Binders for chondroitin sulfate proteoglycan (CSPG4) polypeptides

Abstract

Provided herein are methods and materials involving binding of a binding agent (e.g., an antibody, an antigen-binding fragment, an antibody domain, a CAR, a cell engager, and / or an ADC) to a CSPG4 polypeptide. For example, provided are binding agents (e.g., an antibody, an antigen-binding fragment, an antibody domain, a CAR, a cell engager, and / or an ADC) that bind to a CSPG4 polypeptide, as well as methods and materials for using one or more such binding molecules to treat a mammal (e.g., a human) with cancer.

Description

[Technical field]

[0001] (CROSS REFERENCE TO RELATED APPLICATIONS) This application claims the benefit of U.S. Patent Application No. 63 / 286,719, filed December 7, 2021. The disclosure of the prior application is considered part of (and is incorporated by reference into) the disclosure of this application.

[0002] (STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT) This invention was made with Government support under CA111412 and CA197292 awarded by the National Institutes of Health. The Government has certain rights in this invention.

[0003] (Sequence Listing) This application contains a Sequence Listing that has been submitted electronically as an XML file entitled "09531-0491WO1_SL.XML." This XML file, created on November 17, 2022, is 41000 bytes in size. The contents of the XML file are incorporated herein by reference in their entirety. [Background technology]

[0004] 1.Technical Field The present specification relates to methods and materials involving binding of molecules (e.g., antibodies, antibody fragments, antibody domains, chimeric antigen receptors (CARs), cell engagers, or antibody drug conjugates (ADCs)) to chondroitin sulfate proteoglycan 4 (CSPG4) polypeptides. For example, the present specification provides binding agents (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, cell engagers, or ADCs) that bind to CSPG4 polypeptides, as well as methods and materials for using such binding agents to treat cancer. The present specification also provides cells (e.g., host cells) engineered to express one or more binding agents (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, or cell engagers) capable of binding to CSPG4 polypeptides, as well as methods and materials for using such cells to treat cancer.

[0005] 2. Background information Epithelial ovarian cancer (EOC) is a highly heterogeneous disease that includes a wide range of different molecular subtypes and clinical entities. There is a complex basis for inter- and intra-patient genetic heterogeneity in EOC reflected by different gene signatures associated with different tissue subtypes or genetic / epigenetic changes induced by external stressors such as chemotherapy. See, e.g., Moffitt et al., Int. J Mol Sci., 20(6):1466 (2019). Although many EOC patients initially respond well to surgical removal and adjuvant chemotherapy, the development of chemoresistance is a major obstacle, with 75% of patients experiencing relapse within 5 years. See, e.g., Kroeger et al., Curr Opin Obstet Gynecol., 29(1):26-34 (2017); Lengyel, Am J Pathol., 177(3):1053-64 (2010). Malignant progression also involves extensive intratumor phenotypic heterogeneity associated with dynamic biological demands at various stages of progression. See Lengyel, 2010, supra; Testa et al., Medicines (Basel), 5(1):16 (2018); and Moffitt, 2019, supra. These dynamics include localized changes in growth factors and actively remodeling tumor-associated extracellular matrix that directly contribute to the epigenetic changes associated with malignant progression. Additional factors include the presence of therapy-resistant cancer stem cells, survival of cells within spatially distinct fibrotic or hypoxic microenvironments, and the spread of mutational variants with increased invasive and / or metastatic potential. See, e.g., Habyan et al., Oncogene, 37(37):5127-35 (2018); Meads et al., Nat Rev Cancer, 9(9):665-74 (2009); and Paullin et al., PLoS One, 12(8):e0182930 (2017). The complex and dynamic mechanisms influencing this extensive intratumor phenotypic heterogeneity have hindered the identification of effective prognostic and predictive biomarkers that can be effectively targeted in patients with EOC.

[0006] Ovarian cancer metastasis occurs primarily by the intraperitoneal (IP) route, thus differing from other common carcinomas such as breast and prostate cancer, which primarily utilize the vasculature or lymphatics. See, e.g., Kroeger et al., 2017, supra; Lengyel, 2010, supra; and Habyan et al., 2018, supra. In EOC, individual cells or cell aggregates dissociate from the primary tumor to form multicellular spheroids that are responsible for peritoneal spread, metastasis, and recurrence. See, e.g., Shield et al., Gynecol Oncol., 113(1):143-8 (2009). The survival of individual cells that give rise to spheroids is facilitated by their anchorage-independence and initial resistance to anoikis. Multiple cell adhesion-related pathways (e.g., integrins, cadherins, and claudins) contribute to the strengthening and compaction of these multicellular aggregates and their attachment to mesothelial cells that line multiple organ sites (e.g., omentum) within the pelvis and peritoneum. Increased compaction of cells within spheroids may result in increased therapeutic resistance, in part, by limiting the penetration of chemotherapeutic agents to the more centrally located cells within these spheroids. See, e.g., Habyan et al., 2018, supra; and Shield, 2009, supra. Their subsequent invasion into the submesothelial tissue involves stimulation by growth factors and chemokines within the microenvironment and activation of tumor-associated matrix metalloproteinases that degrade the underlying extracellular matrix.

[0007] Similar to other cancers, malignant progression in EOC tumor cells is associated with a phenotypic shift from epithelial to mesenchymal phenotype (EMT). The EMT program is influenced by multiple complex mechanisms, including changes in the expression / function of multiple signaling pathways (e.g., multiple growth factors, Wnt / β-catenin, and Notch) and multiple adhesion receptors (E-cadherin / N-cadherin, claudins, and integrins). See, e.g., Deng et al., Oncotarget, 7(34):55771-88 (2016); and Yang et al., Nat Rev Mol Cell Biol., 21(6):341-52 (2020). Detachment of tumor cells from the primary tumor and subsequent spheroid formation is associated with increased expression of certain mesenchymal transcription factors, such as ZEB1 and Slug (Snail2). Mesenchymal transition in cancer is often associated with cancer cell "stemness," resistance to apoptosis, and therapy. However, it has been recently emphasized that the pathways involved in regulating EMT are complex and diverse, highlighting the need for caution in linking stemness phenotypes in tumor cells (increased cell survival and drug resistance) to canonical pathways classically associated with EMT (see, e.g., Yang et al., 2020, supra). Summary of the Invention

[0008] The present specification provides methods and materials involving binding of molecules (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, cell engagers, or ADCs) to CSPG4 polypeptides. CSPG4 is a tumor cell surface cancer antigen. As described herein, CSPG4 is an independent risk factor for reduced survival in patients with EOC and can be used, for example, as a diagnostic biomarker in EOC. Furthermore, target cells expressing CSPGs can be used, for example, to limit recurrence and improve outcomes in patients with EOC or other CSPG4+ cancers. For example, as shown herein, CSPG4 promotes resistance to chemotherapy (e.g., cisplatin resistance), promotes tumor invasion and mesenchymal transition, and promotes the formation of multicellular aggregates (spheroids) of tumor cells. These spheroids are involved in the development of peritoneal metastases, which are an important source of recurrence in ovarian cancer patients. Using CRISPR / Cas9 deletion of CSPG4 in multiple ovarian cancer cell lines, it has been demonstrated that CSPG4 functions to promote several distinct phenotypic characteristics associated with malignant progression. These properties include increased tumor invasion, cisplatin resistance and enhanced formation of multicellular spheroids in vitro. CSPG4 also increased the expression of multiple mesenchymal markers and promoted increased proliferation in vivo in xenograft mouse models. Using immunohistochemistry (IHC) analysis of well-defined EOC patient cohorts and mRNA expression levels derived from publicly available patient cohorts including The Cancer Genome Atlas (TCGA), elevated levels of CSPG4 were determined to be associated with poor overall survival in multiple tab types of EOC. These results functionally link CSPG4 expression to multiple distinct cancer-associated phenotypes in EOC cells and associate elevated CSPG4 expression with poor outcomes in EOC patients. Furthermore, monoclonal anti-CSPG4 antibodies inhibit CSPG4-stimulated ZEB1 expression, tumor cell invasion and promote apoptosis of EOC cells in spheroids. The results described herein indicate that CSPG4 is a target for limiting recurrence and improving patient outcomes.

[0009] In some embodiments, the description provides binding agents (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, cell engagers, or ADCs) that bind to CSPG4 polypeptides, as well as methods and materials for using one or more such binding agents to treat a mammal (e.g., a human) with cancer.

[0010] The present specification also provides cells (e.g., host cells) engineered to express one or more binding agents (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, or cell engagers) capable of binding to a CSPG4 polypeptide, as well as methods and materials for using such cells to treat cancer.

[0011] As described herein, a binding agent (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more CARs, one or more cell engagers, and / or one or more ADCs) can be designed to have the ability to bind to a CSPG4 polypeptide. For example, a binding agent (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, or ADC) provided herein can be capable of binding to a polypeptide comprising, consisting essentially of, or consisting of the amino acid sequence of a human CSPG4 polypeptide set forth in SEQ ID NO:33 (see, e.g., FIG. 1).

[0012] In some cases, two sets of three CDRs of the antigen-binding fragments provided herein (e.g., SEQ ID NOs: 1-3 and 9-11 or SEQ ID NOs: 17-19 and 25-27) are selected from the set of three CDRs of CSPG4. + Cells (e.g., CSPG4 + CAR with the ability to target tumor cells + Cells (e.g., CAR + T cells, CAR + Stem cells, e.g. CAR + Induced pluripotent stem cells, or CARs + Natural killer (NK) cells) may be engineered into the CAR to generate CSPG4 + Cells (e.g., CSPG4 +The present invention creates antibodies capable of targeting tumor cells (CSPG4) + It may be engineered into an antibody structure that contains an Fc region to induce antibody-dependent cell-mediated cytotoxicity (ADCC) against cells, and / or CSPG4. + Cells (e.g., CSPG4 + tumor cells) and target CSPG4 + Fc-containing antibodies may be engineered into cell engagers, such as bispecific T cell engagers (e.g., BiTEs), bispecific killer engagers (e.g., BiKEs) and / or trispecific killer engagers (e.g., TriKEs), to create cell engagers capable of inducing one or more immune responses against the cells (e.g., a T cell immune response and / or ADCC using the cell engager in the absence of an Fc-containing antibody). Note that BiKE and TriKE mediated killing are sometimes referred to as ADCC, even though they are not initiated by an Fc domain.

[0013] In addition, as described herein, the binding agents provided herein (e.g., one or more antibodies, one or more antigen-binding fragments, and / or one or more antibody domains) can be used to create a conjugate comprising a binding agent and a drug. For example, an ADC, such as a whole antibody drug conjugate, a Fab drug conjugate, and / or an antibody domain drug conjugate, can be designed to include an appropriate binding agent provided herein to create a conjugate. Such a conjugate can be used to bind to a target cell, such as a cancer cell (e.g., a CSPG4 + They can be used to deliver drug payloads to tumor cells (cancer cells).

[0014] Also as described herein, the binding agents provided herein (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) can be used to treat a mammal (e.g., a human) with cancer. For example, cancer (e.g., CSPG4 +A mammal (e.g., a human) having a cancer may be administered a composition comprising one or more binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) described herein to reduce the number of cancer cells in the mammal, to induce ADCC against cancer cells in the mammal, and / or to prolong survival of the mammal from cancer. The binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) may also be used to reduce tumor cell invasion, limit mesenchymal transition, and / or inhibit spheroid formation.

[0015] Also as described herein, a cell (e.g., a host cell) can be engineered to express one or more binding agents (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, or cell engagers) capable of binding to a CSPG4 polypeptide. For example, a cell, such as a T cell (e.g., a CTL), a stem cell (e.g., an induced pluripotent stem cell), or a NK cell, can be engineered to express one or more CARs capable of binding to a CSPG4 polypeptide. Such cells (e.g., a CSPG4-specific CAR + T cells or NK cells) can be used to treat cancer.

[0016] In one aspect, the present disclosure provides a method for the production of a light chain variable domain or region comprising: (i) a heavy chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO:1 (or SEQ ID NO:1 with the addition, deletion or substitution of one, two or three amino acids), SEQ ID NO:2 (or SEQ ID NO:2 with the addition, deletion or substitution of one, two or three amino acids), and SEQ ID NO:3 (or SEQ ID NO:3 with the addition, deletion or substitution of one amino acid); and (ii) a light chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO:9 (or SEQ ID NO:9 with the addition, deletion or substitution of one, two or three amino acids), SEQ ID NO:10 (or SEQ ID NO:10 with the addition, deletion or substitution of one, two or three amino acids), and (iii) a light chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO:11 (or SEQ ID NO:11 with the addition, deletion or substitution of one, two or three amino acids); The present invention features an antibody comprising a heavy chain variable domain or region comprising the amino acid sequences set forth in SEQ ID NO:17 (or SEQ ID NO:17 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO:18 (or SEQ ID NO:18 with one, two or three amino acid additions, deletions or substitutions), and SEQ ID NO:19 (or SEQ ID NO:19 with one, two or three amino acid additions, deletions or substitutions), and a light chain variable domain or region comprising the amino acid sequences set forth in SEQ ID NO:25 (or SEQ ID NO:25 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO:26 (or SEQ ID NO:26 with one, two or three amino acid additions, deletions or substitutions), and SEQ ID NO:27 (or SEQ ID NO:27 with one, two or three amino acid additions, deletions or substitutions). The antibody may comprise the ability to bind to a human CSPG4 polypeptide (SEQ ID NO:33). In any of the embodiments, the antibody may be a monoclonal antibody or an scFv antibody.

[0017] In some embodiments, the antibody comprises a heavy chain variable domain or region of (i). The heavy chain variable domain or region may comprise an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:8.

[0018] In some embodiments, the antibody comprises a light chain variable domain or region of (i). The light chain variable domain or region may comprise an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:16.

[0019] In some embodiments, the antibody comprises a heavy chain variable domain or region of (ii). The heavy chain variable domain or region may comprise an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:24.

[0020] In some embodiments, the antibody comprises a light chain variable domain or region of (ii). The light chain variable domain or region may comprise an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:32.

[0021] The present specification relates to a method for producing a light chain variable domain or region comprising: (i) a heavy chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO:1 (or SEQ ID NO:1 with the addition, deletion or substitution of one, two or three amino acids), SEQ ID NO:2 (or SEQ ID NO:2 with the addition, deletion or substitution of one, two or three amino acids), and SEQ ID NO:3 (or SEQ ID NO:3 with the addition, deletion or substitution of one amino acid); and (ii) a light chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO:9 (or SEQ ID NO:9 with the addition, deletion or substitution of one, two or three amino acids), SEQ ID NO:10 (or SEQ ID NO:10 with the addition, deletion or substitution of one, two or three amino acids), and SEQ ID NO:11 (or SEQ ID NO:11 with the addition, deletion or substitution of one, two or three amino acids); or Also featured are antigen-binding fragments comprising a heavy chain variable domain or region comprising the amino acid sequences set forth in SEQ ID NO:17 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO:18 (or SEQ ID NO:18 with one, two or three amino acid additions, deletions or substitutions) and SEQ ID NO:19 (or SEQ ID NO:19 with one, two or three amino acid additions, deletions or substitutions), and a light chain variable domain or region comprising the amino acid sequences set forth in SEQ ID NO:25 (or SEQ ID NO:25 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO:26 (or SEQ ID NO:26 with one, two or three amino acid additions, deletions or substitutions) and SEQ ID NO:27 (or SEQ ID NO:27 with one, two or three amino acid additions, deletions or substitutions). The antigen-binding fragment may comprise the ability to bind to SEQ ID NO:33 or SEQ ID NO:34.

[0022] In some embodiments, the antigen-binding fragment comprises a heavy chain variable domain or region of (i). The heavy chain variable domain or region may comprise an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:8.

[0023] In some embodiments, the antigen-binding fragment comprises a light chain variable domain or region of (i). The light chain variable domain or region may comprise an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:16.

[0024] In some embodiments, the antigen-binding fragment comprises a heavy chain variable domain or region of (ii). The heavy chain variable domain or region may comprise an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:24.

[0025] In some embodiments, the antigen-binding fragment comprises a light chain variable domain or region of (ii). The light chain variable domain or region may comprise an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:32.

[0026] In any of the embodiments, the antigen-binding fragment may be monoclonal. In any of the embodiments, the antigen-binding fragment may be a Fab.

[0027] The specification also features a nucleic acid comprising a nucleic acid sequence encoding at least a portion of an antibody or antigen-binding fragment of any of the embodiments described herein, and a host cell comprising such a nucleic acid. The nucleic acid sequence can encode a heavy chain variable domain or region of any of (i)-(ii). The nucleic acid sequence can encode a light chain variable domain or region of any of (i)-(ii). The nucleic acid can be a viral vector or a phagemid.

[0028] Also featured herein is a chimeric antigen receptor comprising an antigen binding domain, a hinge, a transmembrane domain, and one or more signaling domains, wherein the antigen binding domain comprises an antibody or antigen binding fragment of any of the embodiments described herein. The antigen binding domain may comprise an scFv capable of binding to a CSPG4 polypeptide.

[0029] In another aspect, the document features a nucleic acid comprising a nucleic acid sequence encoding a chimeric antigen receptor of any of the embodiments described herein, and a host cell comprising such a nucleic acid. The nucleic acid can be a viral vector or a phagemid.

[0030] Also featured herein is a cell comprising a chimeric antigen receptor of any of the embodiments described herein. The cell can be a T cell, a stem cell, or a NK cell.

[0031] In another aspect, the present specification features a cell engager that includes a first antigen-binding domain, a linker, and a second antigen-binding domain, where the first antigen-binding domain includes an antibody or antigen-binding fragment of any of the embodiments described herein. The first antigen-binding domain can include an scFv capable of binding to a CSPG4 polypeptide. The first antigen-binding domain can be an IgG capable of binding to a CSPG4 polypeptide. The second antigen-binding domain can bind to a polypeptide expressed on the surface of a T cell (e.g., a CD3 polypeptide) or a NK cell (e.g., a CD16a polypeptide). The cell engager can include a third antigen-binding domain, e.g., a third antigen-binding domain that binds to a polypeptide expressed on the surface of a NK cell, such as a CD16a polypeptide.

[0032] In another aspect, the document features a nucleic acid comprising a nucleic acid sequence encoding a cell engager of any of the embodiments described herein, and a host cell comprising such a nucleic acid. The nucleic acid can be a viral vector or a phagemid.

[0033] The present specification also features a host cell that expresses a chimeric antigen receptor or a cell engager described herein. The host cell can be a T cell, a stem cell, or a NK cell.

[0034] In another aspect, the present specification features an antibody drug conjugate (ADC) that includes an antigen-binding domain covalently linked to a drug, where the antigen-binding domain comprises an antibody or antigen-binding fragment of any of the embodiments described herein. The antigen-binding domain can comprise an scFv capable of binding to a CSPG4 polypeptide or an IgG capable of binding to a CSPG4 polypeptide. The drug can be selected from the group consisting of calicheamicin, monomethyl auristatin E (MMAE), emtansine (DM1), and exatecan derivative (Dxd).

[0035] Also featured herein are compositions comprising an antibody or antigen-binding fragment described herein, a cell engager described herein, a cell described herein, or an ADC described herein. The composition can also include a checkpoint inhibitor (e.g., a checkpoint inhibitor selected from the group consisting of cemiplimab, nivolumab, pembrolizumab, JTX-4014, spartalizumab, camrelizumab, sintilimab, tislelizumab, toripalimab, dostallimab, INCMGA00012, AMP-224, AMP-514, avelumab, durvalumab, atezolizumab, KN035, CK-301, AUNP12, CA-170, BMS-986189, and ipilimumab).

[0036] Also featured is a method of treating a mammal (e.g., a human) with cancer. The method includes administering to the mammal (e.g., a human) a composition described herein. The cancer can be a CSPG4+ cancer, such as a CSPG4+ ovarian cancer. The number of cancer cells in the mammal (e.g., a human) can be reduced after the administration step.

[0037] The present specification also features a method for attaching a binding molecule to a CSPG4 polypeptide. The method includes contacting a CSPG4 polypeptide with an antibody or antigen-binding fragment described herein, or contacting a CSPG4 polypeptide with a chimeric antigen receptor, cell engager, or ADC described herein. The contacting step can be performed in vitro or in vivo. For example, the contacting step can be performed in a mammal (e.g., a human) by administering an antibody or antigen-binding fragment to the mammal (e.g., a human). For example, the contacting step can be performed in a mammal (e.g., a human) by administering a chimeric antigen receptor, cell engager, or ADC to the mammal (e.g., a human).

[0038] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this disclosure belongs. Methods and materials are described herein for use in this disclosure; other suitable methods and materials known in the art may also be used. Materials, methods and examples are merely illustrative and are not intended to be limiting. All publications, patent applications, patents, sequences, database entries and other references mentioned herein are incorporated by reference in their entirety. In case of conflict, the present specification, including definitions, will control.

[0039] The details of one or more embodiments of the invention are set forth in the accompanying drawings and the description below. Other features, objects and advantages of the invention will be apparent from the description and drawings, and from the claims. [Brief description of the drawings]

[0040] [Figure 1-1]Figures 1A-1D show that antibody 7H5A2 targets CSPG4. Figure 1A) Amino acid sequence of the recombinant protein target used to generate antibody 7H5A2 (SEQ ID NO: 34). Black text indicates the CSPG4 region (SEQ ID NO: 33) and lighter grey text corresponds to the human FC sequence for protein purification. Figures 1A and 1B) Western blot for CSPG4 antibody 763.74 (Figure 1B) and 7H5A2 (Figure 1C). Cells were harvested in lysis buffer and 40 μg of protein was loaded into each well from mock and CSPG4-CRISPR cell lines as indicated. Alpha-tubulin is included as a loading control. Figure 1D) Immunofluorescence staining of HEY parental cells with antibody 7H5A2. Red bar = 10 μm. Magnification 600x. [Figure 1-2] Figures 1A-1D show that antibody 7H5A2 targets CSPG4. Figure 1A) Amino acid sequence of the recombinant protein target used to generate antibody 7H5A2 (SEQ ID NO: 34). Black text indicates the CSPG4 region (SEQ ID NO: 33) and lighter grey text corresponds to the human FC sequence for protein purification. Figures 1A and 1B) Western blot for CSPG4 antibody 763.74 (Figure 1B) and 7H5A2 (Figure 1C). Cells were harvested in lysis buffer and 40 μg of protein was loaded into each well from mock and CSPG4-CRISPR cell lines as indicated. Alpha-tubulin is included as a loading control. Figure 1D) Immunofluorescence staining of HEY parental cells with antibody 7H5A2. Red bar = 10 μm. Magnification 600x. [Figure 2-1]Figures 2A-2C show CSPG4 protein expression characterized in an ovarian cancer cohort of 126 patients. Figure 2A) Representative images from IHC staining for CSPG4 in an ovarian cancer patient cohort. Staining intensity is scored on a scale of 0-3 [0 (negative), 1 (weak), 2 (moderate) and 3 (strong)]. Tumor tissue staining is indicated by red arrows and stromal staining is indicated by black arrows. The fraction of CSPG4 staining is scored from 0-4, reflecting the percentage of positively stained tumor cells in the sample [0 (0%), 1 (1-25%), 2 (25-50%), 3 (50-75%) and 4 (75-100%)]. Intensity and fractional positivity scores were added together to generate a total score (TS). A TS ≥ 4 was considered (considered) high expression of CSPG4, whereas a TS ≤ 3 was considered low expression. A TS of 0 was considered negative. Red bar = 20 μm. Figures 2B and 2C) Kaplan-Meier curves over 40 months for censored data from this 126-patient ovarian cancer cohort. Figure 2B) High CSPG4 expression (green line; 24 / 29) vs. low CSPG4 expression (blue line) (33 / 48) correlates with decreased progression-free survival (PFS: 22.615 ± 1.754 vs. 16.559 ± 1.940, X2 = 4.316, P = 0.038). Figure 2C) High CSPG4 expression (green line 19 / 29) vs. low CSPG4 expression (blue line 22 / 48) correlates with decreased overall survival (OS: 31.027 ± 1.353 vs. 24.046 ± 2.177, X2 = 7.366, P = 0.007). [Figure 2-2]Figures 2A-2C show CSPG4 protein expression characterized in an ovarian cancer cohort of 126 patients. Figure 2A) Representative images from IHC staining for CSPG4 in an ovarian cancer patient cohort. Staining intensity is scored on a scale of 0-3 [0 (negative), 1 (weak), 2 (moderate) and 3 (strong)]. Tumor tissue staining is indicated by red arrows and stromal staining is indicated by black arrows. The fraction of CSPG4 staining is scored from 0-4, reflecting the percentage of positively stained tumor cells in the sample [0 (0%), 1 (1-25%), 2 (25-50%), 3 (50-75%) and 4 (75-100%)]. Intensity and fractional positivity scores were added together to generate a total score (TS). A TS ≥ 4 was considered high expression of CSPG4, whereas a TS ≤ 3 was considered low expression. A TS of 0 was considered negative. Red bar = 20 μm. Figures 2B and 2C) Kaplan-Meier curves over 40 months for censored data from this 126-patient ovarian cancer cohort. Figure 2B) High CSPG4 expression (green line; 24 / 29) vs. low CSPG4 expression (blue line) (33 / 48) correlates with decreased progression-free survival (PFS: 22.615 ± 1.754 vs. 16.559 ± 1.940, X2 = 4.316, P = 0.038). Figure 2C) High CSPG4 expression (green line 19 / 29) vs. low CSPG4 expression (blue line 22 / 48) correlates with decreased overall survival (OS: 31.027 ± 1.353 vs. 24.046 ± 2.177, X2 = 7.366, P = 0.007). [Figure 3A]Figures 3A-3B show that CRISPR knockout of CSPG4 in A2780 ovarian cancer cells results in reduced tumor growth in vivo. NSG mice were IP injected with 2.0x105 luc+A2780 mock or A2780 CSPG4-CRISPR knockout cells. Figure 3A shows tumor growth monitored by bioluminescence imaging (BLI) on days 6, 13, and 27. Colored scale bars indicate photons / s / cm2 / sr. Figure 3B shows quantification of tumor burden based on BLI total flux (photons / sec). Data are presented as mean ± SD. ****P<0.0001 by ordinary two-way ANOVA with Sidak's multiple comparison test. Mock: n=4, CRISPR, n=5, negative control, n=5. [Figure 3B] Figures 3A-3B show that CRISPR knockout of CSPG4 in A2780 ovarian cancer cells results in reduced tumor growth in vivo. NSG mice were IP injected with 2.0x105 luc+A2780 mock or A2780 CSPG4-CRISPR knockout cells. Figure 3A shows tumor growth monitored by bioluminescence imaging (BLI) on days 6, 13, and 27. Colored scale bars indicate photons / s / cm2 / sr. Figure 3B shows quantification of tumor burden based on BLI total flux (photons / sec). Data are presented as mean ± SD. ****P<0.0001 by ordinary two-way ANOVA with Sidak's multiple comparison test. Mock: n=4, CRISPR, n=5, negative control, n=5. [Figure 4] Figures 4A-4C show CSPG4 expression in parental and CRISPR cell lines by flow cytometry. ES-2 (Figure 4A) and HEY (Figure 4B) parental and CRISPR knockout cells were stained with either regular mouse IgG or anti-CSPG4 antibody 763.74. Figure 4C) A2780 parental and CRISPR knockout cells were stained with either regular mouse IgG2a or anti-CSPG4 antibody 9.2.27. [Figure 5-1]5A-5G show that CSPG4 knockout results in significant loss of invasive capacity and cisplatin resistance in multiple ovarian tumor cell lines. Invasion assay using control (mock) and CSPG4 knockout (CRISPR) HEY cells (FIG. 5A) or A2780 cells (FIG. 5B). Bars represent total number of invasive cells from 5 random fields / wells from triplicate wells, ±SD, from 3 replicate experiments, n=6. P values ​​were determined by Student's t-test with Welch's correction. FIG. 5C shows a Western blot for CSPG4 in ES2 cell lines. Lanes: 1-ES-2 parent, 2-ES-2 mock, 3-ES-2 CSPG4 knockout, 4-ES-2 CSPG4 rescue. FIG. 5D shows an invasion assay using the indicated ES-2 cell lines. Invasion capacity is rescued (rescued) in ES-2 CRISPR knockout lines showing CSPG4 re-expression. P values ​​were determined by Student's t test with Welch's correction. Figures 5E-G show mock and knockout (CRISPR) A2780 cells (Figure 5E), mock and knockout HEY cells (Figure 5F), and mock, knockout and rescue ES-2 cells (Figure 5G) treated with increasing concentrations of cisplatin. Dose-response curves were plotted as percent MTS stained cells vs. untreated cells ± sem for each cell line from triplicate experiments (n = 9). The phenotype was rescued by re-expression of CSPG4 in ES-2 CRISPR cells (ES-2 rescue, in Figure 5G). [Figure 5-2]5A-5G show that CSPG4 knockout results in significant loss of invasive capacity and cisplatin resistance in multiple ovarian tumor cell lines. Invasion assay using control (mock) and CSPG4 knockout (CRISPR) HEY cells (FIG. 5A) or A2780 cells (FIG. 5B). Bars represent total number of invasive cells from 5 random fields / wells from triplicate wells, ±SD, from 3 replicate experiments, n=6. P values ​​were determined by Student's t-test with Welch's correction. FIG. 5C shows a Western blot for CSPG4 in ES2 cell lines. Lanes: 1-ES-2 parent, 2-ES-2 mock, 3-ES-2 CSPG4 knockout, 4-ES-2 CSPG4 rescue. FIG. 5D shows an invasion assay using the indicated ES-2 cell lines. Invasion capacity is rescued (rescued) in ES-2 CRISPR knockout lines showing CSPG4 re-expression. P values ​​were determined by Student's t test with Welch's correction. Figures 5E-G show mock and knockout (CRISPR) A2780 cells (Figure 5E), mock and knockout HEY cells (Figure 5F), and mock, knockout and rescue ES-2 cells (Figure 5G) treated with increasing concentrations of cisplatin. Dose-response curves were plotted as percent MTS stained cells vs. untreated cells ± sem for each cell line from triplicate experiments (n = 9). The phenotype was rescued by re-expression of CSPG4 in ES-2 CRISPR cells (ES-2 rescue, in Figure 5G). [Figure 5-3]5A-5G show that CSPG4 knockout results in significant loss of invasive capacity and cisplatin resistance in multiple ovarian tumor cell lines. Invasion assay using control (mock) and CSPG4 knockout (CRISPR) HEY cells (FIG. 5A) or A2780 cells (FIG. 5B). Bars represent total number of invasive cells from 5 random fields / wells from triplicate wells, ±SD, from 3 replicate experiments, n=6. P values ​​were determined by Student's t-test with Welch's correction. FIG. 5C shows a Western blot for CSPG4 in ES2 cell lines. Lanes: 1-ES-2 parent, 2-ES-2 mock, 3-ES-2 CSPG4 knockout, 4-ES-2 CSPG4 rescue. FIG. 5D shows an invasion assay using the indicated ES-2 cell lines. Invasion capacity is rescued (rescued) in ES-2 CRISPR knockout lines showing CSPG4 re-expression. P values ​​were determined by Student's t test with Welch's correction. Figures 5E-G show mock and knockout (CRISPR) A2780 cells (Figure 5E), mock and knockout HEY cells (Figure 5F), and mock, knockout and rescue ES-2 cells (Figure 5G) treated with increasing concentrations of cisplatin. Dose-response curves were plotted as percent MTS stained cells vs. untreated cells ± sem for each cell line from triplicate experiments (n = 9). The phenotype was rescued by re-expression of CSPG4 in ES-2 CRISPR cells (ES-2 rescue, in Figure 5G). [Figure 5-4]5A-5G show that CSPG4 knockout results in significant loss of invasive capacity and cisplatin resistance in multiple ovarian tumor cell lines. Invasion assay using control (mock) and CSPG4 knockout (CRISPR) HEY cells (FIG. 5A) or A2780 cells (FIG. 5B). Bars represent total number of invasive cells from 5 random fields / wells from triplicate wells, ±SD, from 3 replicate experiments, n=6. P values ​​were determined by Student's t-test with Welch's correction. FIG. 5C shows a Western blot for CSPG4 in ES2 cell lines. Lanes: 1-ES-2 parent, 2-ES-2 mock, 3-ES-2 CSPG4 knockout, 4-ES-2 CSPG4 rescue. FIG. 5D shows an invasion assay using the indicated ES-2 cell lines. Invasion capacity is rescued (rescued) in ES-2 CRISPR knockout lines showing CSPG4 re-expression. P values ​​were determined by Student's t test with Welch's correction. Figures 5E-G show mock and knockout (CRISPR) A2780 cells (Figure 5E), mock and knockout HEY cells (Figure 5F), and mock, knockout and rescue ES-2 cells (Figure 5G) treated with increasing concentrations of cisplatin. Dose-response curves were plotted as percent MTS stained cells vs. untreated cells ± sem for each cell line from triplicate experiments (n = 9). The phenotype was rescued by re-expression of CSPG4 in ES-2 CRISPR cells (ES-2 rescue, in Figure 5G). [Figure 6] Figure 1 shows all enriched GO terms identified in genes differentially expressed between ES-2 CRISPR knockout cells versus mean expression levels in RS-2 parental and mock cell lines. There were a total of 388 enriched GO terms identified. Highly significant GO terms (adjusted p-value < 0.005) of specific functional interest are highlighted. [Figure 7-1]7A-7E show that knockout of CSPG4 reduced anchorage-independent growth and spheroid formation. FIG. 7A) ES-2 parental, mock, CRISPR and stable CSPG4 rescue cells were grown for 7 days in soft agar colony formation assay. Bars represent total number of colonies counted from 5 random fields / wells from replicate wells, ±SE, n=6. P values ​​were determined by Student's t-test with Welch's correction. FIG. 7B and 7C) CSPG4-CRISPR cells form fewer / smaller spheroids when plated in methylcellulose medium. FIG. 7B) Representative images showing larger and more spheroids from mock and smaller and fewer spheroids from CSPG4-CRISPR cultures for three EOC cell lines. Bar=200 μm. FIG. 7C) Spheroid counts ±SD from 5 random fields / wells from 3 replicate experiments. *p<0.001 by Student's t-test with Welch's correction. Spheroids were quantified 7 days after seeding. n=3. Fig. 7D) Cells (mock, M; CRISPR, C; CRISPR rescued with CSPG4 re-expression, R) were cultured in 1.0% methylcellulose / complete medium for 7 days, harvested, and lysates were analyzed by Western blot for FAK expression / phosphorylation. Fig. 7E) Cell proliferation in spheroids. Cells were seeded in duplicate wells in 1% methylcellulose for 7 days, collected from the methylcellulose cultures, and re-seeded overnight in regular growth medium on tissue culture plates. The following day, all cells were harvested from each well and counted. P values ​​were determined by Student's t-test with Welch's correction comparing viable cell numbers (determined by trypan blue exclusion) between mock and CRISPR cells for each cell line. [Figure 7-2]7A-7E show that knockout of CSPG4 reduced anchorage-independent growth and spheroid formation. FIG. 7A) ES-2 parental, mock, CRISPR and stable CSPG4 rescue cells were grown for 7 days in soft agar colony formation assay. Bars represent total number of colonies counted from 5 random fields / wells from replicate wells, ±SE, n=6. P values ​​were determined by Student's t-test with Welch's correction. FIG. 7B and 7C) CSPG4-CRISPR cells form fewer / smaller spheroids when plated in methylcellulose medium. FIG. 7B) Representative images showing larger and more spheroids from mock and smaller and fewer spheroids from CSPG4-CRISPR cultures for three EOC cell lines. Bar=200 μm. FIG. 7C) Spheroid counts ±SD from 5 random fields / wells from 3 replicate experiments. *p<0.001 by Student's t-test with Welch's correction. Spheroids were quantified 7 days after seeding. n=3. Fig. 7D) Cells (mock, M; CRISPR, C; CRISPR rescued with CSPG4 re-expression, R) were cultured in 1.0% methylcellulose / complete medium for 7 days, harvested, and lysates were analyzed by Western blot for FAK expression / phosphorylation. Fig. 7E) Cell proliferation in spheroids. Cells were seeded in duplicate wells in 1% methylcellulose for 7 days, collected from the methylcellulose cultures, and re-seeded overnight in regular growth medium on tissue culture plates. The following day, all cells were harvested from each well and counted. P values ​​were determined by Student's t-test with Welch's correction comparing viable cell numbers (determined by trypan blue exclusion) between mock and CRISPR cells for each cell line. [Figure 7-3]7A-7E show that knockout of CSPG4 reduced anchorage-independent growth and spheroid formation. FIG. 7A) ES-2 parental, mock, CRISPR and stable CSPG4 rescue cells were grown for 7 days in soft agar colony formation assay. Bars represent total number of colonies counted from 5 random fields / wells from replicate wells, ±SE, n=6. P values ​​were determined by Student's t-test with Welch's correction. FIG. 7B and 7C) CSPG4-CRISPR cells form fewer / smaller spheroids when plated in methylcellulose medium. FIG. 7B) Representative images showing larger and more spheroids from mock and smaller and fewer spheroids from CSPG4-CRISPR cultures for three EOC cell lines. Bar=200 μm. FIG. 7C) Spheroid counts ±SD from 5 random fields / wells from 3 replicate experiments. *p<0.001 by Student's t-test with Welch's correction. Spheroids were quantified 7 days after seeding. n=3. Fig. 7D) Cells (mock, M; CRISPR, C; CRISPR rescued with CSPG4 re-expression, R) were cultured in 1.0% methylcellulose / complete medium for 7 days, harvested, and lysates were analyzed by Western blot for FAK expression / phosphorylation. Fig. 7E) Cell proliferation in spheroids. Cells were seeded in duplicate wells in 1% methylcellulose for 7 days, collected from the methylcellulose cultures, and re-seeded overnight in regular growth medium on tissue culture plates. The following day, all cells were harvested from each well and counted. P values ​​were determined by Student's t-test with Welch's correction comparing viable cell numbers (determined by trypan blue exclusion) between mock and CRISPR cells for each cell line. [Figure 8A]8A-8B show that CSPG4 expression is associated with epithelial to mesenchymal plasticity. FIG. 8A) Gene set enrichment analysis (GSEA) of RNA-seq data comparing parental, mock and CRISPR knockout ES-2 cell lines. ES-2 CRISPR cell lines show enrichment in expression of EMT-associated genes when compared to mock and parental cell lines. The normalized enrichment score is 2.190. FIG. 8B) Ovarian cancer TCGA cohort analyzed for mean CSPG4 expression and EMT signature score. [Figure 8B] 8A-8B show that CSPG4 expression is associated with epithelial to mesenchymal plasticity. FIG. 8A) Gene set enrichment analysis (GSEA) of RNA-seq data comparing parental, mock and CRISPR knockout ES-2 cell lines. ES-2 CRISPR cell lines show enrichment in expression of EMT-associated genes when compared to mock and parental cell lines. The normalized enrichment score is 2.190. FIG. 8B) Ovarian cancer TCGA cohort analyzed for mean CSPG4 expression and EMT signature score. [Figure 9-1]9A-9J: CSPG4 expression is associated with epithelial to mesenchymal plasticity and is mediated by CSPG4-associated changes in ZEB1 expression. FIG. 9A) Indicated cell lines, mock (M) and CRISPR (C), were both cultured separately for 7 days in spheroid formation assays, harvested and analyzed by Western blot for various EMT markers. FIG. 9B) Western blot for Zeb1 and CSPG4 expression in EOC cell lines treated for 48 hours with siRNA for ZEB1 or control. FIG. 9C) Western blot for pFAK, FAK and ZEB1 in HEY cells treated for 24 hours with the indicated concentrations of DMSO (control) or FAK inhibitor PND-1186 in regular growth medium. FIG. 9D) Spheroid formation in methylcellulose of EOC cell lines treated for 7 days with siRNA for ZEB1 or control. Bars represent spheroid counts ± SD from 5 random fields / well from triplicate wells. *p<0.02 by Student's t-test with Welch's correction. n=3 from three separate experiments. Figure 9E) Invasion assay of EOC cells treated with ZEB1 siRNA or control siRNA (Methods). *p<0.002 by Student's t-test with Welch's correction. n=3 from three separate experiments. Figures 9F and 9G) Cells were transfected with control or ZEB1 siRNA overnight and plated in cisplatin cytotoxicity assays the following day. MTS data were collected 4 days after cisplatin treatment. Untransfected CSPG4-CRISPR cells for A2780 (Figure 9F) and HEY (Figure 9G) were included as controls. Dose-response curves were plotted as the percentage of MTS stained cells versus untreated cells for each group. Combined data from three independent experiments are shown (n=9). (Figures 9H-9J) Kaplan-Meier curves for ovarian cancer patients in the combined cohort of TCGA and 14 GEO datasets show that tumors expressing CSPG4 (Figure 9H), ZEB1 (Figure 9I), and showing average combined expression of both (Figure 9J) are associated with reduced 5-year survival (red line) when compared to data derived from tumors negative for these markers (black line). [Figure 9-2]9A-9J: CSPG4 expression is associated with epithelial to mesenchymal plasticity and is mediated by CSPG4-associated changes in ZEB1 expression. FIG. 9A) Indicated cell lines, mock (M) and CRISPR (C), were both cultured separately for 7 days in spheroid formation assays, harvested and analyzed by Western blot for various EMT markers. FIG. 9B) Western blot for Zeb1 and CSPG4 expression in EOC cell lines treated for 48 hours with siRNA for ZEB1 or control. FIG. 9C) Western blot for pFAK, FAK and ZEB1 in HEY cells treated for 24 hours with the indicated concentrations of DMSO (control) or FAK inhibitor PND-1186 in regular growth medium. FIG. 9D) Spheroid formation in methylcellulose of EOC cell lines treated for 7 days with siRNA for ZEB1 or control. Bars represent spheroid counts ± SD from 5 random fields / well from triplicate wells. *p<0.02 by Student's t-test with Welch's correction. n=3 from three separate experiments. Figure 9E) Invasion assay of EOC cells treated with ZEB1 siRNA or control siRNA (Methods). *p<0.002 by Student's t-test with Welch's correction. n=3 from three separate experiments. Figures 9F and 9G) Cells were transfected with control or ZEB1 siRNA overnight and plated in cisplatin cytotoxicity assays the following day. MTS data were collected 4 days after cisplatin treatment. Untransfected CSPG4-CRISPR cells for A2780 (Figure 9F) and HEY (Figure 9G) were included as controls. Dose-response curves were plotted as the percentage of MTS stained cells versus untreated cells for each group. Combined data from three independent experiments are shown (n=9). (Figures 9H-9J) Kaplan-Meier curves for ovarian cancer patients in the combined cohort of TCGA and 14 GEO datasets show that tumors expressing CSPG4 (Figure 9H), ZEB1 (Figure 9I), and showing average combined expression of both (Figure 9J) are associated with reduced 5-year survival (red line) when compared to data derived from tumors negative for these markers (black line). [Figure 9-3]9A-9J: CSPG4 expression is associated with epithelial to mesenchymal plasticity and is mediated by CSPG4-associated changes in ZEB1 expression. FIG. 9A) Indicated cell lines, mock (M) and CRISPR (C), were both cultured separately for 7 days in spheroid formation assays, harvested and analyzed by Western blot for various EMT markers. FIG. 9B) Western blot for Zeb1 and CSPG4 expression in EOC cell lines treated for 48 hours with siRNA for ZEB1 or control. FIG. 9C) Western blot for pFAK, FAK and ZEB1 in HEY cells treated for 24 hours with the indicated concentrations of DMSO (control) or FAK inhibitor PND-1186 in regular growth medium. FIG. 9D) Spheroid formation in methylcellulose of EOC cell lines treated for 7 days with siRNA for ZEB1 or control. Bars represent spheroid counts ± SD from 5 random fields / well from triplicate wells. *p<0.02 by Student's t-test with Welch's correction. n=3 from three separate experiments. Figure 9E) Invasion assay of EOC cells treated with ZEB1 siRNA or control siRNA (Methods). *p<0.002 by Student's t-test with Welch's correction. n=3 from three separate experiments. Figures 9F and 9G) Cells were transfected with control or ZEB1 siRNA overnight and plated in cisplatin cytotoxicity assays the following day. MTS data were collected 4 days after cisplatin treatment. Untransfected CSPG4-CRISPR cells for A2780 (Figure 9F) and HEY (Figure 9G) were included as controls. Dose-response curves were plotted as the percentage of MTS stained cells versus untreated cells for each group. Combined data from three independent experiments are shown (n=9). (Figures 9H-9J) Kaplan-Meier curves for ovarian cancer patients in the combined cohort of TCGA and 14 GEO datasets show that tumors expressing CSPG4 (Figure 9H), ZEB1 (Figure 9I), and showing average combined expression of both (Figure 9J) are associated with reduced 5-year survival (red line) when compared to data derived from tumors negative for these markers (black line). [Figure 9-4]9A-9J: CSPG4 expression is associated with epithelial to mesenchymal plasticity and is mediated by CSPG4-associated changes in ZEB1 expression. FIG. 9A) Indicated cell lines, mock (M) and CRISPR (C), were both cultured separately for 7 days in spheroid formation assays, harvested and analyzed by Western blot for various EMT markers. FIG. 9B) Western blot for Zeb1 and CSPG4 expression in EOC cell lines treated for 48 hours with siRNA for ZEB1 or control. FIG. 9C) Western blot for pFAK, FAK and ZEB1 in HEY cells treated for 24 hours with the indicated concentrations of DMSO (control) or FAK inhibitor PND-1186 in regular growth medium. FIG. 9D) Spheroid formation in methylcellulose of EOC cell lines treated for 7 days with siRNA for ZEB1 or control. Bars represent spheroid counts ± SD from 5 random fields / well from triplicate wells. *p<0.02 by Student's t-test with Welch's correction. n=3 from three separate experiments. Figure 9E) Invasion assay of EOC cells treated with ZEB1 siRNA or control siRNA (Methods). *p<0.002 by Student's t-test with Welch's correction. n=3 from three separate experiments. Figures 9F and 9G) Cells were transfected with control or ZEB1 siRNA overnight and plated in cisplatin cytotoxicity assays the following day. MTS data were collected 4 days after cisplatin treatment. Untransfected CSPG4-CRISPR cells for A2780 (Figure 9F) and HEY (Figure 9G) were included as controls. Dose-response curves were plotted as the percentage of MTS stained cells versus untreated cells for each group. Combined data from three independent experiments are shown (n=9). (Figures 9H-9J) Kaplan-Meier curves for ovarian cancer patients in the combined cohort of TCGA and 14 GEO datasets show that tumors expressing CSPG4 (Figure 9H), ZEB1 (Figure 9I), and showing average combined expression of both (Figure 9J) are associated with reduced 5-year survival (red line) when compared to data derived from tumors negative for these markers (black line). [Figure 9-5]9A-9J: CSPG4 expression is associated with epithelial to mesenchymal plasticity and is mediated by CSPG4-associated changes in ZEB1 expression. FIG. 9A) Indicated cell lines, mock (M) and CRISPR (C), were both cultured separately for 7 days in spheroid formation assays, harvested and analyzed by Western blot for various EMT markers. FIG. 9B) Western blot for Zeb1 and CSPG4 expression in EOC cell lines treated for 48 hours with siRNA for ZEB1 or control. FIG. 9C) Western blot for pFAK, FAK and ZEB1 in HEY cells treated for 24 hours with the indicated concentrations of DMSO (control) or FAK inhibitor PND-1186 in regular growth medium. FIG. 9D) Spheroid formation in methylcellulose of EOC cell lines treated for 7 days with siRNA for ZEB1 or control. Bars represent spheroid counts ± SD from 5 random fields / well from triplicate wells. *p<0.02 by Student's t-test with Welch's correction. n=3 from three separate experiments. Figure 9E) Invasion assay of EOC cells treated with ZEB1 siRNA or control siRNA (Methods). *p<0.002 by Student's t-test with Welch's correction. n=3 from three separate experiments. Figures 9F and 9G) Cells were transfected with control or ZEB1 siRNA overnight and plated in cisplatin cytotoxicity assays the following day. MTS data were collected 4 days after cisplatin treatment. Untransfected CSPG4-CRISPR cells for A2780 (Figure 9F) and HEY (Figure 9G) were included as controls. Dose-response curves were plotted as the percentage of MTS stained cells versus untreated cells for each group. Combined data from three independent experiments are shown (n=9). (Figures 9H-9J) Kaplan-Meier curves for ovarian cancer patients in the combined cohort of TCGA and 14 GEO datasets show that tumors expressing CSPG4 (Figure 9H), ZEB1 (Figure 9I), and showing average combined expression of both (Figure 9J) are associated with reduced 5-year survival (red line) when compared to data derived from tumors negative for these markers (black line). [Figure 9-6]9A-9J: CSPG4 expression is associated with epithelial to mesenchymal plasticity and is mediated by CSPG4-associated changes in ZEB1 expression. FIG. 9A) Indicated cell lines, mock (M) and CRISPR (C), were both cultured separately for 7 days in spheroid formation assays, harvested and analyzed by Western blot for various EMT markers. FIG. 9B) Western blot for Zeb1 and CSPG4 expression in EOC cell lines treated for 48 hours with siRNA for ZEB1 or control. FIG. 9C) Western blot for pFAK, FAK and ZEB1 in HEY cells treated for 24 hours with the indicated concentrations of DMSO (control) or FAK inhibitor PND-1186 in regular growth medium. FIG. 9D) Spheroid formation in methylcellulose of EOC cell lines treated for 7 days with siRNA for ZEB1 or control. Bars represent spheroid counts ± SD from 5 random fields / well from triplicate wells. *p<0.02 by Student's t-test with Welch's correction. n=3 from three separate experiments. Figure 9E) Invasion assay of EOC cells treated with ZEB1 siRNA or control siRNA (Methods). *p<0.002 by Student's t-test with Welch's correction. n=3 from three separate experiments. Figures 9F and 9G) Cells were transfected with control or ZEB1 siRNA overnight and plated in cisplatin cytotoxicity assays the following day. MTS data were collected 4 days after cisplatin treatment. Untransfected CSPG4-CRISPR cells for A2780 (Figure 9F) and HEY (Figure 9G) were included as controls. Dose-response curves were plotted as the percentage of MTS stained cells versus untreated cells for each group. Combined data from three independent experiments are shown (n=9). (Figures 9H-9J) Kaplan-Meier curves for ovarian cancer patients in the combined cohort of TCGA and 14 GEO datasets show that tumors expressing CSPG4 (Figure 9H), ZEB1 (Figure 9I), and showing average combined expression of both (Figure 9J) are associated with reduced 5-year survival (red line) when compared to data derived from tumors negative for these markers (black line). [Figure 10-1] 10A-10D show that anti-CSPG4 antibody 7H5A2 inhibits Zeb1 expression and FAK activation, cell invasion and spheroid formation. FIG. 10A) Western blot for pFAK, FAK, and ZEB1 in EOC cells plated in 1% methylcellulose in the presence of 50 μg / ml normal mouse IgG1 (nmIgG1) or anti-CSPG4 antibody 7H5A2, cultured in suspension for 24 h. FIG. 10B) Invasion assay using HEY cells pre-incubated with 50 μg / ml control (nmIgG1) or anti-CSPG4 (763.74 and 7H5A2) antibodies for 2 h and then plated in an invasion chamber for 24 h in the presence of antibodies in the top chamber only. Bars represent the total number of invaded cells counted in 5 random fields / well from duplicate wells, ±SD. P values ​​were determined by Student's t-test with Welch's correction. n=6 from 3 separate experiments. FIG. 10C) Cell proliferation assay using EOC cells seeded in the presence of the indicated antibodies at 50 μg / ml in 1% methylcellulose for 7 days (HEY) or 14 days (A2790 and ES-2). Cells were harvested from the methylcellulose cultures at these time points and re-seeded overnight in regular growth medium in tissue culture plates. The following day, all cells were harvested from each well and counted. P values ​​were determined by Student's t-test with Welch's correction comparing the number of viable cells (determined by trypan blue dye exclusion) from each condition. FIG. 10D) Cells were pretreated with normal mouse IgG1 or antibody 7H5A2 (50 μg / ml) for 1 hour before seeding in 1% methylcellulose supplemented with the indicated antibodies at 50 μg / ml. Spheroids were harvested after 72 hours and assayed for caspase-3 activation by Western blot. Apoptosis of spheroids. [Figure 10-2]10A-10D show that anti-CSPG4 antibody 7H5A2 inhibits Zeb1 expression and FAK activation, cell invasion and spheroid formation. FIG. 10A) Western blot for pFAK, FAK, and ZEB1 in EOC cells plated in 1% methylcellulose in the presence of 50 μg / ml normal mouse IgG1 (nmIgG1) or anti-CSPG4 antibody 7H5A2, cultured in suspension for 24 h. FIG. 10B) Invasion assay using HEY cells pre-incubated with 50 μg / ml control (nmIgG1) or anti-CSPG4 (763.74 and 7H5A2) antibodies for 2 h and then plated in an invasion chamber for 24 h in the presence of antibodies in the top chamber only. Bars represent the total number of invaded cells counted in 5 random fields / well from duplicate wells, ±SD. P values ​​were determined by Student's t-test with Welch's correction. n=6 from 3 separate experiments. FIG. 10C) Cell proliferation assay using EOC cells seeded in the presence of the indicated antibodies at 50 μg / ml in 1% methylcellulose for 7 days (HEY) or 14 days (A2790 and ES-2). Cells were harvested from the methylcellulose cultures at these time points and re-seeded overnight in regular growth medium in tissue culture plates. The following day, all cells were harvested from each well and counted. P values ​​were determined by Student's t-test with Welch's correction comparing the number of viable cells (determined by trypan blue dye exclusion) from each condition. FIG. 10D) Cells were pretreated with normal mouse IgG1 or antibody 7H5A2 (50 μg / ml) for 1 hour before seeding in 1% methylcellulose supplemented with the indicated antibodies at 50 μg / ml. Spheroids were harvested after 72 hours and assayed for caspase-3 activation by Western blot. Apoptosis of spheroids. [Figure 11A]FIG. 11A) HEY mock and CRISPR cells were treated with the indicated anti-CSPG4 monoclonal antibodies or normal mouse IgG1 (nmIgG1). The indicated NK92 cell lines were added at a 1:1 effector:target ratio and ADCC was determined at 4 hours using the DELFIA EuTDA cytotoxicity assay according to the manufacturer's protocol. Bars represent percent specific cell lysis ± sem for triplicate samples. *p<0.05, ****p<0.0001 vs nmIgG1 by two-way ANOVA with Dunnett's multiple comparison test. ADCC of spheroids. FIG. 11B) Replicate of the experiment from FIG. 11A. **p=0.0088, ***p<0.0009 by two-way ANOVA with Dunnett's multiple comparison test compared to nmIgG1 control; ** and ***p<0.0009 vs no Ab control. ADCC of spheroids. FIG. 11C) OVCAR8 mock and CRISPR cells were treated with the indicated anti-CSPG4 monoclonal antibodies or normal mouse IgG1 (nmIgG1). The indicated NK92 cell lines were added at a 1:1 effector:target ratio and ADCC was determined at 4 hours using the DELFIA EuTDA cytotoxicity assay according to the manufacturer's protocol. Bars represent percent specific cell lysis for triplicate samples. **p=0.0038 by two-way ANOVA with Dunnett's multiple comparison test compared to nmIgG1 control; **p=0.0315 vs. no Ab control. ADCC of high grade serous cell spheroids. [Figure 11B]FIG. 11A) HEY mock and CRISPR cells were treated with the indicated anti-CSPG4 monoclonal antibodies or normal mouse IgG1 (nmIgG1). The indicated NK92 cell lines were added at a 1:1 effector:target ratio and ADCC was determined at 4 hours using the DELFIA EuTDA cytotoxicity assay according to the manufacturer's protocol. Bars represent percent specific cell lysis ± sem for triplicate samples. *p<0.05, ****p<0.0001 vs nmIgG1 by two-way ANOVA with Dunnett's multiple comparison test. ADCC of spheroids. FIG. 11B) Replicate of the experiment from FIG. 11A. **p=0.0088, ***p<0.0009 by two-way ANOVA with Dunnett's multiple comparison test compared to nmIgG1 control; ** and ***p<0.0009 vs no Ab control. ADCC of spheroids. FIG. 11C) OVCAR8 mock and CRISPR cells were treated with the indicated anti-CSPG4 monoclonal antibodies or normal mouse IgG1 (nmIgG1). The indicated NK92 cell lines were added at a 1:1 effector:target ratio and ADCC was determined at 4 hours using the DELFIA EuTDA cytotoxicity assay according to the manufacturer's protocol. Bars represent percent specific cell lysis for triplicate samples. **p=0.0038 by two-way ANOVA with Dunnett's multiple comparison test compared to nmIgG1 control; **p=0.0315 vs. no Ab control. ADCC of high grade serous cell spheroids. [Figure 11C]FIG. 11A) HEY mock and CRISPR cells were treated with the indicated anti-CSPG4 monoclonal antibodies or normal mouse IgG1 (nmIgG1). The indicated NK92 cell lines were added at a 1:1 effector:target ratio and ADCC was determined at 4 hours using the DELFIA EuTDA cytotoxicity assay according to the manufacturer's protocol. Bars represent percent specific cell lysis ± sem for triplicate samples. *p<0.05, ****p<0.0001 vs nmIgG1 by two-way ANOVA with Dunnett's multiple comparison test. ADCC of spheroids. FIG. 11B) Replicate of the experiment from FIG. 11A. **p=0.0088, ***p<0.0009 by two-way ANOVA with Dunnett's multiple comparison test compared to nmIgG1 control; ** and ***p<0.0009 vs no Ab control. ADCC of spheroids. FIG. 11C) OVCAR8 mock and CRISPR cells were treated with the indicated anti-CSPG4 monoclonal antibodies or normal mouse IgG1 (nmIgG1). The indicated NK92 cell lines were added at a 1:1 effector:target ratio and ADCC was determined at 4 hours using the DELFIA EuTDA cytotoxicity assay according to the manufacturer's protocol. Bars represent percent specific cell lysis for triplicate samples. **p=0.0038 by two-way ANOVA with Dunnett's multiple comparison test compared to nmIgG1 control; **p=0.0315 vs. no Ab control. ADCC of high grade serous cell spheroids. [Figure 12] Figure 12 shows CSPG4 expression in ovarian cancer cell lines. The indicated cell lines were assayed for CSPG4 expression by Western blot using anti-CSPG4 antibody 9.2.27 (Millipore). Western blots were probed for tubulin as a loading control. Both SKOV-3 and OVCAR-5 cell lines were negative for CSPG4, while the other cell lines expressed various amounts of CSPG4 protein. The Western blot image of OVCAR-8 is from a separate Western blot from a separate experiment from the other cell lines. [Figure 13A]Figures 13A-13B show that CSPG4 TriKE enhances NK cell-mediated killing of CSPG4-positive ovarian cancer tumor spheroids. GFP-expressing OVCAR-8 cells (Figure 13A) or CSPG4-negative SKOV3 ovarian cancer cells (Figure 13B) were cultured for 2 days in ultra-low attachment 96-well round-bottom plates (20,000 cells / well) for spheroid formation. Enriched NK cells were added to wells at a 2:1 effector:target ratio (40,000 cells / well) with no treatment, CSPG4 TriKE 8G5A6 (30 nM), CSPG4 TriKE 7H5A2 (30 nM), or IL-15 (3 nM). Images of each well were taken at the time of NK addition and at 2-hour increments over 4 days using the IncuCyte SX5-Live Cell Analysis Platform. The graph shows the mean fluorescence intensity (normalized to time 0) at each time point as a percentage of untreated tumor spheroids for each condition from three technical replicates. [Figure 13B] Figures 13A-13B show that CSPG4 TriKE enhances NK cell-mediated killing of CSPG4-positive ovarian cancer tumor spheroids. GFP-expressing OVCAR-8 cells (Figure 13A) or CSPG4-negative SKOV3 ovarian cancer cells (Figure 13B) were cultured for 2 days in ultra-low attachment 96-well round-bottom plates (20,000 cells / well) for spheroid formation. Enriched NK cells were added to wells at a 2:1 effector:target ratio (40,000 cells / well) with no treatment, CSPG4 TriKE 8G5A6 (30 nM), CSPG4 TriKE 7H5A2 (30 nM), or IL-15 (3 nM). Images of each well were taken at the time of NK addition and at 2-hour increments over 4 days using the IncuCyte SX5-Live Cell Analysis Platform. The graph shows the mean fluorescence intensity (normalized to time 0) at each time point as a percentage of untreated tumor spheroids for each condition from three technical replicates. [Figure 14]14 is a schematic diagram of an exemplary BiTE designed using the CDR1, CDR2, and CDR3 of the heavy chain provided herein and the CDR1, CDR2, and CDR3 of the light chain provided herein in an Ig format (e.g., IgG1 format). A humanized anti-CD3 scFv (e.g., gOKT3-7 scFv described in U.S. Pat. No. 6,750,325) can be linked to the C-terminus of the light chain via a linker (e.g., a (SGGGG)3-5 (SEQ ID NO: 35) linker). DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0041] Provided herein are binding agents (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, cell engagers, and ADCs) that bind (e.g., specifically bind) to CSPG4 polypeptides (e.g., human CSPG4 polypeptides). For example, provided herein are binding agents (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, cell engagers, and ADCs) that bind (e.g., specifically bind) to polypeptides that comprise, consist essentially of, or consist of the CSPG4 amino acids set forth in FIG. 1 (black text in FIG. 1, SEQ ID NO: 33). The binding agents described herein were generated against the juxtamembrane (D3) region of the CSPG4 polypeptide, a region previously associated with regulation of β1 integrins and cell motility. Targeting CSPG4 polypeptides with the binding agents described herein can inhibit ZEB1 expression, thus limiting the epithelial to mesenchymal shift mediated by CSPG4. In some embodiments, the binding agents described herein can block invasion and promote apoptosis of CSPG4-positive spheroids. This indicates that the anti-CSPG4 binding agents provided herein have an additional function on cell survival and inhibit CSPG4 +ve It can be used to target spheroids containing EOC tumor cells. In some cases, binding agents generated against the juxtamembrane domain can be used as therapeutics to inhibit tumor spread and metastasis in vivo.

[0042] The term "antibody" as used herein includes polyclonal antibodies, monoclonal antibodies, recombinant antibodies, humanized antibodies, human antibodies, chimeric antibodies, multispecific antibodies (e.g., bispecific antibodies) formed from at least two antibodies, diabodies, single chain variable fragment antibodies (e.g., scFv antibodies), and tandem single chain variable fragment antibodies (e.g., taFv). Diabodies can include two chains with heavy and light chain variable domains, respectively, derived from the same or different antibodies (see, e.g., Hornig and Farber-Schwarz, Methods Mol. Biol., 907:713-27 (2012); and Brinkmann and Kontermann, MAbs., 9(2):182-212 (2017)). The two variable regions can be connected by a polypeptide linker (e.g., a polypeptide linker having a length of 5-10 residues). In some cases, there may be an interdomain disulfide bond in one or both of the pairs of heavy and light chain variable domains of the diabody. An scFv is a single-chain polypeptide antibody in which the heavy and light chain variable domains are connected directly or via a polypeptide linker (e.g., a polypeptide linker having a length of 8-18 residues). See also Chen et al., Adv. Drug Deliv. Rev., 65(10):1357-1369 (2013). An scFv can be designed with an orientation in which the heavy chain variable domain is followed by the light chain variable domain, or the light chain variable domain is followed by the heavy chain variable domain. In both cases, an optional linker can be located between the two domains.

[0043] The antibodies provided herein may comprise the CDRs described herein (e.g., described in Table 1) and may be configured to be murine, humanized, or chimeric antibodies. In some cases, the antibodies provided herein may comprise the CDRs described herein (e.g., described in Table 1) and may be monoclonal antibodies. In some cases, the antibodies provided herein may comprise the CDRs described herein (e.g., described in Table 1) and may be configured as scFv antibodies.

[0044] The term "antigen-binding fragment" as used herein refers to a fragment of an antibody (e.g., a humanized antibody fragment, a murine antibody fragment, or a chimeric antibody fragment) that has the ability to bind to an antigen. Examples of antigen-binding fragments include, but are not limited to, Fab, Fab', or F(ab')2 antigen-binding fragments. The antigen-binding fragments provided herein may include CDRs described herein (e.g., as described in Table 1) and may be configured to be murine antigen-binding fragments, humanized antigen-binding fragments, or chimeric antigen-binding fragments. In some cases, the antigen-binding fragments provided herein may include CDRs described herein (e.g., as described in Table 1) and may be monoclonal antigen-binding fragments. In some cases, the antigen-binding fragments provided herein may include CDRs described herein (e.g., as described in Table 1) and may be configured as a Fab antibody. In some cases, the Fab antibody may include a partial hinge sequence for disulfide bonds between the heavy and light chains of the Fab.

[0045] The term "antibody domain" as used herein refers to a domain of an antibody, e.g., a heavy chain variable domain (VH domain) or a light chain variable domain (VL domain), in the absence of one or more other domains of an antibody. In some cases, an antibody domain may be a single antibody domain (e.g., a VH domain or a VL domain) capable of binding to an antigen. The antibody domains provided herein may comprise the CDRs described herein (e.g., as described in Table 1) and may be a murine antibody domain, a human VH domain), a humanized antibody domain (e.g., a humanized VH domain), or a chimeric antibody domain (e.g., a chimeric VH domain). In some cases, the antibody domains provided herein may comprise the CDRs described herein (e.g., as described in Table 1) and may be monoclonal antibody domains. In some cases, the antibody domains provided herein may comprise the CDRs described herein (e.g., as described in Table 1) and may be engineered as a single VH domain or a single VL domain.

[0046] The anti-CSPG4 antibodies, anti-CSPG4 antigen-binding fragments, or anti-CSPG4 antibody domains provided herein may be IgA-, IgD-, IgE-, IgG-, or IgM-type antibody domains, including IgG- or IgM-type, including but not limited to IgG1-, IgG2-, IgG3-, IgG4-, IgM1-, and IgM2-types. In some cases, the antibodies provided herein (e.g., anti-CSPG4 antibodies) may be scFv antibodies. In some cases, the antigen-binding fragments provided herein (e.g., anti-CSPG4 antibody fragments) may be Fab. In some cases, the antibodies provided herein (e.g., anti-CSPG4 antibodies) may be fully intact antibodies. In some cases, the antibody domains provided herein (e.g., anti-CSPG4 antibody domains) may be VH domains.

[0047] The term "chimeric antigen receptor" as used herein refers to a chimeric polypeptide designed to include an optional signal peptide, an antigen-binding domain, an optional hinge, a transmembrane domain, and one or more intracellular signaling domains. As described herein, the antigen-binding domain of a CAR provided herein can be designed to bind a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide). For example, a CAR provided herein can be designed to include components (e.g., combinations of CDRs) of an antibody, antigen-binding fragment, and / or antibody domain described herein as an antigen-binding domain, provided that the antigen-binding domain has the ability to bind a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide). In some examples, a CAR provided herein can be designed to include an antigen-binding domain (e.g., SEQ ID NOs: 1-3 and 9-11, or SEQ ID NOs: 17-19 and 25-27) that includes two sets of three CDRs (e.g., CDR1, CDR2, and CDR3 of the heavy chain, and CDR1, CDR2, and CDR3 of the light chain) of an antigen-binding fragment provided herein. In some cases, the antigen binding domain of a CAR that targets a CSPG4 polypeptide can be designed to include a VH domain described herein or an scFv antibody described herein.

[0048] In some cases, the CAR provided herein can be designed to include a signal peptide.Any suitable signal peptide can be used to design the CAR described herein.Examples of signal peptides that can be used to generate the CAR described herein include, but are not limited to, tPA signal peptide, BiP signal peptide, or CD8α signal peptide.

[0049] In some cases, the CARs provided herein can be designed to include a hinge. Any suitable hinge can be used to design the CARs described herein. Examples of hinges that can be used to create the CARs described herein include, but are not limited to, an Ig-derived hinge (e.g., an IgG1-derived hinge, an IgG2-derived hinge, or an IgG4-derived hinge), an Ig-derived hinge containing a CD2 domain and a CD3 domain, an Ig-derived hinge containing a CD2 domain and lacking a CD3 domain, an Ig-derived hinge containing a CD3 domain and lacking a CD2 domain, an Ig-derived hinge lacking a CD2 domain and lacking a CD3 domain, a CD8α-derived hinge, a CD28-derived hinge, and a CD3ζ-derived hinge. The CARs provided herein can be designed to include any suitable length of hinge. For example, the CARs provided herein can be designed to include a hinge that is about 3 to about 75 (e.g., about 3 to about 65, about 3 to about 50, about 5 to about 75, about 10 to about 75, about 5 to about 50, about 10 to about 50, about 10 to about 40, or about 10 to about 30) amino acid residues in length. In some cases, a linker sequence (e.g., (SGGGG) 2-5 (SEQ ID NO: 36) may be used.

[0050] The CARs provided herein can be designed to include any suitable transmembrane domain. For example, the transmembrane domain of the CARs provided herein can be, but is not limited to, a CD3 zeta transmembrane domain, a CD4 transmembrane domain, a CD8 alpha transmembrane domain, a CD28 transmembrane domain, and a 4-1BB transmembrane domain.

[0051] The CARs provided herein may be designed to include one or more intracellular signaling domains. For example, the CARs provided herein may be designed to include one, two, three, or four intracellular signaling domains. Any suitable intracellular signaling domain or combination of intracellular signaling domains may be used to generate the CARs described herein. Examples of intracellular signaling domains that can be used to generate the CARs described herein include, but are not limited to, CD3ζ intracellular signaling domain, CD27 intracellular signaling domain, CD28 intracellular signaling domain, OX40 (CD134) intracellular signaling domain, 4-1BB (CD137) intracellular signaling domain, CD278 intracellular signaling domain, DAP10 intracellular signaling domain, and DAP12 intracellular signaling domain. In some cases, the CARs described herein may be designed to be first generation CARs with CD3ζ intracellular signaling domain. In some cases, the CARs described herein may be designed to be second generation CARs with CD28 intracellular signaling domain followed by CD3ζ intracellular signaling domain. In some cases, the CARs described herein can be designed to be third generation CARs having (a) a CD28 intracellular signaling domain followed by (b) a CD27 intracellular signaling domain, an OX40 intracellular signaling domain, or a 4-1BB intracellular signaling domain followed by (c) a CD3ζ intracellular signaling domain. See, e.g., Feins, et al., Am J Hematol., 94(S1):S3-S9 (2019).

[0052] In some cases, a CAR that targets a CSPG4 polypeptide comprises a heavy chain variable domain comprising SEQ ID NO:1, SEQ ID NO:2, and SEQ ID NO:3, followed by a linker, such as (SGGGG) 2-5(SEQ ID NO:36), followed by a light chain variable domain comprising SEQ ID NO:9, SEQ ID NO:10, and SEQ ID NO:11, followed by a hinge, e.g., a hinge / linker (e.g., an IgG4-derived hinge, a CD8α hinge, or a linker plus IgG4-derived hinge), followed by a transmembrane domain (e.g., a human CD28 transmembrane domain or a CD8α transmembrane domain), followed by one or more intracellular signaling domains.

[0053] In some cases, a CAR that targets a CSPG4 polypeptide comprises a heavy chain variable domain comprising SEQ ID NO:8, followed by a linker, such as (SGGGG) 2-5 (SEQ ID NO: 36), followed by a light chain variable domain comprising SEQ ID NO: 16, followed by a hinge, e.g., a hinge / linker (e.g., an IgG4-derived hinge, a CD8α hinge, or a linker plus IgG4-derived hinge), followed by a transmembrane domain, followed by one or more intracellular signaling domains, e.g., an scFv having one or more intracellular signaling domains.

[0054] In some cases, a CAR that targets a CSPG4 polypeptide comprises a light chain variable domain comprising SEQ ID NO:9, SEQ ID NO:10, and SEQ ID NO:11, followed by a linker, such as (SGGGG) 2-5 (SEQ ID NO:36), followed by a heavy chain variable domain comprising SEQ ID NO:1, SEQ ID NO:2, and SEQ ID NO:3, followed by a hinge, e.g., a hinge / linker (e.g., an IgG4-derived hinge, a CD8α hinge, or a linker plus IgG4-derived hinge), followed by a transmembrane domain, followed by one or more intracellular signaling domains.

[0055] In some cases, a CAR that targets a CSPG4 polypeptide comprises a light chain variable domain comprising SEQ ID NO: 16, followed by a linker, such as (SGGGG) 2-5(SEQ ID NO:36), followed by a heavy chain variable domain comprising SEQ ID NO:8, followed by a hinge, e.g., a hinge / linker (e.g., an IgG4-derived hinge, a CD8α hinge, or a linker plus IgG4-derived hinge), followed by a transmembrane domain (e.g., a human CD28 transmembrane domain or a CD8α transmembrane domain), followed by one or more intracellular signaling domains.

[0056] In some cases, a CAR that targets a CSPG4 polypeptide comprises a heavy chain variable domain comprising SEQ ID NO: 17, SEQ ID NO: 18, and SEQ ID NO: 19, followed by a linker (SGGGG). 2-5 (SEQ ID NO:36), followed by a light chain variable domain comprising SEQ ID NO:25, SEQ ID NO:26, and SEQ ID NO:27, followed by a hinge (e.g., an IgG4-derived hinge, a CD8α hinge, or a linker plus IgG4-derived hinge), followed by a transmembrane domain (e.g., a human CD28 transmembrane domain or a CD8α transmembrane domain), followed by one or more intracellular signaling domains.

[0057] In some cases, a CAR that targets a CSPG4 polypeptide comprises a heavy chain variable domain comprising SEQ ID NO: 24, followed by a linker, such as (SGGGG) 2-5 (SEQ ID NO: 36), followed by a light chain variable domain comprising SEQ ID NO: 32, followed by a hinge, e.g., a hinge / linker (e.g., an IgG4-derived hinge, a CD8α hinge, or a linker plus IgG4-derived hinge), followed by a transmembrane domain (e.g., a human CD28 transmembrane domain or a CD8α transmembrane domain), followed by one or more intracellular signaling domains.

[0058] In some cases, a CAR that targets a CSPG4 polypeptide comprises a light chain variable domain comprising SEQ ID NO:25, SEQ ID NO:26, and SEQ ID NO:27, followed by a linker, such as (SGGGG) 2-5(SEQ ID NO:36), followed by a heavy chain variable domain comprising SEQ ID NO:17, SEQ ID NO:18, and SEQ ID NO:19, followed by a hinge, e.g., a hinge / linker (e.g., an IgG4-derived hinge, a CD8α hinge, or a linker plus IgG4-derived hinge), followed by a transmembrane domain, followed by one or more intracellular signaling domains.

[0059] In some cases, a CAR that targets a CSPG4 polypeptide comprises a light chain variable domain comprising SEQ ID NO: 32, followed by a linker, such as (SGGGG) 2-5 (SEQ ID NO: 36), followed by a heavy chain variable domain comprising SEQ ID NO: 24, followed by a hinge, e.g., a hinge / linker (e.g., an IgG4-derived hinge, a CD8α hinge, or a linker plus IgG4-derived hinge), followed by a transmembrane domain (e.g., a human CD28 transmembrane domain or a CD8α transmembrane domain), followed by one or more intracellular signaling domains.

[0060] The term "cell engager" as used herein refers to a polypeptide that includes two or more antigen-binding domains (e.g., two, three, or four antigen-binding domains) and has the ability to link two cells together. Examples of cell engagers include, but are not limited to, BiTE, BiKE, and TriKE. Generally, the cell engagers provided herein can be designed to include at least one antigen-binding domain capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) and at least one antigen-binding domain capable of binding to an antigen expressed on the surface of a cell (e.g., a T cell or an NK cell). In some cases, the cell engagers described herein can be designed to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) and an antigen expressed on the surface of a cell (e.g., a T cell or an NK cell) via two or more antigen-binding domains of the cell engager. + Cells (e.g., CSPG4 +A cell engager structure may link one cell (e.g., a cancer cell) to another cell (e.g., a T cell or an NK cell). Examples of cell engager structures of the cell engagers provided herein include, but are not limited to, those structures depicted in FIG. 14. In some cases, the anti-CD3 scFv depicted in FIG. 14 may be replaced with a different antigen binding domain capable of binding to an antigen expressed on the surface of a cell (e.g., a T cell or an NK cell).

[0061] When the cell engager comprises an antigen-binding domain capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) and one or more other antigen-binding domains (e.g., two, three, or four other antigen-binding domains), each of these other antigen-binding domains can bind to a different antigen expressed on the surface of a different cell type, or can bind to a different antigen expressed on the surface of the same cell type. In some embodiments, TriKE can be designed to have a first antigen-binding domain capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide), a second antigen-binding domain capable of binding to a first antigen expressed on the surface of an NK cell (e.g., a CD16 polypeptide, e.g., a CD16a polypeptide), and a third portion, e.g., IL-15, which promotes NK cell-specific proliferation. See, e.g., Arvindam, et al., Leukemia, 35(6):1586-1596 (2021). In some embodiments, the third portion is a third antigen-binding domain capable of binding to a second antigen expressed on the surface of an NK cell (e.g., CD16, NKG2D, NKG2C, NKp46, NKp30, or NKp44).

[0062] As described herein, at least one antigen-binding domain of a cell engager provided herein can be designed to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide). For example, a cell engager provided herein can be designed to include as an antigen-binding domain components (e.g., combinations of CDRs) of an antibody, antigen-binding fragment, and / or antibody domain described herein, provided that the antigen-binding domain has the ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide). In some examples, a cell engager provided herein can be designed to include an antigen-binding domain that includes two sets of three CDRs (e.g., CDR1, CDR2, and CDR3 of the heavy chain, and CDR1, CDR2, and CDR3 of the light chain) of an antigen-binding fragment provided herein (e.g., SEQ ID NOs: 1-3 and 9-11, or SEQ ID NOs: 17-19 and 25-27). In some cases, the antigen-binding domain of a cell engager that targets a CSPG4 polypeptide can be designed to include a VH domain described herein or an scFv / Fab antibody described herein. In some cases, the antigen-binding domain of a CAR described herein that has the ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) can be designed to include a VH domain described herein or an scFv / Fab antibody described herein. + It can be used as an antigen-binding domain of a cell engager to target cells.

[0063] As described herein, a cell engager can be designed to include at least one antigen-binding domain capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) and at least one other antigen-binding domain. The at least one other antigen-binding domain can be capable of binding to any suitable antigen expressed on the surface of a cell. For example, a cell engager, such as a CSPG4 +When designing a BiTE that links a cell to a T cell, the cell engager may include an antigen-binding domain capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) and an antigen-binding domain capable of binding to a polypeptide expressed on the surface of a T cell. Examples of polypeptides expressed on the surface of a T cell that can be targeted by the antigen-binding domain of a cell engager provided herein include, but are not limited to, CD3 polypeptide, NKG2D, TCR, and CD28. Examples of antigen-binding domains capable of binding to a polypeptide expressed on the surface of a T cell that can be used to generate a cell engager (e.g., a BiTE) provided herein include, but are not limited to, anti-CD3 scFv and anti-CD3 VH domain, CD28, TCR, and NKG2D. Further examples of amino acid sequences that can be used as antigen-binding domains capable of binding to a polypeptide expressed on the surface of a T cell (e.g., CD3) are described in U.S. Pat. No. 6,750,325 (see, e.g., the sequence listing in U.S. Pat. No. 6,750,325).

[0064] Cell engagers, e.g. CSPG4 +When designing BiKE or TriKE for linking cells to NK cells, the cell engager may include an antigen-binding domain capable of binding to a CSPG4 polypeptide (e.g., human CSPG4 polypeptide) and one or more (e.g., one, two, or three) antigen-binding domains capable of binding to a polypeptide expressed on the surface of an NK cell. Examples of polypeptides expressed on the surface of an NK cell that can be targeted by the antigen-binding domains of the cell engagers provided herein include, but are not limited to, CD16 polypeptides (e.g., CD16a polypeptide), CD16, NKG2D, NKG2C, NKp46, NKp30, and NKp44. Examples of antigen-binding domains capable of binding to a polypeptide expressed on the surface of an NK cell that can be used to generate the cell engagers provided herein (e.g., BiKE or TriKE) include, but are not limited to, anti-CD16a scFv and anti-CD16a VH domains. Further examples of amino acid sequences that can be used as antigen-binding domains capable of binding to a polypeptide expressed on the surface of a NK cell (e.g., CD16, NKG2D, NKG2C, NKp46, NKp30, or NKp44) are described in McCall et al. (Mol. Immunol., 36(7):433-445 (1999); see, e.g., anti-CD16 scFv sequences); International Patent Application Publication No. PCT / US2017 / 048721 (see, e.g., the CDRs and sequence listing for the anti-CD16a binding domain).

[0065] In some cases, the cell engagers provided herein may be designed to include a linker located between each antigen binding domain. Any suitable linker may be used to design the cell engagers provided herein, including but not limited to, (SGGGG) 3-5(SEQ ID NO: 35). The cell engagers provided herein may be designed to include a linker of any appropriate length. For example, the cell engagers provided herein may be designed to include a linker of about 3 to about 100 amino acid residues in length (e.g., about 3 to about 90, about 3 to about 80, about 3 to about 70, about 3 to about 60, about 3 to about 50, about 3 to about 40, about 3 to about 30, about 3 to about 20, about 3 to about 15, about 5 to about 100, about 10 to about 100, about 20 to about 100, about 30 to about 100, about 40 to about 100, about 50 to about 100, about 60 to about 100, about 70 to about 100, about 10 to about 50, about 10 to about 40, about 10 to about 30, about 10 to about 20, or about 12 to about 17). In some cases, the cell engagers (e.g., BiTEs) provided herein include the following: (SGGGG) 3-5 (SEQ ID NO: 35) linker. In some cases, the hinge of a CAR described herein can be used as a linker to generate a cell engager described herein.

[0066] In some cases, a cell engager (e.g., BiTE) that targets a CSPG4 polypeptide comprises a heavy chain variable domain comprising SEQ ID NO:1, SEQ ID NO:2, and SEQ ID NO:3, followed by a linker, such as the linker described in FIG. 10, followed by a light chain variable domain comprising SEQ ID NO:9, SEQ ID NO:10, and SEQ ID NO:11, followed by the linker (SGGGG). 3-5 (SEQ ID NO:35), followed by an scFv having an antigen binding domain capable of binding to a polypeptide expressed on the surface of a T cell (e.g., an anti-human CD3 scFv).

[0067] In some cases, a cell engager (e.g., BiTE) that targets a CSPG4 polypeptide comprises a light chain variable domain comprising SEQ ID NO:9, SEQ ID NO:10, and SEQ ID NO:11, followed by a linker, such as (SGGGG) 3-5 (SEQ ID NO:35), followed by a heavy chain variable domain comprising SEQ ID NO:1, SEQ ID NO:2, and SEQ ID NO:3, followed by a linker, such as (SGGGG) 3-5(SEQ ID NO:35), followed by an scFv having an antigen binding domain capable of binding to a polypeptide expressed on the surface of a T cell (e.g., an anti-human CD3 scFv).

[0068] In some cases, a cell engager (e.g., a BiTE) that targets a CSPG4 polypeptide comprises a heavy chain variable domain comprising SEQ ID NO:8, followed by a linker, e.g., (SGGGG) 3-5 (SEQ ID NO: 35), followed by a light chain variable domain comprising SEQ ID NO: 16, followed by a linker, e.g., a hinge / linker, followed by an antigen binding domain capable of binding to a polypeptide expressed on the surface of a T cell (e.g., an anti-human CD3 scFv).

[0069] In some cases, a cell engager (e.g., a BiTE) that targets a CSPG4 polypeptide comprises a light chain variable domain comprising SEQ ID NO: 16, followed by a linker, e.g., (SGGGG) 3-5 (SEQ ID NO:35), followed by a heavy chain variable domain comprising SEQ ID NO:8, followed by a linker, e.g., a hinge / linker, followed by an antigen binding domain capable of binding to a polypeptide expressed on the surface of a T cell (e.g., an anti-human CD3 scFv).

[0070] In some cases, a cell engager (e.g., BiKE or TriKE) that targets a CSPG4 polypeptide comprises a heavy chain variable domain comprising SEQ ID NO:1, SEQ ID NO:2, and SEQ ID NO:3, followed by a linker, such as (SGGGG) 3-5(SEQ ID NO:35), followed by a light chain variable domain comprising SEQ ID NO:9, SEQ ID NO:10, and SEQ ID NO:11, followed by a linker, e.g., a hinge / linker, followed by one or more antigen binding domains capable of binding to a polypeptide expressed on the surface of a NK cell (e.g., anti-human CD16a scFv for BiKE, or anti-human CD16a scFv and anti-human CD16, NKG2D, NKG2C, NKp46, NKp30, or NKp44 scFv for TriKE).

[0071] In some cases, a cell engager (e.g., BiKE or TriKE) that targets a CSPG4 polypeptide comprises a light chain variable domain comprising SEQ ID NO:9, SEQ ID NO:10, and SEQ ID NO:11, followed by a linker, such as (SGGGG) 3-5 (SEQ ID NO:35), followed by a heavy chain variable domain comprising SEQ ID NO:1, SEQ ID NO:2, and SEQ ID NO:3, followed by a linker, e.g., a hinge / linker, followed by one or more antigen binding domains capable of binding to a polypeptide expressed on the surface of a NK cell (e.g., anti-human CD16a scFv for BiKE, or anti-human CD16a scFv and anti-human CD16, NKG2D, NKG2C, NKp46, NKp30, or NKp44 scFv for TriKE).

[0072] In some cases, the cell engager that targets a CSPG4 polypeptide (e.g., BiKE or TriKE) comprises a heavy chain variable domain comprising SEQ ID NO:8, followed by a linker, such as (SGGGG) 3-5 (SEQ ID NO:35), followed by a light chain variable domain comprising SEQ ID NO:16, followed by a linker, e.g., a hinge / linker, followed by one or more antigen binding domains capable of binding to a polypeptide expressed on the surface of a NK cell (e.g., an anti-human CD16a scFv for BiKE, or an anti-human CD16a scFv and anti-human CD16, NKG2D, NKG2C, NKp46, NKp30, or NKp44 scFv for TriKE).

[0073] In some cases, the cell engager that targets a CSPG4 polypeptide (e.g., BiKE or TriKE) comprises a light chain variable domain comprising SEQ ID NO: 16, followed by a linker, such as (SGGGG) 3-5 (SEQ ID NO:35), followed by a heavy chain variable domain comprising SEQ ID NO:8, followed by a linker, e.g., a hinge / linker, followed by one or more antigen binding domains capable of binding to a polypeptide expressed on the surface of a NK cell (e.g., an anti-human CD16a scFv for BiKE, or an anti-human CD16a scFv and an anti-human CD16, NKG2D, NKG2C, NKp46, NKp30, or NKp44 scFv for TriKE).

[0074] In some cases, a cell engager (e.g., BiTE) that targets a CSPG4 polypeptide comprises a heavy chain variable domain comprising SEQ ID NO:17, SEQ ID NO:18, and SEQ ID NO:19, followed by a linker, such as (SGGGG) 3-5 (SEQ ID NO:35), followed by a light chain variable domain comprising SEQ ID NO:25, SEQ ID NO:26, and SEQ ID NO:27, followed by a linker, e.g., a hinge / linker, followed by an antigen binding domain capable of binding to a polypeptide expressed on the surface of a T cell (e.g., an anti-human CD3 scFv).

[0075] In some cases, a cell engager (e.g., BiTE) that targets a CSPG4 polypeptide comprises a light chain variable domain comprising SEQ ID NO:25, SEQ ID NO:26, and SEQ ID NO:27, followed by a linker, such as (SGGGG) 3-5 (SEQ ID NO:35), followed by a heavy chain variable domain comprising SEQ ID NO:17, SEQ ID NO:18, and SEQ ID NO:19, followed by a linker, e.g., a hinge / linker, followed by an antigen binding domain capable of binding to a polypeptide expressed on the surface of a T cell (e.g., an anti-human CD3 scFv).

[0076] In some cases, a cell engager (e.g., a BiTE) that targets a CSPG4 polypeptide comprises a heavy chain variable domain comprising SEQ ID NO: 24, followed by a linker, e.g., (SGGGG) 3-5 (SEQ ID NO: 35), followed by a light chain variable domain comprising SEQ ID NO: 32, followed by a linker, e.g., a hinge / linker, followed by an antigen binding domain capable of binding to a polypeptide expressed on the surface of a T cell (e.g., an anti-human CD3 scFv).

[0077] In some cases, a cell engager (e.g., a BiTE) that targets a CSPG4 polypeptide comprises a light chain variable domain comprising SEQ ID NO: 32, followed by a linker, e.g., (SGGGG) 3-5 (SEQ ID NO: 35), followed by a heavy chain variable domain comprising SEQ ID NO: 24, followed by a linker, e.g., a hinge / linker, followed by an antigen binding domain capable of binding to a polypeptide expressed on the surface of a T cell (e.g., an anti-human CD3 scFv).

[0078] In some cases, a cell engager (e.g., BiKE or TriKE) that targets a CSPG4 polypeptide comprises a heavy chain variable domain comprising SEQ ID NO: 17, SEQ ID NO: 18, and SEQ ID NO: 19, followed by a linker, such as (SGGGG) 3-5 (SEQ ID NO:35), followed by a light chain variable domain comprising SEQ ID NO:25, SEQ ID NO:26, and SEQ ID NO:27, followed by a linker, e.g., a hinge / linker, followed by one or more antigen binding domains capable of binding to a polypeptide expressed on the surface of a NK cell (e.g., anti-human CD16a scFv for BiKE, or anti-human CD16a scFv and anti-human CD16, NKG2D, NKG2C, NKp46, NKp30, or NKp44 scFv for TriKE).

[0079] In some cases, a cell engager (e.g., BiKE or TriKE) that targets a CSPG4 polypeptide comprises a light chain variable domain comprising SEQ ID NO:25, SEQ ID NO:26, and SEQ ID NO:27, followed by a linker, such as (SGGGG) 3-5 (SEQ ID NO:35), followed by a heavy chain variable domain comprising SEQ ID NO:17, SEQ ID NO:18, and SEQ ID NO:19, followed by a linker, e.g., a hinge / linker, followed by one or more antigen binding domains capable of binding to a polypeptide expressed on the surface of a NK cell (e.g., anti-human CD16a scFv for BiKE, or anti-human CD16a scFv and anti-human CD16, NKG2D, NKG2C, NKp46, NKp30, or NKp44 scFv for TriKE).

[0080] In some cases, a cell engager that targets a CSPG4 polypeptide (e.g., BiKE or TriKE) comprises a heavy chain variable domain comprising SEQ ID NO: 24, followed by a linker, such as (SGGGG) 3-5 (SEQ ID NO: 35), followed by a light chain variable domain comprising SEQ ID NO: 32, followed by a linker, e.g., a hinge / linker, followed by one or more antigen binding domains capable of binding to a polypeptide expressed on the surface of a NK cell (e.g., an scFv having an anti-human CD16a scFv for BiKE, or an anti-human CD16a scFv and an anti-human CD16, NKG2D, NKG2C, NKp46, NKp30, or NKp44 scFv for TriKE).

[0081] In some cases, the cell engager that targets a CSPG4 polypeptide (e.g., BiKE or TriKE) comprises a light chain variable domain comprising SEQ ID NO: 32, followed by a linker, such as (SGGGG) 3-5(SEQ ID NO:35), followed by a heavy chain variable domain comprising SEQ ID NO:24, followed by a linker, e.g., a hinge / linker, followed by one or more antigen binding domains capable of binding to a polypeptide expressed on the surface of a NK cell (e.g., an anti-human CD16a scFv for BiKE, or an anti-human CD16a scFv and anti-human CD16, NKG2D, NKG2C, NKp46, NKp30, or NKp44 scFv for TriKE).

[0082] In some cases, a cell engager (e.g., BiTE) targeted to a CSPG4 polypeptide can be designed to comprise an IgG (e.g., IgG1) configuration having (a) a heavy chain comprising, consisting essentially of, or consisting of a heavy chain variable domain comprising SEQ ID NO:1, SEQ ID NO:2, and SEQ ID NO:3, an Ig hinge, and a constant domain (e.g., CH1, CH2, and CH3 domains), and (b) a light chain comprising, consisting essentially of, or consisting of a light chain variable domain comprising SEQ ID NO:9, SEQ ID NO:10, and SEQ ID NO:11, a constant domain (e.g., a kappa or lambda constant domain), and an antigen binding domain capable of binding to a polypeptide expressed on the surface of a T cell (e.g., an anti-human CD3 scFv).

[0083] In some cases, a cell engager (e.g., BiTE) targeted to a CSPG4 polypeptide can be designed to comprise an IgG (e.g., IgG1) configuration having (a) a heavy chain comprising, consisting essentially of, or consisting of a heavy chain variable domain comprising SEQ ID NO:8, an Ig hinge, and a constant domain (e.g., CH1, CH2, and CH3 domains), and (b) a light chain comprising, consisting essentially of, or consisting of a light chain variable domain comprising SEQ ID NO:16, a constant domain (e.g., a kappa or lambda constant domain), and an antigen binding domain capable of binding to a polypeptide expressed on the surface of a T cell (e.g., an anti-human CD3 scFv).

[0084] In some cases, a cell engager (e.g., BiKE) targeted to a CSPG4 polypeptide can be designed to comprise an IgG (e.g., IgG1) configuration having (a) a heavy chain comprising, consisting essentially of, or consisting of a heavy chain variable domain comprising SEQ ID NO:1, SEQ ID NO:2, and SEQ ID NO:3, an Ig hinge, and a constant domain (e.g., CH1, CH2, and CH3 domains), and (b) a light chain comprising, consisting essentially of, or consisting of a light chain variable domain comprising SEQ ID NO:9, SEQ ID NO:10, and SEQ ID NO:11, a constant domain (e.g., a kappa or lambda constant domain), and an antigen binding domain capable of binding to a polypeptide expressed on the surface of a NK cell (e.g., an anti-human CD16a scFv).

[0085] In some cases, a cell engager (e.g., BiKE) targeted to a CSPG4 polypeptide can be designed to comprise an IgG (e.g., IgG1) configuration having (a) a heavy chain comprising, consisting essentially of, or consisting of a heavy chain variable domain comprising SEQ ID NO:8, an Ig hinge, and a constant domain (e.g., CH1, CH2, and CH3 domains), and (b) a light chain comprising, consisting essentially of, or consisting of a light chain variable domain comprising SEQ ID NO:16, a constant domain (e.g., a kappa or lambda constant domain), and an antigen binding domain capable of binding to a polypeptide expressed on the surface of a NK cell (e.g., an anti-human CD16a scFv).

[0086] In some cases, a cell engager (e.g., BiTE) targeted to a CSPG4 polypeptide can be designed to comprise an IgG (e.g., IgG1) configuration having (a) a heavy chain comprising, consisting essentially of, or consisting of a heavy chain variable domain comprising SEQ ID NO:17, SEQ ID NO:18, and SEQ ID NO:19, an Ig hinge, and a constant domain (e.g., CH1, CH2, and CH3 domains), and (b) a light chain comprising, consisting essentially of, or consisting of a light chain variable domain comprising SEQ ID NO:25, SEQ ID NO:26, and SEQ ID NO:27, a constant domain (e.g., a kappa or lambda constant domain), and an antigen binding domain capable of binding to a polypeptide expressed on the surface of a T cell (e.g., an anti-human CD3 scFv).

[0087] In some cases, a cell engager (e.g., BiTE) targeted to a CSPG4 polypeptide can be designed to comprise an IgG (e.g., IgG1) configuration having (a) a heavy chain comprising, consisting essentially of, or consisting of a heavy chain variable domain comprising SEQ ID NO:24, an Ig hinge, and a constant domain (e.g., CH1, CH2, and CH3 domains), and (b) a light chain comprising, consisting essentially of, or consisting of a light chain variable domain comprising SEQ ID NO:32, a constant domain (e.g., a kappa or lambda constant domain), and an antigen binding domain capable of binding to a polypeptide expressed on the surface of a T cell (e.g., an anti-human CD3 scFv).

[0088] In some cases, a cell engager (e.g., BiKE) targeted to a CSPG4 polypeptide can be designed to comprise an IgG (e.g., IgG1) configuration having (a) a heavy chain comprising, consisting essentially of, or consisting of a heavy chain variable domain comprising SEQ ID NO:17, SEQ ID NO:18, and SEQ ID NO:19, an Ig hinge, and a constant domain (e.g., CH1, CH2, and CH3 domains), and (b) a light chain comprising, consisting essentially of, or consisting of a light chain variable domain comprising SEQ ID NO:25, SEQ ID NO:26, and SEQ ID NO:27, a constant domain (e.g., a kappa or lambda constant domain), and an antigen binding domain capable of binding to a polypeptide expressed on the surface of a NK cell (e.g., anti-human CD16a scFv or anti-human CD16, NKG2D, NKG2C, NKp46, NKp30, or NKp44 scFv).

[0089] In some cases, a cell engager (e.g., BiKE) targeted to a CSPG4 polypeptide can be designed to comprise an IgG (e.g., IgG1) configuration having (a) a heavy chain comprising, consisting essentially of, or consisting of a heavy chain variable domain comprising SEQ ID NO:24, an Ig hinge, and a constant domain (e.g., CH1, CH2, and CH3 domains), and (b) a light chain comprising, consisting essentially of, or consisting of a light chain variable domain comprising SEQ ID NO:32, a constant domain (e.g., a kappa or lambda constant domain), and an antigen binding domain capable of binding to a polypeptide expressed on the surface of a NK cell (e.g., anti-human CD16a scFv or anti-human CD16, NKG2D, NKG2C, NKp46, NKp30, or NKp44 scFv).

[0090] In one embodiment, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) may comprise: (i) a heavy chain variable domain having a CDR1 having the amino acid sequence set forth in SEQ ID NO: 1 (or a variant of SEQ ID NO: 1 with one or two amino acid modifications), a CDR2 having the amino acid sequence set forth in SEQ ID NO: 2 (or a variant of SEQ ID NO: 2 with one or two amino acid modifications), and a CDR3 having the amino acid sequence set forth in SEQ ID NO: 3 (or a variant of SEQ ID NO: 3 with one or two amino acid modifications); and / or (ii) a light chain variable domain having a CDR1 having the amino acid sequence set forth in SEQ ID NO: 9 (or a variant of SEQ ID NO: 9 with one or two amino acid modifications), a CDR2 having the amino acid sequence set forth in SEQ ID NO: 10 (or a variant of SEQ ID NO: 10 with one or two amino acid modifications), and a CDR3 having the amino acid sequence set forth in SEQ ID NO: 11 (or a variant of SEQ ID NO: 11 with one or two amino acid modifications).

[0091] In some cases, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) having the ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) and having (a) a heavy chain variable domain having a CDR1 having the amino acid sequence set forth in SEQ ID NO: 1 (or a variant of SEQ ID NO: 1 with one or two amino acid modifications), a CDR2 having the amino acid sequence set forth in SEQ ID NO: 2 (or a variant of SEQ ID NO: 2 with one or two amino acid modifications), and a CDR3 having the amino acid sequence set forth in SEQ ID NO: 3 (or a variant of SEQ ID NO: 3 with one or two amino acid modifications), and / or (b) a light chain variable domain having a CDR1 having the amino acid sequence set forth in SEQ ID NO: 9 (or a variant of SEQ ID NO: 9 with one or two amino acid modifications), a CDR2 having the amino acid sequence set forth in SEQ ID NO: 10 (or a variant of SEQ ID NO: 10 with one or two amino acid modifications), and a CDR3 having the amino acid sequence set forth in SEQ ID NO: 11 (or a variant of SEQ ID NO: 11 with one or two amino acid modifications), may comprise any suitable framework region.For example, such binding agents (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, cell engagers and / or ADCs) may comprise: (a) a framework region 1 having the amino acid sequence set forth in SEQ ID NO:4 (or a variant of SEQ ID NO:4 having 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more amino acid modifications); framework region 2 having the amino acid sequence set forth in SEQ ID NO: 6 (or a variant of SEQ ID NO: 6 having 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more amino acid modifications); framework region 3 having the amino acid sequence set forth in SEQ ID NO: 7 (or a variant of SEQ ID NO: 7 having 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more amino acid modifications); and / or (b) a heavy chain variable domain comprising framework region 4 having the amino acid sequence set forth in SEQ ID NO: 12 (or a variant of SEQ ID NO: 12 having 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more amino acid modifications), framework region 1 having the amino acid sequence set forth in SEQ ID NO: 13 (or a variant of SEQ ID NO: 13 having 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more amino acid modifications). and a light chain variable domain comprising framework region 2 having the amino acid sequence set forth in SEQ ID NO: 14 (or a variant of SEQ ID NO: 14 having 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more amino acid modifications), and framework region 4 having the amino acid sequence set forth in SEQ ID NO: 15 (or a variant of SEQ ID NO: 15 having 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more amino acid modifications).

[0092] In some cases, a binding agent (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) having any of the CDRs described in Example 8 or Example 9 may be engineered to include a framework region or may be engineered to include one or more framework regions from another antibody, antibody fragment or antibody domain.

[0093] In some cases, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) may comprise (a) a heavy chain variable domain comprising an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:8, and / or (b) a light chain variable domain comprising an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:16. For example, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) can comprise (a) a heavy chain variable domain comprising an amino acid sequence having at least 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99 percent identity to the amino acid sequence set forth in SEQ ID NO:8, and / or (b) a light chain variable domain comprising an amino acid sequence having at least 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99 percent identity to the amino acid sequence set forth in SEQ ID NO:16. In some cases, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) can comprise (a) a heavy chain variable domain comprising an amino acid sequence having 100 percent identity to the amino acid sequence set forth in SEQ ID NO:8, and / or (b) a light chain variable domain comprising an amino acid sequence having 100 percent identity to the amino acid sequence set forth in SEQ ID NO:16.

[0094] In some cases, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) may comprise: (a) a heavy chain variable domain comprising an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:8, with the proviso that it comprises the amino acid sequence set forth in SEQ ID NO:1, 2, and 3; and / or (b) a light chain variable domain comprising an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:16, with the proviso that it comprises the amino acid sequence set forth in SEQ ID NO:9, 10, and 11. For example, a binding agent (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) provided herein may comprise: (a) a heavy chain variable domain comprising an amino acid sequence having at least 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99 percent identity to the amino acid sequence set forth in SEQ ID NO:8, with the proviso that the heavy chain variable domain comprises an amino acid sequence set forth in SEQ ID NO:1, 2, and 3; and / or (b) a light chain variable domain comprising an amino acid sequence having at least 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99 percent identity to the amino acid sequence set forth in SEQ ID NO:16, with the proviso that the light chain variable domain comprises an amino acid sequence set forth in SEQ ID NO:9, 10, and 11.

[0095] In some cases, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) may comprise (a) a heavy chain variable domain having the amino acid sequence set forth in SEQ ID NO: 8, or the amino acids set forth in SEQ ID NO: 8 with 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid modifications (e.g., amino acid substitutions, amino acid deletions, and / or amino acid additions), and / or (b) a light chain variable domain comprising the amino acid sequence set forth in SEQ ID NO: 16, or the amino acids set forth in SEQ ID NO: 16 with 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid modifications (e.g., amino acid substitutions, amino acid deletions, and / or amino acid additions). For example, the antibodies or antigen-binding fragments provided herein may have the ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) and may comprise a heavy chain variable domain having the amino acid sequence set forth in SEQ ID NO: 8 with one, two, three, four, five, six, seven, eight, nine or ten amino acid modifications (e.g., amino acid substitutions, amino acid deletions and / or amino acid additions), provided that they comprise the amino acid sequences set forth in SEQ ID NOs: 1, 2 and 3, and may comprise a light chain variable domain having the amino acid sequence set forth in SEQ ID NO: 16 with one, two, three, four, five, six, seven, eight, nine or ten amino acid modifications (e.g., amino acid substitutions, amino acid deletions and / or amino acid additions), provided that they comprise the amino acid sequences set forth in SEQ ID NOs: 9, 10 and 11.

[0096] In some cases, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) may comprise: (a) a heavy chain variable domain comprising: (i) a CDR1 comprising, consisting essentially of, or consisting of the amino acid sequence set forth in SEQ ID NO:1; (ii) a CDR2 comprising, consisting essentially of, or consisting of the amino acid sequence set forth in SEQ ID NO:2; and (iii) a CDR3 comprising, consisting essentially of, or consisting of the amino acid sequence set forth in SEQ ID NO:3; and / or (b) a light chain variable domain comprising: (i) a CDR1 comprising, consisting essentially of, or consisting of the amino acid sequence set forth in SEQ ID NO:9; (ii) a CDR2 comprising, consisting essentially of, or consisting of the amino acid sequence set forth in SEQ ID NO:10; and (iii) a CDR3 comprising, consisting essentially of, or consisting of the amino acid sequence set forth in SEQ ID NO:11.

[0097] As used herein, a "CDR1 consisting essentially of the amino acid sequence set forth in SEQ ID NO:1" is a CDR1 that has 0, 1, or 2 amino acid substitutions within SEQ ID NO:1, has 0, 1, 2, 3, 4, or 5 amino acid residues immediately preceding SEQ ID NO:1, and / or has 0, 1, 2, 3, 4, or 5 amino acid residues immediately following SEQ ID NO:1, provided that the binding agent (e.g., antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) maintains the basic ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide).

[0098] As used herein, a "CDR2 consisting essentially of the amino acid sequence set forth in SEQ ID NO:2" is a CDR2 that has 0, 1, or 2 amino acid substitutions within SEQ ID NO:2, has 0, 1, 2, 3, 4, or 5 amino acid residues immediately preceding SEQ ID NO:2, and / or has 0, 1, 2, 3, 4, or 5 amino acid residues immediately following SEQ ID NO:2, provided that the binding agent (e.g., antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) maintains the basic ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide).

[0099] As used herein, a "CDR3 consisting essentially of the amino acid sequence set forth in SEQ ID NO:3" is a CDR3 that has zero or one amino acid substitution within SEQ ID NO:3, has zero, one, two, three, four or five amino acid residues immediately preceding SEQ ID NO:3, and / or has zero, one, two, three, four or five amino acid residues immediately following SEQ ID NO:3, provided that the binding agent (e.g., antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) maintains the basic ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide).

[0100] As used herein, a "CDR1 consisting essentially of the amino acid sequence set forth in SEQ ID NO:9" is a CDR1 that has 0, 1, or 2 amino acid substitutions within SEQ ID NO:9, has 0, 1, 2, 3, 4, or 5 amino acid residues immediately preceding SEQ ID NO:9, and / or has 0, 1, 2, 3, 4, or 5 amino acid residues immediately following SEQ ID NO:9, provided that the binding agent (e.g., antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) maintains the basic ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide).

[0101] As used herein, a "CDR2 consisting essentially of the amino acid sequence set forth in SEQ ID NO:10" is a CDR2 that has 0, 1, or 2 amino acid substitutions within SEQ ID NO:10, has 0, 1, 2, 3, 4, or 5 amino acid residues immediately preceding SEQ ID NO:10, and / or has 0, 1, 2, 3, 4, or 5 amino acid residues immediately following SEQ ID NO:10, provided that the binding agent (e.g., antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) maintains the basic ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide).

[0102] As used herein, a "CDR3 consisting essentially of the amino acid sequence set forth in SEQ ID NO:11" is a CDR3 that has 0, 1, or 2 amino acid substitutions within SEQ ID NO:11, has 0, 1, 2, 3, 4, or 5 amino acid residues immediately preceding SEQ ID NO:11, and / or has 0, 1, 2, 3, 4, or 5 amino acid residues immediately following SEQ ID NO:11, provided that the binding agent (e.g., antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) maintains the basic ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide).

[0103] In another embodiment, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) may comprise: (i) a heavy chain variable domain having a CDR1 having the amino acid sequence set forth in SEQ ID NO: 17 (or a variant of SEQ ID NO: 17 with one or two amino acid modifications), a CDR2 having the amino acid sequence set forth in SEQ ID NO: 18 (or a variant of SEQ ID NO: 18 with one or two amino acid modifications), and a CDR3 having the amino acid sequence set forth in SEQ ID NO: 19 (or a variant of SEQ ID NO: 19 with one or two amino acid modifications); and / or (ii) a light chain variable domain having a CDR1 having the amino acid sequence set forth in SEQ ID NO: 25 (or a variant of SEQ ID NO: 25 with one or two amino acid modifications), a CDR2 having the amino acid sequence set forth in SEQ ID NO: 26 (or a variant of SEQ ID NO: 26 with one or two amino acid modifications), and a CDR3 having the amino acid sequence set forth in SEQ ID NO: 27 (or a variant of SEQ ID NO: 27 with one or two amino acid modifications).

[0104] In some cases, the antibody has the ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) and has (a) a heavy chain variable domain having a CDR1 having the amino acid sequence set forth in SEQ ID NO: 17 (or a variant of SEQ ID NO: 17 with one or two amino acid modifications), a CDR2 having the amino acid sequence set forth in SEQ ID NO: 18 (or a variant of SEQ ID NO: 18 with one or two amino acid modifications), and a CDR3 having the amino acid sequence set forth in SEQ ID NO: 19 (or a variant of SEQ ID NO: 19 with one or two amino acid modifications), and / or (b) an amino acid sequence set forth in SEQ ID NO: 25. The binding agents (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, cell engagers and / or ADCs) provided herein having a light chain variable domain having a CDR1 having the amino acid sequence of SEQ ID NO: 25 (or a variant of SEQ ID NO: 25 with one or two amino acid modifications), a CDR2 having the amino acid sequence set forth in SEQ ID NO: 26 (or a variant of SEQ ID NO: 26 with one or two amino acid modifications), and a CDR3 having the amino acid sequence set forth in SEQ ID NO: 27 (or a variant of SEQ ID NO: 27 with one or two amino acid modifications), can comprise any suitable framework region.

[0105] In some cases, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) may comprise (a) a heavy chain variable domain comprising an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO: 24, and / or (b) a light chain variable domain comprising an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO: 32. For example, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) can comprise (a) a heavy chain variable domain comprising an amino acid sequence having at least 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99 percent identity to the amino acid sequence set forth in SEQ ID NO: 24, and / or (b) a light chain variable domain comprising an amino acid sequence having at least 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99 percent identity to the amino acid sequence set forth in SEQ ID NO: 32. In some cases, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) can comprise (a) a heavy chain variable domain comprising an amino acid sequence having 100 percent identity to the amino acid sequence set forth in SEQ ID NO: 24, and / or (b) a light chain variable domain comprising an amino acid sequence having 100 percent identity to the amino acid sequence set forth in SEQ ID NO: 32.

[0106] In some cases, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) may comprise: (a) a heavy chain variable domain comprising an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO: 24, with the proviso that it comprises the amino acid sequences set forth in SEQ ID NOs: 17, 18, and 19; and / or (b) a light chain variable domain comprising an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO: 32, with the proviso that it comprises the amino acid sequences set forth in SEQ ID NOs: 25, 26, and 27. For example, a binding agent (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) provided herein may comprise: (a) a heavy chain variable domain comprising an amino acid sequence having at least 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99 percent identity to the amino acid sequence set forth in SEQ ID NO: 24, with the proviso that the heavy chain variable domain comprises an amino acid sequence set forth in SEQ ID NO: 17, 18, and 19; and / or (b) a light chain variable domain comprising an amino acid sequence having at least 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99 percent identity to the amino acid sequence set forth in SEQ ID NO: 32, with the proviso that the light chain variable domain comprises an amino acid sequence set forth in SEQ ID NO: 25, 26, and 27.

[0107] In some cases, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) may comprise (a) a heavy chain variable domain having the amino acid sequence set forth in SEQ ID NO: 24, or the amino acids set forth in SEQ ID NO: 24 with 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid modifications (e.g., amino acid substitutions, amino acid deletions, and / or amino acid additions), and / or (b) a light chain variable domain comprising the amino acid sequence set forth in SEQ ID NO: 32, or the amino acids set forth in SEQ ID NO: 32 with 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid modifications (e.g., amino acid substitutions, amino acid deletions, and / or amino acid additions). For example, the antibodies or antigen-binding fragments provided herein may have the ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) and may comprise a heavy chain variable domain having the amino acid sequence set forth in SEQ ID NO: 24 with one, two, three, four, five, six, seven, eight, nine, or ten amino acid modifications (e.g., amino acid substitutions, amino acid deletions, and / or amino acid additions), provided that they comprise the amino acid sequences set forth in SEQ ID NOs: 17, 18, and 19, and may comprise a light chain variable domain having the amino acid sequence set forth in SEQ ID NO: 32 with one, two, three, four, five, six, seven, eight, nine, or ten amino acid modifications (e.g., amino acid substitutions, amino acid deletions, and / or amino acid additions), provided that they comprise the amino acid sequences set forth in SEQ ID NOs: 25, 26, and 27.

[0108] In some cases, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) may comprise: (a) a heavy chain variable domain comprising: (i) a CDR1 comprising, consisting essentially of, or consisting of the amino acid sequence set forth in SEQ ID NO: 17; (ii) a CDR2 comprising, consisting essentially of, or consisting of the amino acid sequence set forth in SEQ ID NO: 18; and (iii) a CDR3 comprising, consisting essentially of, or consisting of the amino acid sequence set forth in SEQ ID NO: 19; and / or (b) a light chain variable domain comprising: (i) a CDR1 comprising, consisting essentially of, or consisting of the amino acid sequence set forth in SEQ ID NO: 25; (ii) a CDR2 comprising, consisting essentially of, or consisting of the amino acid sequence set forth in SEQ ID NO: 26; and (iii) a CDR3 comprising, consisting essentially of, or consisting of the amino acid sequence set forth in SEQ ID NO: 27. As used herein, a "CDR1 consisting essentially of the amino acid sequence set forth in SEQ ID NO: 17" is a CDR1 that has 0, 1, or 2 amino acid substitutions within SEQ ID NO: 17, has 0, 1, 2, 3, 4, or 5 amino acid residues immediately preceding SEQ ID NO: 17, and / or has 0, 1, 2, 3, 4, or 5 amino acid residues immediately following SEQ ID NO: 17, provided that the binding agent (e.g., antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) maintains the basic ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide).

[0109] As used herein, a "CDR2 consisting essentially of the amino acid sequence set forth in SEQ ID NO: 18" is a CDR2 that has 0, 1, or 2 amino acid substitutions within SEQ ID NO: 18, has 0, 1, 2, 3, 4, or 5 amino acid residues immediately preceding SEQ ID NO: 18, and / or has 0, 1, 2, 3, 4, or 5 amino acid residues immediately following SEQ ID NO: 18, provided that the binding agent (e.g., antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) maintains the basic ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide).

[0110] As used herein, a "CDR3 consisting essentially of the amino acid sequence set forth in SEQ ID NO: 19" is a CDR3 that has 0, 1, or 2 amino acid substitutions within SEQ ID NO: 19, has 0, 1, 2, 3, 4, or 5 amino acid residues immediately preceding SEQ ID NO: 19, and / or has 0, 1, 2, 3, 4, or 5 amino acid residues immediately following SEQ ID NO: 19, provided that the binding agent (e.g., antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) maintains the basic ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide).

[0111] As used herein, a "CDR1 consisting essentially of the amino acid sequence set forth in SEQ ID NO:25" is a CDR1 that has 0, 1, or 2 amino acid substitutions within SEQ ID NO:25, has 0, 1, 2, 3, 4, or 5 amino acid residues immediately preceding SEQ ID NO:25, and / or has 0, 1, 2, 3, 4, or 5 amino acid residues immediately following SEQ ID NO:25, provided that the binding agent (e.g., antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) maintains the basic ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide).

[0112] As used herein, a "CDR2 consisting essentially of the amino acid sequence set forth in SEQ ID NO:26" is a CDR2 that has 0, 1, or 2 amino acid substitutions within SEQ ID NO:26, has 0, 1, 2, 3, 4, or 5 amino acid residues immediately preceding SEQ ID NO:26, and / or has 0, 1, 2, 3, 4, or 5 amino acid residues immediately following SEQ ID NO:26, provided that the binding agent (e.g., antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) maintains the basic ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide).

[0113] As used herein, a "CDR3 consisting essentially of the amino acid sequence set forth in SEQ ID NO:27" is a CDR3 that has 0, 1, or 2 amino acid substitutions within SEQ ID NO:27, has 0, 1, 2, 3, 4, or 5 amino acid residues immediately preceding SEQ ID NO:27, and / or has 0, 1, 2, 3, 4, or 5 amino acid residues immediately following SEQ ID NO:27, provided that the binding agent (e.g., antibody, antigen-binding fragment, antibody domain, CAR, cell engager and / or ADC) maintains the basic ability to bind to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide).

[0114] When designing a single chain antibody (e.g., scFv) having a heavy chain variable domain and a light chain variable domain, the two regions may be linked directly or using any suitable linker sequence. For example, a heavy chain variable domain having CDRs of SEQ ID NOs: 1 to 3 or 17 to 19 may be directly linked via a linker sequence to a light chain variable domain having CDRs of SEQ ID NOs: 9 to 11 or 25 to 27, respectively. Examples of linker sequences that can be used to link the heavy chain variable domain and the light chain variable domain to create an scFv include, but are not limited to, (SGGGG) 3-5 (SEQ ID NO: 35).

[0115] As indicated herein, the amino acid sequences described herein may include amino acid modifications (e.g., a concatenated number of amino acid modifications). Such amino acid modifications may include, but are not limited to, amino acid substitutions, amino acid deletions, amino acid additions, and combinations. In some cases, amino acid modifications may be made to improve binding and / or contact with an antigen and / or to improve the functional activity of a binding agent (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) provided herein. In some cases, the amino acid substitutions in the concatenated sequence identifiers may be conservative amino acid substitutions. For example, a conservative amino acid substitution may be made by replacing one amino acid residue with another amino acid residue having a similar side chain. A family of amino acid residues having similar side chains can include amino acids with basic side chains (e.g., lysine, arginine, histidine), acidic side chains (e.g., aspartic acid, glutamic acid), uncharged polar side chains (e.g., glycine, asparagine, glutamine, serine, threonine, tyrosine, cysteine), nonpolar side chains (e.g., alanine, valine, leucine, isoleucine, proline, phenylalanine, methionine, tryptophan), beta-branched side chains (e.g., threonine, valine, isoleucine), and aromatic side chains (e.g., tyrosine, phenylalanine, tryptophan, histidine).

[0116] In some cases, the amino acid substitutions in the concatenated sequence identifiers may be non-conservative amino acid substitutions. Non-conservative amino acid substitutions can be made by replacing one amino acid residue with another amino acid residue having a dissimilar side chain. Examples of non-conservative substitutions include, but are not limited to, replacing (a) a hydrophilic residue (e.g., serine or threonine) with a hydrophobic residue (e.g., leucine, isoleucine, phenylalanine, valine, or alanine), (b) cysteine ​​or proline with any other residue, (c) a residue having a basic side chain (e.g., lysine, arginine, or histidine) with a residue having an acidic side chain (e.g., aspartic acid or glutamic acid), and (d) a residue having a bulky side chain (e.g., phenylalanine) with glycine or other residues having small side chains.

[0117] Methods for generating amino acid sequence variants (e.g., amino acid sequences containing one or more modifications relative to the concatenated sequence identifiers) can include site-directed or random mutagenesis (e.g., by PCR) of a nucleic acid encoding an antibody or fragment thereof. See, e.g., Zoller, Curr. Opin. Biotechnol. 3: 348-354 (1992). Both naturally occurring and non-naturally occurring amino acids (e.g., artificially derivatized amino acids) can be used to generate the amino acid sequence variants provided herein.

[0118] Further set forth in Table 1 are a representative number of binding agents (eg, antibodies, antigen-binding fragments, and / or antibody domains) capable of binding to a CSPG4 polypeptide (eg, a human CSPG4 polypeptide).

[0119] [Table 1]

[0120] The binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, cell engagers, and / or ADCs) can be produced using any suitable method. For example, the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers) can be produced in recombinant host cells. For example, nucleic acids encoding the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers) can be constructed, introduced into an expression vector, and expressed in a suitable host cell. Examples 8 and 9 include nucleic acid sequences encoding the variable domains (e.g., antibodies, antigen-binding fragments, and / or antibody domains) of the exemplary binding agents described herein. In some cases, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, and / or cell engager) can be recombinantly produced in a prokaryotic host, such as E. coli, Bacillus brevis, Bacillus subtilis, Bacillus megaterium, Lactobacillus zeae / casei, or Lactobacillus paracasei.The binding agents (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers) provided herein can be expressed in eukaryotic hosts, such as yeast (e.g., Pichia pastoris, Saccharomyces cerevisiae, Hansenula polymorpha, Schizosaccharomyces pombe, Schwanniomyces occidentalis, Kluyveromyces lactis, or Yarrowia lipolytica), filamentous fungi of the genus Trichoderma (e.g., T. reesei), and Aspergillus (e.g., A. niger). The antibodies can also be recombinantly produced in A. niger and A. oryzae, protozoa such as Leishmania tarentolae, insect cells, or mammalian cells (e.g., mammalian cell lines such as Chinese hamster ovary (CHO) cells, Per.C6 cells, mouse myeloma NS0 cells, baby hamster kidney (BHK) cells, or the human embryonic kidney cell line HEK293). See, e.g., Frenzel et al., Front Immunol., 4:217 (2013).

[0121] In some cases, the antigen-binding fragments or antibody domains provided herein can be produced by proteolytic digestion of intact antibodies. For example, antigen-binding fragments can be obtained by treating antibodies with enzymes such as papain or pepsin. Papain digestion of whole antibodies can be used to produce F(ab)2 or Fab fragments, while pepsin digestion of whole antibodies can be used to produce F(ab')2 or Fab' fragments.

[0122] In some cases, the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, cell engagers, and / or ADCs) may be substantially pure. The term "substantially pure" as used herein with respect to a binding agent (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) refers to a binding agent (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) that is substantially free of other polypeptides, lipids, carbohydrates, and nucleic acids with which it is naturally associated. That is, a substantially pure binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) is any binding agent (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) that is removed from its natural environment and is at least 60% pure. The substantially pure binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, cell engagers, and / or ADCs) may be at least about 65, 70, 75, 80, 85, 90, 95, or 99 percent pure.

[0123] The present specification also provides bispecific binding agents (e.g., bispecific antibodies, bispecific antigen-binding fragments, and / or bispecific antibody domains) that bind to two different epitopes, at least one of which is an epitope of a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide). In some cases, the bispecific binding agents provided herein can be designed to bind to two different epitopes of the same CSPG4 polypeptide (e.g., a human CSPG4 polypeptide). In some cases, the bispecific binding agents provided herein can bind to an epitope on a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) and a different polypeptide (e.g., a CD3 polypeptide). Bispecific binding agents can be produced by chemically conjugating two different binding agents (e.g., antibodies, antigen-binding fragments, and / or antibody domains) together. Bispecific binding agents can also be produced by fusing two antibody-producing cells, e.g., hybridomas, to create a hybrid cell line that produces two different heavy chains and two different light chains in the same cell, resulting in, for example, a bispecific IgG molecule. See Brinkmann and Kontermann, MAbs., 9(2):182-212 (2017).

[0124] In some cases, the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers) can be fused or conjugated (e.g., covalently or non-covalently linked) to another polypeptide or other moiety to obtain a fusion protein or conjugate. For example, the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers) can be conjugated (e.g., covalently or non-covalently linked) to a polymer (e.g., polyethylene glycol (PEG), PEG-modified polyethyleneimine (PEI) (PEI-PEG), and / or polyglutamic acid (PGA) (N-(2-hydroxypropyl) methacrylamide (HPMA) copolymers), hyaluronic acid, a fluorescent agent, a luminescent agent, a hapten, an enzyme, a metal chelate, a drug, a radioisotope, and / or a cytotoxic agent. Any suitable method may be used to conjugate (e.g., covalently or non-covalently link) another polypeptide or other moiety to the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers). For example, the methods described in U.S. Pat. No. 8,021,661 can be used to conjugate another polypeptide or other moiety to the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers).

[0125] In some cases, the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, cell engagers, and / or ADCs) can be modified with a moiety that improves stabilization and / or retention in circulation, e.g., in blood, serum, or other tissues, e.g., at least 1.5-fold, 2-fold, 5-fold, 10-fold, or 50-fold. For example, the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, cell engagers, and / or ADCs) can be conjugated (e.g., covalently or non-covalently conjugated) to a polymer, e.g., a substantially non-antigenic polymer. Examples of substantially non-antigenic polymers that can be used as described herein include, but are not limited to, polyalkylene oxides and polyethylene oxides. In some cases, the polymers used herein can have any suitable molecular weight. For example, polymers having an average molecular weight of about 200 daltons to about 35,000 daltons (e.g., about 1,000 to about 15,000 daltons or about 2,000 to about 12,500 daltons) can be used. In some cases, the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, cell engagers, and / or ADCs) can be conjugated (e.g., covalently or non-covalently conjugated) to a water-soluble polymer. Examples of water-soluble polymers that can be used as described herein include, but are not limited to, hydrophilic polyvinyl polymers, polyvinyl alcohol, polyvinylpyrrolidone, polyalkylene oxide homopolymers, polyethylene glycol (PEG), polypropylene glycol, polyoxyethylenated polyols, and copolymers and / or block copolymers thereof, provided that the water solubility of the copolymer or block copolymer is maintained.

[0126] In some cases, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) can be conjugated (e.g., covalently or non-covalently bound) to one or more polyoxyalkylenes (e.g., polyoxyethylene, polyoxypropylene, or block copolymers of polyoxyethylene and polyoxypropylene), polymethacrylates, carbomers, branched or unbranched polysaccharides, or combinations thereof. For example, a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) can be covalently bound to polyoxyethylene.

[0127] The present specification also provides ADCs. As used herein, the term "ADC" refers to a conjugate comprising (a) an antigen-binding domain and (b) at least one drug directly or indirectly covalently linked to the antigen-binding domain. In some cases, the ADCs described herein may comprise (a) an antigen-binding domain capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) and (b) at least one drug directly or indirectly covalently linked to the antigen-binding domain. Any suitable binding agent (e.g., an antibody, an antigen-binding fragment, and / or an antibody domain) provided herein and capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) can be used as an antigen-binding domain to generate the ADCs described herein. For example, any of the binding agents listed in Table 1 can be used to generate ADCs capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide). Examples of drugs that can be used to make the ADCs described herein include, but are not limited to, calicheamicin, monomethyl auristatin E (MMAE), emtansine (DM1), or exatecan derivative (Dxd). Any suitable ADC linker can be used to covalently link one or more drugs to an antigen-binding domain capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) to form the ADCs provided herein. For example, a cleavable or non-cleavable ADC linker can be used to covalently link one or more drugs to an antigen-binding domain capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) to form the ADCs provided herein.Examples of ADC linkers that can be used to covalently attach one or more drugs to an antigen-binding domain capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) to form the ADCs provided herein include, but are not limited to, ADC disulfide linkers, ADC hydrazone linkers, ADC peptide linkers, ADC thioether linkers, and ADC PEG-containing linkers.

[0128] The present specification also provides nucleic acid molecules (e.g., isolated nucleic acid molecules) having a nucleic acid sequence encoding at least a portion of a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, and / or cell engager). For example, an isolated nucleic acid molecule provided herein can include a nucleic acid sequence encoding a heavy chain variable domain, such as a heavy chain variable domain described in Examples 8 and 9. In another example, an isolated nucleic acid molecule provided herein can include a nucleic acid sequence encoding a light chain variable domain, such as a light chain variable domain described in Examples 8 and 9. In some cases, an isolated nucleic acid molecule provided herein can include a nucleic acid sequence encoding a linker polypeptide (e.g., (SGGGG) 3-5 (SEQ ID NO:35)), and may include a nucleic acid sequence encoding both (a) a heavy chain variable domain and (b) a light chain variable domain. The nucleic acids (e.g., isolated nucleic acid molecules) provided herein can be single-stranded or double-stranded nucleic acids of any suitable type (e.g., DNA, RNA, or DNA / RNA hybrids).

[0129] The present specification also provides vectors (e.g., plasmid vectors or viral vectors) containing one or more nucleic acids provided herein. Examples of plasmid vectors that can be designed to contain one or more nucleic acids having a nucleic acid sequence encoding at least a portion of the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers) include, but are not limited to, phagemids. Examples of viral vectors that can be designed to contain one or more nucleic acids having a nucleic acid sequence encoding at least a portion of the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers) include, but are not limited to, retroviral vectors, parvoviral vectors (e.g., adenoviral vectors and adeno-associated virus (AAV) vectors), lentiviral vectors (e.g., herpes simplex (HSV) vectors), poxviral vectors (e.g., vaccinia virus vectors and fowlpox virus vectors), and hybrid or chimeric viral vectors. For example, viral vectors having lentiviral components, such as adenoviral backbones including those described elsewhere (Zheng et al., Nat. Biotech., 18(2): 176-80 (2000); WO 98 / 22143; WO 98 / 46778; and WO 00 / 17376), or AAV components, such as adenoviral backbones including those described elsewhere (Fisher et al., Hum. Gene Ther., 7:2079-2087 (1996)), can be engineered to include one or more nucleic acids having a nucleic acid sequence encoding at least a portion of a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, and / or cell engager).

[0130] In some cases, a vector provided herein (e.g., a plasmid vector or a viral vector) can include a nucleic acid sequence encoding an scFv or an antibody domain (e.g., a VH domain) provided herein. In some cases, a vector provided herein (e.g., a plasmid vector or a viral vector) can include a nucleic acid sequence encoding a CAR provided herein. In some cases, a vector provided herein (e.g., a plasmid vector or a viral vector) can include a nucleic acid sequence encoding a cell engager provided herein.

[0131] The vectors provided herein (e.g., plasmid vectors or viral vectors provided herein) can include any suitable promoter and other control sequences (e.g., transcription and translation start and stop codons) operably linked to a nucleic acid sequence encoding at least a portion of a binding agent provided herein (e.g., an antibody, an antigen-binding fragment, an antibody domain, a CAR, and / or a cell engager). In some cases, the promoter used to drive expression can be a constitutive promoter or a controllable promoter. Examples of controllable promoters that can be used as described herein include, but are not limited to, inducible promoters, repressible promoters, and tissue-specific promoters. Examples of viral promoters that can be used as described herein include, but are not limited to, adenovirus promoters, vaccinia virus promoters, CMV promoters (e.g., immediate early CMV promoters), and AAV promoters.

[0132] Any suitable method can be used to generate a nucleic acid molecule (or vector, e.g., a plasmid vector or a viral vector) having a nucleic acid sequence encoding at least a portion of a binding agent (e.g., an antibody, an antigen-binding fragment, an antibody domain, a CAR, and / or a cell engager) provided herein. For example, molecular cloning techniques can be used to generate a nucleic acid molecule (or vector, e.g., a plasmid vector or a viral vector) having a nucleic acid sequence encoding at least a portion of a binding agent (e.g., an antibody, an antigen-binding fragment, an antibody domain, a CAR, and / or a cell engager) provided herein, as described elsewhere (see, e.g., Sambrook et al., Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory, NY (1989); and Ausubel et al., Current Protocols in Molecular Biology, Green Publishing Associates and John Wiley & Sons, New York, NY (1994)).

[0133] Also provided herein are host cells comprising a nucleic acid provided herein, e.g., a nucleic acid having a nucleic acid sequence encoding at least a portion of a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, and / or cell engager). Host cells that can be engineered to contain one or more nucleic acids provided herein can be prokaryotic or eukaryotic. Examples of prokaryotic cells that can be engineered to contain a nucleic acid provided herein include, but are not limited to, E. coli (e.g., Tb-1, TG-1, DH5α, XL-Blue MRF (Stratagene), SA2821, or Y1090 cells), Bacillus subtilis, Salmonella typhimurium, Serratia marcescens, or Pseudomonas (e.g., P. aerugenosa) cells. Examples of eukaryotic cells that can be engineered to contain a nucleic acid provided herein include, but are not limited to, insect cells (e.g., Sf9 or Ea4 cells), yeast cells (e.g., S. cerevisiae cells), and mammalian cells (e.g., mouse, rat, hamster, monkey, or human cells). For example, VERO cells, HeLa cells, 3T3 cells, Chinese hamster ovary (CHO) cells, W138 BHK cells, COS-7 cells, and MDCK cells can be engineered to contain a nucleic acid provided herein. Any suitable method may be used to introduce one or more nucleic acids provided herein (e.g., a vector, such as a plasmid vector or a viral vector, having a nucleic acid sequence encoding at least a portion of a binding agent provided herein) into a host cell.For example, calcium chloride-mediated transformation, transduction, conjugation, triparental mating, DEAE, dextran-mediated transfection, infection, membrane fusion with liposomes, high velocity bombardment with DNA-coated microprojectiles, direct microinjection into single cells, electroporation, or combinations thereof can be used to introduce the nucleic acids provided herein into host cells (see, e.g., Sambrook et al., Molecular Biology: A Laboratory Manual, Cold Spring Harbor Laboratory, NY (1989); Davis et al., Basic Methods in Molecular Biology (1986); and Neumann et al., EMBO J., 1:841 (1982)).

[0134] In some cases, cells, such as T cells, stem cells (e.g., induced pluripotent stem cells or mesenchymal stem cells), or NK cells, can be engineered to express one or more nucleic acids encoding a CAR described herein. For example, a population of T cells can be infected with a viral vector designed to express a nucleic acid encoding a CAR described herein (e.g., a CAR having the ability to bind to a CSPG4 polypeptide).

[0135] In some cases, cells, such as T cells, stem cells (e.g., induced pluripotent stem cells or mesenchymal stem cells), or NK cells, can be engineered to express one or more nucleic acids encoding a cell engager described herein. For example, a population of T cells can be infected with a viral vector designed to express a nucleic acid encoding a cell engager described herein (e.g., a cell engager capable of binding to a CSPG4 polypeptide).

[0136] In some cases, a binding agent (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, and / or cell engager) provided herein can be produced using a method that includes the steps of (a) introducing a nucleic acid encoding a polypeptide into a host cell, (b) culturing the host cell in a culture medium under conditions sufficient to express the polypeptide, (c) recovering the polypeptide from the cell or the culture medium, and (d) purifying the polypeptide (e.g., to reach at least 50, 60, 70, 80, 90, 95, 97, 98, or 99 percent purity).

[0137] In some cases, the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, cell engagers, and / or ADCs), nucleic acids provided herein (e.g., nucleic acids encoding antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers provided herein), vectors provided herein (e.g., viral vectors designed to express antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers provided herein), and / or host cells provided herein (e.g., host cells designed to express antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers provided herein) can be formulated as pharmaceutical compositions for administration to a mammal (e.g., a human) with cancer to treat the mammal. In some cases, the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, cell engagers, and / or ADCs), nucleic acids provided herein (e.g., nucleic acids encoding antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers provided herein), vectors provided herein (e.g., viral vectors designed to express antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers provided herein), and / or host cells provided herein (e.g., host cells designed to express antibodies, antigen-binding fragments, antibody domains, CARs, and / or cell engagers provided herein) can be formulated as pharmaceutical compositions for administration to a mammal (e.g., a human) to reduce the number of cancer cells in the mammal and / or increase survival of a mammal afflicted with cancer. For example, the binding agents provided herein (e.g., antibodies, antigen-binding fragments, antibody domains, cell engagers, and / or ADCs) capable of binding to a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) can be formulated as pharmaceutical compositions for administration to a mammal (e.g., a human).In some cases, the pharmaceutical compositions provided herein may include a pharma- ceutically acceptable carrier, such as a buffer, salt, surfactant, sugar, tonicity modifier, or combinations thereof, as described elsewhere (Gervasi et al., Eur. J. Pharmaceutics and Biopharmaceutics, 131:8-24 (2018)). Examples of pharma-ceutically acceptable carriers that can be used to make the pharmaceutical compositions provided herein include, but are not limited to, water, lactic acid, citric acid, sodium chloride, sodium citrate, sodium succinate, sodium phosphate, surfactants (e.g., polysorbate 20, polysorbate 80, or poloxamer 188), dextran 40, or sugars (e.g., sorbitol, mannitol, sucrose, dextrose, or trehalose), or combinations thereof. For example, a pharmaceutical composition designed to include a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, cell engager, and / or ADC) (or a nucleic acid, vector, or host cell provided herein) can be formulated to include a buffer (e.g., acetate, citrate, histidine, succinate, phosphate, or hydroxymethylaminomethane (Tris) buffer), a surfactant (e.g., polysorbate 20, polysorbate 80, or poloxamer 188), and a sugar, such as sucrose. Other components that can be included in the pharmaceutical compositions provided herein include, but are not limited to, amino acids such as glycine or arginine, antioxidants such as ascorbic acid, methionine, or ethylenediaminetetraacetic acid (EDTA), anticancer drugs such as enzalutamide, imatinib, gefitinib, erlotinib, sunitinib, lapatinib, nilotinib, sorafenib, temsirolimus, everolimus, pazopanib, crizotinib, ruxolitinib, axitinib, bosutinib, cabozantinib, ponatinib, regorafenib, ibrutinib, trametinib, perifosine, bortezomib, carfilzomib, batimastat, ganetespib, obatoclax, navitoclax, taxol, paclitaxel, or bevacizumab, or combinations thereof.For example, the pharmaceutical compositions provided herein can be used in combination with one or more checkpoint inhibitors, such as an anti-PD-1 antibody or PD-1 inhibitor (e.g., cemiplimab, nivolumab, pembrolizumab, JTX-4014, spartalizumab, camrelizumab, sintilimab, tislelizumab, toripalimab, dostallimab, INCMGA00012, AMP-224, or AMP-514), an anti-PD-L1 antibody or PD-L1 inhibitor (e.g., avelumab, durvalumab, atezolizumab, KN 035, CK-301, AUNP12, CA-170, or BMS-986189), and / or an anti-CTLA-4 antibody (e.g., ipilimumab), in combination with one or more binding agents provided herein (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cells engineered to express a CAR capable of binding to a CSPG4 polypeptide, one or more cell engagers, and / or one or more ADCs).

[0138] In some cases, when formulating a pharmaceutical composition to include one or more binding agents provided herein (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cells engineered to express a CAR capable of binding to a CSPG4 polypeptide, one or more cell engagers, and / or one or more ADCs), any suitable concentration of the binding agent can be used. For example, the pharmaceutical compositions provided herein contain about 1 mg to about 500 mg (e.g., about 1 mg to about 500 mg, about 10 mg to about 500 mg, about 50 mg to about 500 mg, about 100 mg to about 500 mg, about 0.5 mg to about 250 mg, about 0.5 mg to about 150 mg, about 0.5 mg to about 100 mg, about 0.5 mg to about 50 mg, about 1 mg to about 300 mg, about 2 mg to about 200 mg, about 10 mg to about 300 mg, about 25 mg to about 300 mg, about 50 mg to about 150 mg, or about 150 mg to about 300 mg) of a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, CAR +In another example, the pharmaceutical compositions provided herein may be formulated to be a solid or semi-solid containing about 0.5 mg to about 500 mg (e.g., about 1 mg to about 500 mg, about 10 mg to about 500 mg, about 50 mg to about 500 mg, about 100 mg to about 500 mg, about 0.5 mg to about 250 mg, about 0.5 mg to about 150 mg, about 0.5 mg to about 100 mg, about 0.5 mg to about 50 mg, about 1 mg to about 300 mg, about 10 mg to about 300 mg, about 25 mg to about 300 mg, about 50 mg to about 150 mg, or about 150 mg to about 300 mg) of a binding agent provided herein (e.g., an antibody, antigen-binding fragment, antibody domain, cell engager, and / or ADC). In some cases, pharmaceutical compositions containing a binding agent (e.g., an antibody, antigen-binding fragment, and / or antibody domain) provided herein may be used to treat a patient with a binding agent titer of about 1×10 5 ~Approx. 1×10 12 (For example, about 1×10 5 ~Approx. 1×10 10 , about 1×10 5 ~Approx. 1×10 8 , about 1×10 6 ~Approx. 1×10 12 , about 1×10 6 ~Approx. 1×10 12 , about 1×10 8 ~Approx. 1×10 12 , about 1×10 9 ~Approx. 1×10 12 , about 1×10 6 ~Approx. 1×10 11 , or about 1 × 10 7 ~Approx. 1×10 10 ) may be formulated as a dosage form.

[0139] In some cases, when a pharmaceutical composition is formulated to include one or more nucleic acids (e.g., a vector, e.g., a viral vector) encoding at least a portion of a binding agent (e.g., an antibody, antigen-binding fragment, antibody domain, CAR, and / or cell engager) provided herein, any suitable concentration of nucleic acid can be used. For example, the pharmaceutical compositions provided herein may be formulated to be liquids containing about 0.5 mg to about 500 mg (e.g., about 1 mg to about 500 mg, about 10 mg to about 500 mg, about 50 mg to about 500 mg, about 100 mg to about 500 mg, about 0.5 mg to about 250 mg, about 0.5 mg to about 150 mg, about 0.5 mg to about 100 mg, about 0.5 mg to about 50 mg, about 1 mg to about 300 mg, about 2 mg to about 200 mg, about 10 mg to about 300 mg, about 25 mg to about 300 mg, about 50 mg to about 150 mg, or about 150 mg to about 300 mg) of the nucleic acid provided herein per mL. In another example, the pharmaceutical compositions provided herein may be formulated to be solid or semi-solid containing about 0.5 mg to about 500 mg (e.g., about 1 mg to about 500 mg, about 10 mg to about 500 mg, about 50 mg to about 500 mg, about 100 mg to about 500 mg, about 0.5 mg to about 250 mg, about 0.5 mg to about 150 mg, about 0.5 mg to about 100 mg, about 0.5 mg to about 50 mg, about 1 mg to about 300 mg, about 10 mg to about 300 mg, about 25 mg to about 300 mg, about 50 mg to about 150 mg, or about 150 mg to about 300 mg).

[0140] In some cases, pharmaceutical compositions designed to include a binding agent (e.g., an antibody, antigen-binding fragment, antibody domain, cell engager, and / or ADC) provided herein can be formulated to include one or more agents that can reduce aggregation of the binding agent when formulated. Examples of such agents that can be used as described herein include, but are not limited to, methionine, arginine, lysine, aspartic acid, glycine, glutamic acid, and combinations thereof. In some cases, one or more of these amino acids can be included in the formulation at a concentration of about 0.5 mM to about 145 mM (e.g., about 1 mM to about 145 mM, about 10 mM to about 145 mM, about 100 mM to about 145 mM, about 0.5 mM to about 125 mM, about 0.5 mM to about 100 mM, about 0.5 mM to about 75 mM, or about 10 mM to about 100 mM).

[0141] The pharmaceutical compositions provided herein may be in any suitable form. For example, the pharmaceutical compositions provided herein may be designed to be liquid, semi-solid, or solid. In some cases, the pharmaceutical compositions provided herein may be a liquid solution (e.g., a solution for injection and / or infusion), a dispersion, a suspension, a tablet, a pill, a powder, a microemulsion, a liposome, or a suppository. In some cases, the pharmaceutical compositions provided herein may be lyophilized. In some cases, the pharmaceutical compositions provided herein (e.g., the pharmaceutical compositions provided herein comprising one or more binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs)) may be formulated with a carrier or coating designed to protect against rapid release. For example, the pharmaceutical compositions provided herein may be formulated as controlled or regulated release formulations as described elsewhere (U.S. Patent Application Publication Nos. 2019 / 0241667, 2019 / 0233522, and 2019 / 0233498).

[0142] The present specification also provides a method for the preparation of one or more binding agents provided herein (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) (or a nucleic acid, vector, or host cell provided herein (e.g., a CAR + Also provided are methods of administering a composition (e.g., a pharmaceutical composition provided herein) containing one or more binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) provided herein (or a nucleic acid, vector, and / or host cell (e.g., CAR) provided herein) to a mammal (e.g., a human). + A composition (e.g., a pharmaceutical composition provided herein) containing one or more binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) provided herein (or a nucleic acid, vector, and / or host cell (e.g., a CAR)) can be administered to a mammal (e.g., a human) with cancer to treat the mammal. In some cases, one or more binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) provided herein (or a nucleic acid, vector, and / or host cell (e.g., a CAR)) can be administered to a mammal (e.g., a human) with cancer to treat the mammal. + A composition containing the cancer cells (e.g., a pharmaceutical composition provided herein) can be administered to a mammal (e.g., a human) to reduce the number of cancer cells in the mammal and / or to increase survival of the mammal afflicted with cancer.

[0143] One or more binding agents provided herein (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) (or a nucleic acid, vector, or host cell provided herein (e.g., a CAR +Any suitable cancer can be treated using a composition (e.g., a pharmaceutical composition provided herein) containing one or more binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) provided herein. For example, a mammal (e.g., a human) having cancer can be treated by administering a composition (e.g., a pharmaceutical composition) containing one or more binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) provided herein to the mammal. In some cases, a cancer that can be treated as described herein can be a hematological cancer. In some cases, a cancer that can be treated as described herein can include one or more solid tumors. In some cases, a cancer that can be treated as described herein can be a metastatic cancer. In some cases, a cancer that can be treated as described herein can be a recurrent cancer. In some cases, a cancer that can be treated as described herein can be a chemotherapy-resistant cancer. Examples of cancers that can be treated as described herein include, but are not limited to, ovarian cancer, glioblastoma, melanoma, squamous cell carcinoma, triple-negative breast cancer, mesothelioma, osteosarcoma, AML, chordoma, or other tumors in which elevated CSPG4 expression has been linked to malignant progression or poor outcome. For example, a mammal having ovarian cancer can be administered a composition (e.g., a pharmaceutical composition) containing one or more binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) provided herein to treat the mammal (e.g., reduce the number of cancer cells in the mammal).

[0144] Any suitable method can be used to administer a composition (e.g., a pharmaceutical composition) provided herein to a mammal (e.g., a human). For example, a composition provided herein (e.g., a pharmaceutical composition containing one or more binding agents provided herein, such as one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs provided herein) can be administered to a mammal (e.g., a human) intravenously (e.g., via intravenous injection or infusion), subcutaneously (e.g., via subcutaneous injection), intraperitoneally (e.g., via intraperitoneal injection), orally, via inhalation, or intramuscularly (e.g., via intramuscular injection). In some cases, the route and / or mode of administration of a composition (e.g., a pharmaceutical composition provided herein) can be tailored for the mammal being treated.

[0145] In some cases, one or more binding agents provided herein (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) (or a nucleic acid, vector, or host cell provided herein (e.g., a CAR + An effective amount of a composition (e.g., a pharmaceutical composition provided herein) containing one or more binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) provided herein (or a nucleic acid, vector, or host cell (e.g., CAR) provided herein) can be an amount that reduces the number of cancer cells in a mammal having cancer without causing significant toxicity to the mammal. In some cases, one or more binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) provided herein (or a nucleic acid, vector, or host cell (e.g., CAR) provided herein) can be an amount that reduces the number of cancer cells in a mammal having cancer without causing significant toxicity to the mammal. +An effective amount of a composition (e.g., a pharmaceutical composition provided herein) containing a cellular marker (cells) can be an amount that extends the survival of a mammal having cancer compared to a control mammal having a similar level of cancer and not treated with the composition. For example, an effective amount of a binding agent (e.g., an antibody, antigen-binding fragment, antibody domain, cell engager, and / or ADC) provided herein can be from about 0.001 mg / kg to about 100 mg / kg (e.g., from about 0.001 mg / kg to about 90 mg / kg, from about 0.001 mg / kg to about 80 mg / kg, from about 0.001 mg / kg to about 70 mg / kg, from about 0.001 mg / kg to about 60 mg / kg, from about 0.001 mg / kg to about 100 mg / kg (e.g., from about 0.001 mg / kg to about 100 mg / kg, from about 0.001 mg / kg to about 150 mg / kg, from about 0.001 mg / kg to about 20 mg / kg, from about 0.001 mg / kg to about 30 mg / kg, from about 0.001 mg / kg to about 40 mg / kg, from about 0.001 mg / kg to about 50 mg / kg, from about 0.001 mg / kg to about 60 mg / kg, from about 0.001 mg / kg to about 70 mg / kg, from about 0.001 mg / kg to about 80 mg / kg, from about 0.001 mg / kg to about 90 mg / kg, from about 0.001 mg / kg to about 100 mg / kg). 001mg / kg ~ approx. 50mg / kg, approx. 0.001mg / kg ~ approx. 40mg / kg, approx. 0.001mg / kg ~ approx. 30mg / kg, approx. 0.005mg / kg ~ approx. 100mg / kg, approx. 0.01m g / kg~about 100mg / kg, about 0.05mg / kg~about 100mg / kg, about 0.1mg / kg~about 100mg / kg, about 0.5mg / kg~about 100mg / kg, about 1mg / kg~about 100m g / kg, about 5 mg / kg to about 100 mg / kg, about 0.01 mg / kg to about 25 mg / kg, about 0.1 mg / kg to about 30 mg / kg, about 0.15 mg / kg to about 25 mg / kg, about 0.2 mg / kg~about 20mg / kg, about 0.5mg / kg~about 20mg / kg, about 1mg / kg~about 30mg / kg, about 1mg / kg~about 25mg / kg, about 1mg / kg~about 20mg / kg, about 2mg / kg The effective amount may be about 20 mg / kg to about 20 mg / kg, about 5 mg / kg to about 30 mg / kg, about 10 mg / kg to about 30 mg / kg, about 15 mg / kg to about 30 mg / kg, about 20 mg / kg to about 30 mg / kg, about 3 mg / kg to about 30 mg / kg, about 0.5 mg / kg to about 10 mg / kg, about 1 mg / kg to about 10 mg / kg, about 1 mg / kg to about 5 mg / kg, or about 1 mg / kg to about 3 mg / kg. The effective amount may remain constant or may be adjusted as a sliding scale or variable dose depending on the mammal's response to the treatment. The actual effective amount used for a particular application may be affected by a variety of factors.For example, when treating a mammal with cancer, the severity of the cancer, the route of administration, the age and general health of the mammal, the use of excipients, the possibility of co-administration with other therapeutic or prophylactic treatments, such as the use of other agents (e.g., checkpoint inhibitors), and the judgment of the treating physician may require an increase or decrease in the actual effective amount of the compositions provided herein to be administered (e.g., pharmaceutical compositions containing one or more binding agents provided herein).

[0146] In some cases, one or more binding agents provided herein (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) (or a nucleic acid, vector, or host cell provided herein (e.g., a CAR + An effective frequency of administration of a composition (e.g., a pharmaceutical composition provided herein) containing one or more binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) provided herein (or a nucleic acid, vector, or host cell (e.g., CAR) provided herein) can be a frequency that reduces the number of cancer cells in a mammal having cancer without causing significant toxicity to the mammal. In some cases, an effective frequency of administration of a composition (e.g., a pharmaceutical composition provided herein) containing one or more binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) provided herein (or a nucleic acid, vector, or host cell (e.g., CAR) provided herein) can be a frequency that reduces the number of cancer cells in a mammal having cancer without causing significant toxicity to the mammal. +The effective frequency of administration of a composition (e.g., a pharmaceutical composition provided herein) containing one or more binding agents (cells) may be a frequency that extends the survival of a mammal having cancer compared to a control mammal having a similar cancer and not treated with the composition. For example, the effective frequency of administration of a pharmaceutical composition provided herein, e.g., a pharmaceutical composition containing one or more binding agents provided herein, may be from about twice daily to about once per year (e.g., from about twice daily to about once per month, from about twice daily to about once per week, from about once daily to about once per month, or from once daily to about once per week). In some cases, the frequency of administration of a pharmaceutical composition provided herein, e.g., a pharmaceutical composition containing one or more binding agents provided herein, may be daily. The frequency of administration of a pharmaceutical composition provided herein, e.g., a pharmaceutical composition containing one or more binding agents provided herein, may remain constant or may vary over the duration of the treatment. The actual effective frequency used for a particular application may be influenced by a variety of factors. For example, the severity of the cancer, the route of administration, the age and general health of the mammal, the use of excipients, the possibility of co-administration with other therapeutic or prophylactic treatments, such as the use of other agents (e.g., checkpoint inhibitors), and the judgment of the treating physician may require an increase or decrease in the actual effective frequency of administration of the compositions provided herein (e.g., pharmaceutical compositions containing one or more binding agents provided herein).

[0147] In some cases, one or more binding agents provided herein (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) (or a nucleic acid, vector, or host cell provided herein (e.g., a CAR +The effective duration of administration of a composition (e.g., a pharmaceutical composition provided herein) containing one or more binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) provided herein (or a nucleic acid, vector, or host cell (e.g., CAR) provided herein) can be a duration that reduces the number of cancer cells in the mammal without causing significant toxicity to the mammal. In some cases, the effective duration of administration of a composition (e.g., a pharmaceutical composition provided herein) containing one or more binding agents (e.g., one or more antibodies, one or more antigen-binding fragments, one or more antibody domains, one or more cell engagers, and / or one or more ADCs) provided herein (or a nucleic acid, vector, or host cell (e.g., CAR) provided herein) can be a duration that reduces the number of cancer cells in the mammal without causing significant toxicity to the mammal. + The effective duration of administration of a composition (e.g., a pharmaceutical composition provided herein) containing a cellular component (cells) may be a duration that extends the survival of a mammal having cancer compared to a control mammal having a similar cancer and not treated with the composition. For example, the effective duration of administration of a pharmaceutical composition provided herein, e.g., a pharmaceutical composition containing one or more binding agents provided herein, may vary from a single time point administration to several weeks to several months (e.g., 4-12 weeks). The actual effective duration used for a particular application may be influenced by a number of factors. For example, the severity of the cancer, the route of administration, the age and general health of the mammal, the use of excipients, the possibility of co-administration with other therapeutic or prophylactic treatments, e.g., the use of other agents (e.g., checkpoint inhibitors), and the judgment of the treating physician may require an extension or shortening of the actual effective duration of administration of a composition provided herein (e.g., a pharmaceutical composition containing one or more binding agents provided herein).

[0148] In some cases, the binding agents provided herein (e.g., antibodies, antigen-binding fragments, and / or antibody domains) can be used to detect the presence or absence of a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) in vitro, in situ, or in vivo (e.g., for in vivo imaging in a mammal, e.g., a human). For example, the binding agents provided herein (e.g., antibodies, antigen-binding fragments, and / or antibody domains) can be designed to include a label (e.g., a covalently attached radioactive, enzymatic, chromogenic, or fluorescent label). The labeled binding agent can be used to detect the presence or absence of a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) in a biological sample in vitro. Examples of biological samples that can be evaluated using the binding agents provided herein (e.g., antibodies, antigen-binding fragments, and / or antibody domains) include, but are not limited to, serum samples, plasma samples, tissue samples, biopsy samples, cell line samples, and tissue culture samples. In some cases, biological samples that can be evaluated as described herein can include mammalian body tissues and / or cells that can express a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide), such as white blood cells, ovarian tissues or cells, prostate tissues or cells, cardiac tissues or cells, placental tissues or cells, pancreatic tissues or cells, liver tissues or cells, splenic tissues or cells, lung tissues or cells, breast tissues or cells, head and neck tissues or cells, endometrial tissues or cells, colonic tissues or cells, colorectal tissues or cells, cervical tissues or cells, gastric tissues or cells, or umbilical cord tissues or cells. In some cases, binding agents provided herein (e.g., antibodies, antigen-binding fragments, and / or antibody domains) can be immobilized, for example, on a support, and retention of a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) from a biological sample on the support can be detected, and / or vice versa.In some cases, the binding agents provided herein (e.g., antibodies, antigen-binding fragments, and / or antibody domains) can be used in applications such as fluorescence polarization, microscopy, ELISA, centrifugation, chromatography, and / or cell sorting (e.g., fluorescence activated cell sorting).

[0149] In some cases, binding agents provided herein (e.g., antibodies, antigen-binding fragments, and / or antibody domains) containing a label (e.g., a covalently attached radioactive label) can be used to detect the presence or absence of a CSPG4 polypeptide (e.g., a human CSPG4 polypeptide) in a mammal (e.g., a human). For example, a labeled (e.g., covalently labeled) binding agent provided herein (e.g., antibodies, antigen-binding fragments, and / or antibody domains) having a radiolabel or an MRI-detectable label can be administered to a mammal (e.g., a human), and the mammal can be evaluated using a means for detecting the detectable label. In some cases, the mammal can be scanned to evaluate the location of the labeled binding agent provided herein within the mammal. For example, the mammal can be imaged using NMR or other tomography techniques.

[0150] Examples of labels that can be attached (e.g., covalently or non-covalently attached) to the binding agents (e.g., antibodies, antigen-binding fragments, and / or antibody domains) provided herein include, but are not limited to, radiolabels, e.g. 131 I, 111 In, 123 I, 99m Tc, 32 P, 33 P, 125 I, 3 H, 14 C, and 188These include Rh, fluorescent labels such as fluorescein and rhodamine, nuclear magnetic resonance active labels, positron emitting isotopes detectable by a positron emission tomography ("PET") scanner, chemiluminescent markers such as luciferin, and enzymatic markers such as peroxidase or phosphatase. In some cases, short-range luminescent markers, such as isotopes detectable by short-range detector probes, can be used.

[0151] The invention is further described in the following examples, which do not limit the scope of the invention described in the claims. EXAMPLES

[0152] [Example 1] Methods and Materials Cell Lines: Multiple human ovarian cancer cell lines were used in these studies, including three CSPG4-positive cell lines: HEY (RRID: CVCL_0297), A2780 (RRID: CVCL_0134), and ES-2 (RRID: CVCL_3509). All cell lines were authenticated via STR profiling by the ATCC Cell Line Authentication Service (Manassas, Virginia). Cell lines were routinely screened for mycoplasma using the Universal Mycoplasma Detection Kit (ATCC catalog number 30-1012K). ES-2 and A2780 cells were cultured at 37°C / 5% CO2 in basal DMEM medium (Mediatech catalog no. 10-013-cv) and HEY cells in basal RPMI1640 medium (Gibco catalog no. 11875-093) supplemented with 10% fetal bovine serum (Atlanta Biologicals catalog no. SS11150H, lot no. H1810S), 1% penicillin and streptomycin (Gibco catalog no. 15140-122). Cell lines were routinely used between passages 2 and 15 after thawing. CSPG4-CRISPR knockout and mock stable transfected variants of each cell line were maintained in the appropriate medium supplemented with 0.6 g / mL puromycin (Sigma catalog no. P8833).

[0153] Generation of CSPG4 CRISPR cell lines: The guide RNA (gRNA) target sequences used to generate CSPG4-CRISPR cells were 5' CGAGCGCGGCTCTGCTCCTG 3' (SEQ ID NO: 37) and 5'AGAGACCTGGAGACACCAGG 3' (SEQ ID NO: 38). gRNA was inserted into the plasmid pU6-gRNA. Both gRNA plasmids were co-transfected with a plasmid expressing CAS9 enzyme (pT3.5 Caggs-FLAG-hCas9) and plasmids for puromycin and GFP selection, pcDNA-PB7 and pPB SB-CG-LUC-GFP (Puro)(+CRE). Transfection was performed using Lipofectamine 2000 reagent (Invitrogen catalog number 11668-019) according to the protocol proposed by the manufacturer. Mock cell lines were transfected with selection plasmids only (pcDNA-PB7 and pPB SB-CG-LUC-GFP (Puro)(+CRE)) and selected as a pool by culturing in puromycin-containing medium (0.6 μg / mL). CSPG4-CRISPR knockout cell lines were selected by clonally plating in 96-well plates (based on GFP expression in the transfected population) and culturing in puromycin-containing medium (0.6 μg / mL). Single cell-derived colonies were expanded and screened by genomic PCR for deletion of the CSPG4 gene using primers 5' GGGCCCTTTAAGAAGGTTGA 3' (SEQ ID NO: 39) and 5' GTTTTGACAGCCCAAACCAG 3' (SEQ ID NO: 40). Cell lines were further screened by immunoblot and flow cytometry to verify loss of CSPG4 protein.

[0154] Further antibodies and reagents are listed in Table 2. [Table 2]

[0155] Transfection of siRNA: Small interfering RNA (siRNA) specific for human ZEB1 (catalog number sc-38643) was purchased from Santa Cruz Biotechnology (Dallas, TX), and negative control siRNA was purchased from Qiagen, Inc. (catalog number 1027281, Germantown, MD). Cells were seeded in 6-well plates until they were 60-70% confluent before transfection. Cells were transfected using Lipofectamine RNAimax transfection reagent (catalog number 13778075, ThermoFisher Scientific) according to the protocol suggested by the manufacturer. Cells were harvested 48 hours after transfection for seeding in proliferation, invasion, and spheroid formation assays.

[0156] Anchorage-independent growth assay: Soft agar growth assay was performed as previously described by Yang et al., Cancer Res., 69)19):7538-47 (2009) with the following modifications. Cells were seeded on the upper agarose layer at a final concentration of 0.6% agarose. Colonies were counted in 5 random fields / wells from triplicate wells at the indicated time points (see figure legends) for each cell line. Experiments were performed at least three times, and data shown are the mean number of colonies from 5 fields / wells from triplicate wells from three combined experiments, ±sem. Statistical significance was determined using Student's t-test.

[0157] Xenograft intraperitoneal injection mouse model: All animal studies were approved by the University of Minnesota Institutional Animal Care and Use Committee (IACUC 1908-37330A). NOD / SCID / γc- / - (NSG) 12-week-old female mice (Jackson Laboratories, Bar Harbor, ME) were used. One day before tumor injection, mice were sublethally irradiated (225 cGy). The following day, four female mice were injected with 2x10 5 Inject 2x10 A2780 mock luciferase-expressing tumor cells intraperitoneally into five female mice. 5 A2780 CSPG4-CRISPR luciferase expressing tumor cells were injected intraperitoneally. Tumor burden was monitored after luciferin injection (Goldbio, St. Louis, MO) using bioluminescence imaging (BLI) using an IVIS Spectrum in vivo Imaging system (PerkinElmer) at days 6, 13, and 27 after cell injection. Image analysis was performed using Living Image 4.5 software (PerkinElmer).

[0158] Cell invasion assay: Cells (2.5–5.0x10) in regular growth medium 4 ) were added to the upper chamber and the lower chamber filled with complete medium (ES-2, A2780) or serum-free medium (HEY) in triplicate wells of Matrigel invasion chambers (8 μm, Corning, NY) and cultured for 16–24 h at 37°C in a tissue culture incubator with 5% CO2 atmosphere. The remaining cells in the upper chamber were removed using a cotton swab, and the invaded cells were fixed and stained using a Differential Quick Staining Kit (Electron Microscopy Sciences, Hatfield, PA). The invaded cells were counted under a microscope at 100x magnification from five random fields / well. Each experiment was repeated at least three times.

[0159] Spheroid formation assay: 2x10 cells in 1% high viscosity methylcellulose (Sigma Cat. No. M0512) diluted in complete growth medium 5 Cells were suspended and seeded in poly-HEMA-coated 6-well plates and cultured for 7 days. At the indicated time points, spheroids were imaged and spheroids with diameters greater than 100 μm were counted under a microscope at 100x magnification from 5 random fields / well. Spheroids grown in methylcellulose cultures were harvested by dilution-dispersion in PBS, centrifugation at 400xg for 15, and washed twice in PBS for subsequent analysis. Each experiment was repeated at least three times.

[0160] Cisplatin cytotoxicity assay: Cisplatin stock (3.3 mM) was diluted in growth medium to the required concentration before each experiment. Cells were incubated at 1.0x10 3 Cells / well were seeded in 96-well plates. The following day, medium was removed from wells and replaced with 100 μl of medium containing the indicated treatment or medium alone (baseline) in triplicate wells. 96 hours after treatment, 20 μl of MTS reagent (Promega, Cat. No. G3580) was added to each well and plates were incubated in the dark for 2 hours at 37° C., 5% CO2 atmosphere. Absorbance at 570 nm was collected on a Tecan 200 plate reader. Each experiment was repeated at least three times.

[0161] RNAseq Analysis: Total RNA was isolated from 2 technical replicates derived from ES-2 parental, mock, and CSPG4-CRISPR cell lines using the Rneasy RNA isolation kit (Qiagen, Cat. No. 74104) following the protocol proposed by the manufacturer. RNA samples were submitted to the University of Minnesota Genomics Center for quality control assessment on an Agilent Bioanalyzer and quantification using the fluorometric RiboGreen assay. Strand-specific RNA-seq libraries were generated and sequenced on an IlluminaiSeq 2500 in high output mode, approximately 20 million reads / sample (2 replicate samples) and 2x125bp paired-end reads. Bulk RNAseq samples were processed and aligned using the CHURP version 0.2.2 command line interface framework. A complete description of the CHURP pipeline can be found in Baller et al., CHURP: A lightweight CLI Framework to Enable Novice Users to Analyze Sequencing Datasets in Parallel. Briefly, Trimmomatic version 0.33 (see Bolger et al., Bioinformatics, 30(15):2114-20 (2014)) was used to clean reads for adapter contamination and low quality sequences. FastQC (see Andrews, FastQC: A Quality Control Tool for High Throughput Sequence) was used to generate sequence quality reports for raw and trimmed reads. Samples were aligned to the Genome Reference Consortium H. sapiens build 38 reference genome using HISAT2 version 2.1.0 (see Kim et al., Nat Biotechnol., 37(8):907-15 (2019)). FeatureCounts v1.6.2 was used to count mapped reads to genes. See Liao et al., Bioinformatics, 30(7):923-30 (2014).

[0162] Gene Expression and Pathway Analysis: All differential gene expression and pathway analyses were performed in R v 3.6.3 (R Core Team, 2020). Differential gene expression analysis was performed using EdgeR v 3.28.1. See Robinson et al., Bioinformatics, 26(1):139-140 (2010). Differentially expressed genes were identified between the ES2 CSPG4-CRISPR / Cas9 knockout cell line and the average of the ES-2 parental and mock cell lines. Counts were normalized using a relative log expression normalization method, and only genes showing counts per million greater than 1 in two or more samples were retained. A general linear model approach was used to test for differentially expressed genes. After p-value adjustment, genes were classified as differentially expressed if the p-value was less than 0.01 and the log2 fold change was greater than 1. P-values ​​were adjusted using the Benjamini & Hochberg method. GO term enrichment analysis and gene set enrichment analysis (GSEA) were performed using the ClusterProfler R package. See Yu et al., OMICS, 16(5):284-7 (2012). Hallmark gene sets from the Molecular Signatures Database v 7.1 (gsea-msigdb.org / gsea / msigdb / index.jsp) were used in GSEA.

[0163] Survival analysis was performed using the Kaplan-Meier Plotter web-based informatics tool. See Gyorffy et al., Endocr Relat Cancer. 19(2):197-208 (2012); and Nagy et al., Sci Rep. 8(1):9227 (2018). A total of 15 ovarian cancer patient cohorts were included in the combined analysis, including the ovarian cancer TCGA cohort and 14 additional cohorts from the Gene Expression Omnibus (GEO) database; most cases in these datasets are serous, with a small proportion of endometrioid ovarian cancer. JetSet optimal probes (see Gyorffy et al., Breast Cancer Res Treat, 123(3):725-31 (2010)) were selected for CSPG4 (MCSPG) and ZEB1 expression analysis; 1656 patients had available data for CSPG4 and 355 patients had available data for ZEB1. Cohorts were separated by median for normalized single gene expression or mean normalized combined expression of both genes. Outcomes were censored at 5 years to match the duration of follow-up for immunohistochemistry studies performed separately.

[0164] Analysis of EMT signature enrichment was performed using the web-based Xena informatics tool on the ovarian cancer TCGA dataset (see Goldman et al., Nat Biotechnol. 38(6):675-8 (2020)). Cohorts were separated by mean CSPG4 gene expression and EMT signature scores were calculated using the Xena genomic signature signature. See Salt et al., Cancer Discov. 4(2):186-99 (2014).

[0165] Western Blot: Western blots were performed using standard methods previously described by Yang et al., Cancer Res. 69(19):7538-47 (2009).

[0166] Confocal microscopy: Cells were plated on coverslips for 48 hours, fixed with 4% paraformaldehyde for 15 minutes, permeabilized with 0.05% Triton X-100 for 5 minutes at room temperature, and blocked with 1% donkey serum for 1 hour. Coverslips were incubated with the indicated primary antibodies at 1 g / mL overnight at 4°C, washed twice with PBS+1% BSA for 10 minutes at room temperature, and incubated with Cy3-conjugated anti-mouse secondary antibodies (1:5000) for 1 hour at room temperature. After washing the cells twice with PBS+1% BSA, images were captured as described by Yang et al., J Cell Biol. 165(6):881-91 (2004).

[0167] Flow cytometry: Cells were released in PBS / 5mM EDTA solution and washed twice with FACS buffer (RPMI medium supplemented with 1% goat serum and 5mM HEPES). Cells were incubated with the indicated primary antibodies for 45 min at 4° C., washed three times with FACS buffer, and then incubated with goat anti-mouse phycoerythrin-conjugated secondary antibodies for 30 min at 4° C. Antibody staining was analyzed on a BD Biosciences Accuri C6 flow cytometry system and data were graphed using Accuri C6 software (BD Biosciences).

[0168] Generation of a novel CSPG4 antibody: CSPG4-specific mouse monoclonal antibody 7H5A2 was generated by Promab Biotechnologies Inc (Richmond, CA) by injection of recombinant CSPG4 protein immunogen corresponding to amino acids 1538-2221 of the CSPG4 core protein extracellular domain, expressed from a eukaryotic expression system, and purified (see FIG. 1A). The specificity of the antibody was determined by screening against CSPG4 wild-type and knockout cell lysates by Western blot and cell staining by immunofluorescence (see FIGS. 1B and 1C). 7H5A2 is of isotype IgG1.

[0169] Ovarian cancer patient tissue cohort: The cohort consisted of 126 epithelial ovarian cancer patients with long-term clinical follow-up who underwent initial surgery and treatment at Hunan Cancer Hospital affiliated with Xiangya School of Medicine of Central South University of China, a specialized cancer hospital accredited by Joint Commission International (JCI). Inclusion criteria for the ovarian cancer patient cohort were histologically confirmed epithelial ovarian cancer, including the three major histopathological subtypes (serous, mucinous, and other adenocarcinoma); treatment with platinum / taxane-based chemotherapy after lesion-eliminating surgery; no radiotherapy or biological therapy before surgery; and a Karnofsky Performance Status score of 80 or higher before surgery. Patients were staged according to the International Federation of Gynecology and Obstetrics (FIGO) surgical staging system. Another 16 patients with benign ovarian lesions and 26 hysterectomized patients with normal ovarian tissue were also recruited.

[0170] The protocol was approved by the Ethics Committee of Hunan Cancer Hospital (Changsha, China), and all patients provided informed consent documents on file at the hospital.

[0171] Immunohistochemical analysis: Specimens were paraffin embedded, and tissue sections (4 μm) were dewaxed, rehydrated, blocked with 3% BSA, and subjected to antigen retrieval. After washing, sections were incubated with antibody 9.2.27 (1:1000) against CSPG4 at 4°C overnight. Mouse IgG (Cat. No. A7028, Beyotime, Shanghai, China) was used as a negative control. Bound antibodies were detected using horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG antibody (Beyotime, Cat. No. A0216, China) and visualized by DAB (DAB-2031, Maixin Biotech. Co., Fuzhou, China), followed by counterstaining with hematoxylin (CTS-1090, Maixin Biotech. Co., Fuzhou, China). Results were evaluated under a microscope by two pathologists in a blinded fashion.

[0172] Statistical Analysis: All statistical analyses were performed using Graphpad PRISM 6 (Graphpad Software, San Diego, CA) unless otherwise indicated. Data are graphed as mean ± sem unless otherwise indicated. Comparisons of two independent samples were performed using a two-tailed Student's t test with Welch's correction. Differences in tumor burden over time in xenograft models were analyzed by ordinary two-way ANOVA with Sidak's multiple comparison test. A value of p<0.05 was considered statistically significant. Overall survival and disease-free survival curves for patients with low and high CSPG4 expression in tumor specimens were analyzed using the Kaplan-Meier method and compared using the log-rank test (SPSS 15.0 software, Chicago, IL, USA). Univariate and multivariate analyses were performed using Cox's regression model after adjusting for baseline characteristics. A value of p<0.05 was considered statistically significant.

[0173] [Example 2] CSPG4 is a protein biomarker for poor survival in ovarian cancer patients CSPG4 protein expression was assessed by immunohistochemistry (IHC) in a cohort of 126 ovarian cancer patients (Figure 2). CSPG4 staining intensity and tumor cell positive fraction were scored by a pathologist blinded to patient identity. CSPG4 protein expression was significantly higher (X ≥ 100) in ovarian cancer (39.68%, 50 / 126) compared with benign (12.5%, 2 / 16) or normal ovarian tissue (11.54%, 3 / 26). 2 =11.04, P=0.004) (Table 3).

[0174] [Table 3]

[0175] CSPG4 was detected in a homogenous pattern in tumor cells in contact with tumor-associated stroma, whereas CSPG4 positivity was more heterogeneous in areas distant from the stroma (Figure 2A). Patients with low CSPG4 protein expression had significantly longer progression-free survival (Kaplan-Meier PFS: 22.615 ± 1.754 vs. 16.559 ± 1.940, X 2 =4.316, P=0.038) and improved overall survival (Kaplan-Meier OS: 31.027±1.353 vs. 24.046±2.177, X 2 = 7.366, P = 0.007) (Figures 2B, 2C). The poor prognosis associated with elevated CSPG4 was independent of patient age, tumor subtype (1 or 2), clinical stage (I / II vs. III / IV), differentiation grade, presence of omental or lymphatic metastases, or multiple volumes (see Table 4). Hazard ratio analysis showed that high CSPG4 levels were an independent indicator of poor overall survival in both univariate (HR = 2.33; 95% CI 1.230-4.427; p = 0.009) and multivariate (HR = 2.54; 95% CI 1.255-5.140; p = 0.010) analyses (see Table 5).

[0176] [Table 4] [Table 5]

[0177] [Example 3] CSPG4 promotes EOC tumor expansion, invasion and cisplatin resistance To determine whether CSPG4 expression promotes tumor growth in vivo, an IP xenograft injection model was used. Mice were injected with 2x10 5 Mice were injected with mock-transfected or CSPG4 knockout A2780 cells. Tumor growth, monitored by bioluminescence, was significantly reduced in mice receiving CSPG4 knockout cells (Figures 3A, 3B). By 27 days after injection, CSPG4 expression promoted a growth advantage of nearly an order of magnitude compared to their CSPG4 knockout counterparts (Figures 3A, 3B).

[0178] The functional significance of CSPG4 expression in EOC cells was explored using multiple in vitro correlational analyses of malignant phenotypes. Among 11 cell lines screened by confocal analysis, flow cytometry, and Western blot, three ovarian cell lines with high levels of CSPG4 expression were selected. See Figures 4A-4C. The selected ovarian cancer cell lines (ES-2, HEY, A2780) originated from patients carrying different subtypes of EOC. See Domcke et al., Nat Commun. 4, 2126 (2013). The entire CSPG4 locus was deleted in each of these cell lines using CRISPR / Cas9, and the knockout efficiency was verified using PCR, Western blot, and flow cytometry (see Figures 4A-C).

[0179] Using a standard Matrigel invasion assay, it was determined that loss of CSPG4 expression significantly reduced the invasive phenotype of all three cell lines (Figure 5A-D). Importantly, the invasive phenotype could be rescued by re-expression of CSPG4 in ES-2 CRISPR(CSPG4-deleted) cells (Figure 5C,D).

[0180] Although platinum-based therapy is widely used as a first-line adjuvant therapy for patients with EOC, relapse due to resistant disease is a common complication that reduces overall survival. To investigate whether CSPG4 expression influences cisplatin resistance in EOC, cisplatin sensitivity was assessed in the cell lines using an MTS assay. Loss of CSPG4 led to increased cisplatin sensitivity in all three cell lines, as shown by a 3.6-8.8-fold reduction in IC50 for the knockout cell lines (Figure 5E-G). As observed for the invasive phenotype, re-expression of CSPG4 in the edited cells led to a reversal of cisplatin sensitivity (Figure 5G, purple curve).

[0181] [Example 4] CSPG4 expression enhances cell adhesion, promotes spheroid formation and FAK activation Gene Ontology (GO) term enrichment analysis of the set of differentially expressed genes in ES-2 control and CSPG4 knockout cell lines identifies significant enrichment (adjusted p-value < 0.005) for GO terms associated with regulation of cell adhesion, associated signaling pathways, and extracellular matrix collagen / organization (Figure 6). These associations support previous studies functionally linking CSPG4 to activation of integrins and downstream integrin-stimulated pathways such as focal adhesion kinase activation. Increased tumor cell adhesion may affect malignant progression in several distinct but overlapping pathways. It may promote increased anchorage independence and / or enhance the formation of cell aggregates (spheroids) in the peritoneal cavity. These spheroids, originating from individual cells that have acquired an anchorage-independent phenotype, form metastases by adhering to the mesothelial lining surface and invading the submesothelial tissue.

[0182] As expected, loss of CSPG4 expression in EOC cells inhibited anchorage-independent growth in agarose and spheroid formation by cells cultured in methylcellulose (Figure 7A, B, and C). Consistent with GO term enrichment analysis, CSPG4 expression promoted activation of FAK, which was completely inhibited in CSPG4 gene deletion counterparts (Figure 7D). CSPG4 rescue of gene-edited cells resulted in restoration of FAK activation in gene-edited cells (Figure 7E). Although CSPG4 loss inhibited spheroid formation, there was no significant effect on cell survival in CSPG4 null spheroids (Figure 7F), indicating that the primary effect of inhibiting CSPG4 expression is on cell proliferation within spheroids. These data suggest that CSPG4-mediated activation of cell adhesion-dependent pathways may enhance metastasis by stimulating invasion and spheroid formation in EOC.

[0183] [Example 5] CSPG4 expression is associated with mesenchymal transition in ovarian cancer cells Gene set enrichment analysis (GSEA) of these RNAseq data showed that loss of CSPG4 affected the differential expression of several genes within the hallmark epithelial-mesenchymal transition genes from the Molecular Signatures Database v 7.1 (gsea-msigdb.org / gsea / msigdb / index.jsp) (Figure 8A). Analysis of a large array of gene expression data from ovarian cancer TCGA also showed that high CSPG4 expression was associated with an EMT signature, leading us to conclude that high CSPG4 is predictive of a more mesenchymal phenotype in EOC tumors (Figure 8B).

[0184] To explore this further, the impact of CSPG4 expression was assessed with respect to several mesenchymal / epithelial markers in three EOC cell lines (Figure 9A). CRISPR / Cas9-induced loss of CSPG4 expression reduced expression of multiple mesenchymal biomarkers, including vimentin and the mesenchymal transcription factor SNAI2, ZEB1, while promoting increased expression of the epithelial biomarker, claudin-1 (Figure 9A). We focused on the relationship between CSPG4 and ZEB1 levels, as ZEB1 expression has previously been associated with the development of a mesenchymal phenotype and poor outcome in EOC patients. Inhibition of ZEB1 expression using RNAi (Figure 9B) had no inhibitory effect on CSPG4 levels, suggesting that ZEB1 expression is downstream of CSPG4. Furthermore, CSPG4-stimulated ZEB1 expression is mediated by FAK activation, as ZEB1 expression is inhibited by PND-1186, a small molecule inhibitor of FAK activation (Figure 9C). Inhibition of ZEB1 restricted both spheroid formation (Figure 9D) and invasion (Figure 9E), consistent with studies showing that spheroid formation is associated with mesenchymal transition. However, restricting ZEB1 expression had no detectable effect on cisplatin sensitivity (Figures 9F and 9G), indicating that some of the pro-tumorigenic effects of CSPG4 are independent of ZEB1 transcriptional activation. Aggregated survival analysis of multiple ovarian cancer patient datasets showed that elevated transcriptional expression of both CSPG4 and ZEB1, either individually or in combination, was significantly associated with decreased overall survival at 5 years (Figures 9H-J), supporting the conclusion that both biomarkers may function in concert to promote recurrence and relapse.

[0185] [Example 6] Anti-CSPG4 antibody reduces EOC invasion and spheroid formation by inhibiting the FAK-ZEB1 pathway activated by CSPG4 Although multiple structural / functional domains in the extracellular portion of the CSPG4 core protein have been identified, many of these sites (e.g., collagen and growth factor binding sites, chondroitin sulfate attachment site, laminin G domain) map to the membrane-distal region of the core protein. See, e.g., Price et al., Pigment Cell Melanoma Res. 24(6):1148-57 (2011); Tamburini et al., FASEB J. 33(3):3112-28 (2019). In contrast, less is known about the potential functional importance of domains in the core protein that are membrane-proximal. To investigate the importance of this juxtamembrane protein region (DE) in CSPG4 on EOC cells, a recombinant fragment of the CSPG4 core protein (Q1538-N2221, see Figure 1A) containing this region was purified from a eukaryotic expression system and used in the production of mouse monoclonal antibodies. An antibody clone (7H5A2) was identified that specifically recognizes CSPG4 on the cell surface (see Figures 1B and 1C). The antibody antagonized CSPG4 function and inhibited activation of FAK and ZEB1 expression in all three cell lines, as well as CSPG4 deletion (Figure 10A). The antibody also significantly inhibited EOC invasion to a greater extent than the anti-CSPG4 antibody 763.74 (Figure 10B) (see, e.g., Luo et al., J Immunol 2006; 176:6046-54). Antibody 763.74 has previously been shown to affect CSPG4-mediated invasion of other tumor types, but in contrast to 7H5A2, it recognizes a different membrane-distal epitope on the core protein. The 7H5A2 antibody also significantly inhibited spheroid formation and EOC cell survival of mock-transfected EOC cells expressing CSPG4 (Figure 10C). Gene-editing CSPG4 expression also inhibited spheroid formation, but CRISPR / Cas9-edited cells largely remained viable at the conclusion of methylcellulose culture (Figure 7E), indicating that the 7H5A2 antibody has an additional cytotoxic mechanism independent of its effect on CSPG4-stimulated tumor cell proliferation.

[0186] [Example 7] Anti-CSPG4 antibodies raised against the D3 domain of the CSPG core protein stimulate antibody-dependent cell-mediated cytotoxicity (ADCC) of EOC cells An additional anti-CSPG4 antibody, 8G5A5, was identified using the methods described in Example 6. Preliminary screening was performed using NK cells expressing wild-type CD16a or a mutated hyperactive form of CD16a / S197P that enhances ADCC by resisting proteolytic cleavage after NK cell activation. HEY mock and CRISPR cells were treated with the indicated anti-CSPG4 monoclonal antibodies or normal mouse IgG1 (nmIgG1). The indicated NK92 cell lines were added at a 1:1 effector:target ratio and ADCC was determined at 4 hours using the DELFIA EuTDA cytotoxicity assay according to the manufacturer's protocol. The results show that antibody 8G5A6 can mediate specific ADCC of CSPG4-positive EOC cells mediated by NK cells expressing wild-type CD16a or CD16a / S197P (Figure 11A). The experiment was repeated and the results are shown in Figure 11B. Experiments were also performed with OVCAR8 ovarian cancer cells, a model of the most common high-grade serous ovarian cancer clinically in women (Figure 11C). Again, antibody 8G5A6 was able to mediate specific ADCC of CSPG4-positive cells.

[0187] Taken together, IHC data from this EOC patient cohort demonstrate that high levels of CSPG4 in tumors are an independent risk factor for poor overall survival. CSPG4 expression promotes tumor expansion in vivo and tumor cell invasion, cisplatin resistance and spheroid formation in vitro. CSPG4 does not signal directly by itself, but functions as a co-receptor / plasma membrane scaffold that enhances the strength and duration of multiple stimulated tumorigenic pathways. As we have shown that CSPG4-mediated prolonged activation of Erk causes a shift towards a mesenchymal transcriptome in melanoma cells, promoting their tumorigenic potential, longer duration of signaling activation may result in nuclear changes that affect the transcriptome. Thus, localized elevation of CSPG4 levels in a subpopulation of EOC cells may sustain a tumor cell subpopulation with enhanced tumorigenic signaling that leads to increased proliferation, survival and / or invasive potential. This is consistent with in vivo tumor growth data that associate CSPG4 expression with significantly enhanced tumor expansion compared to CRISPR / Cas9-deleted counterparts. It is also important to note that although high CSPG4 expression levels in EOC tumors negatively impact patient survival, the staining pattern in these tumors is heterogeneous. Because CSPG4 expression is stimulated by hypoxic microenvironmental changes or inflammatory mediators such as TNFα, heterogeneous CSPG4 expression may be associated with these or additional microenvironmental factors in expanding tumors.

[0188] One consequence of CSPG4 expression is that it stimulates a mesenchymal shift in the phenotype of EOC cells, which is associated with spheroid formation and subsequent intraperitoneal metastasis to other organs such as the omentum. More globally, TCGA gene expression data show that elevated CSPG4 levels are associated with an EMT signature, and the present data from multiple EOC cell lines link CSPG4 to expression of ZEB1, a mesenchymal transcription factor, in EOC and other tumors. As a transcriptional regulator, ZEB1 represses the expression of multiple epithelial genes, whereas it stimulates the expression of genes associated with an invasive mesenchymal phenotype. Because TCGA analysis shows that co-expression of CSPG4 and ZEB1 is associated with poor overall survival, the two may function in concert to promote metastasis, as evidenced by their effects on spheroid formation and invasion. Furthermore, tumor cells expressing CSPG4 are also more resistant to cisplatin, suggesting that tumor cells expressing CSPG4 may form a therapy-resistant tumor cell reservoir that promotes relapse after initial standard treatment.

[0189] CSPG4 promotes spheroid formation and invasion by activating FAK and enhancing ZEB1 expression. This is consistent with data from other cell model systems (including fibroblasts isolated from FAK null animals) that have linked FAK activation to ZEB1 expression. As described herein, well-characterized inhibitors of FAK activation limit ZEB1 expression, and the anti-CSPG4 specific antibodies described herein inhibit FAK activation, ZEB1 expression, and tumor cell invasion / spheroid formation. These data are consistent with reports linking CSPG4 function in tumor cells to functionally activating β1 integrin and FAK. Thus, these data support a model in which cell surface CSPG4 interacting with components of the microenvironment (specific ECM components or various growth factors) can enhance the mesenchymal transition of EOC cells.

[0190] However, while limiting CSPG4 expression causes a significant reduction in platinum IC50, specific inhibition of ZEB1 expression has no effect on these values. This leads us to conclude that CSPG4 enhances cisplatin sensitivity by a ZEB1-independent mechanism. As a multifunctional transmembrane signaling node, CSPG4 functions to alter activation by multiple extracellular stimuli (e.g., TGFβ, FGF, HGF), and depending on the cellular context, it can activate multiple oncogenic pathways (e.g., FAK, MAPK, PI3K, NF-κB) in tumor cells. Because reduced cisplatin sensitivity in standard treatments may affect the survival of resistant clones of EOC tumor cells, it is important to further define the mechanism by which CSPG4 alters the response to this treatment. One approach is to rescue CSPG4-null cells using several well-defined CSPG4 structural mutants to identify domains that fail to reverse the loss of cisplatin sensitivity. This approach may result in enhanced targeting by identifying CSPG4 domains that limit cisplatin sensitivity through mechanisms that are consistent with or independent of regulation of ZEB1 expression.

[0191] The present data show that CSPG4 can directly reduce tumor cell sensitivity to cisplatin, as described herein, but in the larger context of tumor tissue, CSPG4 may also affect poor outcomes in EOC patients by contributing to cell adhesion-related mechanisms associated with environmentally mediated drug resistance (EMDR). The concept underlying EMDR is that adherent tumor cell subpopulations that are initially resistant to treatment may form a reservoir of resistant cells that can undergo further mutations that contribute to treatment-resistant flare-ups following standard treatment. This is similar to, but distinct from, the hypothesis that treatment-resistant cancer-initiating stem cells are responsible for treatment failure. Some cell adhesion-related mechanisms (e.g., regulated by integrin and growth factor / cytokine-mediated pathways) may function to promote survival in the absence of transcriptomic profiles that regulate cancer-initiating stem cells. The mesenchymal shift in EOC, induced by factors such as TGF-β, is associated with a collagen remodeling fibrotic gene signature that correlates with metastasis and poor overall survival. The fibrotic signature associated with mesenchymal EOC includes elevated collagen VI, the major ECM ligand for CSPG4, and elevated collagen VI in the tumor parenchyma, which is associated with reduced survival in EOC patients. These studies demonstrated that EOC cells adhering on collagen VI-coated surfaces exhibit increased resistance to cisplatin in vitro. The potential clinical impact is that localized CSPG4 / ECM interactions may trigger the formation of a therapy-resistant adhesive "niche" consisting of deeply embedded EOC populations that may escape detection after standard-of-care surgical lesion removal.

[0192] Targeting CSPG4 with an antibody that binds to the juxtamembrane region of the CSPG4 core protein effectively inhibits ZEB1 expression in spheroids, limiting CSPG4-mediated invasion and promoting apoptosis of EOC cells. Thus, targeting this region of CSPG4 can be used to limit metastasis in patients with EOC, thus improving patient outcomes.

[0193] [Example 8] Exemplary Anti-CSPG4 Antibodies This example shows the amino acid sequences of the heavy and light (kappa) variable domains of the 7H5A2 (murine) antibody. The CDRs are shown in bold and underlined text within the variable domains. The sequences of the respective CDRs and framework regions are also provided, as well as the nucleotide sequences encoding each of the heavy and light chains.

[0194] [Table 6] TIFF2024546096000007.tif110156

[0195] [Example 9] Exemplary Anti-CSPG4 Antibodies This example shows the amino acid sequences of the heavy and light chains (kappa) of the 8G5A6 (murine) antibody. The CDRs are shown in bold and underlined text within the variable domains. The sequences of the respective CDRs and framework regions are also provided, as well as the nucleotide sequences encoding each of the heavy and light chains.

[0196] [Table 7] TIFF2024546096000009.tif153156

[0197] [Example 10] CSPG4 TriKE and Cancer This example shows that CSPG4 TriKE induces NK-mediated killing of CSPG4-positive cancer cells.

[0198] method Peripheral blood mononuclear cell donors and enrichment of natural killer cells Fresh whole blood was obtained from normal volunteers. Peripheral blood mononuclear cells (PBMCs) were isolated using density gradient centrifugation with density gradient medium Lymphoprep™ (STEMCELL Technologies, Cambridge, MA, USA). Natural killer (NK) cells were enriched by magnetic depletion of CD3 and CD19 positive cells using EasySep™ Human NK Cell Enrichment kit (STEMCELL Technologies, Cambridge, MA, USA) according to the manufacturer's recommendations.

[0199] Tumor spheroid killing assay Tumor spheroid killing assays were evaluated in real time using the IncuCyte SX5-Live Cell Analysis platform. 20,000 GFP-expressing ovarian cancer target cells (OVCAR-8 or SKOV3) were seeded into wells of a round-bottom ultra-low attachment 96-well plate (Corning, UK) and allowed to form spheroids for 2 days. 40,000 NK cells magnetically enriched from fresh PBMCs were added to triplicate wells with or without 30 nM CSPG4 TriKE (8G5A6 or 7H5A2) or 3 nM IL-15, and plates were incubated for 4 days at 37°C / 5% CO2 in an IncuCyte SX5. Images from triplicate technical replicate wells for each condition were taken every 2 hours for 96 hours using a 10x objective and then analyzed using IncuCyte™ Basic Software v2018A (Sartorious). Graphed readouts represented target spheroid green fluorescence intensity normalized to untreated spheroid control wells at each time point.

[0200] cell line Ovarian cancer cell lines OVCAR-8, SKOV3, OVCAR-3, A2780, MA-148, and OVCAR-5 were cultured in DMEM medium (Mediatech, Catalog No. 10-013-cv). HEY cells were cultured in RPMI1640 medium (Gibco, Catalog No. 11875-093). All culture media were supplemented with 10% fetal bovine serum (Atlanta Biologicals, Catalog No. SS11150H, Lot No. H1810S) and 1% penicillin / streptomycin (Gibco, Catalog No. 15140-122) at 37°C / 5% CO2.

[0201] Cell lysis and Western blot The indicated ovarian cancer cell lines were cultured in 6-well plates (1.0x10 5 ). Cells were lysed in cell lysis buffer (Cell Signaling, Danvers, MA) and 20 μg protein / sample was fractionated on 4% / 7.5% SDS-PAGE and transferred to PVDF membranes for Western blot analysis using standard techniques. Membranes were probed with anti-CSPG4 antibody 9.2.27 (Millipore) and anti-a-tubulin antibody (Millipore) and appropriate HEP-conjugated secondary reagents. Bands were visualized by incubation with Pierce ECL Western Blotting Substrate (Thermo Fisher Scientific).

[0202] result CSPG4 expression was determined in human ovarian cancer cell lines (Figure 12), and CSPG4-positive and CSPG4-negative cells were treated with NK cells alone, IL-15, or NK cells activated with CSPG4 TriKE (CSPG4 TriKE 8G5A6 or CSPG4 TriKE 7H5A2). Compared with NK cells alone or NK cells activated with IL-15 alone (Figure 13A), TriKEs based on anti-CSPG4 monoclonal antibodies 7H5A2 or 8G5A6 promoted specific NK cell-mediated killing of spheroids composed of human ovarian cancer cells (OVCAR-8) expressing CSPG4. In contrast, these TriKEs were not as effective in promoting killing of CSPG4-negative ovarian cancer cells compared to that observed by NK cells cultured with IL-15 (Figure 13B).

[0203] These results indicate that CSPG4 TRiKE can specifically target CSPG4-expressing ovarian cancer.

[0204] [Example 11] Exemplary embodiments Embodiment 1. (i) a heavy chain variable domain or region comprising the amino acid sequences set forth in SEQ ID NO:1 (or SEQ ID NO:1 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO:2 (or SEQ ID NO:2 with one, two or three amino acid additions, deletions or substitutions) and SEQ ID NO:3 (or SEQ ID NO:3 with one amino acid addition, deletion or substitution), and a light chain variable domain or region comprising the amino acid sequences set forth in SEQ ID NO:9 (or SEQ ID NO:9 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO:10 (or SEQ ID NO:10 with one, two or three amino acid additions, deletions or substitutions) and SEQ ID NO:11 (or SEQ ID NO:11 with one, two or three amino acid additions, deletions or substitutions); or (ii) a heavy chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO: 17 (or SEQ ID NO: 17 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO: 18 (or SEQ ID NO: 18 with one, two or three amino acid additions, deletions or substitutions) and SEQ ID NO: 19 (or SEQ ID NO: 19 with one, two or three amino acid additions, deletions or substitutions), and a light chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO: 25 (or SEQ ID NO: 25 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO: 26 (or SEQ ID NO: 26 with one, two or three amino acid additions, deletions or substitutions) and SEQ ID NO: 27 (or SEQ ID NO: 27 with one, two or three amino acid additions, deletions or substitutions). 4. An antibody comprising:

[0205] Embodiment 2. The antibody of embodiment 1, having the ability to bind to human CSPG4 polypeptide (SEQ ID NO: 33).

[0206] Embodiment 3. The antibody of any one of embodiments 1 to 2, comprising the heavy chain variable domain or region of (i).

[0207] Embodiment 4. The antibody of embodiment 3, wherein the heavy chain variable domain or region comprises an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:8.

[0208] Embodiment 5. The antibody of any one of embodiments 1 to 2, comprising the light chain variable domain or region of (i).

[0209] Embodiment 6. The antibody of embodiment 5, wherein the light chain variable domain or region comprises an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:16.

[0210] Embodiment 7. The antibody of any one of embodiments 1 to 2, comprising the heavy chain variable domain or region of (ii).

[0211] Embodiment 8. The antibody of embodiment 7, wherein the heavy chain variable domain or region comprises an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:24.

[0212] Embodiment 9. The antibody of any one of embodiments 1 to 2, comprising the light chain variable domain or region of (ii).

[0213] Embodiment 10. The antibody of embodiment 9, wherein the light chain variable domain or region comprises an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:32.

[0214] Embodiment 11. (i) a heavy chain variable domain or region comprising the amino acid sequences set forth in SEQ ID NO:1 (or SEQ ID NO:1 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO:2 (or SEQ ID NO:2 with one, two or three amino acid additions, deletions or substitutions) and SEQ ID NO:3 (or SEQ ID NO:3 with one amino acid addition, deletion or substitution), and a light chain variable domain or region comprising the amino acid sequences set forth in SEQ ID NO:9 (or SEQ ID NO:9 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO:10 (or SEQ ID NO:10 with one, two or three amino acid additions, deletions or substitutions) and SEQ ID NO:11 (or SEQ ID NO:11 with one, two or three amino acid additions, deletions or substitutions); or (ii) a heavy chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO: 17 (or SEQ ID NO: 17 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO: 18 (or SEQ ID NO: 18 with one, two or three amino acid additions, deletions or substitutions) and SEQ ID NO: 19 (or SEQ ID NO: 19 with one, two or three amino acid additions, deletions or substitutions), and a light chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO: 25 (or SEQ ID NO: 25 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO: 26 (or SEQ ID NO: 26 with one, two or three amino acid additions, deletions or substitutions) and SEQ ID NO: 27 (or SEQ ID NO: 27 with one, two or three amino acid additions, deletions or substitutions). An antigen-binding fragment comprising:

[0215] Embodiment 12. The antigen-binding fragment of embodiment 11, comprising the ability to bind to SEQ ID NO:33 or SEQ ID NO:34.

[0216] Embodiment 13. The antigen-binding fragment of any one of embodiments 11 to 12, comprising the heavy chain variable domain or region of (i).

[0217] Embodiment 14. The antigen-binding fragment of embodiment 13, wherein the heavy chain variable domain or region comprises an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:8.

[0218] Embodiment 15. The antigen-binding fragment of any one of embodiments 11 to 12, comprising the light chain variable domain or region of (i).

[0219] Embodiment 16. The antigen-binding fragment of embodiment 15, wherein the light chain variable domain or region comprises an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:16.

[0220] Embodiment 17. The antigen-binding fragment of any one of embodiments 11 to 12, comprising the heavy chain variable domain or region of (ii).

[0221] Embodiment 18. The antigen-binding fragment of embodiment 17, wherein the heavy chain variable domain or region comprises an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:24.

[0222] Embodiment 19. The antigen-binding fragment of any one of embodiments 11 to 12, comprising the light chain variable domain or region of (ii).

[0223] Embodiment 20. The antigen-binding fragment of embodiment 19, wherein the light chain variable domain or region comprises an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:32.

[0224] Embodiment 21. The antibody of any one of embodiments 1 to 10, which is a monoclonal antibody.

[0225] Embodiment 22 The antibody of any one of embodiments 1 to 10 and 21, which is an scFv antibody.

[0226] Embodiment 23. The antigen-binding fragment of any one of embodiments 11 to 20, which is monoclonal.

[0227] Embodiment 24. The antigen-binding fragment of any one of embodiments 11 to 20 and 23, which is a Fab.

[0228] Embodiment 25. A chimeric antigen receptor comprising an antigen-binding domain, a hinge, a transmembrane domain, and one or more signaling domains, wherein the antigen-binding domain comprises an antibody or antigen-binding fragment of any one of claims 1 to 24.

[0229] Embodiment 26 The chimeric antigen receptor of embodiment 25, wherein the antigen-binding domain comprises an scFv having the ability of binding to a CSPG4 polypeptide.

[0230] Embodiment 27. A cell comprising the chimeric antigen receptor of any one of embodiments 25 to 26.

[0231] Embodiment 28 The cell of embodiment 27, which is a T cell, a stem cell or a NK cell.

[0232] Embodiment 29. A cell engager comprising a first antigen-binding domain, a linker, and a second antigen-binding domain, wherein the first antigen-binding domain comprises the antibody or antigen-binding fragment of any one of embodiments 1 to 24.

[0233] Embodiment 30 The cell engager of embodiment 29, wherein the first antigen-binding domain comprises an scFv capable of binding to a CSPG4 polypeptide.

[0234] Embodiment 31 The cell engager of embodiment 29, wherein the first antigen-binding domain is an IgG having the ability of binding to a CSPG4 polypeptide.

[0235] Embodiment 32 The cell engager of any one of embodiments 29-31, wherein the second antigen-binding domain binds to a polypeptide expressed on the surface of a T cell.

[0236] Embodiment 33 The cell engager of embodiment 32, wherein said polypeptide expressed on the surface of the T cell is a CD3 polypeptide.

[0237] Embodiment 34 The cell engager of any one of embodiments 29-31, wherein the second antigen-binding domain binds to a polypeptide expressed on the surface of a NK cell.

[0238] Embodiment 35 The cell engager of embodiment 34, wherein said polypeptide expressed on the surface of the NK cell is a CD16a polypeptide.

[0239] Embodiment 36 The cell engager of any one of embodiments 29 to 35, comprising a third antigen-binding domain.

[0240] Embodiment 37 The cell engager of embodiment 36, wherein the third antigen-binding domain binds to a polypeptide expressed on the surface of a NK cell.

[0241] Embodiment 38 The cell engager of embodiment 37, wherein said polypeptide expressed on the surface of the NK cell is a CD16a polypeptide.

[0242] Embodiment 39. A nucleic acid comprising a nucleic acid sequence encoding at least a portion of the antibody or antigen-binding fragment of any one of embodiments 1 to 24.

[0243] Embodiment 40. The nucleic acid of embodiment 39, wherein the nucleic acid sequence encodes the heavy chain variable domain or region of any one of (i) to (ii) of embodiment 1.

[0244] Embodiment 41. The nucleic acid of any one of embodiments 39 to 40, wherein the nucleic acid sequence encodes the light chain variable domain or region of any one of (i) to (ii) of embodiment 1.

[0245] Embodiment 42. The nucleic acid of any one of embodiments 39 to 41, which is a viral vector.

[0246] Embodiment 43. The nucleic acid of any one of embodiments 39 to 41, which is a phagemid.

[0247] Embodiment 44. A nucleic acid comprising a nucleic acid sequence encoding the chimeric antigen receptor of any one of embodiments 25 to 26 or the cell engager of any one of embodiments 29 to 38.

[0248] Embodiment 45. The nucleic acid of embodiment 44, which is a viral vector.

[0249] Embodiment 46 The nucleic acid of embodiment 44, which is a phagemid.

[0250] Embodiment 47. A host cell comprising the nucleic acid of any one of embodiments 39 to 46.

[0251] Embodiment 48. A host cell expressing the chimeric antigen receptor of any one of embodiments 25 to 26 or the cell engager of any one of embodiments 29 to 38.

[0252] Embodiment 49. The host cell of any one of embodiments 47 to 48, which is a T cell, a stem cell or a NK cell.

[0253] Embodiment 50. An antibody drug conjugate (ADC) comprising an antigen-binding domain covalently linked to a drug, wherein the antigen-binding domain comprises the antibody or antigen-binding fragment of any one of embodiments 1 to 24.

[0254] Embodiment 51 The ADC of embodiment 50, wherein the antigen-binding domain comprises an scFv capable of binding to a CSPG4 polypeptide.

[0255] Embodiment 52 The ADC of embodiment 50, wherein the antigen-binding domain is an IgG having the ability to bind to a CSPG4 polypeptide.

[0256] Embodiment 53. The ADC of any one of embodiments 50-52, wherein the drug is selected from the group consisting of calicheamicin, monomethylauristatin E (MMAE), emtansine (DM1), and exatecan derivative (Dxd).

[0257] Embodiment 54. A composition comprising the antibody or antigen-binding fragment of any one of embodiments 1 to 24.

[0258] Embodiment 55. The composition of claim 54 comprising the antibody of any one of embodiments 1 to 10, 21 and 22.

[0259] Embodiment 56. The composition of claim 54, comprising the antigen-binding fragment of any one of embodiments 11 to 20, 23 and 24.

[0260] Embodiment 57. A composition comprising a cell engager of any one of embodiments 29 to 38.

[0261] Embodiment 58. A composition comprising any one of the cells of embodiments 27, 28 and 47 to 49.

[0262] Embodiment 59. A composition comprising the ADC of any one of embodiments 50-53.

[0263] Embodiment 60. The composition of any one of embodiments 54 to 59, comprising a checkpoint inhibitor.

[0264] Embodiment 61. The composition of embodiment 60, wherein the checkpoint inhibitor is selected from the group consisting of cemiplimab, nivolumab, pembrolizumab, JTX-4014, spartalizumab, camrelizumab, sintilimab, tislelizumab, toripalimab, dostallimab, INCMGA00012, AMP-224, AMP-514, avelumab, durvalumab, atezolizumab, KN035, CK-301, AUNP12, CA-170, BMS-986189 and ipilimumab.

[0265] Embodiment 62. A method of treating a mammal having cancer, comprising administering to said mammal a composition of any one of embodiments 54 to 61.

[0266] Embodiment 63 The method of embodiment 62, wherein the mammal is a human.

[0267] Embodiment 64. The cancer is CSPG4 + The method of embodiment 62 or embodiment 63, wherein the treatment is cancer.

[0268] Embodiment 65. The CSPG4 + Cancer, CSPG4 + 72. The method of embodiment 71, wherein the cancer is ovarian cancer.

[0269] Embodiment 66 The method of any one of embodiments 62-65, wherein the number of cancer cells in the mammal is reduced after the administering step.

[0270] Embodiment 67. A method for binding a binding molecule to a CSPG4 polypeptide, comprising contacting the CSPG4 polypeptide with the antibody or antigen-binding fragment of any one of embodiments 1 to 24.

[0271] Embodiment 68 The method of embodiment 67, wherein said contacting step is carried out in vitro.

[0272] Embodiment 69 The method of embodiment 67, wherein said contacting step is performed in vivo.

[0273] Embodiment 70 The method of embodiment 67, wherein the contacting step is performed in a mammal by administering the antibody or antigen-binding fragment to the mammal.

[0274] Embodiment 71 The method of embodiment 70, wherein the mammal is a human.

[0275] Embodiment 72. A method for binding a binding molecule to a CSPG4 polypeptide, comprising the step of contacting the CSPG4 polypeptide with a chimeric antigen receptor of any one of embodiments 25-26, a cell engager of any one of embodiments 29-38, or an ADC of any one of embodiments 50-53.

[0276] Embodiment 73 The method of embodiment 72, wherein the contacting step is carried out in vitro.

[0277] Embodiment 74 The method of embodiment 72, wherein the contacting step is performed in vivo.

[0278] Embodiment 75. The method of embodiment 72, wherein said contacting step is performed in a mammal by administering said chimeric antigen receptor, said cell engager or said ADC to the mammal.

[0279] Embodiment 76 The method of embodiment 75, wherein the mammal is a human.

[0280] Other embodiments While the present invention has been described in conjunction with its detailed description, it should be understood that the foregoing description is for illustrative purposes only and is not intended to limit the scope of the invention as defined by the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.

Claims

1. (i) a heavy chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO: 1 (or SEQ ID NO: 1 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO: 2 (or SEQ ID NO: 2 with one, two or three amino acid additions, deletions or substitutions), and SEQ ID NO: 3 (or SEQ ID NO: 3 with one amino acid addition, deletion or substitution), and a light chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO: 9 (or SEQ ID NO: 9 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO: 10 (or SEQ ID NO: 10 with one, two or three amino acid additions, deletions or substitutions), and SEQ ID NO: 11 (or SEQ ID NO: 11 with one, two or three amino acid additions, deletions or substitutions); or (ii) a heavy chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO: 17 (or SEQ ID NO: 17 with one, two, or three amino acid additions, deletions, or substitutions), SEQ ID NO: 18 (or SEQ ID NO: 18 with one, two, or three amino acid additions, deletions, or substitutions), and SEQ ID NO: 19 (or SEQ ID NO: 19 with one, two, or three amino acid additions, deletions, or substitutions), and a light chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO: 25 (or SEQ ID NO: 25 with one, two, or three amino acid additions, deletions, or substitutions), SEQ ID NO: 26 (or SEQ ID NO: 26 with one, two, or three amino acid additions, deletions, or substitutions), and SEQ ID NO: 27 (or SEQ ID NO: 27 with one, two, or three amino acid additions, deletions, or substitutions). An antibody comprising:

2. (i) a heavy chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO: 1 (or SEQ ID NO: 1 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO: 2 (or SEQ ID NO: 2 with one, two or three amino acid additions, deletions or substitutions), and SEQ ID NO: 3 (or SEQ ID NO: 3 with one amino acid addition, deletion or substitution), and a light chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO: 9 (or SEQ ID NO: 9 with one, two or three amino acid additions, deletions or substitutions), SEQ ID NO: 10 (or SEQ ID NO: 10 with one, two or three amino acid additions, deletions or substitutions), and SEQ ID NO: 11 (or SEQ ID NO: 11 with one, two or three amino acid additions, deletions or substitutions); or (ii) a heavy chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO: 17 (or SEQ ID NO: 17 with one, two, or three amino acid additions, deletions, or substitutions), SEQ ID NO: 18 (or SEQ ID NO: 18 with one, two, or three amino acid additions, deletions, or substitutions), and SEQ ID NO: 19 (or SEQ ID NO: 19 with one, two, or three amino acid additions, deletions, or substitutions), and a light chain variable domain or region comprising the amino acid sequence set forth in SEQ ID NO: 25 (or SEQ ID NO: 25 with one, two, or three amino acid additions, deletions, or substitutions), SEQ ID NO: 26 (or SEQ ID NO: 26 with one, two, or three amino acid additions, deletions, or substitutions), and SEQ ID NO: 27 (or SEQ ID NO: 27 with one, two, or three amino acid additions, deletions, or substitutions). An antigen-binding fragment comprising:

3. An antibody described in claim 1 or an antigen-binding fragment described in claim 2, which has the ability to bind to a CSPG4 polypeptide.

4. The antibody or antigen-binding fragment of claim 1 or claim 2, wherein the antibody or antigen-binding fragment comprises the heavy chain variable domain or region of (i).

5. The antibody or antigen-binding fragment of claim 4, wherein the heavy chain variable domain or region comprises an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:

8.

6. The antibody or antigen-binding fragment of claim 1 or claim 2, wherein the antibody or antigen-binding fragment comprises the light chain variable domain or region of (i).

7. The antibody or antigen-binding fragment of claim 6, wherein the light chain variable domain or region comprises an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:

16.

8. The antibody or antigen-binding fragment of claim 1 or claim 2, wherein the antibody or antigen-binding fragment comprises the heavy chain variable domain or region of (ii).

9. 9. The antibody or antigen-binding fragment of claim 8, wherein the heavy chain variable domain or region comprises an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:

24.

10. The antibody or antigen-binding fragment of claim 1 or claim 2, wherein the antibody or antigen-binding fragment comprises the light chain variable domain or region of (ii).

11. 11. The antibody or antigen-binding fragment of claim 10, wherein the light chain variable domain or region comprises an amino acid sequence having at least 90 percent identity to the amino acid sequence set forth in SEQ ID NO:

32.

12. The antibody of claim 1 , wherein the antibody is a monoclonal antibody or an scFv antibody.

13. The antigen-binding fragment of claim 2, wherein the antigen-binding fragment is monoclonal or is a Fab.

14. A chimeric antigen receptor comprising an antigen-binding domain, a hinge, a transmembrane domain, and one or more signaling domains, wherein the antigen-binding domain comprises the antibody of claim 1 or the antigen-binding fragment of claim 2.

15. A cell, T cell, stem cell or NK cell comprising the chimeric antigen receptor of claim 14.

16. A cell engager comprising a first antigen-binding domain, a linker, and a second antigen-binding domain, wherein the first antigen-binding domain comprises the antibody of claim 1 or the antigen-binding fragment of claim 2.

17. The cell engager of claim 16, wherein the second antigen-binding domain binds to a polypeptide expressed on the surface of a T cell.

18. 18. The cell engager of claim 17, wherein the polypeptide expressed on the surface of a T cell is a CD3 polypeptide.

19. 18. The cell engager of claim 17, wherein the second antigen-binding domain binds to a polypeptide expressed on the surface of an NK cell.

20. 20. The cell engager of claim 19, wherein the polypeptide expressed on the surface of NK cells is a CD16a polypeptide.

21. 18. The cell engager of claim 17, comprising a third antigen-binding domain.

22. 22. The cell engager of claim 21, wherein the third antigen-binding domain binds to a polypeptide expressed on the surface of an NK cell.

23. 23. The cell engager of claim 22, wherein the polypeptide expressed on the surface of NK cells is a CD16a polypeptide.

24. A nucleic acid comprising a nucleic acid sequence encoding at least a portion of the antibody of claim 1 or the antigen-binding fragment of claim 2.

25. A nucleic acid comprising a nucleic acid sequence encoding the chimeric antigen receptor of claim 14.

26. 25. The nucleic acid of claim 24, wherein the nucleic acid is a viral vector or a phagemid.

27. 25. A host cell, T cell, stem cell or NK cell comprising the nucleic acid of claim 24.

28. A host cell expressing the chimeric antigen receptor of claim 14.

29. 1. An antibody drug conjugate (ADC) comprising an antigen-binding domain covalently linked to a drug, wherein the antigen-binding domain comprises the antibody of claim 1 or the antigen-binding fragment of claim 2.

30. 30. The ADC of claim 29, wherein the drug is selected from the group consisting of calicheamicin, monomethyl auristatin E (MMAE), emtansine (DM1), and an exatecan derivative (Dxd).

31. A composition comprising the antibody of claim 1 or the antigen-binding fragment of claim 2.

32. 20. A composition comprising the cell engager of claim 17.

33. A composition comprising the cells of claim 15.

34. 30. A composition comprising the ADC of claim 29.

35. 32. The composition of claim 31, comprising a checkpoint inhibitor.

36. 36. The composition of claim 35, wherein the checkpoint inhibitor is selected from the group consisting of cemiplimab, nivolumab, pembrolizumab, JTX-4014, spartalizumab, camrelizumab, sintilimab, tislelizumab, toripalimab, dostallimab, INCMGA00012, AMP-224, AMP-514, avelumab, durvalumab, atezolizumab, KN035, CK-301, AUNP12, CA-170, BMS-986189, and ipilimumab.

37. The composition of claim 31 for treating a mammal or human having cancer, CSPG4+ cancer, or CSPG4+ ovarian cancer.

Images