AU2003259498A1 - Inducible focal adhesion kinase cell assay - Google Patents

Description

WO 2004/027018 PCT/IB2003/003968 -1 INDUCIBLE FOCAL ADHESION KINASE CELL ASSAY Background of the Invention This invention relates to methods and compositions for inducing the expression of the focal adhesion kinase (FAK) gene, which encodes a signaling protein involved in growth 5 factor response and cell migration and is also implicated in disease. The invention is also directed to the identification of FAK inhibitors. FAK is a cytoplasmic, non-receptor tyrosine kinase. FAK transduces signaling from a diverse group of stimuli (e.g. integrins, cytokines, chemokines, and growth factors) to control a variety of cellular pathways and processes including cell proliferation, migration, 10 morphology, and cell survival. In addition to being expressed in most tissue types, FAK is found at elevated levels in most human cancers, particularly in highly invasive metastases. It has been shown that expression of the dominant-negative FAK-related nonkinase (FRNK) in human tumor cells results in rounded morphology of the cells, the irreversible loss of focal plaques, and subsequent cell death. In addition, the controlled expression of FRNK results in 15 decreased tyrosine phosphorylation of FAK, suggesting that inhibition of FAK phosphorylation may yield a therapeutic index for the treatment of human cancers. Although the exact mechanisms leading to FAK activation are not well defined, it is believed that FAK enzyme activity resulting in phosphorylation at Y397 (tyrosine residue at position 397) is the critical step in integrin signal transduction (Guan, JL, Int. J. 20 Biochem.Cell.Biol. 29: 1085-96, 1997). The transmembrane integrin receptors are important for linking the extracellular matrix (ECM) proteins with the cellular actin cytoskeleton and the nucleus to regulate cell morphology, tissue architecture, and attachment-induced gene expression. It is believed that colocalization of integrin receptors and FAK to sites of focal adhesion leads to FAK phosphorylation at residue Y397, creating a SH2 docking site for Src 25 family tyrosine kinases. Src binding to phosphotyrosine FAKY397 leads to preferential phosphorylation of FAK at various downstream tyrosine residues including Y576, Y577, Y861, and Y925. Phosphorylation of FAK tyrosine residues (Y576/Y577) leads to increased FAK kinase activity and signal transduction to regulate cytoskeletal reorganization, cell proliferation, cell survival, and cell migration. 30 Due to the multiplicity of kinases and substrates involved in the integrin signaling cascade, it is desirable to design assays specific for a particular kinase. An objective of the invention is to design and develop a FAK drug discovery pathway that tracks the biochemical mechanism of FAK. A number of exogenous stimuli can lead to FAK phosphorylation such as (1) integrin binding to ECM ligands (e.g. Integrin P1 to Fibronectin); (2) cytokine or chemokine 35 stimulation (e.g. Endothelinl/2, Bombesin, or PMA); (3) growth factor stimulation of tyrosine kinase receptors (e.g. PDGFBB); and (4) integrin antibody cross-linking (e.g. Anti-P1).

WO 2004/027018 PCT/IB2003/003968 -2 Conversely, the most feasible exogenous control leading to FAK inactivation is the detachment of cell-to-cell and cell-to-ECM contact (e.g. cell suspension). SUMMARY OF THE INVENTION The invention relates to a method for identifying cell-active inhibitors of focal adhesion 5 kinase (FAK) comprising: (a) adding an inducing agent to mammalian cells to induce the expression of a gene encoding FAK, wherein said mammalian cells are stably transfected with said gene, and said gene is expressed in the presence of said inducing agent; (b) adding a test compound; 10 (c) capturing the expressed FAK using a FAK capture agent; and (d) detecting phosphorylation of said FAK. An embodiment of the invention provides a method for measuring the cytotoxicity of a test compound comprising, stably transfecting mammalian cells with a gene encoding FAK, wherein said gene is expressed in the presence of an inducing agent; adding an inducing 15 agent to induce the expression of said gene encoding FAK; adding a test compound; adding a cytotoxicity indicator to said cells; and, detecting the cytoxicity of the test compound. In certain embodiments, the cytoxicity of the test compound is determined by the colorimetric conversion of the cytotoxicity indicator, wherein the amount of converted cytotoxicity indicator is proportional to the number of living cells. 20 An embodiment of the present invention provides a mammalian cell stably transfected with a recombinant nucleic acid molecule, wherein said recombinant nucleic acid molecule is selected from the group consisting of SEQ ID No: 5, SEQ ID No: 6, SEQ ID No: 7 and SEQ ID No: 8, and wherein expression of said sequences requires the presence of an inducing agent. An embodiment of the invention provides a mammalian cell stably transfected with a 25 recombinant nucleic acid molecule that encodes a protein comprising a sequence selected from the group consisting of SEQ ID NOS: 1, 2, 3 and 4, and wherein expression of said protein requires the presence of an inducing agent. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 shows a schematic representation of the detection of phosphorylated FAK 30 using a horseradish peroxidase-conjugated phosphotyrosine antibody (pY 5 4 HRP). Figure 2 shows a schematic representation of the detection of phosphorylated FAK using an unconjugated phosphotyrosine antibody (pY54) followed by a secondary mouse horseradish peroxidase antibody. Figure 3 shows a schematic representation of a FAK inducible cell-based assay of 35 the invention.

WO 2004/027018 PCT/IB2003/003968 -3 DETAILED DESCRIPTION The invention is directed to an inducible cell-based assay for FAK. The cell-based assay exploits the biology of FAK and an inducible gene expression system to exogenously control FAK expression and FAK phosphorylation at the tyrosine residue at position 397 5 (Y397). By using a FAK Y39 7 phosphorylation-specific cell-based assay, rather than a general phosphotyrosine system, the identification of false-positive inhibitors is avoided. The cell based assay of the present invention is flexible and can measure FAK phosphorylation at Y397, total FAK phosphorylation, identify mutant FAK proteins and measure a combination of protein and phosphotyrosine. 10 The inducible FAK cell-based assay of the present invention is advantageous in that it provides tight control over ectopic-basal level expression of FAK and rapid de-repression of FAK gene expression via an exogenous stimulant. The cell-based assay is flexible such that the final read-out is mechanistically relevant to FAK biology as measured for phosphotyrosine

FAK"

'

", total FAK phosphotyrosine profile, FAK or mutant proteins, or some combination of 15 protein and phosphotyrosine. In addition, the present invention has been succesfully used to identify a number of FAK inhibitors. As used herein, the term "tight control" refers to the controlled expression of the FAK gene that occurs in the presence of an exogenous stimulant. In other words, the invention provides an inducible gene expression system for FAK, where the expression of FAK is induced in the presence of an appropriate inducible agent. The 20 present invention provides a method for the inducible expression of FAK, wherein the regulated expression of the FAK gene does not adversely affect cell viability. An embodiment of the invention is directed to a cell-based assay for the screening of FAK inhibitors. The cell-based assay exploits the biology of FAK and an inducible gene expression system to exogenously control FAK expression and FAK phosphorylation at the 25 tyrosine residue at position 397 (Y397). The cell-based assay is mechanistically relevant to FAK biology and measures changes in FAK phosphorylation. By using a FAK Y 1

"

7 phosphorylation-specific cell-based assay, rather than a general phosphotyrosine system, the identification of false-positive inhibitors is avoided. The cell-based assay of the present invention is flexible and can measure FAK phosphorylation, total FAK phosphorylation, 30 identify mutant FAK proteins and measure a combination of protein and phosphotyrosine. An embodiment of the invention provides a method for identifying cell-active inhibitors of FAK comprising, stably transfecting mammalian cells with a gene encoding FAK, wherein said gene is expressed in the presence of an inducing agent; adding an inducing agent to induce the expression of said gene encoding FAK; adding a test compound; capturing the 35 expressed FAK using a FAK capture agent; exposing the captured FAK to an anti-phospho tyrosine antibody; and, detecting the phosphorylation of said FAK. In some embodiments, the extent of phosphorylation of the FAK is determined by the binding of the anti-phospho- WO 2004/027018 PCT/IB2003/003968 -4 tyrosine antibody to the captured FAK, wherein the amount of anti-phospho-tyrosine antibody binding to the captured FAK is proportional to the amount of phosphorylation of said FAK. In certain embodiments of the invention, the method for identifying cell-active inhibitors of FAK comprises an optional step of coating the mammalian cells on a first solid 5 phase. The first solid phase is preferably a well of a first microtiter plate. In other embodiments of the invention, the cells coated on the first solid phase are lysed with a lysis buffer, prior to the capture of the expressed FAK. The lysis buffer optionally comprises a solubilizing detergent. In certain embodiments, the FAK capture agent is coated on a second solid phase, which is preferably a well of a second microtiter plate. 10 In certain embodiments, the test compound inhibits the phosphorylation of FAK at Y397. An embodiment of the invention provides a method for measuring the cytotoxicity of a test compound comprising, stably transfecting mammalian cells with a gene encoding FAK, wherein said gene is expressed in the presence of an inducing agent; adding an inducing 15 agent to induce the expression of said gene encoding FAK; adding a test compound; adding a cytotoxicity indicator to said cells; and, detecting the cytoxicity of the test compound. In certain embodiments, the cytoxicity of the test compound is determined by the colorimetric conversion of the cytotoxicity indicator, wherein the amount of converted cytotoxicity indicator is proportional to the number of living cells. 20 In certain embodiments of the invention, the method for measuring the cytotoxicity of a test compound comprises an optional step of coating the mammalian cells on a solid phase. The solid phase is preferably a well of a first microtiter plate. An embodiment of the invention provides a method for identifying cell-active inhibitors of focal adhesion kinase (FAK) comprising, coating a first solid phase with a homogeneous 25 population of mammalian cells so that the cells adhere to the first solid phase, wherein said cells are stably transfected with a gene encoding FAK, and wherein said gene is expressed in the presence of an inducing agent; adding an inducing agent to induce the expression of said gene encoding FAK; adding a test compound; solubilizing the adhering cells to release the cell lysate; coating a second solid phase with a FAK capture agent so that the FAK capture 30 agent adheres to the second solid phase; exposing the cell lysate to the adhered FAK capture agent so that the FAK capture agent captures FAK; exposing the captured FAK to an anti-phospho-tyrosine antibody; and, measuring binding of the anti-phospho-tyrosine antibody to the captured FAK, wherein the amount of anti-phospho-tyrosine antibody binding to the captured FAK is proportional to the amount of phosphorylation of said FAK. 35 Another embodiment of the present invention provides a mammalian cell stably transfected with a recombinant nucleic acid molecule, wherein said recombinant nucleic acid molecule is selected from the group consisting of SEQ ID No: 5, SEQ ID No: 6, SEQ ID No: 7 WO 2004/027018 PCT/IB2003/003968 -5 and SEQ ID No: 8, and wherein expression of said sequences requires the presence of an inducing agent. An embodiment of the invention provides a mammalian cell stably transfected with a recombinant nucleic acid molecule, that encodes a protein comprising a sequence selected 5 from the group consisting of SEQ ID NOS: 1, 2, 3 and 4, and wherein expression of said protein requires the presence of an inducing agent. FAK is also known as the Protein-Tyrosine Kinase 2, PTK2. Any active FAK variant can be employed in the above assay. Inactive mutants can also be used in the assay for various control purposes. Additional variants of FAK that can be employed in the above 10 described assay include, wild type (WT) human FAK at 153012 with 1052 amino acids (SEQ ID NO:1); splice variants of FAK such as described in Andre, E. & Becker-Andre, M., Expression of an N-terminally truncated form of human focal adhesion kinase in brain. Biochem. Biophys. Res. Commun. 190: 140-147, 1993 (describing an 879 amino acid variant at AAA35819, a 554 amino acid variant at PC1226 and a 431 amino acid variant at PC1227); 15 the 570 catalytic domain of FAK at XP_050337; mouse FAK at NP_032008.1 with 1023/1052 amino acid identity (97%) to human FAK; mouse FAK carboxyl truncated variant at AAH30180.1 with 878/904 amino acid identity (97%) to amino acids 1-903 of human FAK; rat FAK at NP_037213.1 with 1020/1055 amino acid identity (96%) human FAK; FAK variant at JC5494 with 1017/1055 amino acid identity (96%) to human FAK; chicken FAK at Q00944 20 with 988/1054 amino acid identity (93%) to human FAK; chicken FAK variant at A45388 with 965/1029 amino acid identity (93%) to human FAK; synthetic FAK mutants, including the FAK Y397F (SEQ ID NO:2), K454R (SEQ ID NO:3), FRNK (an amino terminal truncant having FAK residues 694-1052 preceded by an initiator MET) (SEQ ID NO:4), and various phosphorylation mimics including FAK Y397D, Y397E, Y577D, Y577E, Y861D, Y861E, 25 Y925D, Y925E and combinations thereof; CD2-FAK fusion (a constitutively active FAK fusion with CD2) described by in Chan P, et al., J Biol. Chem. (1994); 269 (32): 20567-74. As used herein, the term "inducing agent" is an agent, compound, or chemical that produces a signal to noise ratio of at least 6-fold. Examples of inducing agents include but are not limited to Mifepristone (Ru486) and other antiprogestins such as Org31806 and 30 0rg31376. See O'Malley et. al., Cell, 69, 703-713 (1992). As used herein, the term capture agent is an agent, compound or chemical that is capable of capturing any form of focal adhesion kinase, including FAK tagged with histidine residues, streptavidin or other comparable affinity tags. The capture agent includes, but is not limited to, the phosphotyrosine FAK Y397 specific antibody, a general phosphotyrosine antibody, anti-FAK 35 antibody, anti-histidine tag antibody and molecules containing biotin which would facilitate the capture of streptavidin-modified FAK. In certain embodiments of the invention, the capture WO 2004/027018 PCT/IB2003/003968 -6 agent includes a combination of one or more antibodies, including but not limited to the combination of goat anti-rabbit antibody and the phosphotyrosine FAKY3 9 7 specific antibody. As used herein, the term "anti-phospho-tyrosine antibody" includes but is not limited to phosphotyrosine FAK" 397 specific antibody as well as general phosphotyrosine antibodies, 5 where the latter are capable of recognizing any phosphorylated tyrosine residue, including but not limited to the tyrosine residue at position 397 of FAK. As used herein, a cytoxicity indicator is an agent, chemical or compound that is used to assess cell viability. Examples of cytoxicity indicators include, but are not limited to tetrazolium salts (e.g. MTT, XTT, WST-1) that are especially useful for this type of analysis. 10 MTT is a yellow tetrazolium salt that is cleaved to purple formazan crystals by metabolic active cells. The solubilized formazan product may be spectrophotometrically quantified using an ELISA reader or other spectrophotometric device. The number of living directly correlates to the amount of purple formazan crystals formed, as monitored by the absorbance. 15 An embodiment of the present invention is carried out using a inducible FAK gene expression system having the ability to provide consistent and tight regulation of gene expression. In an embodiment of the invention, an inducible FAK gene expression system is provided, which comprises a system for the regulated expression of a transgene, i.e., FAK. FAK expression is "off" in the absence of an inducing agent, but "on" in its presence. The 20 system consists of two genes: one which codes for a regulatory protein, and the other which codes for the inducible transgene of interest. Expression of the regulatory protein is driven by a constitutive promoter. The inducible FAK transgene has a promoter that consists of a minimal promoter linked to multiple copies of a binding site that is capable of binding the regulatory protein. 25 In an embodiment of the invention, the regulatory protein is a transcription factor that consists of the yeast GAL4 DNA binding domain, a truncated human progesterone receptor ligand binding domain, and the human p65 activation domain from NF-KB to facilitate tight regulation of the transgene over basal expression. The plasmid encoding the regulatory protein contains the GAL4 promoter which establishes a positive feedback loop to facilitate 30 rapid response upon ligand treatment. The exogenous control of regulatory protein expression is achieved using a small molecule ligand. Tight regulatioh is achieved by species selective binding of the regulatory protein to the promoter of the transgene, and through ligand-dependent conformational activation of the regulatory protein. Thus, in certain embodiments, the invention has the following advantages: 35 = Induction of FAK in cells using an inducible system, which allows for tight repression of gene expression when not induced, resulting in viable cell clones.

WO 2004/027018 PCT/IB2003/003968 -7 * Detection of phosphorylated FAK may be used to identify inhibitors of FAK kinase activity. Some of the methods of detection of phosphorylated FAK include Polyacrylamide gel electrophoresis-based assays and ELISA-based assays for identification of FAK inhibitors. Other suitable detection systems may also be used. 5 * The assay allows for detection of total FAK protein, total phosphorylated FAK protein, or FAK protein phosphorylated at a given tyrosine (e.g. tyrosine 397). * The cell-based system may be used for in vivo screening of FAK inhibitors, since the transfected cells are tumorigenic and FAK can be induced in vivo by feeding animals Mifepristone. The system's utility in vivo has been demonstrated. 10 COMPARATIVE EXAMPLES Comparative Example 1 Attempts were made to generate stable ectopically expressed FAK or FAK mutant proteins in a variety of cell backgrounds in hopes of improving assay signal and noise. These efforts, however, also proved unsuccessful since most of these cells demonstrated sensitivity 15 to changes in FAK protein levels and never formed viable clones with which to develop a cell based assay. Cells expressing low to moderate levels of endogenous FAK such as NIH3T3 mouse fibroblast or A2058 human metastatic melanoma cells tolerated no more than two-fold expression of FAK over endogenous levels. Consequently, these clones proved insufficient for development of the assay due to similar reasons described above: poor stimulation and 20 reproducibility, high background noise, and not conducive to drug discovery. Therefore, non inducible expression of FAK in cells containing native FAK was not found to be suited for studying the induction of FAK expression. Comparative Example 2 In order to develop a FAK cell-based assay that is both conducive to Medium and 25 High Throughput Screening (ELISA system) and tracks the biochemical mechanism of FAK, a number of approaches were taken to exploit FAK biology. For example, attempts were made to mimic adherent cell stimulation of FAK by plating cells on Extracellular matrix proteins (ECM) proteins such as fibronectin, or collagen, or other ECMs such as matrigel, but this approach proved inefficient, resulted in poor stimulation of FAK, and was entirely dependent 30 on cell type. Furthermore, ECM-matrigel-induced attachment of cells resulted in no stimulation of FAK phosphorylation as measured by pY 5 4 HRP (Horseradish Peroxidase (HRP) conjugated phosphotyrosine antibody) detection, and control detection using secondary Anti MouseHRP ( 2 0MHRP) alone resulted in nonspecific increases in assay signal in wells incubated with lysate derived from cells stimulated on matrigel. The signal and noise ratio (S/N) when 35 using a combination of pY54 HR P and 2 MHRP ranged from 1.0 to 1.7, while a combination of pY54 (unconjugated phosphotyrosine antibody) and 2 oMHRP yielded a SIN ratio of 1.0 to 2.4. This suggested, and it was confirmed, that the nonspecific increase in assay signal was due WO 2004/027018 PCT/IB2003/003968 -8 to cross-reactivity to some matrigel contaminant(s) in the cell lysate. Hence, the above approach to develop an appropriate FAK cell-based assay was not viable. Results also demonstrated a modest and suboptimal two- to three-fold signal over noise stimulation of FAK phosphorylation upon cell attachment to either ECM matrigel or 5 fibronectin. Further optimization of these assay conditions either by varying cell number, cell ECM time for attachment, type of detection and capture antibodies, and combinatorial stimulation using Endothelin-1 and Phorbol 12-Myristate 13-Acetate (PMA) indicated that these non-inducible methods of FAK stimulation were insufficient, irreproducible, and not feasible for assay development. 10 Comparative Example 3 Since integrin clustering leads to FAK phosphorylation, the feasibility of antibody crosslinking of integrin receptors using antibodies specific to p1 integrin was also examined. However these efforts also proved to be unsuccessful because indirect methods of FAK stimulation lead to high variability and irreproduciblity. For example, integrin P31 cross-linking 15 proved to be too cumbersome for even medium throughput screening since optimal stimulation required temperature sensitive steps that were time consuming and difficult to manage. Moreover, maintaining cells on polystyrene plates to lower background noise resulted in decrease cell viability due to poor ECM contact, thereby making these non inducible expression systems infeasible for assay development. 20 Comparative Example 4 To advance the development of the FAK cell-based assay, an attempt was made to generate stable clones expressing either wild-type or mutant FAK proteins in a variety of cell backgrounds. Clones were generated that harbored FAK cDNA transcripts that encode either a constitutively active protein (CD2*FAK), a membrane bound tyrosine-dead protein 25 (CD2FAKY 3 97F), or cytoplasmic FAK mutants that lack key downstream tyrosine residues (e.g., FAK Y 61F , FAKY 925 F, and FAKYM61FY925F). These biological tools were strategically designed based on the knowledge of FAK biology in order to address the issues described above and to necessitate the development of a robust and reproducible cell-based assay conducive to drug discovery. 30 Attempts to generate stable clones overexpressing FAKWT were unsuccessful, even though viable control clones expressing either vector construct or the LacZ protein were obtained. Transfection of FAK mutant constructs designed either to shut down downstream phosphorylation of FAK while preserving FAK phosphorylation at Y397 and kinase activity

(FAK

Y 6 1 , FAKY9 25 0 F, and FAKY"'FY 9 2 5F), or to bypass the requirement for integrin receptors 35 (CD2.FAKwT and CD2*FAKYS 9 7F), also proved unsuccessful. In cells that expressed low to moderate levels of endogenous FAK, such as the human metastatic melanoma A2058, viable WO 2004/027018 PCT/IB2003/003968 -9 clones were identified. However, A2058*FAK w clones exhibit only an approximate two-fold increase in FAK levels which proved insufficient for development of the FAK cell based assay. In addition to the work on exogenous stimulation, control over basal FAK phosphorylation and attempts to ectopically express FAK clones, FAKwT and tyrosine-dead 5 FAK" Y397F cDNAs were subjected to tetracycline control using the Tet-On/Tet-Off inducible systems. However, no FAK wT or FAKY"" 97 F transfectants were detected. Since viable clones could not be identified, the level of "leakiness", i.e., inadequate regulation of basal expression, in the FAK w or FAK Y397F Tet-transfectants could not be determined. The Tet-OnTM/Tet-Off M systems demonstrated variable control over target protein expression, and therefore were not 10 suited for the regulation of lethal genes like FAK. WORKING EXAMPLES The wild-type FAK w (SEQ ID NO: 5), tyrosine-dead FAK ¥ 9 7F (SEQ ID NO: 6), kinase-dead FAK K454 R (SEQ ID NO: 7) and the dominate-negative FAK-related nonkinase (FRNK) (SEQ ID NO: 8) were cloned into the GeneSwitch T M pGeneV5/His A-vector backbone 15 as either BamHI-Apal or KpnI-Apal inserts. As used herein, the term "gene encoding FAK" includes without limitation, SED ID NOS: 1-4. These constructs are based on the published sequence for human FAK at GenBank accession number L13616, see Whitney,G.S. et. al., DNA Cell Biol. 12 (9), 823-830 (1993), and each construct was sequence confirmed and aligned against the published sequence. The DNA constructs were engineered to include 20 specific 5' and 3' restriction sites that are unique and distinct over the prior art. In addition, the 3' ends of certain constructs were also engineered to include tagged epitopes such as V5 and histidine tags. Several cell lines were selected for transfection including the NIH Swiss Mouse fibroblast NIH3T3 (ATCC Accession No. CCL-92), the human epidermoid carcinoma A431 25 (ATCC Accession No. CRL1555), and the human glioblastoma astrocytoma U87MG (ATCC Accession No. HTB-14). Each cell line exhibits unique characteristics that makes its application favorable for overexpression of FAK. For example, NIH3T3 cells offer species specific overexpression of FAK and ease of transfectability, A431 cells express moderate levels of endogenous FAK and provide a tumor background more representative of a FAK 30 native environment, and U87MG cells lack the putative FAK negative regulator PTEN (a tumor suppressor phosphatase). Using Stratagene's GeneJammer T M Transfection reagent, the pSwitch regulatory vector was co-transfected with either the pGenevectorV5His, or the pGeneFAKWTV5His, or the pGeneFAK mutants (pGeneFAKK 4 5 4 RV5His, and pGeneFAKY39 7 FV5His) into A431 cells. Similarly, NIH3T3 and U87MG cells were transformed 35 to co-express the pSwitch regulatory protein and the specific pGene construct of interest. Hygromycin and Zeocin resistant clones were selected and expanded in culture for screening via western and RT-PCR analysis. Several inducible clones including, A431:FAKWTV5His, WO 2004/027018 PCT/IB2003/003968 -10 A431:FAKY S 7 FV5His, A431:FAKK 4 RV5His, A431:Vector, NIH3T3:FAKWTV5His, NIH3T3:FAKY 397 F V5His, NIH3T3:Vector, U87MG:FAKWTV5His, and U87MG:FAKY397FV5His were successfully isolated, selected, confirmed and validated. The FAK inducible assay of this embodiment of the invention measured FAK 5 phosphorylation at Y397. A431:FAKW V5His cells (about 1.0 x 104 to 1.0 x 107 cells) are seeded on U-bottom 96-well plates and allowed to attach at 37 0 C, 5% CO 2 for 4 to 6 hoursprior to overnight incubation in the presence of 0.1 nM Mifepristone. Subsequently, A431:FAKWTV5His induced cells are either left untreated or treated with inhibitory compounds for 30 minutes at 37 0 C, 5% CO 2 and lysed in RIPA lysis buffer (50 mM Tris-HCI, pH7.4, 1% 10 NP-40, 0.25% Na-deoxycholate, 150 mM NaCI, 1 mM EDTA, 1 mM Na 3

VO

4 , 1 mM NaF, and one Complete T M EDTA-free protease inhibitor pellet per 50 ml solution). Approximately 45 pg of total protein (100 pl) is then transferred to goat Anti-rabbit plates coated with 0.35 pg/well of Anti-FAK phosphospecific Y397 antibody for the capture and subsequent detection of phosphorylated FAK proteins (See Figure 3). 15 NIH3T3:FAKWTV5His clones and NIH3T3:FAKY 3 9 7FV5His clones were seeded to approximately 80% confluency in T25 flasks and were either left untreated or stimulated with 0.1 nM Mifepristone for approximately 16 hoursat 37"C, 5%CO 2 . Cell lysates were prepared and immunodepleted of total FAK protein using the Anti-FAK (UBI T M , Lake Placid, N.Y.) monoclonal antibody. FAK immunocomplexes were then subject to SDS-Polyacrylamide gel 20 electrophoresis and immunoblotted with Anti-FAK (A17) polyclonal antibody. Table 1 shows densitometric quantitation measured in arbitrary "band light" units of the autoradiographic film and fold changes in FAK protein levels as compared with unstimulated cells. Table 1: Mifepristone induced FAK protein levels as measured by Densimetric Quantitation 25 Clone FAKwt I FAKwt 3 FAKwt 4 FAKwt 5 number Mifepristone + - + - + + Band Light 2.00E+06 1.00E+06 8.00E+05 5.00E+05 2.00E+06 5.00E+05 3.00E+06 5.00E+05 Units Fold 2 1 2 1 4 1 6 1 Stimulation It was observed that FAKWT inducible clones responded very well to mifepristone stimulation. The above conditions translated easily to a 96-well ELISA format. Table 2 shows 30 NIH3T3 clones induced to express FAKWT or FAK" 397F proteins. FAKWT and FAKY 97 F proteins WO 2004/027018 PCT/IB2003/003968 -11 were captured either to measure total FAK protein or phosphotyrosine FAK Y397 induced by mifepristone. Table 2: Mifepristone induced FAK protein levels as measured by optical density 450 measurements NIH3T3FAK Signal (+) NIH3T3FAKY397F Signal (+) NIH3T3vector Signal (+) wt /Noise(-) /Noise(-) /Noise(-) Ratio Ratio Ratio Mifepristone + - + - + . PhosphoFAK 1.69 0.23 7 0.1 0.12 1 0.08 0.04 2 Y397 Capture FAK protein 1.94 0.33 6 2.56 0.46 6 0.21 0.16 1 Capture 5 It was observed that NIH3T3 clones yield a robust phosphotyrosine FAKY 3 97 and FAK protein signal-to-noise ratio of -7 and -6, respectively. Both the NIH3T3 tyrosine-dead clone and the vector NIH3T3 transfectants confirm assay specificity for phosphoFAKY397. Table 3 shows A431:FAK clones assayed under optimized conditions. 10 Table 3: Mifepristone induced FAK protein levels as measured by optical density 450 measurements A431:FAKwt Signal (+) A31:FAK Signal (+) A31:vector Signal (+) /Noise(-) Y397F /Noise(-) /Noise(-) Ratio Ratio Ratio Mifeprist + - + - + one Phospho 2.09 0.13 17 0.13 0.08 1.4 0.12 0.11 1.0 FAKY397 Capture FAK 2.22 0.21 11 2.31 0.22 11 0.19 0.17 1.0 protein Capture Under optimal conditions, the A431:FAK clones yielded a robust phosphotyrosine

FAK

Y 3 9 7 and FAK protein signal-to-noise ratio of ~17 and -11, respectively. Both the A431 15 tyrosine-dead clone and the vector A431 transfectants confirmed assay specificity for phosphoFAKY397. Induction of phosphotyrosine Y397 and FAK proteins were controllable by WO 2004/027018 PCT/IB2003/003968 -12 varying assay conditions such as mifepristone time of incubation, mifepristone concentration, cell density, and antibody concentrations. The mechanism of FAK activation in A431 and other cell backgrounds was studied. A431:FAK wT cells were either left untreated or stimulated with mifepristone. First, a washout 5 experiment was performed to demonstrate a time-dependent decrease in FAK protein and phosphotyrosine Y397 levels (i.e. cells stimulated with mifepristone overnight (16hs) were subsequently cultured in fresh growth media in the absence of mifepristone over time). Second, A431:FAK w cells stimulated with mifepristone were detached and suspended in fresh growth media for 15 minutes followed by re-plating for 4, 24, 48, and 72 hourson tissue 10 culture treated petri dishes. Table 4, experiment A, shows densimetric quantitation measured in arbitrary "band light" units of the autoradiographic film and fold changes in FAK phosphorylation as compared to a no stimulation control. Table 4: FAK phosphorylation as measured by Densimetric Quantitation Experiment A A A431FAKwt 4hrs 24 hrs 48 hrs 72 hrs Mifepristone - + + + + + Band Light Units 5.7534 57.832 77.375 71.033 27.719 92.834 Fold Increase 1 10 13 12 5 16 over Uninduced 15 Experiment B B A431FAKwt 4hrs 24 hrs 48 hrs 72 hrs Mifepristone - + + + + + Band Light 20723 44791 89444 65946 11473 82273 Units Fold Increase 12 4 3 1 4 over Uninduced In experiment B, a modest 4-fold induction over untreated cells was observed. However, suspending stimulated cells for 15 minutes in fresh growth media followed by a 4 20 hour attachment to plastic led to a decrease in phosphoFAK Y 3 ". Moreover, allowing suspended mifepristone-induced cells to re-attach for 24 hours resulted in the re-activation of phosphotyrosine FAKY 39 7 , suggesting that the mechanism of FAK activation was intact.

WO 2004/027018 PCT/IB2003/003968 -13 Consistent with a time-dependent decrease in FAK protein levels, phosphotyrosine FAKY 397 decreased to unstimulated control levels by 72 hours. Cellular suspension leads to inactivation of FAK phosphorylation. Re-activation of FAK may be achieved by re-plating suspended cells to allow cell attachment. Table 4 5 demonstrates that the mechanism of FAK activation is intact under the regulatory system discussed above. In other words, exogenous stimulation with mifepristone leads to the activation of FAK that is mechanistically relevant. The FAK inducible cell-based assay of the invention was successfully employed in the identification of a number of FAK inhibitors, including PP1 and staurosporin. In addition, 10 the assay was used to determine the cytoxicity of particular compounds on FAK-expressing cells. In these experiments, the test compound was placed in contact with FAK-induced or uninduced control cells. Table 5 shows changes in FAK phosphorylation in OD 450 units upon treatment with increasing concentration of a FAK inhibitor. Table 6 shows densimetric quantitation measured in arbitrary "band light" units of the autoradiographic film and fold 15 changes in FAK phosphorylation in the presence of increasing concentration of the FAK inhibitor. Thus, the assay of the invention can be advantageously used in a FAK drug discovery program. The FAK inducible cell-based assay was used as described in Figure 3 to screen a FAK inhibitor. A 10 pM curve of an FAK inhibitor was performed in 1/2 log dilutions as shown 20 in Table 5. The % inhibition was determined and the inhibition concentration at 50% (ICo) was calculated. A431:FAK w T cells were seeded in T25 flask to -80% confluency and were either left untreated or stimulated overnight with 0.1 nM mifepristone. Subsequently, stimulated cells were treated with the same FAK inhibitor as in Table 5 at the same 1/2 log concentrations of a 10 pM curve for 30 minutes at 37 0 C, 5% CO2. Cell lysates were then 25 prepared and total protein concentration determined by protein assay. Equivalent amounts of total protein were subject to SDS-Polyacrylamide gel electrophoresis and western analysis using the phosphotyrosine FAK 97 specific antibody, the general phosphotyrosine pY20 antibody, and the Anti-FAK (A17) antibody. Table 5: FAK phosphorylation as measured by optical density 450 30 measurements Noise FAK Inhibitor+A431FAKwt WT 10pM 3.3pM 1.1 pM 0.37pM 0.12 pM 0 pM Mifepristone - + + + + + + PhosphoFAKY39 0.28 0.56 0.98 1.65 2.21 2.56 2.77 7 (OD450) Minus 0.000 0.28 0.70 1.37 1.93 2.28 2.49 WO 2004/027018 PCT/IB2003/003968 -14 Noise FAK Inhibitor+A431FAKwt WT 10pM 3.3 pM 1.1 pM 0.37 pM 0.12 pM 0 pM Background %Control 11 28 55 78 92 100 %Inhibition 89% 72% 45% 22% 8% 0% The calculated Inhibition Concentration at 50% for the FAK inhibitor was -0.93 uM. Table 6 shows densimetric quantitation measured in arbitrary "band light" units of the autoradiographic film and fold changes in FAK phosphorylation in the presence of increasing 5 concentration of the FAK inhibitor. These experiments further demonstrate the usefulness of the assay of the invention in a FAK drug discovery program. Table 6: FAK phosphorylation as measured by Densimetric Quantitation A: IB:AntiFAK(A17) Background FAK Inhibitor+A431FAKwt WT 10 pM 3.3 pM 1.1 pM 0.37 puM 0.12 pM 0 pM Mifepristone . + + + + + + BandLight 30.501 119.6 100.2 108.3 128.1 90.087 82.414 Units Minus 0 89 70 78 98 60 52 Background %Control 172 134 150 188 115 100 %Inhibition 0% 0% 0% 0% 0% 0% B: IB:AntiFAKpY397 Background FAK Inhibitor+A431FAKwt WT 10 pM 3.3 pM 1.1 pM 0.37 pM 0.12 pM 0 pM Mifepristone - + + + + + + Band Light 0.34513 3.9238 10.44 20.21 45.27 64.56 68.022 Units Minus 0 4 10 20 45 64 68 Background %Control 5 15 29 66 95 100 %Inhibition 95% 85% 71% 34% 5% 0% WO 2004/027018 PCT/IB2003/003968 -15 C: IB:AntipY20 Background FAK Inhibitor+A431FAKwt WT 10 pM 3.3 pM 1.1 pM 0.37 pM 0.12 pM 0 pM Mifepristone - + + + + + + Band Light 0.7577 5.2206 9.058 21.51 25.19 26.323 33.1 Units 32 Minus 0 4 8 21 24 26 32 Background %Control 14 26 64 75 79 100 %Inhibition 86% 74% 36% 25% 21% 0 As a control for protein loading, the amounts of FAK protein in experiment A were equivalent and incubation with the FAK inhibitor did not have any effect on FAK expression. In experiment B, consistent with the IC50 reported in Table 5, an Anti-FAKpY397 blot showed 5 an estimated IC50 value in the range of 0.37 and 1.1 uM, where the 50% concentration is closer to 1.1 pM. As seen in experiment C, similarly and consistent with FAK biology, total FAK phosphorylation as measured by an Anti-pY20 immunoblot was inhibited by the inhibitor in an estimate IC50 range of 1.1 to 3.3 pM, where the 50% inhibition concentration was closer to 3.3 pM. 10 Exogenous Control of FAK phosphorylation ECM stimulation of FAK was achieved via plating 2.0x10 5 cells/well (e.g. A2058) on commercially available fibronectin (FN) 96-well plates or on FN 96-well plates prepared in house. Cells were allowed to attach for 15, 30, 60, or 90 minutes at 37 0 C, 5%CO 2 to induce integrin engagement and subsequent FAK phosphorylation. Cells lysates were prepare in 15 RIPA lysis buffer (50 mM Tris-HCI, pH7.4, 1% NP-40, 0.25% Na-deoxycholate, 150 mM NaCI, 1 mM EDTA, 1 mM Na 3

VO

4 , 1 mM NaF, and one Complete T M EDTA-free protease inhibitor pellet per 50 ml solution) and transferred to Anti-FAK coated 96-well plates to capture total FAK protein. Phosphotyrosine FAK proteins were measured using the general phosphotyrosine antibody Py54. Similarly, cells were stimulated on commercially available 20 matrigel coated plates or on matrigel plates prepared in-house. For western analysis, cells were allowed to attach to either commercially available ECM coated flask or to ECM coated flask prepared in-house for the indicated times above. Cell lysates were prepared in RIPA lysis buffer and total and phosphotyrosine FAK proteins were either pull-down or immunodetected using Anti-FAK antibodies or Anti-phosphotyrosine antibodies. 25 Cells (e.g. A2058 human metastatic meloanomas) were seeded in growth media (DMEM10%FBS, Pen/Strep/Glu ) on 96-well plates or T25/T75 flasks and allowed to adhere WO 2004/027018 PCT/IB2003/003968 -16 to tissue culture treated plastic for approximately five hours at 370C, 5% CO 2 . Subsequently, they were starved in 0.1% FBS DMEM starvation media overnight at 37'C, 5% CO2, followed by stimulation via treatment with (up to 100 pM) Endothelin I, (up to 100 nM) Bombesin, or (up to 800 nM) PMA. Exposure to cytokines varied from 10 minutes to as long as 60 minutes at 5 37 0 C, 5% CO 2 . Total FAK proteins were either captured in ELISA format or immunoprecipitated using Anti-FAK specific antibodies. Phosphotyrosine FAK proteins were then detected with a general Anti-phopshotyrosine antibody either in ELISA or Western format. Starved Cells (e.g. A2058) were pre-chilled at 40C for 30 minutes prior to incubation 10 with varying concentrations (1:100, 1:330, 1:1000, 1:3,300, or 1:10,000 of stock concentration of 200 pg/ml) of PI3-Integrin antibody (40 C, 30 minutes). P31-integrin clustering was induced with 1:500 dilution of Goat anti-mouse secondary antibody for 30 minutes at 40C. Whole cell lysates were then prepared in lysis buffer (10 mM Tris-HCI, 5-mM NaCI, 10 mM EDTA, 2mM Na Vanadate, 1%NP40, protease inhibitors), and total FAK proteins were immunoprecipitated 15 using Anti-FAK antibody. Phosphotyrosine FAK proteins were measured using the Anti phosphotyrosine antibody Py54. Cloninq of FAKwT and FAK mutants (Noninducible and Tet-OnTM/Tet-OffTM Systems) The full-length FAK W T cDNA was cloned from a T-cell cDNA library using the primers FAK5'Bam: GGATCCATGGCAGCTGCTTACCTTGAC (SEQ ID NO.: 9) and FAK3'Bam: 20 GGATCCTCACTCACTCAGTGTGGTCTCGTCTGCCCA (SEQ ID NO.: 10) in RT-PCR and was sequence confirmed against the accession number L13616. Site-directed mutagenesis of the FAK w template was performed using the Stratagene QuikChangeTM Site-Directed Mutagenesis Kit. To generate plasmids for electroporation into Clontech pre-made HeLa or HEK293 tet-On/tet-Off cell lines, the full-length FAKwT, the tyrosine-dead mutant (FAK" 3 97 F), 25 and the kinase-dead mutant (FAKK 45 4 R) cDNAs were subcloned into the BamHI sites of the pTRE:FLAG vector. The following constructs were generated and electroporated into HeLa Tet-Off, HeLa Tet-On, or 293 Tet-On: pTRE:FLAG Vector, pTRE:FAKwTFLAG, and pTRE:FAKY39 7 1FFLAG. G418 resistant clones were selected and expanded in DMEM+100ug/ml G418 growth media. 30 Cloninq of FAKwT and FAK mutants (GeneSwitchTM System) The full-length Focal Adhesion Kinase (FAK w t ) cDNA was cloned from a T-cell cDNA library using the above primers in RT-PCR and was sequence confirmed against the accession number L13616. Site-directed mutagensis of the FAK wT template was performed using the Stratagene QuikChange T M Site-Directed Mutagenesis Kit. The full-length FAKWT, 35 the tyrosine-dead mutant (FAKY 39 "F), and the kinase-dead mutant (FAKK 45 4 R) cDNAs were subcloned into the BamHI-Apal sites of the pGeneV5/His-A plasmid. The dominate-negative FAK-related nonkinase (FRNK) cDNA was subcloned as a KpnI-Apal insert into the cassette WO 2004/027018 PCT/IB2003/003968 -17 of the pGeneV5/His A-vector. These plasmids were confirmed via DNA sequencing and restriction digestion analysis. Subsequently, these constructs were co-transfected with the pSwitch construct encoding the GeneSwitch T M Protein into NIH3T3, A431, or U87MG cells using Stratagene's GeneJammer' M Transfection reagent. Transfectants were grown and 5 selected in DMEM growth media (10%FBS, Pen/Strep/Glu) spiked with 750 pg/ml of Zeocin and 50 pg/ml of hygromycin antibiotics. Hygromycin and Zeocin resistant clones were selected and expanded in culture for screening in both Western and RT-PCR format. Screening of A431.FAK clones A431*FAK w and A431*FAKY' 1 "F clones were seeded in T25 flasks to near 80% 10 confluency and were either left untreated or treated with 0.1 nM of Mifepristone ligand overnight (~16hrs). A431 transfectants were then lysed in RIPA lysis buffer (50 mM Tris-HCI, pH7.4, 1% NP-40, 0.25% Na-deoxycholate, 150 mM NaCI, 1 mM EDTA, 1 mM Na 3

VO

4 , 1 mM NaF, and one Complete TM EDTA-free protease inhibitor pellet per 50 ml solution) and cell lysates were assayed for total protein concentration. Equivalent total protein concentrations 15 were subject to western analysis immunoblotting for FAK, or FAK mutant, proteins and phosphotyrosine FAK Y3 97 using standard western blotting techniques. Clones tested positive and exhibit at least 3X induction over endogenous FAK levels were selected for optimization and development of the FAK cell-based assay. Clones were also screened and verified via RT-PCR. Cytoplasmic mRNA transcripts 20 were isolated and purified from A431.FAK wild-type and mutant clones for cDNA synthesis using random hexamers. Polymerase chain reactions were subsequently performed on cDNA libraries derived from A431.FAK transfectants using primers specific to the N-terminus, the internal FAK sequences, and the C-terminus of FAK. Primers specific to the antibiotic transcripts Zeocin or Hygromycin and GAPDH were also used to amplify these genes as 25 internal controls for transfection as well as for the quality of cDNA libraries. Optimization of FAK inducible cell-based System A431*pGene ve ct' , A431*FAKwT, A431*FAK K464R, and A431*FAKY 397 clones were seeded in T25 flasks to near 80% confluency and were either left untreated or treated with 0.1, 10, or 100 nM of Mifepristone ligand overnight (~16hrs). A431 transfectants were then 30 lysed in RIPA lysis buffer (described above) and cell lysates were assayed for total protein concentration. Equivalent total protein concentrations were subject to western analysis immunoblotting for FAK, or FAK mutant, proteins and phosphotyrosine FAK 397 using standard western blotting techniques. To determine that optimization of Mifepristone concentration in western format translates into an ELISA system, Mifepristone concentration 35 and time of incubation were also optimized in ELISA format. A431*pGene ve ct' , A431*FAK w , and A431*FAKe 3

"

7 clones were seeded in 96-well U-bottom plates at a cell density of 1.2 x 106 cells/mi. Cells were allowed to sit at 37 0 C, 5% WO 2004/027018 PCT/IB2003/003968 -18

CO

2 for 6 to 8 hours prior to FAK induction with Mifepristone ligand. A431 clones were subsequently left untreated or treated with 0.1, 10, 100, or 110 nM of Mifepristone for -0.5, 1.0, 2.0, 4.0 or 24.0 hoursat 370C, 5% CO2. Cells were then lysed in RIPA lysis buffer and 100 pl of cell lysate (45 pg total protein) were transferred to 96 well plates. FAK or FAK 5 mutant proteins and phosphotyrosine FAK Y39 7 were subsequently captured on goat anti mouse or goat anti-rabbit plates coated with either FAK specific antibodies or phosphotyrosine FAK Y3 97 antibodies, respectively. To capture FAK or FAK mutant proteins, for example, goat anti-mouse plates were coated with 0.5pg/ml of Anti-FAK (UBI) monoclonal antibodies prior to incubation with 100 pl of cell lysate (45 pg total protein). Detection of 10 captured FAK or FAK mutant proteins were then measured by Anti-FAK (A17) polyclonal antibody. Similarly, capture of phosphotyrosine FAK 39 T proteins were performed via coating goat anti-rabbit plates with 3.5 pg/ml of Anti-FAKp[Y397] polyclonal antibody, followed by detection using Anti-FAK (UBI) monoclonal antibody. Parameters can be optimized for the particular high thoughput screening system 15 employed, including evaluating the types of 96-well plate (e.g., anti-rabbit, protein A, or protein G), capture antibody concentration, detection antibody concentration, secondary horseradish peroxidase (HRP) antibody concentration, cell density, blocking buffer (e.g. SuperBlock Blocking TBS, 3%BSA blocking), and time of compound treatment. Using purified GST*FAK protein that is phosphorylated at residue Y397, optimization 20 of capture antibody concentration was first determined by coating 96-well Anti-rabbit plates with increasing concentrations of anti-FAKp[Y397] phosphospecific antibody. A plot of captured phosphospecific GSTFAK e 1 7 vs. Anti-FAKp[Y397] antibody concentration was generated to determine the optimal capture antibody concentration. Using this preliminary Anti-FAKp[Y397] capture concentration, the other parameters (detection and secondary 25 antibody concentrations, cell density, blocking buffer) were optimized and formatted for High Throughput Screening (HTS). Most of these parameters were optimized or evaluated simultaneously either by establishing a 96-well matrix or by cross-comparing changes in signal-to-noise ratios among a number of permutations. For example, on 96-well Anti-rabbit plates coated with 3.5 pg/ml of Anti-FAKp[Y397] capture antibody, increasing concentrations 30 of Detection and SecondaryHRP antibodies were set-up in a 96-well matrix to detect captured phosphotyrosine FAK Y397 proteins to optimized the concentrations of these antibodies. This experiment may be repeated on Protein A plates and/or by varying blocking buffer or other parameters to cross compare the signal-to-noise values among the number of permutations to optimize the system. 35 Validation of FAK cell based assay A431*FAK T cells were seeded in T25 flasks to near 80% confluency and were either left untreated or treated with 0.1 nM of Mifepristone ligand overnight (~16hrs). A431.FAK w WO 2004/027018 PCT/IB2003/003968 -19 uninduced and induced cells were washed with 10 ml of PBS and suspended in 15 ml growth media (DMEM 10%FBS, Pen/Strep/Glu, 750 pg/ml Zeocin, 50 ug/ml Hygromycin) for 30, 60, 90, or 120 minutes at 370C, 5% CO. Cells were subsequently lysed in RIPA lysis buffer (described above) and cell lysates were assayed for total protein concentration. Equivalent 5 total protein concentrations were subject to western analysis immunoblotting for FAK or phosphotyrosine FAK Y 397 proteins using standard western blotting techniques. A431*FAK" r and A431*FAK"" clones were seeded in T25 flasks to near 80% confluency and were either left untreated or treated with 0.1 nM of Mifepristone ligand overnight (~16hrs). A431.FAKwT cells were subsequently treated with increasing 10 concentrations of a FAK inhibitor (half-log dilutions of a 10 pM-starting solution) identified by using the FAK inducible cell-based assay. Cell lysates were then prepared in RIPA lysis buffer and assayed for total protein concentration. Equivalent total protein concentrations were subject to western analysis immunoblotting for FAK, general phosphotyrosine FAK, or phosphotyrosine FAK Y3 9 17 proteins using standard western blotting techniques.

Claims (14)

1. A method for identifying cell-active inhibitors of focal adhesion kinase (FAK) comprising: 5 (a) adding an inducing agent to mammalian cells to induce the expression of a gene encoding FAK, wherein said mammalian cells are stably transfected with said gene, and said gene is expressed in the presence of said inducing agent; (b) adding a test compound; (c) capturing the expressed FAK using a FAK capture agent; and 10 (d) detecting phosphorylation of said FAK.

2. The method of claim 1 wherein said mammalian cells are coated on a first solid phase, and wherein phosphorylation of FAK is detected by exposing phosphorylated FAK to an anti-phospho-tyrosine antibody and detecting the presence of said antibody.

3. The method of claim 1 wherein said FAK capture agent comprises one or 15 more types of antibodies.

4. The method of claim 3, wherein said one or more types of antibodies comprise an anti-phospho-tyrosine antibody.

5. The method of claim 1 wherein said phosphorylation is proportional to the binding of anti-phospho-tyrosine antibody to the captured FAK. 20

6. The method of claim 2 wherein said cells naturally adhere to the first solid phase and wherein said cells are lysed with a lysis buffer prior to capturing the expressed FAK.

7. The method of claim 1 wherein the inducing agent is an agonist for the expression of the gene encoding FAK. 25

8. The method of claim 1 wherein said test compound inhibits the kinase dependent phosphorylation of FAK.

9. A method for measuring the cytotoxicity of a test compound comprising: (a) stably transfecting mammalian cells with a gene encoding FAK, wherein said gene is expressed in the presence of an inducing agent; 30 (b) adding an inducing agent to induce the expression of said gene encoding FAK; (c) adding the test compound; (d) adding a cytotoxicity indicator to said cells; and, (e) detecting the cytoxicity of the test compound. 35

10. A mammalian cell stably transfected with a recombinant nucleic acid molecule, that encodes a protein comprising a sequence selected from the group consisting WO 2004/027018 PCT/IB2003/003968 -21 of SEQ ID NOS: 1, 2, 3 and 4, and wherein expression of said protein requires the presence of an inducing agent.

11. A mammalian cell that is stably transfected with a recombinant nucleic acid that inducibly expresses FAK protein. 5

12. The mammalian cell of claim 11, wherein the nucleic acid encodes proteins selected from the group consisting of human FAK splice variants, catalytic domain of human FAK, mouse FAK, rat FAK and chicken FAK.

13. A mammalian cell stably transfected with a recombinant nucleic acid molecule comprising a polynucleotide selected from the group consisting of SEQ ID NOS. 5, 10 6, 7 and 8, and wherein expression of said polynucleotide requires the presence of an inducing agent.

14. A method for identifying cell-active inhibitors of focal adhesion kinase (FAK) comprising the steps of: (a) coating a first solid phase with a homogeneous population of mammalian 15 cells so that the cells adhere to the first solid phase, wherein said cells are stably transfected with a gene encoding FAK, and wherein said gene is expressed in the presence of an inducing agent; (b) adding an inducing agent to induce the expression of said gene encoding FAK; 20 (c) adding a test compound; (d) solubilizing the adhering cells to release the cell lysate; (e) coating a second solid phase with a FAK capture agent so that the FAK capture agent adheres to the second solid phase; (f) exposing the cell lysate to the adhered FAK capture agent so that the FAK 25 capture agent captures FAK; (g) exposing the captured FAK to an anti-phospho-tyrosine antibody; and, (h) measuring binding of the anti-phospho-tyrosine antibody to the captured FAK, wherein the amount of anti-phospho-tyrosine antibody binding to the captured FAK is proportional to the amount of phosphorylation of said FAK.