IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 1
`
`The
`Human IgG
`Subclasses
`
`Robert G. Hamilton, Ph.D., D. ABMLI
`
`Asthma and Allergy Center
`Johns Hopkins University School of Medicine
`Baltimore, MD 21224
`
`Revised and Edited
`by
`Chandra Mohan,Ph.D.
`
`© Copyright 1987, 1989, 1992, 1994, 1998, 2001 Calbiochem-Novabiochem Corporation
`
`Ex. 2031-0001
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 2
`
`A Word to Our Valued Customers
`We are pleased to present you with this new edition of The Human IgG
`Subclasses Booklet. As part of our continuing commitment to provide useful
`product information and exceptional service to our customers, we have
`compiled this practical resource for investigators who are interested in the
`rapidly expanding field of quantitation of human immunoglobulins, especially
`the IgG subclass proteins. Whether you are just beginning your research or are
`training new researchers in your laboratory, you will find this booklet to be a
`highly useful reference source.
`Calbiochem is a world leader in providing highly innovative products for
`your research needs in Signal Transduction, including the areas of Cancer
`Biology, G-Proteins, Apoptosis, Protein Kinases, and Nitric Oxide-related
`phenomena. Please call us today for a free copy of our LATEST Signal
`Transduction Catalog and Technical Resource.
`If you have used Calbiochem products in the past, we thank you for your
`support and confidence in our products. And, if you are just beginning your
`research career, please call us and give us an opportunity to demonstrate our
`exceptional customer and technical service.
`Please call us and ask for a current listing of our ever expanding Technical
`Resource Library, now with over 50 Calbiochem publications. Or, check out
`our website at http://www.calbiochem.com for even more useful information.
`
`Meddi Awalom
`Sr. Product Manager
`Immunochemicals
`
`A name synonymous with innovative products, high quality, and exceptional service.
`
`2
`
`Ex. 2031-0002
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 3
`
`Table of Contents
`
`Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
`
`Properties of the Human IgG Subclasses . . . . . . . . . . . . . . . . . . . . . . 7
`
`Human IgG Subclass-Specific Monoclonal Antibodies . . . . . . . . . . . 12
`
`Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
`
`Literature Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
`
`Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
`
`Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
`Monoclonal Antibody Conjugates. . . . . . . . . . . . . . . . . . . . . . . . . 35
`Quantitative IgG Subclass Immunoassay Protocol . . . . . . . . . . . 39
`
`References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
`
`CALBIOCHEM®’s Anti-Human Antibodies and Conjugates . . . . . . . . 63
`
`3
`
`Ex. 2031-0003
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 4
`
`4
`
`Ex. 2031-0004
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 5
`
`I. Introduction
`
`The vertebrate immune system consists of well diversified molecules that
`recognize and respond to parasitic invasion in a very complex manner (1). The
`immune system is classified as innate – consisting of barriers to prevent pene-
`tration and spread of infectious agents, and adaptive system – consisting of
`lymphocytes and immunoglobulins. Lymphocytes consist of T cells and B cells
`that regulate immune response and impart cellular and humoral immunity to
`the organism. The B cells develop into plasma cells that secrete antibodies.
`The T cells develop into effector cells that kill infected cells as well as activate
`macrophages and B cells.
`
`Immune System
`
`Immune System
`
`Innate System
`
`Adaptive System
`
`Adaptive System
`Immunoglobulins
`Lymphocytes
`IgG
`Immunoglobulins
`IgG
`IgM
`
`Lymphocytes
`T cells
`B cells
`
`T cells
`
`B cells
`
`Innate System
`Biochemical
`Physical Barrier
`Lysozyme
`Skin
`Biochemical
`Physical Barrier
`Lysozyme
`Skin
`Complement
`Mucosa
`
`Complement
`
`Mucosa
`Phagocytes
`
`IgM
`IgA
`
`Phagocytes
`
`IgA
`IgD
`IgD
`IgE
`IgE
`Figure 1:Organization of the Vertebrate Immune System
`
`Activated
`B cells
`Activated
`B cells
`
`Plasma
`Cells
`Plasma
`Cells
`
`The human immunoglobulins are a group of structurally and functionally
`similar glycoproteins that confer humoral immunity in humans (2). They are
`composed of 82 - 96% protein and 4 - 18% carbohydrate. The immunoglobulin
`protein “backbone” consists of two identical “heavy” and two identical “light”
`chains. Five classes of immunoglobulins (IgG, IgA, IgM, IgD, and IgE) have
`been distinguished on the basis of non-cross-reacting antigenic determinants
`in regions of highly conserved amino acid sequences in the constant regions of
`their heavy chains (3). Four distinct heavy chain subgroups of human IgG were
`first demonstrated in the 1960’s by using polyclonal antisera prepared in
`animals immunized with human myeloma proteins (4-6). A World Health
`Organization (WHO) panel defined them as subclasses 1, 2, 3, and 4 of human
`IgG based on their relative concentration in normal serum and their frequency
`of occurrence as myeloma proteins (Table 1) (7). The structure and function of
`each human IgG subclass protein has been studied extensively, initially with
`polyclonal antisera rendered monospecific by immunoabsorption and more
`recently with monoclonal antibodies.
`
`5
`
`Ex. 2031-0005
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 6
`
`The polyclonal reagents used in IgG subclass studies have not been widely
`available, and they are difficult to prepare, invariably weak, and frequently
`contain a heterogeneous mixture of antibodies specific for immunoglobulin
`subclass-associated allotypes (8, 9). In the 1980’s, murine hybridoma techno-
`logy was used successfully by several groups to produce monoclonal antibod-
`ies specific for the human IgG subclass proteins (8, 10-12). The Human
`Immunoglobulin Subcommittee of the International Union of Immunological
`Societies (IUIS), supported by the WHO, conducted an extensive collaborative
`study of 59 monoclonal antibodies with reported subclass specificity by using a
`variety of immunological assays (13, 14). Highly specific monoclonal antibod-
`ies are now available as research and clinical reagents to facilitate quantitation
`of the level of each IgG subclass in human serum. These antibodies also are
`being applied to the study of IgG subclass antibodies produced in human
`immune responses.
`
`This monograph has been prepared as a general guide for investigators
`who are interested in the rapidly expanding field of quantitation of human IgG
`subclass proteins. The HP-series of immunochemicals discussed in this mono-
`graph includes monoclonal antibodies specific for human IgG PAN, IgG1, IgG2,
`IgG3, IgG4, and the human κ (kappa) and λ (lambda) light chains. This guide is
`intended only as a summary of basic information and not as an all-inclusive
`compendium of facts regarding the human IgG subclasses. First, physical,
`chemical, and biological properties of the human IgG subclasses are summa-
`rized. Second, methods are discussed that are used in the preparation,
`isolation, and quality control of the HP-series monoclonal antibodies. Third,
`applications for these monoclonal antibodies are examined, with emphasis on
`measurement of the level of IgG subclasses 1, 2, 3 and 4 in human serum and
`detection of IgG subclass antibodies by immunoassay. Finally, a bibliography is
`provided that directs the reader to past research and current trends in the
`study of human IgG subclasses in human health and disease.
`
`6
`
`Ex. 2031-0006
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 7
`
`II. Properties of the Human IgG
`Subclasses
`Physical and Chemical Properties
`The human IgG subclasses are glycoproteins (approx. 150 kDa) composed
`of two heavy (2 x 50 kDa) and two light (2 x 25 kDa) chains linked together by
`interchain disulfide bonds (15-17). Intra-chain disulfide bonds are responsible
`for the formation of loops, leading to the compact, domain-like structure of the
`molecule. Schematic diagrams of IgG 1, 2, 3 and 4 are presented in Figure 2.
`There are two types of light chains, which are referred to as lambda (λ) and
`kappa (κ) chains. The ratio of κ to λ varies from species to species, (e.g., in
`
`A
`
`pFc
`
`pFc
`
`C
`
`IgG1
`
`IgG2
`
`B
`
`Fc
`
`Fc
`
`Fab
`
`F(ab')2
`
`IgG3
`
`Fab
`
`F(ab')2
`
`D
`
`Fc
`
`Fc
`
`IgG4
`
`pFc
`
`pFc
`
`F(ab')2
`
`Fab
`
`Fab
`
`F(ab')2
`
`Figure 2.Schematic diagram of the four subclasses of human IgG. The figure shows
`the major pepsin cleavage points (LLLL), major papain cleavage points (•), C1q bind-
`ing site exposed ( ), C1q binding site exposed only in isolated Fc fragments ( ),
`constant region of heavy and light chains ( ), variable region of the heavy and light
`chains that contribute to the antigen binding site ( ) and the carbohydrate side
`chains (
`). Reproduced with permission from Immunology Today, June 1980.
`
`7
`
`Ex. 2031-0007
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 8
`
`mice 20:1, in humans 2:1). This ratio can sometimes be used as a marker of
`immune abnormalities.
`
`The amino terminal regions of the heavy and light chains exhibits highly
`variable amino acid composition (referred as VH and VL respectively). This vari-
`able region is involved in antigen binding. In contrast to the variable region, the
`constant domains of light and heavy chains are referred as CL and CH respec-
`tively. The constant regions are involved in complement binding, placental
`passage, and binding to cell membrane. Differences in the amino acid content
`of the heavy chains and the ratio of κ to λ light chains are characteristic of the
`different subclasses of IgG. While the primary amino acid sequences of the
`constant regions of the IgG subclass heavy chains are greater than 95%
`homologous, major structural differences are found in the hinge region in terms
`of the number of residues and interchain disulfide bonds (Table 1).
`
`The hinge region is the most diverse structural feature of different IgGs. It
`links the two Fab arms to the Fc portion of the IgG molecule and provides flexi-
`bility to the IgG molecule. Also, it forms a connecting structure between the two
`heavy chains. The flexibility of the hinge region is important for the Fab arm to
`interact with differently spaced epitopes, and for the Fc region to adapt differ-
`ent conformations. The disulfide bonds in the middle hinge region are
`important for covalent linking of the heavy chains.
`The IgG1 hinge is 15 amino acid residues long and is freely flexible so that
`the immunoglobulin regions or fragments that bind antigen (Fabs) can rotate
`about their axes of symmetry and move within a sphere centered at the first of
`two interchains disulfide bridges (18). IgG2 has a shorter hinge than IgG1, with
`12 amino acid residues and four disulfide bridges at the Fab base. The hinge
`region of IgG2 also lacks a glycine residue, which together with its shortness
`almost completely prevents rotation and restricts lateral movement of the Fabs
`(19). IgG3 has a unique elongated hinge region containing 62 amino acids (21
`prolines and 11 cysteines) that has been described as an inflexible polyproline
`double helix (16, 19-21). The IgG3 Fabs appear to rotate and wave at a rate
`similar to those in IgG1; however, remoteness of the Fc (crystallizable frag-
`ment) from the Fab causes the Fab to be less frequently near the Fc over time.
`This makes it more readily available for binding of complement component 1q
`(C1q) to the Fc region of IgG3 in solution in comparison with its binding to IgG1
`Fc. Finally, IgG4’s hinge is shorter than that of IgG1, its flexibility is intermediate
`between IgG1 and IgG2 and some rotation may occur around the glycine
`residue in its hinge region. Access of C1q to the IgG4 Fc is hindered by the
`shortness of the IgG4 hinge, which leads to the Fabs spending more time close
`to the Fc (19).
`
`The point of light chain attachment to the heavy chain also differs among the
`subclasses. IgG1 light chains are bound near the midpoint of the heavy chain,
`while those of IgG2, IgG3 and IgG4 are joined one quarter the distance from the
`heavy chain amino termini (19) (Figure 1). Intrachain disulfide bonds of the
`heavy and light chains transform parts of the molecule into compact globular
`regions called domains. These domains participate in the biological functions
`of the immunoglobulin. Unique antigenic determinants are generally found in
`
`8
`
`Ex. 2031-0008
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 9
`
`the Fc region of IgG1 and IgG2, the hinge region of IgG3 and the Fc and Fd
`regions of IgG4 (22).
`Genetic markers (Gm allotypes) are regular minor differences in primary
`amino acid sequences between molecules of one IgG subclass that occur
`throughout a species as a result of gene mutation (23-26). In humans, some
`allotypic markers are restricted to constant region domains of single IgG
`subclasses, while others are shared by several subclasses. Examples of
`shared or isoallotypes are Gm4a, which has been detected on some human
`IgG1, IgG3 and IgG4 molecules, and Gm4b, which is shared by human IgG2 and
`IgG4 molecules (27). In humans, certain allotypes have been associated with
`increased and decreased antibody responses to a variety of bacterial
`pathogens, autoantigens, isoantigens, tumor antigens, and dietary antigens
`(28). Excellent discussions of the human IgG allotypes and their importance
`are presented elsewhere (23, 24, 28-42).
`
`The sedimentation coefficient of the four IgG subclasses is the same (S = 7).
`Early studies indicated that isoelectric focusing (IEF) may be useful in the sepa-
`ration of the four human IgG subclasses based on differences in their net
`charges (43-45). More recent studies using two-dimensional gel electrophoresis
`of serum from patients with monoclonal gammopathies have shown that the IgG
`subclasses are not separated readily by charge alone because their pI ranges
`overlap each other between pH 6.4 and 9.0 (46) (Table 1). Characterization of
`the human IgG subclasses has been accomplished in part by digesting IgG
`subclass preparations with proteolytic enzymes such as papain (47, 48), plas-
`min (49), trypsin (50), and pepsin (51). Papain, in the presence of cysteine,
`digests IgG into two Fab fragments, a Fc fragment, and degradation products.
`Up to two-hour incubation, IgG2 protein appears to be resistant to degradation
`(10-20% digested) with papain, while proteins of the other subclasses are
`completely degraded (52). Pepsin digests IgG into F(ab′)2 with intact antigen-
`binding activity and a pFc′ or small polypeptide chains without antibody activity.
`While IgG3 and IgG4 appear to be relatively more sensitive to pepsin digestion,
`all four subclasses can be digested eventually. Studies of structure and function
`of the proteins of the human IgG subclasses by using enzymatically digested
`fragments are reviewed elsewhere (16, 53).
`Biological Properties
`The biological properties of the human IgG subclasses may be categorized
`as specific reactions of the Fab region with antigen (primary function) and
`effector (secondary) functions. These reactions occur as a result of antigen
`binding and are mediated through interaction of the constant regions of the
`heavy chain, especially the Fc. Principal secondary biological functions of the
`four human IgG subclasses are summarized in Table 2.
`
`The concentration of each immunoglobulin in serum of healthy individuals
`depends in part on the number of plasma cells that produce that particular
`immunoglobulin, the rate of synthesis, catabolism, and the exchange between
`intra- and extravascular spaces. Adults have the highest concentration of IgG1 (5-
`
`9
`
`Ex. 2031-0009
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 10
`
`12 mg/ml), followed by IgG2 (2 - 6 mg/ml), IgA1 (0.5 - 2 mg/ml), IgM (0.5 - 1.5
`mg/ml), IgG3 (0.5 - 1.0 mg/ml), IgG4 (0.2 - 1.0 mg/ml), IgA2 (0 - 0.2 mg/ml), IgD
`(0 - 0.4 mg/ml) and IgE (0 - 0.002 mg/ml) (53, 54). The rare IgG subclasses tend
`to vary more considerably between individuals (2, 15). The IgG concentration of
`a given individual appears to be related to the Gm allotype which indicates that
`genetic factors are one variable that determine the overall IgG subclass concen-
`tration in serum (28, 37, 55). IgG is detected rarely in secretions (2). The 5:1 ratio
`of IgG to IgA in serum contrasts with the 1:20 ratio detected in saliva and other
`secretions. Total IgG is about 100 times lower in cerebrospinal fluid (CSF) than in
`serum (0.8 − 7.5 mg/dl), which represents about 12% of the CSF protein (2).
`IgG exhibits highest synthetic rate and longest biological half-life of any
`immunoglobulin in serum. Studies of clearance rates of radiolabeled IgG
`myeloma proteins in vivohave demonstrated a higher catabolic rate for human
`IgG3 than for IgG1, IgG2, and IgG4 (Table 2) (56). Proteins of all four IgG
`subclasses can pass from the mother to the fetus through the placenta (57-59).
`The transfer of IgG antibodies from mother to the fetus appears to be mediated
`by an active transport mechanism that involves Fc receptors at the syncytiotro-
`phoblast membrane that bind the IgG molecules (60). Differential in vitro
`binding affinity to placental homogenates (IgG1 = IgG3>IgG4>IgG2) suggests
`that the transfer of IgG across the placenta may be a selective process (57,
`61); however, this theory has not been supported by all studies (62, 63).
`Factors involved in the development of serum IgG subclass levels from the
`prenatal through adolescent years are reviewed elsewhere (58, 62, 64-67).
`
`Complement activation is possibly the most important biological function of
`IgG. Activation of the complement cascade by the classical pathway is initiated
`by binding of C1 to sites on the Fc portion of human IgG. IgG subclass activa-
`tion of complement by the alternate pathway has not been demonstrated. The
`globular heads of C1q interact with amino acids 285−292 or 317−340 in the
`second heavy chain constant region (CH2). Reactivity of complement with IgGs
`of the four human subclasses varies as a function of steric interference by the
`Fab arms in the approach of C1q to the CH2 sites (IgG3>IgG1>IgG2>IgG4) (68,
`69). Binding of C1 to IgG3 myeloma is about 40 times greater than binding to
`IgG2, and binding to IgG4 generally is not demonstrable. IgG4 antibodies in fact
`appear to inhibit immune precipitation and binding of C1q to IgG1 in complexes
`containing mixtures of IgG1 and IgG4 (70). IgG4 thus may be considered protec-
`tive against the biological effects of the complement-fixing antibodies (71).
`
`Another vital function of human IgG is its ability to bind to cell surface Fc
`receptors. Once it is fixed to the surface of certain cell types, the IgG antibody
`can complex antigen and facilitate clearance of antigens or immune-complexes
`by phagocytosis. Three classes of human IgG Fc receptors (FcR) on leuko-
`cytes have been reported: the FcR-I, FcR-II, and low-affinity receptor [FcR-Io]
`(72). These are distinguished by their presence on different cell types, by their
`molecular weights and by their differential abilities to bind untreated or aggre-
`gated IgG myeloma protein of the four subclasses. Molecular weights of the
`IgG Fc receptor molecules are reportedly 72 kDa (FcR-I), 40 kDa (FcR-II) and
`50−70 kDa (FcR-Io). The receptors are expressed differentially on overlapping
`populations of leukocytes: FcR-I on monocytes; FcR-II on monocytes,
`10
`
`Ex. 2031-0010
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 11
`
`neutrophils, eosinophils, platelets, and B cells; and FcR-Io on neutrophils,
`eosinophils, macrophages, and killer T cells (72).
`FcR-I reportedly possesses greater affinity for IgG1 and IgG3 (Ka = 108 to
`109 M−1) than for IgG2 or IgG4. IgG4 binds less effectively, and IgG2 proteins
`almost never bind to FcR-I. Estimated cell surface density of FcR-I receptors
`on monocytes is 1 − 4 X 104 per cell. Studies of the FcR-I specificity compare
`well with earlier reports that monocytes have Fc receptors preponderantly for
`IgG1 and IgG3 (73-75). FcR-II specificity has been evaluated only on platelets.
`Aggregated human IgG myeloma proteins of all four subclasses are able to
`release 3H-serotonin from platelets, indicating the presence of receptors for
`all subclasses on the human platelet (76). Use of oligoclonal IgG has shown
`that platelets bind IgG1 = IgG3>IgG2 and IgG4. Addition of complement to the
`medium inhibits the release of serotonin from platelets incubated with aggre-
`gated IgG1 and IgG3, but not with IgG2 and IgG4. This suggests that
`complement binds to IgG aggregates and sterically hinders the reaction of
`IgG Fc with the platelet receptor (2). The low-affinity human IgG receptor has
`not been well defined. Studies of the neutrophil have shown preferential bind-
`ing of IgG1 and IgG3 to FcR-Io. Release of lysosomal enzymes such as
`β-glucuronidase from neutrophils by incubation with aggregated IgG myeloma
`proteins indicates that all subclasses of human IgG can react with the
`neutrophil (75, 77, 78). Study of IgG subclass Fc receptors on human lympho-
`cytes by using human myeloma proteins has demonstrated that IgG1, IgG2
`and IgG3 can bind to lymphocytes and inhibit lymphocyte cytotoxicity (79).
`Human IgG subclasses are known to bind to other proteins. The Fc region of
`human IgG1, IgG2 and IgG4 binds to protein A from Staphylococcus aureus
`(80, 81). A single substitution of arginine for histidine at amino acid 435 in the
`Fc region prevents binding of protein A to IgG3 (82). Patients with cystic fibrosis
`can express a factor in their serum that is a heat- and acid-labile low-molecu-
`lar-weight protein that binds to the constant regions of human IgG1 and IgG2
`(83). Human rheumatoid factors (RF) are IgG, IgA or IgM antibodies that bind
`to the Fc of immunoglobulins (84). In most cases, IgG is also the antigen for
`RF. Human rheumatoid factors react most strongly with IgG1 myeloma proteins
`followed by IgG2 and IgG4. IgG3 appears to be unreactive with RF(85). The
`biological significance of differential binding of the human IgG subclasses to
`human leukocytes and human or foreign proteins is discussed in detail else-
`where (2, 53, 86).
`
`This overview has summarized major differences in the structure and effec-
`tor functions of the four human IgG subclasses. At present, the precise role of
`each IgG subclass protein within the totality of the immune response remains
`to be elucidated. The observation that seemingly healthy individuals may be
`deficient in one IgG subclass challenges the notion the IgG subclass proteins
`have unique and essential roles in an immune response. However, certain anti-
`genic challenges (e.g., bacterial and viral antigens, allergens) elicit a selective
`increase in IgG antibodies of certain subclasses (71, 87, 88). Thus, as has
`been postulated, emergence of the IgG subclasses may permit the efficiency
`of certain effector functions to be optimized within individual subclasses (16).
`
`11
`
`Ex. 2031-0011
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 12
`
`III.Human IgG Subclass-Specific
`Monoclonal Antibodies
`Preparation
`The HP-series of human IgG specific monoclonal antibodies was produced
`from documented hybridoma cell lines that were developed at the Centers for
`Disease Control in Atlanta, Georgia, U.S.A. (89). Hybridomas were maintained
`in cell culture in RPMI with penicillin-streptomycin-fungizone and 10% fetal calf
`serum for 2 − 8 weeks before use. Antibody-containing ascites (in lots of 500−
`1500 ml) was prepared by injecting hybridoma cells (2 − 5 X 106 viable cells
`per mouse) intraperitoneally into 80 to 100-day-old BALB/c mice that had been
`primed 2 weeks earlier with 0.5 − 1 ml of pristane. The ascites was harvested
`5 − 10 days after the injection of cells and immediately centrifuged to remove
`erythrocytes, lipid, and pristane. Filtered ascites was frozen at −70°C without
`azide. The clone number, murine isotype, and pI of the HP-series on mono-
`clonal antibodies are presented in Table 3.
`Isolation
`Monoclonal antibody was purified chromatographically from ascites for cova-
`lent coupling to affinity chromatography matrices; adsorption to immunoassay
`solid phases; conjugation with biotin, enzymes, or fluorescent molecules; or
`labeling with radioiodine. Routine isolation was performed by using DEAE ion
`exchange chromatography (90) followed by hydroxylapatite (Cat. No. 391947
`and 391948) chromatography (91). Protein A affinity purification was avoided to
`eliminate any possibility of contaminating the purified monoclonal antibody with
`protein A, itself a human IgG binding protein (see literature survey.)
`
`Ascites was dialyzed (15,000 M.W. exclusion) overnight against 0.05 M Tris,
`pH 7.7 at +4°C, applied to a column containing DEAE cellulose and eluted with
`a step gradient by using 0.05 M Tris containing 0 to 0.15 M NaCl. Protein peaks
`were monitored by absorbance at 280 nm and fractions around protein peaks
`were analyzed by ELISA (92) to identify immunoreactive monoclonal antibody.
`Column fractions containing antibody were concentrated two- to ten-fold
`(Amicon, YM10 membrane) and analyzed by ELISA, isoelectric focusing (IEF),
`and/or immunoelectrophoresis (IEP) (92, 93).
`
`DEAE-isolated monoclonal antibody that contained any detectable contami-
`nants was subjected to hydroxylapatite chromatography (91). Antibody was
`dialyzed in 0.01 M sodium phosphate buffer, applied to the hydroxylapatite column
`and eluted with 0.01 M sodium phosphate buffer followed by stepwise increase in
`sodium phosphate concentration. The actual salt gradient was designed around
`the known pI of the monoclonal antibody. Analysis by ELISA, IEF and/or IEP was
`repeated, and column fractions containing purified monoclonal antibody were
`pooled, concentrated to 2 − 5 mg/ml (based on optical density), aliquoted and
`frozen at −20°C. Protein content of the IgG was detemined by A280 (E1%/1 cm = 15)
`and by protein assay by using purified mouse IgG standards.
`
`12
`
`Ex. 2031-0012
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 13
`
`Analysis and Quality Control
`Laboratory analysis of each lot of monoclonal antibody was performed in
`three stages from production to final product. First, the cell culture medium was
`analyzed for the presence and relative amount of human IgG-specific antibody
`by ELISA before hybridoma cells were injected into mice. Second, collected
`ascites was analyzed for potency and antibody specificity by using dilutional
`analysis in ELISA and by IEF, often in combination with an affinity immunoblot
`(92, 93). Final purity, quantity, immunoreactivity, and specificity of isolated anti-
`body were documented by using IEF and ELISA.
`Specificity
`Specificity of each lot of antibody was tested and compared to previous lots
`and previous reports of specificity for that clone (13, 89, 92, 94). Serial dilutions of
`ascites samples were analyzed by ELISA by using microtiter plate wells coated
`with human IgG myeloma proteins of the four subclasses. The ratio of reciprocal
`dilutions of monoclonal antibody binding to heterologous vs homologous
`myeloma protein subclasses at 5%, 20%, and/or 50% of the maximum optical
`density (ODmax) was used as a measure of cross-reactivity. Specificity of the HP-
`series of monoclonal antibodies is summarized in Table 4. Results obtained in
`these analyses agree well with similar studies performed at the CDC (89) and in
`the IUIS/WHO collaborative study in laboratories using ELISA and immunofluoro-
`metric assays (13 , 95). Dilution curves generated in a representative cross-
`reactivity study of HP6025 (anti-human IgG4 Fc) are presented in Figure 3.
`
`100
`
`HP 6025
`
`Solid Phase
`
`IgG1 myeloma
`
`IgG2 myeloma
`
`IgG3 myeloma
`
`IgG4 myeloma
`
`1
`
`5
`
`25
`
`50
`
`100
`
`200
`
`400
`
`800
`
`1600
`
`3200
`
`6400
`
`12800
`
`25600
`
`ASCITES DILUTION (x1000)
`
`80
`
`60
`
`40
`
`20
`
`0
`
`%ODMax
`
`Figure 3. Determination of specificity by dilution analysis. Thirteen dilutions of
`HP6025 were analyzed in an ELISA by using microtiter wells coated with human IgG1,
`IgG2, IgG3 or IgG4 myelomas. Cross-reactivity was defined as the ratio of ascites dilu-
`tion that produced the same optical density after binding of monoclonal antibody to
`homologous (IgG4 myeloma) vs heterologous IgG subclass (IgG 1, 2, or 3).
`
`13
`
`Ex. 2031-0013
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 14
`
`Quantitation of Antibody
`The quantity of immunoreactive monoclonal antibody in ascites was
`analyzed by ELISA (96). Microtiter plate wells were coated with one of four
`human IgG subclass myelomas or with bovine serum albumin (BSA, negative
`control). Dilutions of antibody in ascites and purified form were incubated in
`replicate wells coated with human IgG-subclass myelomas. Bound murine
`antibody was then detected by means of enzyme-conjugated polyclonal anti-
`serum to mouse IgG (preabsorbed against human IgG) and developed with
`substrate. Net optical density was plotted as a function of the reciprocal dilu-
`tion of ascites or nanograms per ml of purified antibody standard. Parallel
`ascites dilution curves obtained in a potency study are presented in Figure 4.
`The quantity of antibody was determined either (A) in weight per volume
`units by interpolation from a dose-response curve produced by using a chro-
`matographically purified preparation of the same monoclonal antibody with
`known concentration (mg/ml) of antibody (standard), or (B) in arbitrary units
`as a ratio of the dilution of the test sample vs a reference sample at 50%
`maximum optical density. Approach B was used only in the screening of the
`culture medium and initial evaluation of ascites for the monoclonal antibody
`of interest.
`
`HP6017-aGFc
`
`HP6046-aGfd
`
`HP6001-aG1
`
`HP6002-aG2
`
`HP6050-aG3
`
`HP6025-aG4
`
`Solid Phase
`Human IgG PAN
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`%ODMax
`
`1
`
`5
`
`25
`
`50
`
`100
`200
`400
`800
`1600
`3200
`ASCITES DILUTION (x1000)
`
`6400
`
`12800
`
`25600
`
`Figure 4.Determination of potency by dilution analysis. Binding curves are shown for
`13 dilutions of six monoclonal antibodies (HP6017-aGFc, HP6046-aGFd, HP6001-
`aG1, HP6002-aG2, HP6050-aG3, and HP6025-aG4) analyzed on human IgG PAN
`(1, 2, 3, 4)-coated microtiter wells. The potency of antibody in each ascites sample
`was defined as the reciprocal dilution at 50% maximum response (%ODmax). Results
`from these studies are summarized in Table 3. Reproduced with permission from (92).
`
`14
`
`Ex. 2031-0014
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 15
`
`Purity of Isolated Antibody
`Purity of each lot of monoclonal antibody was assessed by IEF and/or
`crossed-immunoelectrophoresis (XIE). XIEs of unprocessed and partially puri-
`fied ascites and of purified monoclonal antibody are reproduced in Figure 5.
`
`IEF analysis also permitted routine quality control of each lot of antibody in
`terms of its relative purity. Ascites generally contained variable amounts of
`polyclonal host mouse IgG as shown by direct immunoblot analysis of the
`ascites after IEF (Figure 6).
`
`A
`
`B
`
`C
`
`2˚
`
`1˚
`
`UNPROCESSED
`ASCITES
`
`PARTIALLY
`PURIFIED
`ANTIBODY
`
`PURIFIED
`MONOCLONAL
`ANTIBODY
`
`Figure 5.Crossed-immunoelectrophoresis analysis of mouse antibody to human IgG
`Fd (HP6045). Unprocessed ascites (panel A), partially purified antibody (panel B) and
`chromatographically purified monoclonal antibody (panel C) were separated by elec-
`trophoresis in a first dimension (1°) in agarose, cut out, inserted into a second gel and
`subjected to electrophoresis in a second dimension (2°) into a gel containing goat
`antiserum to mouse immunoglobulin. The height of each band reflects the relative
`quantity of IgG, and the number of bands relate to the purity. Panel A shows multiple
`peaks that indicate the presence of major quantities of albumin, transferrin, and other
`mouse proteins in addition to mouse IgG. The partially purified antibody in panel B
`contains a small amount of transferrin contaminant. Panel C shows the purity of the
`chromatographically purified mouse IgG and monoclonal antibody (HP6045), which is
`free from other host protein contaminants.
`
`15
`
`Ex. 2031-0015
`
`

`
`IGG booklet_2001.qxd 4/25/2001 2:42 PM Page 16
`
`– 8.64
`
`– 8.13
`
`– 7.47
`
`– 6.90
`
`– 6.43
`
`– 6.22
`
`– 5.76
`
`– 5.18
`
`pH
`
`HP6000(GPANFc)
`
`HP6017(GPANFc)
`
`HP6046(GPANFd)
`
`HP6025(G4Fc)
`
`HP6023Cx
`
`HP6050(G3H)
`
`HP6047(G3H)
`
`HP6014(G2Fd)
`
`HP6002b(G2Fc)
`
`HP6002a(G2Fc)
`
`Figure 6.Direct immunoblots for detection of mouse IgG of ascites. Ten murine ascites
`samples were focused isoelectrically in a polyacrylamide gel that was subsequently
`overlaid with untreated nitrocellulose paper. Bound murine IgG was detected with
`peroxidase-conjugated antiserum to mouse IgG and developed with substrate.
`Polyclonal murine IgG in the ascites displayed a heterogeneous pI range from pH 5.5-
`8.0. Major (dense