CHARACTERIZATION OF ANTIBODY MIXTURES BY MASS SPECTROMETRY

Related Applications

This application claims priority to United States Provisional Patent Application serial number 61/759,358, filed January 31, 2013, the entirety of which is hereby incorporated by reference.

Background of the Invention

Multi-component therapies based on recombinant antibodies (e.g., recombinant antibody compositions, monoclonal antibody mixtures or cocktails) represent promising new drugs for the treatment of various diseases and disorders including infectious diseases, cancer, neurological disorders, inflammation and immune disorders, and cardiovascular diseases because they simultaneously target multiple epitopes and, therefore, decrease the selective pressure for the development of resistant strains. The effectiveness of these therapies is derived, in part, through the heterogeneity of their individual components. This heterogeneity, however, present obstacles to the thorough characterization of individual antibodies in the multi-component mixtures. Thorough characterization is necessary to document and ensure standardized production, as well as to ensure safe production and therapeutic use.

For example, when proteins are used for diagnostic or therapeutic purposes, their structure has to be thoroughly characterized. Apart from the amino-acid sequence, all modifications, sequence alterations or changes taking place upon production, changes in the production process or upon storage have to be characterized, as described in industry guidelines ICH Q5E and Q6B. One of the most widely -used techniques for this

characterization is mass spectrometry (MS), due to its sensitivity and flexibility.

Currently, there are two main trends in MS protein characterization: bottom-up and top-down analysis. Bottom-up analysis starts typically with an enzymatic, or alternatively chemical, digestion of the intact protein into smaller peptides followed by separation and analysis of these peptides. Top-down analysis performs MS analysis directly on the level of the intact protein.

Bottom-up analysis currently is the technology of choice for rapid protein identification and quantification of large numbers of proteins because of the relatively easy handling of peptides. Peptide separation can easily be achieved on a chromatographic level. However, there are intrinsic limitations in this technique which arise from the so-called "protein inference problem". Bottom-up MS analysis identifies the peptides, not the proteins. The proteins from which these peptides originate are assigned solely based on statistics. This approach also means that mutations in or modifications on the protein sequence can be assigned only if they are explicitly observed. No statement is possible on non-observed deviations from the expected sequence. In addition, due to the upfront digestion, the context between the occurrence of a modification or a sequence mutation and a specific protein isoform is lost (e.g., if a 50% glycosylation on a given sequence position is observed, no statement can be made whether this position is always occupied by 50% or whether there are two equally abundant isoforms, one glycosylated, one completely unglycosylated).

Summary of the Invention

Embodiments of the present invention are based on the insight that the limitations of bottom-up MS analysis are particularly acute if applied to mixtures of antibodies (i.e., polyclonal antibody compositions). The present invention provides effective methods for accurate and comprehensive characterization of mixtures of intact or reduced antibodies based on top-down approaches of mass spectrometry analysis. Characterizations provided by embodiments of the invention include intact mass, amino acid sequence, and post- translational modification including glycosylation form distribution. The present invention is based on the surprising discovery that heterogeneous mixtures of antibody populations can be characterized by mass spectrometry without any digestion step or pre-treatment of the antibodies. This allows for an improved, simpler, more efficient, less expensive and more accurate way to analyze antibody populations through mass spectrometry. Embodiments of the present invention require reliable mass spectrometry techniques with superior resolution and mass accuracy, optionally paired with chromatography (e.g., HPLC). Mass

spectrometers useful in the present invention can include a triple quadrupole, an Orbitrap type MS, an ESI-TOF, a Q-TOF, a TOF, an ion trap or any other instrument of suitable sensitivity and mass resolution. In some embodiments, a Q Exactive Orbitrap MS system with appropriate data analysis is used. In one aspect, the present invention provides a method for characterizing an antibody mixture including a step of subjecting the antibody mixture to mass spectrometry analysis, wherein the antibody mixture comprises a plurality of distinct monoclonal antibody populations. In some embodiments, the antibody mixture comprises at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1, 12, 13, 14, 15, 16, 17, 18, 19, or 20 distinct monoclonal antibody populations.

In some embodiments, an inventive method according to the invention further comprises evaluation of the distribution of molecular masses, post-translational

modifications, and/or purity of the antibody mixture based on the mass spectrometry analysis. In some embodiments, the evaluation step comprises determining the molecular mass of each individual monoclonal antibody population. In some embodiments, the evaluation step comprises determining the charge distribution and/or glycosylation pattern of each individual monoclonal antibody population. In some embodiments, the evaluation step comprises determining the amino acid sequence of the light chain and/or heavy chain of each individual monoclonal antibody population.

In some embodiments, the evaluation step comprises determining the percentage of each individual monoclonal antibody population within the antibody mixture. In some embodiments, the evaluation step comprises determining the ratio(s) between distinct monoclonal antibody populations. In some embodiments, the evaluation step comprises determining the presence and/or amount of aggregated form of monoclonal antibodies (e.g., high molecular weight species). In some embodiments, the evaluation step comprises determining the purity of antibody mixture based on the presence and/or amount of aggregated form of monoclonal antibodies (e.g., high molecular weight species).

In some embodiments, each individual monoclonal antibody population comprises intact antibodies, F(ab')2, F(ab)2, Fab', Fab, ScFvs, diabodies, triabodies and/or tetrabodies. In particular embodiments, each individual monoclonal antibody population comprises intact antibodies (e.g., IgGl).

In some embodiments, the mass spectrometry analysis involves using a three- dimentional quadrupole ion trap mass spectrometer, a linear quadrupole ion trap mass spectrometer, or an Orbitrap mass spectrometer. In some embodiments, the mass

spectrometry analysis involves analyzing the data using one or more deconvolution algorithms. In some embodiments, the mass spectrometry analysis involves top-down fragmentation (e.g., msx HCD). In some embodiments, an inventive method according to the present invention further includes a step of determining if the antibody mixture satisfies a pre-determined quality standard by comparing the mass spectrometry analysis result to a control.

In another aspect, the present invention provides a method of manufacturing an antibody mixture comprising cultivating a polyclonal cell population comprising a plurality of distinct sub-populations, each of which expresses a distinct monoclonal antibody population, to produce an antibody mixture; and characterizing the quality of the antibody mixture using a method described herein. In some embodiments, the characterizing step is conducted before releasing a lot. In some embodiments, a method according to the invention further includes a step of adjusting a manufacturing condition based on the characterization of the antibody mixture. In some embodiments, pre-treatment of the antibody sample is not required before analysis. In some embodiments, pre-treatment comprises denaturing the antibodies. In some embodiments, pre-treatment comprises reduction of the antibody sample. In some embodiments, pre-treatment comprises digestion of the antibody sample. In some embodiments, digestion comprises enzymatic proteolysis.

Among other things, the present invention also provides an antibody mixture manufactured using a method described herein.

As used in this application, the terms "about" and "approximately" are used as equivalents. Any numerals used in this application with or without about/approximately are meant to cover any normal fluctuations appreciated by one of ordinary skill in the relevant art.

Other features, objects, and advantages of the present invention are apparent in the detailed description that follows. It should be understood, however, that the detailed description, while indicating embodiments of the present invention, is given by way of illustration only, not limitation. Various changes and modifications within the scope of the invention will become apparent to those skilled in the art from the detailed description.

Definitions

In order for the present invention to be more readily understood, certain terms are first defined below. Additional definitions for the following terms and other terms are set forth throughout the specification. Antibody: As used herein, the term "antibody" refers to a polypeptide that specifically binds to an epitope or antigen. In some embodiments, an antibody is a polypeptide whose amino acid sequence includes elements characteristic of an antibody -binding region (e.g., an antibody light chain or variable region or one or more complementarity determining regions ("CDRs") thereof, or an antibody heavy chain or variable region or one more CDRs thereof, optionally in presence of one or more framework regions). In some embodiments, an antibody is or comprises a full-length antibody. In some embodiments, an antibody is less than full-length but includes at least one binding site (comprising at least one, and preferably at least two sequences with structure of known antibody "variable regions"). In some embodiments, the term "antibody" encompasses any protein having a binding domain, which is homologous or largely homologous to an immunoglobulin-binding domain. In particular embodiments, an included "antibody" encompasses polypeptides having a binding domain that shows at least 99% identity with an immunoglobulin binding domain. In some embodiments, an "antibody" is any protein having a binding domain that shows at least 70%, 80%, 85%, 90%, or 95% identity with an immunoglobulin binding domain, for example a reference immunoglobulin binding domain. An included "antibody" may have an amino acid sequence identical to that of an antibody that is found in a natural source. As explained in greater detail below, antibodies may be prepared by any available means including, for example, isolation from a natural source, recombinant production in or with a host system, chemical synthesis, etc., or combinations thereof. An antibody may be monoclonal or polyclonal, mono-specific or bi-specific. An antibody may be a member of any

immunoglobulin class, including any of the human classes: IgG, IgM, IgA, IgD, and IgE. Antibodies may be chimeric or humanized mouse monoclonal antibodies. In general, humanized antibodies are human immunoglobulins (recipient antibody) in which residues from a complementary-determining region (CDR) of the recipient are replaced by residues from a CDR of a non-human species (donor antibody) such as mouse, rat or rabbit having the desired specificity, affinity, and capacity. In some embodiments, an antibody may be a human antibody or recombinant human antibody. As used herein, the term "antibody" includes any derivative of an antibody that possesses the ability to bind to an epitope of interest. In certain embodiments, the "antibody" is an antibody fragment that retains at least a significant portion of the full-length antibody's specific binding ability. Examples of antibody fragments include, but are not limited to, Fab, Fab', F(ab')2, scFv, Fv, dsFv diabody, and Fd fragments. Alternatively or additionally, an antibody fragment may comprise multiple chains that are linked together, for example, by disulfide linkages. Antigen: As used herein, the term "antigen" refers to a molecule or entity to which an antibody binds. In some embodiments, an antigen is or comprises a polypeptide or portion thereof. In some embodiments, an antigen is a portion of an infectious agent that is recognized by antibodies. In some embodiments, an antigen is an agent that elicits an immune response; and/or (ii) an agent that is bound by a T cell receptor (e.g., when presented by an MHC molecule) or to an antibody (e.g., produced by a B cell) when exposed or administered to an organism. In some embodiments, an antigen elicits a humoral response (e.g., including production of antigen-specific antibodies) in an organism; alternatively or additionally, in some embodiments, an antigen elicits a cellular response (e.g., involving T-cells whose receptors specifically interact with the antigen) in an organism. It will be appreciated by those skilled in the art that a particular antigen may elicit an immune response in one or several members of a target organism (e.g., mice, rabbits, primates, humans), but not in all members of the target organism species. In some embodiments, an antigen elicits an immune response in at least about 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% of the members of a target organism species. In some embodiments, an antigen binds to an antibody and/or T cell receptor, and may or may not induce a particular physiological response in an organism. In some embodiments, for example, an antigen may bind to an antibody and/or to a T cell receptor in vitro, whether or not such an interaction occurs in vivo. In general, an antigen may be or include any chemical entity such as, for example, a small molecule, a nucleic acid, a polypeptide, a carbohydrate, a lipid, a polymer [in some embodiments other than a biologic polymer (e.g., other than a nucleic acid or amino acid polymer)] etc. In some embodiments, an antigen is or comprises a polypeptide. In some embodiments, an antigen is or comprises a glycan. Those of ordinary skill in the art will appreciate that, in general, an antigen may be provided in isolated or pure form, or alternatively may be provided in crude form (e.g., together with other materials, for example in an extract such as a cellular extract or other relatively crude preparation of an antigen-containing source). In some embodiments, antigens utilized in accordance with the present invention are provided in a crude form. In some embodiments, an antigen is or comprises a recombinant antigen.

Approximately or about: As used herein, the term "approximately" or "about," as applied to one or more values of interest, refers to a value that is similar to a stated reference value. In certain embodiments, the term "approximately" or "about" refers to a range of values that fall within 25%, 20%, 19%, 18%, 17%, 16%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, or less in either direction (greater than or less than) of the stated reference value unless otherwise stated or otherwise evident from the context (except where such number would exceed 100% of a possible value).

Characteristic portion: As used herein, the term a "characteristic portion" of a substance, in the broadest sense, is one that shares some degree of sequence or structural identity with respect to the whole substance. In certain embodiments, a characteristic portion shares at least one functional characteristic with the intact substance. For example, a "characteristic portion" of a protein or polypeptide is one that contains a continuous stretch of amino acids, or a collection of continuous stretches of amino acids, that together are characteristic of a protein or polypeptide. In some embodiments, each such continuous stretch generally contains at least 2, 5, 10, 15, 20, 50, or more amino acids. In general, a characteristic portion of a substance (e.g., of a protein, antibody, etc.) is one that, in addition to the sequence and/or structural identity specified above, shares at least one functional characteristic with the relevant intact substance; epitope-binding specificity is one example. In some embodiments, a characteristic portion may be biologically active.

Expression: As used herein, "expression" of a nucleic acid sequence refers to one or more of the following events: (1) production of an RNA template from a DNA sequence (e.g., by transcription); (2) processing of an RNA transcript (e.g., by splicing, editing, 5' cap formation, and/or 3 ' end formation); (3) translation of an RNA into a polypeptide or protein; and/or (4) post-translational modification of a polypeptide or protein.

Functional: As used herein, a "functional" biological molecule is a biological molecule in a form in which it exhibits a property and/or activity by which it is characterized. A biological molecule may have two functions (i.e., bifunctional) or many functions (i.e., multifunctional).

Functional equivalent or derivative: As used herein, the term "functional equivalent" or "functional derivative" denotes, in the context of a functional derivative of an amino acid sequence, a molecule that retains a biological activity (either function or structural) that is substantially similar to that of the original sequence. A functional derivative or equivalent may be a natural derivative or is prepared synthetically. Exemplary functional derivatives include amino acid sequences having substitutions, deletions, or additions of one or more amino acids, provided that the biological activity of the protein is conserved. The substituting amino acid desirably has chemico-physical properties which are similar to that of the substituted amino acid. Desirable similar chemico-physical properties include, similarities in charge, bulkiness, hydrophobicity, hydrophilicity, and the like.

Improve, increase, or reduce: As used herein, the terms "improve," "increase" or "reduce," or grammatical equivalents, indicate values that are relative to a baseline measurement, such as a measurement in the same individual prior to initiation of the treatment described herein, or a measurement in a control subject (or multiple control subject) in the absence of the treatment described herein. A "control subject" is a subject afflicted with the same form of disease as the subject being treated, who is about the same age as the subject being treated.

In vitro: As used herein, the term "in vitro" refers to events that occur in an artificial environment, e.g., in a test tube or reaction vessel, in cell culture, etc., rather than within a multi-cellular organism.

In vivo: As used herein, the term "in vivo" refers to events that occur within a multicellular organism, such as a human and a non-human animal. In the context of cell-based systems, the term may be used to refer to events that occur within a living cell (as opposed to, for example, in vitro systems).

Isolated: As used herein, the term "isolated" refers to a substance and/or entity that has been (1) separated from at least some of the components with which it was associated when initially produced (whether in nature and/or in an experimental setting), and/or (2) produced, prepared, and/or manufactured by the hand of man. Isolated substances and/or entities may be separated from at least about 10%, about 20%, about 30%, about 40%, about 50%, about 60%, about 70%, about 80%, about 90%, about 95%, about 98%, about 99%, substantially 100%, or 100% of the other components with which they were initially associated. In some embodiments, isolated agents are more than about 80%, about 85%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about

97%, about 98%, about 99%, substantially 100%, or 100% pure. As used herein, a substance is "pure" if it is substantially free of other components. As used herein, the term "isolated cell" refers to a cell not contained in a multi-cellular organism.

Pharmaceutically acceptable: As used herein, the term "pharmaceutically acceptable", refers to substances that, within the scope of sound medical judgment, are suitable for use in contact with the tissues of human beings and animals without excessive toxicity, irritation, allergic response, or other problem or complication, commensurate with a reasonable benefit/risk ratio.

Polypeptide: The term "polypeptide" as used herein refers a sequential chain of amino acids linked together via peptide bonds. The term is used to refer to an amino acid chain of any length, but one of ordinary skill in the art will understand that the term is not limited to lengthy chains and can refer to a minimal chain comprising two amino acids linked together via a peptide bond. As is known to those skilled in the art, polypeptides may be processed and/or modified.

Protein: The term "protein" as used herein refers to one or more polypeptides that function as a discrete unit. If a single polypeptide is the discrete functioning unit and does not require permanent or temporary physical association with other polypeptides in order to form the discrete functioning unit, the terms "polypeptide" and "protein" may be used interchangeably. If the discrete functional unit is comprised of more than one polypeptide that physically associate with one another, the term "protein" refers to the multiple polypeptides that are physically coupled and function together as the discrete unit.

Specific binding: As used herein, the terms "specific binding" or "specific for" or "specific to" refer to an interaction (typically non-covalent) between a target entity (e.g., a target protein or polypeptide) and a binding agent (e.g., an antibody, such as a provided antibody). As will be understood by those of ordinary skill, an interaction is considered to be "specific" if it is favored in the presence of alternative interactions. In many embodiments, an interaction is typically dependent upon the presence of a particular structural feature of the target molecule such as an antigenic determinant or epitope recognized by the binding molecule. For example, if an antibody is specific for epitope A, the presence of a polypeptide containing epitope A or the presence of free unlabeled A in a reaction containing both free labeled A and the antibody thereto, will reduce the amount of labeled A that binds to the antibody. It is to be understood that specificity need not be absolute. For example, it is well known in the art that numerous antibodies cross-react with other epitopes in addition to those present in the target molecule. Such cross-reactivity may be acceptable depending upon the application for which the antibody is to be used. Specificity may be evaluated in the context of additional factors such as the affinity of the binding molecule for the target molecule versus the affinity of the binding molecule for other targets (e.g., competitors). If a binding molecule exhibits a high affinity for a target molecule that it is desired to detect and low affinity for non-target molecules, the antibody will likely be an acceptable reagent for immunodiagnostic purposes. Once the specificity of a binding molecule is established in one or more contexts, it may be employed in other, preferably similar, contexts without necessarily re-evaluating its specificity.

Substantially: As used herein, the term "substantially" refers to the qualitative condition of exhibiting total or near-total extent or degree of a characteristic or property of interest. One of ordinary skill in the biological arts will understand that biological and chemical phenomena rarely, if ever, go to completion and/or proceed to completeness or achieve or avoid an absolute result. The term "substantially" is therefore used herein to capture the potential lack of completeness inherent in many biological and chemical phenomena.

Brief Description of the Drawings

The Figures described below, that together make up the Drawings, are for illustration purposes only, not for limitation.

Figure 1 shows exemplary MS spectra for antibody mixture having three antibodies

(i.e., 42.11.D4, 26.3. E2, and 5.6.H9). (A) Exemplary raw MS spectrum obtained from a bench-top Q Exactive Quadrupole Orbitrap mass spectrometer for antibody mixture having three antibodies. (B) Exemplary zoomed- in image of the raw spectrum for antibody mixture having three antibodies.

Figure 2 depicts exemplary results illustrating deconvoluted molecular mass for each antibody in the mixture having three antibodies (i.e., 42.11.D4, 26.3.E2, and 5.6.H9). (A) Exemplary protein deconvolution spectra obtained using ReSpect algorithm and separately showing the molecular mass for each of the three antibodies in the mixture (shown by arrows). (B) Table showing exemplary results obtained using ReSpect algorithm.

Figure 3 depicts exemplary spectrum of the light chain for antibodies in the mixture having three antibodies (i.e., 42.11.D4, 26.3.E2, and 5.6.H9). (A) Exemplary raw spectrum for the MS analysis of the light chain of the antibodies in the mixture with three antibodies. (B) Exemplary protein deconvolution spectra obtained using ReSpect algorithm and showing molecular mass of the light chain of antibodies in the mixture (shown by ovals).

Figure 4 shows exemplary MS spectra for antibody mixture having five antibodies

(i.e., 42.11.D4, 26.3.E2, 5.6.H9, 5.55.D2, and 42.18.E12) . (A) Exemplary raw MS spectrum obtained from a bench-top Q Exactive Quadrupole Orbitrap mass spectrometer for antibody mixture having five antibodies. (B) Exemplary zoomed-in image of the raw spectrum for antibody mixture having five antibodies.

Figure 5 depicts exemplary results illustrating deconvoluted molecular mass for each antibody in the mixture having five antibodies (i.e., 42.1 1.D4, 26.3. E2, 5.6.H9, 5.55.D2, and 42.18. E12). (A) Exemplary protein deconvolution spectra obtained using ReSpect algorithm and separately showing the molecular mass for each of the five antibodies in the mixture (shown by ovals). (B) Table showing exemplary results obtained using ReSpect algorithm.

Figure 6 depicts exemplary spectrum of the light chain for antibodies in the mixture having five antibodies (i.e., 42.11.D4, 26.3.E2, 5.6.H9, 5.55.D2, and 42.18.E12). (A) Exemplary raw spectrum for the MS analysis of the light chain of the antibodies in the mixture with five antibodies. (B) Exemplary protein deconvolution spectra obtained using ReSpect algorithm and showing molecular mass of the light chain of antibodies in the mixture (shown by ovals).

Figure 7 depicts the exemplary spectra for quantitation of an intact antibody (antibody

5.55.D2). (A) Exemplary base peak chromatograms for an intact antibody analyzed at different concentrations. (B) Exemplary base peak chromatograms of three separate injections at a concentration of 12.5 ng. (C) Exemplary raw MS spectrum obtained from a bench-top Q Extractive Quadrupole Orbitrap Mass Spectrometer for antibody 5.55.D2 over three separate injections at a concentration of 12.5 ng. (D) Exemplary base peak

chromatograms of three separate injections at a concentration of 500 ng. (E) Exemplary raw MS spectrum for antibody 5.55.D2 over three separate injections at a concentration of 500 ng-

Figure 8 depicts processing of the quantitation for an intact antibody using a bench- top Q Extractive Quadrupole Orbitrap Mass Spectrometer. (A) Depicts an exemplary screenshot for the processing method performed on antibody 5.55.D2 using a bench-top Q Extractive Quadrupole Orbitrap Mass Spectrometer. (B) Depicts a calibration curve for antibody 5.55.D2 between concentrations 12.5 ng and 500 ng. (C) Table showing exemplary quantitation results for an intact antibody at a concentration range of 12.5 ng to 500 ng.

Figure 9 depicts exemplary zoomed-in images of the raw spectra for isotopically resolved light chains for four separate antibodies carried in a top-down characterization. (A) Depicts the spectrum for intact 26.3.E2 light chain. (B) Depicts the spectrum for intact 5.6.H9 light chain. (C) Depicts the spectrum for intact 42.18.E12 light chain. (D) Depicts the spectrum for intact 5.55.D2 light chain.

Figure 10 depicts the results of light chain characterization of antibody 26.3. E2 using the top-down approach. (A) Exemplary raw spectrum of the MS analysis of the light chain antibody 26.3. E2. (B) Depicts the location of post-translational modifications based upon observed and theoretical masses obtained for the light chain.

Figure 11 depicts the results of light chain characterization of antibody 5.6.H9 using the top-down approach. (A) Exemplary raw spectrum of the MS analysis of the light chain antibody 5.6.H9. (B) Depicts the location of post-translational modifications based upon observed and theoretical masses obtained for the light chain.

Figure 12 depicts the results of light chain characterization of antibody 42.18.E12 using the top-down approach. (A) Exemplary raw spectrum of the MS analysis of the light chain antibody 42.18.E12. (B) Depicts the location of post-translational modifications based upon observed and theoretical masses obtained for the light chain.

Figure 13 depicts the results of light chain characterization of antibody 5.55.D2 using the top-down approach. (A) Exemplary raw spectrum of the MS analysis of the light chain antibody 5.55.D2. (B) Depicts the location of post-translational modifications based upon observed and theoretical masses obtained for the light chain.

Detailed Description

The present invention provides, among other things, a method for characterizing an antibody mixture including a step of subjecting the antibody mixture to mass spectrometry analysis, wherein the antibody mixture comprises a plurality of distinct monoclonal antibody populations. The present invention also provides methods for characterizing antibody mixtures through mass spectrometry without any treatment or digestion of the mixture prior to spectrometric analysis. Characterization can include detections of micro-heterogeneity among antibodies including post-translational modifications such as glycosylation. Particular embodiments of the invention comprise various combinations of liquid chromatography, mass spectrometry and data processing (i.e., protein deconvolution) to characterize, identify and quantify complex antibody mixtures. Antibody Mixtures

Antibody mixtures (i.e., polyclonal antibody compositions) for use in embodiments of the invention may include any immunoglobulin, whether natural or wholly or partially synthetically produced. All derivatives of antibodies that maintain specific binding ability may also be included. Antibody mixtures as described herein may contain any protein having a binding domain which is homologous or largely homologous to an immunoglobulin- binding domain. Such proteins may be derived from natural sources, or partly or wholly synthetically produced. An antibody in the mixtures may be monoclonal or polyclonal. In some embodiments, the antibody mixtures comprises mixtures of monoclonal antibodies directed to a particular antigen or for treatment of a particular disease. Antibody mixtures may comprise members any immunoglobulin class, including any of the human classes: IgG, IgM, IgA, IgD, and IgE. In certain embodiments, an antibody may be a member of the IgG immunoglobulin class. In some embodiments, the antibodies for analysis can be composed of one or more different antibody subclasses or isotypes, such as human isotypes IgGl, IgG2, IgG3, IgG4, IgAl and IgA2, or isotypes from other species such as murine isotypes IgGl, IgG2a, IgG2b, IgG3, and IgA. Antibody mixtures as described herein may comprise antibody fragments or characteristic portions of an antibody, which are used interchangeably and refer to any derivative of an antibody which is less than full-length. Examples of antibody fragments in antibody mixtures of some embodiments include, but are not limited to, Fab, Fab', F(ab')2, scFv, Fv, dsFv diabody, and Fd fragments. Such fragments may be produced by any means. For example, an antibody fragment may be enzymatically or chemically produced by fragmentation of an intact antibody and/or it may be recombinantly produced from a gene encoding the partial antibody sequence. Alternatively or additionally, an antibody fragment may be wholly or partially synthetically produced. An antibody fragment may optionally comprise a single chain antibody fragment. Alternatively or additionally, an antibody fragment may comprise multiple chains which are linked together, for example, by disulfide linkages. An antibody fragment may optionally comprise a multimolecular complex.

It is contemplated that the present invention may be used to characterize an antibody mixture. As used herein, the term "antibody mixture" refers to a composition that contains a plurality of distinct antibody (e.g., monoclonal antibody) populations. An antibody mixture may contain any number of distinct monoclonal antibody populations. For example, an antibody mixture may comprise at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25, 30, 35, 40, 45, 50, or more distinct monoclonal antibody populations. An antibody mixture may comprise polyclonal antibodies. In some embodiments, an antibody mixture comprises one or more human or humanized antibodies. In some embodiments, the antibody mixtures comprises one or more bispecific antibodies. In some embodiments, an antibody mixture may be capable of binding one, two, three, four, five or more different epitopes. In some embodiments, the antibody mixture comprises two or more antibodies that share a common light chain. In some embodiments, the antibody mixture comprises two or more antibodies that share a common heavy chain.

Mixtures of different antibodies can target different epitopes of the same antigen or different antigens involved in the same disease resulting in a synergistic effect with increased efficacy and potency. The co-expression of mixtures of recombinant monoclonal antibodies have improved pharmacodynamic properties and greater cost reductions.

It is contemplated that the present invention may be used to detect modifications to antibodies such as post-translational modifications including glycosylation. Additional modifications that can be detected on antibodies are: disulfide pairings, N- and C- terminal modifications such as pyroGlu, Lys and Gly clippings. In some embodiments, the present invention can be used to characterize glycosylated antibodies through the use of mass spectrometry without prior digestion or pre-treatment. For example, the methods of the present invention can be used to characterize glycosylated antibodies without proteolytic digestion through the use of mass spectrometers such as ion trap and Orbitrap MS.

It is contemplated that the present invention may be used to detect other modifications including but not limited to isomerization, oxidation, deamidation, and N- and C- terminal variants.

The present invention further relates to methods for characterizing populations of different antibody species in recombinant polyclonal compositions. The methods of the present invention are useful for quantitative analysis and can be used, for example, to analyze consistency between antibody mixes as well as to assess the compositional stability during manufacturing runs and to determine whether a certain mix meets certain standards. The methods of the present invention can also be used to select for clones to create a polyclonal cell bank comprising clones that generate specific preferred antibodies. In some

embodiments, the invention provides for methods for detecting variations between different populations of antibodies in recombinant polyclonal antibody compositions. The present invention relates to methods for simultaneously analyzing the in vivo clearance of individual antibodies constituting a recombinant polyclonal

antibody/composition/product in serum from an individual such as a human being for pharmacokinetic studies. The methods of the present invention can also be used to characterize the polyclonality in a therapeutic/drug product for treatment. In some embodiments, the methods of the present invention are used to quantitate one or more recombinant antibodies in in-process samples.

Production of antibody mixtures

The techniques and protocols described herein are applicable to any antibody mixture and are not limited by the method of antibody production. The present invention may be used to characterize antibody mixtures produced by various methods. Polyclonal antibodies may be produced by injecting a host animal such as rabbit, rat, goat, mouse or other animal with an immunogen of choice. The sera are extracted from the host animal and are screened to obtain polyclonal antibodies which are specific to the immunogen. Methods of screening for polyclonal antibodies are well known to those of ordinary skill in the art such as those disclosed in Harlow & Lane, ANTIBODIES: A LABORATORY MANUAL, (Cold Spring Harbor Laboratories, Cold Spring Harbor, NY: 1988), incorporated herein by reference. In some embodiments, specific monoclonal antibodies may be separately produced by conventional recombinant or hybridoma technologies, purified, and then combined into an antibody mixture. For example, recombinant antibodies may be produced by cloning cDNA or genomic DNA encoding the immunoglobulin light and heavy chains of the desired antibody from a hybridoma or immune cell (e.g., plasma cells or B cells) that encodes, expresses or produces an antibody of interest. The cDNA or genomic DNA encoding those polypeptides is then inserted into recombinant expression vectors so that both genes are operatively linked to their own regulatory sequences, for control of transcription and translation. The expression vector and the regulatory sequences for control of expression are chosen to be compatible for expression in the selected host cell. Typically, both the heavy and light chain genes are inserted into the same expression vector so that expression of both is operatively linked. Prokaryotic or eukaryotic cells may be used as expression hosts. Expression in eukaryotic host cells (e.g., yeast cells, CHO cells, HEK-293 cells) is preferred because such cells are more likely than prokaryotic cells to assemble and secrete properly folded and immunologically active antibody. In some embodiments, an antibody mixture may be produced by cultivating a polyclonal cell population comprising a plurality of distinct sub-populations, each of which expresses a distinct monoclonal antibody population. In particular embodiments, an antibody mixture may be produced by cultivating a stable polyclonal cell population expressing multiple recombinant antibodies without requiring clonal selection (i.e., the use of a selection marker to select an individual cell clone— the progenitor cell— that is then propagated as an individual cell line). In some embodiments, stable cell populations expressing multiple recombinant antibodies can be generated from large numbers of individual subpopulations transformed with vectors encoding distinct antibody or components of an antibody (e.g., light chain and/or heavy chain of an antibody).

In some embodiments, a polyclonal cell population can be generated by first generating individual cell populations that are stable and then mixing the stable individual cell populations. Stable polyclonal cell populations can be generated when sufficient numbers of individual cells are used to produce a cell population. It is contemplated herein that when cell populations with greater than 10,000 individual cells are mixed, the combined cell population yields a stable polyclonal cell population to produce a desired antibody mixture.

In some embodiments, polyclonal antibody mixtures are produced through development of a mixture of monoclonal antibodies through a single production platform (such as a single batch or single production vessel) followed by structural/functional characterization of the mixture. Recombinant polyclonal antibody compositions can be obtained from a single polyclonal cell culture at different time points during cultivation. Recombinant polyclonal antibody compositions can also be obtained from different polyclonal cell cultures at different time points. The use of single production runs to generate antibody mixtures reduces costs and saves production time.

In some embodiments, individual recombinant monoclonal antibody producer cell lines are generated (Rasmussen et al. BIOTECHNOL. LETT., 2007 29:845-852). These cell lines can be collectively drawn upon as a library stock. Selected cell lines can then be mixed to create a polyclonal cell culture used for generation of different target specific recombinant polyclonal antibodies. In some embodiments, mammalian cells are a preferred expression host for recombinant antibody production. Use of mammalian cells allows for proper protein folding, assembly, and post-translational modifications such as glycosylation. Mammalian production cells for generating recombinant polyclonal antibodies include, but is not limited to, Chinese hamster ovary cells lines (CHO) but also mouse myeloma cell lines (SP2/0 and NSO) and human cell lines (embryonic kidney cells (HEK-293) and retinal cells (Per.C6)). The polyclonality of the resulting antibody mixture can be assessed using assays specific for the variable region as a complement to the mass spectrometry techniques of the present invention. Furthermore, through the use of the disclosed mass spectrometry techniques, one could use marker peptides unique to individual antibodies to detect and confirm the presence of different antibodies in the polyclonal mixture.

In exemplary embodiments, stable cell populations are obtained by transforming a population of host cells using preferential integration of at least one vector comprising at least one copy of a nucleic acid sequence encoding an antibody or a component thereof into the host cell genomic DNA to generate large populations of individual transformed cells. In some embodiments, preferential integration can be used to produce cell populations expressing multiple recombinant antibodies. Exemplary vectors for preferential integration are described in in U.S. Patent No. 8,617,881, which is incorporated by reference herein in its entirety.

As will be understood by those of skill in the art, the following reagents, techniques and protocols may be used to produce antibody mixtures.

The term "cell line" as used herein refers to a population of cells derived from a single progenitor cell. Persons skilled in the art would understand that because cells contained in a given cell line are derived from a single progenitor cell that each of the cells in the cell line, at least initially, share the same genomic characteristics (see, e.g., Poste et ah, PNAS 1981 78:6226-6230). A cell line is typically generated by selecting for an individual progenitor cell clone (e.g., using a selection marker, e.g., an antibiotic selection marker, e.g., neomycin, blasticidin, puromycin, or hygromycin). The individual progenitor cell clone is then cultured to expand the cell line capable of producing a gene product (e.g., antibody) of interest.

The terms "cell population", "cell mixture", "transformed cell population" and "recombinant cell population" as used herein refer to a population of cells that share a common genetic background, but are not genetically identical. In a cell population expressing a single recombinant polypeptide, a DNA fragment of interest is integrated into a plurality of cells, where the integration site in each cell may be at one or more locations in the genome, creating cells in the population that are not genetically identical. A cell population differs from a cell line because the cell population is not derived from an individual protein cell and/or is not comprised of cells with the same genetic content. For example, in the cell populations described herein there is no selection process to identify a single progenitor cell followed by clonal expansion of the single progenitor cell. Rather the cell populations described herein are a mixture of cells where each cell in the mixture may have the DNA integrated in one or more locations in the genome. Stability of the cell population expressing a single recombinant polypeptide is achieved by population dynamics.

In exemplary embodiments, the cell population, cell mixture or transformed cell populations described herein may be a polyclonal, e.g., it expresses multiple recombinant antibodies. In a polyclonal cell population more than one DNA fragment of interest is integrated into a plurality of cells, wherein the integration site in each cell may be at one or more locations in the genome, creating cells that are not genetically identical. A polyclonal cell population as described herein may be generated by mixing two or more monoclonal cell populations. Alternatively, a polyclonal cell population maybe generated by a bulk transformation procedure wherein a plurality of DNA fragments of interest is integrated into a plurality of host cells (e.g., one DNA fragment of interest per host cell). A polyclonal cell population as described herein does not require a selection process to identify individual progenitor cells followed by clonal expansion of individual cell lines (e.g., a polyclonal cell population is not a mixture of individual cell lines each derived from an individual progenitor cell). Stability of the polyclonal cell population is achieved by population dynamics.

The term "transformation" as used herein refers to any method for introducing foreign

DNA into a cell. As used herein, "transformation" is a broad term that includes methods for introducing foreign DNA into a cell including transfection, infection, transduction or fusion of a donor cell and an acceptor cell.

In some embodiments, antibody mixtures are produced using random integration of antibody-encoding nucleic acid vectors into host cell genomes. The term "random integration" as used herein refers to the process of integrating a DNA fragment of interest (e.g., a partial or complete DNA encoding a protein of interest) into the genome of a cell, where the fragment of DNA can be integrated in any part of that genome with equal probability. The person skilled in the art would understand that random integration refers to a transformation procedure where nothing is done to guide the expression construct to a predetermined position in a host cell genome. For example, in certain embodiments, random integration refers to an integration process which is performed naturally by the host cell machinery without the aid of extraneously added sequences or enzymes that affect the natural integration site (H. Wurtetle, Gene Therapy, 2003, 10: 1791-1799).

In some embodiments of the invention, recombinant antibodies (e.g., recombinant human antibodies) are produced by site-specific integration of nucleic acid vectors encoding the antibodies into genomes of host cells. The term "site-specific integration" as used herein refers to the process of integrating a DNA fragment of interest (e.g., a partial or complete DNA encoding a protein of interest) into the genome of a cell, where the DNA fragment of interest is targeted to a specific sequence in that genome. The specific target sequence can occur naturally in the genome or may be engineered into the genome of the host cell (e.g., engineering of a FRT-site into the genome of a cell for Flp recombinase mediated integration).

An example of site-specific integration is the use of Flp recombinase to target integration of a DNA fragment of interest to a specific site in a host cell genome. The specific site of integration occurs at a DNA sequence known as the FRT site. FLP is an exception among integrases in that it is highly specific to FRT sites. Site specific integration of DNA to the FRT site can be achieved by including FRT site DNA sequence in the DNA fragment for targeted integration. However, since FRT sites do not occur naturally in most genomes, typically, a FRT site must be integrated into the genome of a host cell of interest before introduction of the DNA fragment of interest.

In some embodiments of the invention, recombinant antibodies (e.g., recombinant human antibodies) are produced by preferential integration of nucleic acid vectors encoding the antibodies into genomes of host cells. The term "preferential integration" as used herein refers to the process of integrating a DNA fragment of interest (e.g., a partial or complete DNA encoding a protein of interest) into the genome of a cell, where the fragment of DNA is targeted to a predetermined region of the genome (but not to a single defined site, e.g., a FRT site). As described herein, preferential integration is not a random integration event because there is an increased probability that the DNA of interest will integrate into a defined region or specific site in the host cell genome. Preferential integration also differs from the site- specific integration because there is variability in the integration site.

Examples of preferential integration include integration using an AAV system by the

AAV rep protein(s). Stable integration into the AAVS 1 site is mediated by inverted terminal repeat or "ITR" sequences, where each ITR seqeunce comprise 145 base pairs (see, e.g., Bohenzky et al, VIROLOGY, 1988, 166(2): 316-27; Wang, XS et al, J. MOL. BIOL., 1995, 250(5):573-80; Weitzman, MD et al, PNAS, 1994, 91(13):5808-12). Recent data suggests the AAV vector integrates into the host cell into a rep targeting sequence ("RTS") such as an AAVS l site located on chromosome 19 with an efficiency of -10%, with the remaining 90% spread over other RTS sites across the human genome (as described in e.g., Smith RH, GENE THERAPY, 2008, 15:817-822 and Huser et al, PLoS PATHOG, 6(7): el000985.

doi: 10.1371/journal.ppat.1000985).

Other examples of preferential integration include use of a retroviral vector system, wherein the retroviral vector integrates into open chromatin domains which encompass about 5% of the total genome; use of a phage integrase < C3 lor lambda integrase, which carries out recombination between the attP site and the attB site (A.C. Groth et al, PNAS, 2000, 97:59995-6000); and use of a Cre recombinase and a variety of lox sites such as loxP from bacteriophage PI or variants or mutants thereof, e.g., lox43, lox 44, lox 66, lox71, lox 75, lox 76, and lox 51 1 (C. Gorman and C. Bullock, CURR. OPINION IN BIOTECHNOLOGY,

2000: 11 :455-460). Differences in preferential integration may also be observed with certain retroviral integrases, e.g., HIV integrates preferably into chromosomal regions rich in expressed genes, MMLV integrates preferably near transcription start sites, and ASLV which integrates with weak preference for active genes, but no preference for transcription start regions (Mitchell et al, PLoS BIOL, 2004, 2:e234).

Exemplary methods for producing antibody mixtures are described in U.S. Patent No.

8,617,881, which is incorporated by reference herein in its entirety.

Liquid Chromatography

In some embodiments, the methods and systems of the present invention comprise liquid chromatography (LC) in combination with mass spectrometry and data processing (i.e., protein deconvolution) to characterize, identify and quantify complex antibody mixtures. LC may be used for one or more of desalting, separation of antibody mixtures and/or separation of light and heavy chains. In preferred embodiments, the antibody mixture is not subjected to enzymatic digestion before or after chromatographic separation. In certain embodiments, the present invention comprises quantitative liquid chromatography tandem mass spectrometry (LC/MS/MS). In some embodiments, two-dimensional or tandem LC is used. Particular embodiments of the invention comprise additional steps of liquid-liquid extractions, dialysis, sample dilution, and/or sample dehydration steps prior to analysis by mass spectrometry.

As used herein, "chromatography" refers to a process in which a chemical mixture carried by a liquid or gas is separated into components as a result of differential distribution of the chemical entities as they flow around or over a stationary liquid or solid phase.

As used herein, "liquid chromatography" (LC) means a process of selective retardation of one or more components of a fluid solution as the fluid uniformly percolates through a column of a finely divided substance, or through capillary passageways. The retardation results from the distribution of the components of the mixture between one or more stationary phases and the bulk fluid, (i.e., mobile phase), as this fluid moves relative to the stationary phase(s). "Liquid chromatography", as used in particular embodiments, includes reverse phase liquid chromatography (RPLC), high performance liquid

chromatography (HPLC), fast performance liquid chromatography (FPLC), low pressure liquid chromatography (LPLC), and high turbulence liquid chromatography (HTLC).

As used herein, the term "HPLC" or "high performance liquid chromatography" refers to liquid chromatography in which the degree of separation is increased by forcing the mobile phase under pressure through a stationary phase, typically a densely packed column. "FPLC" or "fast performance/protein liquid chromatography", as used in particular embodiments, refers to liquid chromatography used primarily to analyze or purify mixtures of proteins. FPLC works similarly to HPLC, however the buffer pressure is much lower (less than 5 bar) and the flow rate is high (1-5 ml/min.). As used herein and in particular embodiments, the term "LPLC" or "low pressure liquid chromatography" refers to an automated, low pressure form of liquid chromatography. As with FPLC, LPLC operates at lower pressure than HPLC. LPLC columns are packed with larger particles than HPLC columns and therefore cannot handle the high back pressure that is generated. As a consequence, LPLC usually provides a lower resolution of analytes than FPLC systems.

In some embodiments, mixtures of intact antibodies or individual antibodies may be reduced prior to chromatographic separation. For example, samples may be incubated for an appropriate time (e.g., about one hour) under appropriate conditions (e.g., temperature of about 60 °C) in 6 M guanidine-HCL containing 5 mM DDT. In some embodiments, mixtures of intact antibodies or individual antibodies may be reduced by treatment with urea prior to chromatographic separation.

The chromatographic column typically includes a medium (i.e., a packing material) to facilitate separation of chemical moieties (i.e., fractionation). The medium may include minute particles. The particles include a bonded surface that interacts with the various chemical moieties to facilitate separation of the chemical moieties such as the biomarker analytes quantified in the experiments herein. One suitable bonded surface is a hydrophobic bonded surface such as an alkyl bonded surface. Alkyl bonded surfaces may include C-4, C- 8, or C- 18 bonded alkyl groups. The chromatographic column includes an inlet port for receiving a sample and an outlet port for discharging an effluent that includes the fractionated sample. A sample may be applied to the column at the inlet port, eluted with a solvent or solvent mixture, and discharged at the outlet port. Different solvent modes may be selected for eluting different analytes of interest. For example, liquid chromatography may be performed using a gradient mode, an isocratic mode, or a polytyptic (i.e. mixed) mode. In some embodiments the column is heated. In particular embodiments, the column is heated to about 80 °C during analysis.

Embodiments of the invention are not limited by the type or number of LC steps used before mass spectrometry. Two or more of the LC techniques described herein may be combined before mass spectrometry (i.e., LC/LC/MS). Additional protein purification techniques may be employed prior to application of the antibody mixture to the LC column.

Some embodiments of the invention employ reversed phase HPLC (RP-HPLC), which has a non-polar stationary phase and an aqueous, moderately polar mobile phase. In some embodiments, the stationary phase is a silica which has been surface-modified with RMe ₂ SiCl, where R is a straight chain alkyl group such as Ci ₈ H ₃₇ or CsHn. With such stationary phases, retention time is longer for molecules which are less polar, while polar molecules elute more readily (early in the analysis). Retention times can be increased by adding more water to the mobile phase; thereby making the affinity of the hydrophobic analyte for the hydrophobic stationary phase stronger relative to the now more hydrophilic mobile phase. Similarly, retention time can be decreased by adding more organic solvent to the eluent. For example, retention can be decreased by adding a less polar solvent

(methanol, acetonitrile) into the mobile phase to reduce the surface tension of water.

Those of skill in the art will appreciate that the structure of the antibodies play an important role in their retention characteristics. In general, an analyte with a larger hydrophobic surface area (C-H, C-C, and generally non-polar atomic bonds, such as S-S and others) is retained longer because it is non-interacting with the water structure. On the other hand, analytes with higher polar surface area (conferred by the presence of polar groups, such as -OH, -NH2, COO- or -NH3+ in their structure) are less retained as they are better integrated into water. Such interactions are subject to steric effects in that very large molecules may have only restricted access to the pores of the stationary phase, where the interactions with surface ligands (alkyl chains) take place. Such surface hindrance typically results in less retention.

Retention time increases with hydrophobic (non-polar) surface area. Branched chain compounds elute more rapidly than their corresponding linear isomers because the overall surface area is decreased. Similarly organic compounds with single C-C-bonds elute later than those with a C=C or C-C-triple bond, as the double or triple bond is shorter than a single C-C-bond. Aside from mobile phase surface tension (organizational strength in eluent structure), other mobile phase modifiers can affect analyte retention. For example, the addition of inorganic salts causes a moderate linear increase in the surface tension of aqueous solutions (ca. 1.5 10-7 J/cm ² per Mol for NaCl, 2.5 10-7 J/cm ² per Mol for (NH ₄ ) ₂ S0 ₄ ), and because the entropy of the analyte-solvent interface is controlled by surface tension, the addition of salts tend to increase the retention time.

Another important factor is the mobile phase pH since it can change the hydrophobic character of the analyte. For this reason most methods use a buffering agent, such as sodium phosphate, to control the pH. Buffers serve multiple purposes: control of pH, neutralize the charge on the silica surface of the stationary phase and act as ion pairing agents to neutralize analyte charge. Ammonium formate may be used to improve detection of certain analytes by the formation of analyte-ammonium adducts. A volatile organic acid such as acetic acid, or most commonly formic acid (FA), may be added to the mobile. Trifluoroacetic acid (TFA) is used infrequently in mass spectrometry applications due to its persistence in the detector and solvent delivery system, but may be used in some embodiments as it a fairly strong organic acid and can be effective in improving retention of analytes such as carboxylic acids. Those of skill in the art will appreciate that the effects of acids and buffers vary by application but generally improve chromatographic resolution.

In some embodiments, the mobile phase comprises 0.1-0.5% formic acid (FA). In particular embodiments, the mobile phase comprises 0.1% FA. In some embodiments, the mobile phase comprises acetonitrile. In some embodiments, the mobile phase comprises acetonitrile and FA. In specific embodiments, the mobile phase comprises 0.1% FA and acetonitrile. In additional embodiments, the mobile phase further comprises 0.1-0.5% trifluoroacetic acid (TFA).

In some embodiments, the flow rate of the mobile phase is between 50-500 μυηήη. In particular embodiments, the flow rate is about 60 μΙ7ηιίη, about 100 μΙ7ηιίη, about 150 μΙ7ηιίη, about 200 μΙ7ηιίη, about 250 μΕ/ιηίη, about 300 μΙ7ηιίη, about 350 μΙ7ηιίη, about 400 μΕ/ιηίη, about 450 μΕ/ιηίη, or about 500 μΕ/ιηίη.

Size exclusion chromatography (SEC) may be applied to separate antibody mixtures under native conditions. It has been routinely used for the characterization and quality control of monoclonal antibody therapeutics in the pharmaceutical industry (Analysis of Reduced Monoclonal Antibodies Using Size Exclusion Chromatography Coupled with Mass Spectrometry, J. AM. SOC. MASS SPECTROM. 20, 2258-2264 (2009)). However, with normal SEC salt containing mobile phases, direct mass spectrometry analysis of antibody fragments is not feasible. Thus, volatile mobile phases should be optimized for the separation of antibodies. In particular embodiments, TFA concentrations are reduced to lower than 0.05%. In particular embodiments, the acetonitrile concentration 10-30%. In specific embodiments, the mobile phase comprises 0.02% TFA, 1% FA and 20% acetonitrile.

Particular embodiments of the invention employ ion-exchange chromatography, including anion-exchange and cation-exchange. Ion-exchange chromatography relies on the affinity of a substance for the exchanger, which affinity depends on both the electrical properties of the material and the relative affinity of other charged substances in the solvent. Bound material can be eluted by changing the pH, thus altering the charge of the material, or by adding competing materials, of which salts are but one example. The principle of ion- exchange chromatography is that charged molecules adsorb to ion exchangers reversibly so that molecules can be bound or eluted by changing the ionic environment. Separation on ion exchangers is usually accomplished in two stages: first, the substances to be separated are bound to the exchanger, using conditions that give stable and tight binding; then the column is eluted with buffers of different pH, ionic strength, or composition and the components of the buffer compete with the bound material for the binding sites.

An ion exchanger is usually a three-dimensional network or matrix that contains covalently-linked charge groups. If a group is negatively charged, it will exchange positive ions and is a cation exchanger. A typical group used in cation exchangers is the sulfonic group, SO ₃ — . If an H+ is bound to the group, the exchanger is said to be in the acid form; it can, for example, exchange on H+ for one Na+ or two H+ for one Ca ²⁺ . The sulfonic acid group is a strongly acidic cation exchanger. Other commonly used groups are phenolic hydroxyl and carboxyl, both weakly acidic cation exchangers. If the charged group is positive— for example, a quaternary amino group— it is a strongly basic anion exchanger. The most common weakly basic anion exchangers are aromatic or aliphatic amino groups. The matrix can be made of various materials. Commonly used materials are dextran, cellulose, agarose and copolymers of styrene and vinylbenzene in which the divinylbenzene both cross-links the polystyrene strands and contains the charged groups. Commercially available ion-exchangers for use in particular embodiments of the invention include SP- Sephadex, CM-Sephadex, QAE-Sephadex, DEAE-Sephadex, CM-cellulose, P-cel, DEAE- cellulose, PEI-cellulose, DEAE(BND)-cellulose, PAB-cellulose, AG 50, AG 1 -Source 15Q, AG 501, Bio-Rex 70, Toso Haas TSK-Gel-Q-5PW, Bio-Rex 40 and AG-3.

There are a number of choices to be made when employing ion exchange chromatography as a technique. The first choice to be made is whether the exchanger is to be anionic or cationic. If the materials to be bound to the column have a single charge (i.e., either plus or minus), the choice is clear. However, many substances (e.g., antibodies), carry both negative and positive charges and the net charge depends on the pH. In such cases, the primary factor is the stability of the substance at various pH values. Most proteins have a pH range of stability (i.e., in which they do not denature) in which they are either positively or negatively charged. Hence, if a protein is stable at pH values above the isoelectric point, an anion exchanger should be used; if stable at values below the isoelectric point, a cation exchanger is required.

The choice between strong and weak exchangers is also based on the effect of pH on charge and stability. For example, if a weakly ionized substance that requires very low or high pH for ionization is chromatographed, a strong ion exchanger is called for because it functions over the entire pH range. However, if the substance is labile, weak ion exchangers are preferable because strong exchangers are often capable of distorting a molecule so much that the molecule denatures. The pH at which the substance is stable must, of course, be matched to the narrow range of pH in which a particular weak exchanger is charged. Weak ion exchangers are also excellent for the separation of molecules with a high charge from those with a small charge, because the weakly charged ions usually fail to bind. Weak exchangers also show greater resolution of substances if charge differences are very small. If a macromolecule has a very strong charge, it may be impossible to elute from a strong exchanger and a weak exchanger again may be preferable. In general, weak exchangers are more useful than strong exchangers.

The Sephadex and Bio-gel exchangers offer a particular advantage for

macromolecules that are unstable in low ionic strength. Because the cross-linking in the support matrix of these materials maintains the insolubility of the matrix even if the matrix is highly polar, the density of ionizable groups can be made several times greater than is possible with cellulose ion exchangers. The increased charge density introduces an increased affinity so that adsorption can be carried out at higher ionic strengths. On the other hand, these exchangers retain some of their molecular sieving properties so that sometimes molecular weight differences annul the distribution caused by the charge differences; the molecular sieving effect may also enhance the separation.

The cellulose ion exchangers have proved to be the most effective for purifying large molecules such as proteins and polynucleotides. This is because the matrix is fibrous, and hence all functional groups are on the surface and available to even the largest molecules. In many cases, however, beaded forms such as DEAE-Sephacel and DEAE-Biogel P are more useful because there is a better flow rate and the molecular sieving effect aids in separation.

Buffers themselves consist of ions, and therefore, they can also exchange, and the pH equilibrium can be affected. To avoid these problems, the rule of buffers is adopted: use cationic buffers with anion exchangers and anionic buffers with cation exchangers. Because ionic strength is a factor in binding, a buffer should be chosen that has a high buffering capacity so that its ionic strength need not be too high. Furthermore, for best resolution, it has been generally found that the ionic conditions used to apply the sample to the column (starting conditions) should be near those used for eluting the column.

Some embodiments of the invention employ hydrophobic interaction chromatography (HIC), in which certain proteins are retained on affinity columns containing hydrophobic spacer arms. Hydrophobic adsorbents including octyl or phenyl groups may be used.

Hydrophobic interactions are strong at high solution ionic strength, as such samples being analyzed need not be desalted before application to the adsorbent. Elution is achieved by changing the pH or ionic strength or by modifying the dielectric constant of the eluant using, for instance, ethanediol. In some embodiments, cellulose derivatized with additional more hydroxyl groups is used to interact with proteins by hydrogen bonding. Samples may be applied to the matrix in a concentrated (over 50% saturated, >2M) solution of ammonium sulphate. Proteins may be eluted by diluting the ammonium sulphate. This introduces more water which competes with protein for the hydrogen bonding sites.

Affinity chromatography may be used in particular embodiments. Examples of commonly used affinity chromatography techniques include immobilized metal affinity chromatography (IMAC), sulfated affinity chromatography, dye affinity chromatography, and heparin affinity. In some embodiments, the chromatographic medium may be prepared using one member of a binding pair, e.g., a receptor/ligand binding pair, or antibody/antigen binding pair (immunoaffinity chromatography). In certain embodiments, affinity chromatography is used to selectively separate antibodies from a mixture by using protein A or G as a functionality of the stationary phase.

In particular embodiments, the LC separation is performed in-line with a mass spectrometry (MS) analysis. Specifically, the LC apparatus is directly connected to the mass spectrometer such that the mass spectrometer directly receives the separated product from the LC column.

Mass Spectrometry

Mass spectrometric techniques can fundamentally be divided into those starting from intact proteins (top-down) and those starting from peptides derived by chemical or, more commonly, enzymatic digestion (bottom up). Digestion, as used herein, refers to any enzymatically or chemically induced proteolysis (i.e., the breakdown of proteins into smaller polypeptides or amino acids by, in general, the hydrolysis of the peptide bonds) of individual antibodies in an antibody mixtures. Some aspects of the invention may utilize a bottom-up approach, in which proteins are enzymatically digested into smaller peptides using a protease such as trypsin. For example, in the standard bottom-up approach to mass spectrometry, proteins (e.g., antibodies) for analysis are enzymatically digested into smaller peptides using proteases such as trypsin. The digestion is carried out by first denaturing the proteins (with urea or guanidine HCl), reducing the disulfide bonds (with dithiothreitol or mercaptoethanol), alkylating the cytseines (through addition of iodoacetamide) and then digesting with a proteolytic enzyme (typically trypsin). In some instances, proteins are analyzed after a very limited proteolytic step prior to introduction into the MS instrument.

While the antibodies of the present invention can be analyzed using either bottom-up or top-down approaches, or combinations thereof, the methods disclosed herein do not require any pre-treatment (i.e., digestion or reduction) prior to mass spectrometric analysis. Stated another way, the methods of the present invention do not require the pre-treatment and/or digestion required for the bottom-up approach to mass spectrometry, regardless of how the distinct antibody populations in the antibody mixture are produced. By removing the steps of pre-treatment and digestion, the methods of the present invention increase efficiency and throughput, reduce cost, and are simpler and more accurate. Less sample handling is involved and the exact masses of all the antibodies of the mixture in their native state can be determined.

In particular embodiments, no enzymatic digestion of the antibody mixture is required, and intact antibodies are transferred directly to the gas phase in the ionic form and are fragmented. Preferred embodiments of the invention, however, utilize the top-down approach in which intact proteins are ionized, and optionally fragmented, and then introduced to a mass analyzer. Subsequently these peptides are identified by peptide mass fingerprinting or tandem mass spectrometry. As described in greater detail below, mass spectrometers useful in the present invention can include a triple quadrupole, an Orbitrap type MS, an ESI- TOF, a Q-TOF, a TOF, an ion trap or any other instrument of suitable sensitivity and mass resolution.

Mass spectrometry (MS) is an analytical technique used to produce spectra of the masses of the atoms or molecules within a sample. The elemental signature of a sample, the masses of particles (and of molecules), and the chemical structures of molecules (such as peptides and other chemical compounds) can all be determined using the resultant MS spectra. Mass spectrometry ionizes chemical compounds to then generate charged molecules or molecule fragments allowing for the measurement of the mass-to-charge (m/z) ratios. In a typical MS procedure, a sample, (in a solid, liquid, or gaseous form), is ionized, for example by bombarding it with electrons. This causes some of the sample's molecules to break into charged fragments where the ions are then separated according to their mass-to-charge ratio, typically by accelerating them and subjecting them to an electric or magnetic field. Ions of the same mass-to-charge ratio will undergo the same amount of deflection. The ions are detected by a mechanism capable of detecting charged particles, typically an electron multiplier. Results are displayed as spectra of the relative abundance of detected ions as a function of the mass-to-charge ratio. The atoms or molecules in the sample can be identified by correlating known masses to the identified masses or through a characteristic

fragmentation pattern.

A mass spectrometer typically includes an ion source, a mass analyzer that measures m/z of the ionized analytes, and a detector that registers the number of ions at each m/z value. The differences in masses of the fragments allows the mass analyzer to sort the ions by their mass-to-charge (m/z) ratio. The detector measures the value of an indicator quantity and thus provides data for calculating the abundances of each ion present. Because mass analysis uses electromagnetic fields in a vacuum, typically molecules are first electrically charged and transferred into the gas phase. There is a wide variety of ionization techniques, depending on the phase (solid, liquid, gas) of the sample and the efficiency of various ionization mechanisms for the unknown species.

The ion source is the part of the mass spectrometer that ionizes the material under analysis. The ions are then transported to the mass analyzer by electric or magnetic fields. Ionization techniques are key to determining what types of samples can be analyzed by mass spectrometry. Electron ionization and chemical ionization are used for gases and vapors. In chemical ionization sources, the analyte is ionized by chemical ion-molecule reactions during collisions in the source. Two techniques often used with liquid and solid biological samples include electrospray ionization (ESI) and matrix-assisted laser desorption/ionization

(MALDI). ESI ionizes the analytes out of a solution and is therefore readily coupled to liquid-based (such as electrophoretic and chromatographic) separation tools. MALDI sublimates and ionizes the sample out of a dry, crystalline matrix with laser pulses. MALDI- MS is normally used to analyze relatively simple peptide mixtures, whereas integrated liquid chromatographic ESI-MS systems (LC-MS) are preferred for the analysis of complex samples. Integrated liquid-chromatography ESI-MS systems (LC-MS) are particularly useful for the analysis of complex samples, such as antibody mixtures.

In embodiments of the present invention, ions can be produced using a variety of methods including, but not limited to, electron ionization, chemical ionization, fast atom bombardment, field desorption, and matrix-assisted laser desorption ionization ("MALDI"), surface enhanced laser desorption ionization ("SELDI"), photon ionization, electrospray ionization (ESI), and inductively coupled plasma.

ESI mass spectrometers are described in further detail in e.g., U.S. Pat. No. 6,673,253 (describing a method of fabricating integrated LC/ESI device and description of ESI mass spectrometry); U.S. Pat. No. 6,642,515 (describing a method and an apparatus for performing electrospray ionization mass spectrometric analysis); U.S. Pat. No. 6,627,883 (describing methods for analyzing samples in a dual ion trap mass spectrometer but provides background description of ESI-MS); U.S. Pat. No. 6,621,075 (describing a device for the delivery of multiple liquid sample streams to a mass spectrometer and provides further guidance of the knowledge of those of skill in the art regarding ESI-MS); and U.S. Pat. No. 6, 188,065

(describing a combination of capillary electrophoresis combined with a mass spectrometer). Each of the foregoing patents are incorporated herein by reference in their entirety as showing the general mechanics of a mass spectrometer and methods of introducing samples for mass spectrometry analysis.

In particular embodiments of the invention, ESI is used as the ionization technique.

ESI allows the antibody ions to be extracted directly from solution, whose composition may be reasonably close to the "native" environment (e.g., pharmaceutically acceptable preparations) in terms of pH and ionic strength. ESI induces ion formation from small droplets, once the ions are formed they are subject to collisions before entering the mass analyzer. These collisions can decluster aggregates, induce fragmentation, and change the charge states by removing protons. The energy with which the ions enter the mass analyzer through the orifice can determine the amount of fragmentation that will take place. This energy can be adjusted by varying the electrospray declustering or fragmentation potential. ESI-generated biopolymer ions are typically produced as multiply charged ions, and the extent of multiple charging is determined by the physical dimensions of the biopolymer molecule in solution (Kaltashov IA, and Abzalimov RR., "Do ionic charges in ESI MS provide useful information on macromolecular structure? '' J AM SOC MASS SPECTROM., 2008, 19: 1239^16). This feature of ESI MS allows it to be used as a means to probe protein higher order structure and detecting large-scale conformational transitions in solution.

Natively folded proteins undergo ESI to produce ions carrying a relatively small number of charges, because the compact shape of a tightly folded polypeptide chain in solution does not allow the accommodation of a significant number of protons on its surface upon transition from solution to the gas phase. As a result, ion peaks in ESI mass spectra of proteins in aqueous solutions at neutral pH typically dominate the high m/z regions of the mass spectra and are almost always characterized by having narrow distribution of charge states

(Kaltashov IA, and Abzalimov RR., J AM SOC MASS SPECTROM., 2008, 19: 1239^16).

Another important feature of ESI is its ability to generate multiply charged species. Multiple charging makes it possible to observe large proteins with mass analyzers that have a relatively small mass range. In addition, observing multiple peaks for the same peptide allows one to make multiple molecular weight calculations from a single spectrum. These values can be averaged to obtain a more accurate molecular weight. Another advantage of generating multiply charged ions with ESI is that multiply charged peptide ions tend to give more complete fragmentation spectra, which is extremely useful in distinguishing between similar species (i.e., various monoclonal antibodies) in a complex mixture (i.e., in a mixture of antibodies). In some embodiments, nanoelectrospray ionization is used for further increases in sensitivity. Adaptable techniques are described in, for example, Wilm, W., et al., "Femtomole sequencing of proteins from poly aery lamide gels by nano-electrospray mass spectrometry " NATURE, 379: 446-469 (1996).

Another important feature of ESI-MS is its ability to directly analyze compounds from aqueous or aqueous/organic solutions. In some embodiments, ESI is used with quadrupole or ion trap analyzers, which allows for MS analysis at relatively high LC flow rates (e.g., 0.5 ml/min) and high mass accuracy (±0.01%). In particular embodiments ESI is used as part of LC/MS or LC tandem mass spectrometry (LC/MS/MS) techniques. Non-volatile electrolytes (which are present in all biopharmaceutical formulations) have a detrimental effect on the quality of ESI MS data even at relatively low concentrations and must be removed by either dialysis or ultrafiltration before the direct ESI MS

measurements are carried out. Sample preparation is generally achieved by dissolving the sample in a protic volatile solvent system that is relatively homogeneous and contains less than one millimolar concentration of salt, although higher concentrations may be used (as high as 100 mM with NH ₄ Ac). However, some salts (alkali and alkaline) and phosphate buffers are more detrimental to signal. Thus, in some embodiments, the antibody mixture is transferred to a volatile electrolyte solution (e.g., ammonium acetate or ammonium bicarbonate).

In some embodiments of the invention, matrix-assisted laser desorption ionization (MALDI) is used as an ionization source. As used herein, MALDI refers to methods in which a non-volatile sample is exposed to laser irradiation, which desorbs and ionizes analytes in the sample by various ionization pathways, including photo-ionization, protonation, deprotonation, and cluster decay. For MALDI, the sample is mixed with an energy-absorbing matrix, which facilitates desorption of analyte molecules. Sample preparation and protocols for analysis of proteins using MALDI are known in the art, as described, for example, in Trauger, S.A., et al, "Peptide and protein analysis with mass spectrometry " SPECTROSCOPY, 2002, 16: 15-28, incorporated by reference herein.

Inductively coupled plasma (ICP) may be used as an ionization technique in particular embodiments. A plasma flame that is electrically neutral overall, but has a large fraction of its atoms ionized by high temperature, is used to atomize introduced sample molecules and to further strip the outer electrons from those atoms. The plasma is usually generated from argon gas, since the first ionization energy of argon atoms is higher than the first of any other elements except He, O, F and Ne, but lower than the second ionization energy of all except the most electropositive metals. The heating is achieved by a radio-frequency current passed through a coil surrounding the plasma.

Other ion source techniques include glow discharge, field desorption (FD), fast atom bombardment (FAB), thermospray, desorption/ionization on silicon (DIOS), Direct Analysis in Real Time (DART), atmospheric pressure chemical ionization (APCI), secondary ion mass spectrometry (SIMS), spark ionization and thermal ionization (TIMS). Ion attachment ionization is an ionization technique that allows for fragmentation free analysis.

Some embodiments of invention employ fragmentation analysis for top-down amino acid sequence analysis of individual antibodies or individual light or heavy chains present in complex antibody mixtures. In such embodiments, intact antibodies are broken into fragments in the gas phase. In some embodiments, collision-induced dissociation (CID) is used, in which analytes are collided in the gas phase with gas atoms or molecules (typically argon or nitrogen). In some embodiments, CID is used in combination with TOF or Orbitrap detection technology. In some embodiments of the invention, high energy CID is used with a unique spectrum multiplexing feature (msx HCD). The multiplexing features refers to a data acquisition mode in which fragment ions produced from several individual HCD events, each on a precursor of a different charge state of an antibody, are detected together in an Orbitrap mass analyzer.

In some embodiments, electronic activity is used to fragment antibodies in the mixture. Techniques that may be used included electron-capture dissociation (ECD) on electrospray FT-ICR-MS; electron-transfer dissociation (ETD) on electrospray ion traps; and in-source decay fragmentation (ISD) on MALDI-TOF instruments. These techniques introduce an electron (ECD), transfer an electron from a reagent radical anion (ETD) or introduce a hydrogen radical (ISD) into the analyte molecule, which creates an electronic disturbance. These changes subsequently induce the cleavage of the N-Ca-bond in the peptide/protein backbone, which is also common to all three techniques, resulting in the formation of c-type and z-type fragments complementary to the b-type and y-type fragments usually observed in CID fragmentation. Differences between the techniques are that ECD and ETD necessarily require multiply-charged analytes for the successful analysis as a charge transfer that takes place would reduce the charge for singly charged analytes to zero, thus making them undetectable by MS. In ISD, the MALDI-generated singly-charged ions are sufficient for the fragmentation as there is no charge transfer, but a radical transfer takes place.

In particular embodiments, the ECD and ETD techniques are used in combination with ESI MS. In some embodiments, ETD is used to fragment antibodies with labile modifications (e.g., glycosylation), while preserving the modification intact on the on the antibody fragment. In some embodiments, ETD induces fragmentation all over the protein backbone, which allows top-down fragmentation of intact antibodies in the mixture.

ISD on MALDI allows fast straight- forward top-down sequence analysis of undigested proteins based on fragmentation of the entire protein chain caused by hydrogen radical transfer from the MALDI-ISD matrix. The technique provides high mass accuracy and enables fast sequencing of terminal domains of monoclonal antibodies in a targeted way (Ayoub et al, mAbs 20135:5, 699-710). The mass analyzer is typically the central component of a mass spectrometer.

Exemplary mass analyzers include ion trap, time-of- flight (TOF), quadrupole and Fourier transform ion cyclotron (FT-MS) analyzers. In some embodiments, these analyzers can be used alone or in tandem to exploit their strengths and minimize any weaknesses. In some embodiments, TOF-MS, or FT-MS is used to analyze an antibody mixture. These two types of instrument typically have wide mass range, and high mass accuracy.

There are several important analyzer characteristics. The mass resolving power is the measure of the ability to distinguish two peaks of slightly different m/z. The mass accuracy is the ratio of the m/z measurement error to the true m/z. Mass accuracy is usually measured in ppm or milli mass units. The mass range is the range of m/z amenable to analysis by a given analyzer. The linear dynamic range is the range over which ion signal is linear with analyte concentration. Speed refers to the time frame of the experiment and ultimately is used to determine the number of spectra per unit time that can be generated.

Particular embodiments of the invention employ a time-of-flight (TOF) analyzer, which uses an electric field to accelerate the ions through the same potential, and then measures the time they take to reach the detector. If the particles all have the same charge, the kinetic energies will be identical, and their velocities will depend only on their masses. Lighter ions will reach the detector first.

Particular embodiments of the invention employ a quadrupole mass analyzer. A typical quadrupole mass analyzer includes four circular rods set parallel to each other, and is responsible for filtering sample ions based on their m/z. A linear series of three quadrupoles can be used to make a triple quadrupole mass spectrometer. In a triple quadrupole mass spectrometer, the first and third quadrupoles typically act as mass filters, and the middle quadrupole typically acts as a collision cell. A quadrupole ion trap can thus exist in linear and three-dimensional formats and refers to an ion trap that uses constant direct current (DC) and radio frequency (RF) oscillating alternate current (AC) electric fields to trap ions.

In some embodiments, a combination of a quadrupole mass filter with a TOF analyzer may be used. Quadrupole TOF instruments achieve peptide separation "in space", i.e., the ions are separated nearly instantaneously by passing through either the quadrupole section, in which only a chosen small mass range has stable trajectories, or by traversing the TOF section.

In some embodiment, an Orbitrap mass analyzer is used. The Orbitrap mass analyzer typically includes a small electrostatic device into which ion peaks are injected at high energies to orbit around a central, spindle-shaped electrode. Image current of the axial motion of the ions is picked up by the detector and its signal is Fourier Transformed (FT) to yield high resolution mass spectra. Thus, a hybrid quadrupole Orbitrap instrument is able to select ions virtually instantaneously due to the fast switching times of quadrupoles and it is able to fragment peptides in High Energy Collision (HCD) mode on a similarly fast time scale. In some embodiments, a combination of a quadrupole mass filter with an Orbitrap analyzer can be used.

Another key part of the mass spectrometer is the detector, which records either the charge induced or the current produced when an ion passes by or hits a surface. In a scanning instrument, the signal produced in the detector during the course of the scan versus where the instrument is in the scan (at what m/Q) will produce a mass spectrum, a record of ions as a function of m/Q. Some type of electron multiplier is used, though other detectors including Faraday cups and ion-to-photon detectors are also used. Because the number of ions leaving the mass analyzer at a particular instant is typically quite small, considerable amplification is often necessary to get a signal. MicroChannel plate detectors are commonly used in modern commercial instruments. In FTMS and Orbitraps, the detector consists of a pair of metal surfaces within the mass analyzer/ion trap region which the ions only pass near as they oscillate. No direct current is produced, only a weak AC image current is produced in a circuit between the electrodes.

Particular embodiments of the invention utilize Fourier transformation (FT) to turn signals picked up by the detection into high resolution mass spectra. Fourier transform mass spectrometry (FTMS), or more precisely Fourier transform ion cyclotron resonance MS, measures mass by detecting the image current produced by ions cyclotroning in the presence of a magnetic field. Instead of measuring the deflection of ions with a detector such as an electron multiplier, the ions are injected into a Penning trap (a static electric/magnetic ion trap) where they effectively form part of a circuit. Detectors at fixed positions in space measure the electrical signal of ions which pass near them over time, producing a periodic signal. Since the frequency of an ion's cycling is determined by its mass to charge ratio, this can be deconvoluted by performing a Fourier transform on the signal. FTMS has the advantage of high sensitivity (since each ion is "counted" more than once) and much higher resolution and thus precision. Ion cyclotron resonance (ICR) is an older mass analysis technique similar to FTMS except that ions are detected with a traditional detector. Ions trapped in a Penning trap are excited by an RF electric field until they impact the wall of the trap, where the detector is located. Ions of different mass are resolved according to impact time. Embodiments of the present invention include analysis of antibody mixtures without pre-treatment or enzymatic digestion of the sample using, for example, Orbitrap MS instruments. With high sensitivity, mass accuracy and resolving power, Orbitrap MS can provide baseline separation of intact half (approximately 75 kDa) and full (approximately 150 kDa) monoclonal antibodies. By analyzing untreated antibodies with the Orbitrap MS in conjunction with deconvolution software (as discussed below), the antibody signal is condensed into a few ion signals in the mass spectrum, allowing for a simpler and more straightforward interpretation of the results and possibly reducing overlap between adjacent monoclonal antibody species. The high resolving power of the Orbitrap MS allows differentiation between antibodies that are very close in mass. In some embodiments, the high resolving power of the Orbitrap MS can be attributed to more efficient desolvation and less adduct formation.

In particular embodiments, a Q Exactive Orbitrap instrument is used, which couples a quadrupole mass filter to an Orbitrap analyzer. The Q Exactive instrument features high ion currents because of an S-lens, and fast high-energy collision- induced dissociation peptide fragmentation because of parallel filling and detection modes. The image current from the detector is processed by an "enhanced Fourier Transformation" algorithm, doubling mass spectrometric resolution. Together with almost instantaneous isolation and fragmentation, the instrument achieves overall cycle times of 1 s for a higher energy collisional dissociation method. Various features, settings and protocols for this instrument are described in

Michalski, A. et al, "Mass Spectrometry-based Proteomics Using Q Exactive, a High- performance Benchtop Quadrupole Orbitrap Mass Spectrometer," MoL CELL PROTEOMICS, 10(9), Sept. 201 1, incorporated by reference herein. In some embodiments, resolution is set at between 17,500 to 140,000 for full MS and 140,000 for top-down tandem MS.

In some embodiments, the methods disclosed herein can characterize and differentiate between at least two different antibodies in an antibody mixture when the antibodies differ by no more than 0.01% of their total mass (e.g., resolving two monoclonal antibodies differing by 15 Da on a total mass of approximately 150 kDa). In some embodiments, the methods disclosed herein can characterize (i.e., identify and quantify) at least two antibodies in an antibody mixture when the antibodies differ by no more than 0.02%, 0.03%, 0.04%, 0.05%, 0.06%, 0.07%, 0.08%, 0.09%, or 0.10% of their total mass. In some embodiments, the methods disclosed herein can characterize (i.e., identify and quantify) at least two, at least three, at least four, at least five, at least six, at least seven, at least eight, at least nine, at least ten, at least eleven, at least twelve, at least thirteen, at least fourteen, or at least fifteen or more individual antibodies in an antibody mixture when the antibodies differ by no more than 0.01%, 0.02%, 0.03%, 0.04%, 0.05%, 0.06%, 0.07%, 0.08%, 0.09%, or 0.10% of their total mass. In some embodiments, the smallest difference between any two antibodies in the mixture is no more than about 10 Da, no more than about 15 Da, no more than about 20 Da, no more than about 25 Da, no more than about 30 Da, no more than about 35 Da, no more than about 40 Da, no more than about 45 Da, or no more than about 50 Da. In some embodiments, the smallest difference between any two antibodies in the antibody mixture is less than about 1.1 kilodalton, less than about 1.0 kilodalton, less than about 900 Da, less than about 800 Da, less than about 700 Da, less than about 600 Da, or less than about 500 Da.

High mass accuracy is critical for unambiguous identification of antibody species in an antibody mixture. In particular embodiments, the mass accuracy for a given antibody in an antibody mixture is less than about 10 ppm, less than about 9 ppm, less than about 8 ppm, less than about 7 ppm, or less than about 5 ppm.

Low mass error (i.e., precision) is also critical for unambiguous identification of antibody species in an antibody mixture. In particular embodiments, the ppm mass error (e.g., determined by analyze antibodies and antibody mixtures several times on different instrument over different days) is less than about 5, less than about 5.5, less than about 6.0, less than about 6.5, less than about 7.0, less than about 7.5, less than about 8.0, less than about 8.5, less than about 9.0, less than about 9.5, or less than about 10.0 ppm.

Embodiments of the present invention also require consistency. In some

embodiments, if analyses are repeated on two separate but otherwise identical MS instruments, the mass difference is less than 10 ppm between instruments; e.g., less than 9 ppm, less than 8 ppm, less than 7 ppm, less than 6 ppm, less than 5 ppm, less than 4 ppm, or less 3 ppm, or less than 2 ppm per instrument.

The analysis of undigested, intact antibodies using Orbitrap MS provides a single highly resolved profile of all protein micro-heterogeneity within a few minutes using a few femto-mole of sample. Therefore, the methods of the present invention make

characterization of antibody mixtures a time- and cost- efficient process. In some embodiments, more than 30 different proteoforms of a single monoclonal antibody can detected using an Orbitrap MS on an undigested sample. Furthermore, the improved performance by an Orbitrap MS leads to increases in the number of identified species, including ones that occur in low abundance. Ionization efficiency is not substantially affected by differences in glycan chains because the glycan side chains remain neutral while basic residues of the protein become protonated. In some embodiments, ESI-tandem MS (ESI-MS/MS) is applied to provide amino acid sequence information of individual antibodies. In some embodiments, the individual antibodies are found within complex mixtures of antibodies. In some embodiments, the antibodies are reduced. In some embodiments, top-down MS/MS is applied using msx HCD. In some embodiments, an Orbitrap MS system is used. The combination of msx HCD and Orbitrap MS can provide improved throughput from spectrum multiplexing, as well as advanced signal processing to provide improved resolution and higher Orbitrap scan speeds. In some embodiments, amino acid sequence information of antibody light chains is provided. In some such embodiments, sequence coverage of over 40% is achieved, including the N- terminal variable region. In some such embodiments, sequence coverage of over 50%, over 60%, over 70%, over 80%, over 90% and up to 100% is achieved, including the N-terminal variable region. Is some embodiments, the mass error is less than 5 ppm (e.g., 5 ppm, 4 ppm, 3 ppm, etc.) for fragment ions.

In some embodiments, the methods of the present invention combine mass spectrometry with spectroscopic methods to obtain secondary and tertiary structural information about antibodies. Circular dichroism and fluorescence spectroscopy can be applied to the analysis of protein conformations in solution. Circular dichroism allows for the rapid evaluation of the secondary structure of proteins. Left and right circularly polarized light in the far-UV region is absorbed by the periodic repeated binding angles of the secondary structure elements to differing extents. Alpha-helical secondary structures have a stronger absorption of light compared to beta strands.

Particular embodiments of the invention employ tandem mass spectrometry (ms/ms). As used herein, tandem mass spectrometry refers to two stages of mass analysis that are employed to examine selectively the fragmentation of particular species (e.g., antibodies) in a mixture. See, e.g., de Hoffmann, E., "Tandem Mass Spectrometry: a Primer," J. MASS

SPECTRO., 31 : 129-137 (1996). Various techniques may be employed for the targeted and untargeted analysis of complex mixtures using tandem mass spectrometry (MS). Particular techniques that may be employed or adapted for use in the present invention are described in, for example, International PCT Publications WO 2013/0931 14 and WO 2013/134165, and U.S. Patent No. 7, 158,903; these publications are incorporated herein by reference.

In some embodiments of the invention, individual antibodies in an antibody may be characterized (i.e., identified and quantified) based on MS spectrum generated solely from the light chain (i.e., a light chain spectrum). In other embodiments, individual antibodies in an antibody may characterized (i.e., identified and quantified) based on MS spectrum generated solely from the heavy chain (i.e., a heavy chain spectrum). In yet other embodiments, individual antibodies in an antibody may be characterized based on MS spectrum generated from both the heavy and light chains.

An additional aspect of the present invention are methods to quantitate individual antibodies in a complex antibody mixture. In particular embodiments, this is done by calculating base chromatographs at different concentrations of a given antibody. In some embodiments, at least five, at least six, at least seven, at least eight or more concentrations are used to calculate the base peak chromatographs. In certain embodiments, the base peak chromatograph at each concentration is calculated in triplicate (i.e., 3 different injections). In certain embodiments, the concentrations used for generating the base peak chromatographs ranges between about 10.0 - 500 ng (e.g., base chromatographs are constructed at 12.5 ng, 25 ng, 50 ng, 100 ng, 125 ng, 200 ng, 250 ng and 500 ng). The base peak chromatographs are then used to construct a calibration curve based on peak area. The amount of a given antibody in a test sample can be extrapolated from the calibration curve by comparing the area of the test antibody's characteristic chromatographic peak.

Additional embodiments of the invention comprise top-down characterization of intact antibody light chains. As the amino acid sequence of the light chains in the antibody mixture is known, comparison of the theoretical and experimental masses provides a direct indication of the homogeneity and the molecular integrity of the antibody light chains. In some embodiments of the invention, intact antibody light chain mass determination is done by MALDI-TOF, ESI-TOF, or ESI in conjunction with FT-based detection (either ICR or Orbitrap). By using average masses, mass accuracy down to a few ppm can be achieved, corresponding to absolute errors of 0-2 Da for intact proteins. In particular embodiments using ESI-TOF or ESI-Orbitrap, resolution of 60,000 to 240,000 can be achieved. In some embodiments, the light chain mass measurement accuracy is less than 1.0 ppm, less than 0.9 ppm, less than 0.8 ppm, less than 0.7 ppm, less than 0.6 ppm, less than 0.5 ppm, less than 0.4 ppm or less than 0.3 ppm.

Top-down characterization of antibodies as described herein may also be used to identify post-translational modifications. For example, both ESI and MALDI may be applied to glycosylated protein analysis. Glycosylation is one of the most common forms of enzymatic post-translational modification. In some embodiments, the analysis begins with a glycoform profiling, where ESI MS or MALDI MS is used to obtain mass distribution of the glycoprotein. If the carbohydrate weight fraction is modest (<5% of the total protein weight), mass distributions can be clearly resolved and provide a useful signature of the glycan size distribution In the case of extensive glycosylation it may become difficult to obtain reliable estimates of the protein mass based on ESI MS data alone; a combination of ESI MS and gas phase chemistry may be needed in order to reduce the complexity of the protein ion ensemble as a pre-requisite for reliable mass measurements.

Glycan release from glycoproteins (i.e., deglycoslyation) can be carried out using peptide N-glycosidase F (PNGase F) and/or PNGase A for N-glycans and chemical methods (e.g., hydrozinolysis) for O-glycans. Methods of tandem MS/MS, especially electron-based ion fragmentation methods (such as electron capture dissociation, ECD, and electron transfer dissociation, ETD), can be used in combination with exoglycosidases to obtain detailed structural information on released glycans.

In particular embodiments of the invention, the high resolving power and mass accuracy of methods described herein allow glycosylation states of individual antibodies in an antibody mixture to be identified without deglycosylation. While a deglycosylated spectrum yields nicely resolved peaks, embodiments of the present invention permit differentiation between the antibody species with the glycans still attached.

Data Processing

Full MS spectra of intact or reduced antibodies may be analyzed using protein deconvolution software, which assists in characterization of antibodies from mass spectrometric data. This software produces accurate results, even for low-abundance proteins, enabling detection of small protein modifications with mass shifts of just a few Da, such as when a few disulfide bonds are reduced. In particular, protein deconvolution software simplifies or condenses the mass spectrum, making interpretation more

straightforward and reducing overlap between adjacent or similar antibody species.

A typical protein deconvolution software employs two different deconvolution algorithms, each optimized for a different type of data, to ensure the highest quality results. Two algorithms are Xtract and ReSpect, which are optimized for isotopically resolved and unresolved data, respectively. For example, using a ReSpect algorithm, the protein deconvolution software deconvolutes isotopically unresolved data, across a wide mass range of up to 160,000 Da, compatible with fast chromatography.

Mass spectra for deconvolution are produced by averaging spectra across the most abundant portion of the elution profile for the antibody. The average spectra are

subsequently deconvoluted using an input m/z range of 2000-4000 m/z, an output mass range of 140,000 to 160,000 Da, a target mass of 150,000 Da, and a minimum of at least 5 or more consecutive charge states from the input m/z spectrum to produce a deconvoluted peak. In some embodiments, at least 6, at least 7, at least 8, at least 9 or at least 10 consecutive charge states from the input m/z spectrum are used to produce a deconvoluted peak.

In some embodiments, data processing may be used to identify glycoforms by comparing the masses of antibodies to their expected masses inclusive of the various combinations of commonly found glycoforms.

In some embodiments, top-down msc HCD spectra are analyzed using appropriate software (e.g., ProSightPC software) under a single protein mode with a fragment ion tolerance of 5 ppm.

Examples

The present invention will be better understood in connection with the following Examples. However, it should be understood that these examples are for illustrative purposes only and are not meant to limit the scope of the invention. Various changes and

modifications to the disclosed embodiments will be apparent to those skilled in the art and such changes and modifications including, without limitation, those relating to the methods and/or formulations of the invention may be made without departing from the spirit of the invention and the scope of the appended claims.

Example 1. Mass spectrometry (MS)-based analysis of a mixture of whole or intact antibodies

Experiments in this Example illustrate Orbitrap mass spectrometer-based analysis of a mixture of monoclonal antibodies (mAbs) at intact protein level. As shown below, a workflow solution that combines high resolution MS, fast chromatography, high throughput msx HCD, and accurate data analysis can be used to effectively and accurately characterize intact individual monoclonal antibodies present in an antibody mixture sample.

Specifically, samples that were analyzed included five individual mAbs, a mixture of three of these antibodies, and a mixture of all five of these antibodies. These samples were desalted using a PLRP-S column (1.0 * 50 mm; 8 μιη particle size; 4000 A pore size) and a five minute gradient. Whole or intact antibodies were eluted from the column and analyzed using a bench-top Q Exactive Quadrupole Orbitrap mass spectrometer. Furthermore, top-down tandem mass spectrometry (MS/MS) was performed on light chains of the antibodies using high energy collision dissociation with a unique spectrum multiplexing feature (msx HCD) in the Q Exactive Quadrupole Orbitrap mass spectrometer.

Protein Deconvulation 1.0 software that utilizes ReSpect algorithm for molecular mass determination was used to analyze full MS spectra of intact antibodies. The top-down msx HCD spectra were analyzed using ProSight PC 2.0 software.

The full MS spectra data for the individual mAbs as well as the antibody mixtures, shows a complete charge envelope distribution of each mAb. Each charge state in this spectra revealed baseline separated major glycosylated forms (glycoforms) of the mAbs, and the deconvulated spectra showed the expected molecular mass of all the major glycoforms for each mAb. The MS spectra for antibody mixture having three antibodies showed complete separation of each individual antibody in the mixture (Figure 1). This allowed the confirmation that the deconvoluted molecular mass for each antibody in the mixture was in agreement with what was measured for each of these antibodies individually (Figure 2). The MS spectra for the antibody mixture having five antibodies (Figure 4) suggested that one minor glycoform of a mAb overlapped with the major glycoform of the next mAb, which resulted in a mass shift of 3 Da on the intact mass after deconvolution of the spectrum (Figure 5).

The isotopic peaks of the light chain were baseline resolved by using the 140 K resolution setting, which resulted in monoisotopic molecular mass determination with an error of less than 5 ppm (Figures 3 and 6).

Furthermore, the amino acid sequence of the antibodies was obtained by applying top- down MS/MS analysis to the light chains of the antibodies using the msx HCD approach. For acquisition of data in this mode, fragment ions produced from several individual HCD events, each on a precursor of a different charge state of the reduced mAb were detected together in the Orbitrap mass analyzer. The results suggested that high resolution and information rich spectra were generated, and excellent sequence coverage, including the N- terminal variable region, was achieved. Thus, the results in this Example showed that precise mass measurement, and extensive and high confidence sequence information can be obtained for intact mAbs either individually or in a mixture. Example 2. Mass spectrometry (MS)-based quantitation of whole or intact antibodies

Experiments in this Example illustrate Orbitrap mass spectrometer-based quantitation of intact monoclonal antibodies (mAbs). An Orbitrap mass spectrometer was used to generate base peak chromatograms for the monoclonal antibody 5.55.D2 at concentrations from 12.5 ng to 500 ng. Monoclonal antibody 5.55.D2 was quantitated by calculating the base chromatograms at concentrations from 12.5 ng to 500 ng (Figure 7A). The base peak chromatogram at each concentration was calculated in triplicate by performing three different injections. Exemplary base peak chromatograms at three injections are depicted for concentrations at 12.5 ng (Figure 7B) and 500 ng (Figure 7D).

Subsequent mass spectra were generated for each injection for all tested

concentrations; mass spectra are shown for 12.5 ng (Figure 7C) and 500 ng (Figure 7E). Calibration curves based upon peak area were calculated (Figure 8B) using processing software from the ThermoFisher Q Exactive Orbitrap mass spectrometer (Figure 8A).

Summarized results confirm highly accurate mass quantitation of intact antibodies using Orbitrap mass spectrometers (Figure 8C). Quantitation results for the monoclonal antibody 5.55.D2 demonstrate high resolution mass analysis throughout the tested concentration range of 12.5 ng to 500 ng.

Example 3. Mass spectrometry (MS)-based characterization of whole or intact antibody light chains using a top-down approach

Experiments in this Example illustrate Orbitrap mass spectrometer-based

characterization of whole or intact antibody light chains using a top-down approach (i.e. without any pre-digestion of the light chains). The intact, undigested light chains from four monoclonal antibodies (26.3.E2; 5.6.H9; 42.18.E12; and 5.55.D2) were characterized in a top-down MS approach to determine if post-translational modifications could be detected on intact antibodies. Figures 9A-9D depicts the raw mass spectra for the isolated resolved light chains of antibodies 26.3.E2; 5.6.H9; 42.18.E12; and 5.55.D2 respectively. The light chain from each antibody was analyzed in an Orbitrap mass spectrometer in High Energy Collision mode (msx) using tandem mass spectrometry. The mass spectrum for the light chain of monoclonal antibody 26.3.E2 is depicted in Figure 10A. Based upon comparison between the theoretical and observed masses, evidence for a series of disulfide bonds were noted in the final sequence of the light chain from 26.3.E2 (Figure 10B). There was also evidence of cysteinylation of C-terminus cysteine. The predicted monoisotopic mass with modifications (2 disulfides, cysteinylation) is 23647.6947; while the experimental monoisotopic mass is 23647.7032. Therefore, the light chain mass measurement accuracy for the monoclonal antibody 26.3. E2 is 0.3 ppm.

The mass spectrum for the light chain of monoclonal antibody 5.6.H9 is depicted in Figure 11 A. Based upon comparison between the theoretical and observed masses, evidence for a series of disulfide bonds were noted in the final sequence of the light chain from 5.6.H9 (Figure 11B). Furthermore, there was also evidence of cysteinylation of C-terminus cysteine. The predicted monoisotopic mass with modifications (2 disulfides, cysteinylation) is 23706.5672; while the observed monoisotopic mass is 23706.5512. Therefore, the light chain mass measurement accuracy for the monoclonal antibody 5.6.H9 is 0.7 ppm.

The mass spectrum for the light chain of monoclonal antibody 42.18.E12 is depicted in Figure 12A. Based upon comparison between the theoretical and observed masses, evidence for a series of disulfide bonds were noted in the final sequence of the light chain from 42.18. E12 (Figure 12B). Furthermore, there was also evidence of cysteinylation of C- terminus cysteine. The predicted monoisotopic mass with modifications (2 disulfides, cysteinylation) is 23408.4031; while the observed monoisotopic mass is 23408.4253.

Therefore, the light chain mass measurement accuracy for the monoclonal antibody

42.18.E12 is 0.9 ppm.

The mass spectrum for the light chain of monoclonal antibody 5.55.D2 is depicted in Figure 13 A. Based upon comparison between the theoretical and observed masses, evidence for a series of disulfide bonds were noted in the final sequence of the light chain from

5.55.D2 (Figure 13B). In addition to evidence of cysteinylation of C-terminus cysteine; there was also support for a post-translational pyro-glu modification of the N-terminus glutamine. The predicted monoisotopic mass with modifications (2 disulfides, cysteinylation) is 23001.2079; while the observed monoisotopic mass is 23001.2280. Therefore, the light chain mass measurement accuracy for the monoclonal antibody 42.18.E12 is 0.9 ppm.

High resolution mass analysis and top-down fragmentation (msx HCD) was conducted for all four observed light chains. Based on these top-down characterization results, 2 disulfide bonds and cysteinylation of the C-terminus cysteine are proposed for the light chains from monoclonal antibodies 256.3. E2; 5.6.H9; and 42.18. E12. A pyro-glu modification of the N-terminus glutamine in addition to a disulfide bond formation was proposed for the light chain for monoclonal antibody 5.55.D2. This example demonstrates a top-down approach to mass spectrometry by analyzing antibodies for post-translational modifications. Differences in theoretical and observed masses led to characterization of post- translational modifications including but not limited to disulfide bonds, pyro-glu

modifications of glutamine and C-terminus cysteinylations. This example further

demonstrates that post-translational modifications other than glycosylation can be detected in whole or intact antibodies applying a top-down approach to mass spectrometry.

EQUIVALENTS AND SCOPE

Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. The scope of the present invention is not intended to be limited to the above Description, but rather is as set forth in the following claims.