`Page 1
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 1
`
`
`
`Applications and Engineering of
`Monoclonal Antibodies
`
`HOSPIRAEX. 1029
`Page 2
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 2
`
`
`
`Applications and Engineering of
`Monoclonal Antibodies
`
`DAVI D J. KING
`
`Ce//tech Therapeutics
`Slough, UK
`
`HOSPIRA EX. 1029
`Page 3
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 3
`
`
`
`UK Taylor & Francis Ltd, I Gunpowder Square, London EC4A 3DE
`USA Taylor & Francis Inc., 325 Chestnut Street, 8th Floor, Philadelphia, PA 19106
`
`Copyright © D.J. King 1998
`All rights reserved. No part of this publication may be reproduced, stored in a retrieval
`system, or transmitted, in at~yjbrnz or by any means, electronic, electrostatic, magnetic
`tape, mechanical, photocopying, recording or otherwise, without the prior pennission of
`the copyright owner
`
`Evemy effort has been made to ensure that the advice and information in this book is true
`and accurate at the time of going to press. However, neither the publisher nor the author
`can accept any legal responsibility or liability for any errors or omissions that may, he
`made. In the case of drug administration, any medical procedure or the use of technical
`equipment mentioned within tlii.r book, you are strongly advised to consult the
`manufacturer's guidelines.
`
`British Library Cataloguing-in-Publication Data
`A catalogue record for this book is available from the British Library
`ISBN 0-7484-0422-8 (fib)
`0-7484-0423-6 (b)
`
`Library of Congress Cataloging Publication Data are available
`
`Cover design by Youngs Design in Production
`Typeset in Times 10/12pt by Graphicraft Limited, Hong Kong
`Printed by T.J. International Ltd, Padstow, UK
`
`HOSPIRA EX. 1029
`Page 4
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 4
`
`
`
`I would like to thank my many colleagues at Ceiltech for their helpful suggestions and
`advice, and particularly Alastair Lawson and Martyn Robinson for their comments on the
`manuscript. Thanks also to Tina Jones for assistance with the figures.
`And lastly, sincere thanks to my wife, Jane, for her support and patience through the
`long hours and weekends spent writing this book.
`
`HOSPIRA EX. 1029
`Page 5
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 5
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Conteits
`
`Preparation, structure and function of monoclonal antibodies
`1.1
`Introduction
`1.2 The role of antibodies in the immune response
`1.3 Structure and function of antibodies
`1.4 The organisation of antibody genes
`1.5 Antigen-binding affinity and avidity
`1.5.1 Affinity
`1.5.2 Avidity
`1.6 Generation of monoclonal antibodies
`1.6.1 Hybridoma technology
`1.6.2 Human monoclonal antibodies
`1.6.3 Human MAbs from transgenic mice
`1.6.4
`Isolation of antibodies by phage display
`
`2 Antibody engineering: design for specific applications
`2.1
`Introduction
`2.2
`Isolation of variable region genes
`2.3 Overcoming immunogenicity
`2.3.1 Chimeric and humanised antibodies
`2.3.2 Antibody fragments to reduce immunogenicity
`2.3.3 Chemical modification to reduce immunogenicity
`2.3.4
`Immunosuppressive therapy
`2.4 Antibody fragments
`2.4.1 Antibody fragments from proteolysis of 1gG
`2.4.2 Recombinant antibody fragments
`Fab-based fragments
`Fv-based fragments
`Multivalent antibody fragments
`2.5 Antibodies with multiple specificities
`2.6 Engineering effector functions
`
`page 1
`1
`2
`2
`11
`13
`13
`14
`14
`14
`18
`20
`21
`
`27
`27
`28
`29
`29
`37
`38
`40
`40
`40
`42
`43
`45
`48
`50
`54
`
`HOSPIRA EX. 1029
`Page 6
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 6
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`viii
`
`Contents
`
`26.1 Engineering natural effector functions
`2.6.2 Attachment of diagnostic or therapeutic agents
`Chemical conjugates
`Site-specific attachment
`Fusion proteins
`2.7 Engineering pharmacokinetics and biodistribution
`2.7.1 Pharmacokinetics of IgG
`2.7.2 Pharmacokinetics of antibody fragments
`2.7.3 Clearance
`2.7.4 Chemi6al modification
`2.7.5 Fe region to extend half-life
`
`3.3
`
`3 Monoclonal antibodies in research and diagnostic applications
`3.1
`Introduction
`Immunoassays in diagnostics and research
`3.2
`3.2.1 Radioimmunoassay
`Immunoradiometric assay
`3.2.2
`3.2.3 Non-isotopic immunoassays
`3.2.4
`Improving sensitivity
`3.2.5 Assay formats
`3.2.6 Advantages of monoclonal antibodies in immunoassay
`Immunosensors
`3.3.1 Mass-detecting irnmunosensors
`3.3.2 Electrochemical immunosensors
`3.3.3 Optical immunoscnsors
`Immunocytochemistry
`3.4
`3.5 Flow cytometry and cell sorting (FACS)
`3.6 Western blotting (immunoblotting)
`3.7
`Immunopurification
`3.8 Antibodies in structural biology
`3.9
`In vivo diagnostics
`3.9.1 Radioimmunodetection of human tuniours
`Tumour-associated antigens
`Form of antibody
`Radioisotopes
`Two- and three-step targeting approaches
`3.9.2 Radioimmunoguided surgery
`3.9.3 Non-tumour radioimmunodetection
`
`4 Monoclonal antibodies in therapeutic applications
`4.1
`Introduction
`4.2 Cancer
`4.2.1 Cancer therapy with unmodified (naked) antibodies
`4.2.2 Anti-idiotype antibodies
`4.2.3 Bispecific antibody-mediated effector cell targeting
`4.2.4 Other approaches to recruit the immune system using MAbs
`4.2.5 Radioimmunotherapy
`Form of antibody
`Radioisotopes
`
`54
`58
`58
`4.3
`65
`67
`68
`71
`74
`75
`75
`
`77
`77
`77
`78
`79
`83
`85
`89
`91
`92
`93
`94
`95
`97
`99
`100
`102
`106
`106
`107
`107
`109
`110
`114
`116
`116
`
`119
`119
`120
`121
`123
`125
`127
`128
`129
`132
`
`HOSPIRAEX. 1029
`Page 7
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 7
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Contents
`
`4.3
`
`Clearance mechanisms and two-step targeting
`Clinical results with RIT
`4.2.6
`Immunotoxins
`4.2.7 Drug conjugates
`4.2.8 Antibody-directed enzyme prodrug therapy (ADEPT)
`4.2.9 Vascular targeting
`Infectious disease
`4.3.1 Antiviral antibodies
`4.3.2 Bacterial sepsis
`4.4 Cardiovascular disease
`4,4.1
`Inhibition of platelet aggregation
`4.4.2 Thrombolysis
`4.5 Disorders of the immune system /inflammatory diseases
`4.5.1 The inflammatory response
`4.5.2 Blocking inflammatory mediators
`Anti-TNF antibodies in rheumatoid arthritis and
`inflammatory bowel disease
`Anti-CS
`4.5,3 Blocking adhesive interactions
`4.5.4 Antibodies which directly inhibit T cell activation and
`proliferation
`4.5.5 Antibody treatment of allergy
`
`5 Production of monoclonal antibodies
`Introduction
`5.1
`5.2 Expression of antibodies in mammalian cells
`5.2,1 Transient expression systems
`5.2.2 Stable expression systems
`5.2.3 Expression of antibody fragments in mammalian cells
`5.3 Expression in Escherichia co/i
`Intracellular expression of antibody fragments in E. co/i
`5.3.1
`5.3.2 Secretion of antibody fragments from E. co/i
`5.4 Expression in other microbial systems
`5.5 Expression in plants
`5.6 Production in transgenic animals
`5.7 Expression in insect cells
`5.8 Production of monoclonal antibodies - cell culture
`5.9 Purification of monoclonal antibodies
`5.9.1 Purification of IgG
`5.9.2 Purification of 1gM
`5.9.3 Purification of monoclonal antibody fragments
`5.9.4 Purification for therapeutic use
`
`6 Prospects for engineered antibodies in biotechnology
`6.1 Gene therapy
`Intracellular antibodies
`6.1.1
`6.1.2 Other applications of MAbs in gene therapy
`6.2 Applications of antibodies in plants
`6.3 Catalytic antibodies
`
`ix
`
`136
`137
`139
`142
`144
`146
`147
`147
`148
`149
`149
`150
`151
`151
`153
`
`153
`154
`154
`
`155
`158
`
`161
`161
`161
`162
`163
`167
`167
`168
`170
`172
`173
`174
`175
`175
`176
`176
`181
`181
`185
`
`187
`187
`187
`188
`189
`189
`
`HOSPIRA EX. 1029
`Page 8
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 8
`
`
`
`6.4 Towards drug design
`Improving affinity
`6.5
`6.6 Summary and prospects
`
`7 References
`
`Index
`
`Contents
`
`190
`191
`192
`
`195
`
`241
`
`HOSPIRAEX. 1029
`Page 9
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 9
`
`
`
`I
`
`Preparation, Structure and Function
`of Monoclonal Antibodies
`
`1.1
`
`Introduction
`
`Antibodies are proteins produced by an individual in response to the presence of a foreign
`molecule in the body. These foreign molecules are known as antigens, and they usually
`result from invading organisms such as bacteria, fungi or viruses. Antibodies bind to
`antigens and elicit a range of effector mechanisms to destroy the invading organism.
`Therefore, the generation of an antibody response is a key step in the immune system
`which has evolved to protect individuals from invading pathogenic organisms. However,
`antibodies are not restricted in specificity to pathogens, but can be formed to a huge
`variety of antigens including proteins, carbohydrates and organic compounds, including
`totally novel structures.
`In 1975 Kohler and Milstein described a method for the 'production of antibodies of
`predefined specificity'. This technical breakthrough allowed, for the first time, the pro-
`duction of antibody molecules of a single specificity which could be characterised and
`defined. Such monoclonal antibodies immediately became valuable research tools, and
`applications in the diagnosis and therapy of human disease began to be widely investig-
`ated. Monoclonal antibodies (MAbs) have become increasingly important as diagnostic
`agents allowing precise molecular structures to be mapped and analysed. However, initial
`enthusiasm for their development as therapeutic agents was premature and many prob-
`lems limited their use in humans (see Chapter 4). A second technical revolution has now
`arrived in the ability to manipulate antibody genes and to design and produce anti-
`body molecules tailor-made for their application. Such redesigned antibody molecules
`are now rapidly becoming valuable reagents for therapy of human diseases as well as
`improved diagnostics and research reagents. The scope of these applications, and how
`antibody molecules are redesigned, or engineered, in the most suitable form for a particular
`application is the subject of this book. To understand the nature of this process requires
`an understanding of antibody structure and function and of how antibody specificities are
`generated.
`
`HOSPIRA EX. 1029
`Page 10
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 10
`
`
`
`Applications and Engineering of Manor/anal Antibodies
`
`1.2 The roe of antibodies in the immune response
`
`The immune system can be considered in two parts. Innate immunity is mediated by a
`variety of physical and biochemical barriers and by cells responding non-specifically to
`the foreign organism or molecule. Innate immunity is characterised by a similar reponse
`on re-exposure to the same foreign agent. Adaptive immunity, the second part of the im-
`mune system, is mediated by cells termed lymphocytes and is characteriscd by improved
`efficiency on re-exposure to the same foreign agent. It is in this system that antibodies
`play a major role.
`Two major types of lymphocytes are involved in the adaptive immune response, T
`lymphocytes and B lymphocytes. T lymphocytes can be subdivided into cytotoxic I
`lymphocytes and 'helper' I lymphocytes. Cytotoxie T lymphocytes bind to foreign or
`infected cells through a surface antigen receptor, the T cell receptor, and lyse them.
`Helper T lymphocytes play a regulatory role in controlling the response of both T and B
`lymphocytes. B lymphocytes exert their effect through producing antibody molecules
`which bind to the foreign agent and invoke specific mechanisms for its elimination.
`Antigen is recognised by B lymphocytes through the use of antibody molecules on the
`surface of the cell. Each B lymphocyte carries antibody molecules on its surface with a
`single specificity as a consequence of the rearrangement of immunoglobulin genes by the
`individual cell during its development in a random, antigen-independent process. Each
`individual will have many millions of B cells at one time thus comprising a 'polyclonal'
`population. When an antigen enters the body it will be recognised by any B cell which has
`antibody molecules able to bind to antigenic determinants (epitopes) present on the anti-
`gen. This recognition leads to activation of the cell leading to proliferation and differ-
`entiation (Figure 1.1).
`In the ease of protein antigens, bound antigen molecules are internalised into the B cell
`and degraded into peptides. Some of these antigen peptides are then bound to major
`histocompatibility complex (MHC) class II molecules to forni a complex which is trans-
`ported to the cell surface and 'presented' by the B cell. Although B cells are relatively
`weak at antigen presentation compared to other cell types such as dcndritie cells, these
`complexes can be recognised by receptors present on T lymphocytes, the T cell receptors.
`If appropriate recognition takes place then the T cell may deliver 'help' in the form of
`signals back to the B cell to stimulate proliferation and differentiation. This process
`therefore gives rise to a 'T-cell-dependent B-cell response'.
`Proliferation results in a clone of identical cells which can differentiate to form either
`plasma cells capable of secreting large amounts of soluble antibody of the same specificity
`as the original activated B cell, or memory cells which mount an accelerated immune
`response on re-exposure to the original antigen. The memory cell-mediated 'secondary'
`response results in the production of higher affinity antibodies due to hypermutation in
`the immunoglobulin gene loci followed by antigen driven selection, a process known as
`affinity maturation. This process of B cell clonal selection and the generation of antibody
`responses is covered in more detail in an excellent review (Rajewsky, 1996).
`
`1.3 Structure and function of antibodies
`
`Antibody molecules have two principal functions, firstly to bind to antigen and secondly
`to trigger its elimination from the body. Antibodies are therefore 'adaptor' molecules
`which have both the ability to bind to the antigen molecule and the ability to bind and
`
`HOSPIRA EX. 1029
`Page 11
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 11
`
`
`
`Preparation, structure and jinction of ' monoclonal antibodies
`
`Clonal
`selection
`
`Antigen
`driven
`proliferation
`
`Differentiation
`
`0
`Memory cells 0
`
`0 0 Plasma cells
`
`Figure 1.1 Differentiation of B cells to antibody-secreting plasma cells
`
`bring into action molecules of the effector system which can then remove the foreign
`material. For effective defence systems to operate it is essential that an individual is able
`to recognise a wide variety of foreign material and thus antibody molecules of many
`different binding specificities are required. However, in each case the antibody must also
`
`HOSPIRA EX. 1029
`Page 12
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 12
`
`
`
`Applications and Engieee;ing of A/fonocionai Ani.iboc!ies
`
`sin
`
`Figure 1.2 Representation of the organisation of protein chains of an IgG molecule
`
`be able to trigger the same effector systems. To achieve this the antibody molecule has
`evolved variable regions which vary in protein sequence and structure to accommodate
`the development of binding specificities to a wide range of different antigens, and con-
`stant regions which are largely the same in each antibody. The constant regions main-
`tain a common structure of the molecule and allow interaction with the effector systems
`such as complement binding and binding to Fc receptors on macrophagcs to activate
`phagocytosis.
`Antibody molecules are also known as immunoglobulins. The term 'immunoglobulin'
`applies to the antibody protein whether or not the binding specificity of the molecule is
`characterised, whereas an antibody is an antigen-specific immunoglobulin. In practice the
`two terms are usually used interchangeably.
`Higher mammals have five classes (isotypes) of immunoglobulin, termed IgG, 1gM,
`IgA, IgE and 1gD. The most abundant of these, which is used in most applications of anti-
`bodies, is the IgG class, the main class of antibody generated by the secondary immune
`response. The IgG molecule consists of four polypeptide chains, two heavy chains of
`approximately 50 kDa and two light chains of approximately 25 kDa (Figure 1.2). Each
`of these is divided into discretely folded structural domains of approximately 110 amino
`acids stabilised by an internal disulphide bond. These are linked together by short regions
`of comparatively flexible protein chain which allow movement of the domains relative to
`one another. Each light chain comprises a variable domain (VL) and a constant domain
`(CL), and each heavy chain a variable domain (Vii) and three constant domains (CH!, CH2
`and Cr13). The heavy and light chain variable domains associate to form the antigen-
`binding site. The IgG molecule thus has two antigen-binding sites and is capable of
`
`HOSPIRA EX. 1029
`Page 13
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 13
`
`
`
`Preparation, structure and tiinction of monoclonal antibodies
`
`F(ab)2
`
`JFc
`
`Figure 1.3 Domain structure of IgG, demonstrating the antigen-binding F(ab'), region and
`the Fc region with carbohydrate attached at the CH2 domain
`
`binding antigens divalently to allow high binding avidity (see Section 1.5). The heavy
`and light chains are linked by a disulphide bond between the CL and Cr-ti domains. A
`flexible region, known as the hinge, links the Cut and CH2 domains of each heavy chain
`and it is at this point that the two heavy chains are linked by disulphide bonds. The Cr-u
`domain is normally glycosylated and the Cr-n domains of each heavy chain associate with
`each other by non-covalent interactions.
`Much early work on the structure and function of antibodies made extensive use of
`proteolysis of the antibody protein and this has led to commonly used terminology for the
`different regions of the antibody (Figure 1.3). The area around the antibody hinge is more
`susceptible to proteolysis than the tightly folded domains and thus this is the point at
`which proteolytic cleavage usually takes place. Proteolysis above the disulphide bonds in
`the hinge region results in monovalent Fab (fragment antigen-binding) fragments which
`comprise light chain together with the N-terminal two domains of the heavy chain. The
`Fab fragment thus contains one functional antigen binding site. Proteolysis immediately
`below one or more of the hinge disulphide bonds results in the divalent F(ab')2 fragment.
`The Cr-n and CH3 domains together make up the Fe fragment (fragment crystalline)
`which contains the sites for binding effector molecules. These fragments were the starting
`points for crystallographic determinations of antibody structure, as the flexibility of the
`intact IgG molecule prevented crystallisation until fairly recently. Thus while there are
`more than 50 structures for Fab fragments in the literature and several for hinge deleted
`immunoglobulins (reviewed by Padlan, 1994) there is only one fully defined structure of
`an intact IgG (Harris et al., 1997).
`Crystallographic structure determinations of many Fab fragments and several Fe frag-
`ments of antibodies have revealed that the folded domains have similar overall structures.
`This structure comprises two stacked t3-sheets twisted into a characteristic fold, termed
`the immunoglobulin fold, and stabilised by a disulphide bond (Poljak et al., 1973). In the
`constant domains one sheet has three and the other four antiparallel beta strands, while in
`the variable domains there are nine strands.
`The primary sequences of several thousand antibody molecules are known, represent-
`ing the largest number of known sequences in one protein family. Analysis of these
`sequences has revealed that the variability of antibody variable domains is largely re-
`stricted to three 'hypervariable' regions in each of the heavy and light chain variable
`domains, with comparatively little variation in the intervening framework' regions (Kabat
`
`HOSPIRA EX. 1029
`Page 14
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 14
`
`
`
`
`
`Applications and Engineering o,fMonoclonal Antibodies
`
`el al., 1991). These hypervariable regions, also known as complen1entarity determining
`regions. or CDRs, form loops at the tip of the Fab structure making up an antigen-binding
`surface of approximately 2800 A2. Structures of several Fab:antigen complexes have
`confirmed that antigen binding takes place at this surface; the hypervariability of these
`loops is therefore responsible for the different binding specificities of different antibody
`molecules. The CDR loops vary in both amino acid sequence and length between anti-
`bodies. This results in the variety of antigen-binding surfaces, some of which contain
`grooves or clefts in the surface to accommodate antigen binding whereas others are
`comparatively smooth. In some cases three-dimensional structures have been determined
`for antibody fragments with and without bound antigen which show that the antigen-
`binding site is not always completely rigid and may move to allow tight antigen binding
`in an 'induced fit' mechanism (Bhat et al., 1990; Rini et at., 1992). In some cases this
`may also lead to some conformational changes at sites in the antibody further away from
`the antigen-binding region (Guddat et al., 1994).
`There are four different subclasses (or isotypes) of IgG, called IgGi, [gG2, IgG3 and
`IgG4 in humans and IgGl, IgG2a, IgG2b and lgG3 in mice. Although all within the IgG
`class, these subclasses vary to some extent in their structure and function. The subclasses
`have different heavy chains termed yl, y2, y3 and y4 in humans and each of these can use
`light chains from one of the two light chain isotypes, K or X. The structures of the
`subclasses vary in their pattern of disulphide bonding (Figure 1.4). The number of
`disulphide bonds between the heavy chains varies from two in human IgG 1 and IgG4 to
`15 in human IgG3. Also, the position of the disulphide bond between the heavy and light
`chains varies from between the CL and Ci-it region to between the CL and Vt-i/Cut
`interdomain region.
`The antibody effector functions are mediated through the constant regions. The effec-
`tor functions can be mediated through complement activation or by cellular interactions
`through specific Fe receptors expressed on a range of cell types, which can result in the
`process of antibody-dependent cellular cytotoxicity (ADCC). Complement activation via
`the classical complement cascade is initiated through binding of the IgG to the comple-
`ment component C1q. This is subject to conformational restraints which are only partially
`understood, and requires two molecules of IgG bound to an antigenic surface to bind Cl q
`efficiently. The binding site for Clq on IgG has been known for some time to be localised
`to the CH2 domain, with evidence that the N-terminal region of CH2 is important for
`binding C1q in the case of human IgGi (Morgan et al., 1995) and also that the C-terminal
`region of CH2 is required for efficient complement lysis (Tao et at., 1993; Greenwood
`et al., 1993). Mutagenesis studies have suggested that three amino acid residues on the
`surface of the CH2 domain of murine IgG2b, numbers 318, 320 and 322, are involved in
`the binding interaction (Duncan and Winter, 1988), although these residues are also
`present on antibodies which cannot activate complement, suggesting that these residues
`are necessary but not sufficient to activate complement. Mutagenesis studies with human
`IgGI have also suggested that C1q binding alone is not sufficient for complement activa-
`tion but that interactions of the IgG molecule with other steps in the complement pathway
`are also required (Tao et at., 1993).
`Fe receptors have been identified for all the classes of immunoglobulin, IgG, 1gM,
`lgD, IgE and I-A. Of these the best characterised are those for IgG and IgE. Three types
`of Fe receptors have been identified for IgG in humans; FcyRI (also known as CD64),
`FcyRII (also known as CDw32), and FcyRlII (also known as CD 16). These receptors are
`structurally related and distributed on various blood cell types. FcyRI is a high affinity
`receptor, capable of binding monomeric IgG, which plays a key role in ADCC and is
`
`HOSPIRA EX. 1029
`Page 15
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 15
`
`
`
`Preparation, structure and function of nionoclonal antibodies
`
`Human Ig
`
`IgGI
`
`IgG2
`
`TgG3
`
`IgG4
`
`Murine IgG
`
`IgGi
`
`IgG2a
`
`tgG2b
`
`1lgG3
`
`Figure 1.4 Variation in the disulphide bonds between heavy chains in IgG subclasses
`
`found on macrophages, monocytes and neutrophils. FoyRll and FcyRIII are low affinity
`receptors which bind aggregated IgG. FcyRlI is found on most leukocytes including
`monocytes, macrophages and neutrophils and FcyRlII is found on neutrophils, macrophages
`and NK cells. These low affinity receptors are also capable of eliciting ADCC and
`phagocytosis. The binding of a single antibody to Fc receptor is reversible and does not
`elicit a response. If several antibodies are clustered together Fc receptor clustering will
`take place and a cellular response will be elicited (Figure 1.5). All three Fey receptors
`recognise sites on the lower hinge/CH2 domain of IgG but not in an identical fashion. In
`particular, the sequence of heavy chain residues at positions 234-237 is involved in the
`Fc'yRl receptor binding site (Duncan et al., 1988; Sarmay et al., 1992).
`The ability to elicit effector functions varies between isotypes of IgG (Table 1. 1), and
`this may reflect important differences in the functions of the individual isotypes. For
`example, human IgGi and IgG3 are highly active isotypes with respect to complement
`activation and elicitation of ADCC responses whereas IgG2 and IgG4 are relatively
`inactive, being only poorly able to activate complement through Clq binding with little
`binding to Fc receptors.
`The presence of carbohydrate attached to the CH2 domain has been shown to be
`required for both Fc receptor binding and complement activation (Tao and Morrison,
`
`HOSPIRA EX. 1029
`Page 16
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 16
`
`
`
`8
`
`Applications and Engineering oJ'Monoclonal Antibodies
`
`Fc receptor bearing
`effector cell
`
`Fc receptor clustering
`
`Signalling
`
`Cellular responses
`(phagocytosis, exocytosis)
`Figure 1.5 Fc receptor clustering due to multiple antibody molecules binding to antigen
`leading to cellular responses
`
`Table 1.1 Effector functions of human and mouse lgG subclasses
`
`Human IgG
`
`Mouse IgG
`
`IgGI
`
`IgG2
`
`1gG3
`
`IgG4
`
`IgGi
`
`IgG2a
`
`IgG2b
`
`IgG3
`
`+++
`
`+
`
`++
`
`-
`
`+
`
`++
`
`+++
`
`++
`
`+++
`
`-
`
`+++
`
`+
`
`++
`
`+
`
`++
`
`-
`
`++
`
`++
`
`-
`
`-
`
`-
`
`+
`
`++
`
`+
`
`+
`
`-
`
`±
`
`+
`
`-
`
`Complement
`activation
`
`Human
`Fc'yRI (CD64)
`
`Human
`FcyR1I (CDw32)
`
`Human
`FcyR1II (CD 16)
`
`HOSPIRA EX. 1029
`Page 17
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 17
`
`
`
`
`
`Preparation, structure and function of inonoclona! antibodies
`
`Table 1.2 Human immunoglobuiins
`
`Antibody
`
`Approximate
`molecular weight
`(kDa)
`
`Heavy chain
`
`Approximate concentration
`in normal human serum
`(mg/ml)
`
`IgGi
`IgG2
`1gG3
`1gG4
`1gM
`IgAl
`1-A2
`1gD
`IgE
`
`146
`146
`170
`146
`970
`160
`160
`200
`200
`
`yl
`y2
`13
`74
`is
`al
`cr2
`6
`S
`
`9
`3
`
`0.5
`1.5
`3
`0.5
`0.03
`0.0001
`
`1989). The role of the carbohydrate is not fully understood but it is believed to he
`important in maintaining the tertiary structure and disposition of the CH2 domains. IgG
`without carbohydrate has been shown to have an altered conformation in the lower hinge
`region and may result in disruption of the interaction sites (Lund et al., 1990).
`IgG represents approximately 70-75% of the total imrnunoglobulin in human serum,
`the remainder being molecules of the other classes mentioned above, 1gM, IgA, IgD and
`IgE. Although the IgG structure is a good general model for antibody structure, these
`irnmunoglobulin classes differ in their structure, reflecting differences in their normal
`functions. The different classes of immunoglobulin use light chains of the same type (is
`or ?) and thus differ predominantly in their heavy chains (Table 1.2).
`1gM is the predominant antibody raised in primary responses to many antigens and
`represents approximately 10% of the immunoglobulin in human serum. 1gM antibodies
`have a pentameric structure of the basic four chain unit such that a total of ten antigen-
`binding sites per molecule are present (Figure 1.6). The j.t heavy chain does not have a
`hinge region but has an extra constant region domain, which is inserted between the CHI
`and CH2 domains in the analogous IgG structure, such that the CH4 domain of 1gM is
`analogous to the CH3 domain of IgG. There is also an extra tail of 19 amino acids on the
`heavy chain involved in polymer assembly. The heavy chains are held together by
`disulphide bonds between CH3 domains and there is also an extra polypeptide chain, the
`J chain, whose role is not fully understood although it may assist the process of assembly
`of the pentameric 1gM molecule. The j.t heavy chain is also more heavily glycosylated
`than the y heavy chain, with sugar attached at five glycosylation sites. Although it is
`presented as a pentameric structure in Figure 1.6, this is not always the case and the
`occurrence of 1gM molecules with different numbers of binding units is now well estab-
`lished. The production of hexamers and tetramers has been reported (Eskeland and
`Christensen, 1975) and in some cases hexamers may be a predominant structure (Cattaneo
`and Neuberger, 1987; Davis et al., 1988). This multivalent structure provides the 1gM
`molecule with a high binding avidity, to help compensate for the relatively low affinity of
`each binding site. Another consequence of the multivalent structure is the ability of one
`molecule of 1gM to activate complement, which is thought to occur via binding to the
`1gM CH3 domain. Therefore I9 molecules are well adapted for their role in the primary
`immune response.
`
`HOSPIRA EX. 1029
`Page 18
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 18
`
`
`
`
`
`
`
`10
`
`Applications and Engineering of iklonoc/onaiAntibodies
`
`1gM
`
`•
`
`ii
`
`IgG
`
`II
`
`1gB
`
`IgE
`
`I I0
`
`J chain
`
`Secretory component
`
`Serum IgA
`
`Secretory IgA
`
`Figure 1.6 Organisation of protein chains in the immunoglobulin classes
`
`There are two isotypes of IgA present in humans (IgAl and IgA2) which together
`make up 15-20% of the total serum immunoglobulin. In addition, a form of IgA, termed
`secretory IgA or slgA, is the predominant immunoglobulin present in external secretions
`such as saliva, milk, tears, genitourinary secretions and tracheobronchial secretions (Tomasi,
`1992). The mucosal surfaces bathed in these secretions are a major site of exposure to the
`environment and thus secretory IgA is an important defence mechanism against invading
`organisms. In terms of quantity, slgA is the major type of Ig produced, with estimates of
`2g per day being produced in humans, more than other Ig forms. The IgA present in
`serum is largely monomeric and consists of two heavy chains and two light chains
`assembled in a similar manner to IgG. IgAl is the predominant isotype, making up
`approximately 90% of the serum IgA. IgA heavy chains have three constant domains
`with IgAl having a longer hinge region than IgA2. In addition, IgA heavy chains also
`
`HOSPIRA EX. 1029
`Page 19
`
`PFIZER and SAMSUNG v. GENENTECH
`IPR2017-01489
`PFIZER EX. 1695, Page 19
`
`
`
`Preparation, structure and/unction of monoclonal antibodies
`
`have a similar tail to 1gM which allows polymerisation, and binding of the J chain to the
`a chain, through a C-terminal cysteinc residue in a similar manner to 1gM. sIgA exists
`predominantly in dirneric form with a molecular weight of approximately 385,000. sigA
`also contains a secretory component which is an extra polypeptide chain of approxim-
`ately 70 kDa, synthesised by epithelial cells and not the plasma cells. The secretory
`component is part of the receptor involved in transport of IgA into mucosal secretions
`which is cleaved during transport across endothelial surfaces to release the 70 kDa sec-
`retory component. This then becomes bound to the IgA by disulphide bonds (Fallgreen-
`Gebauer et al., 1993). sigA is more stable to proteolytic attack than the serum form and
`pa