-
`L
`L'k'
`|-_:Fi§-J.FEYIE‘_-
`'
`IN‘IIIII‘EIII
`W
`"
`"
`-.
`
`_
`
`.
`
`.
`
`3E}??? ULLUDE‘IE‘H oclonal
`
`
`
`I
`
`_
`LWIF'EQE'I—Y‘"
`-|
`IllI-‘|:l
`h
`I
`|
`I.
`I
`
`I
`
`.
`|
`
`I:
`
`Antibodies
`
`Pete Gagnon
`
`Pfizer Ex. 1022
`
`Page 1 of 24
`
`Pfizer Ex. 1022
`Page 1 of 24
`
`

`

`Purification Tools
`for Monoclonal
`Antibodies
`
`by Pete Gagnon
`//
`
`Validated Biosystems
`
`Pfizer Ex. 1022
`Page 2 of 24
`
`

`

`II
`
`Purification Tools for Monoclonal Antibodies
`
`Purification Tools for Monoclonal Antibodies
`by Pete Gagnon
`
`Copyright © 1996 Peter S. Gagnon
`
`All rights reserved. No part of this publication may be repro(cid:173)
`?uced or tran.smitted in any form or by any means-electron(cid:173)
`ic ~r mecha?1cal, including photocopying, recording, taping,
`or mformat10n storage and retrieval systems-without writ(cid:173)
`ten permission of the author and publisher.
`
`First published in 1996 by Validated Biosystems, Inc.,
`5800 N. Kolb Road, Suite 5127, Tucson, AZ 85750
`
`Library of Congess Catalog Card Number
`96-90448
`
`International Standard Book Number
`0-9653515-9-9
`
`Registered names, trademarks, etc. used in this book even
`when not specifically marked as such, are not to be c~nsid­
`ered unprotected by law.
`
`The information contained within this book was obtained
`from sources believed to be reliable. However, while every
`~ffort has bee~ ~ade to ensure its accuracy, no responsibil(cid:173)
`ity ~or loss or m1ury to any person acting or refraining from
`act10n as a result of the information contained herein can
`be accepted by the the publisher, author, or editors.
`In vie':" of ongoing research, equipment modifications,
`changes m governmental regulations, and the constant flow
`of i~formation relating to the use of experimental reagents,
`eqmpment, and devices, readers are urged to review and
`evaluate manufacturer's information for each chemical or
`?e~ce,.taking particular notice of changes in instructions or
`md1cat1on of usage, and for added warnings and precautions.
`
`r;t
`I Yt. 'i.S
`.G:3s
`/111
`
`Foreword
`
`"All truths are easy to understand once they are
`discovered; the point is to discover them"
`-Galileo
`
`O ut of all the biotechnology products in develop(cid:173)
`
`ment, in clinical trials, and on the market, mono(cid:173)
`clonal antibodies are the most numerous. They con(cid:173)
`tinue to be at the forefront of every new field of
`endeavor, including human and veterinary healthcare,
`agriculture, forensics, and environmental monitoring.
`Even as competing product classes arise, technology
`improvements and expanding applications guarantee
`that monoclonals will remain a vital force in the con(cid:173)
`tinuing growth of the industry.
`This places enormous pressure on purification
`process designers. In today's cost conscious environ(cid:173)
`ment, economical purification schemes that support
`the requisite product quality and meet regulatory
`expectations are critical for success. Agressive time-to(cid:173)
`market calendars demand that they be developed ever
`more rapidly. A broad range of purification tools has
`evolved to meet these needs but practical guidance
`concerning their use has lagged.
`This book fills that void. It provides a major re(cid:173)
`source to the downstream processing community by
`integrating comprehensive information on purification
`tools with the specific features and requirements of
`monoclonal antibodies. Knowledge based on years of
`hands-on experience provides process designers from
`both academia and industry with valuable insights that
`will expedite development and assure them that they
`have have taken full advantage of the best that purifi(cid:173)
`cation technology has to offer.
`Gail K. Sofer,
`Director of International Validation
`Pharmacia Biotech, Inc.
`
`Pfizer Ex. 1022
`Page 3 of 24
`
`

`

`iv
`
`Purification Tools for Monoclonal Antibodies
`
`Preface
`
`''There is more to learn by climbing the same
`mountain a hundred times, than by climbing
`a hundred different mountains."
`-Richard Nelson
`
`T he main reason I wrote this book is that I find
`
`monoclonal purification utterly fascintating, and
`I wanted to share that fascination in a way that would
`let others enjoy it as much as I. Certainly, the more
`enjoyable you find a task, the more likely you are to
`be successful. My hope is to facilitate that success by
`removing the obstacles that commonly prevent peo(cid:173)
`ple from fulfilling their monoclonal purification goals.
`To identify those obstacles, I relied on the purifica(cid:173)
`tion questions my clients have asked me over the past
`decade. What are the options? How does a particular
`method work? What are its strengths? What are its
`weaknesses? How do I exploit it to its greatest benefit
`while minimizing the influence of its limitations? How
`does it fit with other methods? Where can I find more
`information?" Every aspect of this book has been
`designed to answer those questions.
`Beyond content and organization, people requested
`a practical hands-on approach; one that integrates the
`unique characteristics of monoclonal antibodies with
`the specific capabilities and limitations of the various
`purification methods. There is an immense difference,
`for example, between ion exchange chromatography in
`general and ion exchange as practiced within the mol(cid:173)
`ecular and regulatory constraints of monoclonal
`purification. There are opportunities and obstacles
`that don't apply to any other group of proteins. This
`applies to every technique.
`In the same spirit, it was frequently requested that
`examples used to illustrate key concepts derive as
`much as possible from industrial applications. I con(cid:173)
`sider this to be one of the book's strongest features, but
`
`Pfizer Ex. 1022
`Page 4 of 24
`
`

`

`vi
`
`Purification Tools for Monoclonal Antibodies
`
`Preface
`
`vii
`
`Volenec at Cytel, Peter Cartier at Rohm and Haas,
`Brandon Price at Microbiological Associates, David
`Buck and Diether Rechtenwald at Am Cell, Paul Smith
`at BioAffinity Systems, Dennis Arvanitis at Dow
`Chemical, Scott Fulton and Dan Freymeyer at
`PerSeptive Biosystems, Peter Grandics and Bruce
`Hoffman at Sterogene Bioseparations, Joel Benner at
`Connective Therapeutics, Anne Moschella at Biopro(cid:173)
`cessing Ltd., Joe Machamer at Bio-Rad, and Andrea
`Knight at Genetics Institute. Their many contribu(cid:173)
`tions over the years have been much appreciated.
`Finally, I offer my deepest gratitude to my clients.
`It's been the experience provided by working with
`them and hundreds of their products that has given me
`the orbital perspective necessary to write this book. I
`now turn this resource over to you, and I look forward
`to seeing where you take the field from here.
`
`PSG
`Tucson, Arizona
`September 27, 1996
`
`it involved a compromise. Much of the information is
`proprietary and permission to use the examples was
`contingent on omitting sensitive details. However
`enough information is included to let you judge for
`yourself if an example applies to your specific circum(cid:173)
`stances, and enough so that you can adapt it.
`Purification-product names are omitted throughout
`the text for another reason. I wanted to avoid inadver(cid:173)
`tent bias favoring or disfavoring any particular suppli(cid:173)
`er, and especially to avoid any implication that acer(cid:173)
`tain set of results require the use of a specific product.
`Assume that the examples are representative of all
`closely related products. Where a specific product is
`needed, it is identified. Where a specific product class
`is required, I identified every product I was aware of.
`I was also asked to keep technical explanations as
`intuitive as possible. Equations have a unique ability to
`integrate the influence of multiple parameters into a
`single expression, but they can be difficult to follow.
`Narrative suffers from linearity but lends itself to
`expressing ideas in more familiar terms. I chose the lat(cid:173)
`ter. This will hopefully provide a larger group of prac(cid:173)
`titioners with an understanding that they can use to
`meet challenges creatively instead of reflexively.
`Numerous people contributed to the development
`of this book in other ways. At the top of the list, Deb
`Neff and all the folks at Becton Dickinson Immuno(cid:173)
`cytometry Systems: Noel Warner, Vernon Oi, Ken
`Davis, Neal Weinstein, Bill Godfrey, Don Ladd,
`George Herrel, Joe Link, the entire Product Devel(cid:173)
`opment, and Manufacturing staffs. I can never thank
`them enough for their indulgence in providing me
`unlimited access to antibodies and laboratory facilities
`to develop ideas into practical methodology.
`Many thanks to Gail Sofer, Jan-Christer Janson,
`Lars Hagel, and Makonnen Belew; some of my oldest
`and most valued friends in the biotech business. Also
`to Duncan Low, Peter Moore, Tim Hooper, Les Bead(cid:173)
`ling, Anne Barry, Al Williams, Lars-Johan Larsson,
`Torgny Lindback, Eric Grund, Ursula Snow and many
`other good friends at Pharmacia. Likewise to Jerry
`
`Pfizer Ex. 1022
`Page 5 of 24
`
`

`

`viii
`
`Purification Tools for Monoclonal Antibodies
`
`Table of Contents
`
`Foreword .................................................. iii
`
`Preface, .................................................... v
`
`Chapter 1. Introduction .................................... 1
`What this bookis about ......................... 1
`How this book is organized ...................... 2
`In practice . . . . . . ............................... 3
`
`Chapter 2. Precipitation .................................... 5
`With inorganic salts ................... . ......... 5
`Mechanism .... . ............................. 5
`Attributes .................................. . . 8
`Limitations .................................. 9
`Method development ....................... . 13
`With polyethylene glycol ...................... . 15
`Mechanism ............ . .................... 15
`Attributes .................................. 16
`Limitations ........ . ............... . ....... . 17
`Method development ........................ 18
`By electrolyte depletion ........................ 19
`Mechanism ............................. .' ... 19
`Attributes .................................. 19
`Limitations ................................. 20
`With octanoic acid ..... . ....................... 20
`Mechanism ............................ . .... 20
`Attributes ............................ . ..... 21
`Limitations ................................. 22
`Method development ........................ 24
`With ethacridine .............................. 24
`Mechanism ................................. 24
`Attributes .................................. 24
`Limitations ........................ . ........ 26
`Methoddevelopment ... . .................... 26
`In perspective ................................. 2 7
`Recommended reading ......................... 28
`References ... . ................................ 2 8
`
`Pfizer Ex. 1022
`Page 6 of 24
`
`

`

`x
`
`Purification Tools for Monoclonal Antibodies
`
`T
`
`Table of Contents
`
`xi
`
`Chapter 3. Size Exclusion Chromatography .............. 3 3
`Mechanism ................................... 3 3
`Attributes ..................................... 38
`Limitations ................................... 42
`Method development .......................... 50
`I
`.
`n perspectrve ................................. 5 3
`Recommended reading ......................... 54
`References .................................... 54
`
`Chapter 4. Ion Exchange Chromatography ............... 57
`Mechanism ................................... 57
`Attributes ..................................... 62
`Limitations ................................... 67
`Method development .......................... 74
`In perspective ................................. 81
`· Recommended reading ......................... 82
`References .................................... 82
`
`Chapter 5. Hydroxyapatite Chromatography ............ 87
`Mechanism ................................... 87
`Attributes ..................................... 90
`L' . .
`1m1tatrons ................................... 94
`Method development .......................... 97
`I
`.
`n perspectrve ................................. 99
`Recommended reading ........................ 100
`References ................................... 100
`
`Chapter 6. Hydrophobic Interaction Chromatograpy . ... 103
`Mechanism .................................. 1O3
`Attributes .................................... 110
`Limitations .................................. 113
`Method development ......................... 120
`In perspective ................................ 123
`Recommended reading ........................ 123
`References ................................... 123
`
`Chapter 7. Immobilized Metal
`Affinity Chromatography .................... 127
`Mechanism .................................. 12 7
`
`Attributes ..................................... 130
`Liinitations .................................. 132
`Method development ......................... 13 5
`In perspective ................................ 13 6
`Recommended reading ........................ 13 7
`References ................................... 13 7
`
`Chapter 8. Other Physicochemically-Based
`Chromatography Methods .................. 139
`Hydrophilic interaction chromatography ........ 139
`Mechanism ................................ 13 9
`Attributes ................................. 141
`Limitations ................................ 142
`Method development ....................... i44
`In perspective .............................. 145
`Euglobulin adsorption chromatography ......... 145
`Thiophilic adsorption chromatography ......... 146
`Dye ligand chromatography ................... 148
`ABx chromatography ......................... 150
`Immobilized boronate chromatography ......... 150
`References ................................... 151
`
`Chapter 9. Protein A Affinity Chromatography ......... 15 5
`Mechanism ............................... : .. 15 5
`Attributes .................................... 163
`Limitations .................................. 168
`Method development ......................... 186
`In perspective ................................ 189
`Recommended reading ........................ 190
`References ................................... 190
`
`Chapter 10. Other Biological Affinity Ligands .......... . 199
`Protein G .................................... 199
`Mechanism ................................ 199
`Attributes ................................. 200
`Limitations ................................ 201
`In perspective .............................. 202
`Protein A/G hybrids .......................... 202
`Other bacterial ligands ........................ 202
`Immunoaffinity .............................. 203
`
`Pfizer Ex. 1022
`Page 7 of 24
`
`

`

`xii
`
`Purification Tools for Monoclonal Antibodies
`
`Purification 1001s tor Monoclonal Ant10oa1es
`
`XIII
`
`Peptide/oligonucloetide affinity ................ 205
`Lectin affinity ................................ 205
`Mechanism ................................ 205
`Attributes ................................. 205
`Limitations ................................ 206
`Alkaline polypeptide ligands ................... 207
`DNA affinity ................................. 207
`Carbohydrate affinity ......................... 208
`Recommended reading ........................ 208
`References ................................... 208
`
`Appendix I. Foundation protocols ....................... 213
`Octanoic acid precipitation followed by
`ammonium sulfate precipitation ............ 213
`Ethacridine precipitation followed by
`size exclusion chromatography ............. 215
`Ammonium sulfate precipitation followed
`by anion exchange chromatography . ~ ....... 217
`Polyethylene glycol precipitation followed by
`euglobulin adsorption chromatography ...... 218
`,
`Immobilized metal affinity followed by
`hydrophobic interaction chromatography ... 219
`Protein A affinity chromatography followed
`by cation exchange chromatography ......... 221
`Ion exchange chromatography with
`hydrophobic interaction chromatography ... 224
`
`Appendix II. Sample Preparation ........................ 225
`
`Index .................................................. 229
`
`Abbreviations ........... ~ .............................. 247
`
`Endnotes .............................................. 251
`
`About the author ...................................... 253
`
`Pfizer Ex. 1022
`Page 8 of 24
`
`

`

`56
`
`Purification Tools for Monoclonal Antibodies
`
`5 6. M. J anado et al, 197 6, J. Biochem., 80 69
`57. L. Politi et al, 1983,J. Chromatogr., 267 403
`58. A. Galkin et al, 1984, Anal. Biochem., 142 252
`59. T. Fujita et al, 1980,J. Biochem., 87 89
`60. J.-C. Janson andJ.-A. Jonsson, 1989, in Protein Purifiction;
`Principles, High resolution Methods, and Applications, G.(cid:173)
`C. Janson and L. Ryden, eds.), p.35, VCH, Cambridge
`61. B. Gelotte, 1960,J. Chromatogr., 3 330
`62. TosoHaas Technical Report #10
`63. C. Middaugh and G. Litman, 1977,J. Biol. Chem.,
`252 8002
`64. C. Tanford, 1968, Adv. Protein Chem., 23 121
`65. W. Jencks, 1969, Catalysis in Chemistry and Enzymology,
`p.323, McGraw-Hill, New York
`66. S. Timasheff and G. Fasman, 1969, Structure and Stability
`of Biological Macromolecules, Marcel Dekker, New York
`67. P. Clezardin et al, 1986,J. Chromatogr., 354 425
`68. P. Clezardin et al, 1986, J. Chromatogr., 358 209
`69. ]. Porath and N. Ui, 1964, Biochim. Biophys. Acta, 90 324
`70. U.-B. Hansson and E. Nilsson, 1973,J. Immunol. Met.,
`2 221
`71. E. Kabat and M. Mayer, Experimental Immunochemistry,
`2nd Ed., p. 264, Charles C. Thomas Publisher, Springfield
`72. J. Porath and B. Olin, 1983, Biochemistry, 22 1621
`73. Y.Katoetal, 1981,J. Chromatogr.,205185
`74. J.-C. Janson and P. Hedman, 1982, in Advances in
`Biochemical Engineering (A. Fietcher, ed.) Vol. 6, p.43,
`Springer-Verlag, New York
`7 5. P. Bristow and J. Knox, 1977, Chromatographia, 10 2 79
`76. L. Hagel, 1985,J. Chromatogr., 324 422
`77. T. Andersson and L. Hagel, 1984, Anal. Biochem., 141 461
`78. C. Middaugh et al, 1978, Clin. Lab. Immunol., 1 141
`79. C. Middaugh et al, 1980, J. Biochem., 25 5 65 3 2
`80. C. Middaugh et al, 1978, Proc. Nat. Acad. Sci. USA, 75 3440
`81. Y. Kato et al, 1981, J. Chromatogr., 206 13 5
`82. Y. Kato et al, 1981,J. Chromatogr., 208 71
`83. -, 1991, Gel Filtration, Principles and Methods, 5th edi-
`tion, Pharmacia, Uppsala
`84. M. Carlsson et al, 1985,J. Immunol. Met., 79 89
`85. B. Malm, 1987,J. Immunol. Met., 104103
`86. G. Perry et al, 1984, Prep. Biochem.,14(5) 431
`87. S. Burchiel et al, 1984,J. Immunol. Met., 69 33
`88. F.-M. Chen et al, 1988, J. Chromatogr., 444 153
`89. A. Wichman and H. Borg, 1977, Biochim. Biophys. Acta,
`490 363
`90. F. Franek, 1986,Met. Enzymol., 121631
`
`Chapter 4
`
`Ion Exchange
`Chromatography
`
`"Don't fight forces, use them!"
`-R. Buckminster Fuller
`
`I on exchange chromatography (IEC) was first suc(cid:173)
`
`cessfully applied to fractionation of plasma proteins
`in the mid-1950s.[1-3] By the 1970s it was firmly
`established as a standard method for purification of
`polyclonal antibodies.[4-10] This provided a founda(cid:173)
`tion for its application to monoclonals and it evolved
`rapidly to become the dominant method in the
`field.[11] Advances in support media, including both
`chromatography and filtration formats, continue to
`expand its capabilities and flexibility.
`Charged residues on protein surfaces include the
`side groups of amino acids, the a-amino and a-car(cid:173)
`boxyl termini of the polypeptide chains, and sialic acid
`residues on glycoproteins. These residues are ampho(cid:173)
`teric, making the sign and net charge on proteins a
`function of pH. (Table 4.l).[12-15] The pH at which a
`protein's positive charge balances its negative charge is
`its isoelectric point (pl). At pH values above their pls,
`proteins become progressively electronegative. Below
`their pis, they become progressively electropositive.
`The traditional ion exchange model suggests that
`pl is the sole determinant of a protein's retention
`behavior.[16,17] A protein should bind to an anion
`exchanger at pH values above its pl and to a cation
`exchanger below its pl. Retention should diminish
`rapidly to zero on both as pH approaches the pl. Anion
`exchange selectivity should precisely mirror selectivity
`on cation exchange.
`
`Mechanism
`
`Pfizer Ex. 1022
`Page 9 of 24
`
`

`

`58
`
`Purification Tools for Monoclonal Antibodies
`
`Chapter 4. Ion Exchange Chromatography
`
`59
`
`Table 4.1. pKs for charged amino
`acids. The altered ranges in proteins
`reflect the influence of local sur-
`face characteristics. Proximity of
`unlike charges elevates the pKs of
`positively charged residues and
`lowers the pKs of negative groups.
`Proximity of like charges or hydro-
`phobic regions have the opposite
`effect. The protein-value for argi-
`nine is predicted. All others were
`measured experimentally.[12-15]
`
`Residue
`
`a-amino
`arginyl
`histidyl
`lysyl
`a-carboxyl
`aspartyl
`glutamyl
`
`pKin
`amino acid
`
`8.8-10.8
`12.5
`6.0
`10.8
`1.8-2.6
`3.9
`4.3
`
`pKin
`protein
`
`6.8-7.9
`;:::12
`6.4-7.4
`5.9-10.4
`3.5-4.3
`4.0-7.3
`4.0-7.3
`
`Figure 4.1. Preferential orientation
`on a cation exchanger as a
`function of pH.
`HIS: histidine.
`ASP: aspartic acid.
`PHE: phenylalanine.
`TYR: tyrosine.
`ARG: arginine.
`LYS: lysine.
`GLU: glutamic acid.
`
`pH5.5
`
`pH 7.5
`
`Rigorous studies show that less than 25% of pro(cid:173)
`teins accurately fulfill the model's predictions.[16-23]
`Some bind a full pH unit below their pl on anion
`exchangers and a full unit above on cation exchangers.
`This indicates that positive and negative sites dominate
`different portions of the molecule. Others fail to bind a
`full unit above their pl on anion exchangers and/or a
`full unit below on cation exchangers. This indicates
`that negative and positive charges are mixed in the
`binding regions, partially neutralizing their respective
`interactions with the exchanger. lsoproteins with dif(cid:173)
`ferent primary sequences but identical pls exhibit dif(cid:173)
`ferent retention properties, while among antibodies,
`differentially charged glycosylation isoforms of a single
`monoclonal usually elute together in a single peak.
`These departures don't indicate that the model is
`fundamentally wrong, merely that it's too general. The
`most important qualification is the distinction between
`the net charge of the protein as a whole and the charge
`characteristics of its actual binding site.[19] The bind(cid:173)
`ing site may include one or more charged residues,
`even multiple areas of a protein, but usually not all of
`its charged residues. An important corollary is that the
`composition of the binding site at one pH may be very
`different from its composition at another (Figure 4.1).
`For example, a histidyl-rich site may dominate bind(cid:173)
`ing on a cation exchanger at pH 5 .5. Retention of the
`same protein at pH 7.5 may rely on a lesser number of
`untitrated residues on the opposite side of the pro-
`
`tein-residues that didn't even contribute to retention
`at pH 5.5. By extension, the respective binding sites of
`a given protein are likely to be different for anion and
`cation exchangers.
`These qualifications have important ramifications
`for selectivity. The contaminant spectrum with which
`an antibody is associated on a cation exchanger is
`almost certain to be different from the spectrum on
`anion exchange. Both are worth exploring in their own
`rights. Even on the same exchanger, altering pH may
`significantly change the spectrum of contaminants
`coeluting with the product. This isn't to suggest that
`multistep purification processes will routinely benefit
`from having 2 steps on the same exchanger. It does
`mean that it's worth exploring a range of pH values to
`determine which provides the best complement to
`other candidate methods in the process.
`Bound proteins can be eluted by altering pH, by
`addition of competing ions, or combinations of the 2.
`Each approach gives different selectivity. Achieving
`elution solely by altering pH at low ionic strength is
`the basis of a technique called chromatofocus(cid:173)
`ing. [17 ,24-28] Proteins elute near their pls. This turns
`out to be more of a liability than an asset with mono-
`
`Pfizer Ex. 1022
`Page 10 of 24
`
`

`

`60
`
`Purification Tools for Monoclonal Antibodies
`
`Chapter 4. Ion Exchange Chromatography
`
`61
`
`Figure 4.2. Analytical chromatofo(cid:173)
`cusing profile of a mouse lgG3
`monoclonal in ascites. pH range
`9-6. The column was eluted with a
`sodium chloride gradient following
`the pH elution. The monoclonal
`profile was determined with a
`separate injection of protein
`A-purified antibody.
`
`--,,-
`~
`
`Figure 4.3. Salt-mediated glyco(cid:173)
`form fractionation as a function of
`pH. Profile A was run at pH 8.5, 8
`at pH 7.0, and Cat pH 5.5. The pl of
`this mouse lgG 1 was 6.4-6.7. Mon-
`oclonal profiles were determined
`by injecting protein A-purified
`antibody in a separate
`series of runs.
`
`A280
`
`clonals because post-translational glycosylation iso(cid:173)
`forms elute in multiple peaks (Figure 4.2).[29,30] Gly(cid:173)
`coform fractionation is also observed with salt elution
`if the pH is near an antibody's pl (Figure 4.3). Mono(cid:173)
`clonals otherwise tend to elute in single peaks.
`Besides controlling selectivity with gradient slope, it
`can be altered by employing different eluting salts. Ions
`of higher valencies exert more effect on retention than
`monovalent ions, but even ions of the same valency
`exert different effects.[16-20,31-35] Their relative abil(cid:173)
`ities in this regard correlate with their respective rank-
`
`Table 4.2. The Hofmeister series of
`lyotropic and chaotropic ions. The
`directions of the arrows indicate
`increasing effectivity of
`eluting ions.
`
`Anions
`
`Cations
`
`Figure 4.4. Titration curves for
`common ion exchange ligands.
`QAE indicates quarternary amino(cid:173)
`ethyl, DEAE indicates Diethyl(cid:173)
`aminoethyl, CM indicates car(cid:173)
`boxymethy, and SP indicates
`sulfopropyl. The stability of the
`QAE and SP curves over a wide
`pH range demonstrates why they
`are called "strong" ion exchang(cid:173)
`ers. This resistance to titration
`translates into level protein bind(cid:173)
`ing capacity over a wide pH range.
`It has no direct significance with
`respect to selectivity.
`
`Ba2+ Ca2+ Mg2+ Li+ Cs+ Na+ K+ Rb+ NH4+
`
`12.0
`
`mEq HCI----.-
`
`mEq NaOH---11-.-
`
`ings in the Hofmeister series (Table 4.2).[35,36]
`Allosteric interactions with specific ions may alter pro(cid:173)
`tein conformation sufficiently to redefine the composi(cid:173)
`tion of the binding site. [20,3 7-40]
`On a practical level, it's not worthwhile to screen a
`wide range of eluting salts. Differences in displacing
`strengths only alter the virtual slope of an elution gra(cid:173)
`dient, and allosteric effects are unpredictable.[17,40]
`Consequently, sodium chloride is used to the virtual
`exclusion of all other candidates.
`Differences among ion exchangers also contribute
`to selectivity.[31] The titration characteristics of the
`ligand determine an exchanger's fundamental charge
`properties (Figure 4.4). Variations in the spacer arm
`contribute to the hydrophobicity of the support overall
`and are the primary determinant of ligand accessibili(cid:173)
`ty. This is why tentacle-type supports provide differ(cid:173)
`ent selectivities than short-spacer supports-the ligand
`has 3-dimensional access to the proteins. [ 41] The sup-
`
`Pfizer Ex. 1022
`Page 11 of 24
`
`

`

`62
`
`Purification Tools for Monoclonal Antibodies
`
`Chapter 4. Ion Exchange Chromatography
`
`63
`
`Attributes
`
`purification
`performance
`
`removal of nonspecific
`antibody
`
`DNA removal
`
`port matrix itself exerts an influence by way of both its
`pore size distribution and surface hydrophobicity. All
`other features being equal, ion exchangers with
`stronger hydrophobicities bind a broader spectrum of
`proteins and bind them more strongly than less
`hydrophobic exchangers.
`IEC is applicable to all monoclonal antibodies,
`regardless of class, species of origin, or production
`method (Figure 4.5).[11,42-63] This is not to say that
`all antibodies are equally suited to preparative fraction(cid:173)
`ation by both exchangers. Some IgGs are too basic to
`support high binding capacities on anion exchangers
`within reasonable pH limits (~8.6). These particularly
`include IgG3s. Acidic antibodies sometimes require
`excessively low pH (<4.5) for high capacity binding on
`cation exchangers.
`Ion exchangers provide 60-95% purity from raw
`production media. Some reports suggest that 1-step
`purifications are feasible, but this is neither representa(cid:173)
`tive nor realistic. These suggestions ignore the real-life
`requirements of commercial purification. High single(cid:173)
`step purity requires narrow peak cutting and the
`reduced recovery is likely to be more costly than the
`addition of a supporting purification step. Such sugges(cid:173)
`tions also fail to acknowledge that even the most robust
`methods are vulnerable to external process variations.
`IEC supports 50-75% removal of host or media
`supplement-derived polyclonal antibody. Combined
`reduction by both exchangers often exceeds 90%.
`IEC usually supports 3-5 logs removal of DNA. It
`binds strongly to anion exchangers by its negatively
`charged phosphate moieties.[42,61-65] The same
`groups cause it to be repelled and pass unretained
`through cation exchangers. An important exception to
`the DNA reduction capabilities of both methods
`occurs when DNA is complexed to the product. DNA
`is known to form stable charge complexes with proteins
`and has even been used as an "affinity" ligand for chro(cid:173)
`matographic purification of antibodies.[66-69] The
`antibodies most frequently and severely affected are
`IgMs and strongly basic IgGs. The low ionic strength
`
`Figure 4.5. Paired anion and cation
`exchange elution characteristics of
`17 monoclonal antibodies. Anion
`exchange results were obtained
`with 0.05M Tris, pH 8.6. Cation
`exchange results were obtained
`with 0.05M MES, pH 5.6.
`
`Antibody
`
`Anion Ex.
`
`Cation Ex.
`
`Mouse lgG1
`Mouse lgG1
`Mouse lgG1
`Mouse lgG1
`Mouse lgG1
`Mouse lgG2a
`Mouse lgG2a
`Mouse lgG2a
`Mouse lgG2b
`Mouse lgG2b
`Mouse lgG3
`Mouse lgG3
`Human lgM
`Mouse lgM
`Mouse lgM
`Mouse lgA
`Human lgE
`
`virus removal
`
`0
`
`0.2
`0.4 0
`0.2
`Molarity NaCl at peak center
`
`0.4
`
`at which IEC is normally carried out favors such com(cid:173)
`plexation. The product can transport high DNA levels
`all the way through such a process (Figure 4.6).[70]
`An important side note: complexation also affects
`DNA assays. Dissociation and removal of interfering
`protein are necessary to obtain accurate measure(cid:173)
`ments.[71,72]
`Anion exchange supports 3-6 logs clearance of
`viruses; cation exchange, somewhat less (Table 4.3).
`[43,44,61,62,64,65,73] Although the clearance capabil(cid:173)
`ities of cation exchange are generally lower, its selec(cid:173)
`tivity is complementary. Combining the 2 chemistries
`increases the net clearance factor, as does combining
`either or both of them with other methods. An impor(cid:173)
`tant caution: design column sanitization procedures
`with sufficient viral-kill capability to ensure that virus(cid:173)
`es left on the column don't contaminate subsequent
`product lots.
`
`Pfizer Ex. 1022
`Page 12 of 24
`
`

`

`64
`
`Purification Tools for Monoclonal Antibodies
`
`Chapter 4. Ion Exchange Chromatography
`
`65
`
`Figure 4.6. DNA transport and
`product loss due to ionic complex(cid:173)
`ation with an lgM. This chro(cid:173)
`matogram illustrates a first-step
`cation exchange separation from
`DNA-contaminated cell culture
`supernatant. Antibody distribution
`is indicated by the shaded areas.
`DNA in pg/ml is indicated by the
`dashed line. From highest to low-
`est, the values range from -1010 to
`-106. See text for discussion.
`
`endotoxin removal
`
`mass recovery
`
`Anion exchangers remove 2-5 logs of endotox(cid:173)
`in.[42,61,62,64,65,74] At alkaline pH, phosphoryl and
`carboxyl moieties from the lipid A and core polysac(cid:173)
`charide region bind strongly while the ethanolamine
`moieties are largely titrated and minimally influen(cid:173)
`tial.[75,76] Since most antibodies are weakly bound to
`anion exchangers, separation is typically good. Cation
`exchangers may eliminate up to 2 logs of endotoxin. At
`acidic pH, the influence of endotoxin negative charges
`is reduced by titration, but the ethanolamines exert
`their full positive charge. Since antibodies are among
`the stronger binding components on cation exchangers,
`they tend to coelute to some degree. Phenol red and
`other negatively charged media addi