Anti-steap2 antibodies, antibody-drug conjugates, and bispecific binding molecules that bind steap2 and CD3, as well as uses thereof

Abstract

To provide antibodies and antigen binding fragments thereof that are specific for six-transmembrane epithelial antigen of prostate 2 (STEAP2) and use methods thereof as well as bispecific antigen binding molecules that bind STEAP2 and CD3, and use methods thereof.SOLUTION: Provided herein is an antibody-drug conjugate (ADC) comprising an anti-STEAP2 antibody or antigen binding fragment thereof and a cytotoxic agent, where the antibody or antigen binding fragment and the cytotoxic agent are covalently bonded by a linker.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] Sequence Listing Reference This application incorporates by reference the Sequence Listing, which was filed in computer readable format as file 10296WO01-Sequence.txt, created on September 22, 2017, and contains 739,964 bytes.

[0002] The present invention relates to antibodies and antigen-binding fragments specific for six-transmembrane epithelial antigen 2 of the prostate (STEAP2), and methods of using them. The present invention also relates to bispecific antigen-binding molecules that bind STEAP2 and CD3, and methods of using them. The present invention further relates to antibody-drug conjugates comprising an anti-STEAP2 antibody or fragment thereof and a therapeutic agent (e.g., a cytotoxic agent). [Background technology]

[0003] Prostate six-transmembrane epithelial antigen 2 (STEAP2), also known as STEAP-2, metalloreductase STEAP2, prostate cancer-associated protein 1, protein upregulated in metastatic prostate cancer, six-transmembrane prostate protein 1 (STAMP1), and 098P4B6, is an integral six-transmembrane protein upregulated in normal and malignant prostate cells. Functioning as a shuttle between the Golgi complex and the plasma membrane, STEAP2 is a metalloreductase that reduces iron and copper, facilitating their entry into the cell. STEAP2 is primarily localized in prostate epithelial cells. STEAP2 is also expressed in normal heart, brain, pancreas, ovary, skeletal muscle, mammary gland, testis, uterus, kidney, lung, trachea, colon, and liver. STEAP2 is overexpressed in cancerous tissues including prostate tumors, bladder tumors, cervical tumors, lung tumors, colon tumors, kidney tumors, breast tumors, pancreatic tumors, gastric tumors, uterine tumors, and ovarian tumors (Non-Patent Document 1, Patent Document 1, Patent Document 2).

[0004] CD3 is a homodimeric or heterodimeric antigen expressed on T cells in conjunction with the T cell receptor complex (TCR) and is required for T cell activation. Functional CD3 is formed from the dimeric association of two of four distinct chains: epsilon, zeta, delta, and gamma. CD3 dimeric configurations include gamma / epsilon, delta / epsilon, and zeta / zeta. Antibodies against CD3 have been shown to cluster CD3 on T cells, thereby triggering T cell activation in a manner similar to TCR engagement by peptide-loaded MHC molecules. Therefore, anti-CD3 antibodies have been proposed for therapeutic purposes, including T cell activation. Furthermore, bispecific antibodies capable of binding to CD3 and a target antigen have been proposed for therapeutic uses, including targeting T cell immune responses to tissues and cells expressing the target antigen.

[0005] Antigen binding molecules targeting STEAP2, including antibody-drug conjugates, and bispecific antigen binding molecules that bind to both STEAP2 and CD3, are useful in therapeutic settings where specific targeting and T cell-mediated killing of cells expressing STEAP2 is desired. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Challita-Eid-PM,et al.,2003,WO03 / 087306 [Patent Document 2] Emtage, PCR, 2005, WO2005 / 079490 [Non-patent literature]

[0007] [Non-Patent Document 1] Gomes,IMet al.,2012,Mol.Cancer Res.10:573-587 Summary of the Invention

[0008] In a first aspect, the present invention provides antibodies and antigen-binding fragments thereof that bind to human STEAP2. Antibodies according to this aspect of the invention are useful, inter alia, for targeting cells expressing STEAP2. The present invention also provides bispecific antibodies and antigen-binding fragments thereof that bind to human STEAP2 and human CD3. Bispecific antibodies according to this aspect of the invention are useful, inter alia, for targeting CD3-expressing T cells and stimulating T cell activation, e.g., in situations where T cell-mediated killing of cells expressing STEAP2 is beneficial or desirable. For example, bispecific antibodies can induce CD3-mediated T cell activation to specific STEAP2-expressing cells, such as prostate tumor cells.

[0009] Exemplary anti-STEAP2 antibodies of the present invention are listed in Tables 1 and 2 herein. Table 1 shows the amino acid sequence identifiers for the heavy chain variable region (HCVR) and light chain variable region (LCVR), as well as the heavy chain complementarity determining regions (HCDR1, HCDR2, and HCDR3) and light chain complementarity determining regions (LCDR1, LCDR2, and LCDR3) of exemplary anti-STEAP2 antibodies. Table 2 shows the sequence identifiers for nucleic acid molecules encoding the HCVR, LCVR, HCDR1, HCDR2, HCDR3, LCDR1, LCDR2, and LCDR3 of exemplary anti-STEAP2 antibodies.

[0010] The present invention provides antibodies or antigen-binding fragments thereof comprising an HCVR comprising an amino acid sequence selected from any of the HCVR amino acid sequences listed in Table 1, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0011] The present invention also provides an antibody or antigen-binding fragment thereof comprising an LCVR comprising an amino acid sequence selected from any of the LCVR amino acid sequences listed in Table 1, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0012] The present invention also provides antibodies or antigen-binding fragments thereof comprising an HCVR and LCVR amino acid sequence pair (HCVR / LCVR) comprising any of the HCVR amino acid sequences listed in Table 1 paired with any of the LCVR amino acid sequences listed in Table 1. According to certain embodiments, the present invention provides antibodies or antigen-binding fragments thereof comprising an HCVR / LCVR amino acid sequence pair contained in any of the exemplary anti-STEAP2 antibodies listed in Table 1. In certain embodiments, the HCVR / LCVR amino acid sequence pair is selected from the group consisting of SEQ ID NOs: 250 / 258 (e.g., H2M11162N).

[0013] The present invention also provides an antibody or antigen-binding fragment thereof comprising a heavy chain CDR1 (HCDR1) comprising an amino acid sequence selected from any of the HCDR1 amino acid sequences listed in Table 1, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

[0014] The present invention also provides a heavy chain CDR2 (HCDR2) comprising an amino acid sequence selected from any of the HCDR2 amino acid sequences listed in Table 1, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity. The present invention provides an antibody or antigen-binding fragment thereof comprising a nucleotide sequence (SEQ ID NO: 1) of the present invention, the nucleotide sequence (SEQ ID NO: 2), a nucleotide sequence (SEQ ID NO: 3), a nucleotide sequence (SEQ ID NO: 4), a nucleotide sequence (SEQ ID NO: 5), a nucleotide sequence (SEQ ID NO: 6), a nucleotide sequence (SEQ ID NO: 7), a nucleotide sequence (SEQ ID NO: 8), a nucleotide sequence (

[0015] The present invention also provides an antibody or antigen-binding fragment thereof comprising a heavy chain CDR3 (HCDR3) comprising an amino acid sequence selected from any of the HCDR3 amino acid sequences listed in Table 1, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

[0016] The present invention also provides an antibody or antigen-binding fragment thereof comprising a light chain CDR1 (LCDR1) comprising an amino acid sequence selected from any of the LCDR1 amino acid sequences listed in Table 1, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

[0017] The present invention also provides an antibody or antigen-binding fragment thereof comprising a light chain CDR2 (LCDR2) comprising an amino acid sequence selected from any of the LCDR2 amino acid sequences listed in Table 1, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

[0018] The present invention also provides an antibody or antigen-binding fragment thereof comprising a light chain CDR3 (LCDR3) comprising an amino acid sequence selected from any of the LCDR3 amino acid sequences listed in Table 1, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

[0019] The present invention also provides antibodies or antigen-binding fragments thereof comprising an HCDR3 and LCDR3 amino acid sequence pair (HCDR3 / LCDR3) comprising any of the HCDR3 amino acid sequences listed in Table 1 paired with any of the LCDR3 amino acid sequences listed in Table 1. According to certain embodiments, the present invention provides antibodies or antigen-binding fragments thereof comprising an HCDR3 / LCDR3 amino acid sequence pair contained in any of the exemplary anti-STEAP2 antibodies listed in Table 1. In certain embodiments, the HCDR3 / LCDR3 amino acid sequence pair is selected from the group consisting of SEQ ID NOs: 256 / 264 (e.g., H2M11162N).

[0020] The present invention also provides antibodies or antigen-binding fragments thereof comprising the set of six CDRs (i.e., HCDR1-HCDR2-HCDR3-LCDR1-LCDR2-LCDR3) contained in any of the exemplary anti-STEAP2 antibodies listed in Table 1. In certain embodiments, the HCDR1-HCDR2-HCDR3-LCDR1-LCDR2-LCDR3 amino acid sequence set is selected from the group consisting of SEQ ID NOs: 252-254-256-260-262-264 (e.g., H2M11162N).

[0021] In related embodiments, the present invention provides an antibody or antigen-binding fragment thereof comprising a set of six CDRs (i.e., HCDR1-HCDR2-HCDR3-LCDR1-LCDR2-LCDR3) contained within the HCVR / LCVR amino acid sequence pair defined by any of the exemplary anti-STEAP2 antibodies listed in Table 1. For example, the present invention includes an antibody or antigen-binding fragment thereof comprising the set of HCDR1-HCDR2-HCDR3-LCDR1-LCDR2-LCDR3 amino acid sequences contained within the HCVR / LCVR amino acid sequence pair selected from the group consisting of SEQ ID NOs: 250 / 258 (e.g., H2M11162N). Methods and techniques for identifying CDRs within HCVR and LCVR amino acid sequences are well known in the art and can be used to identify CDRs within specific HCVR and / or LCVR amino acid sequences disclosed herein. Exemplary rules that can be used to identify CDR boundaries include, for example, the Kabat definition, the Chothia definition, and the AbM definition. In general Generally, the Kabat definition is based on sequence variability, the Chothia definition is based on the location of structural loop regions, and the AbM definition is a compromise between the Kabat and Chothia approaches. See, e.g., Kabat, "Sequences of Proteins of Immunological Interest," National Institutes of Health, Bethesda, Md. (1991); Al-Lazikani et al., J. Mol. Biol. 273:927-948 (1997); and Martin et al., Proc. Natl. Acad. Sci. USA 86:9268-9272 (1989). Public databases are also available for identifying CDR sequences within antibodies.

[0022] The present invention also provides nucleic acid molecules encoding anti-STEAP2 antibodies or portions thereof. For example, the present invention provides nucleic acid molecules encoding any of the HCVR amino acid sequences listed in Table 1, and in certain embodiments, the nucleic acid molecule comprises a polynucleotide sequence selected from any of the HCVR nucleic acid sequences listed in Table 2, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0023] The present invention also provides nucleic acid molecules encoding any of the LCVR amino acid sequences listed in Table 1, and in certain embodiments, the nucleic acid molecule comprises a polynucleotide sequence selected from any of the LCVR nucleic acid sequences listed in Table 2, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0024] The present invention also provides nucleic acid molecules encoding any of the HCDR1 amino acid sequences listed in Table 1, and in certain embodiments, the nucleic acid molecule comprises a polynucleotide sequence selected from any of the HCDR1 nucleic acid sequences listed in Table 2, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0025] The present invention also provides nucleic acid molecules encoding any of the HCDR2 amino acid sequences listed in Table 1, and in certain embodiments, the nucleic acid molecule comprises a polynucleotide sequence selected from any of the HCDR2 nucleic acid sequences listed in Table 2, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0026] The present invention also provides nucleic acid molecules encoding any of the HCDR3 amino acid sequences listed in Table 1, and in certain embodiments, the nucleic acid molecule comprises a polynucleotide sequence selected from any of the HCDR3 nucleic acid sequences listed in Table 2, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0027] The present invention also provides nucleic acid molecules encoding any of the LCDR1 amino acid sequences listed in Table 1, and in certain embodiments, the nucleic acid molecule comprises a polynucleotide sequence selected from any of the LCDR1 nucleic acid sequences listed in Table 2, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0028] The present invention also provides nucleic acid molecules encoding any of the LCDR2 amino acid sequences listed in Table 1. In certain embodiments, the nucleic acid molecule has a polynucleotide sequence selected from, or at least 90% identical to, any of the LCDR2 nucleic acid sequences listed in Table 2. It includes substantially similar sequences having at least 95%, at least 98%, or at least 99% sequence identity.

[0029] The present invention also provides nucleic acid molecules encoding any of the LCDR3 amino acid sequences listed in Table 1, and in certain embodiments, the nucleic acid molecule comprises a polynucleotide sequence selected from any of the LCDR3 nucleic acid sequences listed in Table 2, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0030] The present invention also provides nucleic acid molecules encoding HCVRs, wherein the HCVRs comprise a set of three CDRs (i.e., HCDR1-HCDR2-HCDR3), and the HCDR1-HCDR2-HCDR3 amino acid sequence set is as defined by any of the exemplary anti-STEAP2 antibodies listed in Table 1.

[0031] The present invention provides a nucleic acid molecule encoding an LCVR, wherein the LCVR comprises a set of three CDRs (i.e., LCDR1-LCDR2-LCDR3), and the LCDR1-LCDR2-LCDR3 amino acid sequence set is as defined by any of the exemplary anti-STEAP2 antibodies listed in Table 1.

[0032] The present invention also provides nucleic acid molecules encoding both an HCVR and an LCVR, wherein the HCVR comprises the amino acid sequence of any of the HCVR amino acid sequences listed in Table 1, and the LCVR comprises the amino acid sequence of any of the LCVR amino acid sequences listed in Table 1. In certain embodiments, the nucleic acid molecule comprises a polynucleotide sequence selected from any of the HCVR nucleic acid sequences listed in Table 2, or a substantially similar sequence thereof, having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto, and a polynucleotide sequence selected from any of the LCVR nucleic acid sequences listed in Table 2, or a substantially similar sequence thereof, having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto. In certain embodiments according to this aspect of the invention, the nucleic acid molecule encodes an HCVR and LCVR, wherein both the HCVR and LCVR are derived from the same anti-STEAP2 antibody listed in Table 1.

[0033] The present invention also provides recombinant expression vectors capable of expressing polypeptides comprising the heavy chain variable region or light chain variable region of an anti-STEAP2 antibody. For example, the present invention includes recombinant expression vectors comprising any of the above-described nucleic acid molecules, i.e., nucleic acid molecules encoding any of the HCVR, LCVR, and / or CDR sequences listed in Table 1. Also included within the scope of the present invention are host cells into which such vectors have been introduced, as well as methods for producing antibodies or portions thereof by culturing the host cells under conditions that permit the production of the antibodies or antibody fragments, and methods for recovering the antibodies and antibody fragments so produced.

[0034] The present invention includes anti-STEAP2 antibodies with altered glycosylation patterns. In some embodiments, modifications to remove undesired glycosylation sites or antibodies lacking fucose moieties, for example, to enhance antibody-dependent cellular cytotoxicity (ADCC) function, may be useful (see Shield et al. (2002) JBC 277:26733). In other applications, modifications of galactosylation can be made to modify complement-dependent cytotoxicity (CDC).

[0035] In another aspect, the present invention provides a pharmaceutical composition comprising a recombinant human antibody or fragment thereof that specifically binds to STEAP2 and a pharmaceutically acceptable carrier. In another related aspect, the present invention provides a composition that is a combination of an anti-STEAP2 antibody and a second therapeutic agent. In one embodiment, the second therapeutic agent is any agent that is advantageously combined with an anti-STEAP2 antibody. Additional combination therapies and formulations that include the anti-STEAP2 antibodies of the invention are disclosed elsewhere herein.

[0036] In another aspect, the present invention provides a therapeutic method for targeting / killing STEAP2-expressing tumor cells using an anti-STEAP2 antibody of the present invention, the therapeutic method comprising administering to a subject in need thereof a therapeutically effective amount of a pharmaceutical composition comprising an anti-STEAP2 antibody of the present invention. In some cases, the anti-STEAP2 antibody (or antigen-binding fragment thereof) can be used to treat prostate cancer or may be modified to enhance cytotoxicity by methods including, but not limited to, an engineered Fc domain that increases ADCC (see, e.g., Shield et al. (2002) JBC 277:26733), radioimmunotherapy, antibody-drug conjugates, or other methods to enhance the efficiency of tumor resection.

[0037] The present invention also includes the use of an anti-STEAP2 antibody of the present invention in the manufacture of a medicament for the treatment of a disease or disorder (e.g., cancer) associated with or caused by STEAP2-expressing cells. In one aspect, the present invention relates to a compound comprising an anti-STEAP2 antibody or antigen-binding fragment, or a STEAP2xCD3 bispecific antibody disclosed herein, for use in medicine. In one aspect, the present invention relates to a compound comprising an antibody-drug conjugate (ADC) disclosed herein, for use in medicine.

[0038] In yet another aspect, the present invention provides monospecific anti-STEAP2 antibodies for diagnostic use, eg, as imaging reagents.

[0039] In yet another aspect, the invention provides therapeutic methods for stimulating T cell activation using an anti-CD3 antibody or antigen-binding portion of the antibody of the invention, the therapeutic method comprising administering a therapeutically effective amount of a pharmaceutical composition comprising the antibody.

[0040] In another embodiment, the invention provides an isolated antibody or antigen-binding fragment thereof that binds to STEAP2-expressing C4-2 cells with an EC50 of less than 50 nM as measured by FACS analysis. In another embodiment, the invention provides an isolated antibody or antigen-binding fragment thereof that binds to and is internalized by STEAP2-expressing C4-2 cells.

[0041] The present invention further provides antibodies or antigen-binding fragments that compete with a reference antibody comprising an HCVR / LCVR amino acid sequence pair listed in Table 1 for binding to human STEAP2. In another embodiment, the invention provides an antibody or antigen-binding fragment that competes for binding to human STEAP2 with a reference antibody comprising an HCVR / LCVR amino acid sequence pair selected from the group consisting of SEQ ID NOs: 2 / 10, 18 / 26, 34 / 42, 50 / 58, 66 / 58, 74 / 58, 82 / 58, 90 / 58, 98 / 58, 106 / 114, 122 / 130, 138 / 146, 154 / 162, 170 / 178, 186 / 194, 202 / 210, 218 / 226, 234 / 242, 250 / 258, 266 / 274, 282 / 290, 298 / 306, 314 / 322, 330 / 338, 346 / 354, 362 / 370, and 378 / 386.

[0042] The present invention further provides antibodies or antigen-binding fragments thereof that bind to the same epitope on human STEAP2 as a reference antibody comprising the HCVR / LCVR amino acid sequence pair set forth in Table 1. In another embodiment, the antibody or antigen-binding fragment is selected from the group consisting of SEQ ID NOs: 2 / 10, 18 / 26, 34 / 42, 50 / 58, 66 / 58, 74 / 58, 82 / 58, 90 / 58, 98 / 58, 106 / 114, 122 / 130, 138 / 146, 154 / 162, 170 / 178, 186 / 194, 202 / 210, 218 / 226, 234 / 242, 250 / 258 , 266 / 274, 282 / 290, 298 / 306, 314 / 322, 330 / 338, 346 / 354, 362 / 370, and 378 / 386.

[0043] The present invention further provides an isolated antibody or antigen-binding fragment thereof that binds to human STEAP2, wherein the antibody or antigen-binding fragment comprises a complementarity-determining region (CDR) of a heavy chain variable region (HCVR) having an amino acid sequence set forth in Table 1, and a CDR of a light chain variable region (LCVR) having an amino acid sequence set forth in Table 1. In another embodiment, the isolated antibody or antigen-binding fragment comprises the heavy chain CDRs and light chain CDRs of an HCVR / LCVR amino acid sequence pair selected from the group consisting of SEQ ID NOs: 2 / 10, 18 / 26, 34 / 42, 50 / 58, 66 / 58, 74 / 58, 82 / 58, 90 / 58, 98 / 58, 106 / 114, 122 / 130, 138 / 146, 154 / 162, 170 / 178, 186 / 194, 202 / 210, 218 / 226, 234 / 242, 250 / 258, 266 / 274, 282 / 290, 298 / 306, 314 / 322, 330 / 338, 346 / 354, 362 / 370, and 378 / 386.In yet another embodiment, the isolated antibodies or antigen-binding fragments are selected from the group consisting of SEQ ID NOs: 4-6-8-12-14-16, 20-22-24-28-30-32, 36-38-40-44-46-48, 52-54-56-60-62-64, 68-70-72-60-62-64, 76-78-80-60-62-64, 84-86-88-60-62-64, 92-94-96-6 0-62-64, 100-102-104-60-62-64, 108-110-112-116-118-120, 124-126-128-132-134-136, 140-142-144-148-150-152, 156-158-160-164-166-168, 172-174-176-180-182-184, 188-190-192-196-198-200, 204-206-208-212-214-216, 220-222-224-228-230-232, 236-238-240-244-246-248, 252-254-256-260-262-264, 268-270-272-276-278-280, 284-286-288-292-294-296, 300-302-304-308-310-312, 316- and 380-382-384-388-390-392.

[0044] In another aspect, the invention provides an isolated antibody or antigen-binding fragment thereof that binds to human STEAP2, wherein the antibody or antigen-binding fragment is selected from the group consisting of (a) SEQ ID NOs: 2, 18, 34, 50, 66, 74, 82, 90, 98, 106, 122, 138, 154, 170, 186, 202, 218, 234, 250, 266, 282, 298, 314, 330, 346, 362, and 370. 78; and (b) a heavy chain variable region (HCVR) having an amino acid sequence selected from the group consisting of SEQ ID NOs: 10, 26, 42, 58, 114, 130, 146, 162, 178, 194, 210, 226, 242, 258, 274, 290, 306, 322, 338, 354, 370, and 386. In a further embodiment, the isolated antibody or antigen-binding fragment of claim 10 comprises an HCVR / LCVR amino acid sequence pair selected from the group consisting of SEQ ID NOs: 2 / 10, 18 / 26, 34 / 42, 50 / 58, 66 / 58, 74 / 58, 82 / 58, 90 / 58, 98 / 58, 106 / 114, 122 / 130, 138 / 146, 154 / 162, 170 / 178, 186 / 194, 202 / 210, 218 / 226, 234 / 242, 250 / 258, 266 / 274, 282 / 290, 298 / 306, 314 / 322, 330 / 338, 346 / 354, 362 / 370, and 378 / 386.

[0045] According to another aspect, the present invention provides an antibody-drug conjugate comprising an anti-STEAP2 antibody or antigen-binding fragment thereof and a therapeutic agent (e.g., a cytotoxic agent). In some embodiments, the antibody or antigen-binding fragment and the cytotoxic agent are covalently linked via a linker, as discussed herein. In various embodiments, the anti-STEAP2 antibody or antigen-binding fragment can be any of the anti-STEAP2 antibodies or fragments thereof described herein.

[0046] In some embodiments, the cytotoxic agent is selected from an auristatin, a maytansinoid, a tubulysin, a tomaymycin derivative, or a dolastatin derivative. In some cases, the cytotoxic agent is an auristatin selected from MMAE or MMAF, or a maytansinoid selected from DM1 or DM4. In some embodiments, the cytotoxic agent is a maytansinoid having the structure of Formula (I) or Formula (II), as discussed herein.

[0047] In some embodiments, the cytotoxic agent is a maytansinoid having the following structure: [ka]

[0048] In some embodiments, the cytotoxic agent is a maytansinoid having the following structure: [ka]

[0049] In some embodiments, the antibody-drug conjugate comprises an anti-STEAP2 antibody or fragment thereof and [ka] During the ceremony, [ka] is binding to an anti-STEAP2 antibody or fragment thereof.

[0050] In some embodiments, the antibody-drug conjugate comprises an anti-STEAP2 antibody or fragment thereof and [ka] During the ceremony, [ka] is binding to an anti-STEAP2 antibody or fragment thereof.

[0051] In some embodiments, the antibody-drug conjugate comprises an anti-STEAP2 antibody or fragment thereof and [ka] During the ceremony, [ka] is binding to an anti-STEAP2 antibody or fragment thereof.

[0052] In some embodiments, the bond is contacted with the antibody or fragment thereof through the sulfur moiety of a cysteine ​​residue.

[0053] In some embodiments, the antibody-drug conjugate comprises an anti-STEAP2 antibody or fragment thereof, [ka] and mixtures thereof, During the ceremony, [ka] is the binding to an anti-STEAP2 antibody or fragment thereof.

[0054] In some embodiments, the bond is contacted with the antibody or fragment thereof through the nitrogen moiety of a lysine residue.

[0055] In any of the various embodiments of the antibody-drug conjugate described above or discussed herein, the antibody-drug conjugate can include 1 to 4 cytotoxic agents per anti-STEAP2 antibody or fragment thereof.

[0056] According to another aspect, the present invention provides bispecific antigen-binding molecules (e.g., antibodies) that bind to STEAP2 and CD3. Such bispecific antigen-binding molecules are also referred to herein as "anti-STEAP2 / anti-CD3 bispecific molecules," "anti-CD3 / anti-STEAP2 bispecific molecules," "anti-CD3 / anti-STEAP2 bispecific molecules," "anti-STEAP2 bispecific molecules," "anti-STEAP2 bispecific molecules," "anti-CD3 bispecific molecules," "anti-STEAP2 ... The anti-STEAP2 / anti-CD3 bispecific molecules are also referred to as "bispecific molecules," or "STEAP2 x CD3bsAb." Anti-STEAP2 / anti-CD3 bispecific molecules are useful for targeting cells (e.g., tumor cells) that express STEAP2 (e.g., prostate tumors), and the anti-CD3 portion of the bispecific molecule is useful for activating T cells. Simultaneous binding of STEAP2 on tumor cells and CD3 on T cells promotes direct killing (cytolysis) of the targeted tumor cells by the activated T cells. Thus, the anti-STEAP2 / anti-CD3 bispecific molecules of the invention are useful, inter alia, for treating diseases and disorders associated with or caused by STEAP2-expressing tumors (e.g., prostate cancer).

[0057] The bispecific antigen-binding molecule according to this aspect of the invention comprises a first antigen-binding domain that specifically binds human CD3 and a second antigen-binding domain that specifically binds STEAP2. The invention includes anti-STEAP2 / anti-CD3 bispecific molecules (e.g., bispecific antibodies) in which each antigen-binding domain comprises a heavy chain variable region (HCVR) paired with a light chain variable region (LCVR). In certain exemplary embodiments of the invention, the anti-CD3 antigen-binding domain and the anti-STEAP2 antigen-binding domain each comprise distinct HCVRs paired with a common LCVR. For example, as shown in Example 4 herein, a bispecific antibody was constructed comprising a first antigen-binding domain that specifically binds CD3, the first antigen-binding domain comprising a HCVR from an anti-CD3 antibody paired with a LCVR from an anti-STEAP2 antibody (e.g., the same LCVR contained in the anti-STEAP2 antigen-binding domain), and a second antigen-binding domain that specifically binds STEAP2, the second antigen-binding domain comprising a HCVR / LCVR from an anti-STEAP2 antibody. In other words, in the exemplary molecules disclosed herein, pairing of an HCVR from an anti-CD3 antibody with an LCVR from an anti-STEAP2 antibody generates an antigen-binding domain that specifically binds to CD3 (but not STEAP2). In such embodiments, the first and second antigen-binding domains contain different anti-CD3 and anti-STEAP2 HCVRs but share a common anti-STEAP2 LCVR. In other embodiments, the bispecific antigen-binding molecule contains different anti-CD3 and anti-STEAP2 HCVRs but shares a common LCVR. The amino acid sequence of this LCVR is shown, for example, in SEQ ID NO: 1890, and the amino acid sequences of the corresponding CDRs (i.e., LCDR1-LCDR2-LCDR3) are shown in SEQ ID NOs: 1892, 1894, and 1896, respectively. Genetically modified mice can be used to produce fully human bispecific antigen-binding molecules containing two different heavy chains associated with the same light chain containing variable domains derived from one of two different human light chain variable region gene segments. Alternatively, the variable heavy chains may be paired with one common light chain and expressed recombinantly in a host cell.Thus, antibodies of the present invention can comprise an immunoglobulin heavy chain associated with a single rearranged light chain. In some embodiments, the light chain comprises a variable domain derived from a human Vκ1-39 gene segment or a Vκ3-20 gene segment. In other embodiments, the light chain comprises a variable domain derived from a human Vκ1-39 gene segment rearranged with a human Jκ5 or a human Jκ1 gene segment.

[0058] The present invention provides anti-CD3 / anti-STEAP2 bispecific molecules, wherein a first antigen-binding domain that specifically binds to CD3 comprises any of the HCVR amino acid sequences, any of the LCVR amino acid sequences, any of the HCVR / LCVR amino acid sequence pairs, any of the heavy chain CDR1-CDR2-CDR3 amino acid sequences, or any of the light chain CDR1-CDR2-CDR3 amino acid sequences, as described in U.S. Patent Publication No. 2014 / 0088295, published March 27, 2014, and PCT Publication No. PCT / US2016 / 044732, filed July 29, 2016.

[0059] Furthermore, the present invention provides anti-CD3 / anti-STEAP2 bispecific molecules, wherein the first antigen-binding domain that specifically binds to CD3 comprises any of the HCVR amino acid sequences set forth in Tables 9, 11, and 15 herein. may also comprise any of the LCVR amino acid sequences set forth in Tables 1, 9, 12, and 17 herein. According to particular embodiments, the first antigen-binding domain that specifically binds CD3 comprises any of the HCVR / LCVR amino acid sequence pairs set forth in Tables 9, 11, 12, 15, and 17 herein. The present invention also provides anti-CD3 / anti-STEAP2 bispecific molecules, wherein the first antigen-binding domain that specifically binds CD3 comprises any of the heavy chain CDR1-CDR2-CDR3 amino acid sequences set forth in Tables 9, 11, and 15 herein, and / or any of the light chain CDR1-CDR2-CDR3 amino acid sequences set forth in Tables 1, 9, 12, and 17 herein.

[0060] According to certain embodiments, the present invention provides anti-CD3 / anti-STEAP2 bispecific molecules, wherein the first antigen-binding domain that specifically binds to CD3 comprises a heavy chain variable region (HCVR) having an amino acid sequence set forth in Tables 9, 11, and 15 herein, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

[0061] The present invention also provides anti-CD3 / anti-STEAP2 bispecific molecules, wherein a first antigen-binding domain that specifically binds CD3 comprises a light chain variable region (LCVR) having an amino acid sequence set forth in Tables 1, 9, 12, and 17 herein, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

[0062] The present invention also provides anti-CD3 / anti-STEAP2 bispecific molecules, wherein a first antigen-binding domain that specifically binds CD3 comprises a HCVR and LCVR (HCVR / LCVR) amino acid sequence pair set forth in Tables 9, 11, 12, 15, and 17 herein.

[0063] The present invention also provides anti-CD3 / anti-STEAP2 bispecific molecules, wherein the first antigen-binding domain that specifically binds to CD3 comprises a heavy chain CDR3 (HCDR3) domain having an amino acid sequence set forth in Tables 9, 11, and 15 herein, or a substantially similar sequence thereof with at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto, and a light chain CDR3 (LCDR3) domain having an amino acid sequence set forth in Tables 1, 9, 12, and 17 herein, or a substantially similar sequence thereof with at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0064] In certain embodiments, the first antigen-binding domain that specifically binds to CD3 comprises an HCDR3 / LCDR3 amino acid sequence pair set forth in Tables 9, 11, 12, 15, and 17 herein.

[0065] The present invention also provides anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules, wherein the first antigen-binding domain that specifically binds to CD3 comprises a heavy chain CDR1 (HCDR1) domain having amino acids set forth in Tables 9, 11, and 15 herein, or a substantially similar sequence thereof with at least 90%, at least 95%, at least 98%, or at least 99% sequence identity, and a heavy chain CDR2 (HCDR2) domain having amino acids set forth in Tables 9, 11, and 15, or a substantially similar sequence thereof with at least 90%, at least 95%, at least 98%, or at least 99% sequence identity. a heavy chain CDR3 (HCDR3) domain having an amino acid sequence set forth in Tables 9, 11, and 15, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity; and a light chain CDR1 (LCDR1) domain having an amino acid sequence set forth in Tables 1, 9, 12, and 17 herein, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity. and a light chain CDR2 (LCDR2) domain having an amino acid sequence set forth in Tables 1, 9, 12, and 17 herein, or a substantially similar sequence thereof with at least 90%, at least 95%, at least 98%, or at least 99% sequence identity; and a light chain CDR3 (LCDR3) domain having an amino acid sequence set forth in Tables 1, 9, 12, and 17 herein, or a substantially similar sequence thereof with at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

[0066] Certain non-limiting exemplary anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules of the present invention comprise a first antigen-binding domain that specifically binds to CD3, comprising HCDR1-HCDR2-HCDR3-LCDR1-LCDR2-LCDR3 domains having the amino acid sequences set forth in Tables 9, 11, 12, 15, and 17, respectively, herein.

[0067] The present invention further provides a bispecific antigen-binding molecule, wherein a first antigen-binding domain that specifically binds to human CD3 comprises heavy chain complementarity-determining regions (HCDR1, HCDR2, and HCDR3) derived from a heavy chain variable region (HCVR) comprising the amino acids set forth in Table 9, Table 11, or Table 15, and light chain complementarity-determining regions (LCDR1, LCDR2, and LCDR3) derived from a light chain variable region (LCVR) comprising the amino acid sequences set forth in Table 1, Table 9, Table 12, or Table 17.

[0068] In another aspect, the present invention provides a bispecific antigen-binding molecule, wherein a first antigen-binding domain that specifically binds to human CD3 comprises heavy chain complementarity-determining regions (HCDR1, HCDR2, and HCDR3) from a heavy chain variable region (HCVR) selected from the group consisting of SEQ ID NOs: 1730, 1762, and 1866, and light chain complementarity-determining regions (LCDR1, LCDR2, and LCDR3) from a light chain variable region (LCVR) comprising an amino acid sequence selected from the group consisting of SEQ ID NO: 258.

[0069] The present invention further provides a bispecific antigen-binding molecule, wherein a first antigen-binding domain that specifically binds to human CD3 comprises three heavy chain complementarity-determining regions (A1-HCDR1, A1-HCDR2, and A1-HCDR3) and three light chain complementarity-determining regions (A1-LCDR1, A1-LCDR2, and A1-LCDR3), wherein A1-HCDR1 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1732, 1764, and 1868, A1-HCDR2 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1734, 1766, and 1870, and A1-HCDR3 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1736, 1768, and 1872, A1-LCDR1 comprises the amino acid sequence of SEQ ID NO: 260, A1-LCDR2 comprises the amino acid sequence of SEQ ID NO: 262, and A1-LCDR3 comprises the amino acid sequence of SEQ ID NO: 264.

[0070] In a further aspect, the present invention provides a bispecific antigen-binding molecule, wherein a first antigen-binding domain that specifically binds to human CD3 comprises the heavy chain CDRs and light chain CDRs of an HCVR / LCVR amino acid sequence pair selected from the group consisting of SEQ ID NOs: 1730 / 258, 1762 / 258, and 1866 / 258.

[0071] In another embodiment, the present invention provides an antigen-binding molecule, wherein a first antigen-binding domain that specifically binds to human CD3 comprises three heavy chain complementarity-determining regions (A1-HCDR1, A1-HCDR2, and A1-HCDR3) and three light chain complementarity-determining regions (A1-LCDR1, A1-LCDR2, and A1-LCDR3), and a second antigen-binding domain that specifically binds to human STEAP2 comprises three heavy chain complementarity-determining regions (A2-HCDR1, A2-HCDR2, and A2-HCDR3) and three light chain complementarity-determining regions (A2-LCDR1, A2-LCDR2, and A2-LCDR3), and A1-HCDR1 is selected from the group consisting of SEQ ID NO: A1-HCDR1 comprises the amino acid sequence of SEQ ID NO: 260, A1-LCDR2 comprises the amino acid sequence of SEQ ID NO: 262, and A1-LCDR3 comprises the amino acid sequence of SEQ ID NO: 264, A2-HCDR1 comprises the amino acid sequence of SEQ ID NO: 252, A2-HCDR2 comprises the amino acid sequence of SEQ ID NO: 254, and A2-HCDR3 comprises the amino acid sequence of SEQ ID NO: 256, A2-LCDR1 comprises the amino acid sequence of SEQ ID NO: 260, A2-LCDR2 comprises the amino acid sequence of SEQ ID NO: 262, and A2-LCDR3 comprises the amino acid sequence of SEQ ID NO: 264.

[0072] Certain non-limiting exemplary anti-CD3 / anti-STEAP2 bispecific antigen binding molecules of the present invention comprise a first antigen-binding domain that specifically binds to CD3 comprising a heavy chain comprising a variable domain framework region having an amino acid sequence selected from FR1 (SEQ ID NO: 1903), FR2 (SEQ ID NO: 1904), FR3 (SEQ ID NO: 1905), and FR4 (SEQ ID NO: 1906).

[0073] In a further embodiment, exemplary anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules of the invention include bispecific antigen-binding molecules in which a first antigen-binding domain that specifically binds human CD3 comprises a HCVR comprising HCDR1-HCDR2-HCDR3 having the amino acid sequence of SEQ ID NOs: 1907-1908-1909.

[0074] The present invention also provides anti-CD3 / anti-STEAP2 bispecific molecules, wherein the second antigen-binding domain that specifically binds STEAP2 comprises a heavy chain variable region (HCVR) having an amino acid sequence selected from the group consisting of SEQ ID NOs: 2, 18, 34, 50, 66, 74, 82, 90, 98, 106, 122, 138, 154, 170, 186, 202, 218, 234, 250, 266, 282, 298, 314, 330, 346, 362, and 378, or a substantially similar sequence thereof with at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

[0075] The present invention also provides anti-CD3 / anti-STEAP2 bispecific molecules, wherein the second antigen-binding domain that specifically binds STEAP2 comprises a light chain variable region (LCVR) having an amino acid sequence selected from the group consisting of SEQ ID NOs: 10, 26, 42, 58, 114, 130, 146, 162, 178, 194, 210, 226, 242, 258, 274, 290, 306, 322, 338, 354, 370, and 386, or a substantially similar sequence thereof with at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

[0076] The present invention also provides an anti-CD3 / anti-STEAP2 bispecific molecule, wherein the second antigen-binding domain that specifically binds to STEAP2 comprises the HCVR and LCVR (HCVR / LCVR) amino acid sequence pair of SEQ ID NOs: 250 / 258.

[0077] The present invention also provides an anti-CD3 / anti-STEAP2 bispecific molecule, wherein the second antigen-binding domain that specifically binds STEAP2 comprises a heavy chain CDR3 (HCDR3) domain having the amino acid sequence of SEQ ID NO: 256, or a substantially similar sequence thereof with at least 90%, at least 95%, at least 98%, or at least 99% sequence identity, and a light chain CDR3 (LCDR3) domain having the amino acid sequence of SEQ ID NO: 264, or a substantially similar sequence thereof with at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

[0078] In certain embodiments, the second antigen-binding domain that specifically binds to STEAP2 comprises an HCDR3 / LCDR3 amino acid sequence pair selected from the group consisting of SEQ ID NOs: 256 / 264.

[0079] The second antigen-binding domain that specifically binds to STEAP2 is a heavy chain CDR1 (HCDR1) domain having an amino acid sequence selected from the group consisting of SEQ ID NOs: 4, 20, 36, 52, 68, 76, 84, 92, 100, 108, 124, 140, 156, 172, 188, 204, 220, 236, 252, 268, 284, 300, 316, 332, 348, 364, and 380, or a substantially similar sequence thereof with at least 90%, at least 95%, at least 98%, or at least 99% sequence identity, and a heavy chain CDR1 (HCDR1) domain having an amino acid sequence selected from the group consisting of SEQ ID NOs: 6, 22, 33, 34, 36, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 10 a heavy chain CDR2 (HCDR2) domain having an amino acid sequence selected from the group consisting of SEQ ID NOs: 8, 54, 70, 78, 86, 94, 102, 110, 126, 142, 158, 174, 190, 206, 222, 238, 254, 270, 286, 302, 318, 334, 350, 366, and 382, ​​or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity; heavy chain CDR3 (HCDR3) domains having an amino acid sequence selected from the group consisting of SEQ ID NOs: 12, 28, 44, 60, 116, 132, 148, 164, 180, 196, 212, 228, 244, 260, 276, 292, 308, 324, 340, 356, 372, and 384, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity; a light chain CDR1 (LCDR1) domain having an amino acid sequence selected from the group consisting of SEQ ID NOs: 14, 30, 46, 62, 118, 134, 150, 166, 182, 198, 214, 230, 246, 262, 278, 294, 310, 326, 342, 358, 374, and 390, or a substantially similar sequence thereof having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity;and a light chain CDR3 (LCDR3) domain having an amino acid sequence selected from the group consisting of SEQ ID NOs: 16, 32, 48, 64, 120, 136, 152, 168, 184, 200, 216, 232, 248, 264, 280, 296, 312, 328, 344, 360, 376, and 392, or a substantially similar sequence thereof with at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

[0080] Certain non-limiting exemplary anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules of the present invention comprise a second antigen-binding domain that specifically binds to STEAP2, comprising HCDR1-HCDR2-HCDR3-LCDR1-LCDR2-LCDR3, each having an amino acid sequence selected from the group consisting of SEQ ID NOs: 252-254-256-260-262-264.

[0081] In a related embodiment, the invention comprises an anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule, wherein the second antigen-binding domain that specifically binds STEAP2 comprises heavy and light chain CDR domains contained within heavy and light chain variable region (HCVR / LCVR) sequences selected from the group consisting of SEQ ID NOs: 250 / 258.

[0082] In another aspect, the present invention provides a bispecific antigen-binding molecule comprising a first antigen-binding domain that binds to human CD3 and a second antigen-binding domain that binds to human STEAP2. and the second antigen-binding domain is derived from any one of the anti-STEAP2 antibodies or anti-binding fragments of the present invention. In a further aspect, the present invention provides a bispecific antigen-binding molecule comprising a first antigen-binding domain that specifically binds to human CD3 and a second antigen-binding domain that specifically binds to human STEAP2.

[0083] The present invention further provides a bispecific antigen-binding molecule that binds to human cells expressing human CD3 and cynomolgus monkey cells expressing cynomolgus monkey CD3. In another embodiment, the bispecific antigen-binding molecule binds to human cells expressing human STEAP2.

[0084] In another aspect, the present invention provides bispecific antigen-binding molecules that inhibit tumor growth in immunocompromised mice bearing human prostate cancer xenografts.

[0085] In certain embodiments, the anti-CD3 antibodies, antibody-binding fragments thereof, and bispecific antibodies of the present invention were generated by stepwise replacement of parental amino acid residues based on differences between the germline sequence and the parental antibody sequence.

[0086] In some embodiments, the invention provides a bispecific antigen-binding molecule, wherein the second antigen-binding domain competes for binding to human STEAP2 with a reference antigen-binding protein comprising three heavy chain complementarity-determining regions (A2-HCDR1, A2-HCDR2, and A2-HCDR3) and three light chain complementarity-determining regions (A2-LCDR1, A2-LCDR2, and A2-LCDR3), wherein A2-HCDR1 comprises an amino acid sequence selected from the group consisting of SEQ ID NO: 252, A2-HCDR2 comprises the amino acid sequence of SEQ ID NO: 254, A2-HCDR3 comprises the amino acid sequence of SEQ ID NO: 256, A2-LCDR1 comprises the amino acid sequence of SEQ ID NO: 260, A2-LCDR2 comprises the amino acid sequence of SEQ ID NO: 262, and A2-LCDR3 comprises the amino acid sequence of SEQ ID NO: 264. In some embodiments, the present invention provides a bispecific antigen-binding molecule, wherein the second antigen-binding domain competes for binding to human STEAP2 with a reference antigen-binding protein comprising a heavy chain variable region (HCVR) comprising the amino acid sequence of SEQ ID NO: 250 and a light chain variable region (LCVR) comprising the amino acid sequence of SEQ ID NO: 258.

[0087] In some embodiments, the invention provides a bispecific antigen-binding molecule, wherein a first antigen-binding domain competes for binding to human CD3 with a reference antigen-binding protein comprising three heavy chain complementarity-determining regions (A1-HCDR1, A1-HCDR2, and A1-HCDR3) and three light chain complementarity-determining regions (A1-LCDR1, A1-LCDR2, and A1-LCDR3), wherein A1-HCDR1 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1732, 1764, and 1868, A1-HCDR2 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1734, 1766, and 1870, and A1-HCDR3 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1736, 1768, and 1872, A1-LCDR1 comprises the amino acid sequence of SEQ ID NO: 260, A1-LCDR2 comprises the amino acid sequence of SEQ ID NO: 262, and A1-LCDR3 comprises the amino acid sequence of SEQ ID NO: 264. In some embodiments, the present invention provides a bispecific antigen-binding molecule, wherein a first antigen-binding domain competes for binding to human CD3 with a reference antigen-binding protein comprising a heavy chain variable region (HCVR) comprising an amino acid sequence selected from the group consisting of SEQ ID NOs: 1730, 1762, and 1866, and a light chain variable region (LCVR) comprising the amino acid sequence of SEQ ID NO: 258.

[0088] In some embodiments, the present invention provides a bispecific antigen-binding molecule, wherein the first antigen-binding domain comprises a reference antigen-binding domain comprising, for binding to human CD3, a heavy chain variable region (HCVR) comprising an amino acid sequence selected from the group consisting of SEQ ID NOs: 1730, 1762, and 1866, and a light chain variable region (LCVR) comprising the amino acid sequence of SEQ ID NO: 258. The first antigen-binding domain competes with a reference antigen-binding protein comprising a heavy chain variable region (HCVR) comprising the amino acid sequence of SEQ ID NO: 250 and a light chain variable region (LCVR) comprising the amino acid sequence of SEQ ID NO: 258 for binding to human STEAP2.

[0089] In one embodiment, the present invention provides a pharmaceutical composition comprising an anti-STEAP2 antigen binding molecule or an anti-STEAP2 / anti-CD3 bispecific antigen binding molecule and a pharmaceutically acceptable carrier or diluent. The present invention also provides a method for treating cancer in a subject, comprising administering to the subject a pharmaceutical composition comprising an anti-STEAP2 antigen binding molecule or an anti-STEAP2 / anti-CD3 bispecific antigen binding molecule and a pharmaceutically acceptable carrier or diluent. In some embodiments, the cancer is selected from the group consisting of prostate cancer, bladder cancer, cervical cancer, lung cancer, colon cancer, kidney cancer, breast cancer, pancreatic cancer, gastric cancer, uterine cancer, and ovarian cancer. Optionally, the cancer is prostate cancer. Optionally, the prostate cancer is castration-resistant prostate cancer.

[0090] In another aspect, the present invention provides nucleic acid molecules encoding any of the HCVR, LCVR, or CDR sequences of the anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules disclosed herein, including nucleic acid molecules comprising the polynucleotide sequences set forth in Tables 2, 10, 13, 14, 16, and 18 herein, as well as nucleic acid molecules comprising two or more of the polynucleotide sequences set forth in Tables 2, 10, 13, 14, 16, and 18 in any functional combination or arrangement thereof. Recombinant expression vectors harboring nucleic acids of the invention and host cells into which such vectors have been introduced are also encompassed by the present invention, as are methods of producing the antibodies by culturing the host cells under conditions that allow for the production of the antibodies and recovering the antibodies produced.

[0091] The present invention includes anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules in which any of the above-mentioned antigen-binding domains that specifically bind CD3 are combined, linked, or associated with any of the above-mentioned antigen-binding domains that specifically bind STEAP2 to form bispecific antigen-binding molecules that bind both CD3 and STEAP2.

[0092] The present invention includes anti-CD3 / anti-STEAP2 bispecific antigen binding molecules with altered glycosylation patterns. In some applications, modifications to remove undesired glycosylation sites or antibodies lacking fucose moieties, for example, to enhance antibody-dependent cellular cytotoxicity (ADCC) function, may be useful (see Shield et al. (2002) JBC 277:26733). In other applications, modifications of galactosylation can be made to modify complement-dependent cytotoxicity (CDC).

[0093] In another aspect, the invention provides a pharmaceutical composition comprising an anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule disclosed herein and a pharmaceutically acceptable carrier. In a related aspect, the invention features a composition that is a combination of an anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule and a second therapeutic agent. In one embodiment, the second therapeutic agent is any agent that can be advantageously combined with the anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule. Exemplary agents that can be advantageously combined with the anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule are discussed in detail elsewhere herein.

[0094] In yet another aspect, the present invention provides a therapeutic method for targeting / killing STEAP2-expressing tumor cells using the anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule of the present invention, the therapeutic method comprising administering to a subject in need thereof a therapeutically effective amount of a pharmaceutical composition comprising the anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule of the present invention.

[0095] The present invention also relates to the use of an anti-CD3 / anti-STEAP2 antibody of the invention in the manufacture of a medicament for the treatment of a disease or disorder associated with or caused by STEAP2-expressing cells. This includes the use of multispecific antigen-binding molecules.

[0096] Other embodiments will become apparent from review of the following detailed description. [Brief explanation of the drawings]

[0097] [Figure 1] This shows the efficacy of H1H7814N-7 in a STEAP2-positive prostate cancer xenograft model (SCID mice implanted with C4-2 cells) in which H1H7814N-7 was administered at doses of 10 mg / kg, 20 mg / kg, or 40 mg / kg on day 13 after implantation. [Figure 2] The efficacy of H1H7814N-7 is shown in a STEAP2-positive prostate cancer xenograft model (SCID mice implanted with C4-2 cells) in which H1H7814N-7 was administered at a dose of 20 mg / kg on day 14 after implantation. [Figure 3] This shows the efficacy of H1H7814N-7 in a STEAP2-positive prostate cancer xenograft model (SCID mice implanted with C4-2 cells) in which H1H7814N-7 was administered at a dose of 150 μg / kg on day 17 after implantation. [Figure 4] The efficacy of H1H7814N-60 is shown in a STEAP2-positive prostate cancer xenograft model (SCID mice implanted with C4-2 cells) in which H1H7814N-60 was administered at a dose of 2.5 mg / kg on day 29 post-implantation (DAR 3.6 TV). [Figure 5] Binding of STEAP2xCD3 bispecific antibody to Jurkat cells is shown. [Figure 6] Binding of STEAP2xCD3 bispecific antibody to a human prostate cancer cell line (PC3) engineered to express a STEAP2 / 1 chimeric construct is shown. [Figure 7] Binding of STEAP2xCD3 bispecific antibody to cynomolgus monkey T cells. [Figure 8] Binding of STEAP2xCD3 bispecific antibody to cynomolgus monkey T cells. [Figure 9] 1 shows induction of human PBMC proliferation by STEAP2xCD3 bispecific antibody. [Figure 10] 1 shows induction of proliferation of cynomolgus monkey PBMCs by STEAP2xCD3 bispecific antibody. [Figure 11]Figure 1 shows depletion of C4-2 cells (STEAP2-bearing target cells) in a cytotoxicity assay by a representative STEAP2 x CD3 bispecific antibody in the presence of human PBMCs. [Figure 12] Activation of human T cells by a representative STEAP2xCD3 bispecific antibody is shown and correlates with the observed target cell lysis shown in FIG. 11. DETAILED DESCRIPTION OF THE INVENTION

[0098] Before describing the present invention, it is to be understood that this invention is not limited to the particular methods and experimental conditions described, as such methods and conditions may vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to be limiting, as the scope of the present invention will be limited only by the appended claims.

[0099] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by those skilled in the art to which this invention belongs.As used herein, the term "about" when used in relation to a specific listed numerical value means that this value can vary by 1% or less from the listed value.For example, as used in the present invention, the expression "about 100" includes 99 and 101, and all values ​​therebetween (for example, 99.1, 99.2, 99.3, 99.4, etc.).

[0100] Any methods and materials similar or equivalent to those described herein may be used in the practice of the present invention. or can be used for testing, the preferred methods and materials are now described.All patents, applications, and non-patent publications mentioned herein are incorporated by reference in their entirety.

[0101] definition As used herein, the term "CD3" refers to an antigen expressed on T cells as part of the multimolecular T cell receptor (TCR) and consisting of a homodimer or heterodimer formed from the association of two of the four receptor chains: CD3-epsilon, CD3-delta, CD3-zeta, and CD3-gamma. Human CD3-epsilon comprises the amino acid sequence set forth in SEQ ID NO: 1897, and human CD3-delta comprises the amino acid sequence set forth in SEQ ID NO: 1898. All references herein to proteins, polypeptides, and protein fragments are intended to refer to the human form of the respective protein, polypeptide, or protein fragment, unless specifically identified as being derived from a non-human species. Thus, the term "CD3" refers to human CD3 unless specifically identified as being derived from a non-human species, e.g., "mouse CD3," "simian CD3," etc.

[0102] As used herein, "antibodies that bind CD3" or "anti-CD3 antibodies" include antibodies and antigen-binding fragments thereof that specifically recognize a single CD3 subunit (e.g., epsilon, delta, gamma, or zeta), as well as antibodies and antigen-binding fragments thereof that specifically recognize a dimeric complex of two CD3 subunits (e.g., gamma / epsilon, delta / epsilon, and zeta / zeta CD3 dimers). The antibodies and antigen-binding fragments of the present invention can bind to soluble CD3 and / or cell surface-expressed CD3. Soluble CD3 includes native CD3 protein as well as recombinant CD3 protein variants, such as monomeric and dimeric CD3 constructs, that lack the transmembrane domain or are not associated with the cell membrane.

[0103] As used herein, the phrase "cell surface-expressed CD3" refers to one or more CD3 protein(s) expressed on the surface of a cell in vitro or in vivo, where at least a portion of the CD3 protein is exposed to the extracellular side of the cell membrane and is accessible to the antigen-binding portion of an antibody. "Cell surface-expressed CD3" includes CD3 proteins contained within functional T cell receptors within the cell membrane. The phrase "cell surface-expressed CD3" includes CD3 proteins expressed as part of homodimers or heterodimers on the surface of a cell (e.g., gamma / epsilon, delta / epsilon, and zeta / zeta CD3 dimers). The phrase "cell surface-expressed CD3" also includes CD3 chains expressed by themselves on the surface of a cell, without other CD3 chain types (e.g., CD3-epsilon, CD3-delta, or CD3-gamma). Alternatively, "cell surface-expressed CD3" can comprise or consist of CD3 protein expressed on the surface of a cell that does not normally express human CD3 on its surface but has been artificially engineered to express CD3 on its surface. Alternatively, "cell surface-expressed CD3" can comprise or consist of CD3 protein expressed on the surface of a cell that does not normally express human CD3 on its surface but has been artificially engineered to express CD3 on its surface.

[0104] The term "STEAP2" as used herein refers to the six-transmembrane epithelial antigen 2 of the prostate. STEAP2 is an endogenous six-transmembrane protein that is highly expressed in prostate epithelial cells and is a cell surface marker for prostate cancer. For example, STEAP2 was found to be expressed at significant levels in the LNCaP prostate cell line. (Porkka, et al. Lab Invest 2002, 82:1573-1582). STEAP2 (UniProtKB / Swiss-Prot:Q8NFT2.3) is a 490-amino acid protein encoded by the STEAP2 gene located in human chromosome region 7q21. For example, see the amino acid sequence of human STEAP2 set forth in SEQ ID NO: 1899. It refers to light.

[0105] As used herein, "antibodies that bind to STEAP2" or "anti-STEAP2 antibodies" include antibodies and antigen-binding fragments thereof that specifically recognize STEAP2.

[0106] The term "antigen-binding molecule" includes antibodies and antigen-binding fragments of antibodies, including, for example, bispecific antibodies.

[0107] As used herein, the term "antibody" refers to any antigen-binding molecule or molecular complex that contains at least one complementarity-determining region (CDR) that specifically binds to or interacts with a particular antigen (e.g., STEAP2 or CD3). The term "antibody" includes immunoglobulin molecules that contain four polypeptide chains, two heavy (H) chains and two light (L) chains, inter-connected by disulfide bonds, as well as multimers thereof (e.g., IgM). Each heavy chain contains a heavy chain variable region (referred to herein as HCVR or V). H The heavy chain constant region comprises three domains: C H 1. C H 2, and C H Each light chain comprises a light chain variable region (referred to herein as LCVR or V L The light chain constant region comprises one domain (C L 1) V H Area and V L The regions can be further subdivided into regions of hypervariability called complementarity determining regions (CDRs) interspersed with more conserved regions called framework regions (FRs). H and V L consists of three CDRs and four FRs arranged from the amino terminus to the carboxy terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. In different embodiments of the present invention, the FRs of an anti-STEAP2 antibody or anti-CD3 antibody (or antigen-binding portion thereof) may be identical to human germline sequences or may be naturally or artificially modified. An amino acid consensus sequence may be defined based on a parallel analysis of two or more CDRs.

[0108] As used herein, the term "antibody" also includes antigen-binding fragments of complete antibody molecules. The terms "antigen-binding portion" of an antibody, "antigen-binding fragment" of an antibody, and the like, as used herein, include naturally occurring, enzymatically obtained, synthetic, or genetically engineered polypeptides or glycoproteins that specifically bind antigens to form complexes. Antibody-binding fragments of antibodies can be derived from complete antibody molecules using any suitable standard techniques, such as proteolytic or recombinant genetic engineering techniques involving the manipulation and expression of DNA encoding antibody variable and, optionally, constant domains. Such DNA is known and / or readily available, for example, from commercial sources, DNA libraries (including, for example, phage-antibody libraries), or can be synthesized. DNA can be sequenced and manipulated, for example, chemically or using molecular biology techniques, to arrange one or more variable and / or constant domains with appropriate positional changes, or to introduce codons, generate cysteine ​​residues, modify, add, or delete amino acids, etc.

[0109] Non-limiting examples of antibody-binding fragments include (i) Fab fragments, (ii) F(ab')2 fragments, (iii) Fd fragments, (iv) Fv fragments, (v) single-chain Fv (scFv) molecules, (vi) dAb fragments, and (vii) minimal recognition units consisting of amino acid residues mimicking a hypervariable region of an antibody (e.g., an isolated complementarity-determining region (CDR) such as a CDR3 peptide), or a constrained FR3-CDR3-FR4 peptide. Domain-specific antibodies, single-domain antibodies, domain-deleted antibodies, chimeric antibodies, CDR-grafted antibodies, diabodies, triabodies, tetrabodies, minibodies, nanobodies (e.g., monovalent nanobodies, bivalent nanobodies, etc.), small modular immunopharmaceuticals (SMIPs), and other engineered molecules such as shark variable IgNAR domains are also encompassed by the term "antigen-binding fragment" as used herein.

[0110] Antigen-binding fragments of antibodies typically contain at least one variable domain. A variable domain may be of any size or amino acid composition and generally contains at least one CDR adjacent to or in-frame with one or more framework sequences. L V bound to the domain H In an antibody-binding fragment having a domain, V H Domains and V L The domains can be arranged relative to each other in any suitable configuration. For example, the variable region is a dimer and the V H -V H , V H -V L or V L -V L Alternatively, the antigen-binding fragment of an antibody may comprise a dimer of monomeric V H Domain or V L It may contain domains.

[0111] In certain embodiments, an antigen-binding fragment of an antibody may contain at least one variable domain covalently linked to at least one constant domain. Non-limiting exemplary configurations of variable and constant domains that may be found in an antigen-binding fragment of an antibody of the invention include: (i) VH-CH1, (ii) VH-CH2, (iii) VH-CH3, (iv) VH-CH1-CH2, (v) VH-CH1-CH2-CH3, (vi) VH-CH2-CH3, (vii) VH-CL, (viii) VL-CH1, (ix) VL-CH2, (x) VL-CH3, (xi) VL-CH1-CH2, (xii) VL-CH1-CH2-CH3, (xiii) VL-CH2-CH3, and (xiv) VL-CL. In any configuration of variable and constant domains, including any of the exemplary configurations listed above, the variable and constant domains may be directly linked to each other or may be linked by a complete or partial hinge or linker region. A hinge region may consist of at least two (e.g., 5, 10, 15, 20, 40, 60, or more) amino acids that provide a flexible or semi-flexible connection between adjacent variable and / or constant domains in a single polypeptide molecule. Moreover, antigen-binding fragments of antibodies of the present invention may be linked to each other and / or to one or more monomeric V H Domain or V L In non-covalent association with the domains, the variable domain configurations and constant domain configurations may comprise homodimers or heterodimers (or other multimers) of any of the above-listed variable domain configurations and constant domain configurations.

[0112] Like intact antibody molecules, antibody-binding fragments can be monospecific or multispecific (e.g., bispecific). Multispecific antigen-binding fragments of antibodies will typically comprise at least two different variable domains, each capable of specifically binding to a separate antigen or to a different epitope on the same antigen. Any multispecific antibody format, including the exemplary bispecific antibody formats disclosed herein, can be adapted for use in connection with the antigen-binding fragments of antibodies of the present invention using routine techniques available in the art.

[0113] The antibodies of the present invention can function through complement-dependent cytotoxicity (CDC) or antibody-dependent cell-mediated cytotoxicity (ADCC). "Complement-dependent cytotoxicity" (CDC) refers to the lysis of antigen-expressing cells by the antibodies of the present invention in the presence of complement. "Antibody-dependent cell-mediated cytotoxicity" (ADCC) refers to a cell-mediated reaction in which nonspecific cytotoxic cells expressing Fc receptors (FcRs) (e.g., natural killer (NK) cells, neutrophils, and macrophages) recognize bound antibodies on target cells, thereby resulting in lysis of the target cells. CDC and ADCC can be measured using assays well known and available in the art. (See, e.g., U.S. Patent Nos. 5,500,362 and 5,821,337, and Clynes et al. (1998) Proc. Natl. Acad. Sci. (USA) 95:652-656.) The constant region of an antibody is important in the ability of an antibody to fix complement and mediate cell-dependent cytotoxicity. Thus, the antibody isotype may be selected based on whether it is desirable for the antibody to mediate cytotoxicity.

[0114] In certain embodiments, the anti-STEAP2 monospecific or anti-STEAP2 / anti-CD3 bispecific antibodies of the invention are human antibodies. The term "human antibody," as used herein, is intended to include antibodies having variable and constant regions derived from human germline immunoglobulin sequences. Human antibodies of the invention may include, for example, amino acid residues in the CDRs, particularly CDR3, that are not encoded by human germline immunoglobulin sequences (e.g., mutations introduced by random or site-specific mutagenesis in vitro or by somatic mutation in vivo). However, the term "human antibody," as used herein, is not intended to include antibodies in which CDR sequences derived from the germline of another mammalian species (e.g., a mouse) have been grafted onto human framework sequences.

[0115] The antibodies of the present invention may, in some embodiments, be recombinant human antibodies. As used herein, the term "recombinant human antibody" is intended to include all human antibodies prepared, expressed, generated, or isolated by recombinant means, such as antibodies expressed using a recombinant expression vector transfected into a host cell (described below), antibodies isolated from a recombinant combinatorial human antibody library (described below), antibodies isolated from an animal (e.g., a mouse) that is transgenic for human immunoglobulin genes (e.g., Taylor et al. (1992) Nucl. Res. 20:6287-6295), or antibodies prepared, expressed, generated, or isolated by any other means, including splicing of human immunoglobulin gene sequences into other DNA sequences. Such recombinant human antibodies have variable and constant regions derived from human germline immunoglobulin sequences. In certain embodiments, however, such recombinant human antibodies are subjected to in vitro mutagenesis (or, when animals transgenic for human Ig sequences are used, in vivo somatic mutagenesis), thus modifying the V and V regions of the recombinant antibody. H and V L The amino acid sequence of the region is human germline V H and V L These are sequences that are derived from and related to sequences, but may not naturally occur within the human antibody germline repertoire in vivo.

[0116] Human antibodies can exist in two forms related to hinge heterogeneity. In one form, the immunoglobulin molecule contains a stable four-chain construct of approximately 150-160 kDa in which dimers are held together by interchain heavy chain disulfide bonds. In the second form, the dimers are not linked via interchain disulfide bonds, forming approximately 75-80 kDa molecules consisting of covalently linked light and heavy chains (half antibodies). These forms have been extremely difficult to separate, even after affinity purification.

[0117] The frequency of occurrence of the second form in various intact IgG isotypes is due to, but not limited to, structural differences associated with the antibody hinge region isotype. A single amino acid substitution in the hinge region of a human IgG4 hinge can significantly reduce the occurrence of the second form to levels typically observed with a human IgG1 hinge (Angal et al. al. (1993) Molecular Immunology 30:105). The present invention provides a method for producing a hinge, C H 2 or C H Antibodies with one or more mutations in the three regions are included, which may be desirable, for example, in production to improve the yield of the desired antibody form.

[0118] The antibody of the present invention may be an isolated antibody. As used herein, "isolated antibody" refers to an identified antibody and an antibody that has been separated and / or recovered from at least one component of its natural environment. For example, an antibody that has been separated or removed from at least one component of an organism, or from a tissue or cell in which it naturally occurs or is naturally produced, is an "isolated antibody" for purposes of the present invention. Isolated antibodies also include antibodies in situ within recombinant cells. An isolated antibody is an antibody that has undergone at least one purification or isolation step. According to certain embodiments, an isolated antibody is free from other cellular material and and / or may be substantially free of chemicals.

[0119] The present invention also includes one-arm antibodies that bind to STEAP2. As used herein, "one-arm antibody" refers to an antigen-binding molecule comprising a single antibody heavy chain and a single antibody light chain. The one-arm antibodies of the present invention may comprise any of the HCVR / LCVR or CDR amino acid sequences listed in Table 1.

[0120] The anti-STEAP2 or anti-STEAP2 / anti-CD3 antibodies disclosed herein may contain one or more amino acid substitutions, insertions, and / or deletions in the framework and / or CDR regions of the heavy and light chain variable domains compared to the corresponding germline sequences. Such mutations can be readily ascertained by comparing the amino acid sequences disclosed herein to germline sequences available, for example, from public antibody sequence databases. The present invention includes antibodies and antigen-binding fragments thereof derived from any of the amino acid sequences disclosed herein, in which one or more amino acids in one or more framework and / or CDR regions are mutated to the corresponding residue(s) in the germline sequence from which the antibody was derived, or to the corresponding residue(s) in another human germline sequence, or to a conservative amino acid substitution of the corresponding germline residue(s) (such sequence changes are collectively referred to herein as "germline mutations"). Starting from the heavy and light chain variable region sequences disclosed herein, one skilled in the art can readily generate numerous antibodies and antibody-binding fragments containing one or more individual germline mutations or combinations thereof. In certain embodiments, V H Domain and / or V LAll framework and / or CDR residues within a domain are mutated back to the residue found in the original germline sequence from which the antibody was derived. In other embodiments, only certain residues are mutated back to the original germline sequence, e.g., the mutated residue is found within the first 8 amino acids of FR1, or the mutated residue is found within the last 8 amino acids of FR4, or the mutated residue is found only in CDR1, CDR2, or CDR3. In other embodiments, one or more framework and / or CDR residue(s) are mutated to the corresponding residue(s) in a different germline sequence (i.e., a germline sequence different from the germline sequence from which the antibody was originally derived). Furthermore, antibodies of the present invention may contain any combination of two or more germline mutations within the framework and / or CDR regions, e.g., certain individual residues are mutated to the corresponding residue in a particular germline sequence, while certain other residues that differ from the original germline sequence are maintained or mutated to the corresponding residue in a different germline sequence. Once obtained, antibodies and antibody-binding fragments containing one or more germline mutations can be readily tested for one or more desired properties, such as improved binding specificity, increased binding affinity, improved or enhanced antagonist or agonist biological properties (as the case may be), reduced immunogenicity, etc. Antibodies and antibody-binding fragments obtained by this general method are encompassed within the scope of the present invention.

[0121] The present invention also includes anti-STEAP2 or anti-STEAP2 / anti-CD3 antibodies comprising variants of any of the HCVR, LCVR, and / or CDR amino acid sequences disclosed herein with one or more conservative substitutions. For example, the present invention includes anti-STEAP2 or anti-STEAP2 / anti-CD3 antibodies having HCVR, LCVR, and / or CDR amino acid sequences as set forth in Table 1 herein or as set forth in Tables 9, 11, 12, 15, and 17 herein with, e.g., 10 or fewer, 8 or fewer, 6 or fewer, or 4 or fewer conservative amino acid substitutions relative to any of the HCVR, LCVR, and / or CDR amino acid sequences.

[0122] The term "epitope" refers to an antigenic determinant that interacts with a specific antigen-binding site in the variable region of an antibody molecule, known as a paratope. A single antigen may contain more than one epitope. Thus, different antibodies may bind to different regions on an antigen and have different biological effects. Epitopes can be either conformational or linear. Conformational epitopes are generated by spatially juxtaposed amino acids from different segments of a linear polypeptide chain. Linear epitopes are those generated by adjacent amino acid residues in a polypeptide chain. In certain circumstances, epitopes may include carbohydrate, phosphoryl, or sulfonyl moieties on an antigen.

[0123] The terms "substantial identity" or "substantially identical," when referring to a nucleic acid or fragment thereof, indicate that when optimally aligned with appropriate nucleotide insertions or deletions with another nucleic acid (or its complementary strand), there is at least about 95%, more preferably at least about 96%, 97%, 98%, or 99% nucleotide sequence identity of the nucleotide bases as measured by any well-known algorithm of sequence identity, such as FASTA, BLAST, or Gap, as discussed below. A nucleic acid molecule having substantial identity to a reference nucleic acid molecule can, in certain cases, encode a polypeptide having the same or substantially similar amino acid sequence as the polypeptide encoded by the reference nucleic acid molecule.

[0124] When applied to polypeptides, the terms "substantial similarity" or "substantially similar" mean that two peptide sequences, when optimally aligned using a program such as GAP or BESTFIT with a predetermined gap weight, share at least 95% sequence identity, and even more preferably at least 98% or 99% sequence identity. Preferably, non-identical residue positions differ by conservative amino acid substitutions. A "conservative amino acid substitution" is one in which an amino acid residue is replaced by another amino acid residue having a side chain (R group) with similar chemical properties (e.g., charge or hydrophobicity). Generally, conservative amino acid substitutions do not substantially alter the functional properties of a protein. When two or more amino acid sequences differ from each other by conservative substitutions, the percent sequence identity or degree of similarity may be adjusted upward to correct for the conservative nature of the substitution. Means for making this adjustment are well known to those of skill in the art. See, e.g., Pearson (1994) Methods Mol. Biol. 24:307-331, incorporated herein by reference. Examples of groups of amino acids having side chains with similar chemical properties include: (1) aliphatic side chains: glycine, alanine, valine, leucine, and isoleucine, (2) aliphatic-hydroxyl side chains: serine and threonine, (3) amide-containing side chains: asparagine and glutamine, (4) aromatic side chains: phenylalanine, tyrosine, and tryptophan, (5) basic side chains: lysine, arginine, and histidine, (6) acidic side chains: aspartic acid and glutamic acid, and (7) sulfur-containing side chains: cysteine ​​and methionine. Preferred conservative amino acid substitution groups are valine-leucine-isoleucine, phenylalanine-tyrosine, lysine-arginine, alanine-valine, glutamic acid-aspartic acid, and asparagine-glutamine. Alternatively, a conservative substitution is any change that has a positive value in the PAM250 log-likelihood matrix disclosed in Gonnet et al. (1992) Science 256:1443-1445, which is incorporated herein by reference. A "moderately conservative" substitution is any change that has a non-negative value in the PAM250 log-likelihood matrix.

[0125] Sequence similarity for polypeptides, also called sequence identity, is typically measured using sequence analysis software. Protein analysis software matches similar sequences using similarity measures assigned to various substitutions, deletions, and other modifications, including conservative amino acid substitutions. For example, GCG software contains programs such as Gap and Bestfit, which can be used with default parameters to determine sequence homology or sequence identity between closely related polypeptides, such as homologous polypeptides from organisms of different species, or between a wild-type protein and its mutant protein. For example, GCG 6 See version 6.1. Polypeptide sequences can also be compared using FASTA, a program in GCG version 6.1, with default or recommended parameters. FASTA (e.g., FASTA2 and FASTA3) provides alignments and percent sequence identity of the regions of the best overlap between the query and search sequences (Pearson (2000) supra). Another preferred algorithm for comparing the sequences of the invention to a database containing a large number of sequences from different organisms is the computer program BLAST, particularly BLASTP or TBLASTN, with default parameters. See, for example, Altschul et al. (1990) J. Mol. Biol. 215:403-410 and Altschul et al. (1990) J. Mol. Biol. 215:403-410, each of which is incorporated herein by reference. See, al. (1997) Nucleic Acids Res. 25:3389-402.

[0126] Germline mutations The anti-CD3 antibodies disclosed herein contain one or more amino acid substitutions, insertions and / or deletions in the framework and / or CDR regions of the heavy chain variable domain compared to the corresponding germline sequences.

[0127] The present invention also includes antibodies and antigen-binding fragments thereof derived from any of the amino acid sequences disclosed herein, wherein one or more amino acids in one or more framework and / or CDR regions have been mutated to the corresponding residue(s) in the germline sequence from which the antibody was derived, or to the corresponding residue(s) in another human germline sequence, or to a conservative amino acid substitution of the corresponding germline residue(s) (such sequence changes collectively referred to herein as "germline mutations"), and which exhibit weak or no detectable binding to the CD3 antigen. Some such exemplary antibodies that recognize CD3 are listed in Tables 12 and 18 herein.

[0128] Furthermore, antibodies of the present invention may contain any combination of two or more germline mutations within the framework and / or CDR regions, e.g., certain individual residues are mutated to the corresponding residue in a particular germline sequence, while certain other residues that differ from the original germline sequence are maintained or mutated to the corresponding residue in a different germline sequence. Once obtained, antibodies and antibody-binding fragments containing one or more germline mutations can be tested for one or more desired properties, such as improved binding specificity, weaker or reduced binding affinity, improved or enhanced pharmacokinetic properties, reduced immunogenicity, etc. Antibodies and antigen-binding fragments obtained by this general method, given the guidance of this disclosure, are encompassed within the scope of the present invention.

[0129] The present invention also includes anti-CD3 antibodies comprising variants of any of the HCVR, LCVR, and / or CDR amino acid sequences disclosed herein with one or more conservative substitutions. For example, the present invention includes anti-CD3 antibodies having HCVR, LCVR, and / or CDR amino acid sequences with, for example, 10 or fewer, 8 or fewer, 6 or fewer, or 4 or fewer conservative amino acid substitutions relative to any of the HCVR, LCVR, and / or CDR amino acid sequences set forth in Tables 1, 9, 11, 12, 15, and 17 herein. The antibodies and bispecific antigen-binding molecules of the present invention contain one or more amino acid substitutions, insertions, and / or deletions in the framework and / or CDR regions of the heavy and light chain variable domains compared to the corresponding germline sequences from which the individual antigen-binding domains were derived, while maintaining or improving the desired weak to undetectable binding to the CD3 antigen. A "conservative amino acid substitution" is one in which an amino acid residue is replaced by another amino acid residue having a side chain (R group) with similar chemical properties (e.g., charge or hydrophobicity). Generally, conservative amino acid substitutions do not substantially alter the functional properties of a protein, i.e., the amino acid substitution maintains or improves the desired weak to undetectable binding affinity in the case of anti-CD3 binding molecules. Examples of groups of amino acids having side chains with the following structures include: (1) aliphatic side chains: glycine, alanine, valine, leucine, and isoleucine, (2) aliphatic-hydroxyl side chains: serine and threonine, (3) amide-containing side chains: asparagine and glutamine, (4) aromatic side chains: phenylalanine, tyrosine, and tryptophan, (5) basic side chains: lysine, arginine, and histidine, (6) acidic side chains: aspartic acid and glutamic acid, and (7) sulfur-containing side chains: cysteine ​​and methionine. Preferred conservative amino acid substitution groups are valine-leucine-isoleucine, phenylalanine-tyrosine, lysine-arginine, alanine-valine, glutamic acid-aspartic acid, and asparagine-glutamine. Alternatively, a conservative substitution is any change that has a positive value in the PAM250 log-likelihood matrix disclosed in Gonnet et al. (1992) Science 256:1443-1445, which is incorporated herein by reference. A "moderately conservative" substitution is any change that has a non-negative value in the PAM250 log-likelihood matrix.

[0130] The present invention also includes antigen-binding molecules comprising antigen-binding domains with HCVR and / or CDR amino acid sequences substantially identical to any of the HCVR and / or CDR amino acid sequences disclosed herein, while maintaining or improving the desired weak affinity for the CD3 antigen. When referring to amino acid sequences, the term "substantial identity" or "substantially identical" refers to two amino acid sequences that, when optimally aligned using predefined gap weights, such as by the programs GAP or BESTFIT, share at least 95% sequence identity, and even more preferably at least 98% or 99% sequence identity. Preferably, residue positions that are not identical differ by conservative amino acid substitutions. When two or more amino acid sequences differ from each other by conservative substitutions, the percent sequence identity or degree of similarity may be adjusted upward to correct for the conservative nature of the substitutions. Means for making this adjustment are well known to those skilled in the art. See, e.g., Pearson (1994) Methods Mol. Biol. 24:307-331.

[0131] Sequence similarity for polypeptides, also known as sequence identity, is typically measured using sequence analysis software. Protein analysis software matches similar sequences using similarity measures assigned to various substitutions, deletions, and other modifications, including conservative amino acid substitutions. For example, GCG software contains programs such as Gap and Bestfit, which can be used with default parameters to determine sequence homology or sequence identity between closely related polypeptides, such as homologous polypeptides from different species of organisms, or between a wild-type protein and its mutant protein. See, for example, GCG version 6.1. Polypeptide sequences can also be compared using FASTA, a program in GCG version 6.1, with default or recommended parameters. FASTA (e.g., FASTA2 and FASTA3) provides alignments and percent sequence identity of the regions of best overlap between the query and search sequences (Pearson (2000) supra). Another preferred algorithm for comparing the sequences of the present invention to a database containing a large number of sequences from different organisms is the computer program BLAST, particularly BLASTP or TBLASTN, with default parameters. See, e.g., Altschul et al. (1990) J. Mol. Biol. 215:403-410 and Altschul et al. (1997) Nucleic Acids Res. 25:3389-402.

[0132] Once obtained, the antigen-binding domain containing one or more germline mutations is tested for reduced binding affinity using one or more in vitro assays. Antibodies that recognize a specific antigen are typically screened for their purpose by testing for high (i.e., strong) binding affinity to the antigen, but the antibodies of the present invention exhibit weak or undetectable binding. Antibodies containing one or more antigen-binding domains obtained by this general method are also tested for their reduced binding affinity using one or more in vitro assays. Bispecific antigen-binding molecules are also included within the scope of the present invention and have been found to be advantageous as avidity-driven tumor therapy.

[0133] Unexpected benefits, such as improved pharmacokinetic properties and reduced toxicity to patients, may be realized from the methods described herein.

[0134] Antibody binding properties As used herein, the term "binding," in the context of the binding of either an antibody, immunoglobulin, antibody-binding fragment, or Fc-containing protein to a predetermined antigen, such as, for example, a cell surface protein or fragment thereof, typically refers to an interaction or association between at least two entities or molecular structures, such as an antibody-antigen interaction.

[0135] For example, binding affinities are typically about 10 when determined by surface plasmon resonance (SPR) techniques, e.g., on a BIAcore3000 instrument, using an antigen as the ligand and an antibody, Ig, antibody-binding fragment, or Fc-containing protein as the analyte (or antiligand). -7 M or less, for example, about 10 -8 M or less, for example, about 10 -9 K below M D Cell-based binding strategies such as fluorescence-activated cell sorting (FACS) binding assays are also routinely used, and FACS data correlate well with other methods such as radioligand competitive binding and SPR (Benedict, CA, J Immunol Methods 1997, 201(2):223-31; Geuijen, CA, et al. J Immunol Methods. 2005, 302(1-2):68-77).

[0136] Thus, an antibody or antigen-binding protein of the invention has a K that is at least 10-fold lower than its affinity for binding to a non-specific antigen (e.g., BSA, casein). D According to the present invention, the antibody binds to a given antigen or cell surface molecule with an affinity corresponding to a K value that is 10-fold lower than that of a non-specific antigen. DAlthough antibody affinities corresponding to values ​​can be considered as undetectable binding, such antibodies can be paired with a second antigen-binding arm to produce bispecific antibodies of the invention.

[0137] The term “K D " (M) refers to the dissociation equilibrium constant of a particular antibody-antigen interaction, or the dissociation equilibrium constant of an antibody or antibody-binding fragment binding to an antigen. K D There is an inverse relationship between K and binding affinity, and therefore, K D The smaller the value, the higher, i.e., stronger, the affinity. Thus, the terms "higher affinity" or "stronger affinity" refer to a higher ability to form an interaction, i.e., a smaller K D Conversely, the terms "lower affinity" or "weaker affinity" refer to a lower ability to form an interaction, i.e., a larger K D In some situations, a higher binding affinity (or K) of a particular molecule (e.g., an antibody) to its interaction partner molecule (e.g., antigen X) compared to the binding affinity of the molecule (e.g., an antibody) to another interaction partner molecule (e.g., antigen Y) can be used. D ) is larger than K D A smaller K value (lower, or weaker, affinity) D The binding affinity is expressed as a binding ratio determined by dividing by the binding affinity (higher, or stronger, affinity), e.g., 5-fold or 10-fold higher binding affinity as the case may be.

[0138] The term "k" d " (sec-1 or 1 / s) refers to the dissociation rate constant of a particular antibody-antigen interaction, or the dissociation rate constant of an antibody or antibody-binding fragment. Its value is k off Also called value.

[0139] The term "k" a " (M-1 x sec-1 or 1 / M) refers to the association rate constant of a particular antibody-antigen interaction, or the association rate constant of an antibody or antibody binding fragment.

[0140] The term “K A" (M-1 or 1 / M) is the association equilibrium constant for a particular antibody-antigen interaction, or the association equilibrium constant of an antibody or antibody-binding fragment. The association equilibrium constant is k a k d It is obtained by dividing by

[0141] "EC50" or "EC 50 The term "half-maximal effective concentration" refers to the concentration of antibody that induces a response midway between baseline and maximum after a specific exposure time. 50 essentially represents the concentration of an antibody at which 50% of its maximal effect is observed. In certain embodiments, the EC 50 The EC value is equal to the concentration of an antibody of the invention that confers half-maximal binding to cells expressing CD3 or a tumor-associated antigen, as determined, for example, by a FACS binding assay. Thus, reduced or weak binding is considered to be an EC 50 An increase in the effective concentration of 100 mg / kg or half-maximum effective concentration is observed.

[0142] In one embodiment, the reduction in binding is measured by measuring the EC2 concentration that allows half-maximal binding to target cells. 50 It can be defined as an increase in antibody concentration.

[0143] In another embodiment, EC 50 The values ​​represent the concentration of the antibody of the invention that induces half-maximal depletion of target cells by T cell cytotoxicity. Thus, increased cytotoxicity (e.g., T cell-mediated tumor cell killing) is associated with an EC 50 A decrease in the concentration of α-glucan is observed, or half of the maximum effective concentration value.

[0144] Bispecific antigen binding molecules The antibodies of the present invention can be monospecific, bispecific, or multispecific. Multispecific antibodies can be specific for different epitopes of a single target polypeptide or can contain antigen-binding domains specific for two or more target polypeptides. See, for example, Tutt et al., 1991, J. Immunol. 147:60-69; Kufer et al., 2004, Trends Biotechnol. 22:238-244. The anti-STEAP2 monospecific or anti-STEAP2 / anti-CD3 bispecific antibodies of the present invention can be linked to or coexpressed with another functional molecule, such as another peptide or protein. For example, an antibody or fragment thereof can be functionally linked (e.g., by chemical bonding, genetic fusion, noncovalent association, etc.) to one or more other molecular entities, such as another antibody or antibody fragment, to generate a bispecific or multispecific antibody with a second or additional binding specificity.

[0145] The use of the phrase "anti-CD3 antibody" or "anti-STEAP2 antibody" herein is intended to include both monospecific anti-CD3 or anti-STEAP2 antibodies and bispecific antibodies comprising a CD3-binding arm and a STEAP2-binding arm. Thus, the present invention includes bispecific antibodies in which one immunoglobulin arm binds to human CD3 and the other immunoglobulin arm is specific for human STEAP2. The CD3-binding arm may comprise any of the HCVR / LCVR or CDR amino acid sequences set forth in Tables 1, 9, 11, 12, 15, and 17 herein.

[0146] In certain embodiments, the CD3-binding arm binds to human CD3 and induces human T cell activation. In certain embodiments, the CD3-binding arm weakly binds to human CD3 and induces human T cell activation. In other embodiments, the CD3-binding arm weakly binds to human CD3 and induces tumor-associated antigen-expressing cell killing in the context of a bispecific or multispecific antibody. In other embodiments, the CD3-binding arm weakly binds or associates with human and cynomolgus monkey (monkey) CD3, yet the binding interaction is not detectable by in vitro assays known in the art. The STEAP2-binding arm may comprise any of the HCVR / LCVR or CDR amino acid sequences set forth in Table 1 herein.

[0147] According to certain exemplary embodiments, the present invention provides antibodies that specifically bind to CD3 and STEAP2. Such molecules may be referred to herein as, for example, "anti-CD3 / anti-STEAP2," or "anti-CD3 x STEAP2" or "CD3 x STEAP2" bispecific molecules, or other similar terms (e.g., anti-STEAP2 / anti-CD3).

[0148] As used herein, the term "STEAP2" refers to the human STEAP2 protein unless specified as being from a non-human species (e.g., "mouse STEAP2," "monkey STEAP2," etc.). The human STEAP2 protein has the amino acid sequence set forth in SEQ ID NO: 1899.

[0149] The bispecific antigen-binding molecules that specifically bind to CD3 and STEAP2 have a K of greater than about 40 nM as measured in an in vitro affinity binding assay. D In some cases, the CD3 binding arm may comprise an anti-CD3 antigen binding molecule that binds to CD3 with a weak binding affinity exhibiting a K of greater than about 100 nM, greater than about 200 nM, greater than about 300 nM, greater than about 400 nM, greater than about 500 nM, or greater than about 1 μM. D or EC 50(e.g., as measured in a surface plasmon resonance assay). In some cases, the first antigen-binding domain specifically binds to CD3 (e.g., either or both of human CD3 and cynomolgus monkey CD3 with weak or no measurable affinity).

[0150] As used herein, the phrase "antigen-binding molecule" refers to a protein, polypeptide, or molecular complex comprising or consisting of at least one complementarity-determining region (CDR) alone or in combination with one or more additional CDRs and / or framework regions (FRs), which specifically binds to a particular antigen. In certain embodiments, the antigen-binding molecule is an antibody or an antibody fragment, as those terms are defined elsewhere herein.

[0151] As used herein, the expression "bispecific antigen-binding molecule" refers to a protein, polypeptide, or molecular complex comprising at least a first antigen-binding domain and a second antigen-binding domain. Each antigen-binding domain in a bispecific antigen-binding molecule comprises at least one CDR that specifically binds to a particular antigen, either alone or in combination with one or more additional CDRs and / or FRs. In the context of the present invention, the first antigen-binding domain specifically binds to a first antigen (e.g., CD3), and the second antigen-binding domain specifically binds to a second, different antigen (e.g., STEAP2).

[0152] In certain exemplary embodiments of the present invention, the bispecific antigen-binding molecule is a bispecific antibody. Each antigen-binding domain of a bispecific antibody comprises a heavy chain variable domain (HCVR) and a light chain variable domain (LCVR). In the context of a bispecific antigen-binding molecule (e.g., a bispecific antibody) comprising a first and a second antigen-binding domain, the CDRs of the first antigen-binding domain may be designated with the prefix "A1", and the CDRs of the second antigen-binding domain may be designated with the prefix "A2". Thus, the CDRs of the first antigen-binding domain may be referred to herein as A1-HCDR1, A1-HCDR2, and A1-HCDR3, and the CDRs of the second antigen-binding domain may be referred to herein as A2-HCDR1, A2-HCDR2, and A2-HCDR3.

[0153] The first antigen-binding domain and the second antigen-binding domain may be directly or indirectly linked to each other to form the bispecific antigen-binding molecule of the present invention. Alternatively, the first antigen-binding domain and the second antigen-binding domain may each be linked to a separate multimerization domain. The association of one multimerization domain with another multimerization domain promotes the association between the two antigen-binding domains, thereby forming the bispecific antigen-binding molecule. As used herein, a "multimerization domain" refers to a first antigen-binding domain or a second antigen-binding domain of the same or similar structure or configuration. The multimerization domain may be any macromolecule, protein, polypeptide, peptide, or amino acid capable of associating with two multimerization domains. For example, the multimerization domain may be an immunoglobulin C multimer. H A non-limiting example of a multimerizing component is a polypeptide comprising the Fc portion of an immunoglobulin (C H 2-C H 3 domains), such as the Fc domain of IgG selected from the isotypes IgG1, IgG2, IgG3, and IgG4, as well as allotypes within each isotype group.

[0154] The bispecific antigen-binding molecules of the present invention typically comprise two multimerization domains, e.g., two Fc domains that are each part of a separate antibody heavy chain. The first and second multimerization domains may be of the same IgG isotype, e.g., IgG1 / IgG1, IgG2 / IgG2, or IgG4 / IgG4. Alternatively, the first and second multimerization domains may be of different IgG isotypes, e.g., IgG1 / IgG2, IgG1 / IgG4, or IgG2 / IgG4.

[0155] In certain embodiments, the multimerization domain is an Fc fragment or an amino acid sequence of 1 to about 200 amino acids in length containing at least one cysteine ​​residue. In other embodiments, the multimerization domain is a cysteine ​​residue or a short cysteine-containing peptide. Other multimerization domains include peptides or polypeptides comprising or consisting of leucine zipper, helix-loop-helix, or coiled-coil motifs.

[0156] Any bispecific antibody format or technology can be used to generate the bispecific antigen-binding molecules of the present invention. For example, an antibody or fragment thereof with a first antigen-binding specificity can be functionally linked (e.g., by chemical bonding, genetic fusion, non-covalent bonding, etc.) to one or more other molecular entities, such as other antibodies or antibody fragments with a second antigen-binding specificity, to generate the bispecific antigen-binding molecule. Specific exemplary bispecific formats that can be used in the context of the present invention include, for example, scFv-based or diabody bispecific formats, IgG-scFv fusions, dual variable domain (DVD)-Ig, Quadroma, knobs-into-holes, common light chains (e.g., common light chains with knobs-into-holes), CrossMab, CrossFab, (SEED) bodies, leucine zippers, Duobodies, IgG1 / IgG2, dual acting Fab (DAF)-IgG, and Mab. 2These include, but are not limited to, bispecific formats (for a review of the aforementioned formats, see, e.g., Klein et al. 2012, mAbs 4:6, 1-11, and references cited therein).

[0157] In the context of the bispecific antigen-binding molecules of the present invention, the multimerization domain, e.g., the Fc domain, may contain one or more amino acid changes (e.g., insertions, deletions, or substitutions) compared to a wild-type, naturally occurring Fc domain. For example, the present invention includes bispecific antigen-binding molecules containing one or more modifications in the Fc domain that result in an altered Fc domain with modified binding interactions (e.g., enhanced or decreased) between Fc and FcRn. In one embodiment, the bispecific antigen-binding molecule comprises a C H 2 or C H The Fc domain contains modifications in three regions that enhance the affinity of the Fc domain for FcRn in an acidic environment (e.g., in the endosome at a pH range of about 5.5 to about 6.0). Non-limiting examples of such Fc modifications include modifications at positions 250 (e.g., E or Q), 250 and 428 (e.g., L or F), 252 (e.g., L / Y / F / W or T), 254 (e.g., S or T), and 256 (e.g., S / R / Q / E / D or T), or modifications at positions 428 and / or 433 (e.g., L / R / S / P / Q or K) and / or 434 (e.g., H / F or Y), or modifications at positions 250 and / or 428, or at positions 307 or 308 (e.g., 308F, V308F), and 434. In one embodiment, the modifications include a 428L (e.g., M428L) and a 434S (e.g., N434S) modification, a 428L, 259I (e.g., V259I), and 308F (e.g., V308F) modification, a 433K (e.g., H433K) and a 434 (e.g., 434Y) modification, a 252, 254, and 256 (e.g., 252Y, 254T, and 256E) modification, a 250Q and a 428L modification (e.g., T250Q and M428L), a 307 and / or a 308 modification (e.g., 308F or 308P).

[0158] The present invention also provides a first C H 3 domain and second IgC H a bispecific antigen-binding molecule comprising three domains, a first and a second IgC H The three domains differ from each other by at least one amino acid, and the at least one amino acid difference reduces binding of the bispecific antibody to Protein A compared to a bispecific antibody lacking that amino acid difference. H The 3 domain binds protein A and the second IgC H The 3 domain contains mutations that reduce or abolish Protein A binding, such as the H95R modification (according to IMGT exon numbering, H435R in EU numbering). H 3 may further comprise a Y96F modification (according to IMGT, Y436F in the EU). See, e.g., U.S. Patent No. 8,586,713. H Further modifications that may be found within 3 include D16E, L18M, N44S, K52N, V57M, and V82I for IgG1 antibodies (D356E, L358M, N384S, K392N, V397M, and V422I by EU as per IMGT), N44S, K52N, and V82I for IgG2 antibodies (N384S, K392N, and V422I by EU as per IMGT), and Q15R, N44S, K52N, V57M, R69K, E79Q, and V82I for IgG4 antibodies (Q355R, N384S, K392N, V397M, R409K, E419Q, and V422I by EU as per IMGT).

[0159] In certain embodiments, the Fc domain can be a chimera that combines Fc sequences from two or more immunoglobulin isotypes. For example, a chimeric Fc domain can be a chimeric Fc domain that combines the Fc sequences of human IgG1, human IgG2, or human IgG4. H C derived from 2 regions H Part or all of the 2 sequence, and C derived from human IgG1, human IgG2, or human IgG4H The chimeric Fc domain may comprise some or all of the three sequences. The chimeric Fc domain may also comprise a chimeric hinge region. For example, the chimeric hinge may comprise an "upper hinge" sequence derived from a human IgG1 hinge region, a human IgG2 hinge region, or a human IgG4 hinge region combined with a "lower hinge" sequence derived from a human IgG1 hinge region, a human IgG2 hinge region, or a human IgG4 hinge region. A specific example of a chimeric Fc domain that may be comprised in any of the antigen-binding molecules described herein is a chimeric Fc domain that comprises, from the N-terminus to the C-terminus, a chimeric Fc domain consisting of [IgG4C H Another example of a chimeric Fc domain that may be included in any of the antigen-binding molecules described herein comprises, from the N-terminus to the C-terminus, [IgG1C H 1]-[IgG1 upper hinge]-[IgG2 lower hinge]-[IgG4 CH2]-[IgG1 CH3]. These and other examples of chimeric Fc domains that may be included in any of the antigen-binding molecules of the invention are described in U.S. Patent Publication No. 2014 / 0243504, published August 28, 2014, and incorporated herein in its entirety. Chimeric Fc domains having these general structural arrangements, and variants thereof, can alter Fc receptor binding, which in turn affects Fc effector function.

[0160] In certain embodiments, the invention provides a method for the preparation of a heavy chain constant region (CH) region having an identical identity to any one of SEQ ID NO:1911, SEQ ID NO:1912, SEQ ID NO:1913, SEQ ID NO:1914, SEQ ID NO:1915, SEQ ID NO:1916, SEQ ID NO:1917, SEQ ID NO:1918, SEQ ID NO:1919, or SEQ ID NO:1920, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, at least 100%, at least 101%, at least 102%, at least 103%, at least 104%, at least 105%, at least 106%, at least 107%, at least 108%, at least 109%, at least 1109%, at least 1111, at least 1121, In some embodiments, the antibody heavy chain comprises an amino acid sequence at least 98%, at least 99%, identical to SEQ ID NO: 1911, SEQ ID NO: 1912, SEQ ID NO: 1913, SEQ ID NO: 1914, SEQ ID NO: 1915, SEQ ID NO: 1916, SEQ ID NO: 1917, SEQ ID NO: 1918, SEQ ID NO: 1919, and SEQ ID NO: 1920.

[0161] In other embodiments, the invention provides an antibody heavy chain wherein the Fc domain comprises an amino acid sequence at least 95%, at least 96%, at least 97%, at least 98%, or at least 99% identical to any one of SEQ ID NO: 1921, SEQ ID NO: 1922, SEQ ID NO: 1923, SEQ ID NO: 1924, SEQ ID NO: 1925, SEQ ID NO: 1926, SEQ ID NO: 1927, SEQ ID NO: 1928, SEQ ID NO: 1929, or SEQ ID NO: 1930. In some embodiments, the Fc domain comprises an amino acid sequence selected from the group consisting of SEQ ID NO: 1921, SEQ ID NO: 1922, SEQ ID NO: 1923, SEQ ID NO: 1924, SEQ ID NO: 1925, SEQ ID NO: 1926, SEQ ID NO: 1927, SEQ ID NO: 1928, SEQ ID NO: 1929, and SEQ ID NO: 1930.

[0162] Sequence variants The antibodies and bispecific antigen-binding molecules of the present invention may contain one or more amino acid substitutions, insertions, and / or deletions in the framework and / or CDR regions of the heavy and light chain variable domains compared to the corresponding germline sequences from which the individual antigen-binding domains were derived. Such mutations can be readily identified by comparing the amino acid sequences disclosed herein with germline sequences available, for example, from public antibody sequence databases. The antigen-binding molecules of the present invention may comprise an antigen-binding domain derived from any of the exemplary amino acid sequences disclosed herein, in which one or more amino acids in one or more framework and / or CDR regions are mutated to the corresponding residue(s) in the germline sequence from which the antibody was derived, or to the corresponding residue(s) in another human germline sequence, or to a conservative amino acid substitution of the corresponding germline residue(s) (such sequence changes are collectively referred to herein as "germline mutations"). Starting from the heavy and light chain variable region sequences disclosed herein, one skilled in the art can easily produce numerous antibodies and antibody-binding fragments containing one or more individual germline mutations or combinations thereof. In certain embodiments, V H Domain and / or V LAll framework and / or CDR residues within the domain are mutated back to the residue found in the original germline sequence from which the antigen-binding domain was derived. In other embodiments, only certain residues are mutated back to the original germline sequence, e.g., the mutated residue is found within the first 8 amino acids of FR1, or the mutated residue is found within the last 8 amino acids of FR4, or the mutated residue is found only in CDR1, CDR2, or CDR3. In other embodiments, one or more of the framework and / or CDR residue(s) are mutated to the corresponding residue(s) in a different germline sequence (i.e., a germline sequence that differs from the germline sequence from which the antigen-binding domain was originally derived). Furthermore, the antigen-binding domain may contain any combination of two or more germline mutations within the framework and / or CDR regions, e.g., certain individual residues are mutated to the corresponding residue in a particular germline sequence, while certain other residues that differ from the original germline sequence are maintained or mutated to the corresponding residue in a different germline sequence. Once obtained, antigen-binding domains containing one or more germline mutations can be readily tested for one or more desired properties, such as improved binding specificity, increased binding affinity, improved or enhanced antagonist or agonist biological properties (as the case may be), reduced immunogenicity, etc. Bispecific antigen-binding molecules comprising one or more antigen-binding domains obtained by this general method are encompassed within the scope of the present invention.

[0163] The present invention also provides a method for the preparation of a polypeptide comprising one or both antigen-binding domains having one or more conservative substitutions. The present invention includes antigen-binding molecules comprising variants of any of the HCVR, LCVR, and / or CDR amino acid sequences disclosed herein. For example, the present invention includes antigen-binding molecules comprising an antigen-binding domain having an HCVR, LCVR, and / or CDR amino acid sequence with, for example, 10 or less, 8 or less, 6 or less, or 4 or less conservative amino acid substitutions relative to any of the HCVR, LCVR, and / or CDR amino acid sequences disclosed herein. A "conservative amino acid substitution" is one in which an amino acid residue is replaced with another amino acid residue having a side chain (R group) with similar chemical properties (e.g., charge or hydrophobicity). Generally, conservative amino acid substitutions will not substantially alter the functional properties of a protein. Examples of groups of amino acids having side chains with similar chemical properties include: (1) aliphatic side chains: glycine, alanine, valine, leucine, and isoleucine, (2) aliphatic-hydroxyl side chains: serine and threonine, (3) amide-containing side chains: asparagine and glutamine, (4) aromatic side chains: phenylalanine, tyrosine, and tryptophan, (5) basic side chains: lysine, arginine, and histidine, (6) acidic side chains: aspartic acid and glutamic acid, and (7) sulfur-containing side chains: cysteine ​​and methionine. Preferred conservative amino acid substitution groups are valine-leucine-isoleucine, phenylalanine-tyrosine, lysine-arginine, alanine-valine, glutamic acid-aspartic acid, and asparagine-glutamine. Alternatively, a conservative substitution is any change that has a positive value in the PAM250 log-likelihood matrix disclosed in Gonnet et al. (1992) Science 256:1443-1445, which is incorporated herein by reference. A "moderately conservative" substitution is any change that has a non-negative value in the PAM250 log-likelihood matrix.

[0164] The present invention also includes antigen-binding molecules comprising antigen-binding domains having HCVR, LCVR, and / or CDR amino acid sequences substantially identical to any of the HCVR, LCVR, and / or CDR amino acid sequences disclosed herein. When referring to amino acid sequences, the term "substantial identity" or "substantially identical" means that two amino acid sequences, when optimally aligned using predefined gap weights, such as by the programs GAP or BESTFIT, share at least 95% sequence identity, and even more preferably at least 98% or 99% sequence identity. Preferably, residue positions that are not identical differ by conservative amino acid substitutions. When two or more amino acid sequences differ from each other by conservative substitutions, the percentage of sequence identity or degree of similarity may be adjusted upward to correct for the conservative nature of the substitutions. Means for making this adjustment are well known to those skilled in the art. See, for example, Pearson (1994) Methods Mol. Biol. 24:307-331, incorporated herein by reference.

[0165] Sequence similarity for polypeptides, also known as sequence identity, is typically measured using sequence analysis software. Protein analysis software matches similar sequences using similarity measures assigned to various substitutions, deletions, and other modifications, including conservative amino acid substitutions. For example, GCG software contains programs such as Gap and Bestfit, which can be used with default parameters to determine sequence homology or sequence identity between closely related polypeptides, such as homologous polypeptides from different species of organisms, or between a wild-type protein and its mutant protein. See, for example, GCG version 6.1. Polypeptide sequences can also be compared using FASTA, a program in GCG version 6.1 with default or recommended parameters. FASTA (e.g., FASTA2 and FASTA3) provides alignments and percent sequence identity of the regions of best overlap between the query and search sequences (Pearson (2000) supra). Another preferred algorithm for comparing the sequences of the present invention to a database containing a large number of sequences from different organisms is the computer program BLAST, particularly BLASTP or TBLASTN, using default parameters. For example, , Altschul et al. (1990) J. Mol. Biol. 215:403-410 and Altschul et al., each of which is incorporated herein by reference. See, al. (1997) Nucleic Acids Res. 25:3389-402.

[0166] pH dependent binding The present invention includes anti-STEAP2 antibodies and anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules with pH-dependent binding properties. For example, anti-STEAP2 antibodies of the present invention may exhibit decreased binding to STEAP2 at acidic pH compared to neutral pH. Alternatively, anti-STEAP2 antibodies of the present invention may exhibit enhanced binding to STEAP2 at acidic pH compared to neutral pH. The term "acidic pH" includes pH values ​​less than about 6.2, e.g., about 6.0, 5.95, 5.9, 5.85, 5.8, 5.75, 5.7, 5.65, 5.6, 5.55, 5.5, 5.45, 5.4, 5.35, 5.3, 5.25, 5.2, 5.15, 5.1, 5.05, and 5.0 or less. As used herein, the term "neutral pH" refers to a pH of about 7.0 to about 7.4. The expression "neutral pH" includes pH values ​​of about 7.0, 7.05, 7.1, 7.15, 7.2, 7.25, 7.3, 7.35, and 7.4.

[0167] In some cases, "decreased binding at acidic pH compared to neutral pH" refers to the K of an antibody that binds to its antigen at neutral pH. D K value of an antibody that binds to its antigen at acidic pH D For example, an antibody or antigen-binding fragment thereof may have an acidic / neutral K of about 3.0 or greater. D When a ratio is presented, for purposes of the present invention, the antibody or antigen-binding fragment thereof may be considered to exhibit "reduced binding to STEAP2 at acidic pH compared to neutral pH." In certain exemplary embodiments, the acidic / neutral K of an antibody or antigen-binding fragment of the invention D The ratio can be about 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, 6.0, 6.5, 7.0, 7.5, 8.0, 8.5, 9.0, 9.5, 10.0, 10.5, 11.0, 11.5, 12.0, 12.5, 13.0, 13.5, 14.0, 14.5, 15.0, 20.0, 25.0, 30.0, 40.0, 50.0, 60.0, 70.0, 100.0 or more.

[0168] Antibodies with pH-dependent binding properties can be obtained, for example, by screening a population of antibodies for decreased (or increased) binding to a specific antigen at acidic pH compared to neutral pH. Furthermore, modification of the antigen-binding domain at the amino acid level can produce antibodies with pH-dependent characteristics. For example, by substituting one or more amino acids in the antigen-binding domain (e.g., within the CDR) with histidine residues, an antibody can be obtained that has decreased antigen binding at acidic pH compared to neutral pH.

[0169] Antibodies containing Fc variants According to certain embodiments of the present invention, there are provided anti-STEAP2 antibody bispecific antigen-binding molecules and anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules comprising an Fc domain containing one or more mutations that enhance or decrease antibody binding to the FcRn receptor, e.g., at acidic pH compared to neutral pH. For example, the present invention provides a method for the preparation of an Fc domain comprising the steps of: H 2 or C H Antibodies containing mutations in three regions, where the mutation(s) increase the affinity of the Fc domain for FcRn in acidic environments (e.g., in endosomes where the pH ranges from about 5.5 to about 6.0). Such mutations can result in increased serum half-life of the antibody when administered to an animal. Non-limiting examples of such Fc modifications include, for example, modifications at positions 250 (e.g., E or Q), 250 and 428 (e.g., L or F), 252 (e.g., L / Y / F / W or T), 254 (e.g., S or T), and 256 (e.g., S / R / Q / E / D or T), or modifications at positions 428 and / or 433 (e.g., H / L / R / S / P / Q or K) and / or 434 (e.g., H / F or Y), or modifications at positions 250 and / or 428, or modifications at positions 307 or Modifications include modifications at positions 308 (e.g., 308F, V308F), and 434. In one embodiment, modifications include a 428L (e.g., M428L) and a 434S (e.g., N434S) modification, a 428L, 259I (e.g., V259I), and 308F (e.g., V308F) modification, a 433K (e.g., H433K) and a 434 (e.g., 434Y) modification, a 252, 254, and 256 (e.g., 252Y, 254T, and 256E) modification, a 250Q and a 428L modification (e.g., T250Q and M428L), a 307 and / or a 308 modification (e.g., 308F or 308P).

[0170] For example, the present invention includes anti-STEAP2 antibodies and anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules comprising an Fc domain containing one or more pairs or groups of mutations selected from the group consisting of 250Q and 248L (e.g., T250Q and M248L), 252Y, 254T and 256E (e.g., M252Y, S254T and T256E), 428L and 434S (e.g., M428L and N434S), and 433K and 434F (e.g., H433K and N434F). All possible combinations of the aforementioned Fc domain mutations, and other mutations in the antibody variable domains disclosed herein, are contemplated within the scope of the present invention.

[0171] Biological properties of antibodies and bispecific antigen-binding molecules The present invention relates to antibodies that bind to human STEAP2 with high affinity (e.g., subnanomolar K D The present invention includes antibodies and antigen-binding fragments thereof that bind to the antibody at the target site (a value).

[0172] The present invention also includes anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules that inhibit tumor growth in immunocompromised mice bearing human prostate cancer xenografts (see, e.g., Example 5).

[0173] The present invention includes antibodies and antigen-binding fragments thereof that bind to human CD3 with high affinity. The present invention also includes antibodies and antigen-binding fragments thereof that bind to human CD3 with moderate or low affinity, depending on the therapeutic situation and the specific targeting properties desired. For example, in the context of a bispecific antigen-binding molecule in which one arm binds to CD3 and another arm binds to a target antigen (e.g., STEAP2), it is desirable for the target antigen-binding arm to bind to the target antigen with high affinity, while the anti-CD3 arm binds to CD3 with only moderate or low affinity. In this way, preferential targeting of the antigen-binding molecule to cells expressing the target antigen can be achieved while avoiding general / non-targeted CD3 binding and the resulting adverse side effects associated therewith.

[0174] The present invention includes bispecific antigen-binding molecules (e.g., bispecific antibodies) that can simultaneously bind to human CD3 and human STEAP2. The binding arms that interact with cells expressing CD3 can have weak to undetectable binding when measured in a suitable in vitro binding assay. The degree to which a bispecific antigen-binding molecule binds to cells expressing CD3 and / or STEAP2 can be assessed by fluorescence-activated cell sorting (FACS).

[0175] The present invention also provides a compound having an EC of about 1 nM to 50 nM as measured using the FACS binding assay described in Example 2 or a substantially similar assay. 50 In certain embodiments, the antibodies, antigen-binding fragments, and bispecific antibodies thereof have an EC value of about 50 nM, about 40 nM, about 30 nM, or about 20 nM, as measured using the FACS binding assay described in Example 2 or a substantially similar assay. 50 EC values ​​of less than about 15 nM, less than about 10 nM, less than about 5 nM, less than about 4 nM, less than about 3 nM, less than about 2 nM, and less than about 1 nM 50 values, STEAP2-expressing cells and cell lines (e.g., For example, CA-2 cells.

[0176] The present invention includes antibodies, antibody-binding fragments thereof, and bispecific antibodies thereof that bind to human CD3 with weak (i.e., low) or even undetectable affinity. According to certain embodiments, the present invention provides antibodies, antibody-binding fragments thereof, and bispecific antibodies thereof that bind to human CD3 with a K of greater than about 11 nM as measured by surface plasmon resonance. D In certain embodiments, the antibodies or antigen-binding fragments of the invention have a denaturing activity greater than about 15 nM, greater than about 20 nM, greater than about 25 nM, greater than about 30 nM, greater than about 35 nM, greater than about 40 nM, greater than about 45 nM, greater than about 50 nM, greater than about 55 nM, greater than about 60 nM, greater than about 65 nM, greater than about 70 nM, greater than about 75 nM, greater than about 80 nM, greater than about 85 nM, greater than about 90 nM, greater than about 95 nM, greater than about 100 nM, greater than about 110 nM, greater than about 120 nM, greater than about 130 nM, greater than about 140 nM, greater than about 150 nM, greater than about 160 nM, greater than about 170 nM, greater than about 180 nM, greater than about 190 nM, greater than about 210 nM, greater than about 220 nM, greater than about 230 nM, greater than about 240 nM, greater than about 250 nM, greater than about 260 nM, greater than about 270 nM, greater than about 280 nM, greater than about 290 nM, greater than about 300 nM, greater than about 350 nM, greater than about 350 nM, greater than about 360 nM, greater than about 370 nM, greater than about 380 nM, greater than about 390 nM, greater than about 400 nM, greater than about 450 nM, greater than about 500 nM, greater than about 550 nM, greater than about 600 nM, greater than about 650 nM, greater than about 700 nM, greater than about 75

[0033] The antibody binds to CD3 with a KD of at least 80 nM, greater than about 90 nM, greater than about 100 nM, greater than about 110 nM, at least 120 nM, greater than about 130 nM, greater than about 140 nM, greater than about 150 nM, at least 160 nM, greater than about 170 nM, greater than about 180 nM, greater than about 190 nM, greater than about 200 nM, greater than about 250 nM, greater than about 300 nM, greater than about 400 nM, greater than about 500 nM, or greater than about 1 μM, or without detectable affinity.

[0177] The present invention includes antibodies, antibody-binding fragments thereof, and bispecific antibodies thereof that bind to monkey (ie, cynomolgus) CD3 with weak (ie, low) or even undetectable affinity.

[0178] The present invention includes anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules that bind to and are internalized by human STEAP2-expressing cells (e.g., CA-2 cells) as measured by an assay format defined in Example 3 herein or a substantially similar assay. The present invention includes anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules specific for binding to human STEAP2. In a specific embodiment, an anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule of the invention binds to STEAP-2 transiently expressed in HEK293 cells as measured by an assay format defined in Example 3 herein or a substantially similar assay. In a specific embodiment, an anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule of the invention does not bind to human STEAP1, human STEAP2, or human STEAP4 transiently expressed in HEK293 cells as measured by an assay format defined in Example 3 herein or a substantially similar assay.

[0179] The present invention includes anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules capable of inhibiting C4-2 tumor growth (see, e.g., Example 5). For example, according to certain embodiments, anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules are provided in which a single administration (e.g., about a 0.1 mg / kg or about a 0.01 mg / kg dose) results in a reduction in tumor size compared to animals administered an isotype control bispecific antibody, as detected in a subject using standard caliper measurement methods, e.g., as described in Example 5 herein, measured 46 days after tumor implantation.

[0180] The present invention also includes anti-STEAP2 antibody-drug conjugates that inhibit tumor growth in an in vivo STEAP2-positive prostate cancer xenograft model (see, e.g., Example 7, or a substantially similar assay). In certain embodiments, an anti-STEAP2 antibody-drug conjugate with Compound 7 is provided, in which a single dose of 10, 20, or 40 mg / kg administered on day 13 after tumor implantation inhibits C4-2 tumor growth in an in vivo STEAP2-positive prostate cancer xenograft model. In certain embodiments, an anti-STEAP2 antibody-drug conjugate with Compound 7 is provided, in which a single dose of 5 mg / kg or 20 mg / kg administered on day 14 after implantation inhibits C4-2 tumor growth in an in vivo STEAP2-positive prostate cancer xenograft model. In certain embodiments, a single dose of 150 μg / kg administered on day 17 after implantation inhibits C4-2 tumor growth in an in vivo STEAP2-positive prostate cancer xenograft model. An anti-STEAP2 antibody-drug conjugate with Compound 7 is provided, wherein a single dose of at least 2.5 mg / kg administered on day 29 post-implantation inhibits C4-2 tumor growth in an in vivo STEAP2-positive prostate cancer xenograft model. In another embodiment, an anti-STEAP2 antibody-drug conjugate with Compound 60 is provided, wherein a single dose of at least 2.5 mg / kg administered on day 29 post-implantation inhibits C4-2 tumor growth in an in vivo STEAP2-positive prostate cancer xenograft model.

[0181] Epitope mapping and related techniques The epitopes on CD3 and / or STEAP2 to which the antigen-binding molecules of the present invention bind may consist of a single contiguous sequence of three or more (e.g., 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or more) amino acids of the CD3 or STEAP2 protein. Alternatively, the epitope may consist of multiple non-contiguous amino acids (or amino acid sequences) of CD3 or STEAP2. The antibodies of the present invention may interact with amino acids contained within a single CD3 chain (e.g., CD3-epsilon, CD3-delta, or CD3-gamma) or with amino acids on two or more different CD3 chains. As used herein, the term "epitope" refers to an antigenic determinant that interacts with a specific antigen-binding site in the variable region of an antibody molecule, known as a paratope. A single antigen may have two or more epitopes. Therefore, different antibodies may bind to different regions on an antigen and have different biological effects. Epitopes can be either conformational or linear. Conformational epitopes are generated by spatially juxtaposed amino acids from different segments of a linear polypeptide chain. Linear epitopes are generated by adjacent amino acid residues in a polypeptide chain. In certain circumstances, epitopes may include carbohydrate, phosphoryl, or sulfonyl moieties on an antigen.

[0182] Various techniques known to those skilled in the art can be used to determine whether an antibody antigen-binding domain "interacts with one or more amino acids" in a polypeptide or protein. Exemplary techniques include, for example, the routine cross-blocking assay described in Antibodies, Harlow and Lane (Cold Spring Harbor Press, Cold Spring Harb., NY), alanine scanning mutation analysis, peptide blot analysis (Reineke, 2004, Methods Mol Biol 248:443-463), and peptide truncation analysis. In addition, methods such as epitope excision, epitope extraction, and chemical modification of antigens can be employed (Tomer (2000) Prot. Sci. 9:487-496). Another method that can be used to identify the amino acid in a polypeptide that an antibody antigen-binding domain interacts with is hydrogen / deuterium exchange, detected by mass spectrometry. Generally speaking, hydrogen / deuterium exchange involves deuterium-labeling the protein of interest, and then binding the antibody to the deuterium-labeled protein. The protein / antibody complex is then transferred to water, causing hydrogen-deuterium exchange at all residues except those protected by the antibody (which remain deuterium-labeled). After dissociation of the antibody, the target protein is subjected to protease cleavage and mass spectrometry, thereby revealing the deuterium-labeled residues corresponding to the specific amino acids with which the antibody interacts. See, for example, Ehring (1999) Analytical Biochemistry 267(2):252-259; Engen and Smith (2001) Anal. Chem. 73:256A-265A. X-ray crystallography of antigen / antibody complexes can also be used for epitope mapping purposes.

[0183] The present invention further includes anti-STEAP2 antibodies (antibodies comprising any of the amino acid sequences set forth in Table 1 herein) that bind to the same epitope as any of the specific exemplary antibodies described herein. Similarly, the present invention also includes anti-STEAP2 antibodies (antibodies comprising any of the amino acid sequences set forth in Table 1 herein) that compete for binding to STEAP2 with any of the specific exemplary antibodies described herein. The present invention also includes antibodies containing any of the amino acid sequences of the present invention.

[0184] The present invention also includes bispecific antigen-binding molecules comprising a first antigen-binding domain that specifically binds to human CD3 and / or cynomolgus CD3 with low or detectable binding affinity, and a second antigen-binding domain that specifically binds to human STEAP2, wherein the first antigen-binding domain binds to the same epitope on CD3 as any of the specific exemplary CD3-specific antigen-binding domains described herein, and / or the second antigen-binding domain binds to the same epitope on STEAP2 as any of the specific exemplary STEAP2-specific antigen-binding domains described herein.

[0185] Similarly, the present invention also includes bispecific antigen-binding molecules comprising a first antigen-binding domain that specifically binds human CD3 and a second antigen-binding domain that specifically binds human STEAP2, wherein the first antigen-binding domain competes for binding to CD3 with any of the specific exemplary CD3-specific antigen-binding domains described herein and / or the second antigen-binding domain competes for binding to STEAP2 with any of the specific exemplary STEAP2-specific antigen-binding domains described herein.

[0186] Whether a particular antigen-binding molecule (e.g., an antibody) or its antigen-binding domain binds to the same epitope as a reference antigen-binding molecule of the present invention or competes with the reference antigen-binding molecule of the present invention for binding can be easily determined using routine methods known in the art. For example, to determine whether a test antibody binds to the same epitope on STEAP2 (or CD3) as a reference bispecific antigen-binding molecule of the present invention, the reference bispecific molecule is first bound to the STEAP2 protein (or CD3 protein). The ability of the test antibody to bind to the STEAP2 (or CD3) molecule is then evaluated. If the test antibody can bind to STEAP2 (or CD3) after saturation binding with the reference bispecific antigen-binding molecule, it can be concluded that the test antibody binds to a different epitope on STEAP2 (or CD3) from the reference bispecific antigen. On the other hand, if the test antibody cannot bind to STEAP2 (or CD3) after saturation binding with the reference bispecific antigen-binding molecule, the test antibody may bind to the same epitope on STEAP2 (or CD3) as the epitope bound by the reference bispecific antigen-binding molecule of the present invention. Further routine experiments (e.g., peptide mutation and binding analysis) can then be performed to confirm whether the observed loss of binding of the test antibody is indeed due to binding to the same epitope as the reference bispecific antigen-binding molecule, or whether steric blocking (or another phenomenon) is responsible for the observed loss of binding. This type of experiment can be performed using ELISA, RIA, Biacore, flow cytometry, or any other quantitative or qualitative antibody binding assay available in the art. According to certain embodiments of the present invention, two antigen-binding proteins bind to the same (or overlapping) epitope if, for example, a 1-fold, 5-fold, 10-fold, 20-fold, or 100-fold excess of one antigen-binding protein inhibits binding of the other by at least 50%, but preferably 75%, 90%, or even 99%, as measured in a competitive binding assay.(See, e.g., Junghans et al., Cancer Res. 1990:50:1495-1502.) Alternatively, two antigen-binding proteins are considered to bind the same epitope if essentially all amino acid mutations in the antigen that reduce or eliminate binding of one antigen-binding protein also reduce or eliminate binding of the other. Two antigen-binding proteins are considered to have "overlapping epitopes" if only a subset of the amino acid mutations that reduce or eliminate binding of one antigen-binding protein also reduce or eliminate binding of the other.

[0187] To determine whether an antibody or its antigen-binding domain competes for binding with a reference antigen-binding molecule, the above-described binding method is performed in two ways. In the first way, the reference antigen-binding molecule is bound to the STEAP2 protein (or CD3 protein) under saturating conditions. In the first orientation, the test antibody is allowed to bind to STEAP2 (or CD3) molecules under saturating conditions, and then the binding of the test antibody to the STEAP2 (or CD3) molecule is evaluated. In the second orientation, the test antibody is allowed to bind to STEAP2 (or CD3) molecules under saturating conditions, and then the binding of the reference antigen-binding molecule to STEAP2 (or CD3) molecules is evaluated. In both orientations, if only the first (saturating) antigen-binding molecule can bind to STEAP2 (or CD3) molecules, it is concluded that the test antibody and the reference antigen-binding molecule compete for binding to STEAP2 (or CD3). As will be recognized by those skilled in the art, an antibody that competes for binding with a reference antigen-binding molecule does not necessarily bind to the same epitope as the reference antibody, but may sterically block the binding of the reference antibody by binding to an overlapping or adjacent epitope.

[0188] Preparation of antigen-binding domains and construction of bispecific molecules Antigen-binding domains specific for a particular antigen can be prepared by any antibody production technique known in the art. Once obtained, two different antigen-binding domains specific for two different antigens (e.g., CD3 and STEAP2) can be appropriately positioned relative to one another to produce a bispecific antigen-binding molecule of the present invention using conventional methods. (A discussion of exemplary bispecific antibody formats that can be used to construct bispecific antigen-binding molecules of the present invention is provided elsewhere herein.) In certain embodiments, one or more individual components (e.g., heavy and light chains) of a multispecific antigen-binding molecule of the present invention are derived from a chimeric, humanized, or fully human antibody. Methods for producing such antibodies are well known in the art. For example, one or more heavy and / or light chains of a bispecific antigen-binding molecule of the present invention can be prepared using VELOCIMMUNE™ technology. Using VELOCIMMUNE™ technology (or any other human antibody production technique), a high-affinity chimeric antibody against a specific antigen (e.g., CD3 or STEAP2) is first isolated with a human variable region and a mouse constant region. Antibodies are characterized and selected for desirable characteristics, including affinity, selectivity, epitope, etc. The mouse constant regions are replaced with the desired human constant regions to produce fully human heavy and / or light chains that can be incorporated into the bispecific antigen-binding molecules of the invention.

[0189] Genetically engineered animals can be used to produce human bispecific antigen-binding molecules. For example, genetically modified mice can be used that are unable to rearrange and express endogenous mouse immunoglobulin light chain variable sequences, and the mice express only one or two human light chain variable domains encoded by human immunoglobulin sequences operably linked to mouse kappa constant genes at the endogenous mouse kappa locus. Such genetically modified mice can be used to produce fully human bispecific antigen-binding molecules containing two different heavy chains associated with the same light chain containing variable domains derived from one of two different human light chain variable region gene segments (see, for example, US2011 / 0195454). "Fully human" refers to an antibody, or antigen-binding fragment thereof, or immunoglobulin domain that contains amino acid sequences encoded by DNA derived from human sequences across the entire length of each polypeptide of the antibody, or antigen-binding fragment thereof, or immunoglobulin domain. In some instances, fully human sequences are derived from proteins endogenous to humans. In other instances, fully human proteins or protein sequences include chimeric sequences in which each component sequence is derived from a human sequence. Without being bound by any theory, chimeric proteins or sequences are generally designed to minimize the generation of immunogenic epitopes at the junctions of the component sequences, for example, compared to any wild-type human immunoglobulin region or domain.

[0190] biological equivalent The present invention encompasses antigen-binding molecules having amino acid sequences that differ from those of the exemplary molecules disclosed herein but that retain the ability to bind to CD3 and / or STEAP2. Such variant antibodies may have one or more additions, deletions, or substitutions of amino acids compared to the parent sequence. The bispecific antigen-binding molecules described herein may contain any of the following substitutions, but exhibit biological activity that is essentially equivalent to the biological activity of the bispecific antigen-binding molecules described.

[0191] The present invention includes antigen-binding molecules that are biologically equivalent to any of the exemplary antigen-binding molecules described herein. Two antigen-binding proteins or antibodies are considered to be biologically equivalent if, for example, they are pharmaceutical equivalents or pharmaceutical alternatives that do not show significant differences in absorption rate and extent when administered at the same molar dose, either in a single dose or multiple doses, under similar experimental conditions. Some antigen-binding proteins are considered to be equivalents or pharmaceutical alternatives when their absorption extent is equivalent but their absorption rate is not, and such differences in absorption rate are intentional and reflected in the label, so they can be considered biologically equivalent, for example, they are not essential for achieving effective body drug concentrations in long-term use and are not considered medically significant for the particular drug product tested.

[0192] In one embodiment, two antigen binding proteins are bioequivalent if there are no clinically significant differences in their safety, purity, or potency.

[0193] In one embodiment, two antigen binding proteins are bioequivalent if a patient can be switched one or more times compared to therapy continued without switching between the reference product and the biological product without an expected increase in risk of adverse effects, including clinically significant changes in immunogenicity or reduced efficacy.

[0194] In one embodiment, two antigen binding proteins are biologically equivalent if they both act by a common mechanism or mode of action for a condition or condition of use, to the extent that such mechanism is known.

[0195] Bioequivalence may be demonstrated by in vivo and in vitro methods. Bioequivalence measurements include, for example, (a) in vivo studies in humans or other mammals in which the concentration of an antibody or its metabolites is measured as a function of time in blood, plasma, serum, or other biological fluids, (b) in vitro studies that correlate with and reasonably predict human bioavailability data, (c) in vivo studies in humans or other mammals in which the relevant acute pharmacological effect of the antibody (or its target) is measured as a function of time, and (d) well-controlled clinical trials that establish the safety, efficacy, or bioavailability or bioequivalence of the antigen-binding protein.

[0196] Biologically equivalent variants of the exemplary bispecific antigen-binding molecules shown herein can be constructed, for example, by making various substitutions of residues or sequences or by deleting terminal or internal residues or sequences that are not required for biological activity. For example, cysteine ​​residues that are not essential for biological activity can be deleted or substituted with other amino acids to prevent the formation of unnecessary or incorrect intramolecular disulfide bridges during renaturation. In other contexts, biologically equivalent antigen-binding proteins can include variants of the exemplary bispecific antigen-binding molecules described herein that contain amino acid changes that modify the glycosylation characteristics of the molecule, for example, mutations that eliminate or remove glycosylation.

[0197] Species selectivity and species cross-reactivity According to certain embodiments of the present invention, antigen-binding molecules that bind to human CD3 but not to CD3 from other species are provided. Antigen-binding molecules that bind to human STEAP2 but not to STEAP2 from other species are also provided. The present invention also includes antigen-binding molecules that bind to human CD3 and CD3 from one or more non-human species, and / or antigen-binding molecules that bind to human STEAP2 and STEAP2 from one or more non-human species.

[0198] According to certain exemplary embodiments of the present invention, there are provided antigen-binding molecules that bind to human CD3 and / or human STEAP2, and optionally bind or do not bind to one or more of mouse, rat, guinea pig, hamster, gerbil, pig, cat, dog, rabbit, goat, sheep, cow, horse, camel, cynomolgus monkey, marmoset, rhesus monkey, or chimpanzee CD3 and / or STEAP2. For example, in certain exemplary embodiments of the present invention, there is provided a bispecific antigen-binding molecule comprising a first antigen-binding domain that binds to human CD3 and cynomolgus monkey CD3, and a second antigen-binding domain that specifically binds to human STEAP2.

[0199] Antibody-drug conjugates (ADCs) The present invention provides antibody-drug conjugates (ADCs) comprising an anti-STEAP2 antibody or antigen-binding fragment thereof conjugated to a therapeutic moiety, such as a cytotoxic agent, a chemotherapeutic agent, an immunosuppressant, or a radioisotope. In general terms, ADCs comprise an antibody-drug conjugate (ADC) comprising an A-[LP] y wherein A is an antigen-binding molecule, e.g., an anti-STEAP2 antibody, or a fragment thereof (e.g., a fragment comprising at least one HCDR3 selected from any of the HCDR3 amino acid sequences listed in Table 1), L is a linker, P is a payload or therapeutic moiety (e.g., a cytotoxic agent), and y is an integer between 1 and 30. In various embodiments, the ADC comprises an anti-STEAP2 antibody or antigen-binding fragment thereof comprising CDRs of HCVRs and LCVRs having the amino acid sequences of SEQ ID NOs: 2 and 10 listed in Table 1, or a specific HCVR / LCVR pair (e.g., SEQ ID NO: 2 / 10). In some cases, the anti-STEAP2 antibody or fragment comprises CDRs having the amino acid sequences of SEQ ID NOs: 4-6-8-12-14-16 listed in Table 1. In some cases, the anti-STEAP2 antibody or fragment comprises an HCVR and LCVR having the amino acid sequences of SEQ ID NOs: 2 and 10 listed in Table 1, or a specific amino acid sequence pair (e.g., SEQ ID NO: 2 / 10).

[0200] Cytotoxic agents include any agent that is harmful to cell growth, viability, or proliferation. The antigen-binding molecules or antibodies of the present invention deliver these cytotoxic agents, referred to herein as "payloads," to target cells. Examples of suitable cytotoxic agents and chemotherapeutic agents for forming ADCs are known in the art.

[0201] Examples of suitable cytotoxic and chemotherapeutic agents that can be conjugated to anti-STEAP2 antibodies according to this aspect of the invention include, for example, 1-(2chloroethyl)-1,2-dimethanesulfonylhydrazide, 1,8-dihydroxy-bicyclo[7.3.1]trideca-4,9-diene-2,6-diyn-13-one, 1-dehydrotestosterone, 5-fluorouracil, 6-mercaptopurine, 6-thioguanine, 9-aminocamptothecin, actinomycin D, amanitin, aminopterin, anguidine, anthracycline, anthramycin (AMC), auristatin (monomethyl auristatin E or monomethyl auristatin F), bleomycin, busulfan, butyric acid, calicheamicin, camptothecin, carminomycin, carmustine, cemadotin, cisplatin, colchicine, combretastatin, cyclophosphite, amide, cytarabine, cytochalasin B, dactinomycin, daunorubicin, decarbazine, diacetoxypentyl doxorubicin, dibromomannitol, dihydroxyanthracin dione, disorazole, dolastatin, doxorubicin, duocarmycin, echinomycin, eleutherobin, emetine, epothilone, esperamicin, estramustine, ethidium bromide, etoposide, fluorouracil Rasil, geldanamycin, gramicidin D, glucocorticoids, irinotecan, leptomycin, leurosine, lidocaine, lomustine (CCNU), maytansinoids, mechlorethamine, melphalan, mercaptopurine, methopterin, methotrexate, mithramycin, mitomycin, mitoxantrone, N8-acetylspermidine, podophyllotoxin, procaine, propranolol, pteridine, pyridoxine Included are thromycin, rhizoxin, streptozotocin, tallysomycin, taxol, tenoposide, tetracaine, thioepachlorambucil, tomaymycin, topotecan, tubulysin, vinblastine, vincristine, vindesine, vinorelbine, and derivatives of any of the above.

[0202] According to certain embodiments, the cytotoxic agent conjugated to the anti-STEAP2 antibody is an auristatin such as monomethylauristatin E (MMAE) or monomethylauristatin F (MMAF), a tubulysin such as TUB-OH or TUB-OMOM, a tomaymycin derivative, a dolastatin derivative, or a maytansinoid such as DM1 or DM4. In some embodiments, the cytotoxic agent is a maytansinoid having the structure of Formula (I), including stereoisomers of compounds of Formula (I). [ka] wherein A is arylene or heteroarylene.

[0203] In some embodiments, A is a divalent radical of an optionally substituted benzene, pyridine, naphthalene, or quinolone.

[0204] In some embodiments, A is arylene.

[0205] In some embodiments, A is [ka] and During the ceremony, R 1 is independently at each occurrence alkyl, alkenyl, alkynyl, aryl, alkaryl, aralkyl, halo, heteroaryl, heterocycloalkyl, hydroxyl, cyano, nitro, [ka] or azide, In the formula, RA is alkyl or heteroalkyl; n is an integer from 0 to 4, m is an integer from 0 to 3; p is an integer from 0 to 6, q is an integer of 0 to 5.

[0206] In some embodiments, the compound of formula I is [ka] [ka] [ka] is selected from the group consisting of:

[0207] In one embodiment, the compound of formula (I) is [ka] is.

[0208] In some embodiments, the maytansinoid of Formula (I) is attached to the anti-STEAP2 antibody or antigen-binding fragment thereof via a linker, as shown in Formula (IA) below. [ka] During the ceremony, A is arylene or heteroarylene as described above in relation to formula (I); L is a linker, BA is an anti-STEAP2 antibody or an antigen-binding fragment thereof, k is an integer from 1 to 30.

[0209] In various embodiments, L is: [ka] During the ceremony, SP is a spacer, [ka] is one or more binding to an anti-STEAP2 antibody or fragment thereof, AA 1 is an amino acid, AA 2 is an amino acid.

[0210] In some embodiments, AA 1 -AA 2 is valine-citrulline, citrulline-valine, lysine-phenylalanine, phenylalanine-lysine, valine-asparagine, asparagine-valine, threonine-asparagine, asparagine-threonine, serine-asparagine, asparagine-serine, phenylalanine-asparagine, asparagine-phenylalanine, leucine-asparagine, asparagine-leucine, isoleucine-asparagine, asparagine-isoleucine, glycine-asparagine, asparagine-glycine, glutamate-asparagine, asparagine-glutamate, citrulline-asparagine, asparagine-citrulline, alanine-asparagine, or asparagine-alanine.

[0211] In some embodiments, SP is [ka] and During the ceremony, [ka] is binding to an anti-STEAP2 antibody or fragment thereof, b is an integer of 2 to 8.

[0212] In other embodiments, L is [ka] and During the ceremony, [ka] is binding to an anti-STEAP2 antibody or fragment thereof, b is an integer of 2 to 8.

[0213] In one embodiment, the compound of formula (IA) comprising a linker attached to an anti-STEAP2 antibody or antigen-binding fragment thereof is of the formula: [ka] During the ceremony, [ka] is the binding to an anti-STEAP2 antibody or fragment thereof. Sometimes this moiety is referred to as "compound 10."

[0214] In one embodiment, the compound of formula (IA) comprising a linker attached to an anti-STEAP2 antibody or antigen-binding fragment thereof is of the formula: [ka] During the ceremony, [ka] is the conjugation to an anti-STEAP2 antibody or fragment thereof. Sometimes this moiety is referred to as "compound 60."

[0215] In some embodiments, the cytotoxic agent is a maytansinoid having the structure of Formula (II), including stereoisomers of the compound of Formula (II). [ka] During the ceremony, A 3arepresents an amino acid, a peptide having 2-20 amino acids, alkyl, alkynyl, alkenyl, cycloalkyl, aryl, heteroaryl, heterocyclyl, -CR5R6-, -O-, -C(=O)-, -OC(=O)-, -C(=O)-O-, -OC(=O)-O-, -C(=O)-(CH x)p1 -, -C(=O)-O-(CH x)p1 -, -(CH x)p1 -C(=O)-, -(CH x)p1 -C(=O)-O-, -(O-(CH2) p2 -) p3 -, -((CH2) p2 -O-) p3 -, -C(=S)-, -C(=S)-S-, -C(=S)-NH-, -SC(=S)-, -SC(=S)-S-, -S-, -SO-, -SO2- , -NR4-, -N(R4)-C(=O)-N(R8)-, -N(R4)-C(=O)O-, -N(R4)-C(=O)-, -C(=O)-N( R4)-, -C(=O)-N(R4)-C(=O)-, or -OC(=O)-NR4-, wherein alkyl, alkynyl, alkenyl, cycloalkyl, aryl, heteroaryl, and heterocyclyl are optionally substituted; p1, p2, and p3 each independently represent 0 or an integer of 1 to 100; x is 0, 1, or 2; R4, R5, R6, and R8 are each independently H or substituted or unsubstituted alkyl, alkenyl, alkynyl, aryl, heteroaryl, or heterocyclyl; R 4a is a substituted or unsubstituted alkyl, alkenyl, alkynyl, aryl, heteroaryl, or heterocyclyl.

[0216] In some embodiments, the compound of formula (II) is [ka] is selected from the group consisting of:

[0217] In one embodiment, the compound of formula (II) is [ka] is.

[0218] In some embodiments, the maytansinoid of formula (II) is attached to the anti-STEAP2 antibody or antigen-binding fragment thereof via a linker, as shown in formula (IIA) below. [ka] During the ceremony, BA is an anti-STEAP2 antibody or an antigen-binding fragment thereof, a is an integer from 1 to 30, Z2 is represented by the following structural formula -Z 2A -Z 2B -Z 2C -Z 2D wherein Z 2A , Z 2B , Z 2C , and Z 2D are each independently absent or represent an amino acid, a peptide having 2 to 20 amino acids, alkyl, alkynyl, alkenyl, cycloalkyl, aryl, heteroaryl, heterocyclyl, -CR5R6-, -O-, -C(=O)-, -OC(=O)-, -C(=O)-O-, -OC(=O)-O-, -C(=O)-(CH x ) p1 , -C(=O)-O-(CH x ) p1 , -(CH x ) p1 -C(=O)-, -(CH x ) p1 -C(=O)-O-, -(O-(CH2) p2 -) p3 -, -((CH2) p2 )-O-) p3-, -C(=S)-, -C(=S)-S-, -C(=S)-NH-, -SC(=S)-, -SC(=S)-S-, -S-, -SO-, -SO2-, -NR4-, -N(R4)-C(=O)-N(R8)-, -N(R4)-C(=O)O- , -N(R4)-C(=O)-, -C(=O)-N(R4)-, -C(=O)-N(R4)-C(=O)-, -OC(=O)-N(R4), -OC(=S)-N(R4)-, -C(=S)-N(R4)-, -N=C=S, -N=C=O, [ka] and A is a natural or unnatural amino acid, or a peptide containing 2 to 20 amino acids; W is -O-, -S-, -CR5R6-, or -NR4-; X is aryl, heteroaryl, cycloalkyl, or heterocyclyl, wherein the aryl, heteroaryl, cycloalkyl, and heterocyclyl are optionally substituted; In the formula, A1, A3, and R1 each independently represent an amino acid, a peptide having 2 to 20 amino acids, alkyl, alkynyl, alkenyl, cycloalkyl, aryl, heteroaryl, heterocyclyl, -CR5R6-, -O-, -C(=O)-, -OC(=O)-, -C(=O)-O-, -OC(=O)-O-, -C(=O)-(CH x ) p1 -, -C(=O)-O-(CH x ) p1 -, -(CH x ) p1 -C(=O)-, -(CH x ) p1 -C(=O)-O-, -(O-(CH2) p2 -) p3 -, -((CH2) p2 -O-) p3-, -C(=S)-, -C(=S)-S-, -SC(=S)-, -C(=S)-NH-, -SC(=S)-S-, -S-, -SO-, -SO2-, -NR4-, -N(R4)-C(=O)-N(R8)-, -N(R4)-C(=O)O-, -N(R4)-C(=O)-, -C(=O)-N(R4)-C(=O)-, or -OC(=O)-NR4-, wherein alkyl, alkynyl, alkenyl, cycloalkyl, aryl, heteroaryl, and heterocyclyl are optionally substituted; R 17 are O, S, NR 18 and CR5R6, R 18 is selected from the group consisting of H, alkyl, alkynyl, alkenyl, cycloalkyl, aryl, heteroaryl, heterocyclyl, and acyl, wherein alkyl, alkynyl, alkenyl, cycloalkyl, aryl, heteroaryl, heterocyclyl, and acyl are optionally substituted; R4, R5, R6, and R8 are each independently H or substituted or unsubstituted alkyl, alkenyl, alkynyl, aryl, heteroaryl, or heterocyclyl; R 4a is a substituted or unsubstituted alkyl, alkenyl, alkynyl, aryl, heteroaryl, or heterocyclyl; p1, p2, and p3 each independently represent 0 or an integer of 1 to 100; x is 0, 1, or 2.

[0219] In some embodiments of Formula (IIA), A is a peptide selected from the group consisting of valine-citrulline, citrulline-valine, lysine-phenylalanine, phenylalanine-lysine, valine-asparagine, asparagine-valine, threonine-asparagine, asparagine-threonine, serine-asparagine, asparagine-serine, phenylalanine-asparagine, asparagine-phenylalanine, leucine-asparagine, asparagine-leucine, isoleucine-asparagine, asparagine-isoleucine, glycine-asparagine, asparagine-glycine, glutamic acid-asparagine, asparagine-glutamic acid, citrulline-asparagine, asparagine-citrulline, alanine-asparagine, and asparagine-alanine.

[0220] In one embodiment, the anti-STEAP2 antibody or antigen-binding fragment thereof is linked to the formula (II The compound of A) is of the formula: [ka] During the ceremony, [ka] is the binding to an anti-STEAP2 antibody or fragment thereof. Sometimes this moiety is referred to as "compound 7."

[0221] In some embodiments, the cytotoxic agent conjugated to the anti-STEAP2 antibody or fragment thereof is a pure or substantially pure diastereomer of DM1, [ka] y is an integer from 1 to 0.

[0222] In another embodiment, the ADC is "A-[LP] y " structure, wherein A is an anti-STEAP2 antibody or antigen-binding fragment thereof, and [LP] is [ka] It is a mixture of them, In the formula, y is an integer of 1 to 30, [ka] is the binding to an anti-STEAP2 antibody or fragment thereof.

[0223] Other maytansinoid derivatives are discussed in WO2014 / 145090, WO2016 / 160615, and WO2015 / 031396, each of which is incorporated herein by reference in its entirety.

[0224] In some embodiments, the cytotoxic agent conjugated to the anti-STEAP2 antibody or fragment thereof is MMAE or MMAF.

[0225] Other cytotoxic agents known in the art are contemplated within the scope of the present invention, including, for example, ricin, C. difficile toxin, Pseudomonas exotoxin, diphtheria toxin, botulinum toxin, bryodin, saporin, pokeweed toxin (i.e., phytolactoxin and phytolacugegenin), and protein toxins such as those described in Sapra et al., Pharmacol. & Therapeutics, 2013, 138:452-469.

[0226] A cytotoxic agent ("payload") can be tethered to an anti-STEAP2 antigen-binding molecule or antibody of the invention via a chemical linker that covalently attaches the payload compound to the protein molecule (i.e., antibody). Exemplary embodiments of specific linkers are described above. More generally, as used herein, the term "linker" refers to any bivalent group or moiety that links, connects, or binds a binding agent (e.g., an antibody or antigen-binding fragment thereof) to a payload compound described herein. In general, binder linkers suitable for the antibody conjugates described herein include: The linker is stable enough to take advantage of the circulating half-life of the antibody while simultaneously being able to release its payload after antigen-mediated internalization of the conjugate. The linker can be cleavable or non-cleavable. A cleavable linker is one that is cleaved by intracellular metabolism following internalization, such as hydrolysis, reduction, or enzymatic cleavage. A non-cleavable linker is one that releases the attached payload through lysosomal degradation of the antibody after internalization. Suitable linkers include, but are not limited to, acid-labile linkers, hydrolytically labile linkers, enzymatically cleavable linkers, reduction-labile linkers, self-immolative linkers, and non-cleavable linkers. Suitable linkers also include, but are not limited to, those that are or contain peptides, glucuronides, succinimide-thioethers, polyethylene glycol (PEG) units, hydrazones, mal-caproyl units, dipeptide units, valine-citrulline units, and para-aminobenzyl (PAB) units. In some cases, the linker can be attached to the antibody or antigen-binding fragment through a lysine or cysteine ​​residue (e.g., via cleavage of a disulfide group in the antibody or fragment or via a cysteine ​​residue incorporated into the antibody or fragment). In some cases, the linker can be attached to the antibody or fragment through a glutamine residue, including one derived via transglutaminase-mediated linkage.

[0227] Exemplary linkers that can be used in the context of the present invention include, for example, linkers comprising or consisting of MC (6-maleimidocaproyl), MCC (maleimidomethylcyclohexane-1-carboxylate), MP (maleimidopropanoyl), val-cit (valine-citrulline), val-ala (valine-alanine), ala-phe (alanine-phenylalanine), phe-lys (phenylalanine-lysine), the dipeptide portion of a protease-cleavable linker, PAB (p-aminobenzyloxycarbonyl), SPP (N-succinimidyl 4-(2-pyridylthio)pentanoate), SMCC (N-succinimidyl 4-(N-maleimidomethyl)cyclohexane-1-carboxylate), SIAB (N-succinimidyl(4-iodo-acetyl)aminobenzoate), and variants and combinations thereof. Further examples of linkers that can be used in the context of the present invention are disclosed, for example, in U.S. Patent No. 7,754,681 and Ducrin, Bioconjugate. Chem., 2010, 21:5-13, and the references cited therein, the contents of which are incorporated herein by reference in their entirety. In some cases, the linker is or includes a self-immolative spacer, such as those described in Jin, et al., Bioorganic & Medicinal Chemistry, 2012, 20:3465-3469, and Wu, et al., Bioorganic & Medicinal Chemistry, 2016, 24:2697-2706.

[0228] The payload can be linked to the anti-STEAP2 antibody or antigen-binding fragment via a bond at a specific amino acid within the antibody or antigen-binding molecule. Exemplary amino acid bonds that can be used in the context of this aspect of the invention include, for example, lysine (see, e.g., U.S. Pat. No. 5,208,020, US 2010 / 0129314, Hollander et al., Bioconjugates Chem., 2008, 19:358-361, WO2005 / 089808, U.S. Pat. No. 5,714,586, US2013 / 0101546, and US2012 / 0585592), cysteine ​​(see, e.g., US2007 / 0258987, WO2013 / 055993, WO2013 / 055990, WO2013 / 053873, WO2013 / 053872, WO2011 / 130598, US2013 / 0101546, and U.S. Pat. No. 7,750,116), selenocysteine ​​(see, e.g., WO2008 / 122039, and Hofer et al. al., Proc. Natl. Acad. Sci., USA, 2008, 105:12451-12456), formylglycine (see, e.g., Carrico et al., Nat. Chem. Bi ol., 2007,3:321-322; Agarwal et al., Proc. Natl. Acad. Sci., USA, 2013,110:46-51, and Rabuka et al., Nat. Protocols, 2012,10:1052-1067), unnatural amino acids (e.g., WO2013 / 068874 and WO2012 / 166559), and acidic amino acids (e.g., WO2012 / 05982). Linkers can also be attached to antigen-binding proteins via carbohydrate bonds (e.g., see US2008 / 0305497 and Ryan et al., Food & Agriculture Immunol., 2001,13:127-130) and disulfide linkers. (See, e.g., WO2013 / 085925, WO2010 / 010324, WO2011 / 018611, and Shaunak et al., Nat. Chem. Biol., 2006, 2:312-313).

[0229] The drug-to-antibody ratio (DAR) is the average number of drugs attached to an antibody or antigen-binding fragment and has a significant impact on the efficacy, potency, and pharmacokinetics of an ADC. In various embodiments, the DAR is 1, 2, 3, 4, 5, 6, 7, or 8 drug molecules per antibody. In some embodiments, the DAR is 1 to 4. In certain embodiments, the DAR is 2 to 4. In some cases, the DAR is 2 to 3. In some cases, the DAR is 3 to 4. In some embodiments, the DAR is 1 to 10, 1 to 20, or 1 to 30 (i.e., 1 to 30 drug molecules per antibody or antigen-binding fragment thereof).

[0230] Therapeutic Formulations and Administration The present invention provides pharmaceutical compositions comprising the antigen-binding molecules of the present invention. The pharmaceutical compositions of the present invention are formulated with appropriate carriers, excipients, and other agents that provide improved transfer, delivery, tolerance, etc. Many suitable formulations can be found in formularies known to all pharmaceutical chemists: Remington's Pharmaceutical Sciences, Mack Publishing Company, Easton, PA. These formulations include, for example, powders, pastes, ointments, jellies, waxes, oils, lipid (cationic or anionic) containing vesicles (such as LIPOFECTIN™, Life Technologies, Carlsbad, CA), DNA complexes, anhydrous absorption pastes, oil-in-water and water-in-oil emulsions, emulsion carbowax (polyethylene glycol of various molecular weights), semi-solid gels, and semi-solid mixtures containing carbowax. See also Powell et al., "Compendium of excipients for parenteral formulations," PDA (1998) J Pharm Sci Technol 52:238-311.

[0231] The dose of an antigen-binding molecule administered to a patient may vary depending on the patient's age and size, target disease, pathological condition, route of administration, etc. Preferred doses are typically calculated according to body weight or body surface area. When the bispecific antigen-binding molecule of the present invention is used for the treatment of adult patients, it may be advantageous to administer the bispecific antigen-binding molecule of the present invention intravenously in a single dose of typically about 0.01 to about 20 mg / kg body weight, more preferably about 0.02 to about 7, about 0.03 to about 5, or about 0.05 to about 3 mg / kg body weight. The frequency and duration of treatment can be adjusted depending on the severity of the condition. Effective doses and schedules for administering bispecific antigen-binding molecules can be determined empirically; for example, the patient's progress can be monitored by periodic evaluation and the dose adjusted accordingly. Furthermore, interspecies scaling of dosages can be performed using methods well known in the art (e.g., Mordenti et al., 1991, Pharmaceut. Res. 8:1351).

[0232] Various delivery systems are known and can be used to administer the pharmaceutical compositions of the present invention, such as encapsulation in liposomes, microparticles, microcapsules, recombinant cells capable of expressing mutant viruses, and receptor-mediated endocytosis (see, e.g., Wu et al. 1987, J. Biol. Chem. 262:4429-4432). Methods of introduction include, but are not limited to, intradermal, transdermal, intramuscular, intraperitoneal, intravenous, subcutaneous, intranasal, epidural, and oral routes. The compositions can be administered by any convenient route, for example, by infusion or bolus injection, or by absorption through epithelial or mucocutaneous linings (e.g., oral mucosa, rectal and intestinal mucosa, etc.), and can be administered together with other biologically active agents. Administration can be systemic or local.

[0233] The pharmaceutical composition of the present invention can be delivered subcutaneously or intravenously using a standard needle and syringe. In addition, for subcutaneous delivery, a pen delivery device is easily applicable for delivering the pharmaceutical composition of the present invention. Such a pen delivery device can be reusable or disposable. Reusable pen delivery devices generally use a replaceable cartridge containing the pharmaceutical composition. Once the pharmaceutical composition in the cartridge has been administered and the cartridge is empty, the empty cartridge can be easily discarded and replaced with a new cartridge containing the pharmaceutical composition. The pen delivery device can then be reused. In disposable pen delivery devices, there is no replaceable cartridge. Rather, the disposable pen delivery device is pre-filled with the pharmaceutical composition held in a reservoir within the device. Once the reservoir is empty of the pharmaceutical composition, the entire device is discarded.

[0234] Numerous reusable pen delivery devices and autoinjector delivery devices have application in the subcutaneous delivery of the pharmaceutical compositions of the present invention. Examples include the AUTOPEN™ (Owen Mumford, Inc., Woodstock, UK), the DISETRONIC™ pen (Disetronic Medical Systems, Bergdorf, Switzerland), the HUMALOG MIX 75 / 25™ pen, the HUMALOG™ pen, and the HUMALIN 70 / 30™ pen (Eli Lilly and Co., Indianapolis, IN), the NOVOPEN™ I, II, and III (Novo Nordisk, Copenhagen, Denmark), the NOVOPEN JUNIOR™ (Novo Nordisk, Copenhagen, Denmark), the BD™ pen (Becton Dickinson, Franklin Lakes, NJ), the OPTIPEN™, the OPTIPEN PRO™, and the OPTIPEN™. Examples of disposable pen delivery devices that have application in the subcutaneous delivery of pharmaceutical compositions of the present invention include, but are not limited to, the SOLOSTAR™ pen (Sanofi-Aventis), FLEXPEN™ (Novo), and OPTICLIK™ (Sanofi-Aventis, Frankfurt, Germany), to name just a few. Nordisk), and KWIKPEN™ (Eli Lilly), SURECLICK™ auto-injector (Amgen, Thousand Oaks, CA), PENLET™ (Haselmeier, Stuttgart, Germany), EPIPEN (Dey, LP), and HUMIRA™ pen (Abbott Labs, Abbott Park IL).

[0235] In certain circumstances, pharmaceutical compositions can be delivered in a controlled release system. In one embodiment, a pump can be used (see Langer, supra; Sefton, 1987, CRC Crit. Ref. Biomed. Eng. 14:201). In other embodiments, polymeric materials can be used. Medical Applications of Controlled Release, Langer and Wise (eds.), 1974, CRC Pres., Boca Raton, Florida. In yet another embodiment, a sustained-release system can be placed in proximity to the target of the composition, thereby requiring only a fraction of the systemic dose. (See, e.g., Goodson, 1984, in Medical Applications of Controlled Release, supra, vol. 2, pp. 115-138.) Other sustained-release systems are discussed in the review by Langer, 1990, Science 249:1527-1533.

[0236] Injectable preparations may include dosage forms for intravenous injection, subcutaneous injection, intradermal injection, intramuscular injection, drip infusion, etc. These injectable preparations may be prepared by publicly known methods. For example, injectable preparations may be prepared by dissolving, suspending, or emulsifying the above-mentioned antibody or its salt in a sterile aqueous or oily medium conventionally used for injections. Aqueous media for injection include, for example, physiological saline, glucose-containing isotonic solutions, and other adjuvants, which may be used in combination with appropriate solubilizers such as alcohols (e.g., ethanol), polyalcohols (e.g., propylene glycol, polyethylene glycol), and nonionic surfactants [e.g., polysorbate 80, HCO-50 (polyoxyethylene (50 mole) adduct of hydrogenated castor oil)]. Oily media include, for example, sesame oil and soybean oil, which may be used in combination with solubilizers such as benzyl benzoate and benzyl alcohol. Injections prepared in this manner are preferably filled into appropriate ampoules.

[0237] Advantageously, the pharmaceutical compositions for oral or parenteral use described above are prepared into unit dosage forms suitable for the dosage of the active ingredient. Such unit dosage forms include, for example, tablets, pills, capsules, injections (ampoules), suppositories, etc. The amount of antibody contained is generally about 5 to about 500 mg per unit dosage form, and particularly, in the form of injection, the amount of antibody contained is preferably about 5 to about 100 mg, and for other dosage forms, about 10 to about 250 mg.

[0238] Therapeutic Uses of Antigen-Binding Molecules The present invention includes methods comprising administering to a subject in need thereof a therapeutic composition comprising an anti-STEAP2 antibody or antigen-binding fragment thereof, or a bispecific antigen-binding molecule that specifically binds to CD3 and STEAP2. The therapeutic composition may comprise any of the antibodies or bispecific antigen-binding molecules disclosed herein and a pharmaceutically acceptable carrier or diluent. As used herein, the phrase "subject in need thereof" refers to a human or non-human animal that exhibits one or more symptoms or signs of cancer (e.g., a subject that develops a tumor or suffers from any of the cancers described herein below), or who would otherwise benefit from inhibition or reduction of STEAP2 activity or depletion of STEAP2+ cells (e.g., prostate cancer cells).

[0239] The antibodies and bispecific antigen-binding molecules of the present invention (and therapeutic compositions comprising them) are useful, inter alia, for the treatment of any disease or disorder in which stimulating, activating, and / or targeting the immune response is beneficial. In particular, the anti-STEAP2 antibodies or anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules of the present invention can be used to treat, prevent, and / or ameliorate any disease or disorder associated with or mediated by STEAP2 expression or activity or the proliferation of STEAP2+ cells. The mechanism of action achieved by the therapeutic methods of the present invention involves the killing of STEAP2-expressing cells in the presence of effector cells, for example, by CDC, apoptosis, ADCC, phagocytosis, or a combination of two or more of these mechanisms. Cells expressing STEAP2 that can be inhibited or killed using the bispecific antigen-binding molecules of the present invention include, for example, prostate tumor cells.

[0240] The antigen-binding molecules of the present invention can be used to treat primary and / or metastatic tumors occurring in, for example, the prostate, bladder, cervix, lung, colon, kidney, breast, pancreas, stomach, uterus, and / or ovary. In certain embodiments, the bispecific antigen-binding molecules of the present invention are used to treat one or more of the following cancers: prostate cancer, bladder cancer, cervical cancer, lung cancer, colon cancer, kidney cancer, breast cancer, pancreatic cancer, stomach cancer, uterine cancer, and ovarian cancer. According to certain embodiments of the present invention, anti-STEAP2 antibodies or anti-STEAP2 / anti-CD3 bispecific antibodies are useful for treating patients with castration-resistant prostate cancer. According to other related embodiments of the present invention, methods are provided comprising administering the anti-STEAP2 antibodies or anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules disclosed herein to patients with castration-resistant prostate cancer. Analytical / diagnostic methods known in the art, such as tumor scanning, can be used to determine whether a patient has a castration-resistant tumor.

[0241] The present invention also includes methods for treating residual cancer in a subject. As used herein, the term "residual cancer" refers to the presence or persistence of one or more cancerous cells in a subject after treatment with an anti-cancer therapy.

[0242] According to certain embodiments, the present invention provides methods for treating a disease or disorder associated with STEAP2 expression (e.g., prostate cancer) after a subject has been diagnosed with prostate cancer (e.g., castration-resistant prostate cancer), comprising administering to the subject one or more of the anti-STEAP2 or bispecific antigen-binding molecules described elsewhere herein. For example, the present invention includes methods for treating prostate cancer, comprising administering to the patient an anti-STEAP2 antibody or anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule 1 day, 2 days, 3 days, 4 days, 5 days, 6 days, 1 week, 2 weeks, 3 weeks, or 4 weeks, 2 months, 4 months, 6 months, 8 months, 1 year, or more after the subject has undergone hormone therapy (e.g., antiandrogen therapy).

[0243] Combination Therapies and Formulations The present invention provides methods comprising administering a pharmaceutical composition comprising any of the exemplary antibodies and bispecific antigen-binding molecules described herein in combination with one or more additional therapeutic agents. Exemplary additional therapeutic agents that can be administered in combination with or in combination with the antigen-binding molecules of the present invention include, for example, an EGFR antagonist (e.g., an anti-EGFR antibody [e.g., cetuximab or panitumumab] or a small molecule inhibitor of EGFR [e.g., gefitinib, ceftazidazole ... nib or erlotinib), Her2 / ErbB2, ErbB3, or ErbB4 antagonists of any other EGFR family member (e.g., anti-ErbB2, anti-ErbB3, or anti-ErbB4 antibodies, or small molecule inhibitors of ErbB2, ErbB3, or ErbB4 activity), antagonists of EGFRvIII (e.g., antibodies that specifically bind to EGFRvIII), cMET antagonists (e.g., anti-cMET antibodies), IGF1R antagonists (e.g., anti-IGF1R antibodies), B-raf inhibitors (e.g., vemurafenib, sorafenib, GDC-0879, PLX-4720), PDGFR-α inhibitors (e.g., anti-PDGFR-α antibodies), PDGFR-β inhibitors (e.g., anti-PDGFR-β antibodies), VEGF antagonists (e.g., VEGF traps, see e.g., U.S. Pat. No. 7,087,411 (also referred to herein as "VEGF inhibitory fusion proteins")), anti-VEGF antibodies (e.g., bevacizumab), small molecule kinase inhibitors of VEGF receptors (e.g., sunitinib, sorafenib, or pazopanib), DLL4 antagonists (e.g., anti-DLL4 antibodies disclosed in US2009 / 0142354, such as REGN421), Ang2 antagonists (e.g., anti-Ang2 antibodies disclosed in US2011 / 0027286, such as H1H685P), FOLH1 (PSMA) antagonists, PRLR antagonists (e.g., anti-PRLR antibodies), STEAP1 or STEAP2 antagonists (e.g., anti-STEAP1 antibodies or anti-STEAP2 antibodies), TMPRSS2 antagonists (e.g., anti-TMPRSS2 antibodies), MSLN antagonists (e.g., anti-MSLN antibodies), CA9 antagonists (e.g., anti-CA9 antibodies), uroplakins Other agents that may be beneficially administered in combination with the antigen-binding molecules of the present invention include cytokine inhibitors, including small molecule cytokine inhibitors and antibodies that bind to cytokines such as IL-1, IL-2, IL-3, IL-4, IL-5, IL-6, IL-8, IL-9, IL-11, IL-12, IL-13, IL-17, and IL-18, or their respective receptors. Pharmaceutical compositions of the present invention (e.g., pharmaceutical compositions comprising the anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules disclosed herein) also include anti-CD3 / anti-STEAP2 antibodies, such as "ICE": ifosfamide (e.g., Ifex®), carboplatin (e.g., Paraplatin®), etoposide (e.g., Etopophos®, Toposar®, VePesid®, VP-16), "DHAP": dexamethasone (e.g., Decadron®), cytarabine (e.g., Cytosar- U®, cytosine arabinoside, ara-C), cisplatin (e.g., Platinol®-AQ), and "ESHAP": etoposide (e.g., Etopofos®, Toposar®, VePesid®, VP-16), methylprednisolone (e.g., Medrol®), high-dose cytarabine, cisplatin (e.g., Platinol®-AQ).

[0244] The present invention also includes therapeutic combinations comprising any of the antigen binding molecules described herein and one or more inhibitors of VEGF, Ang2, DLL4, EGFR, ErbB2, ErbB3, ErbB4, EGFRvIII, cMet, IGF1R, B-raf, PDGFR-α, PDGFR-β, FOLH1 (PSMA), PRLR, STEAP1, STEAP2, TMPRSS2, MSLN, CA9, uroplakin, or any of the foregoing cytokines, wherein the inhibitor is an aptamer, antisense molecule, ribozyme, siRNA, peptibody, nanobody, or antibody fragment (e.g., Fab fragment, F(ab')2 fragment). The antigen-binding molecules of the present invention may be administered in combination with and / or co-formulated with antiviral drugs, antibiotics, analgesics, corticosteroids, and / or NSAIDs. The antigen-binding molecules of the present invention may also be administered as part of a treatment regimen that also includes radiation therapy and / or conventional chemotherapy.

[0245] The additional therapeutically active ingredient(s) may be administered immediately before, simultaneously with, or immediately after administration of the antigen-binding molecule of the present invention. (For purposes of this disclosure, such administration regimens will be considered to be administration of the antigen-binding molecule "in combination" with the additional therapeutically active ingredient.

[0246] The present invention includes pharmaceutical compositions in which the antigen-binding molecules of the present invention are co-formulated with one or more of the additional therapeutically active ingredient(s) described elsewhere herein.

[0247] Mode of administration According to certain embodiments of the invention, multiple doses of an antigen-binding molecule (e.g., an anti-STEAP2 antibody or a bispecific antigen-binding molecule that specifically binds to STEAP2 and CD3) can be administered to a subject over a predetermined period of time. The method according to this aspect of the invention comprises sequentially administering multiple doses of an antigen-binding molecule of the invention to a subject. As used herein, "sequentially administering" means that each dose of the antigen-binding molecule is administered to a subject at different time points, e.g., on different days separated by a predetermined interval (e.g., hours, days, weeks, or months). The invention includes methods comprising sequentially administering to a patient a single initial dose of the antigen-binding molecule, followed by one or more secondary doses of the antigen-binding molecule, and then optionally one or more tertiary doses of the antigen-binding molecule.

[0248] The terms "primary dose," "secondary dose," and "tertiary dose" refer to the time sequence of administration of an antigen-binding molecule of the present invention. Thus, a "primary dose" is a dose administered at the beginning of a treatment regimen (also referred to as a "baseline dose"), a "secondary dose" is a dose administered after the primary dose, and a "tertiary dose" is a dose administered after the secondary dose. The primary, secondary, and tertiary doses may all contain the same amount of antigen-binding molecule, but generally may differ from one another in terms of administration frequency. However, in certain embodiments, the amount of antigen-binding molecule contained in the primary, secondary, and / or tertiary doses differs from one another (e.g., adjusted up or down as appropriate) during the course of treatment. In certain embodiments, two or more doses (e.g., two, three, four, or five) are administered at the beginning of a treatment regimen as "loading doses," after which subsequent doses are administered less frequently (e.g., "maintenance doses").

[0249] In certain exemplary embodiments of the invention, each secondary dose and / or tertiary dose is 1 to 26 (e.g., 1, 1 1 / 2, 2, 2 1 / 2, 3, 3 1 / 2, 4, 4 1 / 2, 5, 5 1 / 2, 6, 6 1 / 2, 7, 7 1 / 2, 8, 8 1 / 2, 9, 9 1 / 2, 10, 10 1 / 2, 11, 11 1 / 2, 12, 12 1 / 2, 13 , 13 1 / 2, 14, 14 1 / 2, 15, 15 1 / 2, 16, 16 1 / 2, 17, 17 1 / 2, 18, 18 1 / 2, 19, 19 1 / 2, 20, 20 1 / 2, 21, 21 1 / 2, 22, 22 1 / 2, 23, 23 1 / 2, 24, 24 1 / 2, 25, 25 1 / 2, 26, 26 1 / 2, or more) weeks later). The phrase "immediately preceding dose" as used herein refers to a dose of an antigen-binding molecule in a multiple administration series that is administered to a patient prior to administration of the immediately following dose with no intervening doses.

[0250] The method according to this aspect of the invention may include administering any number of secondary and / or tertiary doses of an antigen-binding molecule (e.g., an anti-STEAP2 antibody or bispecific antigen-binding molecule that specifically binds STEAP2 and CD3) to the patient. For example, in certain embodiments, only a single secondary dose is administered to the patient. In other embodiments, two or more (e.g., 2, 3, 4, 5, 6, 7, 8, or more) secondary doses are administered to the patient. Similarly, in certain embodiments, only a single tertiary dose is administered to the patient. In other embodiments, two or more (e.g., 2, 3, 4, 5, 6, 7, 8, or more) tertiary doses are administered to the patient.

[0251] In embodiments comprising multiple secondary doses, each secondary dose may be administered with the same frequency as the other secondary doses. For example, each secondary dose may be administered to the patient 1-2 weeks after the immediately preceding dose. Similarly, in embodiments comprising multiple tertiary doses, each tertiary dose may be administered with the same frequency as the other tertiary doses. For example, each tertiary dose may be administered to the patient 2-4 weeks after the immediately preceding dose. Alternatively, the frequency with which the secondary and / or tertiary doses are administered to the patient may vary over the course of the treatment regimen. The administration frequency may also be adjusted by the physician during the course of treatment depending on the needs of the individual patient after clinical testing.

[0252] Diagnostic Uses of Antibodies The anti-STEAP2 antibodies of the present invention may also be used to detect and / or measure STEAP2 or STEAP2-expressing cells in a sample, e.g., for diagnostic purposes. For example, an anti-STEAP2 antibody, or a fragment thereof, may be used to diagnose a condition or disease characterized by abnormal expression (e.g., overexpression, underexpression, lack of expression, etc.) of STEAP2. An exemplary diagnostic assay for STEAP2 may include, for example, contacting a sample obtained from a patient with an anti-STEAP2 antibody of the present invention, where the anti-STEAP2 antibody is labeled with a detectable label or reporter molecule. Alternatively, an unlabeled anti-STEAP2 antibody can be used for diagnostic purposes in combination with a secondary antibody that is itself detectably labeled. The detectable label or reporter molecule may be 3 H, 14 C. 32 P, 35 S or 125 The antibody may be a radioisotope such as I, a fluorescent or chemiluminescent moiety such as fluorescein isothiocyanate or rhodamine, or an enzyme such as alkaline phosphatase, β-galactosidase, horseradish peroxidase, or luciferase. 89 For the purpose of non-invasive identification and tracking of tumor cells in a subject, such as Zr-desferrioxamine labeling. 89 Zr-labeled antibodies (e.g., positron emission tomography (PET) imaging) are included. (See, e.g., Tavare, R. et al. Cancer Res. 2016 Jan 1;76(1):73-82, and Azad, B.B. et al. Oncotarget. 2016 Mar 15;7(11):12344-58.) Specific exemplary assays that can be used to detect or measure STEAP2 in a sample include enzyme-linked immunosorbent assay (ELISA), radioimmunoassay (RIA), and fluorescence activated cell sorting (FACS).

[0253] Samples that can be used in the STEAP2 diagnostic assays of the present invention include any tissue or body fluid sample that can be obtained from a patient and that contains a detectable amount of STEAP2 protein or a fragment thereof under normal or pathological conditions. Generally, the level of STEAP2 in a particular sample obtained from a healthy patient (e.g., a patient not suffering from a disease or condition associated with abnormal STEAP2 levels or activity) is measured to first establish a baseline or standard level of STEAP2. This baseline level of STEAP2 can then be compared with the level of STEAP2 measured in a sample obtained from an individual suspected of having a STEAP2-related disease (e.g., a tumor containing STEAP2-expressing cells) or condition. [Example]

[0254] The following examples are presented so as to provide those of ordinary skill in the art with a complete disclosure and description of how to make and use the methods and compositions of the present invention, and are not intended to limit the scope of what the inventors regard as the invention. Efforts have been made to ensure accuracy with respect to numbers used (e.g., amounts, temperatures, etc.), but some experimental error and deviation should be accounted for. Unless otherwise specified, parts are parts by weight, molecular weights are average molecular weights, temperatures are in degrees Celsius, and pressures are at or near atmospheric.

[0255] Example 1: Production of anti-STEAP2 antibodies Anti-STEAP2 antibodies were obtained by immunizing genetically modified mice with human STEAP2 antigen or by immunizing genetically modified mice containing DNA encoding human immunoglobulin heavy chain and kappa light chain variable regions with human STEAP2 antigen.

[0256] Genetically modified mice were immunized with hSTEAP2 antigen (SEQ ID NO: 1899). Following immunization, splenocytes from each mouse were harvested and either (1) fused with mouse myeloma cells to maintain their viability and form hybridoma cells that were screened for STEAP specificity, or (2) B cells were sorted using a human STEAP2 fragment as a selection reagent to bind and identify reactive antibodies (antigen-positive B cells) (as described in US2007 / 0280945A1).

[0257] Chimeric antibodies against STEAP2 with human variable regions and mouse constant regions were first isolated. The antibodies are characterized and selected for desirable characteristics, including affinity, selectivity, etc. If necessary, the mouse constant region is replaced with a desired human constant region, e.g., a wild-type or modified IgG1 or IgG4 constant region, to produce a fully human anti-STEAP2 antibody. The constant region selected can vary depending on the specific application, but high affinity antigen binding and target specificity characteristics reside in the variable region. Antibody designations such as H1H11243N and H1M7804N refer to the fully human antibody "H1H" or the chimeric human variable / mouse constant region antibody "H1 Antibodies identified by hybridoma technology are designated with antibody ID numbers ending in "N" or "N2," while antibodies identified by B cell sorting are designated with antibody ID numbers ending in "P" or "P2."

[0258] The specific biological properties of exemplary anti-STEAP2 antibodies produced according to the methods of this example are described in detail in the Examples below.

[0259] Amino acid and nucleic acid sequences of the heavy and light chain variable regions of anti-STEAP2 antibodies Table 1 shows the amino acid sequence identifiers for the heavy and light chain variable regions and CDRs of selected anti-STEAP2 antibodies of the invention. The corresponding nucleic acid sequence identifiers are shown in Table 2. [Table 1] [Table 2]

[0260] Example 2: Anti-human STEAP2 antibodies selectively bind to STEAP2-expressing cell lines via FACS. The ability of anti-STEAP2 antibodies to selectively bind to cell lines endogenously expressing human prostate six-transmembrane epithelial antigen 2 (STEAP2) was determined by FACS analysis.

[0261] In short, 1×10 5 Cells were incubated with 10 μg / ml anti-STEAP2 antibody or isotype control antibody in antibody dilution buffer for 30 minutes on ice. After one wash with antibody dilution buffer, cells were incubated with 10 μg / ml PE-conjugated anti-human or anti-mouse Fc2 secondary antibody for 30 minutes on ice. After one additional wash, samples were incubated with Cytofix (1% formaldehyde) for 20 minutes. After one final wash, samples were filtered through a Pall 96-well filtration block and placed in a Hyp Electrophoresis was performed on a ForeCyt® cytometer and analyzed by ForeCyt™ (IntelliCyt, Albuquerque, NM). Mean fluorescence intensity (MFI) was expressed as the fold change above unstained levels (background). The mean fold change above background was determined at antibody concentrations ranging from 100 to 300 nM. For cell binding EC50 determination, mAb concentrations ranged from 300 nM to 5 pM, and EC50 values ​​were determined from a four-parameter logistic equation on a 12-point response curve (GraphPad Prism).

[0262] Table 3A and 3B: FACS binding characteristics of anti-STEAP2 antibodies to STEAP2-expressing and STEAP2-negative cell lines [Table 3] [Table 4]

[0263] As shown in Tables 3A and 3B, several anti-STEAP2 antibodies specifically bound to the high-STEAP2-expressing C4-2 prostate adenocarcinoma cell line by FACS at levels greater than 50-fold above background, with low nM EC50s. Some anti-STEAP2 antibodies also weakly bound to low-STEAP2-expressing HEK293 cells. Negligible binding was observed for most anti-STEAP2 antibodies on STEAP2-negative FADU, SK-BR-3, and Raji cells. This example demonstrates the ability of several anti-STEAP2 antibodies of the present invention to specifically and selectively bind to high-STEAP2-expressing cell lines.

[0264] Example 3: Anti-human STEAP2 antibodies exhibit strong internalization and specificity for human STEAP2. The ability of the anti-STEAP2 antibodies of the invention to selectively bind to STEAP2-expressing cell lines has been described (see Example 2 - FACS Binding). Next, the internalization properties of the anti-STEAP2 antibodies of the invention were also evaluated.

[0265] Briefly, 20,000 C4-2 cells were seeded onto PDL-coated 96-well plates. The next day, cells were incubated with anti-human STEAP2 antibody (10 μg / ml) on ice for 30 minutes, followed by two washes with PBS. Cells were then incubated with Alexa 488-conjugated anti-hFcFab secondary antibody on ice for 30 minutes, followed by two more PBS washes. Antibodies were internalized in internalization buffer (PBS + 2% FBS) at 37°C for 1 hour or maintained at 4°C. Cells were fixed in 4% formaldehyde, nuclei were stained with DRAQ5 (Cell signaling), and images were acquired with an ImageXpressmicroXL (Molecular Devices).

[0266] A qualitative visual assessment of total binding intensity and the intensity of antibody internalized within the vesicles was performed and scored according to the following criteria: - (no internalization or binding), + (weak internalization or binding), ++ (moderate internalization), or +++ (robust internalization or binding).

[0267] As shown in Table 4, several antibodies exhibited potent internalization capabilities toward the C4-2 cell line. In general, robust internalization correlated with the highest levels of total binding intensity.

[0268] Selected STEAP2 antibodies were then tested for binding to other human (h)STEAP family members (STEAP1, STEAP3, and STEAP4). To assess anti-STEAP2 antibody specificity, plasmid constructs expressing hSTEAP1, hSTEAP2, hSTEAP3, or hSTEAP4 fused to green fluorescent protein (GFP) were transiently introduced into HEK293 cells using Lipofectamine 2000. After 48 hours, transiently transfected cells were stained with anti-STEAP2 antibody and imaged as described above for the internalization assay. Wells containing GFP-positive cells that bound anti-STEAP2 antibody were scored as positive (+), and wells that did not bind anti-STEAP2 antibody were scored as negative (-). All antibodies tested bound to hSTEAP2-GFP-positive cells but not to STEAP1-GFP, STEAP3-GFP, or STEAP4-GFP-positive cells, confirming the specificity of binding to human STEAP2. The results are summarized in Table 5.

[0269] In summary, several anti-STEAP2 antibodies of the present invention exhibit strong internalization ability and are specific binders for human STEAP2. [Table 5] [Table 6]

[0270] Example 4: Production of bispecific antibodies that bind to STEAP2 and CD3 The present invention provides bispecific antigen-binding molecules that bind to CD3 and STEAP2; such bispecific antigen-binding molecules are also referred to herein as "anti-STEAP2 / anti-CD3 or anti-STEAP2 x CD3 bispecific molecules." The anti-STEAP2 portion of the anti-STEAP2 / anti-CD3 bispecific molecule is useful for targeting tumor cells that express the six-transmembrane epithelial antigen 2 (STEAP2) of the prostate (STEAP2), and the anti-CD3 portion of the bispecific molecule is useful for activating T cells. Simultaneous binding of STEAP2 on tumor cells and CD3 on T cells allows for the direct killing of targeted tumor cells by activated T cells (cell death). Promotes dissolution.

[0271] Bispecific antibodies comprising anti-STEAP2-specific and anti-CD3-specific binding domains were recombinantly constructed using standard molecular cloning methodologies and expressed in CHO cells, where the anti-STEAP2-specific and anti-CD3-specific binding domains each contain distinct HCVRs paired with a common LCVR. In the exemplified bispecific antibody, a heavy chain derived from an anti-CD3 antibody, a heavy chain derived from an anti-STEAP2 antibody, and a common light chain derived from an anti-STEAP2 antibody were used to construct the molecule and express it in CHO cells. In some cases, bispecific antibodies can be constructed using a heavy chain derived from an anti-CD3 antibody, a heavy chain derived from an anti-STEAP2 antibody, and a light chain derived from an anti-CD3 antibody, or an antibody light chain known to be promiscuous or to effectively pair with various heavy chain arms, such as Vκ1-39JK5 or Vκ3-20JK1.

[0272] The bispecific antibodies described in the Examples below are derived from human soluble heterodimeric hCD3ε / δ proteins. Exemplary bispecific antibodies were produced with engineered (chimeric) IgG4 Fc domains, as described in U.S. Patent Application Publication US20140243504A1, published August 28, 2014, consisting of anti-CD3 binding arms with varying binding affinities to protein (described in Example 12 herein) and human STEAP2 (see Examples 1-2 above).

[0273] A summary of the components of the antigen-binding domains of the various anti-STEAP2 x CD3 bispecific antibodies constructed is shown in Table 6. [Table 7]

[0274] The light chains listed in Table 6 were common to both the CD3- and STEAP2-targeting arms of the bispecific antibody. Tables 1 and 2 provide the amino acid and nucleic acid sequence identifiers, respectively, for the various heavy chain variable regions of the anti-STEAP2 arms (i.e., the HCVR and LCVR are derived from H2M11162N) for constructing the bispecific antibody of this example, and their corresponding CDRs. Tables 15 and 16 provide the amino acid and nucleic acid sequence identifiers, respectively, for the various heavy chain variable regions of the anti-CD3 arms of the bispecific antibody of this example, and their corresponding CDRs.

[0275] Example 5: Anti-STEAP2 / anti-CD3 bispecific antibodies exhibit potent anti-tumor effects in vivo. To determine the efficacy of exemplary anti-STEAP2 / anti-CD3 bispecific antibodies in vivo, studies were performed in immunocompromised mice bearing prostate cancer xenografts.

[0276] Efficacy of anti-STEAP2 / anti-CD3 bispecific antibodies in human tumor xenograft models To evaluate the in vivo efficacy of the anti-STEAP2 / anti-CD3 bispecific in a human tumor xenograft study, NODscid gamma (NSG) mice (Jackson Laboratories, Bar Harbor, Maine) were transfected with human prostate cancer C4-2 cells (MD Anderson Cancer Center), which endogenously express STEAP2. Human peripheral blood mononuclear cells (PBMCs, ReachBio LLC., Seattle, WA) were co-transplanted with erythrocytes (IL-1, IL-2, IL-3, IL-4, IL-5, IL-6, IL-8, IL-10, IL-11, IL-12, IL-13, IL-14, IL-15, IL-16, IL-17, IL-18,

[0277] In short, 5.0 x 10 6C4-2 cells were cultured at 1.25 × 10 cells / well in a 50:50 mixture of Matrigel matrix (BD Biosciences, San Jose, CA). 6 Male NSG mice were co-transplanted subcutaneously (sc) with human PBMCs into the right flank of mice. Mice were treated intraperitoneally (ip) with anti-STEAP2 / anti-CD3 bispecific BSSTEAP2 / CD3-001, BSSTEAP2 / CD3-002, or BSSTEAP2 / CD3-003, or an isotype control, at a dose of 0.1 or 0.01 mg / kg (N=5 mice / group) on the day of transplantation (immediate treatment model).

[0278] Measure tumor size twice a week using calipers and calculate tumor volume as volume = (length × width) 2 ) / 2. Data are presented as tumor size (mm ) at the study endpoint, 46 days after tumor implantation. 3 ) (Table 7). [Table 8]

[0279] As the results in Table 7 show, when tumor size was measured at the study endpoint, BSSTEAP2 / CD3-001, BSSTEAP2 / CD3-002, and BSSTEAP2 / CD3-003 significantly suppressed tumor growth compared to the isotype control. Importantly, the anti-STEAP2 / anti-CD3 bispecific antibody was effective in inhibiting C4-2 tumor growth even at the lowest dose of 0.1 mg / kg.

[0280] Example 6: Preparation and characterization of the complex All monoclonal antibodies were expressed in CHO cells and purified by protein A. Isotype controls were prepared similarly. The non-binding isotype control antibody was derived from an immunological antigen unrelated to oncology.

[0281] Antibody (10 mg / ml) in 50 mM HEPES, 150 mM NaCl, pH 7.5 was treated with 1 mM dithiothreitol for 30 minutes at 37°C. After gel filtration (G-25, pH 4.5 sodium acetate), maleimide linker payload derivative compound 7 (1.2 equivalents / SH group) in DMSO (10 mg / ml) was added to the reduced antibody, and the mixture was adjusted to pH 7.0 with 1 M HEPES (pH 7.4). Compound 7 and methods for preparing the compound are described in PCT Publication No. WO2014 / 145090, published September 18, 2014, and incorporated herein by reference in its entirety. After 1 hour, the reaction was quenched with excess N-ethylmaleimide. The conjugates were purified by size-exclusion chromatography and sterile filtered. Protein and linker payload concentrations were determined by UV spectroscopy. Size-exclusion HPLC indicated that all conjugates used were >95% monomeric. RP-HPLC demonstrated <0.5% unbound linker payload. Yields are reported in Table 8 based on protein titer determinations. All conjugated antibodies were analyzed by UV for linker payload loading values ​​according to Hamblett et al., Cancer Res 2004 10 7063. Results are summarized in Table 8.

[0282] Conjugates containing compound 60 can be prepared using similar methods. Compound 60, and methods for making the compound, are described in PCT Publication No. WO2016 / 160615, published October 6, 2016 (Example 20), which is incorporated herein by reference in its entirety. Compound 60 is maytansine-N-methyl-L-alanine-(3-methoxy-4-amino)benzamide-Cit-Val-Cap-Mal. [Table 9]

[0283] Example 7: Anti-STEAP2 antibody-drug conjugates (ADCs) are potent inhibitors of tumor growth in an in vivo STEAP2-positive prostate cancer xenograft model. A. To determine the in vivo efficacy of anti-STEAP2 antibodies conjugated to compound 7, studies were performed in immunocompromised mice bearing STEAP2-positive prostate cancer xenografts.

[0284] For these studies, male SCID mice (Taconic, Hudson NY) were implanted with C4-2 cells that endogenously express STEAP2. Tumors were 200–250 mm 3 Once the mice reached a mean volume of approximately 1500-2000 mm (approximately days 13-17), they were randomized into treatment groups and administered either anti-STEAP2 binding antibody, non-binding antibody, or vehicle. In these in vivo studies, the mice grew to approximately 1500-2000 mm in cohorts that received one dose of antibody followed by vehicle alone. 3 Tumors were monitored until a mean tumor size of 100 mg / kg / day was achieved (approximately 40-50 days). Treatment groups that demonstrated efficacy were maintained for longer periods (80-110 days).

[0285] In the first study, an exemplary anti-STEAP2 antibody conjugated to Compound 7 was examined for efficacy in reducing C4-2 tumor volume. 13 days after implantation, mice were administered a single dose of 10, 20, or 40 mg / kg of anti-STEAP2 and control ADCs. As summarized in Figure 1, H1H7841N-7 (DAR 2.92) significantly reduced tumor growth at all doses tested. At the highest dose (40 mg / kg), H1H784N-7 effectively reduced tumor size, but the non-conjugated control antibody (40 mg / kg) also had an effect on tumor volume. Across all doses tested, H1H784N-7 reduced tumor size more potently than the control conjugated antibody.

[0286] In the second study, STEAP2 ADC was administered at 5 mg / kg and 20 mg / kg, and a control antibody was administered at 20 mg / kg on day 14 post-implantation. As summarized in Figure 2, H1H7841N-7 (DAR 2.92) potently inhibited tumor growth at the 20 mg / kg dose, as in the previous experiment, but showed reduced efficacy at the 5 mg / kg dose. The 20 mg / kg dose of the control antibody showed no difference compared to the vehicle control.

[0287] In further studies, H1H7841N-7 (DAR 2.7) and control antibodies were administered at drug equivalents in μg / kg based on the ADC drug:antibody ratio ("DAR"). The dose was 150 μg / kg on day 17 post-implant (Figure 3). H1H7841N-7 potently inhibited tumor growth at the 150 μg / kg dose, demonstrating tumor regression by day 42 post-implant and day 25 post-injection. At this time point, tumor growth began to rebound. Tumor growth with the control antibody at this dose was not different from vehicle control.

[0288] B. In a similar study, male SCID mice (Taconic, Hudson, NY) were implanted with C4-2 cells that endogenously express STEAP2. An exemplary anti-STEAP2 (H1H7814N) antibody conjugated to Compound 60 was examined for efficacy in C4-2 tumor regression. On day 29 post-implantation, mice received a single dose of 2.5 mg / kg of anti-STEAP2 ADC, isotype control ADC (binding to an irrelevant antigen), or vehicle (PBS). Tumor volume and body weight were recorded on day 0 post-injection (day of injection), as well as days 4, 6, 8, 12, 14, and 20 post-injection. As summarized in Figure 4, H1H7814N-60 (DAR 3.6) potently inhibited tumor growth at the doses tested and demonstrated tumor regression up to 20 days post-injection (49 days post-implantation). The percent change in body weight for the test ADC was less than or equal to -2.01% by day 14 compared to mice treated with the control Ab-ADC, where percent changes in body weight were observed between -4.02% and -11.55% by day 14 (after administration of H1H7814N-60).

[0289] Example 8: Production of anti-CD3 antibodies Anti-CD3 antibodies were obtained by immunizing genetically engineered mice containing DNA encoding human immunoglobulin heavy and kappa light chain variable regions with cells expressing CD3 or DNA encoding CD3. Antibody immune responses were monitored by CD3-specific immunoassays. When the desired immune response was achieved, splenocytes were collected and fused with mouse myeloma cells to maintain their viability and form hybridoma cell lines. The hybridoma cell lines were screened and selected to identify cell lines producing CD3-specific antibodies. Using this technique, several anti-CD3 chimeric antibodies (i.e., antibodies with human variable domains and mouse constant domains) were obtained. Furthermore, as described in US2007 / 0280945A1, several fully human anti-CD3 antibodies were isolated directly from antigen-positive B cells without fusion to myeloma cells.

[0290] The specific biological properties of exemplary anti-CD3 antibodies produced according to the methods of this example are described in detail in the Examples below.

[0291] Example 9: Amino acid and nucleic acid sequences of heavy and light chain variable regions Table 9 shows the amino acid sequence identifiers for the heavy and light chain variable regions and CDRs of selected anti-CD3 antibodies of the invention. The corresponding nucleic acid sequence identifiers are shown in Table 10. Methods for making the anti-CD3 antibodies disclosed herein are also described in U.S. Patent Application Publication No. 2014 / 0129994, published March 27, 2014. This can be found in Patent Publication No. 2014 / 0088295. [Table 10] [Table 11] [Table 12] [Table 13] [Table 14] [Table 15]

[0292] Antibodies are typically referred to herein by the following nomenclature: an Fc prefix (e.g., "H1H," "H1M," "H2M," etc.), followed by a numerical identifier (e.g., "2712," "2692," etc., as shown in Table 1), followed by a "P," "N," or "B" suffix. Thus, according to this nomenclature, antibodies may be referred to herein as, for example, "H1H2712N," "H1M2692N," "H2M2689N," etc. The H1H, H1M, and H2M prefixes in antibody names used herein indicate the particular Fc region isotype of the antibody. For example, an "H1H" antibody has a human IgG1 Fc, an "H1M" antibody has a murine IgG1 Fc, and an "H2M" antibody has a murine IgG2 Fc (all variable regions are fully human, as indicated by the initial "H" in the antibody name). As will be understood by those of skill in the art, an antibody having a particular Fc isotype can be converted to an antibody having a different Fc isotype (e.g., an antibody having a murine IgG1 Fc can be converted to an antibody having a human IgG4 Fc, etc.), but in all events, the variable domains (including the CDRs) indicated by the numerical identifiers shown in Table 1 will remain the same, and the binding characteristics are expected to be the same or substantially similar regardless of the nature of the Fc domain.

[0293] Tables 11 and 12 provide the amino acid sequence identifiers for the heavy chain variable regions (Table 13) and light chain variable regions (Table 14) of additional anti-CD3 HCVRs and LCVRs useful in the anti-STEAP2 x CD3 bispecific antibodies of the invention, and their corresponding CDRs. [Table 16] [Table 17]

[0294] The heavy and light chain variable regions of CD3-VH-F and CD3-VL-F were derived from an anti-CD3 antibody designated "L2K" described in WO2004 / 106380.

[0295] Additionally, Tables 13 and 14 provide sequence identifiers for nucleotide sequences encoding the heavy chain variable regions (Table 13) and light chain variable regions (Table 14) of additional anti-CD3 HCVRs and LCVRs useful in the anti-STEAP2 x anti-CD3 bispecific antibodies of the invention, as well as their corresponding CDRs. [Table 18] [Table 19]

[0296] Control constructs used in the following examples For comparison purposes, various control constructs (anti-CD3 antibodies) were included in the following experiments: the murine monoclonal antibody "OKT-3" against a human T cell surface antigen, available from the American Type Culture Collection (ATCC) under catalog number CRL-8001, and a mouse monoclonal antibody reactive against the epsilon chain of the T3 complex on human T lymphocyte cells, e.g., available from Biolegend, San Diego, CA (Cat No. 302914) or the commercially available mouse monoclonal antibody "SP34" available from BD Pharmagen, Cat. 55052.

[0297] Example 10: Production of additional anti-CD3 antibodies The following procedure aimed to identify antibodies that specifically recognize CD3 (a T cell coreceptor) as an antigen.

[0298] A pool of anti-CD3 antibodies was derived from genetically engineered mice. Briefly, mice were immunized with CD3 antigen to generate B cells containing diverse human VH rearrangements to express a diverse repertoire of high-affinity antigen-specific antibodies. The antibodies listed in Tables 15-18 have the same light chain sequence of VK1-39JK5 (LCVR set forth in SEQ ID NO: 1890).

[0299] The antibodies produced were tested for affinity to human and cynomolgus monkey CD3 antigens in in vitro binding assays, and one CD3 antibody, designated CD3-VH-P (HCVR set forth in SEQ ID NO: 1882), among several others, was identified with an EC50 affinity of 1-40 nM as determined in FACS titrations of Jurkat cells and cynomolgus monkey T cells. 50 These antibodies were found to bind to both human and cynomolgus monkey CD3, respectively, bearing the nucleotide sequence ##STR00001## See, e.g., Example 12 and the FACS binding experiments outlined in PCT Application No. PCT / US2016 / 044732, filed July 29, 2016.

[0300] The germline amino acid residues of CD3-VH-P were then identified, and an antibody designated "CD3-VH-G" was engineered to contain only the germline framework. Other antibody derivatives were engineered by well-known molecular cloning techniques to replace amino acid residues in a stepwise manner based on the differences between the germline and CD3-VH-P sequences. Each antibody derivative is given a "CD3-VH-G" numbering designation. See Table 15.

[0301] While CD3-VH-G and several other engineered antibodies retained their binding affinity as seen in FACS assays, several anti-CD3 antibodies in bispecific formats bind to human or cynomolgus CD3 in vitro with weak to immeasurable binding affinity, such as an EC50 of greater than 100 nM. Bispecific antibodies, including exemplary anti-CD3 antibodies, were subsequently further investigated for binding affinity, binding kinetics, and other biological properties elucidating toxicity and pharmacokinetic (pK) profiles, as were produced according to the methods of this example.

[0302] Example 11: Heavy and light chain variable regions (amino acid and nucleic acid sequences of CDRs) Table 15 shows the amino acid sequence identifiers for the heavy chain variable regions and CDRs of selected anti-CD3 antibodies of the invention. The corresponding nucleic acid sequence identifiers are shown in Table 16.

[0303] The amino acid and nucleic acid sequences were determined for each antibody heavy chain sequence. Each antibody heavy chain derived from the germline sequence (SEQ ID NO: 1910) was assigned the "G" number designation for consistent nomenclature. Table 15 shows the amino acid sequence identifiers for the heavy chain variable regions and CDRs of the engineered anti-CD3 antibodies of the invention. The corresponding nucleic acid sequence identifiers are shown in Table 16. The amino acid and nucleic acid sequence identifiers for the light chain variable regions and CDRs are also identified in Tables 17 and 18, respectively, below. [Table 20] [Table 21] [Table 22] [Table 23]

[0304] Control 1 antibody, designated "CD3-L2K", was constructed based on a known anti-CD3 antibody (ie, the anti-CD3 antibody "L2K" described in WO2004 / 106380).

[0305] The isotype control antibody referred to in the examples herein is an isotype-matched (modified IgG4) antibody that interacts with an irrelevant antigen, namely the FelD1 antigen.

[0306] Example 12: In vitro and in vivo testing of human monoclonal anti-CD3 antibodies In vivo and in vitro testing of human monoclonal anti-CD3 antibodies was performed as described in U.S. Patent Publication No. 2014 / 0088295, published March 27, 2014, and PCT / US2016 / 044732, filed July 29, 2016.

[0307] Some human monoclonal anti-CD3 antibodies of the present invention bind with high affinity to soluble heterodimeric CD3 proteins in either antibody-capture or antigen-capture formats. Soluble heterodimeric CD3 proteins (hCD3-epsilon / hCD3-delta; SEQ ID NOs: 1900 / 1901) were prepared with either a human Fc tag (hFcΔAdp / hFc; SEQ ID NOs: 1931 / 1932) or a mouse Fc tag (mFcΔAdp / mFc; SEQ ID NOs: 1933 / 1934). Heterodimeric CD3 proteins were purified using the method described in Davis et al. (US2010 / 0331527).

[0308] Several human monoclonal anti-CD3 antibodies of the present invention bound to human T cells and induced T cell proliferation. Several human monoclonal anti-CD3 antibodies of the present invention bound to CD2+CD4+ monkey T cells and induced their proliferation. Several human monoclonal anti-CD3 antibodies supported redirected T cell-mediated killing via Fc / FcR interactions in a calcein-based U937 killing assay. The observed killing appears to be dependent on Fc binding of the antibody to Fc receptors on U937 cells, leading to clustering of CD3 on adjacent T cells, and was squelched by the addition of nonspecific human IgG (data not shown).

[0309] Example 13: In vitro studies on human STEAP2xCD3 bispecific antibodies FACS binding titration for Jurkat cells, PC3_STEAP2 / 1 cells, and cynomolgus monkey T cells: Flow cytometry analysis was used to determine the binding of the STEAP2xCD3 bispecific antibody to Jurkat T cells, PC3_STEAP2 / 1 chimeric T cells, and cynomolgus monkey T cells, followed by detection with a phycoerythrin (PE)-conjugated anti-human (IgG) antibody. Briefly, 2 x 10 cells / well were incubated for 30 min at 4 °C with serial dilutions of the STEAP2xCD3 bispecific antibody or a control antibody (a human IgG1 antibody that binds to a feline antigen with no cross-reactivity to STEAP2 or human or cynomolgus monkey CD3) ranging from 66.6 nM to 0.001 nM. After incubation, cells were washed twice with cold PBS containing 1% filtered FBS, and a PE-conjugated anti-human secondary antibody was added to the cells and incubated for an additional 30 min. Wells containing no antibody or only the secondary antibody were used as controls. After incubation, cells were washed, resuspended in 200 μL of cold PBS containing 1% filtered FBS, and analyzed by flow cytometry on a BD FACS Canto II. [Table 24]

[0310] Jurkat cells are derived from a T-cell lymphoblastoid cell line that expresses human CD3. All bispecific antibodies tested (Table 19 and Figure 5) bound to Jurkat cells with EC50s ranging from 1.41E-08 M to 6.15E-10 M. PC3 cells, a human prostate cancer cell line, were engineered to express a STEAP2 / 1 chimeric construct. Several bispecific antibodies bound to PC3_STEAP2 / 1 cells with EC50s ranging from 7.91E-08 M to 3.44E-09 M (Table 19 and Figure 6).

[0311] The binding of STEAP2xCD3 bispecific antibodies to the surface of purified cynomolgus monkey T cells was also tested. Several bispecific antibodies bound with EC50s ranging from 1.73E-08M to 7.27E-09M. Control antibodies did not bind to either cell line. See Table 19 and Figures 7 and 8.

[0312] T cell proliferation assay: Thawed human or freshly isolated monkey PBMCs (50,000 cells / well) were incubated in white flat-bottom 96-well plates with 3-fold serial dilutions (human, concentration range: 5E-10M to 2.82E-15M; cynomolgus, concentration range: 1E-09M to 4.57E-13M) of STEAP2xCD3 bispecific or isotype control in complete medium (RPMI supplemented with 10% FBS, 100 U / mL L-penicillin, 100 μg / mL streptomycin, and 292 μg / mL L-glutamine) and a fixed concentration (human: 200 ng / mL, cynomolgus: 500 ng / mL) of a commercially available anti-CD28 antibody (Biolegend, catalog no. 302914) for 72 hours at 37°C. Isolated monkey PBMCs were from two donors (identified as mk8781M or mk9381M). After incubation, CellTiter Glo® (Promega, Cat. No. 7573) was added, and luminescence, as a readout of cell viability, was measured using a VICTOR X5 multilabel plate reader. Cell titers were calculated by dividing the luminescence of stimulated cells by the baseline luminescence of unstimulated cells.

[0313] All αSTEAP2×αCD3 bispecific antibodies induce human PBMC proliferation in the presence of costimulatory anti-CD28 antibodies (see Table 20 and Figure 9). PBMCs were incubated with serial dilutions of bispecific or control antibodies and a fixed concentration of anti-CD28 for 72 hours, and cell viability was measured in a luminescence assay to detect live cells. Proliferation was measured by comparing the luminescence of bispecific antibody-stimulated cells with cells without antibody. EC 50 The antibody concentration required to induce half-maximal proliferation was 3.68E-1 The range of activity was 3 M to 1.60E-10 M. In contrast, the control antibody showed no activity under the same conditions. [Table 25]

[0314] The bispecific antibodies BSSTEAP2 / CD3-0010 and BSSTEAP2 / CD3-011 also had EC values ​​of 7E-13 and 3.6E-12, respectively. 50 BSSTEAP2 / CD3-004 induced proliferation of cynomolgus monkey PBMCs (donor mk8781M) demonstrating EC activity. BSSTEAP2 / CD3-004 activity was donor dependent. Two additional bispecific antibodies, BSSTEAP2 / CD3-001 and BSSTEAP2 / CD3-006, induced robust proliferation of cynomolgus monkey PBMCs in all donors tested. EC using donor mk9381M 50 The values ​​were 4.6E-12M and 1.53E-11M, respectively (see Table 20 and Figure 10).

[0315] BSSTEAP2 / CD3-007 and BSSTEAP2 / CD3-008 activity was donor-dependent. In contrast, BSSTEAP2 / CD3-005, BSSTEAP2 / CD3-009, and the isotype control showed no activity.

[0316] Cytotoxicity assay targeting C4-2 cells in the presence of anti-STEAP2 x CD3 bispecific antibody and human T cells: To monitor specific killing of STEAP2-bearing target cells by flow cytometry, C4-2 cells were labeled with 1 μM of the fluorescent tracking dye Violet Cell Tracker (Life Technologies kit, #C34557). After labeling, cells were seeded overnight at 37°C. Separately, 1 × 10 human PBMCs were cultured. 6Cells were seeded at 1000 cells / mL in supplemented RPMI medium and incubated overnight at 37°C to enrich for lymphocytes by depleting adherent macrophages, dendritic cells, and some monocytes. The next day, target cells were incubated with adherent cell-depleted naive PBMCs (effector / target cell ratio 4:1) and serially diluted STEAP2 × CD3 bispecific antibody or IgG1 control antibody (not binding to STEAP2) (concentration range: 66.7 nM to 0.25 pM) at 37°C for 48 hours. Cells were removed from the cell culture plate using enzyme-free vesicle dissociation buffer and analyzed by FACS. For FACS analysis, cells were stained with a dead / live far-red cell tracker (Invitrogen). Immediately prior to FACS analysis, 5×10 5 Counting beads were added to each well. 1 x 10 for each sample 5 Beads were collected. To assess the specificity of killing, cells were gated on the live, violet-labeled population. The percentage of the live population was recorded and used to calculate normalized viability.

[0317] T cell activation was assessed by incubating cells with antibodies directly binding to CD2 and CD69 and reporting the percentage of activated (CD69+) T cells among all T cells (CD2+). Several anti-STEAP2 x CD3 bispecific antibodies were tested for their ability to induce naive T cells to kill target cells expressing human STEAP2 (see Table 21 and Figure 11). All tested antibodies activated and induced human T cells to deplete C4-2 cells (a human prostate adenocarcinoma subline derived from LnCap cells). Target cell killing was observed only in the presence of the bispecific antibody, and C4-2 cells were depleted in a dose-dependent manner with an EC50 of 10 pM. Furthermore, the observed target cell lysis was associated with upregulation of CD69 expression on CD2+ T cells with an EC50 of 10 pM (see Table 21 and Figure 12). [Table 26]

[0318] The present invention is not to be limited in scope by the specific embodiments described herein. Indeed, various modifications of the invention in addition to those described herein will be apparent to those skilled in the art from the foregoing description. Such modifications are intended to be included within the scope of the appended claims.

Claims

1. An antibody-drug conjugate (ADC) comprising an anti-STEAP2 antibody or antigen-binding fragment thereof and a cytotoxic agent, wherein the antibody or antigen-binding fragment and the cytotoxic agent are covalently bound via a linker.

2. An isolated antibody or antigen-binding fragment thereof that binds to human prostate six-transmembrane epithelial antigen 2 (STEAP2)-expressing cells with an EC50 of less than about 50 nM as measured by an in vitro FACS binding assay described herein.

3. The isolated antibody or antigen-binding fragment thereof of claim 2, which is internalized by human STEAP2-expressing cells.

4. The isolated antibody of claim 3 , wherein the antibody is fully human.

5. The antibody or antigen-binding fragment of any one of claims 2 to 4, wherein the antibody or antigen-binding fragment thereof competes with a reference antibody comprising an HCVR / LCVR amino acid sequence pair listed in Table 1 for binding to human STEAP2.

6. 6. The antibody or antigen-binding fragment of claim 5, wherein the reference antibody comprises an HCVR / LCVR amino acid sequence pair selected from the group consisting of SEQ ID NOs: 2 / 10, 18 / 26, 34 / 42, 50 / 58, 66 / 58, 74 / 58, 82 / 58, 90 / 58, 98 / 58, 106 / 114, 122 / 130, 138 / 146, 154 / 162, 170 / 178, 186 / 194, 202 / 210, 218 / 226, 234 / 242, 250 / 258, 266 / 274, 282 / 290, 298 / 306, 314 / 322, 330 / 338, 346 / 354, 362 / 370, and 378 / 386.

7. The antibody or antigen-binding fragment thereof of any one of claims 2 to 6, wherein the antibody or antigen-binding fragment thereof binds to the same epitope on human STEAP2 as a reference antibody comprising the HCVR / LCVR amino acid sequence pair listed in Table 1.

8. The antibody or antigen-binding fragment thereof is selected from the group consisting of SEQ ID NOs: 2 / 10, 18 / 26, 34 / 42, 50 / 58, 66 / 58, 74 / 58, 82 / 58, 90 / 58, 98 / 58, 106 / 114, 122 / 130, 138 / 146, 154 / 162, 170 / 178, 186 / 194, 202 / 210, 218 / 226, 234 / 242, 250 / 262, 260 / 270, 272 / 280, 282 / 290, 292 / 300, 306 / 310, 312 / 320, 322 / 330, 336 / 340, 342 / 350, 356 / 360, 370 / 380, 372 / 380, 374 / 390, 382 / 400, 384 / 410, 386 / 420, 388 / 430, 388 / 440, 389 / 450, 392 / 460, 402 / 410, 402 / 420, 404 / 430, 406 / 440, 410 / 410, 412 / 420, 414 / 430, 416 / 440, 418 / 450, 420 / 420, 422 / 430, 424 / 440, 426 / 450, 428 / 460, 430 / 430, 432 / 440, 434 / 450, 440 / The antibody or antigen-binding fragment of claim 7, which binds to the same epitope on human STEAP2 as a reference antibody comprising an HCVR / LCVR amino acid sequence pair selected from the group consisting of 258, 266 / 274, 282 / 290, 298 / 306, 314 / 322, 330 / 338, 346 / 354, 362 / 370, and 378 / 386.

9. An isolated antibody or antigen-binding fragment thereof that binds to human STEAP2, wherein the antibody or antigen-binding fragment comprises: (a) a complementarity-determining region (CDR) of a heavy chain variable region (HCVR) having an amino acid sequence set forth in Table 1; and (b) a CDR of a light chain variable region (LCVR) having an amino acid sequence set forth in Table 1.

10. The antibody or antigen-binding fragment is selected from the group consisting of SEQ ID NOs: 2 / 10, 18 / 26, 34 / 42, 50 / 58, 66 / 58, 74 / 58, 82 / 58, 90 / 58, 98 / 58, 106 / 114, 122 / 130, 138 / 146, 154 / 162, 170 / 178, 186 / 194, 202 / 210, 218 / 226, 234 / 242, 250 / 258, 266 / 274, 282 / 290, 298 / 306, 314 / 322, 330 / 338, 346 / 354 10. The isolated antibody or antigen-binding fragment of claim 9, comprising the heavy chain CDRs and light chain CDRs of an HCVR / LCVR amino acid sequence pair selected from the group consisting of 362 / 370, and 378 / 386.

11. The antibodies or antigen-binding fragments are selected from the group consisting of SEQ ID NOs: 4-6-8-12-14-16, 20-22-24-28-30-32, 36-38-40-44-46-48, 52-54-56-60-62-64, 68-70-72-60-62-64, 76-78-80-60-62-64, 84-86-88-60-62-64, 92-94-96-60-62-64, 100-102-10 4-60-62-64, 108-110-112-116-118-120, 124-126-128-132-134-136, 140-142-144-148-150-152, 156-158-160-164-166-168, 172-174-176-180-182-184, 188-190-192-196-198-200, 204-206-208-212-214- 216, 220-222-224-228-230-232, 236-238-240-244-246-248, 252-254-256-260-262-264, 268-270-272-276-278-280, 284-286-288-292-294-296, 300-302-304-308-310-312, 316-318-320-324-326-328, 332 11. The isolated antibody or antigen-binding fragment of claim 10, comprising a HCDR1-HCDR2-HCDR3-LCDR1-LCDR2-LCDR3 domain selected from the group consisting of: -334-336-340-342-344, 348-350-352-356-358-360, 364-366-368-372-374-376, and 380-382-384-388-390-392.

12. 1. An isolated antibody or antigen-binding fragment thereof that binds to human STEAP2, wherein the antibody or antigen-binding fragment comprises: (a) an amino acid sequence selected from the group consisting of SEQ ID NOs: 2, 18, 34, 50, 66, 74, 82, 90, 98, 106, 122, 138, 154, 170, 186, 202, 218, 234, 250, 266, 282, 298, 314, 330, 346, 362, and 378. and (b) a heavy chain variable region (HCVR) having an amino acid sequence selected from the group consisting of SEQ ID NOs: 10, 26, 42, 58, 114, 130, 146, 162, 178, 194, 210, 226, 242, 258, 274, 290, 306, 322, 338, 354, 370, and 386.

13. 13. The isolated antibody or antigen-binding fragment of claim 12, wherein the antibody or antigen-binding fragment comprises an HCVR / LCVR amino acid sequence pair selected from the group consisting of SEQ ID NOs: 2 / 10, 18 / 26, 34 / 42, 50 / 58, 66 / 58, 74 / 58, 82 / 58, 90 / 58, 98 / 58, 106 / 114, 122 / 130, 138 / 146, 154 / 162, 170 / 178, 186 / 194, 202 / 210, 218 / 226, 234 / 242, 250 / 258, 266 / 274, 282 / 290, 298 / 306, 314 / 322, 330 / 338, 346 / 354, 362 / 370, and 378 / 386.

14. The ADC of claim 1, wherein the anti-STEAP2 antibody or antigen-binding fragment is the antibody or antigen-binding fragment of any one of claims 2 to 13.

15. 15. The ADC of claim 1 or 14, wherein the cytotoxic agent is selected from an auristatin, a maytansinoid, a tubulysin, a tomaymycin derivative, or a dolastatin derivative.

16. 15. The method of claim 1 or 14, wherein the cytotoxic agent is an auristatin selected from MMAE or MMAF, or a maytansinoid selected from DM1 or DM4. The ADC described in

17. 15. The ADC of claim 1 or 14, wherein the cytotoxic agent is a maytansinoid having the structure of Formula (I) or Formula (II):

18. The maytansinoid is 【Chemical 1】 18. The ADC of claim 17, wherein:

19. The maytansinoid is 【Chemistry 2】 18. The ADC of claim 17, wherein:

20. comprising an anti-STEAP2 antibody or fragment thereof and the formula: 【Chemistry 3】 During the ceremony, 【Chemistry 4】 is the binding to the anti-STEAP2 antibody or fragment thereof.

21. comprising an anti-STEAP2 antibody or fragment thereof and the formula: 【Chemistry 5】 During the ceremony, 【Chemistry 6】 is the binding to the anti-STEAP2 antibody or fragment thereof.

22. comprising an anti-STEAP2 antibody or fragment thereof and the formula: 【Chemistry 7】 During the ceremony, 【Chemistry 8】 is the binding to the anti-STEAP2 antibody or fragment thereof.

23. 23. The ADC of any one of claims 20 to 22, wherein the bond contacts the antibody or fragment thereof via a sulfur moiety of a cysteine ​​residue.

24. an anti-STEAP2 antibody or a fragment thereof; 【Chemistry 9】 and mixtures thereof, During the ceremony, 【Chemistry 10】 is the binding to the anti-STEAP2 antibody or fragment thereof.

25. 25. The ADC of claim 24, wherein the bond is contacted with the antibody or fragment thereof via a nitrogen moiety of a lysine residue.

26. 26. The ADC of any one of claims 1 or 14-25, wherein the ADC comprises 1 to 4 cytotoxic agents per anti-STEAP2 antibody or fragment thereof.

27. A pharmaceutical composition comprising the antibody-drug conjugate of any one of claims 1 or 14 to 26, and a pharmaceutically acceptable carrier or diluent.

28. A bispecific antigen-binding molecule comprising a first antigen-binding domain that binds to human CD3 and a second antigen-binding domain that binds to human STEAP2, wherein the second antigen-binding domain is derived from the antibody or antigen-binding fragment of any one of claims 2 to 13.

29. A bispecific antigen-binding molecule comprising a first antigen-binding domain that specifically binds to human CD3 and a second antigen-binding domain that specifically binds to human STEAP2.

30. 30. The dual antibody of claim 28 or claim 29, wherein the first antigen-binding domain binds to a human cell expressing human CD3 and a cynomolgus monkey cell expressing cynomolgus monkey CD3. Specific antigen binding molecules.

31. 30. The bispecific antigen-binding molecule of claim 28 or 29, wherein the second antigen-binding domain binds to a human cell expressing human STEAP2.

32. The bispecific antigen-binding molecule of any one of claims 28 to 31, wherein each of the first antigen-binding domain and the second antigen-binding domain is fully human.

33. The bispecific antigen-binding molecule of claim 28 or 29, wherein the antigen-binding molecule binds to both human CD3 and human STEAP2 and induces T cell-mediated cell killing of STAP2-expressing cells.

34. 30. The bispecific antigen-binding molecule of claim 28 or claim 29, wherein the antigen-binding molecule inhibits tumor growth in immunocompromised mice bearing human prostate cancer xenografts.

35. The bispecific antigen-binding molecule of any one of claims 28 to 34, which is a bispecific antibody or a bispecific antigen-binding fragment thereof.

36. the second antigen-binding domain that specifically binds to human STEAP2 comprises heavy chain complementarity-determining regions (HCDR1, HCDR2, and HCDR3) from a heavy chain variable region (HCVR) selected from the group consisting of SEQ ID NOs: 2, 18, 34, 50, 66, 74, 82, 90, 98, 106, 122, 138, 154, 170, 186, 202, 218, 234, 250, 266, 282, 298, 314, 330, 346, 362, and 378; and light chain complementarity determining regions (LCDR1, LCDR2, and LCDR3) from a light chain variable region (LCVR) comprising an amino acid sequence selected from the group consisting of SEQ ID NOs: 10, 26, 42, 58, 114, 130, 146, 162, 178, 194, 210, 226, 242, 258, 274, 290, 306, 322, 338, 354, 370, and 386.

37. The second antigen-binding domain that specifically binds to human STEAP2 comprises three heavy chain complementarity determining regions (A2-HCDR1, A2-HCDR2, and A2-HCDR3) and three light chain complementarity determining regions (A2-LCDR1, A2-LCDR2, and A2-LCDR3), and A2-HCDR1 is selected from the group consisting of SEQ ID NOs: 4, 20, 36, 52, 68, 76, 84, 92, 100, 108, 124, 140, 156, 172, 188, 204, 220, 236, 252, 26 A2-HCDR2 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 6, 22, 38, 54, 70, 78, 86, 94, 102, 110, 126, 142, 158, 174, 190, 206, 222, 238, 254, 270, 286, 302, 318, 334, 350, 366, and 382; and A2-HCDR3 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 8, 24, 40 , 56, 72, 80, 88, 96, 104, 112, 128, 144, 160, 176, 182, 208, 224, 240, 256, 272, 288, 304, 320, 336, 352, 368, and 384, and A2-LCDR1 is selected from the group consisting of SEQ ID NOs: 12, 28, 44, 60, 116, 132, 148, 164, 180, 196, 212, 228, 244, 260, 276, 292, 308, 324, 340, 356, 372, and and 388, wherein A2-LCDR2 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 14, 30, 46, 62, 118, 134, 150, 166, 182, 198, 214, 230, 246, 262, 278, 294, 310, 326, 342, 358, 374, and 390, and A2-LCDR3 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 16, 32, 48, 64, 120, 136, 152, 168, 184, 200, 216, 232, 248, 262, 278, 294, 310, 326, 342, 358, 374, and 390. The bispecific antigen-binding molecule of any one of claims 28 to 35, comprising an amino acid sequence selected from the group consisting of: 64, 280, 296, 312, 328, 344, 360, 376, and 392.

38. The bispecific antigen-binding molecule of any one of claims 28 to 35, wherein the second antigen-binding domain that specifically binds to human STEAP2 comprises the heavy chain CDRs and light chain CDRs of the HCVR / LCVR amino acid sequence pair selected from the group consisting of SEQ ID NOs: 250 / 258.

39. The bispecific antigen-binding molecule of any one of claims 28 to 35, wherein the first antigen-binding domain that specifically binds to human CD3 comprises heavy chain complementarity determining regions (HCDR1, HCDR2, and HCDR3) derived from a heavy chain variable region (HCVR) comprising an amino acid sequence set forth in Table 9, Table 11, or Table 15, and light chain complementarity determining regions (LCDR1, LCDR2, and LCDR3) derived from a light chain variable region (LCVR) comprising an amino acid sequence set forth in Table 1, Table 9, Table 12, or Table 17.

40. The bispecific antigen-binding molecule of any one of claims 28 to 35, wherein the first antigen-binding domain that specifically binds to human CD3 comprises heavy chain complementarity determining regions (HCDR1, HCDR2, and HCDR3) from a heavy chain variable region (HCVR) selected from the group consisting of SEQ ID NOs: 1730, 1762, and 1866, and light chain complementarity determining regions (LCDR1, LCDR2, and LCDR3) from a light chain variable region (LCVR) comprising the amino acid sequence of SEQ ID NO:

258.

41. the first antigen-binding domain that specifically binds to human CD3 comprises three heavy chain complementarity-determining regions (A1-HCDR1, A1-HCDR2, and A1-HCDR3) and three light chain complementarity-determining regions (A1-LCDR1, A1-LCDR2, and A1-LCDR3); 36. The bispecific antigen-binding molecule of any one of claims 28 to 35, wherein A1-HCDR1 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1732, 1764, and 1868, A1-HCDR2 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1734, 1766, and 1870, A1-HCDR3 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1736, 1768, and 1872, A1-LCDR1 comprises the amino acid sequence of SEQ ID NO: 260, A1-LCDR2 comprises the amino acid sequence of SEQ ID NO: 262, and A1-LCDR3 comprises the amino acid sequence of SEQ ID NO:

264.

42. The bispecific antigen-binding molecule of any one of claims 28 to 35, wherein the first antigen-binding domain that specifically binds to human CD3 comprises the heavy chain CDRs and light chain CDRs of an HCVR / LCVR amino acid sequence pair selected from the group consisting of SEQ ID NOs: 1730 / 258, 1762 / 258, and 1866 / 258.

43. the first antigen-binding domain that specifically binds to human CD3 comprises three heavy chain complementarity-determining regions (A1-HCDR1, A1-HCDR2, and A1-HCDR3) and three light chain complementarity-determining regions (A1-LCDR1, A1-LCDR2, and A1-LCDR3); and the second antigen-binding domain that specifically binds to human STEAP2 comprises three heavy chain complementarity-determining regions (A2-HCDR1, A2-HCDR2, and A2-HCDR3) and three light chain complementarity-determining regions (A2-LCDR1, A2-LCDR2, and A2-LCDR3); A1-HCDR1 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1732, 1764, and 1868; A1-HCDR2 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1734, 1766, and 1870; and A1-HCDR3 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1732, 1764, and 1868. 36, 1768, and 1872, wherein A1-LCDR1 comprises the amino acid sequence of SEQ ID NO: 260, A1-LCDR2 comprises the amino acid sequence of SEQ ID NO: 262, and A1-LCDR3 comprises the amino acid sequence of SEQ ID NO: 264; 36. The bispecific antigen-binding molecule of any one of claims 28 to 35, wherein A2-HCDR1 comprises the amino acid sequence of SEQ ID NO: 252, A2-HCDR2 comprises the amino acid sequence of SEQ ID NO: 254, A2-HCDR3 comprises the amino acid sequence of SEQ ID NO: 256, A2-LCDR1 comprises the amino acid sequence of SEQ ID NO: 260, A2-LCDR2 comprises the amino acid sequence of SEQ ID NO: 262, and A2-LCDR3 comprises the amino acid sequence of SEQ ID NO:

264.

44. The bispecific antigen-binding molecule of any one of claims 28 to 35, wherein the first antigen-binding domain that specifically binds to human CD3 comprises a heavy chain comprising a variable domain framework region having an amino acid sequence selected from FR1 (SEQ ID NO: 1903), FR2 (SEQ ID NO: 1904), FR3 (SEQ ID NO: 1905), and FR4 (SEQ ID NO: 1906).

45. The bispecific antigen-binding molecule of any one of claims 28 to 35, wherein the first antigen-binding domain that specifically binds to human CD3 comprises an HCVR comprising HCDR1-HCDR2-HCDR3 having the amino acid sequence of SEQ ID NOs: 1907-1908-19091.

46. 36. The bispecific antigen-binding molecule of any one of claims 28 to 35, wherein the second antigen-binding domain competes for binding to human STEAP2 with a reference antigen-binding protein comprising three heavy chain complementarity determining regions (A2-HCDR1, A2-HCDR2, and A2-HCDR3) and three light chain complementarity determining regions (A2-LCDR1, A2-LCDR2, and A2-LCDR3), wherein A2-HCDR1 comprises the amino acid sequence of SEQ ID NO: 252, A2-HCDR2 comprises the amino acid sequence of SEQ ID NO: 254, A2-HCDR3 comprises the amino acid sequence of SEQ ID NO: 256, A2-LCDR1 comprises the amino acid sequence of SEQ ID NO: 260, A2-LCDR2 comprises the amino acid sequence of SEQ ID NO: 262, and A2-LCDR3 comprises the amino acid sequence of SEQ ID NO:

264.

47. 36. The bispecific antigen-binding molecule of any one of claims 28 to 35, wherein the second antigen-binding domain competes for binding to human STEAP2 with a reference antigen-binding protein comprising a heavy chain variable region (HCVR) comprising the amino acid sequence of SEQ ID NO: 250 and a light chain variable region (LCVR) comprising the amino acid sequence of SEQ ID NO:

258.

48. the first antigen-binding domain competes for binding to human CD3 with a reference antigen-binding protein comprising three heavy chain complementarity determining regions (A1-HCDR1, A1-HCDR2, and A1-HCDR3) and three light chain complementarity determining regions (A1-LCDR1, A1-LCDR2, and A1-LCDR3), wherein A1-HCDR1 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1732, 1764, and 1868; and A1-HCDR2 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1732, 1764, and 1868. 1734, 1766, and 1870, wherein A1-HCDR3 comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 1736, 1768, and 1872, wherein A1-LCDR1 comprises the amino acid sequence of SEQ ID NO: 260, wherein A1-LCDR2 comprises the amino acid sequence of SEQ ID NO: 262, and wherein A1-LCDR3 comprises the amino acid sequence of SEQ ID NO:

264.

49. 36. The bispecific antigen-binding molecule of any one of claims 28 to 35, wherein the first antigen-binding domain competes for binding to human CD3 with a reference antigen-binding protein comprising a heavy chain variable region (HCVR) comprising an amino acid sequence selected from the group consisting of SEQ ID NOs: 1730, 1762, and 1866, and a light chain variable region (LCVR) comprising the amino acid sequence of SEQ ID NO:

258.

50. 36. The bispecific antigen-binding molecule of any one of claims 28 to 35, wherein the first antigen-binding domain competes for binding to human CD3 with a reference antigen-binding protein comprising a heavy chain variable region (HCVR) comprising an amino acid sequence selected from the group consisting of SEQ ID NOs: 1730, 1762, and 1866, and a light chain variable region (LCVR) comprising the amino acid sequence of SEQ ID NO: 258, and the second antigen-binding domain competes for binding to human STEAP2 with a reference antigen-binding protein comprising a heavy chain variable region (HCVR) comprising the amino acid sequence of SEQ ID NO: 250, and a light chain variable region (LCVR) comprising the amino acid sequence of SEQ ID NO:

258.

51. A pharmaceutical composition comprising the antibody or antigen-binding fragment thereof of any one of claims 2 to 13, or the bispecific antigen-binding molecule of any one of claims 28 to 50, and a pharmaceutically acceptable carrier or diluent.

52. 52. A method of treating cancer in a subject, comprising administering to the subject the pharmaceutical composition of claim 51.

53. 53. The method of claim 52, wherein the cancer is selected from the group consisting of prostate cancer, bladder cancer, cervical cancer, lung cancer, colon cancer, kidney cancer, breast cancer, pancreatic cancer, stomach cancer, uterine cancer, and ovarian cancer.

54. 54. The method of claim 53, wherein the cancer is prostate cancer, optionally castration-resistant prostate cancer.

55. 52. Use of the pharmaceutical composition of claim 27 or claim 51 in the treatment of a disease or disorder associated with expression of STEAP2.

56. 56. The use of claim 55, wherein the disease or disorder is cancer.

57. A compound comprising the antibody or antigen-binding fragment of any one of claims 2 to 13, or the bispecific antigen-binding molecule of any one of claims 28 to 50, for use in medicine.

58. 51. A compound for use in the treatment of prostate cancer, the compound comprising the antibody or antigen-binding fragment of any one of claims 2 to 13, or the bispecific antigen-binding molecule of any one of claims 28 to 50.

59. 51. Use of the antibody or antigen-binding fragment of any one of claims 2 to 13, or the bispecific antigen-binding molecule of any one of claims 28 to 50, for the manufacture of a medicament for use in the treatment of cancer, optionally wherein the cancer is prostate cancer.

60. 27. A compound comprising an ADC according to any one of claims 1 or 14 to 26 for use in medicine.

61. 27. A compound for use in the treatment of prostate cancer, comprising an ADC of any one of claims 1 or 14-26.

62. 30. Use of the ADC of any one of claims 1 or 14 to 26 for the manufacture of a medicament for use in the treatment of cancer, optionally wherein the cancer is prostate cancer.

Images