5
antibodies. The interaction between FGFR4 and MT1-MMP was specific, since no

interactions between V5-tagged IRAKI and HA-tagged MT1-MMP were detected under the
same experimental conditions.
[0202] Since FGFR4 has been reported to be mostly recycled after endocytosis, the effects
of FGFR4 on endosomal localization of MT1-MMP were assessed. Immunofluorescence
analysis with antibodies against endosomal marker proteins (clathrin and EEAl) revealed that
MT1-MMP interaction with FGFR4 coincided with the enhanced levels of MT1-MMP in
intracellular clathrin and EEAl positive endosomal vesicles. By contrast, prominent MT1-
MMP staining was detected on the surface of IRAKI expressing cells. These results are
consistent with potentially enhanced stability of endocytosed MT1-MMP by interaction with
FGFR4.
[0203] To define the contribution of lysosomal degradation for the regulation of MT1 -
MMP activity and protein expression by IRAKI and FGFR4, MDA-MB-231 cells expressing
these kinases were incubated with a lysosomal inhibitor, Bafilomycin A (Calbiochem), and a
proteosome inhibitor, MG132 (MG-132, Z-Leu-Leu-CHO; Peptide Institute Inc., Osaka,
Japan). MDA-MB-231 cells were transfected with empty pCR3.1 expression vector (mock)
and corresponding vectors coding for IRAKI or FGFR4. Cells were immunostained with
anti-MT1-MMP and anti-clathrin antibodies and assessed by confocal imaging. FGFR4
transfected cells were immunostained with anti-MT1-MMP and anti-LAMPl antibodies. The
effect of MG132 on the levels of MT1-MMP was marginal. In contrast, the inhibition of
lysosomal degradation by Bafilomycin A increased MT1-MMP protein levels in control cells,
which is consistent with the reported rapid and constitutive lysosomal degradation of MT1-
MMP. Importantly, FGFR4 specifically increased the levels of MT1-MMP in untreated cells,
thus decreasing the differences in MT1-MMP protein levels between non-treated and
Bafilomycin A treated cells. Accordingly, MT1-MMP localization in LAMP-1 positive
lysosomal structures was markedly decreased in FGFR4 expressing cells as compared to
control cells.
[0204] The results of this Example indicate that FGFR4 inhibits lysosomal sorting and
degradation of MT1-MMP, thus increasing the cellular levels of active MT1-MMP. The
FGFR4-mediated increase in MT1-MMP, by extension, modulates tumor cell invasiveness.
Example 4
[0205] This Example illustrates the functional significance of kinase-mediated MT1-MMP
regulation using a three-dimensional collagen invasion assay, i.e., a tumor cell invasion
assay.
[0206] Pericellular collagen degradation is the major established biological function of
MT1-MMP. Kinase-mediated modulation of MT1-MMP activity was determined in a three-
dimensional collagen invasion assay. MDA-MB-231 human breast cancer cells were
transfected with expression vectors encoding FGFR4, EPHA2, and IRAKI, or with
respective inactivated kinases. The cells were seeded on top of type I collagen gels and
allowed to invade for 5 days. The gels were fixed and embedded in paraffin. Cells that
invaded into the collagen matrix were visualized and counted from hematoxylin and eosin
stained sections.
[0207] The overexpression of each of the active kinases significantly increased the
otherwise relatively slow invasion of unstimulated MDA-MB-231 cells. The expression of
IRAKI, FGFR4, or EphA2 each resulted in over four-fold increased rates of invasion
compared to the mock transfected cells. As expected, inactivated kinases had negligible
effects on cell invasion. Noteworthy, only FGFR4 and EphA2 significantly increased the
number of cells that invaded over 20 urn (12.1 and 9.8 fold, respectively). FGFR4 increased
by 20-fold the number of MDA-MB-231 cells that invaded greater than 100 µm. MT1-MMP
and FGFR4 colocalized at both the leading edge and in intracellular vesicles of the rapidly
invading cells.
[0208] In addition to studying cell invasion, matrix degradation was examined.
Transfected cells were seeded on Alexa 488 gelatin coated coverslips and allowed to attach
and spread in the presence of GM6001 (10 µM) for three hours. After washing out the MMP
inhibitor, cell-mediated gelatin degradation was carried out for 20 minutes in complete
medium. Fixed and permeabilized cells were immunostained with anti-MT1-MMP
antibodies. Degradation was visualized by confocal microscopy and quantified as the ratio
between degradation area and cell number from low magnification figures (mean ± 1 SD,
n = 3). In addition, transfected cells were incubated on crosslinked collagen matrices for
three hours and immunostained for MT1-MMP and FGFR4.
[0209] Consistent with the higher rates of collagen invasion, cells expressing IRAKI,
FGFR4 and EphA2 were also able to degrade fluorescent gelatin substrate more efficiently

than the control cells or cells transfected with KD kinases. Interestingly, FGFR4 expression
in the cells led to markedly polarized foci of pericellular matrix proteolysis that colocalized
with MT1-MMP clusters in one edge of the cells. FGFR4 expressing cells were also able to
degrade and traverse crosslinked collagen matrix within 3 hours. In contrast, IRAKI, which
efficiently increased gelatin degradation and the levels of MT1-MMP at cell-matrix
adhesions, failed to increase the amount of degraded holes in layers of crosslinked 3-D
collagen within the same time period.
[0210] These results suggest a specific role for FGFR4 in inducing highly invasive cancer
cell phenotype, where MT1-MMP functions coordinately with the cellular motile machinery
to drive invasion in cross-linked 3-D collagen.
Example 5
[0211] This Example compares the effects of two FGFR4 alleles on MT1-MMP function
and cell invasion and establishes FGFR4 as a novel target for inhibiting cancer cell invasion.
[0212] A single nucleotide polymorphism in the FGFR4 gene has been linked to poor
prognosis in patients with several types of tumors such as breast, prostate, and colon
adenocarcinomas as well as head and neck squamous cell carcinomas, melanomas, and soft-
tissue sarcomas. In the corresponding FGFR4 variant, glycine 388 in the transmembrane
domain is changed to arginine (R388). Interestingly, sequence analysis revealed that the
FGFR4 cDNA that was included in the kinome library described in Example 1 encoded the
Arg388 variant. To compare the effects of the G388 FGFR4 allele and the R388 FGFR4
allele on MT1-MMP function and cell invasion, cDNA encoding the R388 FGFR of the
kinome library was modified to encode wild-type G388 FGFR4. MDA-MB-231 cells were
transiently transfected with expression vectors encoding the FGFR4 alleles (FGFR4G388-
V5-His and FGFR4R388-V5-His) and corresponding non-functional kinases, as well as
expression vectors encoding HA-tagged MT1-MMP. Cells were incubated with Bafilomycin
A or GM6001 for 16 hours, and the levels of MT1-MMP and FGFR4 were assessed by
immunoblotting. Cell media was harvested and cell lysates prepared as described (Lehti et al.
1998, supra). SDS-PAGE was carried out using 4-20% gradient Laemmli polyacrylamide
gels (Bio-Rad, Hercules, CA). The proteins were then transferred to nitrocellulose
membranes, and their immunodetection was performed as described (Lohi et al., supra).
[0213] Lysosomal degradation inhibition by Bafilomycin A in control MDA-MB-231 cells
increased MT1-MMP protein levels. The expression of the FGFR4 R388 variant relatively
decreased this effect by increasing the levels of MT1-MMP in non-treated cells. In contrast,
MT1-MMP levels were not markedly increased by the expression of FGFR4 G388. In
FGFR4 G388 expressing cells, the increase in MT1-MMP levels after Bafilomycin treatment
was coupled with considerable FGFR4 down-regulation. The effect was less apparent in cells
expressing R388 or either one of the inactivated kinases. Inhibition of MT1-MMP activity by
GM6001, a synthetic MMP inhibitor, correlated with detection of endogenous FGFR4 in
control cells that normally express FGFR4 G388 at levels undetectable by immunoblotting.
[0214] The opposite effects mediated by FGFR4 G388 and FGFR4 R388 on MT1-MMP
protein expression also were evident in co-transfection experiments. FGFR4 Gly388
expression was coupled with decreased MT1-MMP protein levels, while the FGFR4 Arg388
variant enhanced the relative levels of MT1-MMP. FGFR4 Arg388 co-precipitated
prominently with MT1-MMP. Both alleles and the inactive kinase mutants were, however,
detectable in MT1-MMP immunoprecipitates.
[0215] The effect of the two FGFR4 alleles on cell invasion also was examined using
methods similar to those described in Example 4. cDNA coding for the G388 allele was
stably expressed in MDA-MB-231 cells, which were plated atop type 1 collagen gels. While
FGFR4 R388 expressing cells invaded collagen at increased rates compared to mock
transfected cell, the FGFR4 G388 expressing cells invaded at the same rate as, or even more
slowly than, the mock transfected cells. Collagen invasion was abolished by lentiviral MT1-
MMP silencing RNA, confirming the functional link between MT1-MMP and FGFR4 R388
in driving the invasion.
[0216] This Example demonstrates that FGFR4 Gly388 and MT1-MMP are reciprocally
downregulated through a mechanism that depends on the respective kinase and
metalloproteinase activities. Therefore, Gly388 and/or Arg388 FGFR4 variants are targets
for inhibiting tumor cell invasion.
Example 6
[0217] The results of this Example establish that FGFR4 and MT1-MMP are co-expressed
in invasive cancer cells in vivo, further supporting a functional link between the molecules.
[0218] The mRNAs for MT1-MMP and FGFR4 are frequently co-expressed in samples
from different types of human cancers. While expression in individual tissue samples do not
correlate well, the mean expression levels of both MT1-MMP and FGFR4 are frequently
upregulated in different tumor types including colon, testis, uterus and breast carcinomas.
These results suggest that, when both proteins are expressed in the same cells in vivo, their
interaction functionally contributes to the invasiveness of human tumors. To further examine
co-localization of MT1-MMP and FGFR4 in different tumor and stromal cell populations,
frozen tissue arrays containing 40 malignant and normal breast tissue samples were obtained
and immunostained with anti-MT1-MMP and anti-FGFR4 antibodies. The anti-FGFR4
antibodies were well established polyclonal antibodies that do not cross-react with other
FGFRs in cultured cells. Monoclonal anti-MT1-MMP antibodies produced by immunization
of an MT1-MMP-/- mouse (Ingvarsen et al, Biol. Chem., 389: 943-53,2008) were used.
When tested in immunofluorescence staining of MDA-MB-231 breast cancer cells, the anti-
MT1-MMP antibodies readily detected endogenous MT1-MMP, while the staining was
completely blocked after siRNA mediated MT1-MMP knock-down. The specimens were
analyzed with a Leica microscope.
[0219] FGFR4 was observed to be localized predominantly to breast epithelial and
carcinoma cells. FGFR4 was detected in ductal epithelial cells in all four cases of normal
breast analyzed, and the relative levels were frequently increased in ductal carcinoma cells
(strong staining in 20/36 cases). MT1-MMP was significantly upregulated in the reactive
stroma in breast carcinomas (28/36 cases), especially in the myoepithelium adjacent to the
carcinoma cells. This was observed both in invasive and noninvasive areas of the
carcinomas. Importantly, MT1-MMP was specifically detected in the carcinoma cells that
were located in the invasive fronts of tumors and in cells of poorly differentiated breast
carcinomas (10/36 cases), where MT1-MMP prominently co-localized with FGFR4.
Immunohistochemistry of multiple frozen tissue arrays with samples from 14 different tumor
types and corresponding normal tissues revealed that the relative levels of MT1-MMP
staining were also enhanced in most other malignant tissues (10/14 cases). MT1-MMP was
frequently upregulated in the reactive stroma (6/14 cases) and co-expressed with FGFR4 in
tumor cells (8/14 cases). As with breast carcinomas, MT1-MMP in other types of
carcinomas, such as colon adenocarcinomas, was prominently expressed in the tumor cells at
the invasive fronts.

[0220] Expression of FGFR4 alleles (FGFR4 G388 and FGFR4 R388) and MT1-MMP in
vivo was examined using qPCR array coupled with FGFR4 sequencing from 48 human breast
cancer cDNA samples. RNA was extracted with RNeasy Mini Kit (Qiagen) followed by
reverse transcription with random hexamer primers (Invitrogen) and Superscript II reverse
transcriptase (Life Technologies). mRNA expression was quantified as described (Tatti et
al., Exp. Cell Res., 314: 2501-2514, 2008) using TaqMan Universal PCR Master Mix and
validated primers (MT1-MMP; Hs 01037006_gH, MT2-MMP; Hs 00233997_ml, MT3-
MMP; Hs 00234676_ml, MT4-MMP; Hs 00211754ml, MT5-MMP; Hs 00198580ml, MT6-
MMP; Hs 00360861_ml; MMP-9; Hs00957555_ml; FGFR1; Hs00915140_ml, FGFR4;
Hs00242558_ml (Applied Biosystems)). The expression was normalized with TATA-
binding protein (TBP) and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mRNA
expression. The fragments of FGFR4 cDNA containing the Gly388 to Arg388 site (base
pairs 1329-1331) were amplified by PCR (primers: TACCAGTCTGCCTGGCTC (SEQ ID
NO: 17) and AGTACGTGCAGAGGCCTT (SEQ ID NO: 18)) and digested with BstN1
(New England BioLabs). The R388 allele was identified by a specific 126 base pair
fragment. The G388 allele was identified by two fragments (97 and 29 base pairs).
[0221] Of the 48 human breast cancer cDNA samples, 22 had homozygous G/G (46%); 22
had heterozygous G/R (46%); and four had homozygous R/R (8%) alleles of FGFR4.
Consistent with the poor prognosis reported in cancer patients, all four cDNA samples with
homozygous R388 alleles were from the highest grade (3) breast carcinomas. MT1-MMP
mRNA was prominently expressed in all of these tumors co-incident with a lower FGFR4
R388 expression.
[0222] The observations described in this Example strongly suggest that FGFR4 and MT1-
MMP have a synergistic effect on tumor invasiveness, and confirm that the FGFR4 R388
allele strongly correlates with highly invasive cancers.
Example 7
[0223] This Example illustrates the functional significance of FGFR4 R388 activation and
FGFR4 G388 suppression in prostate carcinoma cell invasion of collagen.
[0224] Since MT1-MMP and FGFR4 were most frequently co-expressed in human prostate
cancers, corresponding cell lines were assessed for endogenous MT1-MMP and FGFR4
expression. PC3 and DU145 prostate adenocarcinoma cells expressed notable levels of

FGFR1, MT1-MMP, and either the R388 (PC3 cells) or G388 (DU145 cells) variant of
FGFR4. Notably, only PC3 cells expressing a homozygous FGFR4 R388 allele invaded
collagen gels efficiently, while the corresponding DU145 cells homozygous for the FGFR4
G388 allele did not. FGFR4 siRNAs inhibited 87% of the PC3 cell invasion, which also was
essentially blocked by TIMP-2 -mediated MT1-MMP inhibition. Consistent with the
reported induction of prohferation and motility through both FGFRl and FGFR4 (Sahadevan
et al., J. Pathol, 213: 82-90, 2007), siRNAs transiently targeting either one of these FGFRs
reduced FGF2-induced ERK phosphorylation. However, when FGFRl mRNA expression
was silenced, collagen invasion increased, suggesting that FGFRl has a distinct function.
Notably, although the DU145 cells expressed less MT1-MMP mRNA, these cells also
invaded collagen after transfection with FGFR4 R388.
[0225] Cell growth also was examined. Stable MT1-MMP or FGFR4 knockdown by
lentiviral shRNAs did not significantly alter the normal monolayer growth of either PC3 or
DU145 cells. However, when the cells were plated inside a growth-restricting 3-D matrix
composed of cross-linked collagen I, MT1-MMP silencing significantly decreased PC3 and
DU145 cell growth and invasion. Likewise, FGFR4 R388 knockdown inhibited the growth
and invasion of PC3 cells, as well as decreasing MT1-MMP content and fibroblast receptor
substrate-2 (FRS2) phosphorylation. In contrast, knockdown of FGFR4 G388 in DU145 cells
increased invasive properties of those cells in collagen coincident with increased levels of
endogenous MT1-MMP. FRS2 and ERK phosphorylation also increased in stable FGFR4
G388 and MT1-MMP knockdown cells. Even strong inhibition of FRS2 phosphorylation in
PC3 cells was not associated with ERK activation, suggesting that pathways other than
FGFR4 activation are involved in mitogenic FGF signaling.
[0226] Altogether, the results described above identify FGFR4 R388 as a novel co-factor
in MT1-MMP-driven PC3 tumor cell invasion and growth in 3-D collagen, and suggest that
the G388 and R388 alleles have opposite effects on MT1-MMP-dependent invasion.
Example 8
[0227] This Example demonstrates that FGFR4 R388 induces MT1-MMP phosphorylation
and endosomal stabilization, highlighting an additional potential mechanism of MT1-MMP
regulation by FGFR4.

[0228] MT1-MMP cytoplasmic tail contains a single tyrosine residue that can be
phosphorylated by Src (Nyalendo et al., J. Biol. Chem., 282: 15690-15699, 2007). To
determine if FGFR4 can induce MT1-MMP phosphorylation, FGFR4 G388 and FGFR4
R388 were co-expressed with HA-tagged MT1-MMP followed by MT1-MMP
immunoprecipitation. MT1-MMP tyrosine phosphorylation was repeatedly detected in COS1
cells co-expressing MT1-MMP with either allele of FGFR4, but not in cells expressing
FGFR4 protein having non-functional kinase domains or only MT1-MMP. FGFR4 R388 and
MT1-MMP co-localized mainly in intracellular vesicles. FGF2-treatment increased FGFR4
R388 autophosphorylation, as well as MT1-MMP phosphorylation and endosomal
accumulation, suggesting that activated FGFR4 R388 induces MT1-MMP phosphorylation in
the complexes. This interaction increased the stability of endocytosed MT1-MMP, as
indicated by enhanced co-localization with early endosomal antigen-1 (EEA1) and clathrin,
and reduced co-localization with lysosome-associated membrane protein-1 (LAMP1). In
contrast, the degree of co-localization of MT1-MMP with the weakly/transiently activated
FGFR4 G388 or the kinase-deficient KD proteins in the intracellular vesicles was
significantly lower than that observed with FGFR4 R388.
[0229] MT1-MMP's only intracellular tyrosine residue was mutated to phenylalanine
(MT1-Y/F) to assess the significance of MT1-MMP phosphorylation. The Y573F mutation
did not appear to alter MT1-MMP-activity in HT-1080 cells that express very little
endogenous FGFR4. However, the mutation abrogated the co-localization of wild-type MT1-
MMP and endogenous FGFR4 R388 in cell-cell contacts and intracellular vesicles of MDA-
MB-231 cells. MT1-Y/F was localized predominantly at the cell surface, while FGFR4 R388
translocated into intracellular vesicles. Consistent with enhanced cell-surface MT1-MMP,
the mutation enhanced cell growth and invasion in 3-D collagen. Accordingly, fewer FGFR4
R388/MT1-Y/F complexes were observed compared to FGFR4 R388/MTI-MMP complexes
in transfected MDA-MB-231 cells. While the FGFR4 R388 levels were slightly decreased,
FGFR4 G388 protein and FGFR4 G388/MT1-Y/F complexes were sufficiently suppressed to
be undetectable in cells with high MT1-Y/F content. Consistent with the increased detection
of endogenous FGFR4 G388 after MT1-MMP inhibition, overexpressed FGFR4 G388 also
was suppressed by wild type MT1-MMP, but not by mutant MT1-E/A in which proteinases
activity is abrogated.
[0230] The observations described above suggest that FGFR4 suppression by
unphosphorylated MT1-MMP provides FGFR4 G388-containing cells a feedback mechanism

to sustain pro invasive MT1-MMP activity. While MT1-MMP did not alter the
phosphorylation of FGFR4 G388, FGFR4 R388 phosphorylation was markedly enhanced by
MT1-MMP, but not an inactive MT1-MMP mutant. The interactions, phosphorylation, and
trafficking of MT1-MMP/FGFR4 R388 complexes thus seem to support their synergistic
functions in cell invasion.
Example 9
[0231] This Example demonstrates that endogenous FGFR4/MT1-MMP activity controls
tumor growth and invasion in vivo.
[0232] To determine if targeting of the FGFR4/MT1-MMP axis regulates tumor cell
behavior in vivo, tumor growth, morphology, and extracellular matrix (ECM) composition
were analyzed after subcutaneous injection of PC3 and DU145 cells into SCID mice. PC3
and DU145 cells were lentivirally transduced with a renilla luciferase-green fluorescent
protein (GFP)-fusion reporter protein. Stable cell pools expressing scrambled, MT1-MMP,
and FGFR4-targeting short-hairpin-RNAs were produced by lentiviral transduction followed
by puromycin (Sigma) selection. Greater than 90% knockdown efficiencies were confirmed
by qPCR. The cells (2 x 106) were implanted into the abdominal subcutis of ICR-SCID male
mice (5-7 wks of age; Taconic) and allowed to grow for 6-8 weeks. Tumor size was
measured with a caliper and noninvasive bio luminescence, which was visualized after
intraperitoneal injection of 35 µg/100 µl coelentetrazine using a Xenogen IVIS System
(Xenogen).
[0233] Stable silencing of either MT1 -MMP or FGFR4 R388 dramatically decreased the
growth rates of PC3 tumors and the number of stromal vessels containing intravasated tumor
cells. A fibrous capsule accumulated around the tumor, and intratumoral extracellular matrix
(ECM) separated the MT1-MMP and FGFR4 R388 knock-down tumor cells into small
compartments that showed decreased rates of proliferation. The mitotic index, growth,
invasion, and metastasis of the tumors correlated inversely with collagen and other ECM
protein content. At the same time, cells exhibited increased polarization towards collagen IV,
fibronectin and laminin, and increased acinar lumen formation was detected.
[0234] Consistent with observations from in vitro assays, MT1-MMP silencing also
reduced the growth of DU145 tumors while increasing collagen content. In MT1-MMP
knockdown DU145 tumors, FGFR4 G388 mRNA expression increased between two and

four-fold, suggesting that a transcriptional feedback mechanism was involved in FGFR4
G388 suppression by MT1-MMP. In contrast, FGFR4 G388 silencing produced more
pronounced invasion and extravasation of the DU145 tumor cells while reducing collagen
accumulation inside the tumors and at the tumor edge. No significant changes were detected
in collagen mRNA expression by qPCR.
[0235] The results described above demonstrate that silencing either component of the
MT1-MMP/FGFR4 R388 complex inhibited tumor growth, invasion and metastasis. While
not wishing to be bound to any particular theory, the inhibition appeared to result from
blocking the proteolytic degradation of ECM that physically restricts tumor spread and
promotes epithelial differentiation. Silencing FGFR4 G388 achieved the opposite effect.
Example 10
[0236] This Example provides an exemplary method of generating anti-FGFR4
monoclonal antibodies, such as the antibodies of the present invention. The Example also
provides a method of characterizing the binding affinity of an anti-FGFR4 antibody or
fragment thereof.
[0237] A baculovirus expression vector encoding the extracellular region of FGFR4 linked
to a His-tag was constructed according to methods standard in the art. Recombinant
baculoviruses were generated by co-transfection of Sf9 cells with a recombinant FGFR4
ectodomain coding vector and linearized BACULOGOLD™ DNA (Pharmingen). High Five
cells were infected with the viral stocks obtained from Sf9 cells, and the recombinant His-
tagged FGFR4 protein was purified using nickel columns for immunization. Hybridoma
clones were generated using standard methods and subcloned as required. Ascites fluids or
culture medium from cultured hybridoma cell clones were screened by immunoblotting using
recombinant FGFR4 ectodomain/Fc-fusion protein (R&D systems). Positive clones were
further assessed by immunoblotting using lysates of control and FGFR4-transfected COS-7
cells, as well as by immunofluorescence of corresponding cells. Monoclonal antibodies were
purified using HiTrap Protein G columns according to the instructions (GE Life Sciences).
Three monoclonal antibodies, F85-6C5, F90-3B6 and F90-10C5, were subjected to function
blocking analysis.
[0238] The affinities of the mAbs 3B6, 6C5 and 10C5 for FGFR4-Fc were compared using
a receptor binding assay in the enzyme linked immunoassay format (Figure 2). Recombinant

human FGFR4-Fc, comprising the FGFR4 extracellular domain (amino acid residues 1-369,
Partanen et al, Proc. Natl. Acad. Sci. USA, 87: 8913-8917, 1990) fused to the
carboxyterminal region of human IgG (amino acid residues 100-330) via a polypeptide
linker, was obtained from R&D Systems (Cat # 685-FR). Microtiter plate wells
(ThermoElectron, Cat #95029100) were coated with recombinant human FGF2 (R&D
Systems, Cat# 233-FB) and heparin (Sigma-Aldrich, Cat# H-3149) in 0.1M NaHCO3, pH
9.5. The wells were washed (100 mM Tris, 150 mM NaCl, 0.1% (v/v) Tween 20, pH 7.5)
and available non-specific protein binding sites were blocked with PBS, 0.05 % (v/v) Tween
20, 0.5 % BSA. The wells were washed a second time. FGFR4-Fc was preincubated with a
dilution series of each mAb (3B6, 6C5 or 10C5) in 100 mM Tris, 150 mM NaCl, 0.1%
Tween 20,1% BSA, 0.1 µg/mlheparin, pH 7.5, before competitive binding to the FGF2-
heparin-coated wells. After addition of preincubated FGFR4-Fc and further incubation, the
FGF2 coated wells were washed. Bound FGFR4-Fc was detected using goat anti-human
alkaline phosphatase conjugate (Sigma-Aldrich, Cat# A9544).
[0239] As shown in Figure 2, the potencies of the mAbs 3B6, 6C5 and 10C5 for blocking
the binding of FGFR4-Fc to immobilized FGF2 differed. Maximal FGFR4-FGF2 interaction
is seen with low concentrations of the blockers, while at higher concentrations the interaction
(measured in absorbance units) is inhibited to the level of background absorbance. The
mAbs 3B6 and 6C5 blocked the binding of FGFR4 to immobilized FGF2 with similar
potencies as soluble FGF2 or FGFl. Also, the half-maximal inhibitory concentrations (IC50)
of mAbs 3B6 and 6C5 were close to those of FGF2 and FGFl (1.077 nM, 0.3019 nM, 0.6914
nM, and 0.7334 nM, respectively). mAb 10C5 blocked FGFR4-FGF2 interaction more
weakly than the two other mAbs, as indicated by a shift of the blockage curve toward higher
concentrations of blocker and from a higher half-maximal inhibitory concentration (6.217
nM). The binding of FGFR4 to immobilized FGF2 was specific, indicated by a very low
absorbance level obtained from heparin coated wells (after preincubation with soluble FGF2
or without soluble ligand). Use of the mAbs without prior addition of FGFR4 provided
absorbance values at background level. Thus, the assay measures the blockage of FGFR4-
FGF2 binding specifically.
[0240] Differential binding of mAbs 3B6, 6B5, and 10C5 to a FGFR4-Fc fusion protein
immobilized on a biosensor chip were measured in BIAcore assays, i.e., surface plasmon
resonance using a biosensor (BIAcore 2000®, BIAcore AB). FGFR4-Fc was diluted in 10
mM sodium acetate buffer, pH 4.7, and amine-coupled to a BIAcore sensor chip. The

amount of immobilized FGFR4-Fc used generated a 1000 response unit signal when saturated
with the anti-FGFR4 monoclonal antibodies. An uncoupled biosensor chip channel was used
to measure unspecific background signal, which was subtracted from the signal obtained
from the FGFR4-Fc coupled channel. A dilution series of each mAb was injected over
FGFR4-Fc on a biosensor chip and binding measured in relative response units. Each mAb
(3B6, 6C5 or 10C5) was injected at 10 nM to 240 nM in PBS buffer. The flow rate was
maintained at 20 µl/min and a 5 minute binding phase was used. Following mAb injection,
the flow was exchanged with PBS buffer to determine the rate of dissociation. The sensor
chip was regenerated between cycles with a 30 second pulse of 10 mM glycine, pH 2.2.
Kinetics were analyzed by 1:1 Langmuir fitting using BIAcore evaluation software 3.1. For
comparison, Kd values were also estimated by plotting the maximal relative response units
obtained with a dilution series of each mAb. Figures 4A-C show the concentration of the
mAb on the X-axis and the obtained response units on the Y-axis. The dissociation
equilibrium constant (Kd) of each mAb was estimated, after curve fitting, by the
concentration half-way between the obtained maximal and minimal response units. Based on
these Kd values, the affinity of mAb 6C5 to immobilized FGFR4-Fc (Kd 1.8x10-8 M) is
significantly higher than that of mAb 3B6 or mAb 10C5 which have similar affinities (Kds
2.17 x 10-7 and 1.33 x 10-7 M, respectively).
[0241] In addition, the FGFR4 epitope recognized by F90-10C5 was identified. A PepSpot
array composed of a series peptides covering the amino acid sequence of the extracellular
domain of FGFR4 (excluding signal sequence) was obtained. The peptides were 15 amino
acids in length and comprised sequences that overlapped by three amino acids. An
immunoblotting assay was performed using the monoclonal anti FGFR4 antibodies described
above. mAb 6C5 and 3B6 did not recognize linear epitopes, whereas 10C5 detected the
following peptides: YKEGSRLAPAGRVRG (SEQ ID NO: 5); GSRLAPAGRVRGWRG
(SEQ ID NO: 6); LAPAGRVRGWRGRLE (SEQ ID NO: 7); AGRVRGWRGRLEIAS (SEQ
ID NO: 8); and VRGWRGRLEIASFLP (SEQ ID NO: 9). The peptide comprising SEQ ID
NO: 7 prompted the strongest signal. The location of SEQ ID NOs: 5-9 in the extracellular
region of FGFR4 is illustrated in Figures 3A-3C.
[0242] Hybridoma 10C5, which produces antibody F90-10C5, was transferred to Deutsche
Sammlung von Mikroorganismen und Zellkulturen GmbH (DSMZ), Mascheroder Wep lb.
D-38124, Germany, on September 2, 2008, under the provisions of the Budapest Treaty for
the International Recognition of the Deposit of Microorganisms for the Purpose of Patent

Procedure ("Budapest Treaty"), and assigned Deposit Accession No. DSM ACC2967 on
September 4, 2008. Hybridomas F85-6C5 and F90-3B6, which produce mAb F85-6C5 and
F90-3B6, respectively, also were deposited with DSMZ under the provisions of the Budapest
Treaty, and were assigned Deposit Accession Nos. DSM ACC2966 (F85-6C5) and DSM
ACC2965 (F90-3B6) on September 4, 2008.
Example 11
[0243] This Example examines the contribution of FGFR4 G388 and FGFR4 R388 to
collagen invasion, and describes inhibition of MT1-MMP-mediated cancer cell invasion
using a monoclonal anti-FGFR4 antibody.
[0244] Given the distinct effects of the two FGFR4 variants on the levels of MT1-MMP,
their contribution to collagen invasion was determined. MDA-MB-231 cells were transfected
with an expression vector encoding either FGFR4 G388 or the FGFR4 R388 variant. A
portion of the transfected cells was treated with 10 µg/ml control IgG or the monoclonal anti-
FGFR4 antibodies 3B6, 6C5 and 10C5. The transfected cells were plated on 3-D collagen in
a dual chamber apparatus. FGF2 (25 ng/ml) was used as a chemoattractant to stimulate
invasion. The cells were allowed to invade for 5 days before quantification of the invasive
foci.
[0245] FGFR4 R388-expressing MDA-MB-231 cells invaded at higher rates than either
mock transfected or FGFR4 G388 expressing cells. Invasion of control cells, FGFR4 G388-
expressing cells, and FGFR4 R388-expressing cells was completely blocked by reducing
MT1-MMP mRNA (85% reduction) using lentiviral shRNA against MT1-MMP (Open
Biosystems, Huntsville, AL). These results further support the identified functional link
between FGFR4 and MT1-MMP in cancer cell invasion.
[0246] Monoclonal antibodies F85-6C5, F90-3B6, and F90-10C5, which compete for
ligand binding to FGFR4, were added to both the upper and lower chamber of the dual
chamber collagen assay. FGFR4 R388-induced invasion was efficiently inhibited by the
10C5 monoclonal anti-FGFR4 antibody (Figure 5). In contrast, the 3B6 and 6C5 antibodies
tended to enhance the invasion of both FGFR4 G388 and R388 expressing cells (Figure 5).
[0247] FGFR4 activation in the presence of anti-FGFR4 antibodies was examined to
elucidate possible mechanisms behind inhibition of cell invasion. Serum-starved MDA-MB-
231 cells expressing FGFR4 R388 or FGFR4 G388 (both tagged with V5) were pretreated

with F85-6C5, F90-3B6, and F90-10C5 antibodies (10 µg/ml) overnight and left unstimulated
or incubated with FGF2 (10 ng/ml) for 15 minutes. The cell extracts were subjected to
immunoprecipitation with antibodies against FGFR4, followed by immunoblotting using
antibodies against V5, phosphotyrosine residues, FGFR1, or phosphorylated forms of
ERK1/2. The immunoblot is depicted in Figure 6. Separately, COS-1 cells were transfected
to express V5-tagged FGFR4 G388, V5-tagged FGFR4 R388, and FGFR1 alone or in
combination. After serum starvation, the cells were pretreated with the anti-FGFR4
antibodies (10 µg/ml) for 30 minutes. A portion of the transfected cells were left
unstimulated, while others were incubated with FGF2 (10 ng/ml) for 15 minutes. FGFR4
proteins were immunoprecipitated and immunoblotted with anti-phosphotyrosine and anti-V5
antibodies. The immunoblot is depicted in Figure 7.
[0248] Interestingly, FGFR4 G388 was prominently autophosphorylated in the presence
and absence of ligand stimulation, whereas phosphorylation of the FGFR4 R388 variant was
highly increased after incubation with FGF2. Treatment of FGFR4 G388 expressing cells
with invasion promoting F85-6C5 antibodies (i) suppressed ligand-independent FGFR4
autophosphorylation but (ii) enhanced FGF2-induced phosphorylation. F85-6C5 antibodies
modestly suppressed ligand-independent autophosphorylation of FGFR4 R388, whereas
ligand-stimulated phosphorylation was not notably affected. Of note, the phosphorylation
patterns of mAb 6C5 treated cells expressing either FGFR4 variant were analogous.
Treatment of cells expressing FGFR4 R388 with mAb F90-10C5 reduced Ugand-independent
and ligand-induced phosphorylation of the protein. Phosphorylation was further reduced
when cells were exposed to both mAb 10C5 and mAb 3B6, which binds a different FGFR4
epitope compared to mAb 10C5 (Figure 6).
[0249] Since MDA-MB-231 cells express high levels of endogenous FGFR1, the potential
effects of FGFR4 expression on total FGFR1 levels and the activation of downstream ERK
pathway was analyzed. In both control and FGFR4 R388 expressing cells, FGF2 slightly
increased ERK1/2 phosphorylation without notably affecting FGFR1 levels. In contrast,
FGFR1 levels were markedly decreased in FGFR4 G388 expressing cells coincidentally with
the suppression of FGF2 induced ERK1/2 phosphorylation. Interestingly, the treatment of
FGFR4 G388 expressing cells with mAb 6C5, but not with invasion-blocking 10C5
antibodies, rescued both FGFR1 levels and ERK activation after FGF2 stimulation. This is
consistent with functional co-operation between FGFR1 and FGFR4 that may contribute to
tumor cell invasion and be affected by the invasion-modulating anti-FGFR4 antibodies.

[0250] These results were confirmed using transfected COS-1 cells which do not naturally
express FGFRs. FGFR4 G388 was prominently autophosphorylated in the absence of ligand
stimulation, whereas the phosphorylation of R388 variant was highly increased after 15
minutes incubation with FGF2. The treatment of FGFR4 G388 expressing cells with
invasion promoting mAb 6C5 resulted in suppression of ligand-independent FGFR4
autophosphorylation and enhanced FGF2 induced phosphorylation. Modest ligand-
independent autophosphorylation of FGFR4 R388 was also suppressed by mAb 6C5, whereas
the ligand-stimulated phosphorylation was not notably affected. Antibody 6C5 also reduces
FGFR4/FGFR1 heterodimerization following FGF2 stimulation. Of note, the
phosphorylation patterns of mAb 6C5 treated cells expressing either FGFR4 variant were
analogous. Consistent with the predicted functional FGFR1/FGFR4 interaction, co-
expression of these receptors resulted in markedly increased ligand-independent
phosphorylation of FGFR4 G388 and R388. This was not notably affected by mAb 6C5.
mAb 6C5 may exert its effects on cell invasion through inhibiting constitutive FGFR4
autophosphorylation, and leaving more FGFR4 available for heterotypic interactions with
FGFR1 or ligand-induced homotypic FGFR4 signaling. Treatment with mAb 10C5 resulted
in inhibition of FGF2-induced FGFR4 R388 phosphorylation and FGFR1 downregulation
(Figure 7).
[0251] This Example establishes that certain anti-FGFR4 antibodies block invasion of
cancer cells that express both MT1-MMP and FGFR4 R388.
[0252] All publications, patents and patent applications cited in this specification are
herein incorporated by reference as if each individual publication or patent application were
specifically and individually indicated to be incorporated by reference. Although the
foregoing invention has been described in some detail by way of illustration and example for
purposes of clarity of understanding, it will be readily apparent to those of ordinary skill in
the art in light of the teachings of this invention that certain changes and modifications may
be made thereto without departing from the spirit or scope of the appended claims.
We Claim:
1. An isolated monoclonal antibody, or fragment thereof, or a polypeptide that
comprises a fragment of the antibody;
wherein the antibody, fragment, or polypeptide binds an extracellular epitope of a
fibroblast growth factor receptor-4 (FGFR4) that is expressed by mammalian cells, preferably an
FGFR4 that comprises the amino acid sequence of SEQ ID NO: 1 or 2; and
wherein the antibody, fragment, or polypeptide exhibits at least one activity selected from
the group consisting of:
inhibiting cancer cell invasion in a mammal;
inhibiting fibroblast growth factor 2 (FGF2)-induced phosphorylation of FGFR4
R388 in mammalian cells that express FGFR4 R388;
enhancing FGF2-induced degradation of fibroblast growth factor receptor-1
(FGFR1) in mammalian cells that co-express FGFR4 R388 and FGFR1; and
inhibiting complex formation between FGFR4 and membrane type-1 metalloproteinase
(MT1-MMP) in mammalian cells that co-express FGFR4 R388 and MT1-MMP.
2. The isolated monoclonal antibody, antibody fragment, or polypeptide of claim 1,
wherein the antibody, fragment, or polypeptide binds at least one FGFR4 peptide that consists of
an amino acid sequence selected from the group consisting of SEQ ID NOS: 5-9, and that
preferably binds an FGFR4 peptide consisting of SEQ ID NO: 7, and more preferably further
binds an FGFR4 extracellular epitope that comprises amino acid residues 79-81 of SEQ ID NO:
1 or 2.
3. The isolated monoclonal antibody, antibody fragment, or polypeptide of claim 1
or 2, wherein the antibody is monoclonal antibody F90-10C5 (DSM ACC2967).
4. The isolated antibody fragment or polypeptide of any one of claims 1-3,
comprising a member selected from the group consisting of an ScFv, Fv, Fab', Fab, diabody, and
F(ab')2 antigen-binding fragment of an antibody.
5. The isolated monoclonal antibody, antibody fragment, or polypeptide of any one
of claims 1-4, wherein the antibody, antibody fragment, or polypeptide comprises all
complementarity determine regions (CDR) of monoclonal antibody F90-10C5 (DSM ACC2967),
or comprises the variable regions of monoclonal antibody F90-10C5 (DSM ACC2967).
6. An isolated monoclonal antibody, antibody fragment, or polypeptide that
comprises a fragment of the antibody,
wherein the antibody, fragment, or polypeptide comprises all complementarity
determining regions (CDR) of monoclonal antibody F85-6C5 (DSM ACC2966) or monoclonal
antibody F90-3B6 (DSM ACC2965), or comprises the variable regions of monoclonal antibody
F85-6C5 or F90-3B6, and
wherein the antibody, antibody fragment, or polypeptide binds an extracellular epitope of
FGFR4 that is expressed by mammalian cells.
7. The isolated monoclonal antibody, antibody fragment, or polypeptide of any one
of claims 1-6 that is a humanized antibody, a human antibody, a chimeric antibody, or comprises
a fragment of the human, humanized, or chimeric antibody that binds an extracellular epitope of
FGFR4 that is expressed by mammalian cells.
8. The isolated monoclonal antibody, antibody fragment, or polypeptide of any one
of claims 1-7, further comprising an anti-neoplastic or cytotoxic agent, such as a radionucleotide,
conjugated or attached thereto.
9. A composition comprising the isolated monoclonal antibody, antibody fragment,
or polypeptide of any one of claims 1-8 ("the first monoclonal antibody or fragment thereof)
and a physiologically acceptable carrier.
10. The composition of claim 9, further comprising a second monoclonal antibody or
fragment thereof, or a polypeptide that comprises a fragment thereof ("second monoclonal
antibody or fragment thereof), preferably human or humanized, wherein the second monoclonal
antibody or fragment thereof binds a second epitope of FGFR4 that is different than the epitope
recognized by the first monoclonal antibody or fragment thereof.
11. The composition of claim 9 or 10, further comprising a membrane type-1
metalloproteinase (MT1-MMP) inhibitor.
12. An isolated cell that produces the antibody, antibody fragment, or polypeptide of
any one of claims 1-8, such as a hybridoma cell or a cell transformed or transfected with a
polynucleotide that encodes the antibody, antibody fragment, or polypeptide.
13. Antibody, antibody fragment, or polypeptide of any one of claims 1-8, or the
composition of any one of claims 9-11, for use in treatment a mammalian subject diagnosed with
or treated for cancer or for use in modulating, inhibiting invasion, ingrowth, or metastasis of
cancer cells (such as cells of breast cancer, bladder cancer, melanoma, prostate cancer,
mesothelioma, lung cancer, testicular cancer, thyroid cancer, squamous cell carcinoma,
glioblastoma, neuroblastoma, uterine cancer, colorectal cancer, and pancreatic cancer) in a
mammalian subject, preferably a human subject, wherein the cancer preferably is a cancer
determined to have at least one FGFR4 allele that encodes FGFR4 R388.
14. The antibody, antibody fragment, or polypeptide of any one of claims 1-8, or the
composition of any one of claims 9-11 for use according to claim 13, in combination with one or
more of the following, in treatment of cancer, in modulating or inhibiting invasion, ingrowth, or
metastasis of the cancer:
a second monoclonal antibody or fragment thereof, preferably human or
humanized, wherein the second monoclonal antibody or fragment thereof binds a second
epitope of FGFR4 that is different than the epitope recognized by the first monoclonal
antibody or fragment thereof, wherein the second anti-FGFR4 antibody or fragment
preferably inhibits ligand-independent FGFR4 phosphorylation;
a composition comprising a membrane type-1 metalloproteinase (MT1-MMP)
inhibitor; and a standard of care anti-cancer therapy.
15. An isolated polynucleotide that comprises a nucleotide sequence that encodes at
least one amino acid sequence selected from the group consisting of an antibody heavy chain
variable region (VH) and an antibody light chain variable region (VL), wherein the VH and the VL
comprise complementarity determining regions (CDR) identical to monoclonal antibody F90-
10C5 (DSM ACC2967) CDRs.
16. A composition comprising an adjuvant and an isolated antigenic peptide
consisting of 5-25 amino acids of the amino acid sequence encoding FGFR4, wherein the peptide
comprises the amino acid sequence set forth in any one of SEQ ID NOs: 5-9 or a fragment
thereof.
17. An isolated polynucleotide encoding the antigenic peptide of the composition of
claim 16.
18. A vector comprising the polynucleotide of claim 17.

The invention provides an isolated antibody or antibody fragment thereof that binds an extracellular epitope of a fibroblast
growth factor receptor-4 (FGFR4) that is expressed by mammalian cells and inhibits cancer cell invasion. Optionally, the
antibody or fragment thereof binds an epitope of FGFR4 that is bound by monoclonal antibody F90-10C5, or comprises complementarity
determining regions identical to those of monoclonal antibody F90-10C5. Also provided are methods of using the antibody
or fragment thereof to modulate invasion, ingrowth, or metastasis of cancer cells and treat cancer in a subject. The invention
additionally provides a method of identifying an antibody or antibody fragment that inhibits invasiveness.