System and method for epitope mapping and complex stability determination

Abstract

This disclosure relates methods and systems for the biophysical characterization of protein target and binder interactions, such as antibody and target antigen interactions. The systems and methods allow for the determination of the epitope map at the amino acid level, identify the weak interactions that govern binding and assess the stability7 of the protein-target complex.

Description

Attorney Docket No.: 761349.000125 SYSTEM AND METHOD FOR EPITOPE MAPPING AND COMPLEX STABILITY DETERMINATION Related Applications

[0001] This application claims the benefit of United States Provisional Patent ApplicationNo.63 / 637,027 filed on April 22, 2024 the entire disclosure of which is incorporated herein by reference. Field of the Invention

[0002] This disclosure relates to a system and method for the comprehensive and rapiddetermination of the epitope map of antibodies defined with amino acid level resolution, including for antibodies that are candidates for development as therapeutics (antibody candidates). BACKGROUND

[0003] To date, more than 100 therapeutic proteins have been approved by the FDA. In 2023,22% of the drugs approved belonged to the clinical antibody class. Yet the regulatory approval failure rate of therapeutic protein candidates is still high, within 68- 86%, and the majority of failures occur within the clinical phase II and III trials for varied reasons. Thus, considerable effort has been devoted to the determination of critical quality attributes that have been identified for each therapeutic candidate to aid in ensuring safety and efficacy. However, a greater understanding of the candidate-target interaction is also needed to de-risk even further decision-making of pre-clinical candidate selection. A series of factors must be taken into account to develop a rational approach and the method must be able to determine: (1) the molecular complexity of the candidate as well as the target, (2) the stability of the pure components, both candidate and target, (3) the extent of glycosylation, (4) presence of degradation products of the candidate or target, (5) presence of aggregates whether from the target or the candidate to allow for success in the determination of a functional complex, (6) the target epitope with amino acid level resolution, (7) the optimal mol ratio of the complex, (8) the requirement of glycan binding for complex formation and (9) the stability of the complex. Ideally, the method should be agnostic to therapeutic protein type and similar considerations should govern the target. 1 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125

[0004] However, the current tools available to biopharma and academic researchers ingeneral are labor intensive, time-consuming and expensive, at times resulting in limited success due to the complexity of the proteinaceous sample mixture or the limitations of the analytical technique used for the biophysical characterization of the candidate-target complex. For example, existing analytical techniques such as Cryo-EM, X-ray crystallography, and NMR are some of the established high-resolution techniques that are at times limited due to the extent of glycosylation of both the target and the candidate, the extent of the disordered structure of the target or simply the size and symmetry of the protein- complex. Other techniques such as circular dichroism (CD) can only be used to determine the average secondary structure of the complex and that of the pure protein components, yet no insight towards the interaction interface can be attained. Other techniques commonly used to determine dissociation constants, also referred to as affinity constants, such as surface plasmon resonance (SPR), isothermal titration calorimetry (ITC), biolayer interferometry (BLI), and to a lesser extent phage display provide insight regarding the affinity the candidate has for the target, but do not provide a direct determination of the target epitope site. Also, differential scanning calorimetry (DSC) provides the overall thermal stability and thermodynamic parameters that are used to quantitate the stability of the pure component with respect to the complex. Biochemical methods have also been implemented routinely, they involve generating truncated peptide sequences of the proposed epitope or alanine scanning mutations or the generation of variants to interrogate the binding interface. They all have the limitation of secondary structure uncertainty brought abought by the introduction of the point mutation or the truncation.

[0005] A need exists for systems and methods for the comprehensive and rapid determinationof the epitope map of antibodies defined with amino acid level resolution. SUMMARY

[0006] This disclosure relates to a system and method for the comprehensive and rapiddetermination of the epitope map of antibodies defined with amino acid level resolution, including for antibodies that are candidates for development as therapeutics (antibody candidates). The epitope map can be determined for multiple antibody candidates in an array against the same target / antigen. The comparative analysis is also amenable to the determination of the optimal target / candidate mol ratio, key interactions that stabilize the complex, and the assessment of the candidate-target complex relative stability within the 2 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 array during a controlled perturbation, such as a thermal perturbation. The pure components are also evaluated within the same array ensuring true comparative analysis for decision making. The comparative analysis involves the acquisition of hyperspectral images under controlled perturbation conditions, such as at different set temperatures, and the associated microspectroscopic spectral data of the complex is analyzed against the pure candidate and target. The reference in these studies will be the pure antibody candidate and target, to allow for the determination of its target epitope. The candidate is not limited to monoclonal antibodies, but any new modality thereof. Furthermore, the method is not limited by the target, which can be a druggable or undruggable allowing for an unprecedented level of molecular understanding to enable drug discovery and design.

[0007] In embodiments, this disclosure relates to methods and systems for analyzingantibodies and determining the epitope of an antibody at amino acid level resolution.

[0008] In aspects, the method comprises a) providing spectral data for test samples andreference samples. The test data is preferably obtained using a quantum cascade laser microscope under controlled temperature conditions. The test samples preferably comprise an antibody or antigen-binding fragment thereof and an antigen that the antibody binds with immunological specificity. The reference samples preferably comprise antibody reference samples, which comprises the same antibody or antigen-binding fragment as the test samples but not the antigen. Optionally, and preferably, the reference samples also include antigen reference samples that comprises the same antigen as the test samples but not the antibody or antigen-binding fragment. The spectral data is preferably obtained using a single slide cell array comprising cells that contain test samples and cells that contain reference samples, and the spectral data include spectral images of each test sample and reference sample. The method further comprises b) identifying and selecting in spectral images of a test sample a region of interest and identifying and selecting in spectral images of an antibody reference sample a region of interest; c) obtaining spectral data for the test sample and the antibody reference sample for the region of interest; d) applying a baseline correction to the spectral data for the test sample and the antibody reference sample for the region of interest; e) generating weighted difference spectra for the test sample and antibody reference sample; andf) applying a two-trace two-dimensional correlation to the weighted difference spectra togenerate a synchronous plot , ) and an asynchronous plot , ), and identifyingsignature cross peaks in the asynchronous plot to determine theor epitopes of the antibody or antigen-binding fragment. Preferably the method further comprises g) providing 3 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 baseline corrected spectra for each test sample and each antibody reference sample, and optionally for each antigen reference sample, and generating covariance data, and applying a two-dimensional correlation analysis to the covariance data to generate a synchronous correlation plot and an asynchronous correlation plot for each sample, identifying signature peaks in the plots, and analyze intensity changes of the peaks to define a sequential order of molecular events; and h) applying a two-dimensional co-distribution analysis to the baseline corrected spectra for each test sample and each antibody reference sample, and optionally for each antigen reference sample, to generate joint variance spectra and apply co-distribution functions to generate an asynchronous co-distribution plot and optionally a synchronous co- distribution plot, and identifying cross peaks in the asynchronous co-distribution plots, evaluating cross peaks in the asynchronous co-distribution plot, identify cross peaks associated with key interactions between the antigen and the antibody or antigen-binding fragment, and determine relative stability of the complex of the antibody or antigen-binding fragment bound to the epitope or epitopes.

[0009] In aspects, the method comprises a) providing a slide cell array comprising cells thatcontain test samples and cells that contain reference samples. The test samples preferably comprise an antibody or antigen-binding fragment thereof and an antigen that the antibody binds with immunological specificity. The reference samples preferably comprise antibody reference samples that comprises the same antibody or antigen-binding fragment as test samples but not the antigen. Optionally, and preferably, the reference samples include antigen reference samples that comprises the same antigen as the test samples but not the antibody or antigen-binding fragment. The method further comprises b) acquiring spectral data for the test samples and reference samples using a quantum cascade laser microscope under controlled temperature conditions. The spectral data preferably include spectral images of each test sample and reference sample. The method further comprises c) identifying and selecting in spectral images of a test sample a region of interest and identifying and selecting in spectral images of an antibody reference sample a region of interest; d) obtaining spectral data for the test sample and the antibody reference sample for the region of interest; e) applying a baseline correction to the spectral data for the test sample and the antibody reference sample for the region of interest; f) generating weighted difference spectra for the test sample and antibody reference sample; and g) applying a two-trace two-dimensionalcorrelation to the weighted difference spectra to generate a synchronous plot , ) and anasynchronous plot , ), and identifying signature cross peaks in the asynchronous plot4 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 to determine the epitope or epitopes of the antibody or antigen-binding fragment. Preferably the method further comprises h) providing baseline corrected spectra for each test sample and each antibody reference sample, and optionally for each antigen reference sample, and generating covariance data, and applying a two-dimensional correlation analysis to the covariance data to generate a synchronous correlation plot and an asynchronous correlation plot for each sample, identifying signature peaks in the plots, and analyze intensity changes of the peaks to define a sequential order of molecular events; and i) applying a two- dimensional co-distribution analysis to the baseline corrected spectra for each test sample and each antibody reference sample, and optionally for each antigen reference sample, to generate joint variance spectra and apply co-distribution functions to generate an asynchronous co- distribution plot and optionally a synchronous co-distribution plot, and identifying cross peaks in the asynchronous co-distribution plots, evaluating cross peaks in the asynchronous co-distribution plot, identify cross peaks associated with key interactions between the antigen and the antibody or antigen-binding fragment, and determine relative stability of the complex of the antibody or antigen-binding fragment bound to the epitope or epitopes.

[0010] In embodiments of the methods disclosed herein, the test samples include two or moresamples in which the mol ratio of antigen:antibody differ. For example, the mol ratio of antigen:antibody can be from about 1:1 to about 10:1. Preferably, the mol ratio of antigen:antibody is sufficient to saturate the antibody binding sites in the test sample. The antibody reference samples can comprises samples in which the antibody concentration is the same as the antibody concentration in a corresponding test sample. The antigen reference samples comprises samples in which the antigen concentration is the same as the antigen concentration in a corresponding test sample. The test samples and reference samples are preferably aqueous solutions.

[0011] In embodiments of the methods disclosed herein, the controlled temperatureconditions comprise a temperature from about 4°C to about 95°C. Preferably, the spectral data include data collected at two or more temperatures, e.g., at three or more, four or more, five or more, six or more, seven or more, eight or more, nine or more, ten or more temperatures. The two or more temperatures can be within the range of about 24°C to about 36°C or about 24°C to about 52°C.

[0012] Typically, the antigen is a soluble protein, the extracellular region of a membraneassociated protein or a portion of either of the foregoing. The soluble protein or membrane associate protein can be a glycoprotein. 5 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125

[0013] In embodiments, the methods further comprise determining one or more of Asn and / orGln deamidation, self-association, aggregation and stability of an antibody reference sample or antigen reference sample using both two-dimensional co-distribution and two-dimensional correlation analysis, and / or further comprise applying a two-trace two-dimensionalcorrelation analysis to the test sample, antibody reference sample and / or antigen referencesample to determine the extent of glycosylation, molecular nature of glycan interaction with antibody or antigen-binding fragment during formation of the complex between antibody or antigen-binding fragment and antigen and influence on stability of the complex between antibody or antigen-binding fragment and antigen, and / or further comprise applying a two-trace two-dimensional correlation analysis to the test sample, antibody reference sampleand / or antigen reference sample to determine the molecular nature of antibody or antigen- binding fragment and antigen complex formation and / or dissociation.

[0014] This disclosure also relates to a system for analyzing an antibody and / or determiningthe epitope of an antibody. The system comprises a) a data acquisition module configured to obtain spectral data of test samples and reference samples under controlled temperature conditions. The test samples preferably comprise an antibody or antigen-binding fragment thereof and an antigen that the antibody binds with immunological specificity. The reference samples preferably comprise antibody reference samples, which comprises the same antibody or antigen-binding fragment as the test samples but not the antigen. Optionally, and preferably, the reference samples also include antigen reference samples that comprises the same antigen as the test samples but not the antibody or antigen-binding fragment. The spectral data is preferably obtained using a single slide cell array comprising cells that contain test samples and cells that contain reference samples, and the spectral data include spectral images of each test sample and reference sample. The system further comprises b) a correlation analysis module configured to: generate weighted difference spectra for the testsample and antibody reference sample, apply a two-trace two-dimensional correlation to theweighted difference spectra to generate a synchronous plot , ) and an asynchronousplot , ), and identifying signature cross peaks inplot to determinethe epitope or epitopes of the antibody or antigen-binding fragment. The correlation analysis module can be further configured to apply a two-dimensional correlation analysis to covariance data to generate a synchronous correlation plot and an asynchronous correlation plot for each sample, identify signature peaks in the plots, and analyze intensity changes of the peaks to define a sequential order of molecular events; and / or apply a two-dimensional 6 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 co-distribution analysis to joint variance spectra for each test sample and each antibody reference sample, and optionally for each antigen reference sample, to generate an asynchronous co-distribution plot and optionally a synchronous co-distribution plot, and identifying cross peaks in the asynchronous co-distribution plots, evaluating cross peaks in the asynchronous co-distribution plot, identify cross peaks associated with key interactions between the antigen and the antibody or antigen-binding fragment, and determine relative stability of the complex of the antibody or antigen-binding fragment bound to the epitope or epitopes.

[0015] This disclosure also relates to a non-transitory computer-readable medium comprisinginstructions which, when executed by one or more computers, cause the one or more computers to a) obtain spectral data of test samples and reference samples under controlled temperature conditions. The test samples preferably comprise an antibody or antigen-binding fragment thereof and an antigen that the antibody binds with immunological specificity. The reference samples preferably comprise antibody reference samples that comprises the same antibody or antigen-binding fragment as the test samples but not the antigen, and optionally, and preferably, antigen reference samples that comprises the same antigen as the test samples but not the antibody or antigen-binding fragment. The spectral data is preferably obtained using a single slide cell array comprising cells that contain test samples and cells that contain reference samples; and the spectral data include spectral images of each test sample and reference sample.

[0016] The non-transitory computer-readable medium further comprises instructions which,when executed by one or more computers, cause the one or more computers to b) generate weighted difference spectra for the test sample and antibody reference sample; and c) apply atwo-trace two-dimensional correlation to the weighted difference spectra to generate asynchronous plot , ) and an asynchronous plot , ), and identifying signaturecross peaks in theplot to determine the epitope or epitopes of the antibody or antigen-binding fragment. Optionally, the non-transitory computer-readable medium further comprises instructions which, when executed by one or more computers, cause the one or more computers to d) apply a two-dimensional correlation analysis to covariance data to generate a synchronous correlation plot and an asynchronous correlation plot for each sample, identify signature peaks in the plots, and analyze intensity changes of the peaks to define a sequential order of molecular events; and / or e) apply a two-dimensional co- distribution analysis to baseline corrected spectra for each test sample and each antibody 7 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 reference sample, and optionally for each antigen reference sample, to generate joint variance spectra and apply co-distribution functions to generate an asynchronous co-distribution plot and optionally a synchronous co-distribution plot, and identifying cross peaks in the asynchronous co-distribution plots, evaluating cross peaks in the asynchronous co- distribution plot, identify cross peaks associated with key interactions between the antigen and the antibody or antigen-binding fragment, and determine relative stability of the complex of the antibody or antigen-binding fragment bound to the epitope or epitopes. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] studies disclosed herein.

[0018] FIG. 2A is an illustration of the co-distribution asynchronous plot within the spectralregion of 1780-1490 cm-1

[0019] FIG. 2B is an illustration of the 2D-COS synchronous plot within the spectral regionof 1780-1490 cm-1

[0020] FIG. 2C is an illustration of the 2D-COS asynchronous plot within the spectral regionof 1780-1490 cm-1

[0021] FIG. 3A is an illustration of the co-distribution asynchronous plot within the spectralregion of 1780-1490 cm-1and temperature range of 24°C-52°C for pure mAb1.

[0022] FIG. 3B is an illustration of the 2D-COS synchronous plot within the spectral regionof 1780-1490 cm-1for pure mAb1.

[0023] FIG. 3C is an illustration of the 2D-COS asynchronous plot within the spectral regionof 1780-1490 cm-1for pure mAb1.

[0024] FIG. 4A is an illustration of the co-distribution asynchronous plot within the spectralregion of 1780-1490 cm-1and temperature range of 24°C-52°C for pure mAb2.

[0025] FIG. 4B is an illustration of the 2D-COS synchronous plot within the spectral regionof 1780-1490 cm-1for pure mAb2.

[0026] FIG. 4C is an illustration of the 2D-COS asynchronous plot within the spectral regionof 1780-1490 cm-1for pure mAb2.

[0027] FIG. 5A is an illustration of the co-distribution asynchronous plot within the spectralregion of 1780-1490 cm-1and temperature range of 24°C-52°C for pure mAb3.

[0028] FIG. 5B is an illustration of the 2D-COS synchronous plot within the spectral regionof 1780-1490 cm-1for pure mAb3. 8 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125

[0029] FIG. 5C is an illustration of the 2D-COS asynchronous plot within the spectral regionof 1780-1490 cm-1for pure mAb3.

[0030] FIG. 6A is an illustration of the co-distribution asynchronous plot within the spectralregion of 1780-1490 cm-1and temperature range of 24°C-52°C for pure mAb4.

[0031] FIG. 6B is an illustration of the 2D-COS synchronous plot within the spectral regionof 1780-1490 cm-1for pure mAb4.

[0032] FIG. 6C is an illustration of the 2D-COS asynchronous plot within the spectral regionof 1780-1490 cm-1for pure mAb4.

[0033] FIG. 7 is a table presenting the sequential order of molecular events (SOME) for pure

[0034] FIG. 8A is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb1complex at 24oC comprised of cross peaks that reveal the amino acids associated with theSIRP- epitope for mAb1. The mAb1 epitope involves the Asn, His, Arg and Lys residues inthe loop and random coil regions at 24oC.

[0035] FIG. 8B is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb1complex at 36oC comprised of cross peaks that reveal the amino acids associated with theSIRP- epitope for mAb1. The mAb1 epitope involves the Asn, His, Arg and Lys residues inthe loop and random coil regions at 36oC.

[0036] FIG. 8C is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb1complex at 52oC comprised of cross peaks that reveal the amino acids associated with theSIRP- epitope for mAb1. The mAb1 epitope involves the Asn, His, Arg and Lys residues inthe loop and random coil regions at 52oC. mAb1 began to destabilize at 52oC.

[0037] FIG. 8D is an illustration of the co-distribution asynchronous plot within the amide Iand amide II bands spectral region of 1780-1490 cm-1and temperature range of 24oC-52oCfor the SIRP- / mAb1 complex at 2:1 (SIRP- :mAb1) mol ratio. The SIRP- / mAb1complex exhibits similar stability as pure mAb1.

[0038] FIG. 8E is an illustration of the mAb1 epitope on SIRP- . mAb1 has the highestaffinity of the mAbs analyzed and binds an epitope located at Domain 1 (epitope residues listed in a clockwise direction from bottom right: R46, N51, Q52, K53, and H56, side chains shown) all residues are surface accessible and exposed within random coil and loop structure in accordance with 2T2D results obtained (Table IIA). Q52 was observed at high 9 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 temperature, suggesting it plays a role in stabilizing the complex at high temperature. TheKD of mAb1 for SIRP- is 8 nM.

[0039] FIG. 9A is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb2complex at 24oC comprised of cross peaks that reveal the amino acids associated with theSIRP- epitope for mAb1. The mAb2 epitope involves the Asn, His, Arg and Lys residues inthe loop and random coil regions at 24oC.

[0040] FIG. 9B is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb2complex at 36oC comprised of cross peaks that reveal the amino acids associated with theSIRP- epitope for mAb2. The mAb2 epitope involves the Asn, His, Arg and Lys residues inthe loop and random coil regions at 36oC.

[0041] FIG. 9C is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb2complex at 52oC comprised of cross peaks that reveal the amino acids associated with theSIRP- epitope for mAb2. The cross peaks show disruption of Asn51 and inter-molecular H-bonding interaction at 52oC causing the interaction to be destabilized at high temperature.

[0042] FIG. 9D is an illustration of the co-distribution asynchronous plot within the amide Iand amide II bands spectral region of 1780-1490 cm-1and temperature range of 24oC-52oCfor the SIRP- / mAb2 complex at 2:1 (SIRP- :mAb2) mol ratio. The white circle highlightswith white crosshair indicate cross peaks associated with the perturbation of the Asp / Lys salt-bridge formed in the SIRP- / mAb2 complex that led to instability of the complex.

[0043] FIG. 9E is an illustration of the mAb2 epitope on SIRP- . The mAb2 epitope islocated at Domain 1 (epitope residues listed in a clockwise direction from bottom right: R46, K53, and H56, side chains shown) all residues are surface accessible and exposed within random coil and loop structure in accordance with 2T2D results obtained (Table IIA). Q52 was observed at high temperature, suggesting it plays a role in stabilizing the complex at hightemperature. The KD of mAb2 for SIRP- is 72 nM, about 9x lower affinity than mAb1 forthe same site on SIRP- .

[0044] FIG. 10A is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb3complex at 24oC comprised of cross peaks that reveal the amino acids associated with theSIRP- epitope for mAb3. As seen in FIGs. 10B and 10C the cross peak pattern changes as afunction of temperature, suggesting instability and potentially more than one binding site for the mAb3. 10 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125

[0045] FIG. 10B is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb3complex at 36oC comprised of cross peaks that reveal the amino acids associated with theSIRP- epitope for mAb3.

[0046] FIG. 10C is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb3complex at 52oC comprised of cross peaks that reveal the amino acids associated with theSIRP- epitope for mAb3.

[0047] FIG. 10D is an illustration of the co-distribution asynchronous plot within the amideI and amide II bands spectral region of 1780-1490 cm-1and temperature range of 24oC-52oCfor the SIRP- / mAb3 complex at 2:1 (SIRP- :mAb3) mol ratio. The SIRP- / mAb3complex exhibits similar stability as pure mAb3 with a lysine, glutamate (cross peaks highlighted with a white circle) suggesting a salt bridge interaction that is perturbed during the thermal perturbation.

[0048] FIG. 10E is an illustration of the mAb3 epitope on SIRP- . mAb3 binds a site 1epitope in Domain 1 (site 1 epitope residues listed in a clockwise direction from bottom right: R46, N51, Q52, K53, and H56, side chains shown) and a site two epitope in Domain 1 at higher temperature (site 2 epitope residues listed in a clockwise direction from bottom right: K53, H56, R59 and N80, side chains shown). All epitope residues are surface accessible and exposed. Q52 was not observed above physiological temperature. The results suggest secondary structure changes upon binding of mAb3 to the site 2 epitope. The KD of mAb3for SIRP- is 43 nM, about 5x lower affinity than mAb1.

[0049] FIG. 11A is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb4complex at 24oC comprised of cross peaks that reveal the amino acids associated with theSIRP- epitope for mAb4. The mAb4 epitope at 24oC includes the Asn and His in anantiparallel -sheet and random coil, respectively, suggesting two separate epitope sites for same mAb4.

[0050] FIG. 11B is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb4complex at 36oC comprised of cross peaks that reveal the amino acids associated with theSIRP- epitope for mAb4.

[0051] FIG. 11C is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb4complex at 52oC comprised of cross peaks that reveal the amino acids associated with theSIRP- epitope for mAb4. The mAb4 epitope at 52oC includes the Asn and His in anantiparallel -sheet and random coil, respectively, suggesting two separate epitope sites for 11 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 same mAb4. Also, SIRP- regions at 52oC.

[0052] FIG. 11D is an illustration of the co-distribution asynchronous plot within the amideI and amide II bands spectral region of 1780-1490 cm-1and temperature range of 24oC-52oCfor the SIRP- / mAb4 complex at 2:1 (SIRP- :mAb4) mol ratio. The SIRP- / mAb4complex exhibits lower stability when compared to pure mAb4 and pure SIRP- with lysine,aspartate and tyrosine residues (cross peaks highlighted with a thick line white ellipse) being perturbed due to the thermal stress. Also, highlighted (dash line white ellipse) is the glycosylation in the beta sheet region suggesting greater instability of this complex when compared to the other complexes with higher affinity.

[0053] FIG. 11E is an illustration of the mAb4 epitope in domain 1 of SIRP- (domain 1, site1). The domain 1, site 1 epitope residues listed in a clockwise direction from bottom right:R46, N51, Q52, K53, and H56, side chains shown. The KD of mAb4 for SIRP- is 63 nM.

[0054] FIG. 11F is an illustration of the mAb4 epitope in domain 2 of SIRP- (domain 2, site2). The domain 2, site 2 epitope residues listed in a clockwise direction from bottom right: R182, H192, Q194, and R210, side chains shown. This second epitope has a neighboring N214 N-linked glycosylated residue. All residues are surface accessible and exposed inaccordance with 2T2D secondary structure identified. The KD of mAb4 for SIRP- is 63 nM.

[0055] FIG. 11G is an illustration of the two mAb4 epitopes on SIRP- (domain 1, site 1 anddomain 2, site 2). Site 1 and site 2 are about 40 Angstroms apart.

[0056] FIG. 12 is a table presenting the sequential order of molecular events (SOME) forpure 2:1 and over the temperature range of 24°C-52°C.

[0057] FIG. 13A is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb1complex against the pure mAb1 candidate as reference within the glycan spectral region 1180-1010 cm-1at 24oC.

[0058] FIG. 13B is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb1complex against the pure mAb1 candidate as reference within the glycan spectral region 1180-1010 cm-1at 53oC.

[0059] FIG. 13C is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb4complex against the pure mAb4 candidate as reference within the glycan spectral region 1180-1010 cm-1at 24oC. 12 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125

[0060] FIG. 13D is an illustration of the 2T2D asynchronous plot of the SIRP- / mAb4complex against the pure mAb4 candidate as reference within the glycan spectral region 1180-1010 cm-1at 52oC. The cross peaks that are highlighted define the interaction of mAb4to the glycan region of SIRP- and are: (1148, 1072) CO glycosidic bond, Thr (C-O), and(1118, 1072) (C-O) & Thr (C-O) highlighted with a white ellipse. Also, (1148, 1043) CO glycosidic bond (1118, 1043) (C-O) & CO glycosidic bond highlighted with a white dash ellipse.

[0061] FIG. 14 is a schematic representation of the workflow of embodiments of the methodof this disclosure.

[0062] FIG. 15A is a table summarizing the epitope map and complex stability findings forthe SIRP- / mAb1 complex at a ratio of 2:1 (SIRP- :mAb1) at 24°C, 36°C and 52°C .

[0063] FIG. 15B is a table summarizing the epitope map and complex stability findings forthe SIRP- / mAb2 complex at a ratio of 2:1 (SIRP- :mAb2) at 24°C, 36°C and 52°C withinthe amide I and II band spectral region.

[0064] FIG. 15C is a table summarizing the epitope map and complex stability findings forthe SIRP- / mAb3 complex at a ratio of 2:1 (SIRP- :mAb3) at 24°C, 36°C and 52°C withinthe amide I and II band spectral region.

[0065] FIG. 15D is a table summarizing the epitope map and complex stability findings forthe SIRP- / mAb4 complex at a ratio of 2:1 (SIRP- :mAb4) at 24°C, 36°C and 52°C withinthe amide I and II band spectral region.

[0066] FIG. 16 is a table summarizing the evaluation of SIRP- / mAb complexes at a ratioof 2:1 (SIRP- :mAb) at 24°C, 36°C and 52°C within the target glycan region.DETAILED DESCRIPTION

[0067] This disclosure relates to a system and method for the comprehensive and rapiddetermination of the epitope map of antibodies, defined with amino acid level resolution, including for antibodies that are candidates for development as therapeutics (antibody candidates). The epitope map can be determined for multiple antibody candidates in an array against the same target / antigen. The disclosed comparative analysis can also provide a determination of the optimal target / candidate mol ratio for complex (antibody-target complex) formation, key interactions that stabilize the complex, and the assessment of the candidate-target complex relative stability within the array during a controlled perturbation, 13 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 such as a thermal perturbation. The pure components, antibody and target, can also be evaluated within the same array ensuring true comparative analysis for decision making. The comparative analysis involves the acquisition of hyperspectral images under controlled perturbation conditions, such as at different set temperatures, and analysis of the associated microspectroscopic spectral data of the complex against the pure candidate and pure target. The reference in such analysis is the pure antibody candidate and target, to allow for the determination of the antibody epitope on the target. The candidate is not limited to antibodies and their antigen binding fragment (e.g., Fab, F(ab’)2, scFv, Fv and the like), but any modality thereof including bispecific antibody formats and multispecific antibody formats such as T cell engagers. Furthermore, the method is not limited by the target. It is understood in the field that substantial efforts to discover and develop therapeutic agents against some therapeutic targets using well-known approaches that have been successful in the past have, nonetheless, been unsuccessful. Typically, this is believed to be a result of the physical and chemical properties, and such targets are often referred to as “undruggable.” The system and methods of this disclosure can be used to analyze antibodies against any desired target, including druggable and undruggable targets, allowing for an unprecedented level of molecular understanding to enable drug discovery and design.

[0068] The system and method described herein provides an analytical platform for theaccurate determination of the epitope for an antibody or antigen-binding fragment thereof and the epitope map of the target antigen at amino acid level resolution for an array of antibody candidates. As disclosed herein, two or more antibodies can be analyzed using a single slide cell array. In particular, the system and method acquires spectral images of solutions of pure components (target and candidate) and the respective mixtures and then analyzes spectral data obtained from the spectral images using: (1) two-trace two-dimensional (2T2D) analysis, (2) Co-distribution (2D-CDS) and (3) two-dimensional correlation spectroscopy (2D-COS) to assess the epitope map, the stability of the pure components and the complexes, as well as the molecular behavior (also referred to in the art as the dynamic fingerprint) of the pure components and the sequential molecular changes that occur during complex formation or dissociation, to provide a comprehensive understanding of the key interactions that govern selectivity, affinity, and stability of the array of complexes studied for comparability assessment. The determination of both the paratope and the epitope is performed with amino acid level resolution by the direct determination of the key signature peaks involved in the weak interactions that govern complex formation / dissociation. This evaluation extends to the 14 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 comparative analysis of the glycan composition and its role in stabilizing the pure components as well as the complex. This comparative and comprehensive evaluation serves to determine the target epitope and provide an understanding of the differences in affinity, selectivity and specificity a candidate may have for a specific target. This evaluation is not limited by the secondary structure of the target or the glycosylation state of the components within the complex. The approach also serves to evaluate the stability of the interaction by probing the interaction during thermal stress or other controlled perturbation conditions such as physical or chemical perturbation. The results obtained support the known dissociation constant (KD values) while providing direct molecular evidence of the elements needed for selection based on the therapeutic mechanism of action. This information is beneficial to drug discovery and design.

[0069] Exemplary embodiments of the system and method as used in a case study involvingmultiple antibody candidates with low to mid nM affinities are described herein. The case study demonstrates application of the system and methods in the evaluation of multiplecandidates with known KD’s and the determination of the epitope, structure, influence ofcarbohydrate moiety, and stability of the antibody-epitope interaction to allow for the selection of the optimal pre-clinical candidate or clinical candidate. To the inventor’s knowledge, this disclosure is the first description of a rapid and in-depth analytical tool designed to address such determinations. Thus, the system and method described herein offer significant capabilities to biopharma for drug design, since 80% of the druggable and undruggable targets are membrane protein receptors some of which are highly glycosylated.

[0070] The system and method described herein allows for a comparative assessment of anarray of full-length candidates against the same target allowing for pre-clinical candidate selection under the same thermal stress conditions. There are 5 elements used for the determination of the epitope map: (1) key amino acids and the weak interactions that govern binding obtained using 2T2D asynchronous plots, (2) thermal stability of the weak interactions using both 2T2D and 2D-COS asynchronous plots, (3) the amino acid sequence of the target, (4) the Sequential Order of Molecular Events (SOME) vital information providing the molecular behavior or dynamic fingerprint of the antibody-target complex, determined using 2D-COS synchronous and asynchronous plots, and (5) post-translational modifications (PTMs) that can impact both the interaction interface and the stability of the complex. Furthermore, the analytical platform of the system and method described herein is not limited to a protein target; the target can be any organic molecule entity such as a peptide, 15 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 mRNA, siRNA, DNA or a glycan component. The same is true for the drug modality whether peptides, fusion proteins, monoclonal antibodies (mAbs), Antibody–drug conjugates (ADCs), multi-specifics and a wide range of carriers such as lipid nanoparticles (LNP), liposomes, exosomes or adeno-associated viruses (AAV’s). No limitations based on size, symmetry or post-translational modifications are imposed on this label free highly selective and sensitive platform method. For each candidate against the same target, the binding interface may also be determined while also evaluating the structural changes and stability of the pure components and the complexes thereof during thermal stress, or other controlled perturbation, thus providing a comprehensive evaluation of all the components of interest including the complex.

[0071] Herein, this disclosure describes a combination product solution (a system) thatemploys a vibrational spectroscopy apparatus, such as a quantum cascade laser microscope (QCLM), Raman spectrometer or Raman microscope, Raman amplifier, FT-IR, or light detection and ranging (LIDAR) system in the MID infrared spectral region for the fast acquisition of hyperspectral images, along with a slide cell array to allow for the comparability assessment and a series of dedicated software modules that effectively employ the three 2D-correlation algorithms required to provide both increased spectral and temporal resolution. The slide cell array may be comprised of a polymer, such as polyethylene, calcium fluoride, barium fluoride, or other polymer that is transparent within most of the spectral region of interest. US 11,000,852 and US 11,590,504 which disclose exemplary slide cells suitable for use with the method and system of this disclosure, and each is incorporated herein by reference. The samples to be screened, along with one or more reference standards, or controls, are provided within the cells of the slide cell array, and the QCLM or other is used to acquire spectral images of the slide cell array. A preferred vibrational spectroscopy apparatus for use in the system and method of this disclosure is QCLM, other vibrational spectroscopy apparatus such as a Raman spectrometer or Raman microscope can also be used to acquire spectral images of the slide cell array.

[0072] A schematic representation of the workflow is shown in FIG. 14. As shown in FIG.14, the system and method first acquires a series of hyperspectral images of the slide cell array. The slide cell array includes wells containing the sample being analyzed, and one or more reference samples or controls. Spectral data is obtained from the hyperspectral images and exported for analysis. The baseline corrected spectral data within the spectral region of interest are then subjected to varying methods of spectral subtraction dependent upon the 2D 16 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 algorithm to be applied (see the data analysis description below). In an example, covariance spectra are then generated, and it is to this type of spectra that the different 2D algorithms are then applied for analysis. The algorithms employed are a 2T2D algorithm, a 2D-CDS algorithm, and a 2D-COS algorithm. The 2T2D algorithm is applied for spectra at different set temperatures of interest, and 2T2D asynchronous plots are generated for each set temperature. Signature cross peaks in the asynchronous plot are analyzed to determine the target epitope for the specific candidate of interest at each set temperature. The 2D-CDS algorithm is also applied, and 2D-CDS asynchronous plots are generated for each sample in the array for comparative analysis. Cross peaks in the 2D-CDS plots are evaluated and compared, which provide key inter-molecular interactions that are essential to the target / candidate complex stability. The 2D-COS algorithm is also applied, and 2D-COS synchronous and asynchronous plots are generated. These plots define the spectral intensity changes which are associated with the molecular behavior of the pure components as well as for the target / candidate complexes as they are thermally perturbed. In addition, the cross peaks and their correlations are used to define the peak assignments. Furthermore, the intensity changes of the cross peaks in the 2D-COS plots are evaluated and Noda’s Rules used to define the sequential order of molecular events (SOME). See, I. Noda et al. Generalized two-dimensional correlation spectroscopy, Appl. Spectrosc., 54, 236A-248A (2000); I. Noda, Y. Ozaki, Two-Dimensional Correlation Spectroscopy. Applications in Vibrational and Optical Spectroscopy, John Wiley & Sons, Chichester, 2004; and I. Noda, Frontiers of two-dimensional correlation spectroscopy. Part 1. New concepts and noteworthy developments, J. Mol. Struct., 1069, 3e22 (2014). This collective information allows us to establish differences in stability for both the pure components and the complexes. Table I,Tables IIA-IID, FIGs. 15A-15D and FIG. 16 present data for the case study and peakassignments for each component and complex of interest.

[0073] Further, thermal dependence plots may be evaluated for signature peaks defined using2D-COS vibrational modes to establish secondary structure stability in a comparative manner. The 2T2D evaluation at different set temperatures allows for the evaluation of the stability of the weak interactions that define the epitope. In addition, the 2D-CDS confirms the key weak interactions that allow for binding that in turn define the selectivity. The weak interactions examined are not limited to the amino acids or the backbone, but also the glycan regions of the target. In combination, this approach allows for the understanding of the basis for selectivity, affinity and stability of the interaction which support the KDvalues. 17 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125

[0074] In exemplary embodiments the system and methods of this disclosure involve samplepreparation, acquisition of hyperspectral images, two trace two-dimensional (2T2D) correlation analysis, two-dimensional co-distribution spectroscopy (2D-CDS) analysis and two-dimensional correlation spectroscopy (2D-COS) analysis.

[0075] Sample preparation

[0076] The pure candidates and target (i.e., pure antibody and antigen components), andtarget / candidate complexes are analyzed in the same formulation condition. The analysis is not limited to fixed target:candidate mol ratios, but instead a series of target:candidate mixtures may be and preferably are evaluated to ascertain the optimal target:candidate mol ratio for complex formation. The target:candidate mol ratio can be, for example about 1:1 to about 10:1. Preferably, the target:candidate mol ratio provides an amount of target that is sufficient to saturate the candidate antibody binding sites in the test sample. For example, for a typical bivalent IgG antibody a mole ratio of target:candidate of 2:1 would saturate all antibody binding sites assuming all available target is bound by the antibody. The array of samples typically includes the pure components as well as the complexes to allow for comparative analysis. The negative control is the formulation condition. All samples are and covered with an optically polished cover providing a fixed path length to allow for the quantitative analysis. The slide cell array may be a disposable array. The substrate for the slide cell may be salt or other IR transparent material including polymers, such as polyethylene, calcium fluoride, barium fluoride, or other polymer that is transparent within most of the spectral region of interest. The slide cell may then be assembled into a thermally controlled slide cell holder accessory to ensure a controlled thermal environment for hyperspectral image acquisition. The slide cell holder accessory may be, for example, a controllable heated chamber configured to receive the slide cell. See, e.g., US 11,000,852 and US 11,590,504 regarding slide cells and systems.

[0077] Hyperspectral (HS) Image acquisition

[0078] After the slide cell array is placed in the slide cell holder, spectral images of thesamples within the slide cell array are obtained. An infrared microscope, preferably a quantum cascade laser (QCL) microscope which allows for the rapid collection of a series of HS images with 16x more spectral data per pixel are achieved. The QCL microscope is equipped with a heated accessory and a slide cell array, as described above, to allow for the acquisition of a series of hyperspectral images for the evaluation of the perturbation of the 18 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 weak interactions within the complex that govern binding as well as the stability of both the pure components and the complexes when subject to a thermal perturbation. See, e.g., US 11,000,852 and US 11,590,504. The advantage is the real-time comparability assessment of the samples of interest against the pure components within the MID IR spectral region 1800 – 950 cm-1. HS images may be acquired once thermal equilibrium is reached at different set temperatures within a defined temperature range within 24 – 52oC. Other defined temperature ranges can be used. Typically, the HS images are collected across a range of temperature from about 4°C to about 95°C. In embodiments, HS images are collected at two or more temperatures from about 24°C to about 52°C or about 24°C to about 36°C. Each HS image acquisition occurs within seconds and is comprised of 223,000 QCLM spectra collected at 2 or 4 cm-1spectral resolution and as a result an enhanced signal / noise ratio is obtained when compared to an FT-IR microscope. The entire MID IR spectral region is valid for the analysis. Further, other microscopy techniques may be applied to obtain the spectral images, such as FT-IR or Raman microscopy.

[0079] Two trace two-dimensional (2T2D) correlation analysis

[0080] After the spectral images are acquired, a two-trace two-dimensional (2T2D)correlation process is applied to the spectral data. See, Noda, I., Two-trace two-dimensional (2T2D) correlation spectroscopy- A method for extracting useful information from a pair of spectra. J. Molec. Struct.1160, 471-478 (2018). The goal is to obtain the cross peaks associated with the inter-molecular interaction within the complex. Specifically, determine the amino acid side chains modes that define the epitope. From a pair of spectra a two tracetwo-dimensional correlation can be applied to generate a synchronous spectrum , )and asynchronous spectrum , ). The synchronous spectrum , ) andasynchronous spectrum) are given by:( , ) = [ ( ) ( ) + ( ) ( )] (1)

[0081] , , ofinterest and the second spectrum, ( ), corresponds to the pure component (i.e., candidate),respectively. A 2T2D correlation coefficient is then applied ( , ) to the synchronous and19 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125a disrelation coefficient ( , ) to the asynchronous evaluation resulting in the scaledversion of the 2T2D correlation spectra, as given by: (, ) = ( , ) ( , ) ( , ) (3)where:( , ) + ( , ) = 1 (5)This indicates the complementarity nature of the quantities.

[0082] Two contour plots are generated. These are the synchronous ( ( , )) plot wheredominant spectral components of the sample (target / candidate complex) and reference (purecomponent): ( ) and ( ) are observed. The diagonal is comprised of auto peaks where= . The cross peaks are always positive. The second plot is the asynchronous( ( , )) plot, which is more informative. Albeit for the epitope map determination onlythe asynchronous plot is used, since the focus is on a reduced number of cross peaks that define the side chains included in the epitope. Also, cross peaks of the same intensity and sign correspond to the same component within the sample.

[0083] Two-dimensional co-distribution spectroscopy (2D-CDS)

[0084] 2D-CDS provides the distribution population of the target / candidate complexes orthe pure components in solution, respectively and reveals the main events including weak interactions common to the components in the sample that were perturbed during the thermal perturbation. The spectral correlations observed can be compared directly between pure component samples and the target / candidate complex samples, allowing for determinations of key weak interactions essential for complex formation within the sample. This evaluation is verified by comparing with the pure components, so as to eliminate common intra- molecular interactions that may exist. Used together, the methods can provide an understanding of the structural changes, the inter-molecular, hydrogen bonding and hydrophobic interactions associated with the stability of the target / candidate complex. This does not preclude the glycan region of the target.

[0085] 2D-CDS is especially suited for evaluating the dynamic elements common to anensemble of structures in a solution subjected to a controlled perturbation, such as a thermalperturbation. For a set of spectra, , are obtained as a function of the spectral20 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125variable with = 1, 2, …, and some perturbation variable , with = 1, 2, … , atdefined intervals between and . In our data, the spectral variable is the quantumcascade laser microscope wavenumber (cm-1), and the perturbation variable is temperature.

[0086] The dynamic, is defined as:= , (6)with the average1= , . (7)

[0087] The asynchronous,) is defined as:( , ) ( ) ( )( ) ,, = . (8)

[0088] (, ) = ( , ) ( , ) . (9)

[0089] By convention, the value of ( , ) is set to be zero, when ( ) = 0 or ( ) = 0is encountered, indicating a lack of spectral intensity at either QCLM wavenumber. The asynchronous 2D-CDS intensity is a measure of the difference in the distribution of two spectral signals along the perturbation variable axis based on the mathematical tool known as moment analysis.

[0090] The interpretation of the 2D-CDS asynchronous plot is direct. For a positive crosspeak ( , ) > 0, the presence of the spectral signal at is distributed predominantly at anearlier stage, prior to . In the case of a negative cross peak ( , ) < 0, the order isreversed. Finally, when ( ) 0, the average distributions of the spectral signals aresimilar, and therefore they co-exist. More importantly, because these plots summarize the average behavior of the sample (target / candidate complex or the pure component) then the cross peaks associated with the weak interactions that are key to binding can be discerned.

[0091] Two-dimensional correlation spectroscopy (2D-COS) analysis

[0092] 2D-COS has proven useful in that it provides a detailed molecular description of theeffects of the perturbation on both the side chains and, in turn, the conformational stability of 21 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 the pure components and the target / candidate complexes within the slide cell array during thermal perturbation. The stability of a candidate, target or target / candidate complex is dependent on its sequence, secondary and tertiary structure, post-translational modifications, and formulation conditions. This correlation analysis enhances the spectral and temporal resolution to provide such information as the therapeutic protein (candidate) its target and the complex as they are destabilized under the exact same temperature perturbation conditions, therefore serving as a dynamic fingerprint that provides the key molecular events associated with the stability of the pure component and the target / candidate complex. The 2D-COSalgorithm is defined asà , = , 1 (10)02D correlation intensities ofthe covariance spectral data are defined by:( , ) = à , à , (11)(, ) as a function of two independentwavenumber axes,1and2, is the synchronous plot.

[0094] Asynchronous 2D correlation intensities of the covariance spectral data are definedby:( , ) = Ã , Ã( , ) (12)Hilbert-Noda transformation matrix and isgiven by: 0== (13))

[0096] The cross-correlation function is applied to a covariance spectral dataset, which is adataset obtained from subtracting an initial spectrum from subsequent spectra. For example, the initial spectrum may be acquired at low temperature, and subsequent spectra acquired at higher temperatures. The spectral changes due to temperature increase are then revealed (revealing changes in the candidate or target in their pure and complex forms), which are referred to as covariance spectral data or difference spectra. Applying the cross-correlation 22 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 function to the difference spectral dataset results in two separate, yet symmetrical 2D plots (eqns.11 and 12). The first plot is referred to as the synchronous plot. It contains positive peaks on the diagonal, known as the auto peaks, and provides the overall changes observed in the spectral dataset. The relationship established in this synchronous plot relates to the spectral intensity changes that occur synchronously, hence the name. The second 2D plot is known as the asynchronous plot. This plot relates the asynchronous intensity changes, resulting in enhanced spectral resolution, which can be used to correlate out-of-phase peak intensity changes.

[0097] Both plots contain off-diagonal peaks, which are referred to as cross peaks; thesepeaks correlate the spectral changes observed at the frequencies assigned to the various bands (e.g., backbone and side chain structural elements). Spectral intensity changes observed are due to the incremental thermal perturbation applied to the sample. No a priori knowledge of the system is required for interpretation of the results. The information in both plots allows for determining the sequential order of molecular events (SOME) that occur during the perturbation by following Noda’s rules. These plots are symmetrical in nature and for interpretation purposes, reference is made to the top triangular portion of the plot for analysis. The asynchronous plot is comprised exclusively of cross peaks that relate the out-of-phase peaks. As a result, this plot reveals greater spectral resolution enhancement. The following rules can apply to establish the order of molecular events.

[0098] I. If the asynchronous cross peak 2: if positive, then 2 is perturbed prior to 1 ( 21).

[0099] II. If the asynchronous cross peak 2: if negative, then 2 is perturbed after 1 ( 21).

[0100] III. If the corresponding synchronous cross peak is positive, then the order of theevents is established using the asynchronous plot (rules I and II).

[0101] IV. However, if the corresponding synchronous cross peak is negative and theasynchronous cross peak is positive, the order is reversed.

[0102] V. If the synchronous plot contains negative cross peaks and the correspondingasynchronous cross peak is negative, then the order is maintained.

[0103] An order of events can be established for each peak observed on the 2 axis.

[0104] In summary, the combination of the three algorithms is a powerful tool providing anin-depth comparability assessment of the target / candidate interaction, the candidates selectivity and affinity for the target, the stability of the complex and the epitope. The 23 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 analysis is not limited to the target / candidate complex since the pure components are also evaluated in the array. This design of experiments and data sets obtained provide a unique opportunity to establish differences in secondary structure, intra- and inter-molecular weak interactions and stability, when compared to the complex by: (1) determining of the epitope with amino acid resolution including the influence of post-translational modifications such as the glycans using the 2T2D algorithm, (2) evaluating the key weak interactions that provide stability to the complex using 2D-CDS algorithm, (3) assessing the dynamic fingerprint to allow for the determination of the sequential order of molecular events (SOME) for both the pure components and the target / candidate complexes. Also, the glycan spectral region is also evaluated thus providing a comprehensive evaluation of the target / candidate complex and the target to allow for comparative assessment and the determination of the role if any the glycan region has on complex formation and stability.

[0105] In embodiments, A method for analyzing candidate antibodies in a sample,comprising: a) providing the sample in a slide containing at least one sample well and at least one reference well; b) acquiring at least one spectral image of the sample and at least one spectral image of the at least one reference using a quantum cascade laser microscope under controlled temperature conditions; c) identifying and selecting, in at least one of the acquired first spectral images, a region of interest; d) obtaining spectral data for the sample and the at least one reference for the region of interest; e) applying a baseline correction to the spectral data for the sample and the reference for the region of interest; f) generating weighted difference spectra for the sample and reference; g) applying a two-trace two-dimensionalcorrelation to the weighted difference spectra to generate a synchronous spectrum ( , )and asynchronous spectrum ( , ), and identifying signature cross peaks in theasynchronous plot toa target epitope for a particular candidate antibody; h) applying a two-dimensional correlation analysis to the baseline corrected data and reference spectral data to generate a synchronous correlation plot and an asynchronous correlation plot, identifying signature peaks in the plots, and analyze intensity changes of the peaks to define a sequential order of molecular events; and i) applying a two-dimensional co-distribution analysis to the baseline corrected data to generate a synchronous correlation plot and an asynchronous co-distribution plot, and identifying cross peaks in the plots, and evaluating cross peaks in the co-distribution plots, identify cross peaks associated with key interactions between the target epitope and the particular candidate antibody, and determine relative stability. 24 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 EXEMPLIFICATION

[0106] SIRP- , a membrane associated glycosylated target

[0107] Signal Regulatory Protein alpha (SIRP- ) is a membrane associated glycosylatedtarget protein associated with colon cancer among other indications.

[0108] A comparability assessment of a series of pure antibody candidates (mAb1, mAb2,mAb3 and mAb4) that bind the extracellular portion of SIRP- (FIG.1) against theirrespective complexes using a QCLM for all the samples in a single array under controlled temperature conditions and multiple 2D-correlation algorithms (2T2D, 2D-CDS and 2D- COS) were applied to the spectral data acquired to allow for a comprehensive understanding of the pure components as well as their complexes. The target used is the extracellularcomponent of the membrane protein SIRP- containing three domains with a hexa-histidinetag at the C-terminus (SEQ ID NO:1). SIRP- has a calculated MrW of 38.5 kDa with threeN-linked glycosylation sites.

[0109] Although the hyperspectral image acquisition was done for all samples within a singlearray the spectral analysis was divided into three phases, herein referred to as subtopics. Asummary of the case study SIRP- , a membrane associated glycosylated target is shown inTables I and IIE for the pure components and complexes, respectively. The three analysis phases were: first phase (Phase I) Subtopic: Stability of the pure components; second phase(Phase II) Subtopic: Epitope map of SIRP- and the weak interactions that govern bindingfor a series of SIRP- / candidate complexes and the stability of the interaction focused onthe amino acids within the binding interface and conformational changes that occurred during complex formation within the spectral region of 1780 – 1490 cm-1; and third phase (Phase III)Subtopic: The role of the SIRP- glycan on both the binding and stability of the SIRP- / candidate complexes. This third phase extends the analysis to include the SIRP- (target)glycan spectral region 1180 – 1018 cm-1to verify some of the epitope sites determined and establish if the glycan contributed to the epitope.

[0110] Subtopic: Stability of the pure components (Phase I)

[0111] The analysis involved the full characterization of target (SIRP- ) (FIGs. 1 and 2A-C)and the mAb (1-4) candidates (FIGs.3A-3C, 4A-4C, 5A-5C and 6A-6C) as pure components. The sequential molecular order of events for the pure components is summarized in FIG 7. Also, the comparability assessment for the pure components is summarized in Table I. Below 25 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 is a discussion of key elements of the data that will serve to substantiate the analysis of the complex of interest. Albeit this phase I, served as a solid foundation to allow for thecharacterization of the different SIRP- / candidate complexes. The pure components wereanalyzed for multiple critical quality attributes: Asn / Gln deamidation, self-association, aggregation and stability using both co-distribution and 2D-COS analysis. FIG.1 shows theamino acid sequence of the extracellular domain of SIRP- (target). The target is comprisedof three domains for which domains 2 and 3 contain N-linked glycosylation sites (N214,N239 and N261). The co-distribution plot for SIRP- is shown in FIG. 2A in which aminimal number of cross peaks are observed, suggesting the target is structurally stable. The 2D-COS plots are typical of a protein that has a high -sheet and -turn content, but more importantly the target is aggregate free and no deamidation is observed (FIGs.2B and 2C). Similarly, for the mAb candidates, they were all determined not to deamidate (no Asn or Gln deamidation was observed) within the CDR regions in the temperature range studied (24- 52oC) and were determined to be aggregate free (FIGs 3B, 3C, 4B, 4C, 5B, 5C, 6B and 6C,), suggesting they were all valid for the target / mAb candidate complex evaluation for epitope map determination. The molecular behavior observed during the thermal perturbation as defined by the analysis of the cross-peak pattern within the 2D-COS asynchronous plots (FIGs 3C, 4C, 5C and 6C) was observed to be different for each mAb candidate, suggesting variable conformational stability. The thermal transition temperature (Tm) and the onset of Tmwere determined for each pure component. The stability for these pure mAb candidates was ranked from most stable to least stable as: mAb1 > mAb2 > mAb3 > mAb4 (Table I). Further discussion is provided below for the SOME evaluation and in FIG.7.

[0112] The Sequential Order of Molecular Events (SOME) is based on the application ofNoda’s rules to the cross peak intensity changes observed for both the synchronous and asynchronous plots which in turn describe the molecular behavior of the intact protein in solution. This description is also referred to as the “dynamic fingerprint” of the protein and serves to understand the molecular stability of the protein of interest. For the target (SIRP- ), the sequential order of molecular events does not include the glycosylation region, thussimplifying the dynamic molecular events described herein for SIRP- (target). Nonetheless,a glutamine (1581 cm-1) residues (presumably exposed to the aqueous environment) are first to be perturbed, followed by the weak interactions involving the side chains of aspartates (1572, 1567 cm-1), hydrogen bonding interactions involving tryptophan (1496 cm-1) and 26 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125tyrosine (1516 cm-1) with their aqueous environment or disruption of - stacking(hydrophobic interactions between the aromatic residues) and salt bridge interactions involving lysine residues (1527 cm-1) and aspartates (1567 cm-1) or glutamates (1545 cm-1), then the perturbation of the random coil (1645 cm-1), -sheets (1636 cm-1) and -strands (1623 cm-1), followed by the disruption of hydrogen bonding interactions involving asparagine (1612 cm-1) and histidine residues (1608 cm-1), followed by the -helix (1652 cm-1) and loops (1660 cm-1) being perturbed then hydrogen bonding interactions being disrupted involving glutamines (1581 cm-1) and aspartates (1572 cm-1) are being perturbed, then within the -turns aspartate and glutamates backbone carbonyls (1727, 1712 and 1700 cm-1), the asparagines (1672 cm-1) and finally the perturbation of the -turns (1696, 1688 cm-1).

[0113] In general, the mAb candidates' sequential order of molecular events can be definedas one that occurs in three phases: the initial phase involving the weak interactions including hydrogen bonding, ionic or salt-bridge and hydrophobic interactions involving glutamines (1581 cm-1), aspartates (1572, 1567 cm-1), tryptophan (1496 cm-1), lysines (1527 cm-1), glutamates (1545 cm-1) and tyrosines (1516 cm-1) that are disrupted followed by the propagation phase involving continued hydrogen bonding disruption involving asparagines (1612, 1672 cm-1), histidines (1608, 1596 cm-1), lysines (1627 cm-1) and backbone perturbations random coil, (1645 cm-1), -strand (1623 cm-1), -helix (1653 cm-1), -sheet (1636 cm-1) and ending with the hinge loops (1660 cm-1) being perturbed. However, there is one exception among the candidates that of mAb3 where the -strands (1623 cm-1) are perturbed prior to the random coil (1645 cm-1). The third phase involves the disruption of the hydrogen bonding interactions between the peptide carbonyl of glutamates and aspartates (1727, 1712, 1700 cm-1), and asparagine side chain amides (1672 cm-1) exposed to their aqueous environment, resulting in the perturbation of the -turns and as a result the salt- bridge interactions involving lysines (1527 cm-1) and glutamates within the -turns (1696, 1688 and 1680 cm-1) are disrupted along with the perturbation of the hydrogen binding interaction involving tyrosines (1516 cm-1). This molecular description is a typical dynamic fingerprint for mAbs.

[0114] Subtopic: Epitope map of SIRP- and the weak interactions that govern binding(Phase II).

[0115] This phase provided the number of epitope sites for each SIRP- mAb candidatestudied at a (2.0 / 1.0 (SIRP- / mAb) mol ratio) and the actual epitope map with amino acid27 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 resolution level. For the determination of the epitope map the 2T2D algorithm was employed and the asynchronous plots included the amide I and II bands within the spectral region of 1780 – 1485 cm-1. The resulting negative cross peaks on the upper triangle of the symmetrical contour plot allows for the determination of the amino acid residues within the epitope. In addition, the cross peak coordinates also provide their secondary structure location. Tables IIA -D summarize the cross peak coordinates and their assignments (amino acids and theirinteractions) for each SIRP- / mAb candidate complex evaluated at 24, 36 and 52 oC). Co-distribution plots were used to determine the key target / mab candidate interaction that governed binding (selectivity). In addition, the 2D-COS plots allowed determination of complex formation / dissociation at the molecular level. Also, surface models were generated using the available high-resolution structure PDB ID: 2WNG to make the final determination of the epitope map which needed to comply with the experimental finding using the method described above. We have highlighted the epitope map within the amino acid sequence of theSIRP- (target) for each SIRP- / mAb candidate evaluated. The resulting epitope maps werewithin a stretch of 10-20 amino acids and were also assessed for their stability. They arepresented based on the stability of the SIRP- / mAb candidate complex (from the moststable to least stable complex). FIGs.8E, 9E, 10E, and 11E-11G.

[0116] For the SIRP- mAb1 complex the SIRP- epitope for mAb1 was determined using2T2D asynchronous plots to be within domain 1 (see FIGs.8A-8E) an exposed 10 residuestretch comprised of epitope residues within random coil / loop region of SIRP- aresummarized in Table IIA and are valid for both low (T= 24oC) and high temperatures (T= 52oC) shown in FIGs. 8A-8C, suggesting the SIRP- / mAb1 candidate interaction is stable andtherefore a high affinity complex. The Co-distribution asynchronous plots provide thedistribution population of protein sample in each well for pure SIRP- (FIG. 2A), SIRP- / mAb1 (2.0 / 1.0, mol ratio) complex (FIG.8D) and for the pure mAb1 candidate (FIG.3A) within the amide I and amide II bands spectral region 1780-1490 cm-1. Results from the comparative assessment have yielded that the key interaction that governs selectivity / binding involves a lysine-glutamate salt bridge interaction. The location of the lysine residue is alsowithin a random coil region. The 2D-COS plots we defined the sequential molecular order ofevents (SOME) shown in FIG.12 was used to define complex formation and confirmed the perturbed lysine residue as an early event and that the location of the lysine residue waswithin random coil. Furthermore, we modeled the SIRP- target using the available high-28 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 resolution structure. The cumulative information was used to define the epitope map modeledthe SIRP- using the available high-resolution structure (PDB ID: 2WNG). We identified theepitope residues within the amino acid sequence shown in FIG. 8E as: Arg46, Asn51 / Gln52,Lys53 and His56. Also, when compared to other candidates the SIRP- Lys53 is involved insalt-bridge interaction with the mAb1 candidate and the complex is further stabilized by inter-molecular H-bonding interactions that include SIRP- Asn51 and His56 with residues withinthe candidate that stabilize the protein-protein interaction even further (FIGs.8A-C).

[0117] For the SIRP- / mAb2 complex the epitope was also within domain 1 (see FIGs. 9A-9E and 12 and Table IIB), but exhibited lower affinity / stability than the SIRP- / mAb1complex due to the Arg46 destabilized Lys53 salt-bridge interaction with the second mAbcandidate causing loss of stability within the SIRP- loop region due to the disruption of theSIRP- Asn51 inter-molecular hydrogen bonding interaction with mAb2. The perturbation ofthe Lys salt-bridge is evident as cross peaks in the co-distribution plot shown in FIG.9D. The epitope map includes Arg46, Asn51, Lys53 and His56 has been highlighted in the SIRP- amino acid sequence (FIG.9E).

[0118] For the SIRP- / mAb3 complex two contiguous yet distinct epitope sites weredetermined to be located within domain 1 (see FIGs.10A-10E, 12 and Table IIC). Site 1: theSIRP- epitope includes Asn51 / Gln52, Lys 53and His56 (FIGs. 10A-10B and 10E). Theinclusion of Gln52 suggests that the additional methylene in Gln may cause mAb3 to loseaffinity for SIRP- leading to its release and later binding to the contiguous SIRP- epitopeSite 2 (FIG. 10C) at higher temperature (T= 52 oC). Site 2: the resulting SIRP- epitope iscomprised of His56, Phe57, Arg59, and Asn80. Also, the mAb3 candidate is observed toperturb the SIRP- glycan region (see below subtopic 3 section), thus supporting theexistence of the second contiguous site (FIGs.10A-11C).

[0119] For the SIRP- / mAb4 complex two separate epitope sites were determined (see FIGs.11A-11E, 12 and Table IID). The Co-distribution asynchronous plot for SIRP- / mAb4complex was used to identify the interaction that is the basis for selectivity when compared tothe pure components: SIRP- (FIG. 2A) and mAb4 candidate (FIG. 6A). The selectivityexhibited by mAb4 for SIRP- involved: lysine, aspartate and tyrosine residues (cross peakshighlighted with a thick line white circle, FIG.11D) being perturbed due to the thermal stress. Also, highlighted (dashed line white circle) is the Asp / Glu that are exposed to the aqueous environment within a beta sheet region, suggesting greater instability of this 29 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 complex when compared to the other complexes with higher affinity. Also, the target within oC (FIG.11C). The first site was located within domain 1: the epitope residues are Tyr50 / Phe57, Asn51 / Gln52, Lys53 and His56 have been determined, while for the second binding site located within domain 2 is observed above physiological temperatures: the epitope is comprised of His192, Gln194 and Arg210 with a neighboring Asn214-glycosylated residue (FIGs.11E - 11G). Thissecond site is observed to have SIRP- glycan interaction with mAb4 (FIGs. 13A-13D).

[0120] A comparative stability analysis was performed by evaluating the change inabsorbance within the temperature range of 28-52oC for the series of SIRP- / candidate (2.0 / 1.0 mol ratio) complexes. The rank order of the target / candidate complexes is SIRP- / mAb1 > SIRP- / mAb2 > SIRP- / mAb3 > SIRP- / mAb4 (from the most stable to the leaststable). By monitoring the most abundant secondary structure in both the candidates and the target, the analysis obtained vital information by monitoring the -sheet peak position (1634- 1638 cm-1) as a function of temperature. In each case, both components were observed to be stabilized due to the inter-molecular interaction within the complex.

[0121] The dynamic fingerprint for these complexes within the amide I and II bands spectralregion comprised of 1780 - 1490 cm-1were analyzed to yield the sequential order of molecular events (SOME) (FIG.12) confirm the target / candidate interaction at the molecular level. The common critical molecular interactions and their threshold for the distribution population of each complex were evaluated during the thermal stress. A summaryof the findings is contained in FIGS. 15A-15D. The analysis of the 2D-COS plots followingDr. Noda’s rules allows for the determination of the sequential order of molecular events(SOME) shown in FIG.12. The results provide information about the thermal perturbation effects on the complex without distinction of the components. Except for the events observed after the -turns (1680, 1688 and 1696 cm-1) which are perturbed by the disruption of weak interactions within the -turns of the candidates. At high temperatures this series of events is typically observed for mAbs. This disclosure defines two separate classes of complexes due to their thermal stability and the fingerprint dynamics observed in the SOME.

[0122] FIG. 12 presents the SOME for all four SIRP- / mAb candidate complexes andprovides information about the thermal perturbation effects on the complex without distinction of the components. For the first class a molecular behavior exhibits a combination of both components of the complexes associated with a higher Tm, suggesting greater 30 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125stability: SIRP- / mAb1 and SIRP- / mAb2 for which the molecular dynamics are verysimilar including the stabilization of arginine residues (1676 cm-1) located within the -turnswhen compared to the second class of SIRP- complex (SIRP- / mAb3 and SIRP- / mAb4).However, a key differentiator between SIRP- / mAb1 and SIRP- / mAb2 complexes liewithin the candidate’s -turns (1688 cm-1), whereby for the SIRP- / mAb1 the pre-requisiteof the disruption of hydrogen bonding interaction involving aspartate or glutamate backbone carbonyl (1700 cm-1) is required prior to its destabilization and that is not the case for theSIRP- / mAb2 complex. Furthermore, the second class of SIRP- complex (SIRP- / mAb3and SIRP- / mAb4) exhibits a molecular behavior with strong similarities to the pure mAbcandidates and exhibit lower Tm, suggesting lower stability of the complex. Furthermore, thissecond class of SIRP- complex binds to two different sites as they are thermally stressed.Specifically, SIRP- / mAb3 the second site is contiguous to the first while SIRP- / mAb4complex provides evidence of a second binding site located in domain 2 along withperturbation of SIRP- glycan region. This is not observed for any of the other SIRP- / mAbcandidate complexes. Also, the molecular dynamics of the -turns (1680, 1688, 1696 cm-1) are perturbed at a similar time frame or temperature range. Finally, these results support the KDvalues observed for these complexes.

[0123] Subtopic: The role of the SIRP- glycan on both the binding and stability of theSIRP- / candidate complexes (Phase II).

[0124] In general, the QCL IR microscopy spectral region of 1180 – 1018 cm-1 is comprisedproviding insight towards the configuration of the glycan.

[0125] The glycosylation spectral region was investigated further by 2T2D to provide anunderstanding of the molecular nature of the SIRP- glycan interaction with potential mAbcandidates during complex formation (FIGs.13A - 13D and 16). The glycan peak assignments are preliminary in nature pending information on the glycan composition. It isinteresting to note that the glycan structure of the target (SIRP- ) is highly ordered andstrong correlations are observed although the glycans are highly exposed to their aqueous environment. Evidence of perturbation of the glycans is observed due to long-range effects or simply perturbation of the surrounding aqueous environment within the complex as the temperature increases. Furthermore, the 2T2D asynchronous plot was used to determine 31 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 oC, mAb4 complex by the number of cross peaks suggesting strong perturbation of the spectral presumably associated with the N-linked ASN214 (highlighted with dash end diamond line) presumably involving Thr212 or a threonine within the mAb4 candidate within the binding interface. Table I. Summary of multiple CQA assessments for the pure components (SIRP-a and the mAb (1- 4) candidates for the epitope mapping study. Ons Pureet 2DCOS & 2DCOS &Compone Co- nt Tm Tm Comment: Thermal dependence plots distribution Co-distribution(ooC)(CaggregatN / Q )iondeamidation* SIRP-a 36 40The target exhibits a sigmoidal typecurve, suggesting cooperativity. nonemAb1 48 51 Sigmoidal curve behavior. none none mAb2 44 49 Sigmoidal curve behavior. none none This candidate exhibits a sigmoidal mAb3 44 46 curve behavior, suggesting none none cooperativity. This candidate also exhibits a sigmoidal mAb4 40 47 curve behavior, suggesting none none cooperativity. *The mAb cadidates CDR regions did not deamidate, therefore no impact on target binding. 32 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 exhibits the highest affinity. 2T2D Asynchronous Temperature cross peaks cross peak(oC) coordinates assignment24 - 52 (1643, 1630) random coil and Arg, respectively24 - 52 (1643, 1625) random coil and Lys, respectively24 - 52 (1643, 1612) random coil and Asn, respectively24 - 52 (1643, 1600) random coil and His, respectively24 - 52 (1625, 1600) Lys and His, respectively24 - 52 (1630, 1600) Arg and His, respectively24 - 52 (1660, 1612) loop and Asn, respectively24 - 52 (1660, 1600) loop and His, respectively24 - 52 (1672, 1600) Asn and His, respectively24 - 52 (1676, 1612) Arg and Asn, respectively24 - 52 (1676, 1600) Arg and His, respectivelyNote: The results suggest only one epitope site for this complex. Table IIB. 2T2D SIRP- epitope map for mAb2 candidate as a function of temperature.2T2D Asynchronous Temperature cross peaks cross peak(oC) coordinates assignment24-52 (1643, 1612) random coil and Asn, respectively24-52 (1630, 1600) Arg and His, respectively24-52 (1672, 1600) Asn and His, respectively52 (1667, 1661) destabilizing loop regionNote: The results suggest only one epitope site for this complex. 33 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 Table IIC. 2T2D SIRP- epitope map for mAb3 candidate as a function of temperature.2T2D Asynchronous Temperature cross peaks cross peak(oC) coordinates assignment24-52 (1625, 1600) Lys and His, respectively52 (1630, 1612) Arg and Asn, respectively24-52 (1630, 1600) Arg and His, respectively24-52 (1636, 1600) b-sheet and His, respectively24-52 (1636, 1612) b-sheet and Asn, respectively24-36 (1636, 1581) b-sheet and Gln, respectively24-36 (1645, 1581) random coil and Gln, respectively24-52 (1645, 1612) random coil and Asn, respectively52 (1645, 1630) random coil and Arg, respectively52 (1653, 1612) -helix and Asn, respectively52 (1661, 1658) stabilizing loop52 (1667, 1612) loop and Asn, respectively24-52 (1676, 1600) Arg and His, respectively24-52 (1676, 1612) Arn and Asn, respectivelyNote: The results suggest two epitope sites for this complex. Table IID. 2T2D SIRP- epitope map for mAb4 candidate as a function of temperature.2T2D Asynchronous Temperature cross peaks cross peak(oC) coordinates assignment\\4126-0462-6779 v3Attorney Docket No.: 761349.000125 24-52 (1636, 1612) b-sheet and Asn, respectively52 (1645, 1600) random coil and His, respectively24-36 (1645, 1625) random coil and Lys, respectively52 (1653, 1600) -helix and His, respectively24-52 (1667, 1600) loop and His, respectively52 (1658, 1653) destabilizing -helix region52 (1667, 1661) destabilizing loop region52 (1670, 1600) Gln and His, respectively52 (1670, 1661) Gln and Loop region, respectivelyNote: The results suggest two epitope sites for this complex. EQUIVALENTS

[0126] It will be readily apparent to those skilled in the art that other suitable modificationsand adaptions of the methods of the invention described herein are obvious and may be made using suitable equivalents without departing from the scope of the disclosure or the embodiments. Having now described certain compositions and methods in detail, the same will be more clearly understood by reference to the following examples, which are introduced for illustration only and not intended to be limiting. 35 \\4126-0462-6779 v3

Claims

Attorney Docket No.: 761349.000125 CLAIMS 1. A method for analyzing and determining the epitope of an antibody, comprising a) providing spectral data for test samples and reference samples obtained using aquantum cascade laser microscope under controlled temperature conditions, wherein, i) the test samples comprise an antibody or antigen-binding fragmentthereof and an antigen that the antibody binds with immunological specificity; ii) the reference samples comprising, antibody reference samples thatcomprises said antibody or antigen-binding fragment but not said antigen, and optionally an antigen reference samples that comprises said antigen but not said antibody or antigen-binding fragment; iii) the spectral data is obtained using a single slide cell arraycomprising cells that contain test samples and cells that contain reference samples; and iv) the spectral data include spectral images of each test sample andreference sample; b) identifying and selecting in spectral images of a test sample a region of interest and identifying and selecting in spectral images of an antibody reference sample a region of interest; c) obtaining spectral data for the test sample and the antibody reference sample for the region of interest; f) applying a baseline correction to the spectral data for the test sample and the antibody reference sample for the region of interest; g) generating weighted difference spectra for the test sample and antibody reference sample; h) applying a two-trace two-dimensional correlation to the weighted difference spectrato generate a synchronous plot ( , ) and an asynchronous plot ( , ), andidentifying signature cross peaks in the asynchronous plot toepitope or epitopes of the antibody or antigen-binding fragment; i) providing baseline corrected spectra for each test sample and each antibody reference sample, and optionally for each antigen reference sample, and generating covariance 36 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 data, and applying a two-dimensional correlation analysis to the covariance data to generate a synchronous correlation plot and an asynchronous correlation plot for each sample, identifying signature peaks in the plots, and analyze intensity changes of the peaks to define a sequential order of molecular events; and j) applying a two-dimensional co-distribution analysis to the baseline corrected spectra for each test sample and each antibody reference sample, and optionally for each antigen reference sample, to generate joint variance spectra and apply co-distribution functions to generate an asynchronous co-distribution plot and optionally a synchronous co-distribution plot, and identifying cross peaks in the asynchronous co- distribution plots, evaluating cross peaks in the asynchronous co-distribution plot, identify cross peaks associated with key interactions between the antigen and the antibody or antigen-binding fragment, and determine relative stability of the complex of the antibody or antigen-binding fragment bound to the epitope or epitopes.

2. A method for analyzing and determining the epitope of an antibody, comprising a) providing a slide cell array comprising cells that contain test samples and cellsthat contain reference samples, wherein i. the test samples comprise an antibody or antigen-binding fragmentthereof and an antigen that the antibody binds with immunological specificity, and ii. the reference samples comprising, antibody reference samples thatcomprises said antibody or antigen-binding fragment but not said antigen, and antigen reference samples that comprises said antigen but not said antibody or antigen-binding fragment; b) acquiring spectral data for the test samples and reference samples using a quantumcascade laser microscope under controlled temperature conditions, wherein, the spectral data include spectral images of each test sample and reference sample; c) identifying and selecting in spectral images of a test sample a region of interest and identifying and selecting in spectral images of an antibody reference sample a region of interest; d) obtaining spectral data for the test sample and the antibody reference sample for the region of interest; e) applying a baseline correction to the spectral data for the test sample and the antibody reference sample for the region of interest; 37 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 f) generating weighted difference spectra for the test sample and antibody reference sample; g) applying a two-trace two-dimensional correlation to the weighted difference spectrato generate a synchronous plot ( , ) and an asynchronous plot ( , ), andidentifying signature cross peaks in the asynchronous plot to determine the epitope or epitopes of the antibody orfragment;h) providing baseline corrected spectra for each test sample and each antibody reference sample, and optionally for each antigen reference sample, and generating covariance data, and applying a two-dimensional correlation analysis to the covariance data to generate a synchronous correlation plot and an asynchronous correlation plot for each sample, identifying signature peaks in the plots, and analyze intensity changes of the peaks to define a sequential order of molecular events; and i) applying a two-dimensional co-distribution analysis to the baseline corrected spectra for each test sample and each antibody reference sample, and optionally for each antigen reference sample, to generate joint variance spectra and apply co-distribution functions to generate an asynchronous co-distribution plot and optionally a synchronous co-distribution plot, and identifying cross peaks in the asynchronous co- distribution plots, evaluating cross peaks in the asynchronous co-distribution plot, identify cross peaks associated with key interactions between the antigen and the antibody or antigen-binding fragment, and determine relative stability of the complex of the antibody or antigen-binding fragment bound to the epitope or epitopes.

3. The method of claim 1 or claim 2, wherein the test samples include two or more samples in which the mol ratio of antigen:antibody differ.

4. The method of Claim 3, wherein the mol ratio of antigen:antibody is from about 1:1 to about 10:

1.

5. The method of Claim 3 or 4, wherein the mol ratio of antigen:antibody is sufficient to saturate the antibody binding sites in the test sample.

6. The method of any one of the preceding claims, wherein the antibody reference samples comprises samples in which the antibody concentration is the same as the antibody concentration in a corresponding test sample. 38 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 7. The method of any one of the preceding claims, wherein the antigen reference samples comprises samples in which the antigen concentration is the same as the antigen concentration in a corresponding test sample.

8. The method of any one of the preceding claims, wherein the test samples and reference samples are aqueous solutions.

9. The method of any one of the preceding claims, wherein the controlled temperature conditions comprise a temperature from about 4°C to about 95°C.

10. The method of claim 9, wherein the spectral data include data collected at two or more temperatures.

11. The method of claim 11, wherein the two or more temperatures are in the range of about 24°C to about 36°C or about 24°C to about 52°C.

12. The method of any one of the preceding claims, wherein the antigen is a soluble protein, the extracellular region of a membrane associated protein or a portion of either of the foregoing.

13. The method of claim 12, wherein the soluble protein or membrane associate protein is a glycoprotein.

14. The method of any one of the preceding claims, further comprising determining one or more of Asn and / or Gln deamidation, self-association, aggregation and stability of an antibody reference sample or antigen reference sample using both two-dimensional co-distribution and two-dimensional correlation analysis.

15. The method of any one of the preceding claims, further comprising applying a two-trace two-dimensional correlation analysis to the test sample, antibody reference sample and / orantigen reference sample to determine the extent of glycosylation, molecular nature of glycan interaction with antibody or antigen-binding fragment during formation of the complex between antibody or antigen-binding fragment and antigen an influence on stability of the complex between antibody or antigen-binding fragment and antigen.

16. The method of any one of the preceding claims, further comprising applying a two-trace two-dimensional correlation analysis to the test sample, antibody reference sample and / orantigen reference sample to determine the molecular nature of antibody or antigen-binding fragment and antigen complex formation and / or dissociation.

17. A method for analyzing candidate antibodies in a sample, comprising: 39 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 a) providing the sample in a slide containing at least one sample well and at least one reference well; b) acquiring at least one spectral image of the sample and at least one spectral image of the at least one reference using a quantum cascade laser microscope under controlled temperature conditions; c) identifying and selecting, in at least one of the acquired first spectral images, a region of interest; d) obtaining spectral data for the sample and the at least one reference for the region of interest; e) applying a baseline correction to the spectral data for the sample and the reference for the region of interest; f) generating weighted difference spectra for the sample and reference; f) applying a two-trace two-dimensional correlation to the weighted differencespectra to generate a synchronous spectrum ( , ) and asynchronous spectrum( , ), and identifying signature cross peaks in the asynchronous plot to determinea target epitope for a particular candidate antibody; g) applying a two-dimensional correlation analysis to the baseline corrected data and reference spectral data to generate a synchronous correlation plot and an asynchronous correlation plot, identifying signature peaks in the plots, and analyze intensity changes of the peaks to define a sequential order of molecular events; and h) applying a two-dimensional co-distribution analysis to the baseline corrected data to generate a synchronous correlation plot and an asynchronous co-distribution plot, and identifying cross peaks in the plots, and evaluating cross peaks in the co- distribution plots, identify cross peaks associated with key interactions between the target epitope and the particular candidate antibody, and determine relative stability.

18. A system for analyzing an antibody and / or determining the epitope of an antibody, comprising a) a data acquisition module configured to obtain spectral data of test samples andreference samples under controlled temperature conditions, wherein, 40 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 i) the test samples comprise an antibody or antigen-binding fragmentthereof and an antigen that the antibody binds with immunological specificity; ii) the reference samples comprising, antibody reference samples thatcomprises said antibody or antigen-binding fragment but not said antigen, and optionally an antigen reference samples that comprises said antigen but not said antibody or antigen-binding fragment; iii) the spectral data is obtained using a single slide cell arraycomprising cells that contain test samples and cells that contain reference samples; and iv) the spectral data include spectral images of each test sample andreference sample; and a correlation analysis module configured to: generate weighted difference spectra for the test sample and antibody reference sample; apply a two-trace two-dimensional correlation to the weighted difference spectra togenerate a synchronous plot ( , ) and an asynchronous plot ( , ), and identifyingsignature cross peaks in the asynchronous plot to determine the epitope or epitopes of the antibody or antigen-binding fragment; apply a two-dimensional correlation analysis to covariance data to generate a synchronous correlation plot and an asynchronous correlation plot for each sample, identify signature peaks in the plots, and analyze intensity changes of the peaks to define a sequential order of molecular events; and apply a two-dimensional co-distribution analysis to joint variance spectra for each test sample and each antibody reference sample, and optionally for each antigen reference sample, to generate an asynchronous co-distribution plot and optionally a synchronous co-distribution plot, and identifying cross peaks in the asynchronous co-distribution plots, evaluating cross peaks in the asynchronous co-distribution plot, identify cross peaks associated with key interactions between the antigen and the antibody or antigen-binding fragment, and determine relative stability of the complex of the antibody or antigen-binding fragment bound to the epitope or epitopes. 41 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 19. Non-transitory computer-readable medium comprising instructions which, when executed by one or more computers, cause the one or more computers to: b) obtain spectral data of test samples and reference samples under controlledtemperature conditions, wherein, j) the test samples comprise an antibody or antigen-binding fragmentthereof and an antigen that the antibody binds with immunological specificity; v) the reference samples comprising, antibody reference samples thatcomprises said antibody or antigen-binding fragment but not said antigen, and optionally an antigen reference samples that comprises said antigen but not said antibody or antigen-binding fragment; vi) the spectral data is obtained using a single slide cell arraycomprising cells that contain test samples and cells that contain reference samples; and vii) the spectral data include spectral images of each test sample andreference sample; and generate weighted difference spectra for the test sample and antibody reference sample; apply a two-trace two-dimensional correlation to the weighted difference spectra togenerate a synchronous plot ( , ) and an asynchronous plot ( , ), and identifyingsignature cross peaks in the asynchronous plot to determine the epitope or epitopes of the antibody or antigen-binding fragment; apply a two-dimensional correlation analysis to covariance data to generate a synchronous correlation plot and an asynchronous correlation plot for each sample, identify signature peaks in the plots, and analyze intensity changes of the peaks to define a sequential order of molecular events; and apply a two-dimensional co-distribution analysis to baseline corrected spectra for each test sample and each antibody reference sample, and optionally for each antigen reference sample, to generate joint variance spectra and apply co-distribution functions to generate an asynchronous co-distribution plot and optionally a synchronous co-distribution plot, and identifying cross peaks in the asynchronous co-distribution plots, evaluating cross peaks in the asynchronous co-distribution plot, identify cross peaks associated with key interactions 42 \\4126-0462-6779 v3Attorney Docket No.: 761349.000125 between the antigen and the antibody or antigen-binding fragment, and determine relative stability of the complex of the antibody or antigen-binding fragment bound to the epitope or epitopes. 43 \\4126-0462-6779 v3

Images