Anti-STEAP2 antibody, antibody-drug conjugate, and bispecific antigen-binding molecule that binds STEAP2 and CD3, and their uses

Abstract

To provide novel full-length human IgG antibodies (monospecific antibodies) that bind to human STEAP2.SOLUTION: Provided herein is an isolated antibody or antigen binding fragment thereof that binds to six-transmembrane epithelial antigen of prostate 2 (STEAP2) of human prostate, where the antibody or antigen binding fragment thereof comprises complementarity determining regions (CDRs) contained in an amino acid sequence pair of a heavy chain variable region and a light chain variable region selected from the group consisting of specific sequences. Also provided is a pharmaceutical composition comprising the isolated antibody or antigen binding fragment thereof, and a pharmaceutically acceptable carrier or diluent for treating cancer expressing STEAP2 in a subject.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] Reference to Sequence Listing This application incorporates by reference a Sequence Listing submitted in computer-readable form as file 10296WO01-Sequence.txt, created on September 22, 2017, and containing 739,964 bytes.

[0002] The present invention relates to antibodies and antigen-binding fragments thereof that are specific for prostate six-transmembrane epithelial antigen 2 (STEAP2), and methods of using the same. The present invention also relates to bispecific antigen-binding molecules that bind STEAP2 and CD3, and methods of using the same. 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 Art

[0003] STEAP-2, metalloreductase STEAP2, prostate cancer-associated protein 1, protein upregulated in metastatic prostate cancer, six-transmembrane prostate protein 1 (STAMP1), and prostate six-transmembrane epithelial antigen 2 (STEAP2), also known as 098P4B6, is an endogenous six-transmembrane protein that is upregulated in normal and malignant prostate cells. STEAP2 functions as a shuttle between the Golgi complex and the plasma membrane and is a metalloreductase that reduces iron and copper to facilitate their uptake into cells. STEAP2 is predominantly localized in the epithelial cells of the prostate. STEAP2 is also expressed in normal heart, brain, pancreas, ovary, skeletal muscle, breast, 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, stomach 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 together with the T cell receptor complex (TCR) and is required for T cell activation. Functional CD3 is formed from the dimerization association of two of four different chains: epsilon, zeta, delta, and gamma. The dimer arrangements of CD3 include gamma / epsilon, delta / epsilon, and zeta / zeta. Antibodies against CD3 have been shown to cluster CD3 on T cells, thereby inducing T cell activation in a manner similar to the engagement of the TCR by peptide-loaded MHC molecules. Thus, anti-CD3 antibodies have been proposed for therapeutic purposes including T cell activation. Additionally, bispecific antibodies capable of binding to CD3 and a target antigen have been proposed for therapeutic use including targeting the T cell immune response against tissues and cells expressing the target antigen.

[0005] Antigen-binding molecules that target STEAP2 and include an antibody-drug conjugate, as well as bispecific antigen-binding molecules that bind to both STEAP2 and CD3, are useful in therapeutic settings where it is desired to specifically target cells expressing STEAP2 and kill them in a T cell-mediated manner.

Prior Art Documents

Patent Documents

[0006]

Patent Document 1

Patent Document 2

Non-Patent Documents

[0007]

Non-Patent Document 1

Summary of the Invention

[0008] In a first aspect, the present invention provides an antibody that binds to human STEAP2 and antigen-binding fragments thereof. Antibodies according to this aspect of the invention are useful, inter alia, for targeting cells that express STEAP2. The present invention also provides a bispecific antibody that binds to human STEAP2 and human CD3 and antigen-binding fragments thereof. Bispecific antibodies according to this aspect of the invention are useful, for example, for targeting T cells that express CD3 and stimulating T cell activation in situations where T cell-mediated killing of cells that express STEAP2 is beneficial or desirable. For example, the bispecific antibody can induce CD3-mediated T cell activation in 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 of the heavy chain variable region (HCVR) and the light chain variable region (LCVR), as well as the heavy chain complementarity determining regions (HCDR1, HCDR2, and HCDR3) and the light chain complementarity determining regions (LCDR1, LCDR2, and LCDR3) of the exemplary anti-STEAP2 antibodies. Table 2 shows the sequence identifiers of the nucleic acid molecules encoding the HCVR, LCVR, HCDR1, HCDR2, HCDR3, LCDR1, LCDR2, and LCDR3 of the exemplary anti-STEAP2.

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

[0012] The present invention also provides an antibody or an antigen-binding fragment thereof comprising an amino acid sequence pair of HCVR and LCVR (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 an antibody or an antigen-binding fragment 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 NO: 250 / 258 (e.g., H2M11162N).

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

[0014] The present invention also provides an antibody or an antigen-binding fragment thereof comprising a heavy chain complementarity-determining region 2 (HC DR2) comprising an amino acid sequence selected from any of the HCDR2 amino acid sequences listed in Table 1 or a substantially similar sequence having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0015] The present invention also provides an antibody or an antigen-binding fragment thereof comprising a heavy chain complementarity-determining region 3 (HCDR3) comprising an amino acid sequence selected from any of the HCDR3 amino acid sequences listed in Table 1 or a substantially similar sequence having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

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

[0017] The present invention also provides an antibody or an antigen-binding fragment thereof, comprising a light chain complementarity-determining region 2 (LCDR2) comprising an amino acid sequence selected from any of the LCDR2 amino acid sequences listed in Table 1, or a substantially similar sequence having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0018] The present invention also provides an antibody or an antigen-binding fragment thereof, comprising a light chain complementarity-determining region 3 (LCDR3) comprising an amino acid sequence selected from any of the LCDR3 amino acid sequences listed in Table 1, or a substantially similar sequence having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0019] The present invention also provides an antibody or an antigen-binding fragment thereof, comprising an amino acid sequence pair of heavy chain complementarity-determining region 3 and light chain complementarity-determining region 3 (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 an antibody or an antigen-binding fragment thereof, comprising the 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 NO: 256 / 264 (e.g., H2M11162N).

[0020] The present invention also provides an antibody or an antigen-binding fragment thereof comprising a 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 an 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 an antigen-binding fragment thereof comprising an HCDR1-HCDR2-HCDR3-LCDR1-LCDR2-LCDR3 amino acid sequence set contained within an 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 the CDRs within the specific HCVR amino acid sequences 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. Generally The Kabat definition is based on the variability of the sequences, the Chothia definition is based on the location of the 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 in antibodies.

[0022] The invention also provides a nucleic acid molecule encoding an anti-STEAP2 antibody or a portion thereof. For example, the invention provides a nucleic acid molecule 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 having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0023] The invention also provides a nucleic acid molecule 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 having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto.

[0024] The present invention also provides a nucleic acid molecule encoding any of the HCDR1 amino acid sequences listed in Table 1. 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 a nucleic acid molecule encoding any of the HCDR2 amino acid sequences listed in Table 1. 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 a nucleic acid molecule encoding any of the HCDR3 amino acid sequences listed in Table 1. 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 a nucleic acid molecule encoding any of the LCDR1 amino acid sequences listed in Table 1. 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 a nucleic acid molecule encoding any of the LCDR2 amino acid sequences listed in Table 1. In certain embodiments, the nucleic acid molecule comprises a polynucleotide sequence selected from any of the LCDR2 nucleic acid sequences listed in Table 2, or a substantially similar sequence thereof having at least 90%, It includes substantially similar sequences having at least 95%, at least 98%, or at least 99% sequence identity.

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

[0030] The present invention also provides a nucleic acid molecule encoding an HCVR, wherein the HCVR comprises 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 a nucleic acid molecule encoding both an HCVR and an LCVR, wherein the HCVR comprises an amino acid sequence of any of the HCVR amino acid sequences listed in Table 1, and the LCVR comprises an 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 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 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 an LCVR, both of which are derived from the same anti-STEAP2 antibody listed in Table 1.

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

[0034] The present invention includes anti-STEAP2 antibodies having an altered glycosylation pattern. In some embodiments, modifications to remove unwanted glycosylation sites, or antibodies lacking a fucose moiety to increase, for example, antibody-dependent cell 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 features 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 the anti-STEAP2 antibody. Additional combination therapies and combination formulations that include the anti-STEAP2 antibody of the present invention are disclosed elsewhere herein.

[0036] In another aspect, the present invention provides a method of treatment for targeting / killing tumor cells that express STEAP2 using the anti-STEAP2 antibody of the present invention, the method comprising administering to a subject in need thereof a therapeutically effective amount of a pharmaceutical composition comprising the anti-STEAP2 antibody of the present invention. Optionally, the anti-STEAP2 antibody (or antigen-binding fragment thereof) can be used to treat prostate cancer or can be modified to enhance cytotoxicity by methods including, but not limited to, a modified Fc domain that increases ADCC (e.g., see 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 the 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 for medical use, comprising an anti-STEAP2 antibody or antigen-binding fragment disclosed herein, or a STEAP2×CD3 bispecific antibody. In one aspect, the present invention relates to a compound for medical use, comprising an antibody-drug conjugate (ADC) disclosed herein.

[0038] In yet another aspect, the present invention provides a monospecific anti-STEAP2 antibody for diagnostic use, such as an imaging reagent.

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

[0040] In another aspect, the present 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 aspect, the present 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 an antibody or antigen-binding fragment that competes with a reference antibody comprising the HCVR / LCVR amino acid sequence pairs set forth in Table 1 for binding to human STEAP2. In another aspect, the present invention provides an antibody or antigen-binding fragment that competes 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 for binding to human STEAP2.

[0042] The present invention further provides an antibody or antigen-binding fragment thereof that binds to the same epitope on human STEAP2 as a reference antibody comprising the HCVR / LCVR amino acid sequence pairs set forth in Table 1. In another aspect, the antibody or antigen-binding fragment 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 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 complementarity-determining regions (CDRs) of a heavy-chain variable region (HCVR) having the amino acid sequence set forth in Table 1, and CDRs of a light-chain variable region (LCVR) having the amino acid sequence set forth in Table 1. In another aspect, the isolated antibody or antigen-binding fragment comprises 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 aspect, the isolated antibody or antigen-binding fragment comprises an HCDR1-HCDR2-HCDR3-LCDR1-LCDR2-LCDR3 domain 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, 100-102-104-60-62-64, 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, 316-318-320-324-326-328, 332-334-336-340-342-344, 348-350-352-356-358-360, 364-366-368-372-374-376, and 380-382-384-388-390-392.

[0044] In another aspect, the present invention provides an isolated antibody or antigen-binding fragment thereof that binds to human STEAP2, wherein the antibody or antigen-binding fragment comprises (a) 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; and (b) 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. In a further aspect, the isolated 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, the isolated antibody or antigen-binding fragment according to claim 10.

[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 auristatin, maytansinoid, tubulysin, tomamycin derivative, or dolastatin derivative. Optionally, 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.

Chemical formula

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

Chemical formula

[0049] In some embodiments, the antibody-drug conjugate comprises an anti-STEAP2 antibody or a fragment thereof, and the following formula:

Chemical formula

Chemical formula

[0050] In some embodiments, the antibody-drug conjugate comprises an anti-STEAP2 antibody or a fragment thereof, and the following formula:

Chemical formula

Chemical formula

[0051] In some embodiments, the antibody-drug conjugate comprises an anti-STEAP2 antibody or a fragment thereof, and has the following formula:

Chem.

Chem.

[0052] In some embodiments, the binding contacts the antibody or a fragment thereof via the sulfur moiety of a cysteine residue.

[0053] In some embodiments, the antibody-drug conjugate comprises an anti-STEAP2 antibody or a fragment thereof and

Chem.

Chem.

[0054] In some embodiments, the binding contacts the antibody or a fragment thereof via the nitrogen moiety of a lysine residue.

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

[0056] According to another aspect, the present invention provides a bispecific antigen-binding molecule (e.g., an antibody) that binds to STEAP2 and CD3. Such bispecific antigen-binding molecules are also herein referred to as "anti-STEAP2 / anti-CD3 bispecific molecules", "anti-CD3 / anti-STEAP2 bispecific molecules", or "STEAP2×CD3 bsAb". The anti-STEAP2 / anti-CD3 bispecific molecule is 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 the direct killing (cytolysis) of target tumor cells by activated T cells. Thus, the anti-STEAP2 / anti-CD3 bispecific molecules of the present invention are useful, inter alia, for treating diseases and disorders associated with or caused by STEAP2-expressing tumors (e.g., prostate cancer). bispecific molecules", or "STEAP2×CD3 bsAb" for short. The anti-STEAP2 / anti-CD3 bispecific molecule is 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 the direct killing (cytolysis) of target tumor cells by activated T cells. Thus, the anti-STEAP2 / anti-CD3 bispecific molecules of the present 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 to human CD3 and a second antigen-binding domain that specifically binds to STEAP2. The invention includes an anti-STEAP2 / anti-CD3 bispecific molecule (e.g., a bispecific antibody) 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 and separate HCVRs paired with a common LCVR. For example, as shown in Example 4 herein, a first antigen-binding domain that specifically binds to CD3 comprises an HCVR derived from an anti-CD3 antibody paired with an LCVR derived 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 to STEAP2 comprises an HCVR / LCVR derived from an anti-STEAP2 antibody. In other words, in the exemplary molecules disclosed herein, the pairing of the HCVR derived from an anti-CD3 antibody with the LCVR derived from an anti-STEAP2 antibody generates an antigen-binding domain that specifically binds to CD3 (but not to STEAP2). In such embodiments, the first and second antigen-binding domains comprise different anti-CD3 and anti-STEAP2 HCVRs but share a common anti-STEAP2 LCVR. In other embodiments, the bispecific antigen-binding molecule comprises 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 NO: 1892, 1894, and 1896, respectively. Genetically modified mice can be used to produce a fully humanized bispecific antigen-binding molecule comprising two different heavy chains that associate with the same light chain comprising a variable domain derived from one of two different human light-chain variable region gene segments. Alternatively, the variable heavy chain may be paired with one common light chain and recombinantly expressed in a host cell.Thus, the antibody of the present invention may 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 human Jκ1 gene segment.

[0058] The present invention provides an anti-CD3 / anti-STEAP2 bispecific molecule, wherein the 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 on March 27, 2014 and PCT / US2016 / 044732 filed on July 29, 2016.

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

[0060] According to certain embodiments, the present invention provides an anti-CD3 / anti-STEAP2 bispecific molecule, wherein the first antigen-binding domain that specifically binds to CD3 comprises a heavy chain variable region (HCVR) having the amino acid sequences 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 an anti-CD3 / anti-STEAP2 bispecific molecule, wherein the first antigen-binding domain that specifically binds to CD3 comprises a light chain variable region (LCVR) having the amino acid sequences 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 an anti-CD3 / anti-STEAP2 bispecific molecule, wherein the first antigen-binding domain that specifically binds to CD3 comprises a pair of HCVR and LCVR (HCVR / LCVR) amino acid sequences set forth in Tables 9, 11, 12, 15, and 17 herein.

[0063] The present invention also provides an anti-CD3 / anti-STEAP2 bispecific molecule, wherein the first antigen-binding domain that specifically binds to CD3 comprises a heavy chain CDR3 (HCDR3) domain having the amino acid sequences 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 therewith, and a light chain CDR3 (LCDR3) domain having the amino acid sequences 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 therewith.

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

[0065] The present invention also provides an anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule, wherein the first antigen-binding domain that specifically binds to CD3 comprises a heavy chain CDR1 (HCDR1) domain having the amino acids described 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, a heavy chain CDR2 (HCDR2) domain having the amino acids described 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, a heavy chain CDR3 (HCDR3) domain having the amino acids described 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, a light chain CDR1 (LCDR1) domain having the amino acid sequences described 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, a light chain CDR2 (LCDR2) domain having the amino acid sequences described 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 CDR3 (LCDR3) domain having the amino acid sequences described 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.

[0066] A particular non-limiting and exemplary anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule of the present invention comprises a first antigen-binding domain that specifically binds to CD3, which comprises an HCDR1-HCDR2-HCDR3-LCDR1-LCDR2-LCDR3 domain having the amino acid sequences described in Tables 9, 11, 12, 15, and 17 herein, respectively.

[0067] The present invention further provides a bispecific antigen-binding molecule, 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 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 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) selected from the group consisting of SEQ ID NOs: 1730, 1762, and 1866, and light-chain complementarity-determining regions (LCDR1, LCDR2, and LCDR3) derived 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 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), 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.

[0070] In a further aspect, the present invention provides a bispecific antigen-binding molecule, wherein the first antigen-binding domain that specifically binds to human CD3 comprises 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 aspect, the present invention provides an antigen-binding molecule, wherein 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), and 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, 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, 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] A specific non-limiting and exemplary anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule of the present invention 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), and a first antigen-binding domain that specifically binds to CD3.

[0073] In a further embodiment, an exemplary anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule of the present invention comprises a bispecific antigen-binding molecule, 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: 1907-1908-1909.

[0074] The present invention also provides an anti-CD3 / anti-STEAP2 bispecific molecule, wherein the second antigen-binding domain that specifically binds to 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 having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

[0075] The present invention also provides an anti-CD3 / anti-STEAP2 bispecific molecule, wherein the second antigen-binding domain that specifically binds to 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 having 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 amino acid sequence pair of HCVR and LCVR (HCVR / LCVR) of SEQ ID NO: 250 / 258.

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

[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 NO: 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 NO: 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 having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto, and a heavy-chain CDR2 (HCDR2) domain having an amino acid sequence selected from the group consisting of SEQ ID NO: 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, or a substantially similar sequence having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto, and a heavy-chain CDR3 (HCDR3) domain having an amino acid sequence selected from the group consisting of SEQ ID NO: 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, or a substantially similar sequence having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto, and a light-chain CDR1 (LCDR1) domain having an amino acid sequence selected from the group consisting of SEQ ID NO: 12, 28, 44, 60, 116, 132, 148, 164, 180, 196, 212, 228, 244, 260, 276, 292, 308, 324, 340, 356, 372, and 388, or a substantially similar sequence having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity thereto, and a light-chain CDR2 (LCDR2) domain having an amino acid sequence selected from the group consisting of SEQ ID NO: 14, 30, 46, 62, 118, 134, 150, 166, 182, 198, 214, 230, 246, 262, 278, 294, 310, 326, 342, 358, 374, and 390, or at least 90%,A light chain CDR3 (LCDR3) domain having at least 95%, at least 98%, or at least 99% sequence identity with its substantially similar sequence, and 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 having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with its substantially similar sequence.

[0080] Certain non-limiting and exemplary anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules of the invention comprise a second antigen-binding domain that specifically binds to STEAP2 and comprises 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 related embodiments, the invention provides an anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule, wherein the second antigen-binding domain that specifically binds to 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 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 antibody or antigen-binding fragment of the anti-STEAP2 antibodies of the invention. In a further aspect, the 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 aspect, the bispecific antigen-binding molecule binds to human cells expressing human STEAP2.

[0084] In another aspect, the present invention provides a bispecific antigen-binding molecule that inhibits tumor growth in immunodeficient mice bearing human prostate cancer xenografts.

[0085] In certain embodiments, the anti-CD3 antibodies, antibody-binding fragments, and bispecific antibodies of the present invention were made 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 present invention provides a bispecific antigen-binding molecule, wherein the second antigen-binding domain competes 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) for binding to human STEAP2, 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 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.

[0087] In some embodiments, the present invention provides a bispecific antigen-binding molecule, wherein the first antigen-binding domain competes 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) for binding to human CD3. 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. In some embodiments, the present invention provides a bispecific antigen-binding molecule, wherein the first antigen-binding domain competes 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 for binding to human CD3.

[0088] In some embodiments, the present invention provides a bispecific antigen-binding molecule, wherein the first antigen-binding domain competes 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 for binding to human CD3, and the second 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. The second 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 aspect, 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 further 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 aspects, 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 a nucleic acid molecule encoding any of the HCVR, LCVR, or CDR sequences of the anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule disclosed herein, a nucleic acid molecule comprising the polynucleotide sequences described in Tables 2, 10, 13, 14, 16, and 18 herein, and a nucleic acid molecule comprising two or more of the polynucleotide sequences described in Tables 2, 10, 13, 14, 16, and 18 in any of their functional combinations or arrangements. A recombinant expression vector carrying the nucleic acid of the present invention, and a host cell into which such a vector has been introduced are also included in the present invention in the same manner as a method for producing an antibody by culturing the host cell under conditions that allow the production of the antibody and recovering the produced antibody.

[0091] The present invention includes an anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule in which any of the aforementioned antigen-binding domains that specifically bind to CD3 is combined, linked, or associated with any of the aforementioned antigen-binding domains that specifically bind to STEAP2 to form a bispecific antigen-binding molecule that binds to CD3 and STEAP2.

[0092] The present invention includes anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules having an altered glycosylation pattern. In some applications, modifications to remove unwanted glycosylation sites, or antibodies lacking a fucose moiety to increase, for example, antibody-dependent cell 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 present invention provides a pharmaceutical composition comprising an anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule disclosed herein and a pharmaceutically acceptable carrier. In related aspects, the present 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 is 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 method of treatment for targeting / killing tumor cells that express STEAP2 using the anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule of the present invention, the method of treatment 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 includes the use of the anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule of the present invention in the manufacture of a medicament for the treatment of a disease or disorder associated with or caused by STEAP2-expressing cells. double specific antigen-binding molecule.

[0096] Other embodiments will become apparent from a review of the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0097]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

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

[0099] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. As used herein, the term "about," when used in reference to a specific recited numerical value, means that the value can vary by no more than 1% from the recited value. For example, as used in the present invention, the expression "about 100" includes 99 and 101, as well as all values in between (e.g., 99.1, 99.2, 99.3, 99.4, etc.).

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

[0101] Definitions As used herein, the expression "CD3" refers to an antigen that is expressed on T cells as part of the multi-molecular T cell receptor (TCR) and consists of a homodimer or heterodimer formed from the association of two of the four receptor chains, namely CD3-epsilon, CD3-delta, CD3-zeta, and CD3-gamma. Human CD3-epsilon contains the amino acid sequence set forth in SEQ ID NO: 1897, and human CD3-delta contains 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 each protein, polypeptide, or protein fragment, unless specifically identified as being from a non-human species. Thus, the expression "CD3" means human CD3 unless specifically identified as being derived from a non-human species, such as "mouse CD3", "monkey CD3", etc.

[0102] As used herein, "antibody that binds to CD3" or "anti-CD3 antibody" includes 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 that lack a transmembrane domain or are not associated with the cell membrane, such as monomeric and dimeric CD3 constructs.

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

[0104] As used herein, the expression "STEAP2" refers to six-transmembrane epithelial antigen of the prostate 2. 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 has been 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 on human chromosome region 7q21. For example, refer to the amino acid sequence of human STEAP2 set forth in SEQ ID NO: 1899. For reference.

[0105] As used herein, "antibody that binds to STEAP2" or "anti-STEAP2 antibody" includes an antibody that specifically recognizes STEAP2 and antigen-binding fragments thereof.

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

[0107] As used in the present invention, the term "antibody" means any antigen-binding molecule or molecular complex that includes at least one complementarity-determining region (CDR) that specifically binds to or interacts with a specific antigen (e.g., STEAP2 or CD3). The term "antibody" includes immunoglobulin molecules comprising four polypeptide chains interconnected by disulfide bonds, two heavy (H) chains and two light (L) chains, and multimers thereof (e.g., IgM). Each heavy chain includes a heavy chain variable region (abbreviated herein as HCVR or V H for short) and a heavy chain constant region. The heavy chain constant region includes three domains, C H 1, C H 2, and C H 3. Each light chain includes a light chain variable region (abbreviated herein as LCVR or V L for short) and a light chain constant region. The light chain constant region includes one domain (C L 1). The V H region and the V L region can be further subdivided into hypervariable regions called complementarity-determining regions (CDRs) interspersed with more conserved regions called framework regions (FRs). Each V H and V L consists of three CDRs and four FRs arranged in the order of FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4 from the amino terminus to the carboxy terminus. In different embodiments of the present invention, the FRs of the anti-STEAP2 antibody or anti-CD3 antibody (or antigen-binding portion thereof) may be identical to the human germline sequences, or may be modified naturally or artificially. Amino acid consensus sequences can be defined based on the alignment analysis of two or more CDRs.

[0108] When used in the context of the present invention, the term "antibody" includes antigen-binding fragments of whole antibody molecules. The terms "antigen-binding portion of an antibody", "antigen-binding fragment of an antibody", and the like, as used herein, include polypeptides or glycoproteins that are natural, enzymatically obtainable, synthetic, or genetically engineered and that specifically bind an antigen to form a complex. Antibody-binding fragments of an antibody can be derived from whole antibody molecules using any suitable standard techniques, such as protein digestion techniques or recombinant genetic engineering techniques related to the manipulation and expression of DNA encoding antibody variable domains and optionally constant domains. Such DNA is known and / or readily available from, for example, commercial sources, DNA libraries (including phage-antibody libraries), or can be synthesized. The DNA can be sequenced and manipulated using chemical or molecular biological techniques, for example, to arrange one or more variable domains and / or constant domains in appropriate configurations, introduce codons, generate cysteine residues, modify, add, or delete amino acids, and the like.

[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) amino acid residues mimicking the hypervariable regions of an antibody (e.g., isolated complementarity-determining regions (CDRs) such as CDR3 peptides), or minimal recognition units consisting of constrained FR3-CDR3-FR4 peptides. Also included within the meaning of the expression "antigen-binding fragment" as used herein are other engineered molecules such as 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 shark variable IgNAR domains.

[0110] An antigen-binding fragment of an antibody typically comprises at least one variable domain. The variable domain can be of any size or amino acid composition and generally contains at least one CDR that is adjacent to or in-frame with one or more framework sequences. V L domain bound to the V H domain in an antibody-binding fragment having a V H domain and a V L domain can be arranged relative to each other in any suitable arrangement. For example, the variable region is a dimer and contains V H -V H , V H -V L or V L -V L dimers. Alternatively, the antigen-binding fragment of an antibody can contain a monomeric V H domain or a V L domain.

[0111] In certain embodiments, an antigen-binding fragment of an antibody can contain at least one variable domain covalently attached to at least one constant domain. Non-limiting and exemplary configurations of variable and constant domains that can be found within the antigen-binding fragments of the antibodies of the present 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 conformation of variable and constant domains, including any of the exemplary conformations listed above, the variable and constant domains may be directly linked to each other or may be linked by a full or partial hinge region or linker region. The hinge region can consist of at least two (e.g., 5, 10, 15, 20, 40, 60 or more) amino acids that provide a flexible or semi-flexible linkage between adjacent variable and / or constant domains in a single polypeptide molecule. Moreover, the antigen-binding fragments of the antibodies of the present invention can include homodimers or heterodimers (or other multimers) of any of the variable domain conformations and constant domain conformations listed above in non-covalent association with each other and / or with one or more monomeric V H domains or V L domains.

[0112] Similar to full antibody molecules, antibody binding fragments can be monospecific or multispecific (e.g., bispecific). Multispecific antigen-binding fragments of antibodies typically will include at least two different variable domains, each of which can specifically bind to a distinct 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 the antibodies of the present invention using conventional techniques available in the art.

[0113] The antibodies of the present invention can function via 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 non-specific cytotoxic cells (e.g., natural killer (NK) cells, neutrophils, and macrophages) expressing Fc receptors recognize the bound antibody on the target cell, thereby resulting in the lysis of the target cell. CDC and ADCC are well-known in the art and can be measured using available assays. (See, e.g., U.S. Pat. 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 the antibody is important in the ability of the antibody to fix complement and mediate cell-dependent cytotoxicity. Thus, the isotype of the antibody may be selected based on whether it is desirable for the antibody to mediate cytotoxicity.

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

[0115] In some embodiments, the antibody of the present invention can be a recombinant human antibody. As used herein, the term "recombinant human antibody" refers to all human antibodies (described below) prepared, expressed, produced or isolated by recombinant means, such as antibodies expressed using a recombinant expression vector transfected into a host cell, antibodies isolated from a recombinant combinatorial human antibody library (described below), antibodies isolated from an animal (e.g., mouse) 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 to 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, if transgenic animals for human Ig sequences are used, in vivo somatic mutagenesis), and thus the amino acid sequences of the V H and V L regions of the recombinant antibody are related to and derived from the human germline V H and V L sequences, but are sequences that do not naturally occur in 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 comprises a stable four-chain construct of about 150-160 kDa in which the dimers are held together by interchain heavy chain disulfide bonds. In a second form, the dimers are not linked via interchain disulfide bonds and a molecule of about 75-80 kDa consisting of covalently linked light and heavy chains is formed (half antibody). These forms were 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 related to the hinge region isotype of the antibody. A single amino acid substitution in the hinge region of the human IgG4 hinge can significantly reduce the occurrence of the second form to levels typically observed using the human IgG1 hinge (Angal et al. (1993) Molecular Immunology 30:105). The present invention encompasses antibodies having one or more mutations in the hinge, C H 2 or C H 3 region, which may be desirable, for example, in production to improve the yield of the desired antibody form.

[0118] The antibodies of the present invention can be isolated antibodies. As used herein, "isolated antibody" means an antibody that has been separated and / or recovered from an identified antibody and 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 the antibody naturally occurs or is naturally produced, is an "isolated antibody" for the 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, the isolated antibody may be substantially free of other cellular materials and / or chemical substances.

[0119] The present invention also includes single-arm antibodies that bind to STEAP2. As used herein, "single-arm antibody" means an antigen-binding molecule comprising a single antibody heavy chain and a single antibody light chain. The single-arm antibodies of the present invention can comprise any of the HCVR / LCVR or CDR amino acid sequences set forth 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 regions and / or CDR regions of the heavy chain variable domain and the light chain variable domain as compared to the corresponding germline sequences. Such mutations can be readily identified by comparing the amino acid sequences disclosed herein to, for example, germline sequences available 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, wherein one or more amino acids within one or more framework and / or CDR regions are mutated to the corresponding residue(s) of the germline sequence from which the antibody is derived, or to the corresponding residue(s) of 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"). One of ordinary skill in the art can readily produce a number of antibodies and antibody-binding fragments containing one or more individual germline mutations or combinations thereof starting from the heavy chain variable region sequences and light chain variable region sequences disclosed herein. In certain embodiments, V H domain and / or V LAll framework and / or CDR residues within the domain are mutated back to the residues 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 residues are within the first 8 amino acids of FR1, or the mutated residues are within the last 8 amino acids of FR4, or the mutated residues are only found within CDR1, CDR2 or CDR3. In other embodiments, one or more of the framework and / or CDR residues are mutated to the corresponding residues of a different germline sequence (i.e., a germline sequence different from the germline sequence from which the antibody was originally derived). Further, the antibodies of the invention may contain any combination of two or more germline mutations within the framework and / or CDR regions, e.g., certain individual residues mutate to the corresponding residues of a particular germline sequence, while certain other residues different from the original germline sequence are maintained or mutate to the corresponding residues of 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 (where appropriate) biological properties of an antagonist or agonist, reduced immunogenicity, etc. Antibodies and antibody binding fragments obtained in this general manner are encompassed within the scope of the invention.

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

[0122] The term "epitope" refers to an antigenic determinant that interacts with the 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 and may. Thus, different antibodies can bind to different regions on an antigen and can have different biological effects. Epitopes can be conformational or linear. Conformational epitopes are produced by amino acids that are spatially juxtaposed from different segments of a linear polypeptide chain. Linear epitopes are produced by adjacent amino acid residues in a polypeptide chain. In certain situations, an epitope can include a portion of a carbohydrate, phosphoryl group, or sulfonyl group on an antigen.

[0123] The term "substantial identity" or "substantially identical" when referring to a nucleic acid or a fragment thereof, when optimally aligned with another nucleic acid (or its complementary strand) with appropriate nucleotide insertions or deletions, as discussed below, when measured by any well-known algorithm for sequence identity, such as FASTA, BLAST, or Gap, indicates that there is at least about 95%, more preferably at least about 96%, 97%, 98%, or 99% nucleotide sequence identity of nucleotide bases. 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 share at least 95% sequence identity, more preferably at least 98% or 99% sequence identity when optimally aligned by programs such as GAP or BESTFIT using a default gap weight. Preferably, the non-identical residue positions differ only 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). In general, conservative amino acid substitutions are not expected to substantially alter the functional properties of the protein. If two or more amino acid sequences differ from each other only by conservative substitutions, the percentage of 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, for example, Pearson (1994) Methods Mol. Biol. 24:307-331, which is 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 having 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 having a non-negative value in the PAM250 log-likelihood matrix.

[0125] Sequence similarity to a polypeptide, also called sequence identity, is typically measured using sequence analysis software. Protein analysis software matches similar sequences using similar measurements assigned to various substitutions, deletions, and other modifications, including conservative amino acid substitutions. For example, GCG software can be used with default parameters for determining 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, and programs such as Gap and Bestfit. See, for example, GCG version 6 .1. Polypeptide sequences can also be compared using FASTA with default or recommended parameters, which is a program in GCG version 6.1. FASTA (e.g., FASTA2 and FASTA3) provides the alignment of the best overlapping regions between the query sequence and the search sequence and the percent sequence identity (Pearson (2000) supra). Another preferred algorithm when 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. See, for example, Altschul et al. (1990) J. Mol. Biol. 215:403-410 and Altschul et al. (1997) Nucleic Acids Res. 25:3389-402, each of which is incorporated herein by reference.

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

[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 within one or more framework and / or CDR regions are mutated to the corresponding residue(s) of the germline sequence from which the antibody is derived, or to the corresponding residue(s) of another human germline sequence, or to conservative amino acid substitutions of the corresponding germline residue(s) (such sequence changes are collectively referred to herein as "germline mutations"), and the detectable binding to the CD3 antigen is weak or absent. Some such exemplary antibodies that recognize CD3 are described in Tables 12 and 18 herein.

[0128] Furthermore, the antibodies of the present invention may contain any combination of two or more germline mutations within the framework and / or CDR regions. For example, certain individual residues mutate to the corresponding residues of a particular germline sequence, while certain other residues different from the original germline sequence are maintained or mutate to the corresponding residues of 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, reduced or decreased binding affinity, improved or enhanced pharmacokinetic properties, and reduced immunogenicity. Antibodies and antigen-binding fragments obtained in this general manner in consideration of the guidance of the present 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 amino acid sequences, LCVR amino acid sequences, and / or CDR amino acid sequences disclosed herein having one or more conservative substitutions. For example, the present invention provides HCVR amino acid sequences, LCVR amino acid sequences, and / or CDR amino acid sequences having conservative amino acid substitutions, such as 10 or fewer, 8 or fewer, 6 or fewer, or 4 or fewer conservative amino acid substitutions, respectively, relative to any of the HCVR amino acid sequences, LCVR amino acid sequences, and / or CDR amino acid sequences described 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 as compared to the corresponding germline sequences from which the individual antigen-binding domains are derived, but maintain or improve the desired weak to undetectable binding to the CD3 antigen. "Conservative amino acid substitution" means that 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 change the functional properties of the protein, i.e., the amino acid substitution maintains or improves the desired weak to undetectable binding affinity in the case of an anti-CD3 binding molecule. Amino acids with similar chemical properties Examples of groups of amino acids having a side chain 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 substituents are valine-leucine-isoleucine, phenylalanine-tyrosine, lysine-arginine, alanine-valine, glutamic acid-aspartic acid and asparagine-glutamine. Alternatively, a conservative substitution is any change having 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 having a non-negative value in the PAM250 log-likelihood matrix.

[0130] The present invention also includes an antigen-binding molecule comprising an antigen-binding domain having an HCVR and / or CDR amino acid sequence that is 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. As used herein, the terms "substantial identity" or "substantially identical" in reference to amino acid sequences means that two amino acid sequences share at least 95% sequence identity, more preferably at least 98% or 99% sequence identity when optimally aligned using a program such as GAP or BESTFIT with a defined gap weight. Preferably, the residue positions that are not identical differ only by conservative amino acid substitutions. When two or more amino acid sequences differ from each other only 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 of skill in the art. See, for example, Pearson (1994) Methods Mol. Biol. 24:307-331.

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

[0132] Once obtained, antigen-binding domains containing one or more germline mutations were tested for reduced binding affinity using one or more in vitro assays. Antibodies that recognize a particular antigen are typically screened for their intended 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. Containing one or more antigen-binding domains obtained by this general method Bispecific antigen-binding molecules are also included within the scope of the present invention and have been found to be advantageous as binding activity-driven tumor therapies.

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

[0134] Binding properties of the antibody As used herein, the term "binding" in the context of an antibody, immunoglobulin, antibody binding fragment, or Fc-containing protein binding to a given antigen, such as 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, when the binding affinity is determined using an antigen as a ligand and an antibody, Ig, antibody binding fragment, or Fc-containing protein as an analyte (or anti-ligand), for example, by surface plasmon resonance (SPR) technology in a BIAcore3000 instrument, it is typically about 10 -7 M or less, such as about 10 -8 M or less, such as about 10 -9 M or less K D value. Cell-based binding strategies such as fluorescence-activated cell sorting (FACS) binding assays are also routinely used, and FACS data correlates well with other methods such as radioligand competition 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, the antibodies or antigen-binding proteins of the present invention bind to a given antigen or cell surface molecule having an affinity corresponding to a K D value that is at least 10-fold lower than its affinity for binding to non-specific antigens (e.g., BSA, casein). According to the present invention, a K that is 10-fold or less lower than that for non-specific antigens DThe affinity of the antibody corresponding to the value can be regarded as undetectable binding, but such an antibody can pair with the second antigen-binding arm for producing the bispecific antibody of the present invention.

[0137] The term "K D "(M) refers to the dissociation equilibrium constant of a specific antibody-antigen interaction, or the dissociation equilibrium constant of an antibody or antibody-binding fragment that binds to an antigen. There is an inverse relationship between K D and the binding affinity. Thus, the smaller the K D value, the higher, i.e., stronger, the affinity. Therefore, the terms "higher affinity" or "stronger affinity" relate to a higher ability to form an interaction, i.e., a smaller K D value. Conversely, the terms "lower affinity" or "weaker affinity" relate to a lower ability to form an interaction, i.e., a larger K D value. Depending on the situation, a higher binding affinity (or K D ) of a specific molecule (e.g., an antibody) for an interaction partner molecule (e.g., antigen X) compared to the binding affinity of the molecule (e.g., an antibody) for another interaction partner molecule (e.g., antigen Y) is represented as a binding ratio determined by dividing the larger K D value (lower, or weaker, affinity) by the smaller K D (higher, or stronger, affinity), and is represented, for example, as 5-fold or 10-fold higher binding affinity in some cases.

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

[0139] The term "k a "(M-1×sec-1 or 1 / M) refers to the association rate constant of a specific 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 of a specific antibody - antigen interaction, or the association equilibrium constant of an antibody or antibody - binding fragment. The association equilibrium constant is k a divided by k d to obtain.

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

[0142] In one embodiment, the decrease in binding can be defined as an increase in the EC 50 antibody concentration that enables half - maximal binding to target cells.

[0143] In another embodiment, the EC 50 value represents the concentration of an antibody of the present invention that induces half - maximal depletion of target cells by T - cell cytotoxic activity. Thus, an increase in cytotoxic activity (e.g., T - cell - mediated tumor cell death) is observed as a decrease in EC 50 , or half of the maximum effective concentration value.

[0144] Bispecific antigen - binding molecule The antibodies of the present invention can be monospecific, bispecific or multispecific. Multispecific antibodies can be specific for different epitopes of one 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 antibody or anti-STEAP2 / anti-CD3 bispecific antibody of the present invention can be linked to or co-expressed with another functional molecule, such as another peptide or protein. For example, the antibody or fragment thereof can be functionally linked (e.g., by chemical conjugation, genetic fusion, non-covalent association, etc.) to one or more other molecular entities, such as another antibody or antibody fragment, to generate a bispecific or multispecific antibody having a second or additional binding specificity.

[0145] The use of the terms "anti-CD3 antibody" or "anti-STEAP2 antibody" herein is intended to include both monospecific anti-CD3 or anti-STEAP2 antibodies and bispecific antibodies containing a CD3-binding arm and a STEAP2-binding arm. Thus, the present invention includes bispecific antibodies in which one arm of the immunoglobulin binds to human CD3 and the other arm of the immunoglobulin is specific for human STEAP2. The CD3-binding arm can comprise any of the HCVR / LCVR or CDR amino acid sequences described 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 death 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 cannot be detected in in vitro assays known in the art. The STEAP2 binding arm can comprise any of the HCVR / LCVR or CDR amino acid sequences set forth in Table 1 herein.

[0147] According to certain exemplary embodiments, the invention includes a bispecific antigen-binding molecule that specifically binds to CD3 and STEAP2. Such molecules can be referred to herein, for example, as "anti-CD3 / anti-STEAP2", or "anti-CD3×STEAP2" or "CD3×STEAP2" bispecific molecules, or other similar terms (e.g., anti-STEAP2 / anti-CD3). When used in the context of the present invention, 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.

[0148] The foregoing bispecific antigen-binding molecule that specifically binds to CD3 and STEAP2 can comprise an anti-CD3 antigen-binding molecule that binds to CD3 with a weak binding affinity exhibiting a K greater than about 40 nM as measured in an in vitro affinity binding assay. Optionally, the CD3 binding arm has a K 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

[0149] or EC D D 50It binds to CD3. (For example, as a measure in surface plasmon resonance assays). In some cases, the first antigen-binding domain specifically binds to CD3 (e.g., with weak or undetectable affinity to either or both human CD3 and cynomolgus CD3).

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

[0151] As used herein, the expression "bispecific antigen-binding molecule" means a protein, polypeptide or molecular complex comprising at least a first antigen-binding domain and a second antigen-binding domain. Each antigen-binding domain within the bispecific antigen-binding molecule comprises at least one CDR that specifically binds to a particular antigen, 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 the 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 a bispecific antigen-binding molecule. As used herein, a "multimerization domain" refers to a first multimerization 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 the multimerization domain of the two. For example, the multimerization domain may be the multimerization domain of the immunoglobulin C. H A non-limiting example of a multimerizing component is 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 molecule of the present invention typically comprises two multimerization domains, for example, two Fc domains that are each part of a separate antibody heavy chain. The first and second multimerization domains can be of the same IgG isotype, such as IgG1 / IgG1, IgG2 / IgG2, IgG4 / IgG4, etc. Alternatively, the first and second multimerization domains can be of different IgG isotypes, such as IgG1 / IgG2, IgG1 / IgG4, IgG2 / IgG4, etc.

[0155] In certain embodiments, the multimerization domain is an amino acid sequence of 1 to about 200 amino acids in length that contains an Fc fragment or 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 that contain or consist of leucine zippers, helix-loop-helix, or coiled-coil motifs.

[0156] The bispecific antigen-binding molecules of the invention can be made using any bispecific antibody format or technology. For example, an antibody or fragment thereof having a first antigen-binding specificity can be functionally linked (e.g., by chemical conjugation, genetic fusion, non-covalent binding, etc.) to one or more other molecular entities, such as another antibody or antibody fragment having a second antigen-binding property, to generate a bispecific antigen-binding molecule. Specific exemplary bispecific formats that can be used in the context of the present invention include, for example, scFv-based formats or diabody bispecific formats, IgG-scFv fusions, dual variable domain (DVD)-Ig, Quadroma, knobs-into-holes, common light chain (e.g., common light chain with knobs-into-holes), CrossMab, CrossFab, (SEED) body, leucine zipper, Duobody, IgG1 / IgG2, dual action Fab (DAF)-IgG, and Mab 2Bispecific formats are included, but not limited to these (for a review of the formats described above, see, for example, Klein et al. 2012, mAbs 4:6, 1-11, and the references cited therein).

[0157] In the context of the bispecific antigen-binding molecules of the invention, the multimerization domain, such as the Fc domain, may contain one or more amino acid changes (e.g., insertions, deletions or substitutions) compared to the wild-type, naturally occurring Fc domain. For example, the invention encompasses bispecific antigen-binding molecules that include one or more modifications in the Fc domain that result in a modified binding interaction (e.g., enhanced or decreased) between Fc and FcRn. In one embodiment, the bispecific antigen-binding molecule contains modifications in the C H 2 or C H 3 region that enhance the affinity of the Fc domain for FcRn in an acidic environment (e.g., within an endosome in the pH range of about 5.5 to about 6.0). Non-limiting examples of such Fc modifications include, for example, modifications at position 250 (e.g., E or Q), positions 250 and 428 (e.g., L or F), position 252 (e.g., L / Y / F / W or T), position 254 (e.g., S or T), and position 256 (e.g., S / R / Q / E / D or T), or modifications at position 428 and / or 433 (e.g., L / R / S / P / Q or K) and / or position 434 (e.g., H / F or Y), or modifications at position 250 and / or 428, or positions 307 or 308 (e.g., 308F, V308F), and position 434 In one embodiment, the modifications include modifications of 428L (e.g., M428L) and 434S (e.g., N434S), modifications of 428L, 259I (e.g., V259I), and 308F (e.g., V308F), modifications of 433K (e.g., H433K) and 434 (e.g., 434Y), modifications of 252, 254, and 256 (e.g., 252Y, 254T, and 256E), modifications of 250Q and 428L (e.g., T250Q and M428L), modifications of 307 and / or 308 (e.g., 308F or 308P).

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

[0159] In certain embodiments, the Fc domain can be a chimeric combination of Fc sequences derived from two or more immunoglobulin isotypes. For example, the chimeric Fc domain can be a C H 2 region-derived C H 2 sequence, part or all, and a C derived from human IgG1, human IgG2, or human IgG4H It can include some or all of the 3 - arrays. The chimeric Fc domain can also include a chimeric hinge region. For example, the chimeric hinge can include 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 can be included in any of the antigen - binding molecules described herein is, from the N - terminus to the C - terminus, [IgG4C H 1]-[IgG4 upper hinge]-[IgG2 lower hinge]-[IgG4CH2]-[IgG4CH3]. Another example of a chimeric Fc domain that can be included in any of the antigen - binding molecules described herein is, from the N - terminus to the C - terminus, [IgG1C H 1]-[IgG1 upper hinge]-[IgG2 lower hinge]-[IgG4CH2]-[IgG1CH3]. These and other examples of chimeric Fc domains that can be included in any of the antigen - binding molecules of the present invention are described in U.S. Patent Publication No. 2014 / 0243504, published on August 28, 2014, the entire disclosure of which is incorporated herein by reference. Chimeric Fc domains having these general structural arrangements, and variants thereof, can alter Fc receptor binding, which affects Fc effector functions.

[0160] In certain embodiments, the present invention provides that the heavy - chain constant region (CH) region has at least 95%, at least 96%, at least 97%, at least Provided is an antibody heavy chain comprising an amino acid sequence that is at least 98%, at least 99% identical. In some embodiments, the heavy chain constant region (CH) region comprises an amino acid sequence selected from the group consisting 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, and SEQ ID NO: 1920.

[0161] In other embodiments, the present invention provides an antibody heavy chain wherein the Fc domain comprises an amino acid sequence that is at least 95%, at least 96%, at least 97%, at least 98%, 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 variant The antibodies and bispecific antigen-binding molecules of this specification 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 as compared to the corresponding germline sequences from which the individual antigen-binding domains are derived. Such mutations can be readily identified by comparing the amino acid sequences disclosed herein to germline sequences available, for example, from public antibody sequence databases. The antigen-binding molecules of the invention may comprise antigen-binding domains derived from any of the exemplary amino acid sequences disclosed herein, and one or more amino acids within one or more frameworks and / or CDR regions may be mutated to the corresponding residue(s) of the germline sequence from which the antibody is derived, or to the corresponding residue(s) of 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"). One of ordinary skill in the art can readily generate a number of antibodies and antibody-binding fragments containing one or more individual germline mutations or combinations thereof, starting from the heavy chain variable region sequences and light chain variable region sequences disclosed herein. In certain embodiments, V H domain and / or V LAll framework and / or CDR residues within the domain mutate back to the residues 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 residues are found within the first 8 amino acids of FR1, or the mutated residues are found within the last 8 amino acids of FR4, or the mutated residues are found only within CDR1, CDR2 or CDR3. In other embodiments, one or more of the framework and / or CDR residue(s) mutate to the corresponding residue(s) of a different germline sequence (i.e., a germline sequence different from the germline sequence from which the antigen-binding domain was originally derived). Further, 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 mutate to the corresponding residues of a particular germline sequence, while certain other residues different from the original germline sequence are maintained or mutate to the corresponding residues of a different germline sequence. Once obtained, an antigen-binding domain 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 (where appropriate) biological properties of an antagonist or agonist, 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 that one or both of the antigen-binding domains have one or more conservative substitutions An antigen-binding molecule comprising a variant of any of the HCVR amino acid sequence, LCVR amino acid sequence, and / or CDR amino acid sequence disclosed herein. For example, the present invention provides an antigen-binding domain having an HCVR amino acid sequence, LCVR amino acid sequence, and / or CDR amino acid sequence having, 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 amino acid sequence, LCVR amino acid sequence, and / or CDR amino acid sequence disclosed herein, and an antigen-binding molecule comprising the same. "Conservative amino acid substitution" means that 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 are not expected to substantially change the functional properties of the 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 having 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 having a non-negative value in the PAM250 log-likelihood matrix.

[0164] The present invention also includes an antigen-binding molecule comprising an antigen-binding domain having an HCVR, LCVR, and / or CDR amino acid sequence that is substantially identical to any of the HCVR amino acid sequences, LCVR amino acid sequences, and / or CDR amino acid sequences disclosed herein. When referring to amino acid sequences, the terms "substantial identity" or "substantially identical" mean that two amino acid sequences share at least 95% sequence identity, more preferably at least 98%, or 99% sequence identity when optimally aligned using a program such as GAP or BESTFIT with a defined gap weight. Preferably, the residue positions that are not identical differ only by conservative amino acid substitutions. When two or more amino acid sequences differ from each other only 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 of skill in the art. See, for example, Pearson (1994) Methods Mol. Biol. 24:307-331, which is incorporated herein by reference.

[0165] Sequence similarity to a polypeptide, also called sequence identity, is typically measured using sequence analysis software. Protein analysis software matches similar sequences using similar measurements assigned to various substitutions, deletions, and other modifications, including conservative amino acid substitutions. For example, GCG software can be used with default parameters for determining 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 proteins, and programs such as Gap and Bestfit. See, for example, GCG version 6.1. Polypeptide sequences can also be compared using FASTA with default or recommended parameters, which is a program in GCG version 6.1. FASTA (e.g., FASTA2 and FASTA3) provides an alignment of the best overlapping regions between the query sequence and the search sequence and percent sequence identity (Pearson (2000), supra). Another preferred algorithm when 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 , see Altschul et al. (1990) J. Mol. Biol. 215:403-410 and Altschul et al. (1997) Nucleic Acids Res. 25:3389-402, which are incorporated herein by reference, respectively.

[0166] pH-dependent binding The present invention includes anti-STEAP2 antibodies having pH-dependent binding characteristics and anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules. For example, the anti-STEAP2 antibodies of the present invention may exhibit a decrease in binding to STEAP2 at acidic pH compared to neutral pH. Alternatively, the anti-STEAP2 antibodies of the present invention may exhibit enhanced binding to STEAP2 at acidic pH compared to neutral pH. The expression "acidic pH" includes pH values of less than about 6.2, such as 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, 5.0 or less. As used herein, the expression "neutral pH" means a pH of from 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, "a decrease in binding... at acidic pH compared to neutral pH" is represented by the ratio of the K D value of the antibody that binds to the antigen at acidic pH to the K D value of the antibody that binds to the antigen at neutral pH. (Or vice versa). For example, if the antibody or its antigen-binding fragment exhibits an acidic / neutral K D ratio of about 3.0 or more, for the purposes of the present invention, the antibody or its antigen-binding fragment can be considered to exhibit "a decrease in binding to STEAP2 at acidic pH compared to neutral pH". In certain exemplary embodiments, the acidic / neutral K D ratio of the antibody or antigen-binding fragment of the present invention is 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 a decrease (or enhancement) in binding to a specific antigen at acidic pH compared to neutral pH. Additionally, 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 in which antigen binding is decreased at acidic pH relative to neutral pH.

[0169] Antibodies comprising an Fc variant According to certain embodiments of the invention, for example, provided are anti-STEAP2 antibody bispecific antigen-binding molecules, and anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules, comprising an Fc domain comprising one or more mutations that enhance or decrease antibody binding to the FcRn receptor at acidic pH compared to neutral pH. For example, the invention encompasses antibodies comprising mutations in the C H 2 or C H 3 region of the Fc domain, the mutation(s) increasing the affinity of the Fc domain for FcRn in an acidic environment (e.g., in an endosome where the pH ranges from about 5.5 to about 6.0). Such mutations can result in an increase in the serum half-life of the antibody when administered to an animal. Non-limiting examples of such Fc modifications include, for example, modifications at position 250 (e.g., E or Q), positions 250 and 428 (e.g., L or F), position 252 (e.g., L / Y / F / W or T), position 254 (e.g., S or T), and position 256 (e.g., S / R / Q / E / D or T), or modifications at position 428 and / or 433 (e.g., H / L / R / S / P / Q or K) and / or position 434 (e.g., H / F or Y), or modifications at position 250 and / or 428, or at position 307 or Modifications at the 308th position (e.g., 308F, V308F) and the 434th position are included. In one embodiment, the modifications include modifications of 428L (e.g., M428L) and 434S (e.g., N434S), modifications of 428L, 259I (e.g., V259I), and 308F (e.g., V308F), modifications of 433K (e.g., H433K) and 434 (e.g., 434Y), modifications of 252, 254, and 256 (e.g., 252Y, 254T, and 256E), modifications of 250Q and 428L (e.g., T250Q and M428L), and modifications of 307 and / or 308 (e.g., 308F or 308P).

[0170] For example, the present invention includes anti-STEAP2 antibodies and anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules that include an Fc domain comprising 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 within 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 includes antibodies and antigen-binding fragments thereof that bind to human STEAP2 with high affinity (e.g., a K D value below nanomolar).

[0172] The present invention also includes anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules that inhibit tumor growth in immune-deficient mice bearing human prostate cancer xenografts. (See, for example, 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 treatment situation and the specific targeting properties desired. For example, in the context of a bispecific antigen-binding molecule (e.g., a bispecific antibody) where one arm binds to CD3 and the other 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 only with 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 associated harmful side effects.

[0174] The present invention includes bispecific antigen-binding molecules (e.g., bispecific antibodies) that can bind simultaneously to human CD3 and human STEAP2. The binding arm that interacts with cells expressing CD3 can have weak to undetectable binding as measured in an appropriate in vitro binding assay. The extent to which the bispecific antigen-binding molecule binds to cells expressing CD3 and / or STEAP2 can be evaluated by fluorescence-activated cell sorting (FACS).

[0175] The present invention also includes antibodies, antigen-binding fragments thereof, and bispecific antibodies thereof that bind to STEAP2-expressing cells and cell lines (e.g., CA-2 cells) with an EC 50 value of about 1 nM to 50 nM as measured using the FACS binding assay described in Example 2 or a substantially similar assay. In certain embodiments, the antibodies, antibody-binding fragments thereof, and bispecific antibodies thereof have an EC 50 value of about 50 nM, about 40 nM, about 30 nM, about 20 nM, and an EC 50 value 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, less than about 1 nM, and bind to STEAP2-expressing cells and cell lines (e.g., CA-2 cells).

[0176] The present invention includes antibodies that bind to human CD3 with weak (i.e., low) or even undetectable affinity, antibody binding fragments thereof, and bispecific antibodies thereof. According to certain embodiments, the present invention has a K D binding to human CD3 (e.g., at 37° C.) of greater than about 11 nM as measured by surface plasmon resonance and antigen-binding fragments of the antibody. In certain embodiments, the antibody or antigen-binding fragment of the present invention binds to CD3 with a KD of 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, 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 as measured by surface plasmon resonance (e.g., mAb capture or antigen capture format) or a substantially similar assay, or binds to CD3 with no detectable affinity.

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

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

[0179] The present invention includes an anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule that can inhibit C4-2 tumor growth (see, e.g., Example 5). For example, according to certain embodiments, an anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule is provided, e.g., as described in Example 5 of this specification, which, when detected in a subject using standard caliper measurement methods at day 46 post-tumor implantation, a single administration (e.g., at a dose of about 0.1 mg / kg or about 0.01 mg / kg) results in a decrease in tumor size compared to animals administered an isotype control bispecific antibody.

[0180] The present invention also includes an anti-STEAP2 antibody-drug conjugate that inhibits tumor growth in an in vivo STEAP2-positive prostate cancer xenograft model (see, e.g., Example 7, or assays substantially similar thereto). In certain embodiments, an anti-STEAP2 antibody-drug conjugate with Compound 7 is provided, wherein 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, wherein 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, an anti-STEAP2 antibody-drug conjugate with Compound 7 is provided, wherein 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. In other embodiments, 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 after implantation inhibits C4-2 tumor growth in an in vivo STEAP2-positive prostate cancer xenograft model. Single-dose administrations as described above inhibit 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 molecule of the present invention binds may consist of a single contiguous sequence of amino acids 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) CD3 or STEAP2 proteins. Alternatively, the epitope may consist of a plurality of non-contiguous amino acids (or amino acid sequences) of CD3 or STEAP2. The antibodies of the present invention can 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 in the present invention, the term "epitope" refers to an antigenic determinant that interacts with the 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. Thus, different antibodies can bind to different regions on the antigen and can have different biological effects. The epitope may be conformational or linear. A conformational epitope is produced by amino acids that are spatially juxtaposed from different segments of a linear polypeptide chain. A linear epitope is produced by adjacent amino acid residues in a polypeptide chain. In certain situations, the epitope may include moieties of carbohydrates, phosphoryl groups, or sulfonyl groups on the antigen.

[0182] Using various techniques known to those skilled in the art, it is possible to determine whether the antigen-binding domain of an antibody "interacts with one or more amino acids" within a polypeptide or protein. Exemplary techniques include, for example, routine cross-blocking assays described in Antibodies, Harlow and Lane (Cold Spring Harbor Press, Cold Spring Harb., NY), alanine scanning mutagenesis, peptide blot analysis (Reineke, 2004, Methods Mol Biol 248:443-463), and peptide cleavage analysis. In addition, methods such as epitope excision, epitope extraction, and chemical modification of the antigen can be employed (Tomer (2000) Prot. Sci. 9:487-496). Another method that can be used to identify the amino acids within a polypeptide with which the antigen-binding domain of an antibody interacts is hydrogen / deuterium exchange detected by mass spectrometry. Generally speaking, the hydrogen / deuterium exchange method involves labeling the protein of interest with deuterium and then binding the antibody to the deuterium-labeled protein. Next, the protein / antibody complex is transferred to water to allow hydrogen-deuterium exchange to occur 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 the antigen / antibody complex may also be used for epitope mapping purposes.

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

[0184] The present invention also includes a bispecific antigen-binding molecule 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 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, wherein the first antigen-binding domain competes with any of the specific exemplary CD3-specific antigen-binding domains described herein for binding to CD3, and / or the second antigen-binding domain competes with any of the specific exemplary STEAP2-specific antigen-binding domains described herein for binding to STEAP2.

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

[0187] To determine whether an antibody or its antigen-binding domain competes with a reference antigen-binding molecule for binding, the binding methods described above are performed in two directions. In the first direction, after binding the reference antigen-binding molecule to the STEAP2 protein (or CD3 protein) under saturation conditions, the binding of the test antibody to the STEAP2 (or CD3) molecule is evaluated. In the second direction, after binding the test antibody to the STEAP2 (or CD3) molecule under saturation conditions, the binding of the reference antigen-binding molecule to the STEAP2 (or CD3) molecule is evaluated. In both directions, if only the first (saturating) antigen-binding molecule can bind to the STEAP2 (or CD3) molecule, 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, antibodies that compete for binding with a reference antigen-binding molecule do not necessarily bind to the same epitope as the reference antibody, but can sterically block the binding of the reference antibody by binding to overlapping or adjacent epitopes.

[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 arranged relative to each other and the bispecific antigen-binding molecules of the invention can be produced using conventional methods. (A discussion of exemplary bispecific antibody formats that can be used to construct the bispecific antigen-binding molecules of the invention is provided elsewhere herein). In certain embodiments, one or more individual components (e.g., heavy and light chains) of the multispecific antigen-binding molecules of the invention are derived from chimeric antibodies, humanized antibodies or fully human antibodies. Methods for making such antibodies are well known in the art. For example, one or more heavy and / or light chains of the bispecific antigen-binding molecules of the invention can be prepared using VELOCIMMUNE™ technology. Using VELOCIMMUNE™ technology (or any other human antibody production technology), high-affinity chimeric antibodies against a particular antigen (e.g., CD3 or STEAP2) are first isolated having human variable regions and murine constant regions. The antibodies are characterized and selected for desirable features including affinity, selectivity, epitope, etc. The murine 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] The genetically engineered animals can be used to produce human bispecific antigen-binding molecules. For example, genetically modified mice that are unable to rearrange and express endogenous mouse immunoglobulin light chain variable sequences can be used, and the mice express only one or two human light chain variable domains encoded by human immunoglobulin sequences operably linked to the mouse kappa constant gene of the endogenous mouse kappa locus. Such genetically modified mice can be used to produce fully human-like bispecific antigen-binding molecules that contain two different heavy chains that associate with the same light chain containing a variable domain derived from one of two different human light chain variable region gene segments. (See, for example, US2011 / 0195454). By "fully human-like" is meant an antibody, or an antigen-binding fragment or immunoglobulin domain thereof, that contains an amino acid sequence encoded by DNA derived from a human sequence over the entire length of each polypeptide of the antibody or antigen-binding fragment or immunoglobulin domain. In some instances, the fully human-like sequences are derived from proteins that are endogenous to humans. In other instances, the fully human-like protein or protein sequence contains chimeric sequences where each component sequence is derived from a human sequence. Without being bound by any theory, chimeric proteins or chimeric sequences are generally designed to minimize the generation of immunogenic epitopes at the junctions of the component sequences, for example, as compared to any wild-type human immunoglobulin region or domain.

[0190] Biological equivalents The present invention encompasses antigen-binding molecules having amino acid sequences that, although different from those of the exemplary molecules disclosed herein, retain the ability to bind to CD3 and / or STEAP2. Such variant antibodies contain one or more additions, deletions, or substitutions of amino acids as compared to the parental sequence, but exhibit a biological activity that is essentially equivalent to the biological activity of the bispecific antigen-binding molecules described. and exhibit a 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, for example, pharmaceutical equivalents or pharmaceutical alternatives that do not show a significant difference in absorption rate and extent of absorption when administered at the same molar dose, either as a single dose or multiple doses, under similar experimental conditions, and are considered to be biologically equivalent. Some antigen-binding proteins are considered to be equivalents or pharmaceutical alternatives when these degrees of absorption are equivalent but the absorption rates are not, and such differences in absorption rate can be considered to be biologically equivalent because they are intentional and reflected in the labeling, for example, not essential for achieving an effective body drug concentration for long-term use and not considered medically significant for the specific drug product tested.

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

[0193] In one embodiment, two antigen-binding proteins are biologically equivalent if the patient can be switched one or more times compared to a therapy that is sustained without switching between a reference product and a biological product without an expected increase in the risk of adverse effects, including a clinically significant change in immunogenicity or a decrease in efficacy.

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

[0195] Biological equivalence can be demonstrated by in vivo and in vitro methods. Biological equivalence measurement methods include, for example, (a) in vivo tests in humans or other mammals where the concentration of an antibody or its metabolite is measured as a function of time in blood, plasma, serum, or other biological fluids, (b) in vitro tests that reasonably predict this data and correlate with human bioavailability data, (c) in vivo tests in humans or other mammals where the appropriate acute pharmacological effect of an 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 biological equivalence of an antigen - binding protein.

[0196] Biologically equivalent variants of the exemplary bispecific antigen - binding molecules shown herein can be constructed, for example, by causing 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 unwanted or inaccurate intramolecular disulfide bridges during regeneration. 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 properties of the molecule, such as mutations that eliminate or remove glycosylation.

[0197] Species selectivity and species cross - reactivity According to certain embodiments of the 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 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, provided are 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, provided is 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 conjugate (ADC) The present invention provides an antibody-drug conjugate (ADC) comprising an anti-STEAP2 antibody or an antigen-binding fragment thereof conjugated to a therapeutic moiety such as a cytotoxic agent, a chemotherapeutic agent, an immunosuppressive agent or a radioisotope. Generally speaking, an ADC is A-[L-P] y wherein A is an antigen-binding molecule, such as 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 a therapeutic moiety (e.g., a cytotoxic agent), and y is an integer from 1 to 30. In various embodiments, the ADC comprises an anti-STEAP2 antibody or an antigen-binding fragment thereof comprising the CDRs of the HCVR and LCVR having the amino acid sequences of the SEQ ID NOs described in Table 1 (e.g., SEQ ID NOs 2 and 10), or a specific HCVR / LCVR pair (e.g., SEQ ID NOs 2 / 10). Optionally, the anti-STEAP2 antibody or fragment comprises CDRs having the amino acid sequences of the SEQ ID NOs described in Table 1 (e.g., SEQ ID NOs 4-6-8-12-14-16). Optionally, the anti-STEAP2 antibody or fragment comprises HCVR and LCVR having the amino acid sequences of the SEQ ID NOs described in Table 1 (e.g., SEQ ID NOs 2 and 10), or a specific pair of amino acid sequences (e.g., SEQ ID NOs 2 / 10).

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

[0201] According to this aspect of the invention, examples of suitable cytotoxic agents and chemotherapeutic agents that can be conjugated to the anti-STEAP2 antibody include, for example, 1-(2-chloroethyl)-1,2-dimethanesulfonylhydrazide, 1,8-dihydroxy-bicyclo[7.3.1]trideca-4,9-diene-2,6-diyne-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, calminomycin, carmustine, cedax, cisplatin, colchicine, combretastatin, cyclophosphamide, cytarabine, cytochalasin B, dactinomycin, daunorubicin, dacarbazine, diacetoxypentyl doxorubicin, dibromomannitol, dihydroxyanthracenedione, disorazole, dostarlimab, doxorubicin, duocarmycin, echinomycin, ellipticine, emetine, epothilone, esperamicin, estramustine, ethidium bromide, etoposide, fluorouracil, geldanamycin, gramicidin D, glucocorticoid, irinotecan, leptomycin, leurosine, lidocaine, lomustine (CCNU), maytansinoid, mechlorethamine, melphalan, mercaptopurine, methopterin, methotrexate, mitramycin, mitomycin, mitoxantrone, N8-acetylspermidine, podophyllotoxin, procaine, propranolol, pteridine, p It includes euromycin, risoxin, streptozotocin, talisomycin, taxol, tenoposide, tetracaine, thioepachlorambucil, tomyamycin, topotecan, tubulysin, vinblastine, vincristine, vindesine, vinorelbine, or 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 monomethyl auristatin E (MMAE) or monomethyl auristatin F (MMAF), a tubulysin such as TUB-OH or TUB-OMOM, a tomyamycin 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 the compound of formula (I).

Chemical formula

[0203] In some embodiments, A is a divalent group of optionally substituted benzene, pyridine, naphthalene, or quinoline.

[0204] In some embodiments, A is an arylene.

[0205] In some embodiments, A is

Chemical formula

Chemical formula

[0206] In some embodiments, the compound of formula I is

Chemical formula

Chemical formula

Chemical formula

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

Chemical formula

[0208] In some embodiments, the maytansinoid of formula (I) is conjugated to an anti-STEAP2 antibody or an antigen-binding fragment thereof via a linker as shown in the following formula (IA).

Chemical formula

[0209] In various embodiments, L is of the following formula

Chemical formula

[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 [Chemical formula] and wherein [Chemical formula] is a binding to an anti-STEAP2 antibody or a fragment thereof, b is an integer from 2 to 8.

[0212] In other embodiments, L is [Chemical formula] and wherein [Chemical formula] is a binding to an anti-STEAP2 antibody or a fragment thereof, b is an integer from 2 to 8.

[0213] In one embodiment, the compound of formula (IA) comprising a linker that binds to an anti-STEAP2 antibody or an antigen-binding fragment thereof has the following formula:

Chemical formula

Chemical formula

[0214] In one embodiment, the compound of formula (IA) comprising a linker that binds to an anti-STEAP2 antibody or an antigen-binding fragment thereof has the following formula:

Chemical formula

Chemical formula

[0215] In some embodiments, the cytotoxic agent is a maytansinoid having the structure of formula (II) including stereoisomers of the compound of formula (II).

Chemical formula

[0216] In some embodiments, the compound of formula (II) is

Chemical formula

[0217] In one embodiment, the compound of formula (II) is [Chemical Formula] as follows.

[0218] In some embodiments, the maytansinoid of formula (II) is attached via a linker to an anti-STEAP2 antibody or an antigen-binding fragment thereof, as shown in formula (IIA) below. [Chemical Formula] In the formula, BA is an anti-STEAP2 antibody or an antigen-binding fragment thereof, a is an integer from 1 to 30, Z2 is 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 are an amino acid, a peptide having 2 to 20 amino acids, alkyl, alkynyl, alkenyl, cycloalkyl, aryl, heteroaryl, heterocyclyl, -CR5R6-, -O-, -C(=O)-, -O-C(=O)-, -C(=O)-O-, -O-C(=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-, -S-C(=S)-, -S-C(=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)-, -O-C(=O)-N(R4), -O-C(=S)-N(R4)-, -C(=S)-N(R4)-, -N=C=S, -N=C=O,

Chem.

[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, glutamate-asparagine, asparagine-glutamate, citrulline-asparagine, asparagine-citrulline, alanine-asparagine, and asparagine-alanine.

[0220] In one embodiment, the compound of formula (II A) that binds to the anti-STEAP2 antibody or antigen-binding fragment thereof has the following formula:

Chemical formula

Chemical formula

[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,

Chemical formula

[0222] In another embodiment, the ADC comprises the "A-[L-P] y " structure, wherein A is an anti-STEAP2 antibody or antigen-binding fragment thereof, and [L-P] is

Chemical formula

Chemical formula

[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 toxins (i.e., phytolaccatoxin and phytolaccagenin), and protein toxins such as those described in Sapra et al., Pharmacol. & Therapeutics, 2013, 138:452 - 469.

[0226] The cytotoxic agent (“payload”) can be linked to the anti-STEAP2 antigen-binding molecule or antibody of the present invention via a chemical linker that covalently attaches the payload compound to a protein molecule (i.e., an antibody). Exemplary embodiments of specific linkers are described above. More generally, as used herein, the term “linker” refers to any divalent group or moiety that links, connects, or couples a binder (e.g., an antibody or an antigen-binding fragment thereof) to the payload compounds described herein. Generally, binder linkers suitable for the antibody conjugates described herein It is stable enough to utilize the circulating half-life of the antibody and at the same time can release its payload after internalization of the antigen-mediated complex. The linker can be cleavable or non-cleavable. A cleavable linker is a linker that is cleaved by intracellular metabolism following internalization, such as cleavage by hydrolysis, reduction, or enzymatic reaction. A non-cleavable linker is a linker that releases the conjugated payload via lysosomal degradation of the antibody after internalization. Suitable linkers include, but are not limited to, acid-labile linkers, hydrolysis-labile linkers, enzymatically cleavable linkers, reduction-labile linkers, self-immolative linkers, and non-cleavable linkers. Suitable linkers include, but are not limited to, peptides, glucuronides, succinimide-thioethers, polyethylene glycol (PEG) units, hydrazones, mal-caproyl units, dipeptide units, valine-citrulline units, and para-aminobenzyl (PAB) units or those containing them. In some cases, the linker can bind to the antibody or antigen-binding fragment via a lysine residue or a cysteine residue (e.g., via cleavage of the disulfide group of the antibody or fragment, or via a cysteine residue incorporated into the antibody or fragment). In some cases, the linker can bind to the antibody or fragment via a glutamine residue, including those induced via transglutaminase-mediated binding.

[0227] Exemplary linkers that can be used in the context of the present invention include, for example, 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 moiety 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 linkers comprising or consisting of 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 Ducry, 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 as 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 binding at specific amino acids within the antibody or antigen-binding molecule. Exemplary amino acid linkages that can be used in the context of this aspect of the invention include, for example, lysine (see, e.g., U.S. Patent No. 5,208,020, US2010 / 0129314, Hollander et al., Bioconjugate Chem., 2008, 19:358-361, WO2005 / 089808, U.S. Patent 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. Patent No. 7,750,116), selenocysteine (see, e.g., WO2008 / 122039, and Hofer et 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 (see, e.g., WO2013 / 068874, and WO2012 / 166559), and acidic amino acids (see, e.g., WO2012 / 05982). The linker can also be attached to the antigen-binding protein via a carbohydrate linkage (see, e.g., US2008 / 0305497, and Ryan et al., Food & Agriculture Immunol., 2001, 13:127-130) and via a disulfide linker. (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 conjugated to an antibody or antigen-binding fragment and has a significant impact on the efficacy, potency, and pharmacokinetics of the 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 from 1 to 4. In certain embodiments, the DAR is from 2 to 4. In some cases, the DAR is from 2 to 3. In some cases, the DAR is from 3 to 4. In some embodiments, the DAR is from 1 to 10, from 1 to 20, or from 1 to 30 (i.e., 1 to 30 drug molecules per antibody or its antigen-binding fragment).

[0230] Therapeutic Formulations and Administration The present invention provides a pharmaceutical composition comprising an antigen-binding molecule of the present invention. The pharmaceutical compositions of the present invention are formulated with suitable carriers, excipients, and other agents that provide improved mobility, delivery, resistance, etc. Many suitable formulations can be found in the formulary 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, lipids (cationic or anionic) containing vesicles (such as LIPOFECTIN™, Life Technologies, Carlsbad, CA, etc.), DNA complexes, anhydrous absorbent pastes, oil-in-water emulsions and water-in-oil emulsions, emulsion carbowaxes (polyethylene glycols of various molecular weights), semi-solid gels, and semi-solid mixtures containing carbowaxes. See also Powell et al. “Compendium of excipients for parenteral formulations” PDA (1998) J Pharm Sci Technol 52:238-311.

[0231] The dosage of the antigen-binding molecule administered to a patient can vary depending on the patient's age and size, the target disease, the disease state, the route of administration, etc. Preferred dosages 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, the bispecific antigen-binding molecule of the present invention is usually administered intravenously in a single dose of 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. Depending on the severity of the condition, the frequency and duration of treatment can be adjusted. The effective dosage and schedule for administering the bispecific antigen-binding molecule are determined empirically. For example, the patient's course can be monitored by regular evaluation and the dosage adjusted accordingly. Furthermore, interspecies scaling of dosages can be carried out 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, 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 composition can be administered by any convenient route, for example, by injection or bolus injection, by absorption through epithelial or mucosal linings (such as 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 standard needles and syringes. In addition, with respect to subcutaneous delivery, pen delivery devices are readily applicable in delivering the pharmaceutical composition of the present invention. Such pen delivery devices can be reusable or disposable. Reusable pen delivery devices generally use replaceable cartridges containing the pharmaceutical composition. Once all of the pharmaceutical composition in the cartridge has been administered and the cartridge is empty, this empty cartridge can be easily discarded and easily 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, disposable pen delivery devices are pre-filled with the pharmaceutical composition held in a reservoir within the device. Once the reservoir is empty with respect to the pharmaceutical composition, the entire device is discarded.

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

[0235] In certain situations, the pharmaceutical composition can be delivered in a sustained 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, a polymeric material can be used. See Medical Applications of Controlled Release, Langer and Wise (eds.), 1974, CRC Pres., Boca Raton, Florida. In yet another embodiment, the sustained release system can be placed in the vicinity of the target of the composition, whereby only a fraction of the systemic dose is required. (See, for example, Goodson, 1984, in Medical Applications of Controlled Release, supra, vol. 2, pp. 115-138). Other sustained release systems are discussed in a review by Langer, 1990, Science 249:1527-1533.

[0236] Injectable preparations can include dosage forms for intravenous injection, subcutaneous injection, intradermal injection, intramuscular injection, infusion, etc. These injectable preparations can be prepared by publicly known methods. For example, an injectable preparation can be prepared by dissolving, suspending, or emulsifying the above-described antibody or a salt thereof in a sterile aqueous medium or an oily medium conventionally used for injection. Examples of injectable aqueous media include physiological saline, glucose-containing isotonic solutions, and other adjuvants, which can be used in combination with appropriate solubilizing agents such as alcohols (e.g., ethanol), polyalcohols (e.g., propylene glycol, polyethylene glycol), nonionic surfactants [e.g., polysorbate 80, HCO-50 (polyoxyethylene (50 mol) adduct of hydrogenated castor oil)]. As the oily medium, for example, sesame oil, soybean oil, etc. are used, and they can be used in combination with solubilizing agents such as benzyl benzoate, benzyl alcohol. The injection thus prepared is preferably filled into an appropriate ampoule.

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

[0238] Therapeutic Use of Antigen-Binding Molecules The present invention includes a method comprising administering to a subject in need thereof a therapeutic composition comprising an anti-STEAP2 antibody or an 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 expression "a subject in need thereof" means a human or non-human animal (e.g., a subject expressing a tumor or a subject suffering from any of the cancers described hereinafter) showing one or more symptoms or signs of cancer, or alternatively, one who would benefit from the inhibition or reduction of STEAP2 activity or the depletion of STEAP2+ cells (e.g., prostate cancer cells).

[0239] The antibodies and bispecific antigen-binding molecules of the present invention (and therapeutic compositions containing the same) are useful, inter alia, for the treatment of any disease or disorder in which stimulation, activation and / or targeting of 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 for the treatment, prevention and / or amelioration of any disease or disorder associated with or mediated by STEAP2 expression or activity or the proliferation of STEAP2+ cells. The mechanism of action by which the therapeutic methods of the present invention are achieved includes the killing of cells expressing STEAP2 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, for example, to treat primary and / or metastatic tumors that occur in 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 for the treatment of one or more of the cancers of 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 suffering from castration-resistant prostate cancer. According to other related embodiments of the present invention, there is provided a method comprising administering to a patient suffering from castration-resistant prostate cancer an anti-STEAP2 antibody or an anti-CD3 / anti-STEAP2 bispecific antigen-binding molecule disclosed herein. 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 a method for treating residual cancer in a subject. As used herein, the term "residual cancer" means the presence or persistence of one or more cancerous cells in a subject after treatment with anti-cancer therapy.

[0242] According to certain aspects, the present invention provides a method for treating a disease or disorder associated with STEAP2 expression (e.g., prostate cancer), comprising administering to a subject one or more of the anti-STEAP2 or bispecific antigen-binding molecules described elsewhere herein after the subject has been determined to have prostate cancer (e.g., castration-resistant prostate cancer). For example, the present invention provides a method for treating prostate cancer, comprising administering to a patient an anti-STEAP2 antibody or an 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 received hormonal therapy (e.g., anti-androgen therapy).

[0243] Combination Therapies and Formulations The present invention provides a method 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 concomitantly with the antigen-binding molecules of the present invention include, for example, EGFR antagonists (e.g., anti-EGFR antibodies [e.g., cetuximab or panitumumab] or small molecule inhibitors of EGFR [e.g., gefitinib nib or erlotinib]), Her2 / ErbB2, ErbB3, or ErbB4 Antagonists of any other EGFR family members (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 trap, see e.g., US7,087,411 (also referred to herein as "VEGF inhibitory fusion protein")), 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, e.g., REGN421), Ang2 antagonists (e.g., anti-Ang2 antibodies disclosed in US2011 / 0027286, e.g., 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), uroplakin It includes antagonists (e.g., anti-uroplakin antibodies), etc. Other agents that can be beneficially administered in combination with the antigen-binding molecules of the present invention include 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, IL-18, or small molecule cytokine inhibitors and antibodies that bind to their respective receptors, including cytokine inhibitors. The pharmaceutical compositions of the present invention (e.g., pharmaceutical compositions containing the anti-CD3 / anti-STEAP2 bispecific antigen-binding molecules disclosed herein) can also be administered as part of a treatment regimen that includes one or more therapeutic combinations selected from "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 any of the antigen-binding molecules described herein, and a therapeutic combination comprising 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 aforementioned cytokines, and the inhibitors are aptamers, antisense molecules, ribozymes, siRNAs, peptibodies, nanobodies or antibody fragments (e.g., Fab fragments, F(ab')2 fragments a fragment, Fd fragment, Fv fragment, scFv, dAb fragment, or diabody, triabody, tetrabody, minibody, and minimal recognition unit). The antigen-binding molecule of the present invention can also be administered and / or co-formulated in combination with an antiviral agent, an antibiotic, an analgesic, a corticosteroid, and / or an NSAID. The antigen-binding molecule of the present invention may also be administered as part of a treatment regimen that includes radiation therapy and / or conventional chemotherapy.

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

[0246] The present invention includes a pharmaceutical composition in which the antigen-binding molecule of the present invention is co-formulated with one or more of the additional therapeutic active ingredient(s) described elsewhere herein.

[0247] Mode of Administration According to a particular embodiment of the present 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. The method according to this aspect of the present invention includes continuously administering multiple doses of the antigen-binding molecule of the present invention to a subject. As used herein, "administering continuously" means that each dose of the antigen-binding molecule is administered to the subject on different days separated by different time points, e.g., a predetermined interval (e.g., several hours, several days, several weeks, or several months). The present invention includes methods that include continuously administering 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 to a patient.

[0248] The terms "first dose", "second dose", and "third dose" refer to the chronological sequence of administration of the antigen-binding molecules of the present invention. Thus, the "first dose" is the dose administered at the start of the treatment regime (also referred to as the "baseline dose"), the "second dose" is the dose administered after the first dose, and the "third dose" is the dose administered after the second dose. The first dose, second dose, and third dose may all contain the same amount of antigen-binding molecules, but generally may differ from each other with respect to the frequency of administration. However, in certain embodiments, the amounts of antigen-binding molecules contained in the first dose, second dose and / or third dose may differ from each other (e.g., are adjusted up or down as appropriate) during the course of treatment. In certain embodiments, two or more (e.g., two, three, four, or five) doses are administered at the start of the treatment regime as a "loading dose", after which subsequent doses are administered at a lower frequency basis (e.g., a "maintenance dose").

[0249] In certain exemplary embodiments of the present invention, each second dose and / or third dose is administered 1 to 26 (e.g., 1, 1 and 1 / 2, 2, 2 and 1 / 2, 3, 3 and 1 / 2, 4, 4 and 1 / 2, 5, 5 and 1 / 2, 6, 6 and 1 / 2, 7, 7 and 1 / 2, 8, 8 and 1 / 2, 9, 9 and 1 / 2, 10, 10 and 1 / 2, 11, 11 and 1 / 2, 12, 12 and 1 / 2, 13, 13 and 1 / 2, 14, 14 and 1 / 2, 15, 15 and 1 / 2, 16, 16 and 1 / 2, 17, 17 and 1 / 2, 18, 18 and 1 / 2, 19, 19 and 1 / 2, 20, 20 and 1 / 2, 21, 21 and 1 / 2, 22, 22 and 1 / 2, 23, 23 and 1 / 2, 24, 24 and 1 / 2, 25, 25 and 1 / 2, 26, 26 and 1 / 2, or more) weeks after the immediately preceding dose. The phrase "immediately preceding dose" as used herein, in the context of a plurality of successive administrations, means the dose of the antigen-binding molecule that is administered to a patient prior to the administration of the dose that is immediately after the intervening dose(s).

[0250] The method according to this aspect of the invention may comprise administering to a patient any number of secondary and / or tertiary 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). 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 at the same frequency as the other secondary doses. For example, each secondary dose may be administered to the patient 1 to 2 weeks after the previous dose. Similarly, in embodiments comprising multiple tertiary doses, each tertiary dose may be administered at the same frequency as the other tertiary doses. For example, each tertiary dose may be administered to the patient 2 to 4 weeks after the previous dose. Alternatively, the frequency at which the secondary and / or tertiary doses are administered to the patient may vary over the course of the treatment regimen. The dosing frequency may also be adjusted by a physician during the course of treatment according to the needs of an individual patient following clinical examination.

[0252] Diagnostic use of the antibody The anti-STEAP2 antibody of the present invention may also be used to detect and / or measure STEAP2 or STEAP2-expressing cells in a sample, for example, for diagnostic purposes. For example, an anti-STEAP2 antibody, or a fragment thereof, may be used to diagnose a pathological condition or disease characterized by abnormal expression of STEAP2 (e.g., overexpression, underexpression, lack of expression, etc.). An exemplary diagnostic assay for STEAP2 may, for example, comprise contacting a sample obtained from a patient with the anti-STEAP2 antibody of the present invention, wherein 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. Detectable labels or reporter molecules can be 3 H, 14 C, 32 P, 35 radioisotopes such as S or 125 I, fluorescent moieties or chemiluminescent moieties such as fluorescein isothiocyanate or rhodamine, or enzymes such as alkaline phosphatase, β-galactosidase, horseradish peroxidase, or luciferase. Another exemplary diagnostic use of the anti-STEAP2 antibody of the present invention is 89 for non-invasive identification and tracking of tumor cells in a subject, such as 89 Zr-labeled antibodies (e.g., positron emission tomography (PET) imaging) such as Zr-desferrioxamine labeling. (See, for example, 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 sorter (FACS).

[0253] Samples that can be used in the STEAP2 diagnostic assay according to the present invention include any tissue or body fluid sample that can be obtained from a patient and contains a detectable amount of STEAP2 protein or a fragment thereof under normal or pathological conditions. Generally, the level of STEAP2 in a specific 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 to the level of STEAP2 measured in a sample obtained from an individual suspected of having a disease (e.g., a tumor containing STEAP2-expressing cells) or condition associated with STEAP2.

Example

[0254] The following examples are presented to provide those skilled in the art with a complete disclosure and description of how to make and use the methods and compositions of the invention and are not intended to limit the scope that the inventors regard as the invention. Efforts have been made to ensure the accuracy with respect to the numerical values (e.g., amounts, temperatures, etc.) used, but some experimental errors and deviations should be taken into account. Unless otherwise indicated, parts are by weight, molecular weights are average molecular weights, temperatures are in degrees Celsius, and pressures are at or near atmospheric pressure.

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

[0256] Genetically modified mice were immunized with the hSTEAP2 antigen (SEQ ID NO: 1899). After immunization, splenocytes were collected from each mouse and either (1) fused with mouse myeloma cells to maintain their viability, form hybridoma cells, and screen for STEP specificity, or (2) B cells were sorted using a human STEAP2 fragment as a sorting reagent that binds to reactive antibodies (antigen-positive B cells) and identifies them (as described in US2007 / 0280945A1).

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

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

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

Table 1

Table 2

[0260] Example 2: The anti-human STEAP2 antibody selectively binds to STEAP2-expressing cell lines via FACS. The ability of the anti-STEAP2 antibody to selectively bind to cell lines that endogenously express the six-transmembrane epithelial antigen of the prostate 2 (STEAP2) in humans was determined by FACS analysis.

[0261] Briefly, 1×10 5 cells were incubated with 10 μg / ml of anti-STEAP2 antibody, or isotype control antibody, on ice for 30 minutes in antibody dilution buffer. After washing once with antibody dilution buffer, the cells were incubated with 10 μg / ml of PE-conjugated anti-human or anti-mouse Fc secondary antibody on ice for 30 minutes. After further washing once, the samples were incubated with Cytofix (1% formaldehyde) for 20 minutes. After the last wash, the samples were filtered through a Pall 96-well filtration block and run on a Hyp ercyt (registered trademark) cytometer and analyzed with ForeCyt (trademark) (IntelliCyt, Albuquerque, NM). The mean fluorescence intensity (MFI) was expressed as the fold change over the unstained level (background). The mean fold over background was determined at antibody concentrations of 100 - 300 nM. For cell-binding EC50 determination, the mAb concentration ranged from 300 nM to 5 pM, and the EC50 value was determined from a four-parameter logistic equation on a 12-point response curve (GraphPad Prism).

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

[0263] As shown by the results in Tables 3A and 3B, several anti-STEAP2 antibodies specifically bound to high STEAP2-expressing C4-2 prostate adenocarcinoma cell lines with an EC50 in the low nM range and at levels more than 50-fold above background, as determined by FACS. Several anti-STEAP2 antibodies also bound weakly 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 invention to bind specifically and selectively to high-expressing STEAP2 cell lines.

[0264] Example 3: Anti-human STEAP2 antibodies demonstrate potent internalization and specificity for human STEAP2. The ability of the anti-STEAP2 antibodies of the invention to bind selectively 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 into PDL-coated 96-well plates. The next day, the cells were incubated on ice for 30 minutes with anti-human STEAP2 antibody (10 μg / ml), then washed twice with PBS. The cells were then incubated on ice for 30 minutes with Alexa 488-conjugated anti-hFc Fab secondary antibody, followed by two additional PBS washes. The antibody was allowed to internalize for 1 hour at 37°C or maintained at 4°C in internalization buffer (PBS + 2% FBS). The cells were fixed in 4% formaldehyde, nuclei were stained with DRAQ5 (Cell signaling), and images were acquired with ImageXpressmicroXL (Molecular Devices).

[0266] A qualitative visual assessment of the total binding intensity and the intensity of the antibody internalized into 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 by the results in Table 4, several antibodies showed strong internalization ability against the C4-2 cell line. Generally, robust internalization correlated with the highest level of total binding strength.

[0268] Next, the selected STEAP2 antibodies were tested for binding to other human (h) STEAP family members (STEAP1, STEAP3, and STEAP4). To evaluate anti-STEAP2 antibody specificity, plasmid constructs expressing hSTEAP1, hSTEAP2, hSTEAP3, or hSTEAP4 fused to green fluorescent protein (GFP) were transiently introduced into HEK293 cells by Lipofectamine 2000 based on the methodology. After 48 hours, the transiently transfected cells were stained with the anti-STEAP2 antibody and imaged as described above for the internalization assay. Wells containing GFP-positive cells bound by the anti-STEAP2 antibody were recorded as positive (+), and wells not bound by the anti-STEAP2 antibody were recorded 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, some of the anti-STEAP2 antibodies of the present invention show strong internalization ability and are specific binding agents for human STEAP2. [Table 5] [Table 6]

[0270] Example 4: Production of Bispecific Antibodies that Bind to STEAP2 and CD3 The present invention provides a bispecific antigen-binding molecule that binds to CD3 and STEAP2, and such bispecific antigen-binding molecules are also referred to herein as "anti-STEAP2 / anti-CD3 or anti-STEAP2×CD3 bispecific molecules". The anti-STEAP2 portion of the anti-STEAP2 / anti-CD3 bispecific molecule is useful for targeting tumor cells that express prostate six-transmembrane epithelial antigen 2 (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 promotes the direct killing of target tumor cells by activated T cells (cell lysis).

[0271] A bispecific antibody comprising an anti-STEAP2 specific binding domain and an anti-CD3 specific binding domain was recombinantly constructed by standard molecular cloning methodologies and expressed in CHO cells, wherein the anti-STEAP2 specific binding domain and the anti-CD3 specific binding domain each comprise a different and separate HCVR paired with a common LCVR. In the exemplified bispecific antibody, the molecule was constructed using the heavy chain derived from an anti-CD3 antibody, the heavy chain derived from an anti-STEAP2 antibody, and the common light chain derived from an anti-STEAP2 antibody, and expressed in CHO cells. Optionally, the bispecific antibody may be constructed using the heavy chain derived from an anti-CD3 antibody, the heavy chain derived from an anti-STEAP2 antibody, and the light chain derived from an anti-CD3 antibody, or may be indiscriminate, or may be constructed using antibody light chains known to effectively pair with various heavy chain arms such as Vκ1-39JK5 or Vκ3-20JK1.

[0272] The bispecific antibodies described in the following examples consist of anti-CD3 binding arms with various binding affinities for human soluble heterodimer hCD3ε / δ tan protein (described in Example 12 herein) and human STEAP2 (see Examples 1-2 above). An exemplary bispecific antibody having a modified (chimeric) IgG4 Fc domain, as described in US Patent Application Publication US20140243504A1, published August 28, 2014, was produced.

[0273] Table 6 shows a summary of the components of the antigen-binding domains of various anti-STEAP2×CD3 bispecific antibodies constructed.

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 show the amino acid identifiers and nucleic acid sequence identifiers for the various heavy chain variable regions of the anti-STEAP2 arm (i.e., HCVR and LCVR are derived from H2M11162N) for constructing the bispecific antibodies of this example, and their corresponding CDRs. Tables 15 and 16 show the amino acid identifiers and nucleic acid sequence identifiers for the various heavy chain variable regions of the anti-CD3 arm of the bispecific antibodies of this example, and their corresponding CDRs.

[0275] Example 5: Anti-STEAP2 / anti-CD3 bispecific antibody shows a potent antitumor effect in vivo. To determine the efficacy of an exemplary anti-STEAP2 / anti-CD3 bispecific antibody in vivo, tests were conducted in immunodeficient mice bearing prostate cancer xenografts.

[0276] Efficacy of anti-STEAP2 / anti-CD3 bispecific antibody in human tumor xenograft model To evaluate the in vivo efficacy of anti-STEAP2 / anti-CD3 bispecificity in human tumor xenograft tests, NODscid gamma (NSG) mice (Jackson Laboratories, Bar Harbor, Maine) were co-transplanted with human prostate cancer C4-2 cells (MD Anderson Cancer Ce nter, Houston TX) that endogenously express STEAP2, together with human peripheral blood mononuclear cells (PBMC, ReachBio LLC., Seattle, WA).

[0277] Briefly, 5.0×10 6Individual C4-2 cells were co-transplanted subcutaneously (s.c.) into the right flank of male NSG mice with 1.25×10 6 human PBMCs. Mice were treated intraperitoneally (i.p.) with anti-STEAP2 / anti-CD3 bispecific BSSTEAP2 / CD3-001, BSSTEAP2 / CD3-002, or BSSTEAP2 / CD3-003, or 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] Tumor size was measured twice a week using calipers, and tumor volume was calculated as volume = (length × width 2 ) / 2. Data are shown as tumor size (mm 3 ) at the study endpoint, 46 days after tumor implantation (Table 7).

Table 8

[0279] As shown by the results in Table 7, when tumor size was measured at the study endpoint, BSSTEAP2 / CD3-001, BSSTEAP2 / CD3-002, and BSSTEAP2 / CD3-003 significantly inhibited 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 Conjugates All monoclonal antibodies were expressed in CHO cells and purified by Protein A. The isotype control was also prepared in the same manner. The unbound isotype control antibody was derived from an immunological antigen unrelated to oncology.

[0281] Antibodies (10 mg / ml) in 50 mM HEPES, 150 mM NaCl, pH 7.5 were treated with 1 mM dithiothreitol at 37 °C for 30 minutes. After gel filtration (G-25, sodium acetate pH 4.5), 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 the method for making the compound, are described in PCT Publication No. WO2014 / 145090, published September 18, 2014, which is incorporated herein by reference in its entirety. After 1 hour, the reaction was quenched with excess N-ethylmaleimide. The complex was purified by size exclusion chromatography and sterile filtered. Protein and linker payload concentrations were determined by UV spectral analysis. Size exclusion HPLC demonstrated that all complexes used were >95% monomeric and RP-HPLC demonstrated that there was <0.5% unbound linker payload. Yields are reported in Table 8 based on protein titer determination. All conjugated antibodies were analyzed by UV for linker payload loading values according to Hamblett et al, Cancer Res 2004 10 7063. The results are summarized in Table 8.

[0282] Complexes containing compound 60 can be prepared using a similar method. Compound 60, and the method for making the compound, are described in PCT Publication No. WO2016 / 160615 (Example 20), published October 6, 2016, 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 conjugate (ADC) is a potent inhibitor of tumor growth in an in vivo STEAP2-positive prostate cancer xenograft model. To determine the in vivo efficacy of the anti-STEAP2 antibody conjugated to Compound 7, studies were conducted in immunodeficient mice bearing STEAP2-positive prostate cancer xenografts.

[0284] For these studies, male SCID mice (Taconic, Hudson NY) were transplanted with C4-2 cells that endogenously express STEAP2. When the tumors reached an average volume of 200 - 250 mm 3 (around day 13 - 17), the mice were randomized into treatment groups and administered either the anti-STEAP2 binding antibody, non-binding conjugated antibody, or vehicle. In these in vivo studies, the antibody was administered once, and then the tumors were monitored until an average tumor size of approximately 1500 - 2000 mm 3 was achieved (around day 40 - 50) in the cohort that was then administered vehicle alone. Treatment groups showing efficacy were maintained for a longer period (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. On day 13 post-transplantation, mice were administered a single dose of 10, 20, or 40 mg / kg of anti-STEAP2 and control ADC. As summarized in Figure 1, H1H7841N-7 (DAR 2.92) potently inhibited tumor growth at all doses tested. At the highest dose (40 mg / kg), H1H784N-7 efficiently reduced tumor size, but the non-binding control antibody (40 mg / kg) also showed an effect on tumor volume. Across all doses examined, 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 on day 14 post-transplantation, and the control antibody was administered at 20 mg / kg. As summarized in Figure 2, H1H7841N-7 (DAR 2.92) potently inhibited tumor growth at a dose of 20 mg / kg as in the previous experiment, but showed a decrease in efficacy at a dose of 5 mg / kg. The control antibody at the 20 mg / kg dose showed no difference compared to the vehicle control.

[0287] In further tests, H1H7841N-7 (DAR2.7) and control antibodies were administered at drug equivalent μg / kg based on the ADC drug:antibody ratio (“DAR”). The dose was 150 μg / kg on day 17 post-implantation (Figure 3). H1H7841N-7 strongly inhibited tumor growth at a dose of 150 μg / kg and showed tumor regression by day 42 post-implantation and 25 days post-injection. At this point, tumor growth began to rebound. Tumor growth with the control antibody at this dose was no different from the vehicle control.

[0288] B. In a similar test, 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 were administered a single dose of anti-STEAP2 ADC, isotype control ADC (binding to an irrelevant antigen), or vehicle (PBS) at 2.5 mg / kg. Tumor volume and body weight were recorded on day 0 (injection day) post-injection, as well as on days 4, 6, 8, 12, 14, and 20 post-injection. As summarized in Figure 4, H1H7814N-60 (DAR3.6) strongly inhibited tumor growth at the tested doses and showed tumor regression by day 20 post-injection (day 49 post-implantation). The rate of change in body weight for the test ADCs was ≤ -2.01% by day 14 compared to mice treated with control Ab-ADCs where the rate of change in body weight was observed to be from -4.02% to -11.55% by day 14 (after administration of H1H7814N-60).

[0289] Example 8: Production of Anti-CD3 Antibodies The anti-CD3 antibodies were obtained by immunizing transgenic mice containing DNA encoding the variable regions of the human immunoglobulin heavy and kappa light chains with cells expressing CD3 or DNA encoding CD3. The immune response of the antibodies was monitored by a CD3-specific immunoassay. When the desired immune response was achieved, spleen cells 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 having human variable domains and mouse constant domains) were obtained. Further, as described in US2007 / 0280945A1, several fully human anti-CD3 antibodies were isolated directly from antigen-positive B cells without fusing them to myeloma cells.

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

[0291] Example 9: Amino Acid Sequences and Nucleic Acid Sequences of the Heavy Chain Variable Region and the Light Chain Variable Region Table 9 shows the amino acid sequence identifiers of the heavy chain variable region, the light chain variable region, and the CDRs of selected anti-CD3 antibodies of the present invention. The corresponding nucleic acid sequence identifiers are shown in Table 10. The method for producing the anti-CD3 antibodies disclosed herein can also be found in US Patent Publication No. 2014 / 0088295, published on March 27, 2014. It can be found in US 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 suffix of "P", "N", or "B". Thus, according to this nomenclature, an antibody may be referred to herein, for example, as "H1H2712N", "H1M2692N", "H2M2689N", etc. The H1H, H1M, and H2M prefixes of the antibody names used herein indicate the specific Fc region isotype of the antibody. For example, the "H1H" antibody has a human IgG1 Fc, the "H1M" antibody has a mouse IgG1 Fc, and the "H2M" antibody has a mouse IgG2 Fc (all variable regions are fully human type as indicated by the first "H" in the antibody name). As understood by those skilled in the art, an antibody having a specific Fc isotype can be converted to an antibody having a different Fc isotype (e.g., an antibody having a mouse IgG1 Fc can be converted to an antibody having a human IgG4, etc.), but in any event, the variable domains (including the CDRs) indicated by the numerical identifiers shown in Table 1 will remain the same, and the binding properties are expected to be identical or substantially similar regardless of the nature of the Fc domain.

[0293] Tables 11 and 12 show the amino acid sequence identifiers of additional anti-CD3 HCVRs and LCVRs useful for the anti-STEAP2×CD3 bispecific antibodies of the present invention (Table 13) and light chain variable regions (Table 14), as well as their corresponding CDRs.

Table 16

Table 17

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

[0295] Furthermore, Tables 13 and 14 show the sequence identifiers of nucleotide sequences encoding additional anti-CD3 HCVR and LCVR heavy chain variable regions (Table 13) and light chain variable regions (Table 14) useful for the anti-STEAP2×anti-CD3 bispecific antibodies of the present 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. Namely, the mouse monoclonal antibody "OKT-3" against human T cell surface antigen available from the American Type Culture Collection (ATCC) under catalog number CRL-8001, and, for example, 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 was aimed at identifying antibodies that specifically recognize CD3 (T cell receptor) as an antigen.

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

[0299] The antibodies produced were tested for affinity to human and cynomolgus CD3 antigens in an in vitro binding assay, and among others, one CD3 antibody designated CD3-VH-P (HCVR described in SEQ ID NO: 1882) was identified and had an EC 50 affinity of 1-40 nM and was found to bind to both human and cynomolgus CD3, as determined by FACS titration of Jurkat cells and cynomolgus T cells. See, for example, the FACS binding experiments outlined in Example 12 and PCT / US2016 / 044732, filed July 29, 2016.

[0300] Subsequently, the germline amino acid residues of CD3-VH-P were 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 substitute amino acid residues in a stepwise manner based on the differences between the germline sequence and the CD3-VH-P sequence. Each antibody derivative was given a "CD3-VH-G" designation. See Table 15.

[0301] As seen in the FACS assay, CD3-VH-G and several other engineered antibodies retained their binding affinity, but some anti-CD3 antibodies in the bispecific format bound to human or cynomolgus CD3 in vitro with weak to unmeasurable binding affinities, such as an EC50 greater than 100 nM. Subsequently, bispecific antibodies containing exemplary anti-CD3 antibodies were further investigated for binding affinity, binding kinetics, and other biological properties, such as toxicity and pharmacokinetic (pK) profiles, and 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 of the heavy chain variable regions and CDRs of the selected anti-CD3 antibodies of the present 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. Each antibody heavy chain derived from the germline sequence (SEQ ID NO: 1910) was assigned a "G" number designation for consistent nomenclature. Table 15 shows the amino acid sequence identifiers of the heavy chain variable regions and CDRs of the engineered anti-CD3 antibodies of the present invention. The corresponding nucleic acid sequence identifiers are shown in Table 16. The amino acid identifiers and nucleic acid sequence identifiers of the light chain variable regions and CDRs are also identified in Tables 17 and 18 below, respectively. [Table 20] [Table 21] [Table 22] [Table 23]

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

[0305] The isotype control antibodies referred to in the examples herein are isotype-matched (modified IgG4) antibodies that interact with an irrelevant antigen, i.e., the FelD1 antigen.

[0306] Example 12: In vitro and in vivo tests on human monoclonal anti-CD3 antibodies In vivo and in vitro tests on human monoclonal anti-CD3 antibodies were conducted as described in US Patent Publication No. 2014 / 0088295, published on March 27, 2014, and PCT / US2016 / 044732, filed on July 29, 2016.

[0307] Some of the human monoclonal anti-CD3 antibodies of the present invention bind to soluble heterodimeric CD3 protein with high affinity in either an antibody capture format or an antigen capture format. The soluble heterodimeric CD3 protein (hCD3-epsilon / hCD3-delta; SEQ ID NO: 1900 / 1901) was prepared using either a human Fc tag (hFcΔAdp / hFc; SEQ ID NO: 1931 / 1932) or a mouse Fc tag (mFcΔAdp / mFc; SEQ ID NO: 1933 / 1934). The heterodimeric CD3 protein was purified using the method described by Davis et al. (US2010 / 0331527).

[0308] Some of the human monoclonal anti-CD3 antibodies of the present invention bound to human T cells and induced T cell proliferation. Some of the human monoclonal anti-CD3 antibodies of the present invention bound to CD2+CD4+ rhesus T cells and induced their proliferation. Some human monoclonal anti-CD3 antibodies supported redirected T cell-mediated killing via Fc / FcR interaction in a calcein-based U937 killing assay. The observed killing was thought to depend on the Fc binding of the antibody to the Fc receptor on U937 cells, which results in clustering of CD3 on adjacent T cells, and was abrogated by the addition of non-specific human IgG (data not shown).

[0309] Example 13: In Vitro Tests on Human STEAP2xCD3 Bispecific Antibody FACS Binding Titration to Jurkat Cells, PC3_STEAP2 / 1 Cells, and Rhesus T Cells: After determining the binding of the STEAP2xCD3 bispecific antibody to Jurkat T cells, PC3_STEAP2 / 1 chimeric T cells, and rhesus T cells using flow cytometry analysis, detection was performed with a phycoerythrin (PE)-labeled anti-human (IgG) antibody. Briefly, 2×105 cells / well were incubated at 4°C for 30 minutes using serial dilutions of the STEAP2×CD3 bispecific antibody or control antibody (human IgG1 antibody that binds to a feline antigen with no cross-reactivity to STEAP2 or human or rhesus CD3) in the range of 66.6 nM to 0.001 nM. After incubation, the cells were washed twice with cold PBS containing 1% filtered FBS, and the PE-conjugated anti-human secondary antibody was added to the cells and incubated for an additional 30 minutes. Wells containing only the antibody or secondary antibody were used as controls. After incubation, the cells were washed and resuspended in 200 μL of cold PBS containing 1% filtered FBS and analyzed by flow cytometry using a BD FACS Canto II.

Table 24

[0310] Jurkat cells are derived from a T cell lymphoblast cell line that expresses human CD3. All bispecific antibodies tested (Table 19 and Figure 5) bound to Jurkat cells with an EC50 in the range of 1.41E-08 M to 6.15E-10 M. The human prostate cancer cell line, PC3 cells, were engineered to express the STEAP2 / 1 chimeric construct. Some bispecific antibodies bound to PC3_STEAP2 / 1 cells with an EC50 in the range of 7.91E-08 M to 3.44E-09 M (Table 19 and Figure 6).

[0311] Binding of the STEAP2×CD3 bispecific antibody to the surface of purified cynomolgus T cells was also tested. Some of the bispecific antibodies bound with an EC50 in the range of 1.73E-08 M to 7.27E-09 M. The control antibody did not bind to any of the cell lines. See Table 19 and FIGS. 7 and 8.

[0312] T cell proliferation assay: Thawed human or freshly isolated monkey PBMCs (50,000 cells / well) were incubated for 72 hours at 37° C. in white flat-bottom 96-well plates with serial dilutions (human, concentration range: 5E-10 M to 2.82E-15 M; cynomolgus monkey, concentration range: 1E-09 M to 4.57E-13 M) of the STEAP2×CD3 bispecific or isotype control and a fixed concentration (human: 200 ng / mL, cynomolgus monkey: 500 ng / mL) of a commercially available anti-CD28 antibody (Biolegend, catalog number 302914) in complete medium (RPMI supplemented with 10% FBS, 100 U / mL of L-penicillin, 100 μg / mL of streptomycin, 292 μg / mL of L-glutamine). The isolated monkey PBMCs were from two donors (identified as mk8781M or mk9381M). After incubation, CellTiter Glo® (Promega, catalog number 7573) was added and luminescence as a readout of cell viability was measured using a VICTOR X5 multilabel plate reader. The cell titer was calculated by dividing the luminescence of the stimulated cells by the baseline luminescence of the unstimulated cells.

[0313] All αSTEAP2×αCD3 bispecific antibodies induced human PBMC proliferation in the presence of the costimulatory anti-CD28 antibody (see Table 20 and FIG. 9). PBMCs were incubated for 72 hours with serial dilutions of the bispecific antibody or control antibody and a fixed concentration of anti-CD28, and cell viability was measured in a luminescence assay to detect live cells. Proliferation was measured by comparing the luminescence of the bispecific antibody-stimulated cells to that of the cells without antibody. The EC 50 value (defined as the concentration of antibody required to produce half-maximal proliferation) was 3.68E-1 It was in the range of 3M to 1.60E-10M. In contrast, the control antibody did not show activity under the same conditions.

Table 25

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

[0315] The BSSTEAP2 / CD3-007 and BSSTEAP2 / CD3-008 activities were donor-dependent. In contrast, BSSTEAP2 / CD3-005, BSSTEAP2 / CD3-009, and the isotype control did not show activity.

[0316] Cytotoxicity assay targeting C4-2 cells in the presence of anti-STEAP2×CD3 bispecific antibody and human T cells: To monitor the specific killing of STEAP2-bearing target cells by flow cytometry, C4-2 cells were labeled with 1 μM fluorescent tracking dye Violet Cell Tracker (Life Technologies kit, #C34557). After labeling, the cells were seeded at 37°C overnight. Separately, human PBMCs were seeded at 1×10 6Cells were seeded in supplemented RPMI medium at cells / mL and incubated overnight at 37°C to deplete adherent macrophages, dendritic cells, and some monocytes to enrich lymphocytes. The next day, target cells were incubated with non-adherent cell-depleted naive PBMCs (effector / target cell ratio of 4:1) and serially diluted STEAP2×CD3 bispecific antibody or IgG1 control antibody (non-binding to STEAP2) (concentration range: 66.7 nM to 0.25 ρM) for 48 hours at 37°C. Cells were removed from the cell culture plates using enzyme-free cell 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 count beads were added to each well. 1×10 5 beads were collected for each sample. To evaluate the specificity of killing, cells were gated on the live violet-labeled population. The percentage of the live population was recorded and used for calculation of the normalized survival rate.

[0317] T cell activation was evaluated by incubating cells with antibodies directly conjugated to CD2 and CD69 and reporting the percentage of activated (CD69+) T cells among total T cells (CD2+). Several anti-STEAP2×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 antibodies tested activated and induced human T cells to deplete C4-2 cells (a human prostate adenocarcinoma subline derived from LnCap cells). Killing of target cells was observed only in the presence of the bispecific antibody, and C4-2 cells were depleted in a dose-dependent manner with a pM EC50. Furthermore, the observed target cell lysis was associated with upregulation of CD69 cells on CD2+ T cells with a pM EC50 (see Table 21 and Figure 12).

Table 26

[0318] The present invention should not be limited in scope by the specific embodiments described herein. In fact, various modifications of the present 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. 1. An isolated antibody or antigen-binding fragment thereof that binds to six transmembrane epithelial antigen 2 (STEAP2) of the human prostate, comprising: The antibody or antigen-binding fragment comprises a heavy chain variable region (HCVR) that comprises three complementarity determining regions, HCDR1, HCDR2, and HCDR3, and a light chain variable region (HCVR) that comprises three light chain complementarity determining regions, LCDR1, LCDR2, and LCDR3; wherein HCDR1, HCDR2, HCDR3, LCDR1, LCDR2 and LCDR3 are, respectively, SEQ ID NOs: 4, 6, 8, 12, 14 and 16; 20, 22, 24, 28, 30 and 32; 36, 38, 40, 44, 46 and 48; 52, 54, 56, 60, 62 and 64; 92, 94, 96, 60, 62 and 64; 100 268, 270, 272, 276, 278, and 280; and 284, 286, 288, 292, 294, and 296; and The antibody or antigen-binding fragment thereof is internalized by human STEAP2-expressing cells. The above isolated antibody or antigen-binding fragment thereof.

2. 2. The isolated antibody or antigen-binding fragment thereof of claim 1, which is fully human.

3. The HCVR and the LCVR comprise an amino acid sequence pair selected from the group consisting of SEQ ID NOs: 2 / 10, 18 / 26, 34 / 42, 50 / 58, 90 / 58, 98 / 58, 186 / 194, 250 / 258, 266 / 274, and 282 / 290.

3. An isolated antibody or antigen-binding fragment thereof according to claim 1 or 2.

4. A pharmaceutical composition comprising the isolated antibody or antigen-binding fragment thereof of any one of claims 1 to 3, and a pharma- ceutically acceptable carrier or diluent.

5. An antibody-drug conjugate (ADC) comprising an isolated antibody or antigen-binding fragment thereof described in any one of claims 1 to 3 and a cytotoxic agent, wherein the antibody or antigen-binding fragment and the cytotoxic agent are covalently bonded via a linker.

6. A pharmaceutical composition comprising the ADC described in claim 5 and a pharma- ceutically acceptable carrier or diluent.

7. 10. The pharmaceutical composition of claim 4 or 6 for use in treating a cancer that expresses STEAP2 in a subject.

8. The pharmaceutical composition described in claim 7, wherein the cancer expressing STEAP2 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.

9. The pharmaceutical composition described in claim 8, wherein the cancer expressing STEAP2 is prostate cancer.

10. The pharmaceutical composition described in claim 9, wherein the prostate cancer is castration-resistant prostate cancer.

Images