WO2025177185A1 - Materials and methods for biologic affinity analysis - Google Patents

Abstract

A method for determining the affinity of an antibody to a cell-surface antigen. The method generally includes incubating the antibody with the cell-surface antigen and separating free antibody in solution from antibody that is bound to the cell-surface antigen. The free antibody in solution may then be added to a solid support having a fixed antigen and incubated. A labeled agent that binds to antibody that is bound to the fixed antigen is then added, and unbound label is removed. Electrochemiluminescence of the labeled agent that is bound to the antibody that is bound to the fixed antigen is then measured and that data is used to calculate a binding affinity of the antibody to the cell-surface antigen. The method may be implemented in high throughput format, may be automated and/or continuous, and further may be controlled by a central processor, for example a computer.

Description

MATERIALS AND METHODS FOR BIOLOGIC AFFINITY ANALYSIS

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of priority under 35 U.S.C. § 119 to U.S. Provisional Patent Application Serial Nos. 63/555,631 and 63/555,627, both filed on February 20, 2024, the disclosures of which are incorporated by reference herein in their entirety.

FIELD

[0002] Materials and methods for biologic affinity analysis, including the field of biologic affinity analysis, and more particularly to methods for determining the affinity of an antibody to a cell-surface antigen.

BACKGROUND

[0003] Bispecific antibodies (BsAbs) are being developed as an alternative to combination therapy of two or more mAbs or as a means to redirect immune effector cells to target cells (Jarantow, 2015). BsAbs have been shown to have potential as therapeutics for some cancers, such as acute myeloid leukemia immunotherapy (Kontermann, 2012; Chichili, 2015) and can be used as a tool for diagnosis and imaging of cancer and infectious diseases (Thakur, 2016). Methods to assess the binding affinity of the individual binding arms of a bispecific Ab are used in the process of selecting candidate molecules and selecting human doses.

SUMMARY

[0004] The inventors of the inventions disclosed herein recognize that the process of identifying biologically relevant targets, developing pharmaceutical treatments, and ultimately registering and commercializing biotherapeutic molecules requires the use of sensitive, accurate, and technically advanced analytical methods. These methods must enable the clear selection of optimal candidates. Beyond in vitro functional assays and in vivo models demonstrating pharmacological efficacy and safety, a critical factor in the pharmaceutical advancement of biologic therapeutics is the intrinsic affinity of therapeutic antibodies for their antigenic targets. The modulation of target function, among other factors, depends on the binding to the relevant conformation of the target protein. Therefore, the inventors appreciate the importance of performing affinity-based selection of candidates early in therapeutic antibody discovery using biologically relevant target conformations and doing so in a time-efficient manner.

[0005] The inventions disclosed herein address the need for analytical methods that enable the study of affinities between cell-surface targets and antibodies in whole cells, particularly within a high-throughput framework. This approach enables the measurement of antibody-antigen interactions in the nanomolar (nM) to sub-nanomolar (sub-nM) range, while also allowing the determination of antibody affinity to its antigen in its native cellular environment.

[0006] Various methods have been employed for affinity analysis of antibodies binding to cell- surface-expressed targets, including radioimmunoassays (Rogers, 1983), enzyme-linked immunoassays (Bator, 1989; Debbia, 2004), fluorometric assays (Bondza, 2017; Encarnacao, 2018), flow cytometry (Benedict, 1997; Siiman, 2000), surface plasmon resonance (SPR) (Navratilova, 2011; Wang, 2012; Shepherd, 2014), kinetic exclusion assays (KinExA; Xie, 2005; Rathanaswami, 2008), and surface plasmon resonance microscopy (SPRm) along with other SPR-microscopy hybrid technologies (Zhang, 2022). While these methods are highly sensitive and accurate, they can be cumbersome, requiring receptor solubilization, protein labeling, or they may have limited sample capacity.

[0007] High-throughput solution-based affinity assays using meso scale discovery (MSD) technologies have been developed to identify high-affinity antibodies (Estep, 2013; Della Ducata, 2015). These assays enable equilibrium affinity measurements for soluble or solubilized antigens with greater throughput. However, many cell surface antigens, particularly receptors with multiple transmembrane domains, are difficult to solubilize and purify. Moreover, solubilized antigens may adopt different conformations compared to their native, membranebound forms. For instance, G protein-coupled receptors (GPCRs) require extensive purification and solubilization efforts to maintain their functional and structural integrity (Lundstrom, 2005). While some methods have made progress in addressing these challenges, they often suffer from low capacity and/or require immobilization or fixation of cells (Bondza, 2017; Jing, 2020).

[0008] These challenges in the prior art are overcome by the inventions disclosed herein. More specifically, the disclosed inventions address the need for more effective methods for determining the affinity of an antibody to a cell-surface antigen.

[0009] Accordingly, provided is a method for determining the affinity of an antibody to a cellsurface antigen comprising: (a) means for determining the affinity of the antibody to the cellsurface antigen by specific interaction of the antibody with the cell-surface antigen; and (b) providing a labeled agent that binds to the antibody.

[0010] Further provided is a method for determining the affinity of an antibody to a cell-surface antigen comprising: (a) incubating the antibody with the cell-surface antigen; (b) separating free antibody in solution from antibody that is bound to the cell-surface antigen; (c) adding the free antibody in solution to a solid support comprising a fixed antigen and incubating; (d) adding a labeled agent that binds to antibody that is bound to the fixed antigen; (e) removing free labeled agent; and (f) measuring electrochemiluminescence of the labeled agent that is bound to the antibody that is bound to the fixed antigen.

[0011] In some embodiments, the methods may be in high throughput format. In some embodiments, the methods may be automated and/or continuous. In further embodiments, the automation is controlled by a central processor, for example a computer. In yet further embodiments, the automation is robotic.

[0012] In some embodiments, the methods may further comprise measuring electrochemiluminescence of the labeled agent that is bound to the antibody.

[0013] In some embodiments, the methods may further comprise calculating the binding affinity of the antibody to the cell-surface antigen by analyzing electrochemiluminescence data. In some embodiments, the electrochemiluminescence data is used to calculate percent free antibody.

[0014] In some embodiments, the methods may further comprise (g) calculating binding affinity of the antibody to the cell-surface antigen by analyzing the electrochemiluminescence data, wherein the calculating comprises an equation set forth in: wherein the electrochemiluminescence data is used to calculate percent free antibody, wherein the percent free antibody is calculated according to an equation comprising: . > _ 100.

[0015] In some embodiments, the cell-surface antigen in step (a) is expressed on a cell.

[0016] In some embodiments, the fixed antigen comprises an epitope that is also on the cellsurface antigen.

[0017] In some embodiments, the fixed antigen is biotinylated and is bound to streptavidin that is bound to a solid surface or plate.

[0018] In some embodiments, the labeled agent is a ruthenylated antibody.

[0019] In some embodiments, the antibody to a cell-surface antigen is a monoclonal antibody, a Fab fragment, an scFv, a bispecific antibody, or a multispecific antibody. In some embodiments, the multispecific antibody has more than two target antigens. In further embodiments, the antibody to a cell-surface antigen is a human antibody. [0020] In some embodiments, the cell-surface antigen is a receptor. In further embodiments, the receptor is a cytokine receptor, a lymphokine receptor, a T-cell receptor, or a chimeric antigen receptor (CAR), or a G-protein coupled receptor (GPCR).

[0021] In some embodiments, the step (b) of separating free antibody in solution from antibody that is bound to the cell-surface antigen is done by centrifugation.

[0022] In some embodiments, the method further comprises, after step (c), washing with a buffer and optionally repeating the wash 2-3 times.

[0023] In some embodiments, the method further comprises, after step (d), washing with a buffer and optionally repeating the wash 2-3 times.

[0024] In some embodiments, the method further comprises, after step (e), washing with a buffer and optionally repeating the wash 2-3 times.

[0025] In some embodiments, the electrochemiluminescence is measured by an electrochemiluminescence reader.

[0026] In some embodiments, the binding affinity of the antibody to the cell-surface receptor is calculated by a central processor, for example a computer.

[0027] In some embodiments, the binding affinity of the antibody to the cell-surface receptor may be calculated by a central processor, for example a computer, wherein the binding affinity of the antibody is calculating using equation:

[0028] In some embodiments, the cell-surface antigen concentration may be calculated by a central processor, for example a computer, wherein the calculation is performed using an equation: wherein a fitted parameter “n” is used to calculate the cell-surface antigen concentration by multiplying an estimated concentration of antigen that was entered at a beginning of a data fitting procedure by “n.” DESCRIPTION OF THE FIGURES

[0029] FIG. 1 illustrates a schematic representation of a method of the present disclosure, specifically the MSD-CAT procedure.

[0030] FIGS. 2A-2F illustrate representative binding curves of an MSD-CAT affinity analysis for cell surface expressed CD123 on SP1 cells, wherein the Y-axis shows % free anti-receptor molecules (BsAb, Fab, or mAh) or CD123 negative control (I3CB15) while the X-axis shows the concentration of cell surface antigen (or cells/mL for parental cells). Each curve represents a different but fixed concentration of mAh, bispecific, or Fab as indicated, wherein FIG. 2A is BsAl, FIG. 2B is BsA2, FIG. 2C is BsA3, FIG. 2D is mAbA, FIG. 2E is FabA, and FIG. 2F is I3CB15.

[0031] FIGS. 3A-3F illustrate representative binding curves of an MSD-CAT affinity analysis for cell surface expressed CD 123 on SP2 cells, wherein the Y-axis shows % free anti-receptor molecules (BsAb, Fab, or mAh) or CD123 negative control (I3CB15) while the X-axis shows the concentration of cell surface antigen (or cells/mE for parental cells). Each curve represents a different but fixed concentration of mAh, bispecific, or Fab as indicated, wherein FIG. 3A is BsAl, FIG. 3B is BsA2, FIG. 3C is BsA3, FIG. 3D is mAbA, FIG. 3E is FabA, and FIG. 3F is mAbA on parental cells.

[0032] FIGS. 4A-4D illustrate representative binding curves for MSD-CAT affinity analysis for cyno CD123. The Y-axis shows % free anti-receptor molecules (Fab in FIG. 4A or mAh in FIG. 4C) while the X-axis shows the concentration of cell surface antigen. Each curve represents a different but fixed concentration of mAh or Fab as indicated. The figures also illustrate a representative binding curve of the control commercial antibody, 7G3 (FIGS. 4B and 4D), which shows cross reactivity to human CD123 (SP1).

DETAILED DESCRIPTION

[0033] While the general inventive concepts are susceptible of embodiment in many forms, there are shown in the drawings, and will be described herein in detail, specific embodiments thereof with the understanding that the present disclosure is to be considered an exemplification of the principles of the general inventive concepts. Accordingly, the general inventive concepts are not intended to be limited to the specific embodiments illustrated herein.

[0034] It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting. [0035] The articles “a” and “an” are used herein to refer to one or more than one (i.e., to at least one) of the grammatical object of the article. By way of example, “a cell” means one cell or more than one cell.

[0036] “About” as used herein when referring to a measurable value such as an amount, a temporal duration, and the like, is meant to encompass variations of ±5%, preferably ±1%, and still more preferably ±0.1 % from the specified value, as such variations are appropriate to perform the disclosed methods.

[0037] The terms “antibody” and "antibodies" as used herein are meant in a broad sense and include immunoglobulin molecules including polyclonal antibodies, monoclonal antibodies including murine, human, human-adapted, humanized, and chimeric monoclonal antibodies, antibody fragments, bispecific or multispecific antibodies, dimeric, tetrameric or multimeric antibodies, and single chain antibodies.

[0038] Immunoglobulins can be assigned to five major classes, namely IgA, IgD, IgE, IgG and IgM, depending on the heavy chain constant domain amino acid sequence. IgA and IgG are further sub-classified as the isotypes IgAl, IgA2, IgGl, IgG2, IgG3, and IgG4. Antibody light chains of any vertebrate species can be assigned to one of two clearly distinct types, namely kappa (K) and lambda (X), based on the amino acid sequences of their constant domains.

[0039] The term "antibody fragments" refers to a portion of an immunoglobulin molecule that retains the heavy chain and/or the light chain antigen binding site, such as heavy chain complementarity determining regions (HCDR) 1, 2, and 3, light chain complementarity determining regions (LCDR) 1, 2, and 3, a heavy chain variable region (VH), or a light chain variable region (VL). Antibody fragments include a Fab fragment, a monovalent fragment consisting of the VL, VH, CL, and CHI domains; a F(ab)2 fragment, a bivalent fragment comprising two Fab fragments linked by a disulfide bridge at the hinge region; a Fd fragment consisting of the VH and CHI domains; a Fv fragment consisting of the VL and VH domains of a single arm of an antibody; a domain antibody (dAb) fragment, which consists of a VH domain. VH and VL domains can be engineered and linked together via a synthetic linker to form various types of single chain antibody designs where the VH/VL domains pair intramolecularly, or intermolecularly in those cases when the VH and VL domains are expressed by separate single chain antibody constructs, to form a monovalent antigen binding site, such as single chain Fv (scFv) or diabody; described for example in PCT Inti. Publ. Nos. W01998/44001, WO1988/01649, WO1994/13804, and W01992/01047. These antibody fragments are obtained using well known techniques known to those of skill in the art, and the fragments are screened for utility in the same manner as are full length antibodies.

[0040] The phrase "isolated antibody" refers to an antibody or antibody fragment that is substantially free of other antibodies having different antigenic specificities (e.g., an isolated antibody specifically binding CD38 is substantially free of antibodies that specifically bind antigens other than human CD38). An isolated antibody that specifically binds CD38, however, can have cross-reactivity to other antigens, such as orthologs of human CD38, such as Macaca fascicularis (cynomolgus monkey) CD38. Moreover, an isolated antibody may be substantially free of other cellular material and/or chemicals.

[0041] "Humanized antibody" refers to an antibody in which the antigen binding sites are derived from non-human species and the variable region frameworks are derived from human immunoglobulin sequences. Humanized antibodies may include substitutions in the framework regions so that the framework may not be an exact copy of expressed human immunoglobulin or germline gene sequences.

[0042] "Human antibody" refers to an antibody having heavy and light chain variable regions in which both the framework and the antigen binding sites are derived from sequences of human origin. If the antibody contains a constant region, the constant region also is derived from sequences of human origin. A human antibody comprises heavy or light chain variable regions that are "derived from" sequences of human origin wherein the variable regions of the antibody are obtained from a system that uses human germline immunoglobulin or rearranged immunoglobulin genes. Such systems include human immunoglobulin gene libraries displayed on phage, and transgenic non-human animals such as mice carrying human immunoglobulin loci as described herein. A human antibody may also contain amino acid differences when compared to the human germline or rearranged immunoglobulin sequences due to for example naturally occurring somatic mutations or intentional introduction of substitutions in the framework or antigen binding sites. Typically, a human antibody is at least about 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical in amino acid sequence to an amino acid sequence encoded by a human germline or rearranged immunoglobulin gene.

[0043] Isolated humanized antibodies may be synthetic. Human antibodies, while derived from human immunoglobulin sequences, may be generated using systems such as phage display incorporating synthetic CDRs and/or synthetic frameworks, or can be subjected to in vitro mutagenesis to improve antibody properties, resulting in antibodies that do not naturally exist within the human antibody germline repertoire in vivo.

[0044] The term "recombinant antibody" as used herein, includes all antibodies that are prepared, expressed, created or isolated by recombinant means, such as antibodies isolated from an animal (e.g., a mouse) that is transgenic or transchromosomal for human immunoglobulin genes or a hybridoma prepared therefrom, antibodies isolated from a host cell transformed to express the antibody, antibodies isolated from a recombinant, combinatorial antibody library, and antibodies prepared, expressed, created, or isolated by any other means that involve splicing of human immunoglobulin gene sequences to other DNA sequences, sequences, or antibodies that are generated in vitro using Fab arm exchange such as bispecific antibodies.

[0045] The term "monoclonal antibody" as used herein refers to a preparation of antibody molecules of single molecular composition. A monoclonal antibody composition displays a single binding specificity and affinity for a particular epitope, or in a case of a bispecific monoclonal antibody, a dual binding specificity to two distinct epitopes.

[0046] The term "epitope" as used herein means a portion of an antigen to which an antibody specifically binds. Epitopes usually consist of chemically active (such as polar, non-polar or hydrophobic) surface groupings of moieties such as amino acids or polysaccharide side chains and can have specific three-dimensional structural characteristics, as well as specific charge characteristics. An epitope can be composed of contiguous and/or discontiguous amino acids that form a conformational spatial unit. For a discontiguous epitope, amino acids from differing portions of the linear sequence of the antigen come in close proximity in 3-dimensional space through the folding of the protein molecule.

[0047] The term “chimeric antigen receptor” or “CAR” as used herein means a synthetic or recombinant receptor comprising an antigen specific domain, a costimulatory domain and an intracellular signaling domain. In some embodiments, the CAR further comprises an extracellular hinge or spacer region, a transmembrane domain, or combinations thereof. In some embodiments, the antigen specific domain is an scFv.

[0048] The term “chimeric antigen receptor T cell” or “CAR-T” as used herein means a T cell expressing a CAR.

[0049] In some embodiments of any of the compositions or methods described herein, a range is intended to comprise every integer or fraction or value within the range. [0050] Embodiments described herein as “comprising” one or more features may also be considered as disclosure of the corresponding embodiments “consisting of’ and/or “consisting essentially of’ such features.

[0051] To date, the determination of affinity of an antibody to a cell-surface antigen may have required solubilization of receptors, labeling of proteins, and/or may have had limited sample capacity. Provided are methods that do not have the same limitations as the previously described methods.

METHODS

[0052] Provided is a method for determining the affinity of an antibody to a cell-surface antigen. The method generally comprises: (a) incubating the antibody with the cell-surface antigen; (b) separating free antibody in solution from antibody that is bound to the cell-surface antigen; (c) adding the free antibody in solution to a solid support comprising a fixed antigen and incubating; (d) adding a labeled agent that binds to antibody that is bound to the fixed antigen; (e) removing free labeled agent; and (f) measuring electrochemiluminescence of the labeled agent that is bound to the antibody that is bound to the fixed antigen.

[0053] In some embodiments, the method is in high throughput format. In some embodiments, the method is automated and/or continuous. In further embodiments, the automation is controlled by a central processor, for example a computer. In yet further embodiments, the automation is robotic.

[0054] In some embodiments, the step (b) of separating free antibody in solution from antibody that is bound to the cell-surface antigen may be done by centrifugation.

[0055] In some embodiments, the method further comprises, after step (c), washing with a buffer. The washing step may optionally be repeated 2-3 times.

[0056] In some embodiments, the method further comprises, after step (d), washing with a buffer. The washing step may optionally be repeated 2-3 times.

[0057] In some embodiments, the method further comprises, after step (e), washing with a buffer. The washing step may optionally be repeated 2-3 times.

Cell-Surface Antigen

[0058] The antigen used in step (a) may be an antigen present on the surface of a cell, i.e., cellsurface antigen. In some embodiments, the cell-surface antigen is a receptor. In further embodiments, the receptor is a cytokine receptor, a lymphokine receptor, a T-cell receptor, or a chimeric antigen receptor (CAR).

[0059] The cell-surface antigen may be cell surface expressed. In some embodiments, the cell is transfected with a vector that encodes the cell-surface antigen. In further embodiments, the vector further encodes an antibiotic resistance cassette, for example a neomycin resistance cassette.

[0060] Cells may be generated that express the cell-surface antigen as a full-length isoform or as a truncated isoform. In some embodiments, the cell-surface antigen is human CD123-SP1 (full- length isoform) or CD123-SP2 (truncated isoform). CD123 is the alpha chain of the interleukin 3 receptor (IL-3Ra), which is a protein on the surface of cells that plays a role in the production and function of blood cells. CD123 is expressed in a number of hematologic malignancies, including leukemia and lymphoma, and is a marker that can be useful for prediction of clinical outcomes and prognosis in leukemia patients, particularly relapsed or refractory leukemia.

Antibody to the cell-surface antigen

[0061] In some embodiments, the antibody to a cell-surface antigen is a monoclonal antibody, a Fab fragment, an scFv, a bispecific antibody, or a multispecific antibody. In some embodiments, the multispecific antibody has more than two target antigens. In further embodiments, the antibody to a cell-surface antigen is a human antibody.

[0062] In some embodiments, the antibody to a cell-surface antigen is generated using biopanning, a technique that screens a mixture of bacteriophages to identify peptides or antibodies that bind to a target. It is an affinity selection technique that mimics the process of antibody selection.

[0063] In some embodiments, the antibody to a cell-surface antigen is generated by human framework adaptation and affinity maturation of an antibody. Framework adaptation modifies the structural framework, i.e., non-antigen binding regions, of the antibody to optimize its binding affinity to a specific antigen (e.g., make it more compatible with the immune system of the target subject, such as the human immune system). Affinity maturation increases an antibody's binding strength to its target antigen through genetic mutations in the complementarity determining regions (CDRs) within the antibody molecule.

[0064] In some embodiments, the antibody to a cell-surface antigen is a bispecific antibody. The bispecific antibody may be generated by joining two monospecific antibodies via a Fab arm exchange process. Solid Support

[0065] In some embodiments, the solid support is an electrochemiluminescence plate. In some embodiments, the plate has streptavidin attached to it. In further embodiments, the plate is a meso scale discovery (MSD) Streptavidin plate (L15SA-1, Mesoscale Discoveries, MD).

Fixed Antigen

[0066] The fixed antigen may be conjugated to the solid support directly or indirectly. In some embodiments, the antigen is biotinylated. In further embodiments, the biotinylated antigen binds to streptavidin that is attached to the solid support.

Labeled Agent

[0067] In some embodiments, the labeled agent is a labeled antibody that binds the antibody that is bound to the fixed antigen. In some embodiments, the antibody that is bound to the fixed antigen is a human IgG, and the labeled agent is an anti-human IgG antibody.

[0068] In some embodiments, the labeled agent is a ruthenylated antibody. In some embodiments, the ruthenylated antibody is an anti-human IgG antibody.

Electrochemiluminescence Reader

[0069] In some embodiments, the method comprises measuring electrochemiluminescence of the labeled agent that is bound to the antibody that is bound to the fixed antigen.

[0070] In some embodiments, the electrochemiluminescence of the labeled agent is measured using an electrochemiluminescence reader. The electrochemiluminescence reader may be commercially available, for example an MSD Sector Imager 6000™ Reader (MesoScale Discovery, Rockville, MD). In some embodiments, the electrochemiluminescence reader is built from individual components that are commercially available.

Data Analysis

[0071] In some embodiments, the method further comprises calculating the binding affinity of the antibody to the cell-surface antigen by analyzing electrochemiluminiscence data. In some embodiments, the electrochemiluminescence data is used to calculate percent free antibody.

[0072] In some embodiments, the binding affinity of the antibody to the cell-surface receptor is calculated by a central processor, for example a computer.

[0073] In some embodiments, to obtain equilibrium dissociation constants (KD, referred elsewhere herein as affinity) by the methods described herein, the data is processed in four different steps. In the first step the raw data (electrochemiluminescence signal) is plotted against an estimated or nominal antigen concentration and fitted to obtain maximum signal (Smax) for every curve. In the second step the Smax resulting from this fitting is used to calculate % free antibody and the Y-axes for all curves are normalized in terms of % free antibody. In the third step the normalized data for a selected antibody is fitted to obtain a correction factor “n” that would enable conversion of the nominal or estimated antigen concentration to actual molar antigen concentration. In the fourth step, % Free antibody plotted against actual antigen concentration is fitted to obtain equilibrium dissociation constants.

[0074] In some embodiments, fitting is performed using non-linear least square analysis using the Graph Pad software (Prism). Exemplary equations employed are described in the Examples.

[0075] Advantages of the methods provided herein include but are not limited to not requiring solubilization of cell-surface antigens, not requiring labeling of proteins or not having limited sample capacity.

EXAMPLE: Method for determining the affinity of an antibody to a cell-surface antigen

[0076] A. Materials

[0077] A.i. Generation o f monoclonal monospeci fic antibodies

[0078] For generation of monoclonal anti-CD123 antibodies, recombinant human CD123 SP1 ECD-His tag protein (R&D Systems, 301-R3/CF) was used for phage panning and hit screening. CD123 was biotinylated prior to use in the phage panning studies.

[0079] For generation of monoclonal anti-CD3 antibodies, monoclonal antibody SP34, mouse IgG3/lambda isotype (BD Biosciences Pharmingen, 556611) and monoclonal antibody, SP34-2, mouse IgGl/lambda isotype (BD Biosciences, 551916) were purchased.

[0080] A.ii. Recombinant CD 123 proteins

[0081] Recombinant human CD123 SP1 ECD-His tag protein was obtained from R&D Systems (301-R3/CF). Recombinant human CD123 SP2 ECD-His tagged protein was purified from baculovirus (custom produced by BlueSky in High 5 cells).

[0082] Anti-CD123 Fab-His was expressed in HEK293F and purified as described by the methods of Zhao and co-workers (Zhao, 2009).

[0083] A. Hi. Surface Plasmon Resonance

[0084] Goat anti-human IgG Fey fragment-specific Ab (cat # 109-005-098) and sheep anti human IgG (Fd) antibody (Cat # PC075) were obtained from Jackson ImmunoResearch laboratories (West Grove, PA) and Binding Site (San Diego, CA), respectively. Biacore T200, series S CM5 sensor chips (GE Healthcare cat # BR-1005-30) and reagents for preparation of the capture surfaces were obtained from Biacore (GE healthcare, Piscataway, NJ) or from BioRad Life Sciences (Bio-Rad, Hercules, CA). Experiments were run using 0.2 pm filter-sterilized and degassed PBS pH 7.4, supplemented with 3 mM EDTA and 0.005% Tween 20 (Bio-Rad cat # 176-2730) as the Biacore running buffer (BRB).

[0085] A.iv. MSD-CAT

[0086] The panel of human bispecific antibodies, Fabs, and mAbs used in this study are listed in Table 1. Cell culture medium DMEM Glutamax (Life Technology # 10569-044), cell dissociation buffer (Gibco, cat # 13150-016), and Dulbecco’s phosphate buffered saline (PBS) containing 10 mM phosphate buffer, pH 7.5, and 0.15 M NaCl (Gibco # 14190) were purchased from Life Technologies (Grand Island, NY, USA). MSD Streptavidin Standard plates (L15SA- 1), MSD Read Buffer 4X (R92TC-1), and Sulfo-TAG NHS Ester (MSD R91AN-1) were products of MesoScale Discovery (Rockville, MD). Unconjugated donkey anti-human IgG (H+L) (cat # 709-006-149) was procured from Jackson ImmunoResearch (West Grove, PA, USA). Magnet GENIE ROTATOR GENIE, 1-2200 (VWR 89202-306) and Universal Clip Plates SI- 1134 (VWR 14222-140) were purchased from VWR. Super Sealer Automated CapMat (Thermo 4110), sterile CapMats (Thermo 4412-11), Sulfo-NHS-LC-B iotin (Thermo 21327), and Zeba Desalt Spin Columns (Thermo 89889) were purchased from Thermo Scientific (Rockford, IL).

[0087] B. Methods

[0088] B.i. Generation of monoclonal anti-CD123 antibodies

[0089] Anti-CD123 monoclonal antibodies were generated using a phage panning campaign against rhCD123 SP1.

[0090] Human CD123 protein was subjected to biopanning using the de novo pIX Fab libraries made by Shi and co-workers (Shi, 2010). Biotinylated CD123 was captured on streptavidin magnetic beads (Dynal), then exposed to the de Novo pIX Fab libraries at a final concentration of 100 nM. Phages were amplified from these cells overnight and panning was repeated for a total of four rounds. Monoclonal Fabs were screened for binding to CD123 in an ELISA where Fab was captured on an ELISA plate by anti-Fd antibody and biotinylated CD 123 was added to captured Fabs, followed by detection of CD123 with streptavidin/HRP (Pierce). Clones that demonstrated binding to CD 123 were sequenced in the heavy (HC) and light chain (LC) variable regions. Unique Fab sequences identified via phage panning were ultimately converted to IgG for further characterization.

[0091] The expression vectors containing the genes encoding the light chain (LC) and heavy chain (HC) were transiently expressed in Expi293F cell using standard procedures. These are human IgG4_PAA antibody molecules with the respective F405L and K409R mutations. The individual culture supernatants expressing parental mAbs were purified using pre-packed 1 ml or 5 mL HiTrap MabSelect SuRe Protein A columns. Columns were equilibrated with 5 column volumes (CV) of D-PBS, pH 7.2 prior to loading the culture supernatants. Unbound proteins were removed by washing with 10 column volumes (CV) of D-PBS, pH 7.2. Bound protein was eluted with 5 CV 0.1 M Na-acetate, pH 3.5 and neutralized with 2.5M Tris pH 7.2. The neutralized protein solution was dialyzed into D-PBS, pH 7.2.

[0092] B .ii. Generation of monoclonal anti-CD3 antibodies

[0093] Anti-CD3 monoclonal antibodies were generated by human framework adaptation and affinity maturation of a public domain antibody known as SP34 (Zorn, 2022). SP34 sequence was elucidated in-house, and the resulting antibody human was adapted and affinity matured. Three antibodies were selected from the affinity maturation effort to be paired with mAbA for the generation of bispecific antibodies.

[0094] B .Hi. Generation of bispecific antibodies

[0095] The lead CD 123 and CD3 antibodies are joined together post-purification via the Fab Arm Exchange process. This results in a monovalent binding, bi-functional antibody. Bispecific human antibodies were generated by the controlled Fab-arm exchange process from lead anti- CD3 with lead anti-CD123 molecules, using the method of Labrijn and co-workers (Labrijn, 2013), as modified by others (Paul, 2016; Gramer, 2013). Purified anti-CD3 parental mAh and anti-CD123 parental mAh were mixed in a 1:1.08 mass ratio in D-PBS to minimize the residual anti-CD3 mAh. 2-MEA dissolved in D-PBS was added to a final concentration of 75 mM. The mixture was rotated for 5 min at room temperature before being transferred to a 31 °C incubator where the exchange was allowed to continue for 5 h without agitation. The 2-MEA was removed by dialyzing the samples into lx DPBS, pH 7.2 using a 20K MWCO dialysis cassette.

[0096] The protein concentration for each purified mAh and bispecific antibody was determined by measuring the absorbance at 280nm on a NanoDroplOOO spectrophotometer. Extinction coefficients were calculated based on the amino acid sequence. Standard SE-HPLC and Caliper LabChip GXII methods were used to analyze the proteins. All bispecifics were greater than 95% monomeric by SE-HPLC or analytical-CEX.

[0097] B .iv. Generation of CD123 cell lines (human CD123 SP1 and human CD123 SP2 and cyno CD 123}

[0098] Generation of pDisplay Vectors for human CD123 SP1, human CD123 SP2, and cyno CD123 SP1

[0099] A set of pDisplay vectors presenting human CD123 SP1 ECD, human CD123 SP2 ECD, and cyno CD123 ECD were generated for use as screening tools to assess the anti-CD123 leads. Invitrogen’s pDisplay vector was used to construct three different cell lines with CD123 expression on the cell surface. Transfected 293F adherent cells were selected for stable plasmid integration, then single cell-sorted and the CD 123 surface receptor expression was quantified by FACS. A set of 10 single cell clones were selected for screening for each cell line and three were quantified for CD 123 ECD expression. The cell lines used for subsequent hit screening were CD123 hSPl Clone 9, CD123 hSP2 Clone 5 and CD123 cynoSPl Clone 3, with surface expression of approximately 500,000 CD 123 ECD copies per cell.

[0100] B.v. Surface Plasmon Resonance Procedure

[0101] Surface plasmon resonance (SPR) experiments were performed using a Biacore T200 optical biosensor (Cytiva Life Sciences) and all experiments were run in PBS pH 7.4, supplemented with 3 mM EDTA and 0.005% Tween 20 at 25 °C. Biosensor surfaces were prepared by coupling goat anti -human IgG Fey fragment specific or sheep anti-human IgG Fd antibody in 10 mM sodium acetate, pH 4.5 to the carboxymethylated dextran surface of a CM5 chip using the manufacturer’s instructions for amine-coupling chemistry. An average of 6000 response units (RU) of for each antibody was immobilized in each of four flow cells. The antibodies were captured (~ 40-70 RU) onto the anti-human Fey antibody-modified sensor chip surfaces while Fabs were captured (~ 50 RU) onto the anti -human Fd antibody-modified sensor chip surfaces. Following antibody capture, monomeric recombinant human full-length antigen (huCD123 SP1) or truncated antigen-variant (huCD123 SP2) in solution (1.6 nM to 400 nM with 4-fold dilutions) were injected. The association was monitored for 3 minutes (120 pL injected at 40 pL/min) and dissociation was monitored for 10 minutes. Regeneration of the sensor surface was achieved with a 20 second injection of 0.85% H3PO4 followed by 20 second injection of 50 mM NaOH. [0102] Data were processed using the Biacore T200 evaluation software. Double reference subtraction of the data was performed by subtracting the curves generated by buffer injection from the reference- subtracted curves for analyte injections to correct for bulk refractive index changes to the signal and systematic instrument noise in the flow cell (Myszka, 1999; Myszka, 2000). After data processing, the data generated for kinetic and affinity determination were analyzed using the Biacore T200 Evaluation software - version 2, and Biacore T200 Kinetic summary. The response data were globally fitted using a 1 : 1 interaction model.

[0103] B.vi. Cell Culture

[0104] Cells over-expressing human CD123 SP1 (full-length isoform), CD123-SP2 (truncated isoform), and negative control of HEK293F parental cells were maintained in DMEM Glutamax supplemented with 10% fetal bovine serum (FBS), 1% sodium pyruvate, 1% Glutamine. Cells expressing full-length and truncated receptor isoforms have a neomycin resistance cassette and were periodically selected with 400 pg/mL of Geneticin.

[0105] B .vii. Preparation of MSD-CAT labeled reagents for capture and detection

[0106] Unconjugated donkey anti-human IgG (H+L) was labeled with Sulfo-TAG NHS Ester with a ratio of 1:12, respectively. Unconjugated recombinant human antigen was biotinylated with Sulfo-NHS-LC-Biotin with a ratio of 1:3, respectively. The labeling reactions were placed in 2 mL polypropylene tubes covered with aluminum foil, mixed gently, and allowed to incubate at ambient temperature for 1 hour in the dark. The unconjugated free tag was removed with Zeba Spin columns equilibrated with PBS.

[0107] B.viii. MSD-CAT procedure

[0108] A schematic representation of the MSD-CAT procedure is depicted in FIG. 1. To determine affinity by this MSD based method it is necessary to keep antibody concentration at, or lower than KD for a more accurate affinity measurement and better assay sensitivity. It is also important to find the optimal concentration of antigen on the plate that produces the optimized versus background signal reproducibly.

[0109] HEK293F cells uniquely over-expressing either human CD 123 SP1 or human CD 123 SP2 were used to assess the binding affinity of bispecific antibodies, mAh, and Fab. A commercially available antibody (anti-CD123, Clone: 7G3, BD Biosciences) was used as a positive control on the human CD123-SP1 and cyno CD123 experiments. A constant concentration of antibodies was allowed to incubate with varying concentrations of either cell line in incubation buffer (DMEM Glutamax medium, 0.05% Azide, 1% BSA, 3 mM EDTA). Following equilibration, cells with bound antibodies were removed by centrifugation and the concentration of free antibodies in the supernatant was determined and the data were analyzed for affinity. For these studies the reaction mixtures were prepared by adding an equal volume of antibodies or Fab or mAbs at four different but fixed concentrations (800, 160, 32, and 6 pM, or 1000, 200, 40, 8 pM) and the cells were 3-fold serially diluted starting at 2xl0⁷ cells/mL in 96- well polypropylene plates. Super Sealer was used to seal 96 well plates with a sterile Cap Mat and plates were agitated briefly. The mixtures were allowed to reach equilibrium by incubating with gentle rotation of the plates on a rotator for ~24 hours at 4°C. The following day, plates were spun for 10 minutes at 2000 rpm to separate supernatant from cells and cell-bound antibodies. The Cap Mat was carefully removed to avoid spillage or cross contamination between wells and the concentration of free antibodies in the supernatant was measured by the Electrochemiluminescence Immunoassays (ECLIA) as described below, using aliquots of 50 pL/well of the supernatant.

[0110] To detect free (unbound) antibody in solution using the (ECLIA) assay, MSD SA-STD plates were blocked with 50 pL per well of assay buffer (PBS containing 0.2% BSA, 0.05% Tween 20) for at least 5 minutes. Plates were turned over to remove assay buffer and tapped on paper towels and 50 pL per well of 0.7 pg/mL biotinylated human CD123 SP1 or human CD123 SP2 in assay buffer were added and the plates were allowed to incubate overnight in the refrigerator. The next morning, 150 pL of assay buffer was added to block each well of the plates without removing the biotinylated CD 123 and the plates were allowed to incubate for ~ 60 minutes. The plates were washed three times with wash buffer (PBS containing 0.05% Tween 20) and were tapped lightly on paper towels to remove residual wash buffer. Fifty microliters of samples containing free antibodies were added to each well, and the plates were allowed to incubate for 1 hour with gentle vortex at ambient temperature. Plates were washed three times with wash buffer, and 50 pl per well of 0.7 pg/mL ruthenium-labeled donkey antihuman IgG (H+L) were added and allowed to incubate for 1 hour with gentle vortex at ambient temperature. Finally, the plates were washed three times with wash buffer, and 150 pL of read buffer were added to each well and immediately read on the MSD Sector Imager 6000™ Reader for luminescence levels.

[0111] For the detection of negative control antibody (I3CB15) the capture was performed by capturing with anti-human IgG in place of CD 123.

[0112] B .ix. MSD- CAT data analysis [0113] To obtain equilibrium dissociation constants (KD, referred in other places in this paper as affinity) by MSD-CAT analysis, the data was processed in four different steps. In the first step the raw data (electrochemiluminescence MSD signal) was plotted against an estimated or nominal antigen concentration and fitted to obtain maximum signal (Smax) for every curve. In the second step the Smax resulting from this fitting was used to calculate % free antibody and the Y- axes for all curves were normalized in terms of % free antibody. In the third step the normalized data for a selected antibody were fitted to obtain a correction factor “n” that would enable conversion of the nominal or estimated antigen concentration to actual molar antigen concentration. The selected antibodies were mAh A for human CD123 and 7G3 for cyno CD123. These antibodies were selected based on the criteria of containing both Ko-controlled and stoichiometrically controlled curves as described by Bee and collaborators (Bee, 2012). This factor “n” was used both to calculate actual membrane antigen concentration in the prep and to calculate receptor number per cell. In the fourth and last step, % Free antibody plotted against actual antigen concentration was fitted to obtain equilibrium dissociation constants. All fitting was performed using non-linear least square analysis using the Graph Pad software (Prism) with equations derived in similar fashion as those described by Ohmura and collaborators (Ohmura, 2001).

[0114] In more detail, to determine KD the necessary equations were derived using the mass balance equation, eq. 1 , describing the equilibrium dissociation constant, KD. where [B] is free receptor concentration, [A] is free antibody binding site concentration, and [AB] is the concentration of receptor/antibody complex.

[0115] According to the law of mass conservation we can write total receptor concentration, BT, and total antibody binding site concentration, AT, as

Eqs. 2 and 3 can be rearranged to give [B] and [AB], thus yielding eqs. 4 and 5,

[B] = [B_r] - [AB] (4)

[AB] = [A_r] - [A] (5)

Eq. 5 can be substituted into eq. 4 to give an equation for [B] in terms of [BT], [AT], and [A] yielding eq. 6.

[B] = B_T - A_T + [A] (6) By substitution of eqs. 5 and 6 into eq. 1, a quadratic equation, eq. 7, results that is the solution for [A],

A² + (B_T - A_T + K_D)A - A_TK_D = 0 (7)

The quadratic formula is used to solve for [A] as shown in eq. 8,

The MSD signal can be related to [A] by eq. 9 where P is a proportionality constant that relates concentration in molarity to signal and a term for nonspecific or background binding, NSB, at zero concentration of [BT].

MSD signal=P*[A] + NSB (9)

The proportionality constant can be defined in terms of Smax, NSB, and [AT] where Smax is the maximum binding signal as shown in eq. 10, p _ (Smax-NSB')

Mr] ^J

[0116] Finally, eq. 11 is arrived at by substitution of eqs. 8 and 10 into eq. 9. The dilution factor, DF, shown in eqs. 11 and 12 is to allow for fitting of multiple titration curves globally with different constant antibody concentrations. DF is held constant during nonlinear fitting. For example, if two titration curves were analyzed globally with a 10-fold difference in constant antibody concentration, the highest concentration of antibody used would be assigned a DF of “1” and the titration curve with the lower concentration of constant antibody would be assigned a DF of “10.” Eq. 11 (receptor concentration known) is identical to eq. 12 (receptor concentration unknown) except for the fitted parameter “n” in all the terms containing [BT].

[0117] When fitting with an unknown receptor concentration, the nonlinear fitting is done by holding the concentrations of antibody constant from multiple titrations and fitting for “n” and KD globally across all curves and locally for Smax, and NSB for each curve. The fitted parameter “n” adjusts the initial estimated concentrations of receptor that were entered at the beginning of the data fitting procedure. The active concentration of receptor is n x the initial estimated concentration.

[0118] To perform the first of the four steps delineated above, the data were fitted using equation 11 to obtain maximum signal (Smax). Smax was obtained for every curve by locally fitting for Smax while plotting the MSD signal against an estimated or nominal antigen concentration. This estimated nominal concentration was arbitrarily chosen based on known receptor density expression levels for transfected cell lines. Smax obtained in this manner was used for the second step which was performed to calculate % free antibody in the equilibrated reaction using Equation 13. . > _ 100 (13)

[0119] This second step is a Y-axis normalization step performed for easier representation, visualization, and quality inspection of the data when plotted. After this normalization step, the third step was performed to obtain the fitted parameter “n” by fitting the Y-axis normalized data to equation 12 (antigen unknown). This fit was performed only for the data for the selected antibodies containing both Ko-controlled and stoichiometrically controlled, which are required for accurate “n” determination, as described above. The selected antibodies were mAh A for human CD123 and 7G3 for cyno CD123. The value of “n” obtained by fitting mAh A CD123 and 7G3 for human and cyno CD123, respectively, for each cell line was used to convert estimated receptor (CD123) concentration to actual molar receptor concentration as described by Xie and collaborators (Xie, 2005). Following this, the X-axis was expressed in terms of molar membrane CD123 concentration and the fourth and last step was performed for all antibodies, including those that showed weak or no binding or that did not have both Ko-controlled and stoichiometrically controlled curves. The affinities obtained from the fit using Equation 12 are reported here.

[0120] C. Results

[0121] C.i. Generation o f monospecific and bispecific antibodies

[0122] Parental monoclonal antibody 1 (mAh A) was generated targeting both CD 123 isoforms (CD 123 SP1 and CD 123 SP2) using phage display and the Fab of mAh A (Fab A) also prepared. Bispecific human antibodies were generated by the controlled Fab-arm exchange process from lead anti-CD3 molecules with a single lead anti-CD123 molecules, using the method of Eabrijn and co-workers (Labrijn, 2013; Paul, 2016) to generate three bispecific antibodies (BsAbs) designated as BsAl, BsA2, and BsA3. In addition, a negative control antibody (I3CB15) targeting an irrelevant antigen instead of CD 123 was also prepared and evaluated.

[0123] In all BsAb preparations, the residual anti-CD3 parental was not detected as determined by analytical ion exchange analysis (analytical-CEX) while residual anti-CD123 homodimer was 1.9% to 4.4% (data not shown).

[0124] C. ii. Affinities and On/Off Rates to purified recombinant antigen determined by SPR

[0125] Biacore studies were performed in three independent experiments using the recombinantly expressed extracellular domains of the CD123 SP1 (CD123 SP1) and CD123 SP2 (CD123 SP2), which were presented in monomeric form. To obtain monovalent affinities for the interactions, the parental antibody, its Fab and the BsAbs were captured onto the sensor surface via anti-Fc or anti-Fd (for Fab) antibody followed by injection of recombinant CD123 SP1 or CD 123 SP2. The kinetics and affinity for all forms of antibodies for both receptor target isoforms determined by surface plasmon resonance are shown in Table 1. mAh A and Fab A exhibit high binding affinities to both recombinant CD123 SP1 and CD123 SP2. However, mAbA and FabA, bind tighter to CD123 SP2 than to CD123 SP1. The data in Table 1 indicates that mAbA and FabA display a 3.6-fold and 4.6-fold higher affinity for CD123 SP2, respectively, compared to CD123 SP1 as determined by Biacore. However, it is important to note that another SPR experiment run in HTP format on a different instrument, ProteOn XPR36, showed affinities of mAh A to CD123 SP1 and CD123 SP2 of 0.6 nM and 0.4 nM, respectively. This ProteOn experiment, although following a similar trend in CD123 SP1 versus CD123 SP2 binding, shows however, only 1.5-fold higher affinity towards CD123 SP2. This could be due to potential differences in interactions between the recombinant materials with the different sensor chip surfaces in the two instrument which we have observed for other antigens.

[0126] These data indicate that the parental antibody and its Fab recognizes epitopes that are present on both recombinant receptor isoforms. The less than 2-fold differences in the SPR affinity of the mAh and its corresponding Fab indicate that the interaction of mAbA with either recombinant receptor isoform is monovalent. The binding of the BsAbs and I3CB15 to CD123 were compared to that of MabA to determine the influence of the anti-CD3 arm on CD 123 binding. The BsAbs bound to CD123 with equivalent affinity to the parent mAh indicating no influence of the CD3 arm on CD 123 binding. The I3CB15 control did not bind to either receptor isoform as expected (Table 1). [0127] In addition to mAbA and its Fab, Table 1 also describes SPR results for a series bispecifics containing different CD3 binding arms. In general, the data reveal that the bispecific antibodies exhibited similar SPR affinities for both recombinant CD123 SP1 and CD123 SP2 receptor isoforms and similar to the parental mAh, indicating that the variable second Fab arm of the bispecific pair did not significantly impact the affinity of mAbA. The parental mAbA and FabA have affinities (KD) for recombinant CD123 SP1 of 1.01 and 1.8 nM, respectively, while binding to CD123 SP2 with affinities of 0.28 and 0.39 nM, respectively. The bispecific antibodies bound to CD 123 SP1 and CD 123 SP2 with affinities between 1.34-1.49 nM (CD 123 SP1) and 0.31-0.36 nM (CD123 SP2), respectively (Table 1).

[0128] C.iii. Affinities of antibodies to cells expressing CD123 SP1 or CD123 SP2 determined by MSD-CAT

[0129] In the MSD-CAT procedure, transfected HEK293F cells expressing CD 123 SP1 or CD123 SP2 were employed to assess the affinity of all prepared antibodies. A schematic representation of the MSD-CAT procedure is depicted in FIG. 1 and is fully described in the Methods section. mAbA was used to determine receptor densities for the CD123 SP1 and CD123 SP2 cell lines of (8.07+ 4.47) x 10⁵ and (1.19+0.85) x 10⁶ receptors per cell, respectively. These numbers represent the average of three independent experiments. The receptor density was used to calculate molar concentration of receptors in the reaction mixture (see methods). Differences in receptor expression was observed between different cell batches as reported by others (Xie, 2005); therefore, data analysis was performed with the corresponding receptor density for the batch.

[0130] Affinity data for the various antibodies and representative binding profiles of several antibodies are presented in Table 1 (far right column) and FIGS. 2A-2F and 3A-3F, respectively. In contrast to the SPR studies, during the MSD-CAT studies, the antibodies and Fabs can move freely in solution while the target (receptor) is tethered to a surface (in this case the cell-surface). This provides opportunity for bivalent binding (the avidity effect) for the monospecific bivalent parental antibodies by cross-linking to neighboring target molecules on the cell-surface. Fabs and monovalent BsAbs which are unable to crosslink antigen were used to obtain monovalent affinities. The MSD-CAT affinities for the parental antibodies (Table 1) in comparison to the data for the Fab and BsAbs demonstrate that in a cellular context, the binding of the parental antibody (mAbA) is influenced by avidity. mAbA show at least 3 -fold higher apparent affinities relative to its Fab and BsAbs. [0131] Analogous to the Biacore results, the variable second arm of the bispecific pair did not seemingly impact the affinity of the bispecifics. The mAbA-derived BsAbs (BsAl, BsA2 and BsA3) exhibited relative affinities for cell-bound CD123 SP1 and CD123 SP2 of around 0.07 nM and 0.12 nM, respectively (<2-fold difference). The affinities of BsAs targeting cell- associated CD 123 SP1 are similar to the affinity for Fab A (0.08 nM) indicating that the bispecific affinities are representative of the antibody monovalent affinities. The affinity of FabA to cellular CD 123 SP2 was 0.13 nM, indicating that mAh A (0.007 nM) binds to CD 123 SP2 bivalently.

[0132] To evaluate the potential advantages of the MSD-CAT method when confronting difficult to purify antigens, the method was applied to determine cross- reactivity of mAbA and FabA to cynomolgus CD 123 which could not be obtained recombinantly due to instability during purification. HEK293F cells transfected with the full-length non-human primate version (cyno-CD123) of CD 123 SP1 were generated. The data show (Table 1, FIGS. 4A-4D) that mAbA and FabA showed weak binding to cells expressing the cynic CD123 with affinity > 11 nM which was important to support toxicological study design. On the other hand, 7G3, a commercial mouse antibody known to bind to both human and monkey CD 123, bound with 0.097 nM affinity to cyno CD123 SP1 and with comparably affinity (Table 1) to human CD123 SP1 expressed on cells, clearly demonstrating the differences in affinity for this antigen which could not be expressed recombinantly.

[0133] The strength of the affinity of antibodies to target antigens is driven in large part by the dissociation rate of the antibody/antigen complex. Due to availability, and ease of use, purified recombinant antigens are routinely used to measure affinity of antibodies. One critical element that is overlooked when using recombinant antigen is whether the protein is functionally active and whether the recombinant protein is a faithful representation of the conformation and the posttranslational modifications of the target in the cellular context. For soluble antigens, this can be readily addressed in cellular functional assays that can report out activity. For membranebound proteins and receptors the issue is more complex and much more problematic. In general, extracellular domains (ECD) of cell-bound receptors are usually expressed and utilized to monitor affinity of antibodies targeting the protein without fully appreciating whether the target antigen is being presented in a conformationally relevant manner. In most instances, it is unclear whether the native ligand can even bind to the representative ECD with the same or similar affinities as expected in whole cells and the problem is exacerbated when the antigens are orphan receptors due to the lack of knowledge of the receptor ligand which limits the testing of the purified receptor activity. [0134] The power of MSD for HTP analysis was applied in complex matrices for affinity evaluation of cell-surface targets in the physiologically relevant context of the cell. The inventors evaluated whether the affinities obtained for the recombinantly expressed antigen are representative of the affinities in whole cells and whether the relative affinities of various monospecific bivalent antibodies, its Fab, as well as engineered bispecific antibodies recognized recombinant receptor or cell surface associated receptor equivalently. To accomplish this, the MSD-CAT assay was developed which is capable of measuring affinity utilizing whole cells that present the relevant receptor.

[0135] Antibodies, one Fab, and derived BsAbs (mAbA, FabA, and the corresponding bispecifics) were evaluated using this procedure and the data were compared to those obtained by more traditional methods such as SPR. The comparison showed that while for one isoform of the target (SP2) the affinities to the cell-surface antigen were within 30 % of the affinities of the recombinant material, for the other isoform (SP1), the affinities for the cell- surface antigen were tighter than the recombinant material (Table 1). Without wishing to be bound by theory, these data support the supposition that the membrane-associated form of CD123 SP1 is presented in a different conformational structure (which is in the presence of membrane-associated lipids and proteins) than the recombinant form of CD123 SP1 used in SPR, which could account for the higher affinity for the membrane-associated antigen. CD123 SP1, in contrast to CD123 SP2, was poorly expressed as a recombinant protein and with an additional 79 amino acids involving 3 glycosylation sites, CD123 SP1 could have been potentially more difficult to express in a form representative of the protein on the cell surface than CD 123 SP2.

[0136] However, although this seems a plausible explanation, it is important to note that differences in these affinities could also be associated with different methodologies. This is suggested by the observation that differences in mAbA affinities for recombinant CD123 SP1 were observed between Biacore and ProteOn which are two SPR platforms. These instruments use different sensor chip matrices which could have different interactions with antigens and can influence association rates as we have observed for other antigens and others have observed between different sensor types (Drake, 2012). The MSD-CAT affinity data of mAbA for CD123 SP1 and CD123 SP2 are equivalent, and this suggests that both isoforms are presented in a similar structural conformation within the context of the cell membrane and that mAbA, recognizes both receptor isoforms similarly.

[0137] The MSD-CAT methodology presented in this report enables affinity determinations for cell-surface antigens that are contextually and physiologically relevant in the intact cell while taking advantage of the high-throughput capabilities of automated liquid handlers for sample preparation and the MSD Sector Imager 6000 with a high-capacity stacker for multi-plate detection. This procedure not only eliminates the need for protein purification for affinity determination purposes, but also eliminates the need for plasmid generation, transfection, and protein expression processes in situations where cells endogenously expressing the targets are available. Most importantly, MSD-CAT enables affinity determination of antigens for which production of properly folded and active recombinant form may not to be feasible.

Table 1: Biacore and MSD-CAT affinity data for the binding of anti-CD123 to human antigen-full length (“SP1”) and its truncated variant (“SP2”).

SPR SPR SPR MSD-CAT

Sample ID Antigen _{1 1 1} k _n (M s ) ± (SD) k _ff (s ) ± (SD) K_D (nM) ± SD K_D (nM) ± SD

SP1 (7.40 ± 0.64) x 10⁵ (7.48 ± 0.44) x 10'⁴ 0.94 ± 0.08 *0.023 ± 0.005 mAbA

SP2 (3.27 ± 0.95) x 10⁶ (9.12 ± 0.67) x 10⁴ 0.23 ± 0.06 *0.007 ± 0.005

SP1 (4.89-4.97) x 10⁵ (7.31-0.10) x 10⁴ 1.76 ± 0.41 0.079 ± 0.021

FabA

SP2 (2.44 ± 0.20) x 10⁶ (9.59 ± 0.33) x 10⁴ 0.39 ± 0.03 0.129 ± 0.086

SP1 (5.64± 0.38) x 10⁵ (8.30 ± 0.47) x 10⁴ 1.47 ± 0.13 0.078 ± 0.040

BsA1

SP2 (3.12± 0.63) x 10⁶ (1.10 ± 0.05) x 10³ 0.36± 0.07 0.145 ± 0.105

SP1 (5.62 ± 0.45) x 10⁵ (8.40 ± 0.53) x 10⁴ 1.49 ± 0.15 0.072 ± 0.034

BsA2

SP2 (3.33 ± 0.94) x 10⁶ (1.10 ± 0.02) x 10³ 0.34 ± 0.10 0.118 ± 0.0.082

SP1 (5.87 ± 0.46 x 10⁵ (7.90 ± 0.50) x 10⁴ 1.34 ± 0.14 0.072 ± 0.034

BsA3

SP2 (3.57 ± 0.98) x 10⁶ (1.10 ± 0.06) x 10³ 0.31 ± 0.09 0.124 ± 0.072

SP1 No binding No binding No binding No binding

I3CB15

SP2 No binding No binding No binding No binding mAbA Cy CD123 ND ND ND >11

SP1 ND ND ND 0.058

7G3

Cy CD123 ND ND ND 0.097

The values reported here correspond to the average and standard deviation obtained in three independent experiments, with exception of the Surface Plasmon Resonance (SPR) measurements for FabA binding to CD 123 SP1 which is the average of 2 experiments.

* This is called Apparent KD because it could be affected by avidity due to bivalent binding. ND = Not determined. Cyno CD 123 could not be expressed recombinantly.

[0138] Aspects of the present disclosure [0139] The present disclosure describes various aspects of the invention related to novel methods and compositions for analyzing antibody affinities to cell-surface targets. Key aspects include the development of high-throughput techniques for measuring antibody-antigen interactions in their native cellular environment, overcoming the challenges posed by solubilization and conformational integrity of cell-surface proteins. The invention also covers the use of advanced technologies for affinity determination that do not require complex receptor solubilization or immobilization of cells, offering a more efficient and scalable approach for identifying high-affinity therapeutic antibodies. Additionally, the invention addresses the need for accurate and sensitive assays that can be applied in both early-stage therapeutic antibody discovery and large-scale screening processes, ensuring the selection of optimal candidates for biotherapeutic development.

[0140] In an implementation, the disclosure provides a method for determining the affinity of an antibody to a cell-surface antigen comprising: (a) means for determining the affinity of the antibody to the cell-surface antigen by specific interaction of the antibody with the cell-surface antigen; and (b) providing a labeled agent that binds to the antibody.

[0141] In an implementation, the disclosure provides a method for determining the affinity of an antibody to a cell-surface antigen. The method generally comprises (a) means for determining the affinity of the antibody to the cell-surface antigen by specific interaction of the antibody with the cell-surface antigen; and (b) providing a labeled agent that binds to the antibody. For example, the method may comprise (a) incubating the antibody with the cell-surface antigen; (b) separating free antibody in solution from antibody that is bound to the cell-surface antigen; (c) adding the free antibody in solution to a solid support comprising a fixed antigen and incubating; (d) adding a labeled agent that binds to antibody that is bound to the fixed antigen; (e) removing free labeled agent; and (f) measuring electrochemiluminescence of the labeled agent that is bound to the antibody that is bound to the fixed antigen to generate electrochemiluminescence data.

[0142] According to aspects of the implementations, the method is in high throughput format.

[0143] According to any aspect of the implementations, the method is automated and/or continuous.

[0144] According to any aspect of the implementations, the automation is controlled by a central processor, for example a computer.

[0145] According to any aspect of the implementations, the automation is robotic. [0146] According to any aspect of the implementations, the method further comprises calculating the binding affinity of the antibody to the cell-surface antigen by analyzing the electrochemiluminescence data. For example, the electrochemiluminescence data may be used to calculate percent free antibody.

[0147] According to any aspect of the implementations, the method further comprises (g) calculating binding affinity of the antibody to the cell-surface antigen by analyzing the electrochemiluminescence data, wherein the calculating comprises an equation set forth in: wherein the electrochemiluminescence data is used to calculate percent free antibody, wherein the percent free antibody is calculated according to an equation comprising: . > _ 100.

[0148] According to any aspect of the implementations, the cell-surface antigen in step (a) is expressed on a cell.

[0149] According to any aspect of the implementations, the fixed antigen comprises an epitope that is also on the cell-surface antigen.

[0150] According to any aspect of the implementations, the fixed antigen is biotinylated and is bound to streptavidin that is bound to a solid surface or plate.

[0151] According to any aspect of the implementations, the labeled agent is a ruthenylated antibody.

[0152] According to any aspect of the implementations, the antibody to a cell-surface antigen is a monoclonal antibody, a Fab fragment, an scFv, a bispecific antibody, or a multispecific antibody.

[0153] According to any aspect of the implementations, the antibody to a cell-surface antigen is a human antibody.

[0154] According to any aspect of the implementations, the cell-surface antigen is a receptor.

[0155] According to any aspect of the implementations, separating free antibody in solution from antibody that is bound to the cell-surface antigen, i.e., step (b), is done by centrifugation. [0156] According to any aspect of the implementations, the method further comprises washing with a buffer after step (c), and optionally repeating the wash 2-3 times.

[0157] According to any aspect of the implementations, the method further comprises washing with a buffer after step (d), and optionally repeating the wash 2-3 times.

[0158] According to any aspect of the implementations, the method further comprises washing with a buffer after step (e), and optionally repeating the wash 2-3 times.

[0159] According to any aspect of the implementations, electrochemiluminescence is measured by an electrochemiluminescence reader.

[0160] According to any aspect of the implementations, the binding affinity of the antibody to the cell-surface receptor is calculated by a central processor, for example a computer.

[0161] According to any aspect of the implementations, the binding affinity of the antibody to the cell-surface receptor may be calculated by a central processor, for example a computer, wherein the binding affinity of the antibody is calculating using equation:

[0162] According to any aspect of the implementations, the cell-surface antigen concentration may be calculated by a central processor, for example a computer, wherein the calculation is performed using an equation: wherein a fitted parameter “n” is used to calculate the cell-surface antigen concentration by multiplying an estimated concentration of antigen that was entered at a beginning of a data fitting procedure by “n.”

REFERENCES

[0163] Bator, J. M. & Reading, C. L. Measurement of antibody affinity for cell surface antigens using an enzyme-linked immunosorbent assay. J Immunol Methods 125, 167-176, doi: 10.1016/0022-1759(89)90090-2 (1989). [0164] Bee, C. A., YN; Stone, DM; Collier, S; Lindquist, KC; Pinkerton, AC; Pons, J; Rajpal, A. Exploring the Dynamic Range of the Kinetic Exclusion Assay in Characterizing Antigen- Antibody Interactions. PLoS ONE 7, e36261 (2012).

[0165] Benedict, C. A., MacKrell, A. J. & Anderson, W. F. Determination of the binding affinity of an anti-CD34 single-chain antibody using a novel, flow cytometry based assay. J Immunol Methods 201, 223-231, doi:10.1016/s0022-1759(96)00227-x (1997).

[0166] Bondza, S. et al. Real-time Characterization of Antibody Binding to Receptors on Living Immune Cells. Frontiers in Immunology 8, doi:10.3389/fimmu.2017.00455 (2017).

[0167] Chen, J. et al. A new isoform of interleukin-3 receptor {alpha} with novel differentiation activity and high affinity binding mode. J Biol Chem 284, 5763-5773, doi: 10.1074/jbc.M808197200 (2009).

[0168] Chichili, G. R. et al. A CD3xCD123 bispecific DART for redirecting host T cells to myelogenous leukemia: preclinical activity and safety in nonhuman primates. Sci Transl Med 7, 289ra282, doi:10.1126/scitranslmed.aaa5693 (2015).

[0169] Debbia, M. & Lambin, P. Measurement of anti-D intrinsic affinity with unlabeled antibodies. Transfusion 44, 399-406, doi: 10.1111/j.1537-2995.2004.00651. x (2004).

[0170] Della Ducata, D. et al. Solution Equilibrium Titration for High-Throughput Affinity Estimation of Unpurified Antibodies and Antibody Fragments. J Biomol Screen 20, 1256-1267, doi: 10.1177/1087057115595002 (2015).

[0171] Drake, A. W. et al. Biacore surface matrix effects on the binding kinetics and affinity of an antigen/antibody complex. Anal Biochem 429, 58-69, doi:10.1016/j.ab.2012.06.024 (2012).

[0172] Encarnacao, J. C., Schulte, T., Achour, A., Bjorkelund, H. & Andersson, K. Detecting ligand interactions in real time on living bacterial cells. Appl Microbiol Biotechnol 102, 4193- 4201, doi:10.1007/s00253-018-8919-3 (2018).

[0173] Estep, P. et al. High throughput solution-based measurement of antibody-antigen affinity and epitope binning. MAbs 5, 270-278, doi:10.4161/mabs.23049 (2013).

[0174] Gramer, M. J. et al. Production of stable bispecific IgGl by controlled Fab-arm exchange: scalability from bench to large-scale manufacturing by application of standard approaches. MAbs 5, 962-973, doi:10.4161/mabs.26233 (2013).

[0175] Itoh, N. et al. Cloning of an interleukin-3 receptor gene: a member of a distinct receptor gene family. Science 247, 324-327, doi:10.1126/science.2404337 (1990). [0176] Jarantow, S. W. et al. Impact of Cell-surface Antigen Expression on Target Engagement and Function of an Epidermal Growth Factor Receptor x c-MET Bispecific Antibody. J Biol Chem 290, 24689-24704, doi:10.1074/jbc.M115.651653 (2015).

[0177] Jing, W., Hunt, A., Tao, N., Zhang, F. & Wang, S. Simultaneous Quantification of Protein Binding Kinetics in Whole Cells with Surface Plasmon Resonance Imaging and Edge Deformation Tracking. Membranes (Basel) 10, doi:10.3390/membranesl0090247 (2020).

[0178] Jordan, C. T. et al. The interleukin-3 receptor alpha chain is a unique marker for human acute myelogenous leukemia stem cells. Leukemia 14, 1777-1784, doi:10.1038/sj.leu.2401903 (2000).

[0179] Kontermann, R. E. Dual targeting strategies with bispecific antibodies. MAbs 4, 182-197, doi: 10.4161/mabs.4.2.19000 (2012).

[0180] Labrijn, A. F. et al. Efficient generation of stable bispecific IgGl by controlled Fab-arm exchange. Proc Natl Acad Sci U SA 110, 5145-5150, doi:10.1073/pnas.1220145110 (2013).

[0181] Lundstrom, K. Structural genomics of GPCRs. Trends Biotechnol 23, 103-108, doi: 10.1016/j .tibtech.2004.12.006 (2005) .

[0182] Myszka, D. G. Improving biosensor analysis. J Mol Recognit 12, 279-284 (1999).

[0183] Myszka, D. G. in Methods in Enzymology Vol. 323 325-340 (Academic Press, 2000).

[0184] Navratilova, I., Besnard, J. & Hopkins, A. L. Screening for GPCR Ligands Using Surface Plasmon Resonance. ACS Med Chem Lett 2, 549-554, doi:10.1021/ml2000017 (2011).

[0185] Ohmura, N., Lackie, S. J. & Saiki, H. An immunoassay for small analytes with theoretical detection limits. Anal Chem L , 3392-3399, doi:10.1021/ac001328d (2001).

[0186] Paul, S. et al. An efficient process of generating bispecific antibodies via controlled Fabarm exchange using culture supernatants. Protein Expression and Purification 121, 133-140, (2016).

[0187] Rathanaswami, P., Babcook, J. & Gallo, M. High-affinity binding measurements of antibodies to cell-surface-expressed antigens. Anal Biochem 373, 52-60, doi: 10.1016/j.ab.2007.08.014 (2008).

[0188] Rogers, G. T., Rawlins, G. A., Kardana, A. & Gibbons, A. R. Binding studies on two different monoclonal antibodies raised against CEA. Eur J Cancer Clin Oncol 19, 629-639, doi: 10.1016/0277-5379(83)90179-7 (1983). [0189] Shepherd, C. A., Hopkins, A. L. & Navratilova, I. Fragment screening by SPR and advanced application to GPCRs. Prog Biophys Mol Biol 116, 113-123, doi: 10.1016/j .pbiomolbio.2014.09.008 (2014).

[0190] Shi, L. et al. De Novo Selection of High- Affinity Antibodies from Synthetic Fab Libraries Displayed on Phage as pIX Fusion Proteins. Journal of Molecular Biology 397, 385- 396 (2010).

[0191] Siiman, O. & Burshteyn, A. Cell surface receptor- antibody association constants and enumeration of receptor sites for monoclonal antibodies. Cytometry 40, 316-326 (2000).

[0192] Testa, U., Pelosi, E. & Frankel, A. CD 123 is a membrane biomarker and a therapeutic target in hematologic malignancies. Biomarker Research 2, 4, doi:10.1186/2050-7771-2-4 (2014).

[0193] Testa, U., Pelosi, E. & Castelli, G. CD 123 as a Therapeutic Target in the Treatment of Hematological Malignancies. Cancers (Basel) 11, doi: 10.3390/cancersl 1091358 (2019).

[0194] Thakur, A. & Lum, L. G. "NextGen" Biologies: Bispecific Antibodies and Emerging Clinical Results. Expert Opin Biol Ther 16, 675-688, doi:10.1517/14712598.2016.1150996 (2016).

[0195] Wang, W. et al. Label-free measuring and mapping of binding kinetics of membrane proteins in single living cells. Nat Chem 4, 846-853, doi:10.1038/nchem.l434 (2012).

[0196] Xie, L. et al. Measurement of the functional affinity constant of a monoclonal antibody for cell surface receptors using kinetic exclusion fluorescence immunoassay. J Immunol Methods 304, 1-14, doi:10.1016/j.jim.2005.04.009 (2005).

[0197] Zhang, P. et al. In Situ Analysis of Membrane-Protein Binding Kinetics and Cell-Surface Adhesion Using Plasmonic Scattering Microscopy. Angew Chem Int Ed Engl 61, e202209469, doi: 10.1002/anie.202209469 (2022).

[0198] Zhao, Y. et al. Two routes for production and purification of Fab fragments in biopharmaceutical discovery research: Papain digestion of mAb and transient expression in mammalian cells. Protein Expression and Purification 67, 182-189, (2009).

[0199] Zorn, J. A. et al. Humanization of a strategic CD3 epitope enables evaluation of clinical T-cell engagers in a fully immunocompetent in vivo model. Sei Rep 12, 3530, doi: 10.1038/s41598-022-06953-7 (2022). [0200] All publications and patents referred to herein are incorporated by reference. Various modifications and variations of the described subject matter will be apparent to those skilled in the art without departing from the scope and spirit of the invention. Although the invention has been described in connection with specific embodiments, it should be understood that the invention as claimed should not be unduly limited to these embodiments. Indeed, various modifications for carrying out the invention are obvious to those skilled in the art and are intended to be within the scope of the following claims.

Claims

CLAIMS What is claimed is:

1. A method for determining affinity of an antibody to a cell-surface antigen comprising:

(a) incubating the antibody with the cell-surface antigen;

(b) separating free antibody in solution from antibody that is bound to the cell-surface antigen;

(c) adding the free antibody in solution to a solid support comprising a fixed antigen and incubating;

(d) adding a labeled agent that binds to antibody that is bound to the fixed antigen;

(e) removing free labeled agent; and

(f) measuring electrochemiluminescence of the labeled agent that is bound to the antibody that is bound to the fixed antigen to generate electrochemiluminescence data.

2. The method of claim 1, wherein the method is in high throughput format.

3. The method of claim 1, wherein the method is automated and/or continuous.

4. The method of claim 3, wherein the automation is robotic.

5. The method of claim 3, wherein the automation is controlled by a central processor, for example a computer.

6. The method of claim 5, wherein the automation is robotic.

7. The method of claim 1, further comprising:

(g) calculating binding affinity of the antibody to the cell-surface antigen by analyzing the electrochemiluminescence data, wherein the calculating comprises an equation set forth in:

8. The method of claim 7, wherein the electrochemiluminescence data is used to calculate percent free antibody, wherein the percent free antibody is calculated according to an equation comprising: . > _ 100.

9. The method of claim 1, wherein the cell-surface antigen in step (a) is expressed on a cell.

10. The method of claim 1, wherein the fixed antigen comprises an epitope that is also on the cell-surface antigen.

11. The method of claim 1, wherein the fixed antigen is biotinylated and is bound to streptavidin that is bound to a solid surface or plate.

12. The method of claim 1, wherein the labeled agent is a ruthenylated antibody.

13. The method of claim 1, wherein the antibody to a cell-surface antigen is a monoclonal antibody, a Fab fragment, an scFv, a bispecific antibody, or a multispecific antibody.

14. The method of claim 13, wherein the antibody to a cell-surface antigen is a human antibody.

15. The method of claim 1, wherein the cell-surface antigen is a receptor.

16. The method of claim 1, wherein in step (b) separating free antibody in solution from antibody that is bound to the cell-surface antigen is done by centrifugation.

17. The method of claim 1, further comprising washing with a buffer after step (c), optionally repeating the wash 2-3 times.

18. The method of claim 1, further comprising washing with a buffer after step (d), optionally repeating the wash 2-3 times.

19. The method of claim 1, further comprising washing with a buffer after step (e), optionally repeating the wash 2-3 times.

20. The method of claim 1, wherein the electrochemiluminescence is measured by an electrochemiluminescence reader.

21. The method of claim 1, wherein binding affinity of the antibody to the cell-surface receptor is calculated by a central processor, for example a computer, wherein the binding affinity of the antibody is calculating using equation:

, (Smax-NSB) ,_r ^AT-i rr. .

MSD Signal = — — * ([— ] - K_D

²I-DFI - [B_r]) + + NSB.

22. The method of claim 1, wherein the cell-surface antigen concentration is calculated by a central processor, for example a computer, wherein the calculation is performed using an equation: wherein a fitted parameter “n” is used to calculate the cell-surface antigen concentration by multiplying an estimated concentration of antigen that was entered at a beginning of a data fitting procedure by “n.”