Antigen binding protein and its use as addressing product for the treatment of cancer

Abstract

The present invention relates to a novel antigen binding protein, in particular a monoclonal antibody, capable of binding specifically to the protein Axl as well as the amino and nucleic acid sequences coding for said protein. From one aspect, the invention relates to a novel antigen binding protein, or antigen binding fragments, capable of binding specifically to Axl and, by inducing internalization of Axl, being internalized into the cell. The invention also comprises the use of said antigen binding protein as an addressing product in conjugation with other anti-cancer compounds, such as toxins, radio-elements or drugs, and the use of same for the treatment of certain cancers.

Claims

1. An in vitro method for the screening of an antigen-binding protein, or a binding fragment thereof, capable of delivering or internalizing a molecule of interest into a mammalian cell expressing at its surface the Axl protein, said molecule of interest being covalently linked to said antigen binding protein, said method comprising the steps of: a) selecting an antigen binding protein which is capable of specifically binding the Axl protein, wherein said antigen-binding protein has no significant activity on the proliferation of tumor cells; b) covalently linking the molecule of interest to the antigen binding protein selected in step a) to form a complex; c) contacting the complex obtained in step b), with a mammalian cell expressing at its surface the Axl protein; d) determining whether said complex has been intracellularly delivered or internalized into the said mammalian cell expressing at its surface the Axl protein; and e) selecting said antigen binding protein as a compound capable of delivering or internalizing a molecule of interest into a mammalian cell expressing at its surface the Axl protein.

2. The method of claim 1, wherein the antigen binding protein selected in step a) is capable of binding the human protein Axl.

3. The method of claim 1, wherein the antigen binding protein selected in step a) is capable of binding the human protein Axl extracellular domain (ECD) with an EC50 of at least 10.sup.?9 M.

4. The method of claim 1, wherein the determination in step d) is performed by a method selected from the group of FACS, Immunofluorescence, flow cytometry, western-blot and cytotoxicity/cytostatic evaluations methods.

5. An in vitro method for the preparation of a cytotoxic or cytostatic complex capable of delivering a cytotoxic compound into a mammalian cell expressing at its surface the Axl protein, said method comprising the step of: covalently linking a cytotoxic or cytostatic agent to an antigen binding protein, wherein said antigen binding protein is: i) capable of specifically binding the Axl protein, ii) devoid of any significant activity on the proliferation of tumor cells, and iii) internalized into a mammalian cell following its binding to said Axl protein when said Axl protein is expressed at the surface of said mammalian cell.

6. The in vitro method according to claim 5, wherein said antigen binding capable of specifically binding the Axl protein and internalized into a mammalian cell following its binding to said protein Axl when said Axl protein is expressed at the surface of said mammalian cell, is an antigen binding protein selected by an in vitro method comprising the steps of: a) selecting an antigen binding protein which is capable of specifically binding the Axl protein, wherein said antigen-binding protein has no significant activity on the proliferation of tumor cells; b) covalently linking the molecule of interest to the antigen binding protein selected in step a) to form a complex; c) contacting the complex obtained in step b), with a mammalian cell expressing at its surface the Axl protein; d) determining whether said complex has been intracellularly delivered or internalized into the said mammalian cell expressing at its surface the Axl protein; and e) selecting said antigen binding protein as a compound capable of delivering or internalizing a molecule of interest into a mammalian cell expressing at its surface the Axl protein.

7. An in vitro method for the preparation of a cytotoxic or cytostatic complex capable of delivering a cytotoxic compound into a mammalian cell, said method comprising the step of: covalently linking a cytotoxic agent to a compound which is: i) capable of specifically binding the Axl protein, ii) devoid of any significant activity on the proliferation of tumor cells, and iii) is internalized into a mammalian cell following its binding to said protein Axl when said Axl protein is expressed at the surface of said mammalian cell.

8. The in vitro method according to claim 7, wherein said compound capable of specifically binding the Axl protein and internalized into a mammalian cell following its binding to said protein Axl when said Axl protein is expressed at the surface of said mammalian cell, is an antigen binding protein selected by an in vitro method comprising the steps of: a) selecting an antigen binding protein which is capable of specifically binding the Axl protein, wherein said antigen-binding protein has no significant activity on the proliferation of tumor cells; b) covalently linking the molecule of interest to the antigen binding protein selected in step a) to form a complex; c) contacting the complex obtained in step b), with a mammalian cell expressing at its surface the Axl protein; d) determining whether said complex has been intracellularly delivered or internalized into the said mammalian cell expressing at its surface the Axl protein; and e) selecting said antigen binding protein as a compound capable of delivering or internalizing a molecule of interest into a mammalian cell expressing at its surface the Axl protein.

9. The method of claim 1, wherein the antigen binding protein selected in step a) is capable of binding the extracellular domain (ECD) of the human Axl protein, said ECD having the sequence SEQ ID No. 31 or 32.

10. The method of claim 1, wherein the antigen binding protein selected in step a) is capable of binding the human Axl protein extracellular domain (ECD) with an EC50 comprised between 10.sup.?9 and 10.sup.?12 M.

11. The method of claim 5, wherein the Axl protein of step a) is the human Axl protein.

12. The method of claim 7, wherein the Axl protein of step a) is the human Axl protein.

Description

FIGURE LEGENDS

(1) FIG. 1: in vitro cytotoxicity assay using Mab-zap conjugated secondary antibody on SN12C cells.

(2) FIG. 2A, FIG. 2B and FIG. 2C: Binding specificity of 1613F12 on the immobilized rhAxl-Fc protein (FIG. 2A), rhDtk-Fc (FIG. 2B) or rhMer-Fc (FIG. 2C) proteins by ELISA.

(3) FIG. 3: FACS analysis of the 1613F12 binding on human tumor cells

(4) FIG. 4: ELISA on the immobilized rmAxl-Fc protein (rm for murine recombinant).

(5) FIG. 5: 1613F12 binding on COS7 cells as determined by indirect labelling protocol using flow cytometry method.

(6) FIG. 6: Competition ELISA of Gas6 binding using 1613F12.

(7) FIG. 7: Epitope binding analysis by western Blot using SN12C cell lysate. NH (no heat); NR (no reduction); H (heat); R (reduction). GAPDH detection attests to the correct sample loading on the gel.

(8) FIG. 8A and FIG. 8B: Study of Axl downregulation after 1613F12 binding on SN12C cells by Western Blot with FIG. 8AWestern blot image representative of the 3 independent experiments performed (The western blot analysis was performed after a 4 h and 24 h incubation of the 1613F12 on SN12C cells); and FIG. 8BOptical density quantification of the presented film using QuantityOne software.

(9) FIG. 9A, FIG. 9B and FIG. 9C: Immunofluorescence microscopy of SN12C cells after incubation with the 1613F12 FIG. 9APhotographs of the mIgG1 isotype control conditions both for the membrane and the intracellular staining. FIG. 9BMembrane staining. FIG. 9CIntracellular staining of both Axl receptor using the 1613F12 and of the early endosome marker EEA1. Image overlays are presented bellow and co-localizations visualized are indicated by the arrows.

(10) FIG. 10: Effect of 1613F12 on in vitro SN12C cells proliferation compared to the effect of the mIgG1 isotype control antibody.

(11) FIG. 11A through FIG. 11K: Direct cytotoxicity assays of the 1613F12-saporin immunoconjugate using various human tumor cell lines. FIG. 11ASN12C, FIG. 11BCalu-1, FIG. 11CA172, FIG. 11DA431, FIG. 11EDU145, FIG. 11FMDA-MB435S, FIG. 11GMDA-MB231, FIG. 11HPC3, FIG. 11I-NCI-H226, FIG. 11JNCI-H125, FIG. 11KPanc1.

(12) FIG. 12A, FIG. 12B, FIG. 12C: Antibody binding competition between FIG. 12A 427B7, FIG. 12B 11D12 and FIG. 12C 320G10 obtained by Biacore.

(13) FIG. 13: Direct cytotoxicity assays using SN12C cells (in presence of 1A4-saporin immunoconjugate and an isotype control coupled to saporin).

EXAMPLES

Example 1: Axl Receptor Internalization

(14) As an immunoconjugate approach is more efficient when the targeted antigen is an internalizing protein, Axl receptor internalization using Mab-Zap cytotoxicity assay on human tumor cell lines was studied. More precisely, the Mab-Zap reagent is a chemical conjugate including an affinity purified goat anti-mouse IgG and the ribosome-inactivating protein, saporin. If internalization of the immune complex occurs, saporin breaks away from the targeting agent and inactivates the ribosomes, resulting in protein synthesis inhibition and, ultimately, cell death. Cell viability determination after 72 hours of incubation with the anti-Axl antibody or with mIgG1 isotype control antibody on Axl-positive cells allows concluding on the anti-Axl antibody-induced Axl receptor internalization. For this study, several other murine anti-Axl antibodies were tested.

(15) For this example highly Axl-positive cells, as determined using Qifikit reagent (Dako), were used. Data are presented in the following table 5.

(16) TABLE-US-00005 TABLE 5 Antigen binding capacity of the MAB154 anti-Axl commercial antibody determined for the human renal cancer SN12C cells RTK AXL Cell line Antibody MAB154 SN12C >100 000

(17) In the following example, the SN12C cells were used as non limitative example. Any other cell line expressing appropriate level of Axl receptor on its cell surface could be used.

(18) Concentration ranges of either the 1613F12 or any other tested murine anti-Axl antibody or the mIgG1 isotype control antibody were pre-incubated with 100 ng of Mab-Zap (Advanced targeting systems) secondary antibody in cell culture medium for 30 min at RT. These mixtures were loaded on sub-confluent SN12C cells plated in white 96-well plate microplate. Plates were incubated for 72 h at 37? C. in presence of 5% CO.sub.2. Cell viability was determined using a Cell Titer Glo cell proliferation method according to the manufacturer's instructions (Promega). Several controls are performed: i) without any secondary immunoconjugate and ii) without primary antibody. In parallel, assays are performed with a mIgG1 isotype control.

(19) Data obtained with the 1613F12 antibody are presented in the FIG. 1 and the data obtained with the other murine anti-Axl antibodies are presented in Table 6. The 1613F12 shows a maximal cytotoxic effect on the SN12C cells of ?36%. The percentages of maximum cytotoxicity obtained with other anti-Axl antibodies are listed in Table 6.

(20) No cytotoxic effect was observed in presence of the 9G4 antibody, considered as mIgG1 isotype control in the experiment. No cytotoxicity was observed in wells containing only primary antibodies (data not shown).

(21) Using different anti-Axl antibodies, maximal cytotoxicity percentages are comprised between 36.3% and 69% after 3 days of incubation with the antibodies.

(22) Referring to the obtained cytotoxicity effect in this experiment, the Axl receptor appears to be a convenient antigen to target for an immunoconjugate approach as the various immune complexes comprising Axl-Axl anti-antibody-MabZap can trigger an effective cytotoxicity of the targeted cells.

(23) Thus the Axl receptor appears to be a convenient antigen to target for an immunoconjugate approach as the immune complex comprising Axl-1613F12-MabZap, or all other anti-Axl antibodies of the invention, triggers an effective cytotoxicity of the targeted cells.

(24) TABLE-US-00006 TABLE 6 Percentages of the maximum cytotoxicity obtained in indirect SN12C cytotoxicity assays % cytotoxicity max 1A4 direct* 11D12 54.2% 320G10 41.4% 427B7 36.3% 517F1 66.5% 530C8 50.2% 534G9 49.9% 535F7 55.6% 547A1 69.0% 1023F10 49.7% 1614G5 65.1% *Axl antibody tested in direct in vitro SN12C cytotoxicity assay

Example 2: Generation of an Antibody Against rhAxl-ECD

(25) Various strategies were used to generate murine antibodies targeting the human Axl extracellular domain.

(26) The detailed protocol used to obtain the 1613F12 antibody is described bellow.

(27) To generate murine monoclonal antibodies (Mabs) against human extracellular domain (ECD) of the Axl receptor, 5 BALB/c mice were immunized 5-times s.c. with 15-20.Math.10.sup.6 CHO-Axl cells and twice with 20 ?g of the rh Axl-ECD. The first immunization was performed in presence of Complete Freund Adjuvant (Sigma, St Louis, Md., USA). Incomplete Freund adjuvant (Sigma) was added for following immunizations.

(28) Three days prior to the fusion, immunized mice were boosted with both 20.Math.10.sup.6 CHO-Axl cells and 20 ?g of the rhAxl-ECD with IFA.

(29) To generate hybridomas, splenocytes and lymphocytes were prepared by perfusion of the spleen and by mincing of the proximal lymph nodes, respectively, harvested from 1 out of the 5 immunized mice (selected after sera titration) and fused to SP2/0-Ag14 myeloma cells (ATCC, Rockville, Md., USA). The fusion protocol is described by Kohler and Milstein (Nature, 256:495-497, 1975). Fused cells are then subjected to HAT selection. In general, for the preparation of monoclonal antibodies or their functional fragments, especially of murine origin, it is possible to refer to techniques which are described in particular in the manual Antibodies (Harlow and Lane, Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor N.Y., pp. 726, 1988).

(30) Approximately 10 days after the fusion, colonies of hybrid cells were screened. For the primary screen, supernatants of hybridomas were evaluated for the secretion of Mabs raised against the Axl ECD protein using an ELISA. In parallel, a FACS analysis was performed to select Mabs able to bind to the cellular form of Axl present on the cell surface of both wt CHO and Axl expressing CHO cells (ATCC).

(31) As soon as possible, selected hybridomas were cloned by limit dilution and subsequently screened for their reactivity against the Axl ECD protein. Cloned Mabs were then isotyped using an Isotyping kit (cat #5300.05, Southern Biotech, Birmingham, Ala., USA). One clone obtained from each hybridoma was selected and expanded. Once produced the anti-Axl antibodies were further studied for their ability to be internalized following Axl binding on the cell-surface. In parallel, the hybridomas are deposited at the CNCM.

(32) ELISA assays are performed as followed either using pure hybridoma supernatant or, when IgG content in supernatants was determined, titration was realized starting at 5 ?g/ml. Then a ? serial dilution was performed in the following 11 rows. Briefly, 96-well ELISA plates (Costar 3690, Corning, N.Y., USA) were coated 50 ?l/well of the rh Axl-Fc protein (R and D Systems, cat No. 154-AL) or rhAxl-ECD at 2 ?g/ml in PBS overnight at 4? C. The plates were then blocked with PBS containing 0.5% gelatin (#22151, Serva Electrophoresis GmbH, Heidelberg, Germany) for 2 h at 37? C. Once the saturation buffer discarded by flicking plates, 50 ?l of pure hybridoma cell supernatants or 50 ?l of a 5 ?g/ml solution were added to the ELISA plates and incubated for 1 h at 37? C. After three washes, 50 ?l horseradish peroxidase-conjugated polyclonal goat anti-mouse IgG (#115-035-164, Jackson Immuno-Research Laboratories, Inc., West Grove, Pa., USA) was added at a 1/5000 dilution in PBS containing 0.1% gelatin and 0.05% Tween 20 (w:w) for 1 h at 37? C. Then, ELISA plates were washed 3-times and the TMB (#UP664782, Uptima, Interchim, France) substrate was added. After a 10 min incubation time at room temperature, the reaction was stopped using 1 M sulfuric acid and the optical density at 450 nm was measured.

(33) For the selection by flow cytometry, 10.sup.5 cells (CHO wt or CHO-Axl) were plated in each well of a 96 well-plate in PBS containing 1% BSA and 0.01% sodium azide (FACS buffer) at 4? C. After a 2 min centrifugation at 2000 rpm, the buffer was removed and hybridoma supernatants or purified Mabs (1 ?g/ml) to be tested were added. After 20 min of incubation at 4? C., cells were washed twice and an Alexa 488-conjugated goat anti-mouse antibody 1/500? diluted in FACS buffer (#A11017, Molecular Probes Inc., Eugene, USA) was added and incubated for 20 min at 4? C. After a final wash with FACS buffer, cells were analyzed by FACS (Facscalibur, Becton-Dickinson) after addition of propidium iodide to each tube at a final concentration of 40 ?g/ml. Wells containing cells alone and cells incubated with the secondary Alexa 488-conjugated antibody were included as negative controls. Isotype controls were used in each experiment (Sigma, ref M90351MG). At least 5000 cells were assessed to calculate the mean value of fluorescence intensity (MFI).

(34) More precisely, the fusion was performed with 300.10.sup.6 of harvested splenocytes and 300.Math.10.sup.6 myeloma cells (1:1 ratio). Two hundred thousand cells of the resulting cell suspension were then plated at 2.Math.10.sup.6 cell/ml in 30 96-well plates.

(35) A first screen (around Day 14 after fusion) both by ELISA on the rhAxl-ECD protein and by FACS analysis using the both wt CHO and Axl expressing CHO cells allowed to select 10 hybridomas presenting optical densities (ODs) above 1 on the rh Axl-ECD coating and MFI bellow 50 on wt CHO cells and above 200 on CHO-Axl cells.

(36) These 10 hybridomas were expanded and cloned by limit dilution. One 96-well plate was prepared for each code. Nine days after plating, supernatants from cloning plates were first screened by ELISA for their binding specificity for the extracellular domain of the rh Axl-ECD protein. Three clones of each code were expanded and isotyped. Once produced the anti-Axl antibodies were further studied for their ability to be internalized following Axl binding on the cell-surface.

(37) The protocols applied to give rise to the other anti-Axl antibodies are summarized in Table 7.

(38) TABLE-US-00007 TABLE 7 Immunization protocols Injections Immunogen Site of injection Organ source Hybridoma screen Antibody 3-5 immunizations + 5-20 ?g of rh Axl protein s.c. spleen ELISA rhAxl 1A4, 320G10, 1 boost (i.v or i.p.) (monomeric or dimeric) proximal lymph FACS DU145 11D12, 517F1, 530C8, nodes 534G9, 535F7, 547A1, 427B7, 1023F10 5 immunizations 15-20 10.sup.6 CHO-Axl cells s.c. spleen ELISA rh Axl 1613F12 2 immunization + 20 ?g rhAxl protein proximal lymph FACS CHO Axl/CHO wt 1614G5 1 boost (i.v or i.p.) nodes

(39) Moreover, the isotype of the anti-Axl antibodies are given in the following Table 8.

(40) In addition, the epitope diversity of the obtained Axl antibodies was assessed by antibody binding competition. Epitope clustering experiments were run out on a Biacore X device. This instrument based on the optical phenomenon of surface plasmon resonance (SPR) enables the detection and measurement of protein-protein interactions in real time, without any labelling.

(41) Briefly, the experiments were realized using a sensor chip CM5 as the biosensor. An anti-polyhistidine mouse IgG1 antibody (R and D Systems, MAB050) was coupled on both flow cells (FC1 and FC2) using amine coupling chemistry.

(42) The experiments were carried out at a flow rate of 10-20 ?l/min in a HBS-EP buffer at 25? C. The recombinant hAxl-Fc chimera protein (R and D Systems, 154-AL-100) was injected during one minute at the concentration of 5-10-20 ?g/ml on both flowcells. Typically, this injection captured between 300 to 500 RU of protein. A first antibody (culture supernatant or purified antibody at the concentration of 50 ?g/ml) is injected on the flowcell 1 during 1-2 min. in order to reach a saturation of the captured antigen. The second antibody was the injected on both flowcells (culture supernatant or purified antibody at the concentration of 50 ?g/ml) during 60 s. Both antibodies are classed in the same epitopic group when the signal of the second antibody on the flowcell 1 is at least 10 times lower than the response on the flowcell 2.

(43) At the end of each cycle, the surfaces are regenerated by injecting a 10 mM glycine hydrochloride pH1.5 solution during 30 s.

(44) The epitopic group of the anti-Axl antibodies is precised in Table 8.

(45) Examples of Biacore data are reported in FIG. 12 which illustrates the competition between 427B7, 11D12 and 320G10. The same way of analysis was applied for the other anti-Axl antibodies of the invention.

(46) TABLE-US-00008 TABLE 8 Isotype and Epitopic group of the Axl antibodies. Isotype Epitopic group CNCM 1A4 IgG1 K group #B I-3947 11D12 IgG1 K group #B I-4116 320G10 IgG1 K group #C I-4514 427B7 IgG1 K group #A I-4518 517F1 IgG1 K group #B I-4502 530C8 IgG1 K group #B1 I-4728 534G9 IgG1 K group #B2 I-4729 535F7 IgG1 K group #B I-4510 547A1 IgG1 K group #B I-4506 1023F10 IgG1 K group #B I-4507 1613F12 IgG1 K group #B I-4505 1614G5 IgG1 K group #A I-4519

(47) They spread over 5 epitopic clusters.

Example 3: Axl Binding Specificity

(48) In this example, the binding of the anti-Axl antibodies was first studied on the rhAxl-Fc protein. Then, their binding on the two other members of the TAM family, rhDtk-Fc and rhMer-Fc, was studied.

(49) Briefly, the recombinant human Axl-Fc (R and D systems, cat No. 154AL/CF), rhDtk (R and D Systems, cat No. 859-DK) or rhMer-Fc (R and D Systems, cat No. 891-MR) proteins were coated overnight at 4? C. to Immulon II 96-well plates and, after a 1 h blocking step with a 0.5% gelatine solution, purified anti-Axl antibodies were added for an additional 1 h at 37? C. at starting concentration of 5 ?g/ml (3.33 10.sup.?8M). Then ? serial dilutions were done over 12 columns. Plates were washed and a goat anti-mouse (Jackson) specific IgG-HRP was added for 1 h at 37? C. Reaction development was performed using the TMB substrate solution. The commercial anti-Axl Mab 154 antibody was also used in parallel (data not shown). Coating controls were performed in presence of a goat anti-human IgG Fc polyclonal serum labelled with HRP (Jackson, ref 109-035-098) and/or in presence of a HRP-coupled anti-Histidine antibody (R and D Systems, ref: MAB050H). No non specific binding was observed in absence of primary antibody (diluant). Results obtained with the 1613F12 antibody are graphically represented in FIGS. 2A, 2B and 2C, respectively. Data obtained for the other anti-Axl antibodies are reported in the Table 9.

(50) TABLE-US-00009 TABLE 9 Binding specificty of the Axl antibodies assessed by ELISA using human recombinant proteins* rhAxl-Fc rhMer-Fc rhDtk-Fc 1A4 2.816 0.298 0.142 11D12 2.994 0.281 0.104 320G10 2.737 0.287 0.107 427B7 2.840 0.284 0.144 517F1 2.862 0.300 0.112 530C8 2.712 0.250 0.115 534G9 2.629 0.226 0.108 535F7 2.721 0.235 0.115 547A1 2.691 0.227 0.099 1023F10 2.676 0.247 0.114 1613F12 2.806 0.252 0.121 1614G5 2.592 0.275 0.122 *Values of ODs (450 nm) obtained in presence of a 5 ?g/ml concentration of Axl antibody are reported here.

(51) This example shows that the 1613F12 antibody only binds to the rhAxl-Fc protein and does not bind on the two other members of the TAM family, rhDtk or rhMer. No cross-specificity of binding of the 1613F12 antibody is observed between TAM members. Similar results are observed for all the other tested anti-Axl antibodies. Indeed, they present a specificity of binding restricted to Axl; they do not recognize the other members of the TAM family.

Example 4: 1613F12 and Other Anti-Axl Antibodies Recognized the Cellular Form of Axl Expressed on Tumor Cells

(52) Cell surface Axl expression level on human tumor cells was first established using a commercial Axl antibody (R and D Systems, ref: MAB154) in parallel of calibration beads to allow the quantification of the Axl expression level. Then, binding of the cell-surface Axl was studied using the anti-Axl antibodies of the invention. In both cases, the experimental conditions were as briefly described bellow.

(53) For cell surface binding studies, two fold serial dilutions of a 10 ?g/ml (6.66 10.sup.?8 M) primary antibody solution (anti-Axl antibodies, MAB154 anti-Axl commercial antibody or mIgG1 isotype control 9G4 Mab) are prepared on 10 points and applied on 2.10.sup.5 cells for 20 min at 4? C. After 3 washes in phosphate-buffered saline (PBS) supplemented with 1% BSA and 0.01% NaN.sub.3, cells were incubated with the secondary Goat anti-mouse Alexa 488 antibody (1/500? dilution) for 20 minutes at 4? C. After 3 additional washes in PBS supplemented with 1% BSA and 0.1% NaN.sub.3, cells were analyzed by FACS (FACS, Becton-Dickinson). At least 5000 cells were assessed to calculate the mean value of fluorescence intensity.

(54) For quantitative ABC determination using the MAB154 Axl antibody, QIFIKIT? calibration beads are used. In parallel with the QIFIKIT? beads, cells are incubated, with a Polyclonal Goat Anti-Mouse Immunoglobulins/FITC (Goat F(ab).sub.2) at a saturating concentration. The number of antigenic sites on the studied cells is then determined by interpolation of the calibration curve (the fluorescence intensity of the individual bead populations against the number of Mab molecules on the beads).

(55) 4.1. Quantification of Cell-Surface Axl Expression Level

(56) Axl expression level on the surface of human tumor cells was determined by flow cytometry using indirect immunofluorescence assay (QIFIKIT? method (Dako, Denmark), a quantitative flow cytometry kit for assessing cell surface antigens. A comparison of the mean fluorescence intensity (MFI) of the known antigen levels of the beads via a calibration graph permits determination of the antibody binding capacity (ABC) of the cell lines.

(57) Table 10 presents Axl expression level detected on the surface of various human tumor cell lines (SN12C, Calu-1, A172, A431, DU145, MDA-MB435S, MDA-MB231, PC3, NCI-H226, NCI-H125, MCF7, Panc1) (ATCC, NCI) as determined using QIFIKIT? using the Axl commercial antibody MAB154 (R and D Systems). Values are given as Antigen Binding Complex (ABC).

(58) TABLE-US-00010 TABLE 10 MCF7 NCI-H125 MDA-MB-435S Panc1 MDA-MB-231 Calu-1 SN12C Tumor Breast NSCLC Breast Pancreas Breast Lung Renal type/organ ABC 71 5 540 17 814 36 809 61 186 >100 000 >100 000 (Qifikit) A172 A431 DU-145 PC3 NCI-H226 Tumor glioblastoma Epidermoid Prostate prostate NSCLC type/organ carcinoma ABC 52421 3953 55268 8421 32142 (Qifikit)

(59) Results obtained with a commercial Axl monoclonal antibody (MAB154) showed that Axl receptor is expressed at various levels depending of the considered human tumor cell.

(60) 4.2. Axl Detection by 1613F12 and Anti-Axl Antibodies According to the Invention on Human Tumor Cells

(61) Axl binding obtained using the 1613F12 was studied more extensively on a panel of different human tumor cells. Data are graphically presented in FIG. 3.

(62) The binding of the other anti-Axl antibodies was documented using the SN12C human renal tumor cells only.

(63) Anti-Axl antibody dose response curves were run. MFIs obtained with the 1613F12 antibody using the various human tumor cells were then analysed with Prism software. Data obtained with the 1613F12 are presented in FIG. 3. The EC.sub.50 of the binding on SN12C cells obtained for the other tested anti-Axl antibodies are reported in Table 11.

(64) TABLE-US-00011 TABLE 11 Binding of the Axl antibodies on SN12c cells studied by flow cytometry EC.sub.50 SN12C binding (M) 1A4 1.E?09 11D12 2.E?10 320G10 4.E?10 427B7 5.E?10 517F1 7.E?10 530C8 1.E?09 534G9 4.E?10 535F7 8.E?11 547A1 2.E?10 1023F10 1.E?10 1614G5 9.E?10

(65) As illustrated in FIG. 3, data indicate that the 1613F12 binds specifically to the membrane Axl receptor as attested by the saturation curve profiles. Similar profiles were obtained with the other tested antibodies (data not shown). The values of the EC.sub.50 of anti-Axl antibody binding on SN12C cells are comprised between 10.sup.?9M and 10.sup.?12 M.

Example 5: 1613F12 Inter-Species Crosspecificity

(66) To address the species cross-specificity of the 1613F12, two species were considered: mouse and monkey. First the binding on the recombinant mouse (rm) Axl receptor is studied by ELISA (FIG. 4). Then flow cytometry experiments were performed using monkey COS7 cells as these cells express the Axl receptor on their surface (FIG. 5). The COS7 cell line was obtained by immortalizing a CV-1 cell line derived from kidney cells of the African green monkey with a version of the SV40 genome that can produce large T antigen but has a defect in genomic replication.

(67) rmAxl-Fc ELISA

(68) Briefly, the recombinant mouse Axl-Fc (R and D systems, cat No. 854-AX/CF) proteins were coated overnight at 4? C. to Immulon II 96-well plates and, after a 1 h blocking step with a 0.5% gelatine solution, the 1613F12 purified antibody was added for one additional hour at 37? C. at starting concentration of 5 ?g/ml (3.33 10.sup.?8 M). Then ? serial dilutions were done over 12 columns. Plates were then washed and a goat anti-mouse (Jackson) specific IgG HRP was added for 1 h at 37? C. Reaction development was performed using the TMB substrate solution. The commercial mouse anti-Axl Mab 154 antibody is also used in parallel. Coating controls are performed in presence of a goat anti-human IgG Fc polyclonal serum coupled with HRP (Jackson, ref 109-035-098) and/or in presence of a HRP-coupled anti-Histidine antibody (R and D Systems, ref: MAB050H). No specific binding is observed in the absence of primary antibody (diluant).

(69) Results are represented in FIG. 4. This figure shows that the 1613F12 Mab described in the present invention does not bind to the murine Axl ECD domain.

(70) FACS COS7

(71) For 1613F12 cellular binding studies using COS7 cells, 2.Math.10.sup.5 cells were incubated with an antibody concentration range prepared by ? serial dilution (12 points) of a 10 ?g/ml (6.66 10.sup.?8 M) antibody solution of 1613F12 or m9G4 (mIgG1 isotype control Mab) for 20 min at 4? C. After 3 washes in phosphate-buffered saline (PBS) supplemented with 1% BSA and 0.01% NaN.sub.3, cells were incubated with secondary antibody goat anti-mouse Alexa 488 (dilution 1/500) for 20 minutes at 4? C. After 3 additional washes in PBS supplemented with 1% BSA and 0.1% NaN.sub.3, cells were analyzed by FACS (Facscalibur, Becton-Dickinson). At least 5000 cells were assessed to calculate the mean value of fluorescence intensity. Data are analyzed using Prism software.

(72) Results are represented in FIG. 5. The titration curve established on COS7 cells using either 1613F12 or mIgG1 isotype control confirms that 1613F12 is able to recognize the monkey cellular form of the Axl receptor expressed on the surface of the COS7 cells. Plateau is reached for 1613F12 concentrations above 0.625 ?g/ml (4.2 10.sup.?10 M). As expected no binding is observed in presence of the mIgG1 isotype control.

(73) This example illustrates the fact that the 1613F12 does not cross-react with the mouse Axl receptor. In contrast it strongly binds to the monkey Axl receptor expressed on the surface of COS7 cells.

Example 6: Gas6 Competition Experiments Performed in Presence of the 1613F12

(74) To further characterize the anti-Axl Mabs, Gas6 competition assays were performed. In this assay, the free rhAxl-Fc protein and the anti-Axl antibody are incubated to form antigen-antibody complex and then the complexes are loaded on Gas6-coated surface in the assay plate. The unbound antibody-antigen complexes are washed out before adding enzyme-linked secondary antibody against the human Fc portion of the rhAxl-Fc protein. The substrate is then added and the antigen concentration can be determined by the signal strength elicited by the enzyme-substrate reaction.

(75) Briefly reaction mixture comprising the rhAxl-Fc protein in the presence or not of the anti-Axl Mabs to be tested, are prepared on a separate saturated (0.5% gelatin in PBS 1?) plate. Serial 1: 2 dilutions (starting from 80 ?g/ml on 12 columns) of murine anti-Axl antibodies are performed. Then 0.5 ?g/ml of the rhAxl-Fc protein is added (R and D Systems, ref 154AL/CF), except to the negative control line that contains only ELISA diluant (0.1% gelatin, 0.05% Tween 20 in PBS 1?). After homogenisation, the competition samples are loaded on Gas6-coated plates with a 6 ?g/ml rhGas6 solution in PBS (R and D systems cat No. 885-GS-CS/CF). After incubation and several washes, bound rhAxl-Fc proteins are detected using a goat anti-Human IgG-HRP (Jackson, ref 109-035-098). Once bound, the TMB substrate is added to the plates. The reaction is stopped by addition of 1M H.sub.2SO.sub.4 acid solution and the obtained optical densities read at 450 nm using a microplate reader instrument.

(76) This experiment (FIG. 6) shows that the 1613F12 is able to compete with the rhAxl-Fc binding on its immobilized ligand. Competition with Gas6 binding occurs in presence of 1613F12 antibody concentrations above 2.5 ?g/ml (1.67 10.sup.?8 M). No more binding of the rhAxl-Fc on the immobilized Gas6 is observed in presence of a 1613F12 concentration above 10 ?g/ml (6.67 10.sup.?8 M). The 1613F12 blocks Gas6 binding to rhAxl-Fc.

Example 7: Epitope Recognition by Western Blot

(77) To determine if the 1613F12 recognizes a linear or a conformational epitope, western blot analysis was done using SN12C cell lysates. Samples were differently treated to be in reducing or non reducing conditions. If a band is visualized with reduced sample, the tested antibody targets a linear epitope of the ECD domain; If not, it is raised against a conformation epitope of the Axl ECD.

(78) SN12C cells were seeded in RPMI+10% heat inactivated FBS+2 mM L-glutamine at 5.10.sup.4 cells/cm.sup.2 in T162 cm.sup.2 flasks for 72 h at 37? C. in a 5% CO.sub.2 atmosphere. Then the cells were washed twice with phosphate buffered saline (PBS) and lysed with 1.5 ml of ice-cold lysis buffer [50 mM Tris-HCl (pH7.5); 150 mM NaCl; 1% Nonidet P40; 0.5% deoxycholate; and 1 complete protease inhibitor cocktail tablet plus 1% antiphosphatases]. Cell lysates were shaken for 90 min at 4? C. and cleared at 15 000 rpm for 10 min. Protein concentration was quantified using BCA. Various samples were loaded. First 10 ?g of whole cell lysate (10 ?g in 20 ?l) were prepared in reducing conditions (1? sample buffer (BIORAD)+1? reducing agent (BIORAD)) and loaded on a SDS-PAGE after 2 min incubation at 96? C. Secondly two other samples of 10 ?g of whole cell lysate were prepared in non-reducing conditions (in 1? sample buffer (BIORAD) only). Prior to be loaded on the SDS-PAGE gel, one of these two last samples is heated 2 min incubation at 96? C.; the other one is kept on ice. After migration, the proteins are transferred to nitrocellulose membrane. Membranes were saturated for 1 h at RT with TBS-tween 20 0.1% (TBST), 5% non-fat milk and probed with the 1613F12 at 10 ?g/ml overnight at 4? C. Antibodies were diluted in Tris-buffered saline-0.1% tween 20 (v/v) (TBST) with 5% non-fat dry milk. Then membranes were washed with TBST and incubated with peroxydase-conjugated secondary antibody (dilution 1/1000) for 1 h at RT. Immunoreactive proteins were visualized with ECL (Pierce #32209). After Axl visualization, membranes were washed once again with TBST and incubated for 1 h at RT with mouse anti-GAPDH antibody (dilution 1/200 000). Then membranes were washed in TBST and incubated with peroxydase-conjugated secondary antibodies, for 1 h at RT. Membranes were washed and GAPDH was revealed using ECL.

(79) Results are represented in FIG. 7.

(80) The 1613F12 mainly recognizes a conformational epitope as a specific band is essentially observed in non-reduced conditions. However a faint signal is detected in the denaturating migrating condition of the SN12C cell lysate indicating 1613F12 is able to weakly bind to a linear epitope.

Example 8: Measurement of Axl Down-Regulation Triggered by the 1613F12 and Other Anti-Axl Antibodies by Western Blot

(81) In the following example, the human renal cell carcinoma cell line SN12C (ATCC) was selected to address the activity of Axl antibodies on Axl receptor expression. The SN12C cell line overexpresses the Axl receptor. The Axl down-regulation was studied by Western-Blot on whole cell extracts.

(82) SN12C cells were seeded in RPMI+10% heat inactivated FBS+2 mM L-glutamine at 6.Math.10.sup.4 cells/cm.sup.2 in six-well plates for 48 h at 37? C. in a 5% CO.sub.2 atmosphere. After two washes with phosphate buffer saline (PBS), cells were serum-starved in a medium containing either 800 ng/ml recombinant mouse gas6 ligand (R and D Systems, ref: 986-GS/CF) or 10 ?g/ml of a mIgG1 isotype control antibody (9G4) or 10 ?g/ml of the Axl antibody of the present invention and incubated for 4 h or 24 additional hours. Then the medium was gently removed and cells washed twice with cold PBS. Cells were lysed with 200 ?l of ice-cold lysis buffer [50 mM Tris-HCl (pH7.5); 150 mM NaCl; 1% Nonidet P40; 0.5% deoxycholate; and 1 complete protease inhibitor cocktail tablet plus 1% antiphosphatases]. Cell lysates were shaken for 90 min at 4? C. and cleared at 15 000 rpm for 10 min. Protein concentration was quantified using BCA method. Whole cell lysates (10 ?g in 20 ?l) were separated by SDS-PAGE and transferred to nitrocellulose membrane. Membranes were saturated for 1 h at RT with TBS-Tween 20 0.1% (TBST), 5% non-fat milk and probed with a commercial M02 Axl antibody at 0.5 ?g/ml (AbNova H00000558-M02) overnight at 4? C. Antibodies were diluted in Tris-buffered saline-0.1% tween 20 (v/v) (TBST) with 5% non-fat dry milk. Then membranes were washed with TBST and incubated with peroxydase-conjugated secondary antibody (dilution 1/1000) for 1 h at RT. Immunoreactive proteins were visualized with ECL (Pierce #32209). After Axl visualization, membranes were washed once again with TBST and incubated for 1 h at RT with mouse anti-GAPDH antibody (dilution 1/200000). Then membranes were washed in TBST and incubated with peroxydase-conjugated secondary antibodies, for 1 h at RT. Membranes were washed and GAPDH was revealed using ECL. Band intensity was quantified by densitometry.

(83) Experimental data describing the effect of the isotype control, rmGas6 and the 1613F12 antibody are provided in FIGS. 8A-8B. They are representative of 3 independent experiments. Then the percentage of downregulated Axl receptor in the SN12C cell extract at 24 h obtained in presence of other Axl antibodies is given in Table 12.

(84) The results show that 1613F12 is able to down-regulate Axl in an Axl-overexpressing human tumor cell line. At 4 h, the 1613F12 triggers a 66% Axl down-regulation, and up to 87% after a 24 hour incubation with the 1613F12. For the other tested anti-Axl antibodies, the percentage of downregulated Axl in SN12C cells after a 24 h incubation of Axl antibody is comprised between 48.7% and 89%.

(85) TABLE-US-00012 TABLE 12 Axl downregulation after Axl antibody binding on SN12C cells studied by Western Blot % downregulated Axl at 24 h 1A4 69.0 11D12 64.4 320G10 71.0 427B7 62.0 517F1 68.0 530C8 48.7 534G9 64.5 535F7 80.0 547A1 80.0 1023F10 70.0 1614G5 63.0

Example 9: Flow Cytometry Study of the 1613F12 and Other Anti-Axl Antibodies Effect on Cell Surface Axl Expression

(86) Flow cytometry technique allows labelling of cell-surface Axl receptor. The use of this technique can highlight the effect of antibody on the membrane Axl expression. Human renal tumor SN12C cells that express high levels of Axl were used in this example.

(87) SN12C tumor cell line was cultured in RMPI1640 with 1% L-glutamine and 10% of FCS for 3 days before experiment. Cells were then detached using trypsin and plated in 6-multiwell plate in RPMI1640 with 1% L-glutamine and 5% FBS. The next day, antibodies of interest were added at 10 ?g/ml. Untreated wells were also included. The cells are incubated at 37? C., 5% CO.sub.2. Twenty four hours later, cells were washed with PBS, detached and incubated with the same antibodies of interest in FACS buffer (PBS, 1% BSA, 0.01% sodium azide). Untreated wells were also stained with the same antibody in order to compare the signal intensity obtained with the same Mab on the treated and the non-treated cells. Cells were incubated for 20 minutes at 4? C. and washed three times with FACS buffer. An Alexa 488-labeled goat anti-mouse IgG antibody was incubated for 20 minutes and cells were washed three times before FACS analysis on propidium iodide negative cell population.

(88) Two parameters are determined: (i) the difference of the fluorescent signal detected on the surface of untreated (no Ab) cells compared to the Ab-treated cells at T24 h and (ii) the percentage of remaining Axl on the cell surface. The percentage of remaining Axl is calculated as follows:
% remaining Axl=(MFI.sub.Ab 24 h/MFI.sub.no Ab 24 h)?100

(89) Data from one representative experiment (out of 3) are presented in Table 13a. The results were reproduced in three independent experiments.

(90) The difference of MFI between the staining of a Mab in the untreated cell and the treated condition with the same antibody reflects a down-regulation of the Axl protein on the cell surface of the cells due to the binding of the considered Mab. Conditions without antibody gave similar results to conditions in presence of the isotype control antibody (m9G4).

(91) TABLE-US-00013 TABLE 13a MFI at ? (MFI.sub.No Ab 24 h ? Labelling Treatment T24 h MFI.sub.Ab 24 h) % remaining Axl 1613F12 No Ab 938 514 45.2 1613F12 424 9G4 No Ab 11 ?2 117 9G4 13 MAB154 No Ab 950 ND ND 9G4 ND

(92) The data demonstrate that the mean fluorescence intensity detected on the surface of the cells treated with 1613F12 for 24 hours is reduced (?514) compared to the MFIs obtained with untreated cells labelled with the 1613F12. After a 24 h incubation with the 1613F12 antibody, 45.2% of the cell-surface Axl receptor remains at the SN12C cell-surface.

(93) Data obtained with other anti-Axl antibodies are presented in Table 13b.

(94) TABLE-US-00014 TABLE 13b Antibody internalization study by flow cytometry usnig SN12C cells ?MFI 1A4 497 11D12 438 320G10 355 427B7 256 517F1 303 530C8 334 534G9 257 535F7 352 547A1 473 1023F10 380 1614G5 305

Example 10: Anti-Axl Antibodies Internalization Study Using Fluorescent Immunocytochemistry Labelling

(95) Complementary internalization results are obtained by confocal microscopy using indirect fluorescent labelling method.

(96) Briefly, SN12C tumor cell line was cultured in RMPI1640 with 1% L-glutamine and 10% of FCS for 3 days before experiment. Cells were then detached using trypsin and plated in 6-multiwell plate containing coverslide in RPMI1640 with 1% L-glutamine and 5% FCS. The next day, the anti-Axl antibody was added at 10 ?g/ml. Cells treated with an irrelevant antibody were also included. The cells were then incubated for 1 h and 2 h at 37? C., 5% CO.sub.2. For T 0 h, cells were incubated for 30 minutes at 4? C. to determine antibody binding on cell surface. Cells were washed with PBS and fixed with paraformaldehyde for 15 minutes. Cells were rinsed and incubated with a goat anti-mouse IgG Alexa 488 antibody for 60 minutes at 4? C. to identify remaining antibody on the cell surface. To follow antibody penetration into the cells, cells were fixed and permeabilized with saponin. A goat anti-mouse IgG Alexa 488 (Invitrogen) was used to stained both the membrane and the intracellular antibody. Early endosomes were identified using a rabbit polyclonal antibody against EEA1 revealed with a goat anti-rabbit IgG-Alexa 555 antibody (Invitrogen). Cells were washed three times and nuclei were stained using Draq5. After staining, cells were mounted in Prolong Gold mounting medium (Invitrogen) and analyzed by using a Zeiss LSM 510 confocal microscope.

(97) Photographs are presented in FIGS. 9A-9C.

(98) Images were obtained by confocal microscopy. In presence of the mIgG1 isotype control (9G4), neither membrane staining nor intracellular labelling is observed (FIG. 9A). A progressive loss of the membrane anti-Axl labelling is observed as soon as after 1 h incubation of the SN12C cells with the 1613F12 (FIG. 9B). Intracellular accumulation of the 1613F12 Axl antibody is clearly observed at 1 h and 2 h (FIG. 9C). Intracellular antibody co-localizes with EEA1, an early endosome marker. These photographs confirm the internalization of the 1613F12 into SN12C cells.

Example 11: In Vitro Anti-Axl Mediated Anti-Tumoral Activity

(99) SN12C Proliferation Assay

(100) Ten thousand SN12C cells per well were seeded in FCS-free medium on 96 well plates over night at 37? C. in a 5% CO.sub.2 atmosphere. The next day, cells were pre-incubated with 10 ?g/ml of each antibody for 1 h at 37? C. Cells were treated with or without rmGas6 (R and D Systems, cat No. 986-GS/CF), by adding the ligand directly to the well, and then left to grown for 72 h. Proliferation was measured following .sup.3H thymidine incorporation.

(101) Data are presented in FIG. 10. No effect was observed with the 1613F12 which is silent when added to SN12C cells.

Example 12: Cytotoxicity Potency of 1613F12-Saporin and of 1A4-Saporin Immunoconjugates in Various Human Tumor Cell Lines

(102) In the present example, is first documented the cytotoxicity potency of the saporin coupled-1613F12. For this purpose direct in vitro cytotoxicity assays using a large panel of human tumor cell lines were performed (FIGS. 11A-11K). This tumor cell line panel offers various Axl expression levels. Secondly, is evaluated another anti-Axl targeting immunoconjugate, 1A4-saporin for its direct cytotoxic effect on human SN12C renal tumor cells (FIG. 13).

(103) Briefly, 5000 cells were seeded in 96 well culture plates in 100 ?l of 5% FBS adequate culture medium. After 24 hour incubation in a 5% CO.sub.2 atmosphere at 37? C., a range of concentrations of the anti-Axl antibody or the immunoconjugate (anti-Axl antibody-saporin or 9G4-saporin or the naked anti-Axl antibody or 9G4) is applied to the cells. The immunoconjugates were used in a large range of concentrations. Culture plates are then incubated at 37? C. in a humidified 5% CO.sub.2 incubator for 72 hours.

(104) At D4, the cell viability is assessed using the CellTiter-Glo? Luminescent Cell Viability kit (Promega Corp., Madison, Wis.) that allows determining the number of viable cells in culture based on quantification of the ATP present, an indicator of metabolically active cells. Luminescent emissions are recorded by a luminometer device.

(105) From luminescence output is calculated the percentage of cytotoxicity using the following formula:
% cytotoxicity=100?[(RLU.sub.Ab-sap?100)/RLU.sub.No Ab]

(106) On FIGS. 11A-11K are put together graphs presenting cytotoxicity percentage in function of the immunoconjugate concentration obtained in distinct in vitro cell cytotoxicity assays with (A) SN12C, (B) Calu-1, (C) A172, (D) A431, (E) DU145, (F) MDA-MB-435S, (G) MDA-MB-231, (H) PC3, (I) NCI-H226, (J) NCI-H125 or (K) Panc1 tumor cells treated with a range of 1613F12-saporin immunoconjugate concentrations.

(107) FIGS. 11A-11K shows that the 1613F12-saporin immunoconjugate triggered cytotoxicity in these different human tumor cell lines. The potency of the resulting cytotoxicity effect depends on the human tumor cell line.

(108) FIG. 13 presents the cytotoxicity activity triggered by the 1A4-saporin immunoconjugate using SN12C human renal cells. This Axl-targeting immunoconjugate exhibited a dose-dependent cytotoxic effect with a maximum cytotoxic effect of 71%,

Example 13: Binding Kinetics of Axl Antibodies to Human Axl ECD

(109) Affinity measurement of the 1613F12 was then determined using Biacore.

(110) A Biacore X is used to measure the binding kinetics of Axl antibodies on human Axl ECD.

(111) The instrument based on the optical phenomenon of surface plasmon resonance (SPR) used by Biacore systems enables the detection and measurement of protein-protein interactions in real time, without the use of labels.

(112) Briefly, the experiments were realized using a sensor chip CM5 as the biosensor. Rabbit IgGs were immobilized on the flow cells 1 and 2 (FC1 and FC2) of a CM5 sensor chip at a level of 9300-10000 response units (RU) using amine coupling chemistry to capture antibodies.

(113) Binding is evaluated using multiple cycles. Each cycle of measure is performed using a flow rate of 30 ?l/min in a HBS-EP buffer. Then the Axl antibody to be tested is captured on the chip for 1 min on FC2 only to reach a mean capture value of 311.8 RU (SD=5.1 RU) for the 1613F12 antibody. The analyte (Axl ECD antigen) is injected starting at 200 nM and using two-fold serial dilutions to measure rough ka and kd in real time.

(114) At the end of each cycle, the surfaces are regenerated by injecting a 10 mM glycine hydrochloride pH1.5 solution to eliminate the antibody-antigen complexes and the capture antibody as well. The considered signal corresponds to the difference of the signals observed between FC1 and FC2 (FC2?FC1). Association rates (ka) and dissociation rates (kd) were calculated using a one-to-one Langmuir binding model. The equilibrium dissociation constant (KD) is determined as the ka/kd ratio. The experimental values were analyzed in the Biaevaluation software version 3.0. A ?2 analysis will be performed to assess the accuracy of the data.

(115) Data are summarized in the following Table 14.

(116) TABLE-US-00015 TABLE 14 Binding kinetics and affinity of 1613F12 to human Axl ECD Antibody Ka (1/Ms) Kd (1/s) KD (M) Chi.sup.2 1613F12 1.06 10.sup.5 2.42 10.sup.?4 2.29 10.sup.?9 0.71 (0.6%)

(117) To produce the human extracellular domain (ECD) of Axl, the human cDNAs coding for the human soluble AXL receptor was first cloned into the pCEP4 expression vector by PCR. The purified product was then digested with restriction enzymes HindIII and BamHI and ligated into pCEP4 expression vector which had been precut with the same enzymes. Finally, the identified recombinant plasmid pCEP[AXL]His.sub.6 was further confirmed by DNA sequencing.

(118) Then suspension adapted cells HEK293E were cultivated in Ex-cell 293 (SAFC Biosciences) medium with 4 mM glutamine. All transfections were performed using linear 25 kDa polyethyleneimine (PEI). The transfected cultured were maintained at 37? C. in an incubateur shaker with 5% CO2 and with agitation at 120 rpm for 6 days. The cells were collected by centrifugation, and the supernatant containing the recombinant His-tagged protein was treated for purification on a Ni-NTA agarose column.