Monoclonal antibody (mAb)

254 Pages·2012·3.7 MB·English

Monoclonal antibody (mAb)
purification by Counter Current
Chromatography (CCC)

A thesis submitted to Brunel University

for the degree of

Doctor of Philosophy

by

Samantha Fernando

Institute for Bioengineering, Brunel University, London

December 2011
Abstract
Counter current chromatography (CCC) is a form of liquid liquid chromatography, which
the Brunel Institute for Bioengineering (BIB) team have developed to process scale. In this
thesis, its application has been successfully extended to the rapid, scalable purification of
monoclonal antibodies (mAb) from mammalian cell culture, using aqueous two-phase
systems (ATPS) of inorganic salts and polymer. A polyethylene glycol (PEG) and sodium
citrate system was found to be the most appropriate by robotic phase system selection. The
search for an economical alternative to protein A HPLC is a substantial bioprocessing
concern; in this work CCC has been investigated.
Initial studies showed that unpredictably, despite separation from impurities being
achieved, some loss in the IgG‘s ability to bind to Protein A was seen, as confirmed by
Protein A BiaCore analysis. CCC machines were seen to adversely affect IgG
functionality. This led to a systematic investigation of the effect of CCC phase mixing on
IgG functionality in a number of different CCC instruments, allowing direct comparisons
of modes of CCC (hydrodynamic and hydrostatic CCC) and their associated mixing (wave-
like and cascade, respectively). The varying g forces produced within the CCC column
were determined using a recently developed model to calculate g force range. The effect of
interfacial tension was also studied using a custom built ‗g‘ shaker.
The optimum CCC mode was identified to be the non synchronous CCC,
operated in a hydrodynamic mode but allowing bobbin to rotor speed (Pr ratio) to be
controlled independently. In a normal synchronous J type centrifuge a Pr of 1 is fixed, this
is where the bobbin and rotor speed are identical I.e. one bobbin rotation (where mixing
occurs) to one rotor revolution (where settling occurs). Constraints were seen with this 1:1
ratio and the separation of mAb using ATPS. This work has shown with the use of the non
synchronous CCC at a Pr of 0.33, mixing is reduced and rotor rotations increased.
Consequently the associated g force range is decreased. Furthermore, by the extension of
settling time, the clear separation of the mAb from impurities has been achieved with
retention of biological activity.
This thesis demonstrates the importance of settling time for ATPS in
phase separation and documents the fundamental requirements for the successful
separation of biologics. Purified non synchronous CCC samples have additionally
undergone rigorous quality control testing at Lonza Biologics by their purification
1
scientists. This work has ultimately showed that with optimisation, the non synchronous
CCC can be used to produce biological samples that are of industry standard.

2
Table of contents

1) Introduction and literature review 10
1.1.1 Monoclonal antibody (mAb) discovery 11
1.1.2 mAb cell culture production 12
1.1.3 mAb structure 14
1.1.4 Monoclonal antibody market 15
1.1.5 Modes of action 17
1.1.6 Upstream/downstream processing of mAbs 18
1.1.7 Protein A: Structure 19
1.1.8 Protein A chromatography 20
1.1.9 Protein A purification issues 21
1.2 Chromatographic alterative to Protein A 22
1.2.1 Cation exchange chromatography 22
1.2.2 Hydrophobic interaction chromatography 24
1.2.3 Affinity chromatography 25
1.3 Non chromatographic alternatives 27
1.3.1 Bulk separations 28
1.3.2 Field based separations 29
1.3.3 Absorptive separation 29
1.4 Aqueous two phase system (ATPS) extraction and Counter 30
current chromatography (CCC)
1.4.1 Monoclonal antibody purification by ATPS extraction 35
1.4.2 CCC machines for separation of proteins 42
1.4.3 CCC VS Protein A affinity chromatography 44
1.4.4 CCC column design for protein separation 44
1.5 Aims and objectives 57
2) Robotic solvent system selection 58
2.1 Summary 59
2.2 Introduction 60
2.3 Method and materials 61
2.3.1 Creation of solvent system table: obtained from Lukasz Grudzien 61
2.3.2 Preparation of solvents for the robot 64
2.3.3 Sample used 65
2.3.4 Stock solution preparation 65
2.3.5 Robot programming 66
2.3.6 Initial run with model proteins 67
2.3.7 Changes made to robotic handling due to the nature of the ATPSs 68
2.4 Results and discussion 69
2.4.1 Investigation of differing protein assays 69
2.4.2 Development of Protein A HPLC analytical method for two phase 76
samples
2.4.3 Robotic run using crude Cell Culture Supernatant (CCS), and 81
result analysis by developed protein A HPLC
2.5 Conclusions 89
3) Separation of monoclonal antibody from cell culture supernatant (CCS)
using Centrifugal partitioning chromatography (CPC) 90
3.1 Summary 91
3.2 Introduction 92
3.3 Method and materials 94
3
3.3.1 Preparation of phase system 94
3.3.2 Sample preparation 94
3.3.3 CCC operation 95
3.4 Results and discussion 96
3.4.1 mAb separation from crude Cell Culture Supernatant (CCS) using
CPC: CPC run 3-1 96
3.4.2 SDS PAGE results 99
3.4.3 Investigation of the formation Protein A non binding IgG 102
3.4.4 Study into specie interconversion 104
3.4.5 Sample loading capacity: CPC run 3-2 106
3.4.6 Stationary phase retention and stripping 107
3.4.7 Prediction of retention time 109
3.4.7 Precipitation issues associated with the ammonium sulphate system 109
3.4.8 Comparison of work with the Aires Barros group (Lisbon, Portugal):
PEG 1000 14%/14% potassium phosphate system. 110
3.4.9 PEG 1000 17.5%/17.5% sodium citrate system vs. PEG 335011%/11%
ammonium sulphate system 113
3.4.10 CPC run with the PEG 1000 17.5%/17.5% sodium citrate system 117
3.5 Conclusion 122
4) Separation of monoclonal antibody from cell culture supernatant (CCS)
using toroidal coil CCC (TC CCC) centrifuge 123
4.1 Summary 124
4.2 Introduction 125
4.3 Method and materials 126
4.3.1 Preparation of phase system 126
4.3.2 TC CCC operation 126
4.4 Results and discussion 127
4.4.1 PEG 1000 17.5%/17.5% sodium citrate run on the TC CCC: Run 4-1 127
4.4.2 SDS PAGE results: silver stain 128
4.4.3 Determination of purity: host cell ELISA and concentration by stirred
cell 130
4.4.4 Effect of flow rate, peak width and spin speed 133
4.4.5 Theoretical predictability: CCC2 modelling programme 135
4.4.6 Overview 136
4.5 Conclusions 138
5) Critical look at IgG functionality 140
5.1 Summary 141
5.2 Introduction 142
5.3 Method and materials 142
5.3.1 Aim of using Biacore from this work 142
5.3.2 Biacore theory: Surface Plasmon resonance 142
5.4 Results and discussion 143
5.4.1 Effect of phase system incubation on IgG functionality 145
5.4.2 Investigation of HPLC column saturation 148
5.4.3 Biacore analysis 149
5.4.4 Investigation of shaking vigour using a bench top centrifuge 154
5.4.5 Investigation of shaking vigour at one ‗g‘ using a custom built shaker 155
5.5 Conclusions 162

4
6) Separation of monoclonal antibodies using J-type CCC centrifuges:
the need for gentler mixing 164
6.1 Summary 164
6.2 Introduction 165
6.3 Method and materials 166
6.3.1 Sample preparation 166
6.3.2 CCC processing 166
6.4 Results and discussion 167
6.4.1 Effect of rotational speed in IgG functionality 167
6.4.2 Investigation of IgG transit time of functionality 168
6.4.3 Investigation into sample loading capacity 169
6.4.4 Mini CCC separation using CCS 170
6.4.5 Investigation of model proteins 174
6.4.6 Conclusions drawn from the literature 177
6.4.7 Overview of results 178
6.5 Conclusions 182
7) Successful monoclonal antibody separation: Solution as provided by
the non synchronous CCC 183
7.1 Summary 184
7.2 Introduction 185
7.3 Method and materials 189
7.4 Results and discussion 189
7.4.1 Investigation of differing Pr ratios using PA purified IgG:
effect on recovery 189
7.4.2 Investigation of differing Pr ratios using CCS 191
7.4.3 Investigation of sample loading capacity using CCS 194
7.4.4 Increased fraction collection: effect on purity 196
7.4.5 Throughput: Studies into sample loading capacity 197
7.4.6 Effect of mass balance, purity and IgG recovery on throughput. 202
7.4.7 Investigation of associated g force with Non synchronous processing 204
7.5 Conclusions 207
8) Characterization of non synchronous CCC purified samples using Lonza
quality control assays 209
8.1 Summary 210
8.2 Introduction 211
8.3 Method and materials 211
8.3.1 Investigation of sample clean up using PD 10 size exclusion columns 211
8.3.2 Investigation of ultrafiltration columns for sample clean up 213
8.3.3 Triplicate non synchronous CCC runs with sample clean up using a
5 kDa ultrafiltration column. 215
8.3.4 Child‘s assay for the removal of PEG from samples post ultrafiltration 218
8.3.5 QC assays conducted at Lonza by purification scientist Samit Patel 220
8.3.6 Summary of results 226
8.4 Conclusions 228

9) General discussion, conclusions and future work 231

5
Acknowledgements
Firstly, I would like to acknowledge all my supervisors for giving me the opportunity to
accomplish a personal ambition with the studying of a PhD. Foremost I would like to thank
my first supervisor Dr Ian Garrard, for all his invaluable support, ideas, enthusiasm,
encouragement and advice. It was a real pleasure to work alongside him and I immensely
enjoyed his company on a personal level too. I would like to thank Professor Derek Fisher
for his expertise, insight, scientific support and tuition on biochemistry aspects.
A deepest of thank you is extended to my industrial sponsor Dr Lee Allen and his
team for their help. I thank Dr Lee Allen for all his knowledge, approachability,
promptness and general manner, which has made this collaboration so beneficial to me.
For all the industrial insight and laboratory skills I have acquired, I am very grateful. In
particular, I would like to greatly thank Mr Samit Patel for all his expertise, experience and
training within the laboratory. Most specifically for the entire QC testing he personally
conducted and his continual willingness to help or add clarity.
Great thanks and appreciation is given to all my colleagues at the Brunel Institute
for Bioengineering. In particular, to Mr Lukasz Grudzien for his experimental advice,
training, assistance within the laboratory at times of difficulty (leaks, machine blockages,
pumps breakages etc!), scientific discussion and most of all friendship. Professor Ian
Sutherland (for invaluable insight and engineering advice on all CCC machines), Dr
Remco van den Heuvel (for experimental advice and the use of his g force program), Mrs
Caroline Kenealy (for ordering and organisation of numerous conferences so efficiently)
and Mr Tony Bunce (for making the custom built shaker).
I wish to extend a heartfelt thank you to all the other PhD students at BIB, who
have now become great friends. Mostly for all the camaraderie, entertainment, and laughter
we have shared, through what has been an extremely challenging time.
Within the Division of Biosciences at Brunel I thank Professor Robert Evans for
the use of his laboratories and also knowledge and advice regarding protein chemistry. I
also thank Dr Ian Boluton, for the training and insight into the Biacore technology.
To my fiancé, Chris, I sincerely thank him for all his infinite patience,
personal support and love, without which this journey would have been so much harder.
6
Finally I dedicate this thesis to my parents without whom this would not have been
possible. In particular for all the unconditional love and financial support they have
tirelessly and unfalteringly given me. I thank my mother for her endless support and
persistent encouragement at the hardest of times. I thank my father, for his constant
reassurance and reasoning.

7
Abbreviations
ATPS Aqueous two phase system
BCA Bicinchoninic acid
BIB Brunel Institute for Bioengineering
CCC Counter current chromatography
CCD Counter-current distribution
CCS Cell culture supernatant
CDR Complementarily determining regions
CEX Cation exchange chromatography
CHO Chinese Hamster Ovary
CPC Centrifugal partitioning chromatography
DNA Deoxyribonucleic acid
ELISA Enzyme linked immunosorbent assay
Fab Fragment of antigen binding
Fc Constant fragment
FDA Food and Drug Administration
HCP Host cell proteins
HPLC High pressure liquid chromatography
HIC Hydrophobic interaction chromatography
IgG Immunoglobulin G
K Partition coefficient
kDa Kilodalton
mAb Monoclonal antibody
PA Protein A
PEG Polyethylene glycol
Pr Planetary drive ratio
PTFE Polytetrafluoroethylene
QC Quality control
qPCR Quantitative polymerase chain reaction
8
rpm Revolutions per minute
SDS PAGE Sodium dodecyl sulfate polyacrylamide gel electrophoresis
Sf Stationary phase retention
TC CCC Toroidal coil counter current chromatography
VH Heavy chain variable domain
VL Light chain variable domain

The following abbreviations have been used to ease representation in phase system
composition:
K citrate Potassium citrate and citric acid

KPO di-Potassium monohydrogen phosphate and potassium dihydrogen
4
phosphate

Na citrate Sodium citrate and citric acid

NaPO di-Sodium monohydrogen phosphate and sodium dihydrogen phosphate
4

NaSO Sodium sulphate
4

NH PO di-Ammonium hydrogen orthophosphate and ammonium di-hydrogen
4 4
orthophosphate

NH SO Ammonium sulphate
4 4

9