ANTI-WT1/HLA-SPECIFIC ANTIBODIES

Abstract

The present invention relates to antibodies or fragments thereof binding to human WT1/HLA. In particular, the present invention relates to antibodies or fragments thereof that have combined improved and/or beneficial properties, and are therefore suited for clinical development.

Claims

1. An isolated antibody comprising (a) a heavy chain comprising an HCDR1 region consisting of SEQ ID NO: 17, an HCDR2 region consisting of SEQ ID NO: 18, an HCDR3 region consisting of SEQ ID NO: 19, and a light chain comprising a LCDR1 region consisting of SEQ ID NO: 20, a LCDR2 region consisting of SEQ ID NO: 21, and a LCDR3 region consisting of SEQ ID NO: 22, or (b) a heavy chain comprising an HCDR1 region consisting of SEQ ID NO: 27, an HCDR2 region consisting of SEQ ID NO: 28, an HCDR3 region consisting of SEQ ID NO: 29, and a light chain comprising a LCDR1 region consisting of SEQ ID NO: 30, a LCDR2 region consisting of SEQ ID NO: 31, and a LCDR3 region consisting of SEQ ID NO: 32, or (c) a heavy chain comprising an HCDR1 region consisting of SEQ ID NO: 37, an HCDR2 region consisting of SEQ ID NO: 38, an HCDR3 region consisting of SEQ ID NO: 39, and a light chain comprising a LCDR1 region consisting of SEQ ID NO: 40, a LCDR2 region consisting of SEQ ID NO: 41, and a LCDR3 region consisting of SEQ ID NO: 42, or (d) a heavy chain comprising an HCDR1 region consisting of SEQ ID NO: 47, an HCDR2 region consisting of SEQ ID NO: 48, an HCDR3 region consisting of SEQ ID NO: 49, and a light chain comprising a LCDR1 region consisting of SEQ ID NO: 50, a LCDR2 region consisting of SEQ ID NO: 51, and a LCDR3 region consisting of SEQ ID NO: 52, or (e) a heavy chain comprising an HCDR1 region consisting of SEQ ID NO: 57, an HCDR2 region consisting of SEQ ID NO: 58, an HCDR3 region consisting of SEQ ID NO: 59, and a light chain comprising a LCDR1 region consisting of SEQ ID NO: 60, a LCDR2 region consisting of SEQ ID NO: 61, and a LCDR3 region consisting of SEQ ID NO: 62, wherein said antibody specifically binds to a complex of HLA-A0201/RMFPNAPYL (SEQ ID NO: 1).

2. The isolated antibody of claim 1, wherein (a) the heavy chain comprises a heavy chain variable region (VH) comprising at least 95% sequence identity to SEQ ID NO: 15, and the light chain comprises a light chain variable region (VL) comprising at least 95% sequence identity to SEQ ID NO: 16, or (b) the heavy chain comprises a heavy chain variable region (VH) comprising at least 95% sequence identity to SEQ ID NO: 25, and the light chain comprises a light chain variable region (VL) comprising at least 95% sequence identity to SEQ ID NO: 26, or (c) the heavy chain comprises a heavy chain variable region (VH) comprising at least 95% sequence identity to SEQ ID NO: 35, and the light chain comprises a light chain variable region (VL) comprising at least 95% sequence identity to SEQ ID NO: 36, or (d) the heavy chain comprises a heavy chain variable region (VH) comprising at least 95% sequence identity to SEQ ID NO: 45, and the light chain comprises a light chain variable region (VL) comprising at least 95% sequence identity to SEQ ID NO: 46, or (e) the heavy chain comprises a heavy chain variable region (VH) comprising at least 95% sequence identity to SEQ ID NO: 55, and the light chain comprises a light chain variable region (VL) comprising at least 95% sequence identity to SEQ ID NO: 56.

3. The isolated antibody of claim 1, wherein (a) the heavy chain comprises a heavy chain variable region (VH) comprising SEQ ID NO: 15, and the light chain comprises a light chain variable region (VL) comprising SEQ ID NO: 16, or (b) the heavy chain comprises a heavy chain variable region (VH) comprising SEQ ID NO: 25, and the light chain comprises a light chain variable region (VL) comprising SEQ ID NO: 26, or (c) the heavy chain comprises a heavy chain variable region (VH) comprising SEQ ID NO: 35, and the light chain comprises a light chain variable region (VL) SEQ ID NO: 36, or (d) the heavy chain comprises a heavy chain variable region (VH) comprising SEQ ID NO: 45, and the light chain comprises a light chain variable region (VL) comprising SEQ ID NO: 46, or (e) the heavy chain comprises a heavy chain variable region (VH) comprising SEQ ID NO: 55, and the light chain comprises a light chain variable region (VL) comprising SEQ ID NO: 56.

4. The antibody of claim 1, wherein the antibody is in an Fab format.

5. The antibody of claim 4, wherein the antibody has an EC50 of less than 10 nM.

6. The antibody of claim 1, wherein the antibody is in an immunoglobulin format.

7. The antibody of claim 4, wherein the antibody has an EC50 of less than 10 nM.

8. The antibody of claim 1, wherein the antibody binds to the complex of HLA-A0201/RMFPNAPYL (SEQ ID NO: 1) with a dissociation rate of less than 0.5 nM.

9. The antibody of claim 1, wherein the antibody is humanized.

10. The antibody of claim 1, wherein said antibody binds to a complex of a peptide consisting of RMFPNAPYL (SEQ ID NO: 1) with HLA-A0201 with an EC50 that is at least 10 times lower than the EC50 for a complex of a peptide consisting of RMFPGEVAL (SEQ ID NO: 64) with HLA-A0201.

11. The antibody of claim 1, wherein said antibody does not bind a complex of a peptide consisting of RMFPGEVAL (SEQ ID NO: 64) with HLA-A0201.

12. The antibody of claim 1, wherein said antibody has an EC50 of less than 10 nM in a Fab format and in an immunoglobulin format.

13. The antibody of claim 1, further comprising a cytotoxic moiety linked thereto.

14. A pharmaceutical composition comprising the antibody according to claim 1, and a pharmaceutically acceptable carrier and/or excipient.

15. An isolated antibody or an antigen-binding fragment thereof which specifically binds to a complex of HLA-A0201/RMFPNAPYL (SEQ ID NO: 1), wherein said antibody or antigen-binding fragment thereof comprises: (a) a heavy chain variable region (VH) comprising CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 15, and a light chain variable region (VL) comprising CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 16, or (b) a heavy chain variable region (VH) comprising CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 25, and a light chain variable region (VL) comprising CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 26, or (c) a heavy chain variable region (VH) comprising CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 35, and a light chain variable region (VL) comprising CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 36, or (d) a heavy chain variable region (VH) comprising CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 45, and a light chain variable region (VL) comprising CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 46, or (e) a heavy chain variable region (VH) comprising CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 55, and a light chain variable region (VL) comprising CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 56.

16. The isolated antibody or an antigen-binding fragment thereof of claim 15, wherein (a) the CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 15 comprise the amino acid sequences of SEQ ID NO: 17, 18, and 19, respectively, and the CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 16 comprise the amino acid sequences of SEQ ID NO: 20, 21, and 22, respectively, or (b) the CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 25 comprise the amino acid sequences of SEQ ID NO: 27, 28, and 29, respectively, and the CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 26 comprise the amino acid sequences of SEQ ID NO: 30, 31, and 32, respectively, or (c) the CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 35 comprise the amino acid sequences of SEQ ID NO: 37, 38, and 39, respectively, and the CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 36 comprise the amino acid sequences of SEQ ID NO: 40, 41, and 42, respectively, or (d) the CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 45 comprise the amino acid sequences of SEQ ID NO: 47, 48, and 49, respectively, and the CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 46 comprise the amino acid sequences of SEQ ID NO: 50, 51, and 52, respectively, or (e) the CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 55 comprise the amino acid sequences of SEQ ID NO: 57, 58, and 59, respectively, and the CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 56 comprise the amino acid sequences of SEQ ID NO: 60, 61, and 62, respectively.

17. The isolated antibody or an antigen-binding fragment thereof of claim 15, wherein (a) the CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 15 consist of the amino acid sequences of SEQ ID NO: 17, 18, and 19, respectively, and the CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 16 consist of the amino acid sequences of SEQ ID NO: 20, 21, and 22, respectively, or (b) the CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 25 consist of the amino acid sequences of SEQ ID NO: 27, 28, and 29, respectively, and the CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 26 consist of the amino acid sequences of SEQ ID NO: 30, 31, and 32, respectively, or (c) the CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 35 consist of the amino acid sequences of SEQ ID NO: 37, 38, and 39, respectively, and the CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 36 consist of the amino acid sequences of SEQ ID NO: 40, 41, and 42, respectively, or (d) the CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 45 consist of the amino acid sequences of SEQ ID NO: 47, 48, and 49, respectively, and the CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 46 consist of the amino acid sequences of SEQ ID NO: 50, 51, and 52, respectively, or (e) the CDR regions HCDR1, HCDR2 and HCDR3 of SEQ ID NO: 55 consist of the amino acid sequences of SEQ ID NO: 57, 58, and 59, respectively, and the CDR regions LCDR1, LCDR2, and LCDR3 of SEQ ID NO: 56 consist of the amino acid sequences of SEQ ID NO: 60, 61, and 62, respectively.

18. The isolated antibody or an antigen-binding fragment thereof of claim 15, wherein the antibody or an antigen-binding fragment thereof is in an Fab format.

19. The isolated antibody or an antigen-binding fragment thereof of claim 15, wherein the antibody or an antigen-binding fragment thereof is in an immunoglobulin format.

20. A method of treating a patent who has cancer that presents RMFPNAPYL (SEQ ID NO: 1) in a complex with HLA-A0201, comprising administering to the patient the antibody of claim 1, wherein the cancer is selected from the group consisting of chronic myelocytic leukemia, multiple myeloma, acute lymphoblastic leukemia, acute myeloid/myelogenous leukemia, myelodysplastic syndrome, mesothelioma, ovarian cancer, gastrointestinal cancers, breast cancer, prostate cancer, and glioblastoma.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0125] The present invention shall now be explained in the following examples with reference to the figures, nevertheless, without being limited thereto. For the purpose of the present invention, all references as cited herein are incorporated by reference in their entireties.

[0126] FIG. 1 shows the binding of six reference antibodies to the RMF/HLA complex.

[0127] FIG. 2 shows the binding of six reference antibodies to the RHAMM/HLA complex.

[0128] FIG. 3 shows the binding of six reference antibodies to the PIGQ/HLA complex.

[0129] FIG. 4 shows the binding of the antibodies of the present invention to the RMF/HLA complex.

[0130] FIG. 5 shows the binding of the antibodies of the present invention to the RHAMM/HLA complex.

[0131] FIG. 6 shows the binding of the antibodies of the present disclosure to the PIGQ/HLA complex.

EXAMPLES

Example 1: Generation of Antigens

[0132] The RMF/HLA complex was used as a target for antibody selections. The target was produced as described in Dao et al. (Sci Transl Med. 2013 Mar. 13; 5(176): 176ra33). Biotinylated WT1/HLA-A0201 and RHAMM-R3/HLA_A0201 complexes were synthesized by refolding the peptides with recombinant HLA-A2 and 2-microglobulin within the Immunology Department (as described in Altman, et al. 1996. Phenotypic analysis of antigen specific T lymphocytes. Science 274:94, and Jung, G., Ledbetter, J. A., and Mller-Eberhard, H. J. (1987). Induction of cytotoxicity in resting human T lymphocytes bound to tumor cells by antibody heteroconjugates. Proceedings of the National Academy of Sciences of the United States of America, 84(13), 4611-4615, with small modifications).

[0133] As a negative control, the RHAMM/HLA complex was generated (Dao et al., 2013). The RHAMM peptide has an amino sequence of ILSLELMKL (SED ID NO: 63). The RHAMM peptide was complexed to HLA-A0201.

[0134] As another negative control, a PIGQ/HLA complex was generated. The PIGQ peptide was identified via the XPRESIDENT (immatics biotechnologies GmbH, Germany) target identification platform (amino acid sequence of RMFPGEVAL (SED ID NO: 64)). It was also complexed to HLA-A0201. The PIGQ peptide occurs ubiquitously in healthy human tissue. Binding to the PIGQ/HLA complex is therefore highly undesirable. Five out of nine amino acids of the PIGQ peptide are identical to the WT1 peptide.

[0135] All antigens were produced in biotinylated and in non-biotinylated form. Purity (>95%) and monomer portion (>97-99%) of all peptides and HLA antigens was confirmed via SDS-PAGE, analytical size exclusion chromatography and dynamic light scattering.

Example 2: Generation of Reference Antibodies

[0136] Six RMF/HLA complex-binding antibodies are described on WO2012/135854 (Memorial Sloan-Kettering). Five of these six antibodies are also described in the Dao et al. referenced (see above). Only Refmab #18 (see below) is not described in Dao et al. (2013).

[0137] The sequences of these six antibodies (Refmab #3, Refmab #5, Refmab #13, Refmab #15, Refmab #18 and Refmab #23) were cloned, and human IgG1f antibodies were produced in HKB11 cells. Antibodies were purified and subjected to quality control (yield, concentration, purity and monomer content).

Example 3: Binding of Reference Antibodies to Isolated Antigens

[0138] In this experiment, the specificities of the reference antibodies were investigated. Biotinylated peptide/HLA complexes were coated on Neutravidin plates at 50 nM. The isolated reference antibodies of Example 2 were titrated on the antigens: Binding was tested in ELISA utilizing an AP-conjugated goat anti-human IgG secondary antibody.

[0139] All reference antibodies bound to the RMF/HLA complex. Refmab #23 also unspecifically bound to the RHAMM-HLA complex, as has been described in the literature.

[0140] Interestingly all reference antibodies except for Refmab #18 also bound to the PIGQ/HLA complex, i.e. a peptide presented on healthy human tissue. Only Refmab #18 was specific for the WT1-HLA complex in ELISA. Results are shown in FIGS. 1-3.

Example 4: Binding of Reference Antibodies to Antigens on Cells

[0141] Since binding to an isolated antigen does not necessarily coincide with binding to the antigen when presented on intact cells it was investigated whether or not the control antibodies also bound to the RMF/HLA complex on cells expressing this antigen.

[0142] SET2 cells (DSMZ No. ACC 608) and BV173 cells (DSMZ No. ACC 20) were used as antigen-positive cancer cells (see Dao et al.; 2013). Binding was measured by flow cytometry utilizing a PE-conjugated goat anti-human IgG secondary antibody, and EC50 values were determined. Results are shown in Table 2. ++ designates binding with an EC50 of below 10 nM, + designates binding with an EC50 of more than 10 nM, and designates no binding to the target on cells.

TABLE-US-00003 TABLE 2 Binding of the reference antibodies (IgG) to cells EC50 on SET2 EC50 on BV173 cells cells Binder [nM] [nM] Refmab #3 Refmab #5 + + Refmab #13 ++ ++ Refmab #15 + + Refmab #18 Refmab #23 + +

[0143] As can be seen in Table 2, two out of the six reference antibodies do not bind to cells expressing the target. These antibodies (including Refmab #18) are therefore no suitable for therapeutically development.

Example 5: Identification of Superior RMF/HLA Complex Binders

[0144] For antibody generation the Ylanthia library (MorphoSys AG, Germany) was used to select Fab fragments against the WT1/HLA complex. The Ylanthia library (Tiller et al. mAbs 5:3, 1-26; May/June (2013) and U.S. Pat. No. 8,728,981) is a commercially available phagemid library and employs the CysDisplay technology for displaying the Fab on the phage surface (Lohning et al., WO 2001/05950).

[0145] In order to isolate RMF/HLA complex-specific antibodies, different panning strategies were used (solution panning, plated-based panning). Each panning strategy comprised at least 3 individual rounds of panning against the RMF/HLA complex. The selection of unspecific binders was inhibited by pre-blocking with the PIGQ/HLA complex as a counter target.

[0146] The isolated binders were subjected to primary screening on an Intellicyt HTFC Screening System utilizing fluorescent beads. Three different antigens (WT1/HLA complex, RHAMM/HLA complex and PIGQ/HLA complex) were tested in parallel. Hits that were positive on the WT1-HLA complex antigen but negative on the other two antigens were isolated and subjected to a secondary screening using a more stringent ELISA assay. A selection of positive clones was converted into IgG format. Six of the most promising candidates were purified and characterized further. The sequences of the six binders are shown in Table 1.

[0147] The binders were further subjected to in-depth characterization. They were also directly compared to the antibodies of the prior art. The EC50 of the binders was measured in a monovalent Fab format, and as full-length immunoglobulins (IgG1).

Example 6: Characterization of the Binders in the Fab Format

[0148] The binders were subjected to an in-depth characterization. They were directly compared to the antibodies of the prior art in an ELISA assay using Neutravidin plates as described in Example 3. For selected reference antibodies and binders of the present invention, the EC50 were determined. Results are summarized in Table 3. ++ designates binding with an EC50 of below 10 nM, + designates binding with an EC50 of more than 10 nM, and designates no binding.

[0149] The EC50 values on RMF/HLA are significantly lower (better) than the EC50 values of the reference antibodies of the prior art. All binders tested demonstrated an EC50 which was at least 3.3-fold better than that of the best binder of the prior art. Three out of the four binders tested demonstrated an EC50 which was at least 5.5-fold better than that of the best binder of the prior art. Two out of the four binders tested demonstrated an EC50 which was at least 7.5-fold better than that of the best binder in the prior art. Furthermore, none of the binders of the present invention show any binding to the counter targets, in particular no binding to the PIGQ/HLA complex. Some degree of binding was observed for some of the reference antibodies.

[0150] Affinities of the Fab fragments of binders Aali and Refmab #13 for the antigen RMF/HLA were also measured with an Octet system using the antigen immobilized to Streptavidin. The K.sub.D values were determined as 66 nM for Aali and 590 nM for Refmab #13, also confirming that the antibodies of the present invention have a higher affinity.

TABLE-US-00004 TABLE 3 Summary of Fab characterization EC50 on EC50 on EC50 on RMF/HLA RHAMM/HLA PIGQ/HLA Binder [nM] [nM] [nM] Aali ++ Cyprus ++ Daniel ++ Fiwi + Refmab #3 + + Refmab #13 + + Refmab #18 + Refmab #23 +

Example 7: Specificity of the Binders in an Immunoglobulin Format

[0151] In this experiment the specificities of the antibodies of the present disclosure were investigated as described in Example 3. Results are shown in FIGS. 4-6.

[0152] It was found that all binders are highly specific for the RMF/HLA complex. None of the binders showed any cross-reactivity with either the RHAMM-HLA complex or the PIGQ/HLA complex. This is in striking contrast to the antibodies of the prior art (see Example 3), which all (with the exception of Refmab #18), showed cross-reactivity with a peptide/HLA-complex (PIGQ/HLA) that is expressed on healthy human tissue.

Example 8: Affinities and EC50's of the Binders in an Immunoglobulin Format

[0153] EC50 values were determined for the binders of the present disclosure and the binders of the prior art (ELISAs were performed on Neutravidin plates). Results are summarized in Table 4. ++ designates binding with an EC50 of below 10 nM, + designates binding with an EC50 of more than 10 nM, and designates no binding.

TABLE-US-00005 TABLE 4 Summary of EC50's of the binders (IgG) of the present invention EC50 on EC50 on EC50 on RMF/HLA RHAMM/HLA PIGQ/HLA Binder [nM] [nM] [nM] Aali ++ Bibi ++ Cyprus ++ Daniel ++ Elfu ++ Fiwi ++ Refmab #3 ++ + Refmab #5 ++ + + Refmab #13 ++ ++ Refmab #15 ++ ++ Refmab #18 ++ Refmab #23 ++ + +

[0154] All binders (the binders of the present invention, as well as the binders of the prior art) showed very low EC50 values towards the RMF/HLA complex (with the exception of Refmab #3 and Refmab #5 all EC50's were in fact <1 nM). However, all binders of the prior art (except Refmab #18) also showed binding to the PIGQ/HLA counter antigens. Two binders of the prior art even showed binding to the RHAMM/HLA counter antigen.

[0155] The apparent affinity of the binders Aali and Refmab #13 was measured for the RMF/HLA antigen with an Octet system. The apparent affinity value for Refmab #13 was determined as 8.2 nM. The binding of Aali was characterized by a very slow dissociation rate, which exceeded the specification limits that were set in the system. Compared to Refmab #13 (8.2 nM), the apparent affinity of Aali is therefore <0.5 nM.

Example 9: Binding of the Antibodies of the Present Invention to Cells

[0156] Example 4 was repeated with the antibodies of the present disclosure, i.e. binding to RNF/HLA-positive SET2 cells (DSMZ No. ACC 608) and BV173 cells (DSMZ No. ACC 20) was tested by flow cytometry. EC50 values were determined. Results are shown in Table 5. ++ designates binding with an EC50 of below 10 nM, + designates binding with an EC50 of more than 10 nM, and designates no binding to the target on cells.

TABLE-US-00006 TABLE 5 Summary of IgG characterization by flow cytometry EC50 on SET2 cells EC50 on BV173 cells Binder [nM] [nM] Aali ++ + Bibi + + Cyprus + + Daniel ++ + Elfu + + Fiwi + +

[0157] All antibodies of the present invention show binding to WT1/HLA-expressing cells. This was also the case for most (4 out of 6) of the prior art antibodies, but most notably not for Refmab #18, the only reference antibody that does not bind to PIGQ/HLA.

[0158] The antibodies and binders of the present invention are therefore characterized in that they specifically bind to the RMF/HLA-complex without demonstrating any cross-reactivity to the PIGQ/HLA complex. In addition, the antibodies of the present disclosure also do bind to WT1/HLA expressing cell lines.

Example 7: Binding to Leukemic Patient Samples

[0159] The binders of the present invention are investigated for their ability to detect the RMF epitope on primary AML cells. The binders are expected to bind to AML blasts of patients. Results are confirmed by flow cytometry analysis. Results confirm that the level of RMF/HLA-A0201 on the surface of leukemia cells is adequate to allow for a reactivity with the binders of the present invention. The results also confirm that the levels of the target molecule on negative healthy cells are insignificant.

Example 8: Mediation of ADCC Activity

[0160] The binders of the present invention are also investigated for their potential to mediate ADCC, one of the major effector mechanisms of therapeutic antibodies in humans. In the presence of human PBMCs, the binders mediate a dose-dependent PBMC ADCC against T2 cells (an antigen-processing-deficient cell line, see, for example, WO 2012/135854) loaded with RMF peptide, but not T2 cells alone or T2 cells pulsed with a control peptide. The binders are also able to mediate ADCC against naturally presented RMF epitope by HLA-A0201 molecule on tumor cells, such as the mesothelioma cell line, JMN and the leukemia cell line BV173, but not to HLA-A2 negative cells, such as MSTO or HL-60. These results demonstrate that the binders mediates specific ADCC against cells that naturally express RMF and HLA-A0201 complex at physiologic levels as well as on cell lines.

Example 9: Elimination of Human Leukemia Cells in NSG Mice

[0161] The binders of the present invention are further investigated in an in vivo NOD SCID gamma (NSG) mice xenograft model. Mice are xenografted intravenously 6 days previously with BV173 bcr/abl positive acute lymphoblastic leukemia. At the time of treatment, mice developed leukemia in their liver, spleen, and BM as visible by luciferase imaging. The binders of the present invention dramatically reduce the tumor burden for at least 30 days. Results are confirmed by titrating the dose of the antibody.