PRODUCTION OF BIOSIMILAR TRASTUZUMAB IN PLANTS: EXPRESSION,

PURIFICATION, AND ANALYSIS OF FUNCTION

A Thesis

Presented to

The Faculty of Graduate Studies

of

The University of Guelph

by

BRITTANY MARIE GROHS

In partial fulfilment of requirements

for the degree of

Masters of Science

January, 2011

© Brittany M. Grohs, 2011

1*1 Library and Archives Canada

Published Heritage Branch

395 Wellington Street OttawaONK1A0N4 Canada

Bibliotheque et Archives Canada

Direction du Patrimoine de Pedition

395, rue Wellington OttawaONK1A0N4 Canada

Your Tile Votre rSference ISBN: 978-0-494-80078-2 Our file Notre r6f6rence ISBN: 978-0-494-80078-2

NOTICE: AVIS:

The author has granted a non­exclusive license allowing Library and Archives Canada to reproduce, publish, archive, preserve, conserve, communicate to the public by telecommunication or on the Internet, loan, distribute and sell theses worldwide, for commercial or non­commercial purposes, in microform, paper, electronic and/or any other formats.

L'auteur a accorde une licence non exclusive permettant a la Bibliotheque et Archives Canada de reproduce, publier, archiver, sauvegarder, conserver, transmettre au public par telecommunication ou par I'lnternet, preter, distribuer et vendre des theses partout dans le monde, a des fins commerciales ou autres, sur support microforme, papier, electronique et/ou autres formats.

The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission.

L'auteur conserve la propriete du droit d'auteur et des droits moraux qui protege cette these. Ni la these ni des extraits substantiels de celle-ci ne doivent etre imprimes ou autrement reproduits sans son autorisation.

In compliance with the Canadian Privacy Act some supporting forms may have been removed from this thesis.

Conformement a la loi canadienne sur la protection de la vie privee, quelques formulaires secondaires ont ete enleves de cette these.

While these forms may be included in the document page count, their removal does not represent any loss of content from the thesis.

Bien que ces formulaires aient inclus dans la pagination, il n'y aura aucun contenu manquant.

1*1

Canada

ABSTRACT

PRODUCTION OF BIOSIMILAR TRASTUZUMAB IN PLANTS: EXPRESSION, PURIFICATION, AND ANALYSIS OF FUNCTION

Brittany Marie Grohs Advisor: University of Guelph, 2011 Professor J. C. Hall

This thesis is an investigation of the production and purification of biosimilar

trastuzumab from Nicotiana benthamiana. N. benthamiana plants were engineered to

express trastuzumab and purification of this antibody was performed. Plant-produced

trastuzumab was determined to have the same specificity for HER2 as the innovator drug

Herceptin® and was also as effective at inhibiting the proliferation of HER2-

overexpressing cancer cells. A scalable purification scheme (hollow fiber tangential flow

microfiltration, Protein A chromatography and SP Sepharose cation exchange

chromatography) was developed to improve the purification of plant-produced

trastuzumab. Immunoblot analyses revealed the purity and antibody-banding pattern of

plant-produced trastuzumab to be comparable to Herceptin®. The results of this thesis

provide further evidence that plants can be used to produce biosimilar therapeutic

antibodies.

PREFACE

This thesis contains one chapter, portions of which will be published in a book

and one chapter that has been published in a peer-reviewed scientific journal.

CHAPTER TWO

Meyers, A. J., Grohs, B.M., and Hall, J.C. Antibody Production inplanta. In: Comprehensive Biotechnology, 2nd Edition (Butler, M., Webb, C, Moreira, A., Grodzinski, B., Cui, Z.F., and Moo-Young, M., eds.). Oxford, UK: Elsevier. In press.

CHAPTER THREE

Grohs, B.M., Niu, Y., Veldhuis, L.J., Trabelsi, S., Garabagi, F., Hassell, J.A., McLean, M.D., and Hall, J.C. (2010). Plant-produced trastuzumab inhibits the growth of HER2 positive cancer cells in vitro. J. Agric. Food Chem., 58, 10056-10063.

Reprinted with permission from Grohs et al., 2010. Copyright 2010 American Chemical Society.

l

ACKNOWLEDGEMENTS

First, I would like to thank my advisor Dr. J. Christopher Hall. Four years ago, I

was an undergraduate student with no practical experience and no desire to conduct

research. Through Chris' mentoring, I have not only developed a passion for research, but

have gained valuable insight, research and writing skills, and the ability to think

critically. For this, I will be forever grateful.

Thanks to Dr. Michael D. McLean for his endless patience, guidance and support.

I will always appreciate Mike's willingness to make time to discuss new ideas and

troubleshoot problems.

Thanks to my committee members, Drs. Raja Ghosh and Donald I.H. Stewart for

their guidance in planning laboratory experiments and for critical review of this thesis.

Sincere gratitude and thanks to Yongqing Niu and Linda Veldhuis for their

friendship, for their patience in answering my endless questions, and for providing me

with the training and tools that have allowed me to excel in the lab.

Thanks to Ashley J. Meyers for her friendship, for her help in the lab, and for our

problem-solving conversations. I will always cherish your friendship.

Thanks to all my fellow members of the Hall lab group from 2003-2010. You

have made my time in the lab not only memorable, but truly enjoyable.

Finally, I would like to thank my friends and family for their continuous love and

support. I am especially grateful to Brenda Grohs, Brian Grohs, Kirsten Grohs, Valerie

Sutter, Devin Woods, and Melissa Bassoriello - 1 would not be where I am today with

out you.

ii

LIST OF ABBREVIATIONS

ADCC

AEX

Akt

ATCC

ATPS

BCIP

BSA

CDK

CDR

CEX

CH

CHO

CHT

CL

CRC

DMEM

d.p.i.

EBA

ECD

EDTA

EGF

EGFR

Antibody dependent cellular cytotoxicity

Anion exchange chromatography

Protein kinase B

American type culture collection

Aqueous two-phase system

5-Bromo-4-Chloro-3-Indolyl Phosphate

Bovine serum albumin

Cyclin-dependent kinase

Complementarity determining region

Cation exchange chromatography

Constant domain of the heavy chain

Chinese hamster ovary

Ceramic hydroxyapatite

Constant domain of the light chain

Canada research chair

Dulbecco's modified Eagle's medium

Days post infiltration

Expanded-bed adsorption

Extracellular domain

Ethylenediaminetetraacetic acid

Epidermal growth factor

Epidermal growth factor receptor

i i i

ELISA

ELP

ER

Fab

FBS

Fc

FDA

FPLC

GalT

GMP

HB-EGF

HCIC

HDEL

HER2

HIC

HIV

HRP

HSV

I.D.

IEC

IEF

IgA

IgD

Enzyme-linked immunosorbent assay

Elastin like-polypeptide

Endoplasmic reticulum

Fragment, antigen binding

Fetal bovine serum

Fragment, crystallizable

United States Food and Drug Administration

Fast-performance liquid chromatography

(31,4-galactosyltransferase

Good manufacturing practice

Heparin-binding epidermal growth factor

Hydrophobic charge induction chromatography

Histidine-aspartate-glutamate-leucine

Human epidermal growth factor receptor 2

Hydrophobic interaction chromatography

Human immunodeficiency virus

Horseradish peroxidase

Herpes simplex virus

Internal diameter

Ion exchange chromatography

Isoelectric focusing

Immunoglobulin alpha (a) isotype antibody

Immunoglobulin delta (8) isotype antibody

IgG

IgE

IgM

IMAC

ITC

IV

IQ

kDa

KDEL

LMH

mAb

MES

mRNA

MW

MS

mumAb

NBT

NF-KB

NK

NMR

NOSp

NOSt

NRG

Immunoglobulin gamma (y) isotype antibody

Immunoglobulin epsilon (s) isotype antibody

Immunoglobulin mu (u) isotype antibody

Immobilized metal affinity chromatography

Inverse transition cycling

Intravenous

Dissociation equilibrium constant

Kilodalton

Lysine-aspartate-glutamate-leucine

Litres per square meter per hour

Monoclonal antibody

1 -(iV-morpholino)ethanesulphonic acid

messenger ribonucleic acid

Molecular weight

Mass spectrometry

Murine monoclonal antibody

Nitro blue tetrazolium chloride

Nuclear factor-kappa B

Natural killer

Nuclear magnetic resonance

Nopaline synthase promoter

Nopaline synthase terminator

Neuregulin

NSERC

OD600

OMAFRA

PBK

PBS

PBST

PCR

PEG

Pi

PBK

PS

PVDF

PVX

RNAi

RP

RP-HPLC

RPMI

RTK

scFv

SDS-CE

SDS-PAGE

SP

SPR

Natural Sciences and Engineering Research Council of Canada

Optical density at 600 nm

Ontario Ministry of Agriculture, Food, and Rural Affairs

Protein kinase B

Phosphate buffered saline

Phosphate buffered saline Tween

Polymerase chain reaction

Poly(ethylene glycol)

Isoelectric point

Phosphatidylinositol-3 kinase

Polysulphone

Polyvinylidene fluoride

Potato virus X

RNA interference

Reverse phase

Reverse phase high performance liquid chromatography

Roswell Park Memorial Institute

Receptor tyrosine kinases

Single chain variable fragment

Sodium dodecyl sulfate-capillary electrophoresis

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis

Sulphopropyl

Surface plasmon resonance

TAG

TFF

TGF

TMV

TMP

TNF

TSP

Tt

VEGF

VH

v L

WNV

Triacylglycerol

Tangential flow filtration

Transforming growth factor

Tobacco mosaic virus

Transmembrane pressure

Tumor necrosis factor

Total soluble protein

Transition temperature

Vascular endothelial growth factor

Variable domain of the heavy chain

Variable domain of the light chain

West Nile virus

vn

TABLE OF CONTENTS

PREFACE i

ACKNOWLEDGEMENTS ii

LIST OF ABBREVIATIONS iii

TABLE OF CONTENTS viii

LIST OF TABLES xii

LIST OF FIGURES xiii

1 INTRODUCTION, RESEARCH OBJECTIVES, AND HYPOTHESES 1

2 LITERATURE REVIEW 5

2.1 Trastuzumab 5

2.1.1 Human Epidermal Growth Factor Receptor 2 (HER2) 5

2.1.2 Engineering of Trastuzumab 9

2.1.3 Mechanism of Action 11

2.1.3.1 Cytostatic Effects 11

2.1.3.1.1 Arrest of Cell-Cycle Progression 12

2.1.3.2 Cytolytic Effects 14

2.2 Plant Biopharming of Therapeutic Antibodies 15

2.2.1 Expression of Antibodies in Plants 15

2.2.1.1 Glycosylation 16

2.2.2 Purification of Antibodies from Plants 16

2.2.2.1 Grinding and Extraction 17

2.2.2.2 Clarification and Enrichment 18

2.2.2.2.1 Aqueous Two-Phase Partitioning System (ATPS) 19

2.2.2.2.2 ELP Fusion Protein (ELPylation) 20

viii

2.2.2.2.3 Filtration 21

2.2.2.3 Antibody Capture and Purification 23

2.2.2.4 SemBioSys Oilbody Purification Platform 24

2.2.2.5 Polishing 25

2.2.3 Characterization of Plant-Produced Antibodies 26

2.3 Affinity Chromatography 27

2.3.1 Affinity Purification Scheme 27

2.3.1.1 Affinity Purification Scheme - Binding 30

2.3.1.2 Affinity Purification Scheme - Elution 32

2.3.1.3 Factors Affecting Solute Retention 32

2.3.1.3.1 Reaction Kinetics 33

2.3.2 Affinity Resins 35

2.3.2.1 Affinity Ligands 35

2.3.2.1.1 Biological Ligands 35

2.3.2.1.2 Bioengineered Ligands 39

2.4 Conclusion 40

3 RESEARCH CHAPTER 1: PLANT PRODUCED TRASTUZUMAB INHIBITS THE GROWTH OF HER2 POSITIVE CANCER CELLS IN VITRO 41

3.1 Abstract 41

3.2 Introduction 41

3.3 Material and Methods 45

3.3.1 Cell Lines and Plasmids 45

3.3.2 Vector Construction and Plant Infiltration 46

3.3.3 SDS-PAGE and Western Blot Analyses 49

3.3.4 Quantitative ELISA 50

3.3.5 Antibody Purification 51

ix

3.3.6 N-Terminal Sequence Analysis 52

3.3.7 Cell Culture 52

3.3.8 Cell Proliferation Assay 53

3.4 Results 54

3.4.1 Accumulation of Trastuzumab in N. benthamiana Plants 54

3.4.2 Purification and Characterization of Plant-Produced Trastuzumab 55

3.4.3 Specificity of Plant-Produced Trastuzumab 56

3.4.4 Inhibition of Tumor Cell Proliferation 58

3.5 Discussion 60

3.6 Acknowledgements 62

4 RESEARCH CHAPTER 2: PURIFICATION OF A PLANT-PRODUCED ANTIBODY USING HOLLOW FIBER TANGENTIAL FLOW MICROFILTRATION 64

4.1 Abstract 64

4.2 Introduction 65

4.3 Materials and Methods 68

4.3.1 Plant Material 68

4.3.2 Extraction and Clarification 68

4.3.3 Chromatography 70

4.3.4 SDS PAGE and Immunoblot Analyses 71

4.3.5 Quantitative ELISA 71

4.4 Results 72

4.4.1 Purification of Trastuzumab from TV. benthamiana 72

4.4.2 Polishing of Plant-Purified Trastuzumab 77

4.4.3 Characterization of Antibody Integrity and Purity 77

4.5 Discussion 80

x

4.6 Acknowledgements 83

5 GENERAL CONCLUSIONS AND FUTURE DIRECTIONS 84

6 LITERATURE CITED 88

XI

LIST OF TABLES

Table 2.1 Characteristics and specifications of trastuzumab (Herceptin®) 11

Table 2.2 Immunoglobulins bound by the bacterial cell wall proteins: Protein A, G, and L 37

Table 3.1 Nucleotide sequences of the primers used in the construction of pTrasHC 48

Table 3.2 Nucleotide sequences of the primers used in the construction of pTrasLC 48

Table 4.1 Specifications for the hollow fiber tangential flow filtration module (M10S-320-01P) used in the purification of trastuzumab from N. benthamiana 70

Table 4.2 Analysis of the recovery of trastuzumab from 100 g of A7! benthamiana . 75

xn

LIST OF FIGURES

Figure 2.1 The ligand-free conformation of each of the ErbB receptors (A) and the process of ErbB receptor dimerization (B) 6

Figure 2.2 Representation of the effects of epidermal growth factor receptor (EGF)

and neuregulin 4 (NRG4) on the ErbB signalling network 8

Figure 2.3 The mammalian cell cycle 13

Figure 2.4 Schematic representation of the two types of micro filtration: dead-end

filtration (A) and tangential flow filtration (B) 22 Figure 2.5 Affinity chromatography purification scheme represented by an affinity

column (A) and typical chromatograph (B) 29

Figure 2.6 Affinity chromatograms showing the effect of reaction kinetics on solute retention 34

Figure 2.7 Ribbon diagram showing the IgG-binding sites of the B domain of Protein A (A) and the C2 domain of Protein G (B) 38

Figure 3.1 Schematic diagram of the constructs for expression of trastuzumab in N.

benthamiana; pTrasHC (A) and pTrasLC (B) 47

Figure 3.2 Quantification of trastuzumab expression in N. benthamiana 55

Figure 3.3 Analysis of the purity of plant-produced trastuzumab. Reducing,

Coomassie stained SDS-PAGE (A) and immunoblot (B) 57 Figure 3.4 Qualitative analysis of the binding of plant-produced trastuzumab to HER2

ligand 58

xin

Figure 3.5 Effect of plant-produced trastuzumab on the proliferation of human breast tumor cells that overexpress HER2 59

Figure 4.1 Schematic of the follow fiber tangential flow microfiltration system flow-path 69

Figure 4.2 Structural integrity of plant-produced trastuzumab purified using a combination of ammonium sulfate precipitation, Protein G, and Protein A chromatography 73

Figure 4.3 Purification scheme for the recovery of trastuzumab from 100 g of frozen N. benthamiana tissue (A) 74

Figure 4.4 Purification of trastuzumab from clarified N. benthamiana extract using Protein A affinity chromatography 76

Figure 4.5 Polishing of plant-purified trastuzumab by SP Sepharose cation exchange chromatography 78

Figure 4.6 Non-reducing immunoblot analysis of the fractions collected throughout the SP Sepharose purification of plant-produced trastuzumab (A-F) 79

Figure 4.7 Non-reducing (A) and reducing (B) immunoblot analyses of the purity of plant-produced trastuzumab 80

xiv

1 INTRODUCTION, RESEARCH OBJECTIVES, AND HYPOTHESES

Monoclonal antibodies (mAb) are valuable biopharmaceuticals that are used in a

variety of therapeutic applications including immunomodulation, oncology, and the

treatment of pathogenic infections (Nissim and Chernajovsky, 2008). Trastuzumab

(Herceptin®, Genentech, Inc., San Francisco, CA) is one therapeutic mAb that is used for

the treatment of metastatic breast cancer. By targeting cells that overexpress human

epidermal growth factor receptor 2 (HER2), Herceptin® regulates the uncontrolled

growth of HER2 overexpressing cells through the induction of both cytostatic and

cytolytic effects (i.e. arrest of cell cycle progression and targeted cell lysis) (Baselga and

Albanell, 2001; Beano et al., 2008; Varchetta et al , 2007). However, treatment of human

disease with mAbs such as Herceptin® requires large quantities of these

biopharmaceuticals. In one treatment cycle, a single patient is administered a loading

dose of 4 mg of Herceptin®/kg body weight followed by a weekly maintenance dose of 2

mg/kg (Cobleigh et al., 1999). Over one year, 5-10 g of Herceptin® is required to treat

one patient. An efficient expression system is thus required to meet market demands.

Therapeutic mAbs are produced by conventional mammalian cell expression

systems. However, alternative expression systems that would allow the production of

biosimilar antibodies (follow-on biopharmaceuticals derived from innovator drugs) are

being investigated (Birch and Racher, 2006; Covic and Kuhlmann, 2007; Gottlieb, 2008;

Karg and Kallio, 2009). Plants are a promising alternative to traditional mammalian

expression systems due to their ease of handling, scalability, and because their protein

folding and post-translational modifications are similar to those of mammalian systems.

Furthermore, plant-produced antibodies retain biological activities (i.e., specificity,

1

cytotoxicity, and neutralization activity) that are similar to parental antibodies produced

by mammalian cell culture (reviewed in De Muynck et al., 2010; Fischer et al., 2009).

The large-scale agricultural production of antibodies is currently limited by plant-specific

N- and 0-glycosylation profiles; however, several strategies have been developed to

humanize plant glycosylation profiles (Gomord et al., 2010).

Antibody production in plants can be divided into two stages: the expression of

antibodies in plants (upstream production) and the post-harvest processing and

purification of plant-produced antibodies (downstream processing). With the advent of

improved expression technologies (i.e. magnlCON®), high antibody bioaccumulation

levels have been achieved in plants (Bendandi et al., 2010; Giritch et al., 2006).

Furthermore, the established infrastructure for large-scale agriculture and the scale-up

potential of plants pushes the upstream production of antibodies in plants to the forefront

of alternative antibody production systems (Twyman et al., 2007). However, attaining

maximum yields of plant-produced antibodies will also require improved extraction and

purification processes since downstream processing is currently the bottleneck of plant

biopharming.

Downstream processing currently accounts for over half of the total

manufacturing costs associated with the production of therapeutic proteins in any

expression system (Roque et al., 2004). For plant-produced antibodies, post-harvest

processing and purification procedures could account for more than 80% of the total cost

of plant biopharming (Evangelista et al., 1998; Hassan et al., 2008; Mison and Curling,

2000). Therefore, despite the extensive research that has been conducted to improve the

upstream production of antibodies in plants (i.e. antibody bioaccumulation and

2

humanized glycosylation patterns), downstream costs limit the success of plant

biopharming. A simplified processing and purification scheme that is similar to, or better

than, current mammalian processes would reduce downstream processing costs (Woodard

et al., 2009). In plant processing, multiple low efficiency clarification and concentration

steps can lead to greater product losses, longer processing times, and higher costs in

comparison to antibody-purification from mammalian cell culture (Aguilar and Rito-

Palomares, 2010; Platis and Labrou, 2006; Pujol et al., 2005). Thus, to attain maximum

yields of plant-produced antibodies, the inefficiency of initial post-harvest clarification

and concentration procedures must be addressed.

The aim of this thesis was to demonstrate the effectiveness of plants as an

alternative expression system for the production of biosimilar antibodies. The research

objectives for this thesis are as follows:

• To express biosimilar trastuzumab in Nicotiana benthamiana plants using the

magnlCON® viral-based transient expression system

• To characterize plant-produced trastuzumab and compare to the innovator drug

Herceptin®

• To improve the purity of plant-produced trastuzumab

Based on the objectives listed above, the hypotheses for this thesis are:

Research Chapter 1: Plant-produced trastuzumab will be just as effective as Herceptin® at

inhibiting the proliferation of HER2-overexpressing cancer cells.

3

Research Chapter 2: A combination of hollow fiber tangential flow filtration, Protein A

affinity chromatography, and SP Sepharose cation exchange chromatography will

improve the purification of plant-produced trastuzumab.

4

2 LITERATURE REVIEW

2.1 Trastuzumab

Trastuzumab (Herceptin®) is an anti-HER2/neu humanized IgGlK antibody that is

used to treat metastatic breast cancer (Molina et al., 2001; Suzuki et al., 2007).

Herceptin® specifically targets the extracellular domain (ECD) of human epidermal

growth factor receptor 2 (HER2), a transmembrane tyrosine kinase receptor that is

overexpressed in 25-30% of breast cancers (Baselga et al., 1998; Carter et al., 1992;

Slamon et al., 1987, 1989). In clinical treatment, Herceptin® has been successfully used

as both a monotherapy (Vogel et al., 2002) and in combination with chemotherapy

(Marty et al., 2005; Slamon et al, 2001; Suzuki et al., 2007).

2.1.1 Human Epidermal Growth Factor Receptor 2 (HER2)

HER2 (pl85HER2) is a 182 kDa transmembrane tyrosine kinase receptor encoded

by the HER2 proto-oncogene (also known as neu, human homologue of the rat

neuroblastoma proto-oncogene product, or c-erbB-2, similar to the avian erythroblastosis

viral oncogene B (v-erbB) product) (Kumar et al., 1991; Lewis et al., 1993; Molina et al.,

2001; Sahin and Wiemann, 2009; Slamon et al., 1987). As a member of the ErbB family

of transmembrane receptor tyrosine kinases (RTK), HER2 is directly involved in the

regulation of normal cell growth and differentiation (Baselga and Swain, 2009; Hynes

and Stern, 1994; Suzuki et al., 2007; Yakes et al., 2002). The ErbB network consists of

four receptors, ErbBl (EGFR, HER1), ErbB2 (HER2), ErbB3 (HER3), and ErbB4

(HER4) (Baselga and Swain, 2009). Each of the four ErbB receptors is comprised of an

extracellular ligand-binding domain, a transmembrane domain, and an intracellular

5

tyrosine kinase domain (Figure 2.1A) (Fry et al., 2009; Yarden and Sliwkowski, 2001).

The extracellular ligand-binding domain consists of four subdomains (I-IV) (Citri and

Yarden, 2006). The leucine-rich repeats of subdomains I and III are specifically involved

in ligand-binding (Citri and Yarden, 2006).

Figure 2.1 The ligand-free conformation of each of the ErbB receptors (A) and the process of ErbB receptor dimerization (B). ErbB receptor dimerization begins when an extracellular ligand binds to subdomains I and III of an ErbB receptor. The interaction between a ligand and an ErbB receptor induces a conformational change that exposes the dimerization domain (subdomain II) of the ErbB receptor (Lemmon, 2009). Two ligand-bound receptors can subsequently dimerize through subdomain II (Lemmon, 2009). Receptor dimerization allows trans-phosphorylation of C-terminal tyrosine residues, which are required for activation of intracellular signalling cascades (Baselga and Swain, 2009). Reprinted by permission from Macmillan Publishers Ltd: [NATURE REVIEWS CANCER] (Baselga and Swain, 2009), copyright 2009.

6

Fourteen extracellular ligands, with conserved epidermal growth factor (EGF)

domains, can bind to the extracellular domain of the ErbB receptors (Citri and Yarden,

2006; Lazzara and Lauffenburger, 2009). Upon ligand-binding, three of the four ErbB

receptors (EGFR/Erbl, ErbB3/HER3, ErbB4/HER4) become activated and undergo a

conformational change that promotes receptor homo- or hetero-dimerization (Figure

2.IB) (Fry et al., 2009; Schmitz and Ferguson, 2009). Dimerization triggers the intrinsic

tyrosine kinase activity of the ErbB receptors, causing the trans-autophosphorylation of

tyrosine residues on the C-terminus of the receptors (Fry et al., 2009; Yakes et al., 2002).

Binding of signal transducers and adapter molecules to these phosphorylated sites

initiates various intracellular cascades involved in the regulation of cell-cycle progression

and cell proliferation (Fry et al., 2009; Mohsin et al., 2005; Yakes et al., 2002).

Control of cell growth and differentiation through ErbB signal transduction is

highly complex (Figure 2.2) (Alroy and Yarden, 1997; Baselga and Albanell, 2001).

Fourteen different extracellular ligands can bind and activate ErbB receptor dimerization

(Alroy and Yarden, 1997; Citri and Yarden, 2006). In turn, the various combinations of

ligand-induced receptor dimers determine which intracellular signalling pathways are

triggered (Baselga and Albanell, 2001; Alroy and Yarden, 1997). HER2 and ErbB3 are

non-autonomous receptors (Citri and Yarden, 2006): HER2 exists in a conformational

state that mimics a ligand-bound receptor conformation (Citri and Yarden, 2006; Klapper

et al., 1999) and ErbB3 lacks the intrinsic tyrosine kinase activity required for signal

transmission within the cell (Citri and Yarden, 2006; Guy et al., 1994) (Figure 2.1A).

7

0 u , P u t [Apoptosia ] [ Migration ) [ Growth ] [ Adhesion ] j Difleramiatiorij

Figure 2.2 Representation of the effects of epidermal growth factor receptor (EGF) and neuregulin 4 (NRG4) on the ErbB signalling network, (a) Binding of a ligand (green boxes) to the extracellular domain of an ErbB receptor (green circles) induces a conformational change that allows receptor homo- or heterodimerization to occur (input layer; green). The ErbB ligands in this figure are transforming growth factor-a (TGF-a), epidermal growth factor receptor (EGF), epiregulin, (3-cellulin, heparin-binding EGF (HB-EGF), amphiregulin, and neuregulin la, ip, 2a, 20, 3 and 4 (NRGla, IP, 2a, 2p, 3, and 4, respectively). The ErbB ligands biregulin and epigen (Lazzara and Lauffenburger, 2009) are not represented in this figure. The receptor(s) to which a particular ligand binds is (are) represented in parentheses within the green boxes. ErbB2 (HER2) does not bind any ligands and is represented by a complete green circle. ErbB3 (HER3) does not posses intrinsic tyrosine kinase activity and is represented by a complete green circle with an X. (b) Binding of adaptors and enzymes to phosphorylated tyrosine residues on the C-terminus of the ErbB dimers initiates various intracellular signalling cascades (signal-processing layer; blue) that (c) regulate cell adhesion, apoptosis, differentiation, growth, and migration (output layer; yellow). For simplicity, only a portion of the cascades and transcription factors involved in the ErbB signaling network are shown in this figure. Reprinted by permission from Macmillan Publishers Ltd: [NATURE REVIEWS MOLECULAR CELL BIOLOGY] (Yarden and Sliwkowski, 2001), copyright 2001.

8

Although not directly involved in ligand binding, HER2 is the preferred

heterodimeric partner for other ErbB receptors (Fry et al., 2009; Schmitz et al., 2009;

Yakes et al., 2002). As reviewed by Citri and Yarden (2006), HER2 heterodimers induce

stronger mitogenic signals due to stronger ligand-receptor interactions, recruit a more

diverse group of phosphotyrosine-binding proteins for signal transmission within the cell

(Jones et al., 2006), and undergo slower signal attenuation (i.e. receptor endocytosis and

recycling) (Baulida et al., 1996; Lenferink et al., 1998; Worthylake et al., 1999).

The occurrence of HER2 related breast cancer results from the amplification of

the gene encoding the HER2 protein (i.e. multiple copies of that gene) causing

overexpression of HER2 (Slamon et al., 1989; Yakes et al., 2002). Overexpression of

HER2 in the cell membrane results in the spontaneous formation of HER2 multimers,

which become constitutively autophosphorylated (Samanta et al., 1994; Yakes et al.,

2002). As a result, normal HER2 signalling pathways become disrupted, causing the loss

of cell growth regulation and the development of resistance to apoptosis (Le et al., 2003;

Zhou et al., 2001). Clinical studies have shown a correlation between the extent of HER2

overexpression and the survival time of a patient (Molina et al., 2001; Slamon et al.,

1987, 1989), where patients with higher levels of HER2 typically have more aggressive

tumors and are at a higher risk for relapse or death (Slamon et al., 1987, 1989).

2.1.2 Engineering of Trastuzumab

HER2 is an ideal candidate for targeted cancer therapy using mAbs (Molina et al.,

2001) since HER2 is overexpressed in 25-30% of human breast and ovarian tumors

(Slamon et al., 1987, 1989), and has a highly accessible ECD (Molina et al., 2001). A

murine anti-HER2 mAb (mumAb 4D5), was first developed against the ECD of HER2

9

(Fendly et al., 1990; Hudziak et al., 1989). Briefly, BALB/c mice were immunized

through a series of intraperitoneal injections with HER2-overexpressing NIH 3T3/HER2-

3400 cells (Fendly et al., 1990; Hudziak et al., 1989). Mice were subsequently boosted

with intraperitoneal injections of partially-purified HER2, followed by a final intravenous

injection with enriched HER2 (Fendly et al., 1990; Hudziak et al., 1989). Splenocytes

from mice with high antibody titers were harvested and fused with mouse myeloma cells

(cell line X63-Ag8.653) to generate hybridomas expressing anti-HER2 antibodies

(Fendly et al., 1990; Hudziak et al., 1989). Of the various anti-HER2 mAbs that were

generated, murine mAb (mumAb) 4D5 was the most effective at inhibiting HER2-

overexpressing cancer cell growth in vitro (Fendly et al., 1990; Hudziak et al., 1989;

Lewis et al., 1993). However, the potential for mumAb 4D5 to be used in human therapy

was limited because of the immunogenicity of murine antibodies in humans (Baselga et

al., 1998). mumAb 4D5 also lacked the ability to mediate cytolytic effects through

recruitment of immune effector cells (Lewis et al., 1993). A humanized mAb was thus

created by engineering the complementary determining regions (CDR) of mumAb 4D5

into the framework of a human IgGl (Carter et al., 1992). The resulting humanized anti-

HER2 antibody (trastuzumab, Herceptin®; Table 2.1) was found to be as effective as

mumAb 4D5 in inhibiting the growth of cells that were overexpressing HER2 and could

also mediate cytolytic effects on tumor cells, but lacked the immunogenicity associated

with mumAb 4D5 (Carter et al., 1992; Lewis et al., 1993). Further clinical studies

demonstrated that Herceptin® is effective in treating HER2-overexpressing breast cancer

both alone (Baselga et al., 1996; Cobleigh et al., 1999; Vogel et al., 2002) and in

combination with chemotherapy (Pegram et al., 1998; Slamon et al., 2001).

10

Table 2.1 Characteristics and specifications of trastuzumab (Herceptin ) (DrugBank, 2009; United States Food and Drug Administration (FDA) Center for Drug Evaluation and Research, 2009).

Product Background

Product name Manufacturing company FDA approval date Drug bank accession

Trastuzumab; Herceptin® Genentech, Inc. 1998 DB00072, BIOD00098, BTD00098

Product Description

Type of antibody Source Production host Target

IgGlK Humanized murine antibody Chinese hamster ovary (CHO) cells Extracellular domain (ECD) of HER2

Protein Specifications

Chemical structure Binding affinity Isoelectric point (pi) Molecular weight (MW)

C6470H10012N1726O2013S42 Kd = 5 nM 8.45 145.5 kDa

Therapy

Administration Loading dose Maintenance dose Half life

Intravenous (IV) infusion 4 mg/kg body weight 2 mg/kg body weight 2-12 days

2.1.3 Mechanism of Action

Herceptin® regulates the uncontrolled growth of cells overexpressing HER2 in

vivo through several direct and indirect molecular mechanisms, including the induction of

both cytostatic and cytolytic effects (Beano et al., 2008; Varchetta et al., 2007).

2.1.3.1 Cytostatic Effects

The cytostatic effects induced by Herceptin® include accelerated receptor

endocytosis and degradation (Baselga and Albanell, 2001; De Santes et al., 1992; Sarup

et al., 1991), prevention of HER2 ECD cleavage (reviewed by Baselga and Albanell,

11

2001) as well as reduced tyrosine phosphorylation (Kumar et al., 1991), which disrupts

initiation of intracellular cascades involved in cell-cycle progression and cell proliferation

(Mohsin et al., 2005; Molina et al., 2001; Varchetta et al., 2007; Yakes et al., 2002).

As reviewed by Shepard et al. (2008), some of the additional effects of

Herceptin® include an increased response to tumor necrosis factor-a (TNF-a) mediated

growth inhibition (Hudziak et al., 1989), decreased production of interleukin (IL-8) and

vascular endothelial growth factor (VEGF) (both proangiogenic factors) (Wen et al.,

2006), and decreased expression of CXCR4, a chemokine receptor involved in metastasis

(Li et al., 2004).

2.1.3.1.1 Arrest of Cell-Cycle Progression

Herceptin® mediates the arrest of cell-cycle progression by blocking HER-

overexpressing cells from transitioning from the first gap phase (Gl) to the DNA

synthesis phase (S) of the cell cycle (Figure 2.3) (Lane et al., 2001). Cell-cycle

progression is controlled within the cell by the family of cyclin-dependent kinases (CDK)

(Lane et al., 2001). Specifically, cyclin ECDK2 is involved in regulating the Gl/S

transition of the cell cycle (Lane et al., 2001; Sherr and Roberts, 1999). p27&pl is a CDK-

inhibitor that regulates the activity of cyclin ECDK2 (Le et al., 2003). By binding to the

cyclin ECDK2 complex during the Gl phase of the cell cycle, p27IOpl prevents the

initiation of DNA synthesis (Le et al., 2003; Sherr and Roberts, 1999). CDK-inhibition

by p27Kjpl depends on cellular levels of the inhibitor (Le et al., 2003). Cell proliferation

will proceed when p27Kjpl levels are low; however, if p27Kipl levels are high, cell

proliferation is arrested and the cell exits the cell cycle (Le et al., 2003). A correlation

between HER2 overexpression and low cellular levels of p27Kjpl was found in primary

12

breast tumor samples (Yang et al., 2000). Specifically, HER2 mediated signals have been

linked to the increased degradation of p27Kipl within the cell (Yang et al., 2000).

Overexpression of HER2 thus leads to lower levels of p27Kipl and the excessive cell

growth associated with tumors (Yang et al., 2000). Researchers have shown that

Herceptin® arrests the proliferation of HER2-overexpressing cells by affecting receptor

signalling that leads to the sequestration of p27KipI and that prevents the formation of

p27Kipl/Cdk2 (Lane et al., 2001; Le et al., 2000, 2003). As reviewed by Baselga and

Albanell (2001), treatment with mumAb 4D5 or Herceptin® increases the number of

HER2-overexpressing cells in the Go/Gi phase by 10%, while decreasing the number of

cells in the S phase (Sliwkowski et al., 1999). In another study, Herceptin® was found to

block the transition from Gl to S in almost all HER2-overexpressing cells (Lane et al.,

2001).

M G2 ^ ^ ^ ^

UP Figure 2.3 The mammalian cell cycle. Go, Gap phase 0/resting phase; Gi and G2, Gap

phases 1 and 2, respectively; M, mitosis; S, DNA synthesis. Reprinted with modifications by permission from Macmillan Publishers Ltd: [NATURE REVIEWS MOLECULAR CELL BIOLOGY] (Reed, 2003), copyright 2003.

13

2.1.3.2 Cytolytic Effects

Herceptin® mediates cytolytic effects by triggering antibody-dependent cellular

cytotoxicity (ADCC) (Arnould et al., 2006; Beano et al., 2008; Suzuki et al., 2007;

Varchetta et al., 2007). Within the body, antibody-coated cells become targeted for lysis

by cytotoxic cells such as natural killer (NK) cells, macrophages, monocytes, and

neutrophils (Kindt et al., 2007; Suzuki et al., 2007). Binding of Herceptin® to the ECD of

HER2 specifically causes increased infiltration of NK cells to the site of the tumor

(Arnould et al., 2006). Targeted cell lysis is triggered by the interaction between the Fey

receptor Ilia on cytotoxic cells of the immune system and the Fc region of Herceptin®

(Kindt et al., 2007; Suzuki et al., 2007). The release of lytic enzymes, perforin,

granzymes and cytokines by the infiltrated NK cells leads to the destruction of the

targeted cells (Kindt et al., 2007; Shepard et al., 2008; Varchetta et al., 2007).

HER2-overexpression in NIH 3T3 cells has been shown to decrease sensitivity to

TNF-a and cytotoxicity induced by activated macrophages (Hudziak et al., 1988). HER2-

overexpressing breast tumor cell lines also showed resistance to TNF-a induced

cytotoxicity (Hudziak et al, 1989). Stimulation of the apoptotic pathway by TNF-a is the

natural antitumor response of the innate immune system (Shepard and Lewis, 1988;

Vivanco and Sawyers, 2002). Under normal cellular conditions, transphosphorylation of

HER2 leads to the activation of the phosphatidyhnositol 3-kinase (PI3K)/protein kinase B

(PKB; Akt) pathway (Vivanco and Sawyers, 2002; Zhou et al., 2000). The PI3K/Akt

pathway is highly regulated due to its involvement in cell growth, proliferation, adhesion

and motility (Vivanco and Sawyers, 2002; Zhou et al, 2000). Upregulation of the

PI3K/Akt pathway and subsequent activation of Akt and nuclear factor-kappa B (NF-KB)

14

leads to the resistance to apoptotic stimuli induced by TNF-a (Vanhaesebroeck and

Waterfield, 1999; Vivanco and Sawyers, 2002; Zhou et al., 2000). By binding to HER2,

Herceptin® prevents signalling to the PI3K/Akt pathway and subsequently sensitizes cells

to TNF-a induced apoptosis (Hudziak et al., 1989; Mohsin et al., 2005; Yakes et al.,

2002).

2.2 Plant Biopharming of Therapeutic Antibodies

Mammalian cell expression systems are traditionally used for the production

therapeutic antibodies; however, these systems are limited by time consuming culturing

processes and are associated with high cost. Plants have proven to be successful for the

production of recombinant antibodies (Hiatt et al., 1989) and in comparison to

mammalian systems, offer the advantages of minimal upstream production costs, ease of

handling (i.e. crop maintenance and scalability), and biological safety due to the absence

of human pathogens such as prions or viruses (Hiatt et al., 1989; Roque et al., 2004;

Schillberg et al, 2003).

2.2.1 Expression of Antibodies in Plants

Antibody expression in plants can be achieved using three different systems:

direct gene transfer, Agrobacterium-medi&tQd stable transformation or viral-based

expression. Using these systems, antibody expression has been achieved in a wide variety

of hosts including alfalfa, algae, duckweed, lettuce, maize, moss, potato, wheat, rice,

soybean, and tobacco (Nikolov et al., 2009). Antibody expression can also be targeted to

different cellular locations (i.e. apoplast, cytosol, ER, and plastids). A comprehensive list

of the various antibodies that have been expressed in plants is provided in De Muynck et

al. (2010) and Fischer et al. (2009).

15

2.2.1.1 Glycosylation

Plant and mammalian cells have similar secretory pathways. Therefore, plant-

produced antibodies undergo protein folding and post-translational modifications

resembling those that occur in mammalian cells (Hiatt et al., 1989; Schillberg et al.,

2003). However, plant-specific N- and O-glycosylation patterns may affect antibody

efficacy and/or be immunogenic in human therapy (Chargelegue et al., 2000). There are

currently three main strategies to address the differences between plant and mammalian

glycosylation patterns: incorporation of endoplasmic reticulum (ER) retention signals,

inactivation of endogenous plant glycosyltransferases and expression of mammalian

glycosyltransferases in plants (reviewed in Gomord et al., 2010). ER retention signals

(i.e. lysine-aspartate-glutamate-leucine, KDEL, or histidine-aspartate-glutamate-leucine,

HDEL) can be added to the C-terminal ends of the heavy and light chains to retain the

antibody in the ER, (i.e. where protein glycosylation is more conserved between plants

and mammals) (Gomord et al., 2010). In contrast, RNA interference (RNAi) technology

can be used to inactivate endogenous plant glycosyltransferases such as (3(1,2)-

xylosyltransferase and a(l,3)-fucosyltransferase to prevent the addition of the plant-

specific glycans (31,2-xylose and cd,3-fucose (Cox et al., 2006; Strasser et al., 2008,

2009). Finally, mammalian glycosyltransferases such as pi,4-galactosyltransferase

(GalT) can be expressed in plants for the addition of the mammalian-specific glycan

pl,4-galactose (Bakker et al., 2006; Frey et al., 2009; Vezina et al., 2009).

2.2.2 Purification of Antibodies from Plants

Post-harvest processing and purification procedures can account for more than

80% of the total cost of plant biopharming, thus making extraction of antibodies and

16

other proteins from plants one of the greatest barriers to using plants as bioreactors

(Evangelista et al., 1998; Hassan et al., 2008; Mison et al., 2000). Inefficiency arises from

the numerous purification steps that are required to separate plant-produced antibodies

from the complex mixture of plant proteins, alkaloids, pigments, polyphenols, and

mucilages (Platis et al., 2008; Valdes et al., 2003). There is currently no universal

strategy for the purification of plant-produced antibodies; however, most schemes employ

the same series of steps (grinding and extraction, clarification and enrichment, capture

and purification, and polishing).

2.2.2.1 Grinding and Extraction

Downstream processing and purification strategies are largely dependent on the

composition of the starting material (Roque et al., 2004). Initial antibody extraction is

thus the most important post-harvest step since it releases the antibody from the plant

tissue (Hassan et al., 2008; Menkhaus et al., 2004). The extraction conditions will also

determine the ratio of the target antibody to unwanted plant contaminants such as

pigments, proteins, and enzymes that may cause antibody degradation (Hassan et al.,

2008).

Antibody expression has been achieved in a wide variety of plant hosts, tissues

and cellular locations (i.e. the ER, the cytoplasm, the chloroplast or the apoplast)

(Menkhaus et al., 2004). Consequently, the selection of appropriate extraction techniques

and conditions can depend on where the antibody is expressed. Efficient disruption of the

cell wall and membrane are required, for example, to ensure maximum antibody

extraction (Cox et al., 2009). Small-scale plant extraction can be achieved through

manual grinding of fresh or nitrogen-frozen tissue in a mortar with a pestle or through

17

mechanical grinding with a mixer mill or blender (Hassan et al., 2008; Menkhaus et al.,

2004). In contrast, large-scale plant extraction is typically conducted using mechanical

disruption equipment including hammer mills, high-shear rotor-stator mixers, and/or

high-pressure homogenizers (Nikolov et al., 2009). The disruption of plant tissue is

usually conducted in the presence of an extraction buffer; however, the ratio of buffer to

tissue must be optimized since larger volumes increase the total process time and cost.

Fresh leaf tissue from plants such as Nicotiana tabacum are composed of 80-90% water

and can thus be ground with the addition of little to no buffer (Valdes et al., 2003). In

contrast, seeds have a much lower water content (-10%) and thus require more buffer

and/or additional steps to ensure efficient extraction (Nikolov et al., 2009).

Buffer properties and composition are very important parameters to consider for

antibody extraction from plants, for example, high pH can increase antibody degradation

or reduce extraction efficiency (Hassan et al., 2008; Menkhaus et al., 2004). The addition

of detergents to the extraction buffer can increase the solubility of the target antibody and

plant proteins, while protease inhibitor cocktails and antioxidants can prevent

modification of the target antibody through degradation and/or oxidation. The

temperature, pH, and ionic strength of the extraction buffer will affect antibody stability

and solubility and thus the efficiency of antibody extraction from plant tissues (Menkhaus

et al., 2004; Nikolov et al., 2009). Buffer properties and composition can also improve

the selective extraction of the antibody over that of the contaminants.

2.2.2.2 Clarification and Enrichment

The purpose of adding a clarification step to a purification scheme is to remove

extraneous particles and macromolecules from the feed sample. Particulates and other

18

extraneous particles may cause fouling in a chromatography column. In addition,

phenolic compounds and anionic proteins present in crude plant extracts can interfere

with the performance of a chromatography column by decreasing its binding capacity, or

by causing the resin to develop ion-exchange properties due to the non-specific binding

of proteins and oligosaccharides (Platis et al., 2006). Clarification of crude plant extracts

thus extends the lifetime of a column by reducing contaminants within the feed sample.

Antibody recovery can also be enhanced by selectively isolating and enriching the

subcellular compartment containing the antibody, e.g. expression in chloroplasts followed

by their selective isolation by centrifugation (Seon et al., 2002).

Several different techniques have been employed, alone or in combination, to

clarify crude plant extracts and/or enrich the target antibody. These techniques include

acid precipitation (Woodard et al., 2009), ammonium sulfate precipitation (Grohs et al.,

2010; Huang et al., 2010), aqueous two-phase partitioning systems (ATPS) (Platis and

Labrou, 2006; Platis et al., 2008), centrifugation, dead-end filtration, ELP fusion proteins

(ELPylation) (Floss et al., 2010), and flat panel tangential flow filtration (Fischer et al.,

1999; Yuetal., 2008a).

2.2.2.2.1 Aqueous Two-Phase Partitioning System (ATPS)

The extraction of an antibody from a crude plant extract can be achieved using

aqueous two-phase partitioning (Platis and Labrou, 2006; Platis et al., 2008). Antibody

separation by ATPS is achieved by mixing two immiscible reagents, such as

poly(ethylene glycol) (PEG), dextran, other polymers, or salts, to form two distinct

phases (Balasubramaniam et al., 2003; Platis and Labrou, 2006). When a crude plant

extract is mixed with the two immiscible reagents, plant proteins and contaminants

19

separate into one of the two phases based on size, conformation, charge and

hydrophobicity. System properties such as polymer molecular mass and concentration,

salt concentration, ionic strength and pH can be optimized to minimize partitioning of

contaminants into the same phase as the antibody (Balasubramaniam et al., 2003; Platis

and Labrou, 2006). A PEG-potassium phosphate system has been specifically developed

to facilitate the purification of therapeutic antibodies (anti-HIV mAbs 2F5, 2G12, and

4E10) from N. tabacum plants (Platis and Labrou, 2006; Platis et al., 2008). Due to the

phenol-complexing properties of PEG, clarification of plant extracts by ATPS

successfully separated phenols and alkaloids from the target antibodies. Overall, plant

contaminants were reduced by 2-4 fold, while achieving 84-95% antibody recovery

(Platis and Labrou, 2006; Platis et al, 2008).

2.2.2.2.2 ELP Fusion Protein (ELPylation)

Elastin like-polypeptides (ELP) are synthetic polypeptides composed of the

repeating amino acid sequence Val-Pro-Gly-Xaa-Gly, where Xaa is any amino acid

except Pro (Conley et al., 2009; Floss et al., 2010). In a process called inverse transition

cycling (ITC), ELPs undergo temperature-dependent reversible transitions between

soluble monomers (below the transition temperature (Tt)) and insoluble aggregates

(above the Tt) (Floss et al., 2010). The purification of antibodies can be facilitated

through fusion of the antibody to the C-terminal ends of ELPs (ELPylation). Increasing

the temperature above the Tt induces the formation of ELP-antibody aggregates, which

can be separated from other proteins and compounds through centrifugation. Decreasing

the reaction temperature below the Tt allows the ELP-antibody fusions in the pellet to be

resolubilized in buffer. The antibody is subsequently released from the ELP through

20

proteolytic cleavage, a pH/temperature shift or treatment with dithiothreitol. In planta,

ELPylation has been specifically shown to increase the stability and expression levels of

both full-length antibodies and antibody fragments. ELPylation has also been used to

facilitate the purification of plant-produced antibodies and antibody fragments through

selective enrichment of the antibodies prior to chromatography (Floss et al., 2010).

2.2.2.2.3 Filtration

Microfiltration is a commonly used technique for the clarification, separation, and

purification of proteins (Saxena et al., 2009). There are two types of filtration: dead-end

filtration and tangential flow filtration (TFF). In dead-end filtration, the sample passes

perpendicular to the membrane (Figure 2.4A) (Ballew et al., 2002; Belfort et al., 1994).

The transmembrane pressure (TMP) drop, or the pressure difference between the two

sides of the membrane, is the force that drives sample permeation (Belfort et al., 1994;

Czermak et al., 2007). In dead-end filtration, the TMP is determined by the equation

(Ballew et al., 2002):

1 JVL.P— .r inlet — xpermeate

In this equation, Piniet and Pretentate represent the pressure of the feed sample and retentate,

respectively (Ballew et al., 2002). During filtration, increasing the TMP can lead to a

proportional increase in permeation rate; however, solutes in the feed sample begin to

adsorb to the surface of the membrane and block the pores (Ballew et al., 2002; Belfort et

al., 1994). As a result, the permeation rate decreases and the membrane begins to foul

(Ballew et al., 2002; Belfort et al., 1994).

21

Inlet B

ir ' t \t ••*>• \t \ m, Inlet ~> Retentate

• III •

Permeate Permeate

Figure 2.4 Schematic representation of the two types of microfiltration: dead-end filtration (A) and tangential flow filtration (B). Arrows indicate the direction of the sample relative to the membrane. Reprinted with modifications by permission from Spectrum Laboratories: (Ballew et al. 2002), copyright 2002.

In TFF, the sample passes parallel or tangential to the membrane (Figure 2.4B)

(Ballew et al., 2002; Belfort et al., 1994). In TFF, the TMP is also dependent on the

retentate pressure (PR) where (Ballew et al., 2002):

TMP=P"1'et+Pretentate-Ppemeate

Similar to dead-end filtration, the permeation rate of the sample (flux) is

proportional to the TMP (Ballew et al., 2002). Increasing the pressure exerted on the

membrane can thus also lead to the adsorption of solutes to the surface of the membrane;

however, increasing the velocity of the sample (shear rate) can help to prevent the build­

up of solutes. In TFF, the permeate flux is driven by a balance between the TMP drop and

the shear rate, but is also affected by factors such as temperature, feed concentration and

membrane selectivity (Ballew et al., 2002).

22

TFF can be used to clarify hard-to-filter solutions such as crude plant extracts,

which quickly cause membrane fouling when conventional dead-end filtration methods

are used (Stoger et al., 2004). There are several different tangential flow module

configurations: hollow fiber, tubular, flat plate, spiral wound and rotating (reviewed in

Hill and Bender, 2007; Zeman and Zydney, 1996b). Each configuration offers different

advantages such as lower manufacturing and processing costs, increased surface area,

lower shear rates, and scalability (reviewed in Hill and Bender, 2007; Zeman and

Zydney, 1996b). In contrast to other techniques, clarification by tangential flow

microfiltration does not require any phase changes or chemical additives, which may

affect the stability and structural activity of the antibody (i.e. through protein

denaturation, deactivation, and/or degradation) (Zeman and Zydney, 1996a). To date,

only flat panel TFF has been used to clarify crude plant extracts for the purification of

plant-produced antibodies (Fischer et al., 1999; Yu et al., 2008a).

2.2.2.3 Antibody Capture and Purification

Similar to antibody purification from mammalian cell cultures, conventional

techniques such as bioaffinity (Protein A, G, and L) chromatography, expanded-bed

adsorption (EBA) chromatography, ion-exchange chromatography (IEC), immobilized

metal affinity chromatography (IMAC) (reviewed in Hussack et al., 2010; Nikolov et al.,

2009), and membrane chromatographic processes (Yu et al., 2008a, 2008b) have all been

used to purify plant-produced antibodies. Fusion of affinity tags (i.e. a polyhistidine or c-

MYC tag) can also facilitate the purification of plant-produced antibodies and antibody

fragments, although it must be remembered that these tags may affect the properties of

therapeutic antibodies (i.e. protein folding, stability, and immunogenicity).

23

Due to the high selectivity of the technique, Protein A chromatography still

remains the most commonly used technique for the purification of full-length antibodies

(Platis et al., 2008). However, Protein A chromatography is associated with high cost and

is subject to instability of the ligand (Platis et al., 2008; Terman, 1985a, 1985b). As a

result, numerous strategies have also been devised to circumvent the high costs associated

with traditional column chromatography and include, for example, the use of engineered

affinity ligands (Hussack et al., 2010) and oilbodies (Seon et al., 2002).

2.2.2.4 SemBioSys Oilbody Purification Platform

Oilbodies are plant-seed organelles that store lipids (Seon et al., 2002). For

example, in safflower seeds, triacylglycerols (TAG) are encapsulated by a phospholipid

monolayer containing structural proteins called oleosins; this structure is known as an

oilbody. The hydrophobic central core of the oleosin protein is found embedded in the

phospholipid membrane surrounding the oil, while the hydrophilic N- and C-termini are

found on the cytoplasmic side of the oilbody. Fusion of recombinant proteins to either the

N- or C-terminal of the oleosin polypeptide can be achieved without affecting oilbody

structure (Seon et al., 2002). Oilbodies have specifically been exploited for the

purification of plant-produced antibodies by creating oilbodies that display recombinant

Protein A-oleosin fusions (Gomord et al., 2004). Antibody expression can be

subsequently targeted to the seeds of plants expressing the recombinant oilbodies. During

seed extraction, the oilbodies remain intact and bind plant-produced antibodies through

the Protein A molecules fused to oleosin. Centrifugation of the crude aqueous extract

results in the formation of an immiscible oilbody layer on the surface of the crude

aqueous extract (Parmenter et al., 1995; Rooijen and Moloney, 1995). Following removal

24

of the aqueous extract, plant-produced antibodies that are bound to the oilbodies via

Protein A are eluted under acidic conditions (Gomord et al., 2004). The purification of

plant-produced antibodies from leaf tissues may also be achieved by combining the crude

aqueous extract from antibody-expressing leaf tissue with the recombinant oilbodies

previously extracted from seeds.

2.2.2.5 Polishing

The cost efficiency of plant biopharming is highly dependent on the required level

of antibody purity (Twyman et al., 2005). Plant contaminants and antibody-related

impurities can affect the safety and efficacy of therapeutic antibodies by causing

sensitization and allergic reactions, toxic effects on patients and/or product instability and

altered biodistribution (Miele, 1997). Leaching of Protein A during purification is also an

issue since Protein A is both immunogenic and toxic (Platis et al., 2008; Terman, 1985a,

1985b). Plant-produced, as well as mammalian cell-produced, therapeutic antibodies

must therefore undergo one or two polishing steps to ensure that the antibody

preparations are devoid of significant levels of all contaminants (e.g. antibody aggregates

and fragments, contaminating proteins and peptides, DNA, endotoxins, leached Protein

A, and plant contaminants) (Nikolov et al., 2009). Antibody-related impurities and other

contaminants can be removed using combinations of anion exchange chromatography

(AEX), cation exchange chromatography (CEX), ceramic hydroxyapatite (CHT)

chromatography, hydrophobic charge induction chromatography (HCIC), and

hydrophobic interaction chromatography (HIC) (reviewed in Nikolov et al., 2009).

25

2.2.3 Characterization of Plant-Produced Antibodies

The fixed requirement of any expression system and purification scheme is to

ensure the consistency, purity, and safety of the antibody. In addition, the structure and

function of the antibody must not be altered. Antibody variants that differ in charge,

hydrophobicity, and/or size are easily generated during antibody expression, processing,

and purification (Liu et al., 2008). Antibody heterogeneity arises from chemical

modifications (i.e. disulfide bonds, deamidation, glycosylation, oxidation, and

truncation), noncovalent interactions, and/or aggregation (Liu et al., 2008). Several

different biochemical and biological assays must thus be conducted to characterize a

therapeutic antibody.

Biochemical techniques are used to examine the molecular integrity of the

antibody. Isoelectric focusing (IEF) and ion exchange chromatography, for example, can

be used to determine charge variants, while HIC and reverse phase (RP) chromatography

are used to determine differences in hydrophobicity. SEC, sodium dodecyl sulfate-

polyacrylamide gel electrophoresis (SDS-PAGE), sodium dodecyl sulfate-capillary

electrophoresis (SDS-CE) and mass spectrometry (MS) are all techniques that identify

size variants (Beck et al., 2005; Liu et al., 2008). Furthermore, enzyme-linked

immunosorbent assays (ELISA) and surface plasmon resonance (SPR) can be used to

examine antibody affinity and specificity. In contrast, biological assays are used to

measure antibody-induced biological effects in vitro (i.e. metabolic activity, downstream

signalling, and immune effector functions). For biosimilar antibodies, biochemical and

biological assays are also used to confirm antibody identity and efficacy compared to the

innovator drug.

26

2.3 Affinity Chromatography

Among the various purification processes, affinity chromatography is the most

widely used in the isolation of recombinant antibodies (Roque et al., 2004, 2007). First

described by Cuatrecasas and colleagues in 1968, affinity chromatography exploits the

specific and reversible interactions between biological molecules (Clonis, 2006; Hage

and Ruhn, 2006; Roque et al., 2007). The biological interactions most commonly used in

affinity chromatography include the binding of an enzyme and substrate, an antibody and

an antigen, and a hormone to its receptor (Hage and Ruhn, 2006; Phillips and Dickens,

2000a). As a result of these highly specific interactions, affinity chromatography

decreases non-specific interactions while maximizing product yields (Roque et al., 2007).

The bio-molecular recognition between a solute and ligand also facilitates protein

purification from dilute solutions (Roque et al., 2007). Affinity chromatography can thus

be used in the isolation of recombinant antibodies (Roque et al., 2007).

2.3.1 Affinity Purification Scheme

Protein purification by affinity chromatography is dependent on the highly

specific and reversible interactions between biological molecules (Clonis, 2006; Hage

and Ruhn, 2006; Roque et al., 2007). By immobilizing of one of the interacting molecules

onto a solid support the target binding partner can be purified from a complex mixture of

compounds (Hage and Ruhn, 2006). In affinity chromatography, the ligand refers to the

immobilized binding partner, while the solute refers to the target molecule to be purified

(Hage and Ruhn, 2006).

27

The affinity chromatography purification scheme begins when the mobile phase, a

crude extract containing the target solute, is applied to an affinity column (Figure 2.5)

(Hage and Ruhn, 2006; Roque et al., 2007). Crude extracts can be derived from a variety

of sources including blood fractions, cell culture supernatants or cell extracts from

bacterial, mammalian, and plant sources (Huse et al., 2002). Application of the crude

extract to the column requires prior mixing with an application buffer, which provides the

pH and the ionic strength required to promote binding between the solute to be purified

and the stationary phase (the affinity resin with immobilized affinity ligand) (Hage and

Ruhn, 2006).

During sample application, the target solute can interact with the stationary phase

through non-covalent interactions such as ion attractive interactions, hydrogen bonds,

hydrophobic interactions and/or van der Waals forces (Janson and Jonsson, 1998; Phillips

et al., 2000a). The solutes that interact with the ligand are subsequently retained on the

affinity resin (Hage and Ruhn, 2006). The solutes that do not interact with the ligand

remain in the mobile phase and pass through the column to form the non-retained peak of

an affinity chromatogram (Hage and Ruhn, 2006). To ensure the removal of all

contaminating solutes, a series of washings are performed, by passing application buffer

through the column (Roque et al., 2007).

Elution of the target solute from the column requires the application of an eluant

to disrupt the specific interaction between the ligand and the target solute (Hage and

Ruhn, 2006). Once the target solute enters the mobile phase, it passes from the column

and forms the elution peak on an affinity chromatogram (Hage and Ruhn, 2006). Finally,

28

the application buffer is passed through the column to remove the harsh elution buffer

and regenerate the affinity resin (Hage and Ruhn, 2006; Ostrove, 1990).

A

o n o

-a ii - & a in -^ iv ^ < —> ^a° —> * — • «

i ii iii iv v

EFFLUENT VOLUME •

Figure 2.5 Affinity chromatography purification scheme represented by an affinity column (A) and typical chromatograph (B). The crude extract containing the target solute is applied to the column (i). The solutes that interact with the stationary phase are retained on the column, while all other solutes (i.e. contaminants) in the mobile phase pass through to form the non-retained peak (ii). The column is washed to remove any remaining contaminants (iii). An eluant is applied to the column (iv) to elute the target solute (v). An introduction to affinity chromatography, by Hage and Ruhn, in Handbook of Affinity Chromatography, Second Edition. Copyright 2006 by Taylor & Francis Group. Reproduced with modifications with permission of Taylor & Francis Group via Copyright Clearance Center. Figure originally appeared in Hage (1998).

29

2.3.1.1 Affinity Purification Scheme - Binding

The highly specific and reversible interaction between a target solute (S) and an

immobilized ligand (L) is represented by (Hage et al., 2006b; Mohr and Pommerening,

1985a; Nelson and Cox, 2005):

ka S + L - SL

ka

SL represents the complex formed between the solute and the ligand, ka represents the

second-order association constant and ka represents the first order dissociation constant

(Hage et al., 2006b; Mohr and Pommerening, 1985a; Nelson et al., 2005). The association

equilibrium constant (Ka) can be used to quantify the affinity of the solute for the ligand

(Hage et al., 2006b; Nelson and Cox, 2005).

^ k a [S][L]

kd [SL]

At the reaction equilibrium, [S] is the concentration of the solute remaining in the mobile

phase, [L] is the concentration of immobilized ligand without bound solute and [SL] is

the concentration of the immobilized ligand with bound solute (Hage et al., 2006b). The

dissociation equilibrium constant (IQ) is also commonly used to represent the strength of

the interaction between a solute and ligand (Nelson and Cox, 2005).

Ka

30

Solute-ligand binding occurs through a combination of noncovalent interactions,

which are dependent on the collision of the solute and ligand in the right orientation

(Janson and Jonsson, 1998; Mohr and Pommerening, 1985a). Although each individual

interaction may only contribute approximately 1 kcal/mol of binding energy, the

combination of several noncovalent interactions forms a strong binding interaction

between the solute and ligand (Janson and Jonsson, 1998). The relationship between the

free energy change of the reaction (AG0) and binding strength is represented by the

following equation (Janson and Jonsson, 1998; Phillips and Dickens, 2000a):

AG°=RTlnKd

In this equation R is the gas constant (8.314 Jmol"1 K"1) and Kj (M) is the dissociation

equilibrium constant (Janson and Jonsson, 1998; Phillips and Dickens, 2000a). T

represents the reaction temperature (in degrees Kelvin), which has a direct effect on the

rotation of the solute in solution (Janson and Jonsson, 1998; Mohr and Pommerening,

1985a). The lowest dissociation equilibrium constant required to retain a solute on an

affinity column is equal to 10"5 M (Janson and Jonsson, 1998). Using the equation above,

the binding energy required for the interaction between a solute and ligand at standard

temperature (298 K) and pressure (1 atm) is thus 7 kcal/mol (Janson and Jonsson, 1998).

As each binding interaction contributes around 1 kcal/mol of binding energy, four to

eight noncovalent interactions are required to retain a solute on an affinity column

(Janson and Jonsson, 1998).

31

2.3.1.2 Affinity Purification Scheme — Elution

Elution of the target solute from a column requires the application of an eluant to

disrupt the specific interactions between the ligand and the target solute (Hage and Ruhn,

2006). Solute elution can be accomplished by using either a specific or a non-specific

eluant (Hage and Ruhn, 2006; Phillips and Dickens, 2000b). Although both elution

schemes are effective, choosing the wrong eluant can cause damage to the target solute or

to the immobilized ligand (Phillips and Dickens, 2000b).

In competitive elution, the specific eluant is a solvent that contains an analogous

solute, which will compete for binding to the immobilized ligand (Phillips and Dickens,

2000b; Roque et al., 2007). When the specific eluant is applied to the column, the target

solute is displaced from the ligand and is eluted from the column (Hage and Ruhn, 2006;

Roque et al., 2007). Competitive elution will not affect the structural integrity of the

target solute or ligand (Phillips and Dickens, 2000b).

A non-specific eluant is a solvent that disrupts the non-covalent binding

interactions between the target solute and the immobilized ligand through an increase in

ionic strength, temperature, an extreme pH or by chaotropic dissociation (Hage and Ruhn,

2006; Larsson, 1993; Roque et al., 2007). Applying a nonspecific eluant to the column

causes the target solute to dissociate from the ligand and elute from the column (Phillips

and Dickens, 2000b; Roque et al., 2007).

2.3.1.3 Factors Affecting Solute Retention

Numerous factors can affect the retention of a solute on an affinity resin (Hage et

al., 2006b; Mohr and Pommerening, 1985a; Phillips and Dickens, 2000b). The most

32

significant factors include the concentration of ligand per unit of resin and the reaction

kinetics that define the interaction between the solute and ligand (Hage et al., 2006b).

2.3.1.3.1 Reaction Kinetics

The dissociation equilibrium constant (Ka) is commonly used to represent the

strength of the interaction between a solute and ligand (Nelson and Cox, 2005). When

there is little to no affinity between the target solute and ligand (Ka> 10" M), the target

solute will elute from the column with the contaminating solutes (non-retained peak)

(Figure 2.6A; (Mohr and Pommerening, 1985a)). As the affinity between the target

solute and ligand increases (Kj< 10"3 M), the target solute will weakly interact with the

immobilized ligand. However, this weak binding affinity is not strong enough to allow

for sufficient separation of the target solute from the contaminants (Figure 2.6B). As the

affinity between the target solute and ligand increases (IQ values between 10"4 and

10"3 M), the solute will be retained on the column, thus allowing for solute separation

(Figure 2.6C). Optimal Ka values for protein separation on an affinity column range

between 10" and 10" M. These binding affinities correspond to multiple non-covalent

interactions between the target solute and the immobilized ligand. The target solute is

thus firmly retained on the column, and can only be removed by applying a specific or

nonspecific eluant (Figure 2.6D). The greater the affinity between a target solute and

ligand (Kd< 10" M), the more difficult it will be to elute the target solute from the

column (Mohr and Pommerening, 1985a).

33

t s e o 00 csi <u ( j

c €3

S-r

O VI

<

c f

I \

\ / \

Effluent Volume

Figure 2.6 Affinity chromatograms showing the effect of reaction kinetics on solute retention. No affinity between the target solute and immobilized ligand; inefficient separation of the target solute (elution peak - dashed line) from the contaminants (nonretained peak - solid line) (A). Weak affinity (Ka< 10"3 M) (B), moderate affinity (Kd = 10"4 - 10"3 M) (C) and high affinity (104 and 10"8 M) (D) between the target solute and ligand. General considerations of the adsorption and elution step by Mohr and Pommerening, in Affinity Chromatography. Copyright 1985 by Marcel Dekker, Inc. Reproduced with modifications with permission of Marcel Dekker, Inc. via Copyright Clearance Center.

Solute retention on an affinity resin can be determined by the following equation (Hage et

al., 2006b; Janson and Jonsson, 1998):

k = (Ka)(mL)

V M

(tM)

In the first equation, k represents the retention factor of the solute of interest (Hage et al,

2006b; Janson and Jonsson, 1998). mL is the amount of active ligand that is immobilized

on the affinity resin and VM represents the volume of solvent in the column (also known

as void volume) (Hage et al., 2006b; Janson and Jonsson, 1998). In the second equation,

34

tR represents the retention time of the target solute that was retained on the column and tM

represents the retention time of the solutes that were not retained on the column (Hage et

al., 2006b; Sanbe and Haginaka, 2003). Efficient protein resolution occurs when the

retention time of the target solute is much higher than the retention time of contaminating

solutes (Janson and Jonsson, 1998).

2.3.2 Affinity Resins

A wide range of ready-to-use affinity resins are currently available for the

purification of antibodies (Mohr and Pommerening, 1985b). Selection of an appropriate

resin requires careful consideration of the source of the antibody to be purified, the

antibody isotype and the required degree of purity of the final product (Huse et al., 2002;

Roque et al., 2007). The cost associated with the production or purchase of the affinity

resin is also an important factor to be considered (Roque et al., 2007). If no suitable

commercial resin is available, it may be necessary to develop a more efficient affinity

resin (Mohr and Pommerening, 1985b).

2.3.2.1 Affinity Ligands

2.3.2.1.1 Biological Ligands

Biological affinity ligands used in the purification of antibodies are isolated from

natural sources and include antigens, anti-antibodies and lectins (Clonis, 2006; Huse et

al., 2002; Roque et al., 2007). However, the most widely used group of natural ligands

are the bacterial cell wall proteins (Roque et al., 2007). Found on the cell surface of some

species of bacteria, these proteins bind immunoglobulins in the blood to prevent the host

immune system from recognizing the bacteria as a foreign agent (Hage et al., 2006a;

Huse et al., 2002).

35

Protein A, a cell wall associated protein isolated from Staphylococcus aureus, is

one of the most widely used bacterial cell wall proteins for the purification of antibodies

(Bjorck and Kronvall, 1984; Hober et al., 2007). In its native form, Protein A consists of

five homologous immunoglobulin-binding domains (E, D, A-C), each with the ability to

bind the Fc region of IgGl, IgG2, and IgG4 (Table 2.2) (Hober et al., 2007; Kronvall and

Williams, 1969; Linhult et al., 2004; Turkova, 1999). Immobilization of Protein A to a

solid support; however, results in the inactivation of the three immunoglobulin-binding

domains located at the N-terminus of Protein A (Phillips et al., 1985; Turkova, 1999).

Therefore, when used as an affinity ligand, native Protein A can only bind two

immunoglobulins (Phillips et al., 1985; Turkova, 1999).

The structures of three domains of Protein A (B, D, and E) have been solved by

protein nuclear magnetic resonance (NMR) (Gouda et al., 1992; Starovasnik et al., 1996;

Torigoe et al., 1990) and x-ray crystallography (Graille et al., 2000). Structural analysis

reveals that these domains adopt similar immunoglobulin-binding structures, each

comprised of three a-helices (Graille et al., 2000; Karimi et al., 1999). The structure of

the B domain of Protein A in complex with the Fc region of an IgG has also been

examined through x-ray crystallography (Deisenhofer, 1981). The crystal structure

indicates that two of the three helices of the B domain of Protein A bind within the region

connecting the CH2 and CH3 domains of the antibody heavy chain (Figure 2.7A)

(Deisenhofer, 1981; Starovasnik et al., 1996).

Protein G, a surface receptor isolated from groups C and G Streptococci, has a

high affinity for the Fc region of an immunoglobulin (Bjorck and Kronvall, 1984;

Phillips, 2006; Roque et al., 2004, 2007). In its native form, Protein G consists of three

36

Table 2.2 Immunoglobulins bound by the bacterial cell wall proteins: Protein A, G, and L. Abbreviations: Strong binding (+++), mild binding (++), weak binding (+), no binding (-), unknown (U). Bioaffinity chromatography, by Hage et al. (2006a), in Handbook of Affinity Chromatogarphy, Second Edition. Copyright 2006 by Taylor & Francis Group. Reproduced with modifications with permission of Taylor & Francis Group via Copyright Clearance Center.

Source

Bovine

Goat

Human

Mouse

Rabbit

Isotype Protein A Protein G Protein L

IgGi

IgG2

IgGi

IgG2

IgGj

IgG2

IgG3

IgG4

IgM

IgA,

IgA2

IgE

IgD

IgG,

IgG2a

IgG2b

IgG3

IgG

+

+++

+

+++

+++

+++

+

+++

+

+

+

++

-

+

+++

+++

+++

+++

+++

+++

+++

+++

+++

+++

+++

+++

-

-

-

-

-

++

+++

+++

+++

+++

-

-

-

-

+++

+++

+++

+++

+++

+++

+++

+++

+++

+++

+++

+++

+++

+

trThe Fc region of an antibody is bound by Protein A and G, while KI, KIII, and KIV light chains and bound by Protein L

immunoglobulin-binding domains (C1-C3), each with the ability to bind the Fc region of

an antibody (Hage et al., 2006a; Stahl et al., 1993). In addition, native Protein G contains

two albumin-binding domains, each with the ability to bind serum albumin from various

mammalian species (Hage et al., 2006a; Stahl et al., 1993; Wideback and Kronvall, 1982;

Wideback et al., 1982). In recombinant forms of Protein G, the albumin binding sites

37

CH2

CH3

CH2

CH3

Figure 2.7 Ribbon diagram showing the IgG-binding sites of the B domain of Protein A (A) and the C2 domain of Protein G (B). The B domain of Protein A is formed of three a-helices, two of which (red) bind to the CH2 side of the region joining the CH2 and CH3 domains of IgG (A; Protein Data Bank ID 1FC2; Deisenhofer, 1981). The C2 domain of Protein G is formed of an a-helix and a four-stranded P-sheet (blue), which bind within the hinge region joining the CH2 and CH3 domains of IgG (B; Protein Data Bank ID 1FCC; Sauer-Eriksson et al., 1995).

have been removed along with one immunoglobulin-binding region (C2) (Hage et al.,

2006a; Sauer-Eriksson et al., 1995). As a result, the recombinant form of Protein G

consists of only two Fc binding regions, each with the potential to bind one

immunoglobulin (Hage et al., 2006a; Jurado et al., 2006).

Protein L, a cell wall protein isolated from Peptostreptococcus magnus, is also

used in the purification of antibodies (Akerstrom and Bjorck, 1989; Jurado et al., 2006;

Phillips, 2006). In its native form, Protein L consists of five domains (B1-B5), each with

the ability to bind kappa light chain containing antibodies (Table 2.1; (Wikstrom et al.,

1994)). The Bl domain of Protein L contains an a-helix and a four-stranded P-sheet,

which bind to a conserved backbone conformation in the variable region of kappa light

38

chain containing antibodies, specifically kappa I, kappa III, and kappa IV but not kappa

IV or lambda (Graille et al., 2001; Nilson et al., 1992). As such, Protein L can be used in

the purification of both whole antibody molecules and immunoglobulin fragments such as

single-chain variable fragments (scFv) and Fab fragments without affecting the antigen

binding properties of the antibody (Akerstrom and Bjorck, 1989; Hage et al., 2006a).

2.3.2.1.2 Bioengineered Ligands

During the chromatographic process, affinity ligands are subject to the harsh

conditions required for solute elution (Linhult et al., 2004). One of the downfalls of using

biological ligands in affinity chromatography is their susceptibility to protein

denaturation or chemical cleavage during these harsh treatments (Linhult et al., 2004;

Narayanan, 1994). Bioengineered ligands resemble natural ligands such a Protein A and

Protein L, but have been engineered to provide enhanced selectivity or chemical stability

(Roque et al., 2007).

The Z domain is a recombinant form of the native B domain of Protein A, which

has been engineered to increase protein stability under alkaline conditions (Roque et al.,

2004). In alkaline conditions, proteins with asparagine residues are subject to

deamidation or backbone cleavage, where the degree of susceptibility is relative to the

primary structure of a protein (Linhult et al., 2004). A protein is particularly susceptible

to alkaline conditions when its primary structure contains an asparagine residue followed

by a glycine residue (Linhult et al., 2004). A protein that contains asparagine-glycine

residues is also susceptible to cleavage by hydroxylamine, a chemical used to cleave

proteins that have been expressed as fusions to Protein A (Hober et al., 2007; Linhult et

al., 2004). Although the B domain of Protein A is relatively stable under alkaline

39

conditions, the Z domain was engineered through two amino acid substitutions (Alai to

Val and Gly29 to Ala) to provide additional stability and resistance to cleavage by

hydroxylamine (Linhult et al., 2004; Nilsson et al., 1987; Tashiro et al., 1997). Despite

slight structural differences between the Z domain and the native B domain of Protein A,

both proteins exhibit similar binding capacities for the Fc region of antibodies (Hober et

al , 2007; Jendeberg et al , 1995; Zheng et al, 2004). The Z domain of Protein A,

however, does not bind to the Fab region of antibodies with as high of an affinity as the

native B domain (Jansson et al., 1998). The Z domain is thus a more stable purification

reagent than the B domain for the elution of recombinant antibodies under alkaline

conditions (Roque et al., 2004; Uhlen and Moks, 1990).

2.4 Conclusion

Herceptin® is an important biopharmaceutical used in the treatment of human

metastatic breast cancer. Although Herceptin® is produced by traditional mammalian cell

culture, production of a biosimilar trastuzumab in plants could alleviate the cost and

growing demand for antibody-based therapeutics. The agricultural production of

biosimilar trastuzumab was thus examined in the first research chapter of this thesis.

The large-scale agricultural production of antibodies remains limited by

inefficient downstream processing and purification procedures. A scalable scheme that

combines hollow fiber tangential flow microfiltration with Protein A affinity

chromatography and cation exchange chromatography was thus developed in the second

research chapter of this thesis.

40

3 RESEARCH CHAPTER 1: PLANT PRODUCED TRASTUZUMAB INHIBITS THE GROWTH OF HER2 POSITIVE CANCER CELLS IN VITRO

Reprinted with permission from Grohs et al., 2010. Copyright 2010 American Chemical Society.

3.1 Abstract

To study the agricultural production of biosimilar antibodies, trastuzumab

(Herceptin®) was expressed in Nicotiana benthamiana using the magnlCON® viral-based

transient expression system. Immunoblot analyses of crude plant extracts revealed that

trastuzumab accumulates within plants mostly in the fully assembled tetrameric form.

Purification of trastuzumab from N. benthamiana was achieved using a scheme that

combined ammonium sulfate precipitation with affinity chromatography. Following

purification, the specificity of the plant-produced trastuzumab for the HER2 receptor was

compared with Herceptin® and confirmed by western immunoblot. Functional assays

revealed that plant-produced trastuzumab and Herceptin® have similar in vitro anti­

proliferative effects on breast cancer cells that overexpress HER2. Results confirm that

plants may be developed as an alternative to traditional antibody expression systems for

the production of therapeutic mAbs.

3.2 Introduction

Antibody research over the past 30 years has led to the development of valuable

biopharmaceuticals for the diagnosis and treatment of human disease (Nissim and

Chernajovsky, 2008). To date, the United States Food and Drug Administration (FDA)

has approved 22 monoclonal antibodies (mAb) for clinical use, while hundreds of others

are in clinical trials (Chames et al., 2009; Dimitrov and Marks, 2009). Antibodies

41

currently approved for clinical therapy have a wide range of applications, including the

treatment of microbial infections, autoimmune diseases, and cancer (Chadd and Chamow,

2001; Stoger et al., 2005). The advantage of using antibodies in therapeutic applications

is their low toxicity and high specificity for a target antigen (Ko et al., 2009); however, to

ensure the efficacy of some treatments, high antibody serum concentrations must be

maintained over a period of several months (Mori et al., 2007). One treatment cycle for a

single patient can require hundreds of milligrams to gram quantities of mAbs (Leong and

Chen, 2008; Mori et al., 2007). Therapeutic mAbs are thus among the most lucrative

products within the biopharmaceutical industry (Karg and Kallio, 2009). From 2004 to

2006, market sales of the top five therapeutic mAbs (Rituxan®, Remicade®, Herceptin®,

Humira®, and Avastin®) increased from $6.4 billion to $11.7 billion (Dimitrov and

Marks, 2009). During 2010, the market value of these antibodies is predicted to rise to

over $30 billion (Ko et al., 2009). In the past, such high market demands for

biopharmaceuticals have led to a manufacturing bottleneck (Karg and Kallio, 2009).

Therapeutic mAbs have traditionally been produced in mammalian cell systems;

however, these systems are associated with high production costs and are hindered by

time-consuming culturing processes (Birch and Racher, 2006; Roque et al., 2007). In an

attempt to meet rising market demands, pharmaceutical companies are working to

improve the efficiency of existing biopharmaceutical production systems (Birch and

Racher, 2006; Karg and Kallio, 2009) as well as increase the number of antibody

production facilities (Karg and Kallio, 2009). Following construction, these facilities

must be validated under Good Manufacturing Practice (GMP), a process that can take an

average of three years (Vezina et al., 2009). Although some improvements have been

42

made to increase antibody production, pharmaceutical companies still may not be able to

meet future demands. As a result, alternative expression systems that would allow the

production of biosimilar antibodies (follow-on biopharmaceuticals that have been proven

to be similar to innovator drugs) are also being investigated (Birch and Racher, 2006;

Covic and Kulmann, 2007; Gottlieb, 2008; Karg and Kallio, 2009).

Agricultural production of therapeutic proteins (biopharming) is one alternative to

traditional mammalian cell expression systems for the large-scale production of

therapeutic mAbs. In comparison to mammalian systems, plant bioreactors offer many

advantages for the pharmaceutical industry, including lower upstream production costs,

speed of manufacturing, indefinite scalability, and ease of handling (Arntzen, 2008;

Churchill et al., 2002). Plants also offer the advantage of biological safety, as there is no

health risk from contamination with zoonotic pathogens and toxins (Hefferon, 2010;

Huang et al., 2010). Plant biopharming is also beneficial for the agricultural industry

since biopharmed crops can be maintained and harvested using current agricultural

practices (Twyman et al., 2007), and thus provide valuable new markets for farmers

(Hussack et al., 2010). Conversely, the limitations of plant bioreactors include higher

downstream processing and purification costs, and the addition of plant-specific N-

glycans to the recombinant antibodies. Full-length recombinant antibodies were first

successfully expressed in tobacco plants in 1989 (Hiatt et al., 1989). Since then, a wide

variety of transgenic plant hosts have been successfully used for recombinant antibody

production (Fischer et al., 2009), including plants that have been genetically modified to

express recombinant antibodies with humanized A -̂glycan profiles (Gomord et al., 2010).

The expression of antibodies in plants has also been achieved using different expression

43

platforms, including both stable and transient plant transformation technologies (Ko et

al., 2009).

To achieve regulatory affirmation of plant-produced biosimilar therapeutics,

researchers must be able to demonstrate that plant-produced antibodies maintain the

identical structural and functional integrity as their mammalian counterparts (Stoger et

al., 2005). Plant-produced antibody preparations must also be analyzed to ensure that they

are homogeneous, not adversely immunogenic, and devoid of significant contaminants

(Stoger et al., 2005). No study has been conducted to date to compare a plant-produced

antibody with a clinically approved therapeutic mAb.

Trastuzumab (Herceptin®, Genentech Inc., San Francisco, CA) is a humanized

murine immunoglobulin G1K antibody that is used in the treatment of metastatic breast

cancer. Trastuzumab binds to the extracellular domain of human epidermal growth factor

receptor 2 (HER2), a member of the ErbB family of transmembrane tyrosine kinase

receptors, that is overexpressed in 20-30% of metastatic breast cancer patients (Ben-

Kasus et al., 2009; Slamon et al, 1987, 1989). Under normal cell conditions, HER2 is

directly involved in the activation of signaling pathways that mediate cell growth and

differentiation (Ben-Kasus et al., 2009; Molina et al., 2001; Suzuki et al., 2007).

Overexpression of HER2 results in the disruption of normal signaling pathways, causing

the loss of cell growth regulation and the development of resistance to apoptosis (Le et

al., 2003; Zhou et al., 2001). By targeting cells that overexpress HER2, trastuzumab

mediates the arrest of cell proliferation and the lysis of cancer cells by antibody-

dependent cellular cytotoxicity (ADCC) (Arnould et al., 2006; Beano et al., 2008; Suzuki

et al., 2007). In treatment, patients with HER2-overexpressing metastatic breast cancer

44

are administered a loading dose of 4 mg of trastuzumab/kg followed by a weekly

maintenance dose of 2 mg/kg (Cobleigh et al., 1999). Upon the basis of market demand,

treatment of human metastatic breast cancer with trastuzumab thus requires kilogram

quantities of this biopharmaceutical. An alternative expression system may therefore be

required to supply future worldwide need for therapeutic antibodies such as trastuzumab.

Genetically modified plants have proven successful for the large-scale production

of mAbs; however, no study has yet been conducted to characterize and compare a plant-

produced antibody having a primary structure identical to a clinically approved

therapeutic mAb. In this study, trastuzumab was expressed in N. benthamiana, a relative

of tobacco, using a viral-based transient expression system (Giritch et al., 2006).

Trastuzumab expression in N. benthamiana plants was quantified, and plant-purified

trastuzumab was characterized in comparison to Herceptin®. Plant-produced and

commercial trastuzumab were found to bind the same ligand and have similar in vitro

anti-proliferative effects on breast cancer cells that overexpress HER2. These results

indicate that agricultural biopharming could be developed for effective use as an

alternative to mammalian cell systems for the large-scale production of biosimilar

antibodies.

3.3 Material and Methods

3.3.1 Cell Lines and Plasmids

All cloning procedures were performed using Top 10 F' Escherichia coli cells

(Invitrogen, Burlington, Canada). Plasmids used in the magnlCON® viral-based transient

expression system (pICH21595, pICH25433 pICH20111, pICH24180 and pICH14011)

were obtained from Icon Genetics GmbH (Halle, Germany (Giritch et al., 2006)). All

45

mammary adenocarcinoma cell lines (MCF-7, SK-BR-3 and BT-474) were obtained from

American Type Culture Collection (ATCC; Rockville, MD) and cultured according to

ATCC specifications unless stated otherwise.

3.3.2 Vector Construction and Plant Infiltration

The variable coding regions of the heavy (VH) and light (VL) chains of

trastuzumab (Carter et al., 1992) were synthesized as gene segments by the PBI/NRC

DNA/Peptide Synthesis Laboratory of the National Research Council of Canada

(Saskatoon, Canada), both incorporating preferred plant codons (Almquist et al., 2006;

McLean et al., 2007; Olea-Popelka et al., 2005), a 24 amino acid N-terminal murine SP

(GenBank accession no. AAA38889.1), and 5' Xbal and 3' Notl restriction sites. The

complete heavy chain coding sequence was assembled by subcloning murine SP-VH into

the XbaVNotl sites of pMM29 (McLean et al., 2007), a plasmid containing the tobacco

optimized coding region of a human gamma-1 heavy chain constant (CH) domain fused to

a six-Histidine and a KDEL tag. Three amino acids (ASP359, Leu36i, and Lys45o) differed

between the human gamma-1 heavy chain constant domain and the heavy chain of

trastuzumab. Site directed mutagenesis was used to change ASP359 to Glu and Leu36i to

Met and to remove the Notl site. The complete heavy chain coding sequence including

murine SP was then amplified by PCR to remove Lys45o, the six-Histidine, and KDEL C-

terminal tags using primers that contained Bsal sites and was subcloned into pICH21595

(Giritch et al., 2006; Icon Genetics) to generate pMTrasHC. The complete light chain

coding sequence was assembled by subcloning murine SP-VL into the XbaVNotl sites of

pMM29 (McLean et al., 2007). The Notl site was removed by site directed mutagenesis,

and the complete light chain coding sequence including murine SP was PCR amplified

46

using primers containing Bsal sites and subcloned into pICH25433 (Giritch et al., 2006;

Icon Genetics) to generate pMTrasLC. The Arabidopsis basic Chitinase SP (Samac et al.,

1990) later replaced the murine SP in both pMTrasHC and pMTrasLC, generating

pTrasHC and pTrasLC, respectively (Figure 3.1). All primers used for the development

of pTrasHC and pTrasLC are listed in Tables 3.1 and 3.2, respectively.

LB

Lf NOSt npt II NOSp int - SP HC

RB

3'TMV NOSt

B

LB

Lf NOSt npt II NOSp mt -SP LC

RB

3' PVX NOSt u

Figure 3.1 Schematic diagram of the constructs for expression of trastuzumab in N. benthamiana; pTrasHC (A) and pTrasLC (B). Both expression constructs contain the npt II gene under the control of the nopaline synthase promoter (NOSp). NOSt: nopaline synthase terminator; LB and RB: left and right borders, respectively; AttB: recombination site; int: intron; SP: Arabidopsis basic chitinase signal peptide; HC: coding sequence of the heavy chain of trastuzumab; LC: coding sequence of the light chain of trastuzumab; 3'TMV: 3' untranslated region; 3'PVX: 3' untranslated region.

47

Table 3.1 Nucleotide sequences of the primers used in the construction of pTrasHC.

Name Type Nucleotide Sequence

Removal of Notl site

TrasHC-NotI Forward 5'-GTGACAGTATCAAGTGCTTCCACCAAGGGACCAAGC-3'

Reverse 5 '-GCTTGGTCCCTTGGTGGAAGCACTTGATACTGTCAC-3'

Amino acid modification (ASPTSQ —» Glu; L e u ^ —» Met)

TrasHC-2AA Forward 5'-CACTTCCACCTTCTAGGGAAGAAATGACAAAGAACCAAGTG AGCC-3'

Reverse 5 '-GGCTC ACTTGGTTCTTTGTCATTTCTTCCCTAGAAGGTGGAA GTG-3'

Subcloning into pICH21595 (addition of Arabidopsis basic chitinase SP, removal of Lys4S0, 6xHis and KDEL tags)

TrasHC-S Forward 5'-TTTGGTCTCAAGGTATGGCTAAAACAAATCTCTTTTTATTCTT GATTTTCTCCCTTTTACTTTCCTTAAGCTCAGCGGAAGTTCAACT TGTTGAGAGTG-3'

Reverse 5 '-TTTGGTCTCAAAGCTCATTATCCTGGGCTAAGGCTAAG-3'

Table 3.2 Nucleotide sequences of the primers used in the construction of pTrasLC.

Name Type Nucleotide Sequence

Removal of Notl site

TrasLC-NotI Forward 5'-CAAAGTTGAGATCAAGAGGACCGTGGCTGCACCAAG-3'

Reverse 5 '-CTTGGTGCAGCCACGGTCCTCTTGATCTCAACTTTG-3'

Subcloning into pICH25433 (addition of Arabidopsis basic chitinase SP)

TrasLC-S Forward 5'-TTTGGTCTCAAGGTATGGCTAAAACAAATCTCTTTTTATTCTT GATTTTCTCCCTTTTACTTTCCTTAAGCTCAGCGGACATTCAAAT GACTCAATCCC-3'

Reverse 5 '-TTTGGTCTCAAAGCTCATTAACACTCTCCTCTATTGA-3'

48

The TMV-based 5' module (pICH20111), PVX-based 5' module (pICH24180),

and integrase (pICH14011) vectors (Giritch et al., 2006; Icon Genetics) were unaltered.

All five plasmids (pICHHOl 1, pICH20111, pICH24180, pTrasHC, and pTrasLC) were

introduced into the Agrobacterium tumefaciens strain At542 by electroporation. N.

benthamiana plants were vacuum infiltrated according to the protocol described in ref

(Marillonnet et al., 2005) with several modifications. Briefly, all cultures were grown at

28°C and 220 rpm to a final optical density at 600 nm (OD6oo) of 1.8. Equal volumes

were combined and pelleted by centrifugation at 8,000 rpm for 4 min, resuspended, and

diluted by 10 in infiltration buffer (10 mM l-(7V-morpholino)ethanesulphonic acid

(MES) at pH 5.5 and 10 mM MgSC^). The aerial parts of six-week-old N. benthamiana

plants were submerged in a desiccator containing the A. tumefaciens resuspension

solution under vacuum (0.5 to 0.9 bar) for 90 s followed by a slow release of the vacuum,

after which plants were returned to the greenhouse for 8 days before being harvested.

3.3.3 SDS-PAGE and Western Blot Analyses

Fresh leaf biomass from three N. benthamiana plants was harvested 8 days post-

infiltration (d.p.i), ground separately under liquid nitrogen, and combined with two

volumes of cold extraction buffer [40 mM phosphate buffer, 50 mM ascorbic acid, and 10

mM ethylenediaminetetraacetic acid (EDTA) disodium salt dihydrate, pH 7.0]. Crude

extracts were clarified by centrifugation at 10,000 rpm for 30 min and then 5,000 rpm for

10 min at 4°C. Total soluble protein (TSP) concentration was determined using the Bio-

Rad Protein Assay (Mississauga, Canada). Bovine serum albumin (BSA; Thermo

Scientific, Nepean, Canada) was used as the protein standard. Western immunoblots were

performed as described (Almquist et al., 2006), using a mixture of goat anti-human IgG

49

y- and K-chain specific probes conjugated to alkaline phosphatase (Sigma-Aldrich,

Oakville, Canada), diluted to 1:2500 in phosphate-buffered saline, pH 7.4, containing

0.05% Tween-20 (PBST).

3.3.4 Quantitative ELISA

Ninety-six-well microtiter plates (High-binding; Corning Inc. Life Sciences,

Lowell, MA) were coated overnight at 4°C with 0.3125 ug/mL of mouse anti-human IgG

y-chain specific antibody (Sigma-Aldrich) diluted in phosphate-buffered saline (PBS) at

pH 7.4. Plates were blocked with 4% (w/v) skim milk (EMD Biosciences, Newark, NJ)

dissolved in PBS for 24 h at 4°C and then washed five times with PBST. Serial dilutions

of clarified extract from N. benthamiana plants expressing trastuzumab were added to the

plate, which was incubated at 37°C for 1 h. Serial dilutions of human myeloma IgGl

(Athens Research & Technology Inc., Athens, GA), which is of the same antibody

isotype as Herceptin , were used as a quantification standard with 10 jug of TSP from

untreated N. benthamiana plants. The plate was washed five times with PBST before

adding polyclonal rabbit anti-human IgG (H + L)-horseradish peroxidase (HRP)

conjugate (Abeam, Cambridge, MA), diluted to 1 (ag/mL in PBS, for 1 h at 37°C. The

plate was washed five times with PBST before development with 1-Step™ Turbo TMB-

ELISA (Thermo Scientific). Color development was stopped with 1.8 M sulfuric acid,

and optical densities were measured at 450 nm using an En Vision 2100 Multilabel

microtiter plate reader (Perkin-Elmer, Woodbridge, Canada). Quantitative ELISAs to

determine the expression of antibody in plants were performed in triplicate. Three

independent expression experiments were performed, involving a total of 11 plants.

50

Overall expression of the antibody is reported as the average ± standard error of the mean

for all 11 plants.

3.3.5 Antibody Purification

Infiltrated N. benthamiana leaf tissue was harvested 8 d.p.i and stored at -80°C.

Frozen leaf tissue (250 g) was combined with two volumes (500 mL) of cold extraction

buffer in a food processor (Morphy Richards Inc., Mexborough, South Yorkshire, United

Kingdom) and disrupted for three 30 s pulses. Disrupted tissue was collected and

homogenized further using a benchtop Polytron® homogenizer (PT10/35, Kinematica

Inc., Bohemia, NY). Large plant debris was removed from the homogenate by dead-end

filtration through miracloth (Calbiochem, San Diego, CA). Solid ammonium sulfate was

slowly added to the filtered homogenate to a final concentration of 20%. The plant

homogenate was then incubated at 4°C for 1 h with gentle stirring. Insoluble material was

pelleted by centrifugation at 10,000 rpm for 30 min at 4°C and the resulting supernatant

collected. The concentration of ammonium sulfate in the resulting supernatant was

subsequently increased to 60%, incubated at 4°C for 2 hs with gentle stirring, and

centrifuged at 10,000 rpm for 30 min at 4°C. Pelleted protein was resuspended in 250 mL

of 20 mM sodium phosphate, pH 7.0, and then passed through a series of filters (2.7 urn

glass microfiber (GF/D), 1.2 urn glass microfiber (GF/C), 0.8 urn cellulose acetate, 0.45

um cellulose acetate; Whatman, Piscataway, NJ). The protein solution was dialyzed and

concentrated in a 250-mL Amicon ultrafiltration stirred cell (Millipore, Billerica, MA)

fitted with a molecular cutoff membrane of 30 kDa (Millipore), then applied (4 mL/min)

to a chromatography column (ID = 2.5 cm; Bio-Rad) containing 10 mL of Protein G

Sepharose 4 Fast Flow affinity media (GE Healthcare, Baie d'Urfe, Canada) pre-

51

equilibrated with 20 mM phosphate buffer at pH 7.0. A series of washings were

performed with 20 mM phosphate buffer, pH 7.0, to ensure the removal of all

contaminating solutes from the Protein G column. The antibody was eluted from the

column with 0.1 M glycine, pH 2.2, and immediately buffered with 1 M TrisCl, pH 9.0.

The buffered eluate was subsequently applied (2.5 mL/min) to a Protein A affinity

column (5 mL HiTrap™ Protein A HP column, GE Healthcare) connected to an AKTA-

FPLC (Amersham Pharmacia Biotech, Uppsala, Sweden). To ensure the removal of all

contaminating solutes from the Protein A column, a series of washings were performed

with 20 mM phosphate buffer, pH 7.0. The antibody was eluted from the column with 0.1

M glycine, pH 2.2, and immediately buffered with 1 M TrisCl, pH 9.0. The antibody

eluate was dialyzed against 20 mM phosphate buffer, pH 7.0, and concentrated using

polyethylene glycol 35,000. Coomassie-stained SDS-PAGE gels and western

immunoblots were used to analyze the purity and structural integrity of plant-produced

trastuzumab.

3.3.6 N-Terminal Sequence Analysis

Plant-purified trastuzumab (3 |u,g) was separated by reducing 12% SDS-PAGE

and then transferred to a Sequi-Blot™ PVDF membrane (Bio-Rad) which was treated

with Coomassie blue G-250. N-terminal sequencing analysis (Edman degradation) was

performed at the Hospital for Sick Children's Research Institute (The Advanced Protein

Technology Centre, University of Toronto, Canada).

3.3.7 Cell Culture

MCF-7, SK-BR-3 and BT-474 cell lysates were prepared from cell lines grown to

95% confluence. Cells were treated with a lx trypsin-EDTA solution (0.25% trypsin,

52

0.1% EDTA; SAFC Biosciences, Lenexa, KS) for dissociation from the cell culture

flasks, washed twice with ice-cold PBS, and lysed with NP40 cell lysis buffer (50 mM

Tris, pH 7.4, 250 mM NaCl, 5 mM EDTA, 50 mM NaF, 1 mM Na3V04, 1% Nonidet

P40, and 0.02% NaN3; Invitrogen) supplemented with 1 mM phenylmethanesulfonyl

fluoride solution (PMSF; Sigma Aldrich) and 10% protease inhibitor cocktail (4-[2-

aminoethyljbenzenesulfonyl fluoride, N-ftoms'-epoxysuccinylJ-L-leucine 4-

guanidinobutylamide, bestatin hydrochloride, leupeptin hemisulfate salt, aprotinin and

sodium EDTA; Sigma-Aldrich). TSP concentration was determined for each lysate using

the BCA Protein Assay (Thermo Scientific). Through western immunoblot analysis,

HER2 was detected in the cell lysate preparations using 0.1 |a,g/mL of either Herceptin®

or plant-produced trastuzumab in PBST. Antibody samples were detected using a mixture

of goat anti-human IgG y- and K-chain specific probes conjugated to alkaline phosphatase

(Sigma-Aldrich), diluted to 1:2500 in PBST.

3.3.8 Cell Proliferation Assay

MCF-7 and SK-BR-3 cell lines were cultured in Dulbecco's modified Eagle's

medium (DMEM) supplemented with 1 mg/mL fungizone, 1% penicillin/streptomycin

(all from Invitrogen), and 10% fetal bovine serum (FBS; Sigma-Aldrich). The BT-474

cell line was cultured in Roswell Park Memorial Institute (RPMI) 1640 basal medium

(Invitrogen) supplemented with 1 mg/mL fungizone, 1% penicillin/streptomycin, and

10% FBS. SK-BR-3, BT-474 and MCF7 cells were seeded into 6-well plates (Corning,

Lowell, MA) (5 xlO4 cells/well). After allowing the cells to adhere, the cells were treated

with 2 ug/mL of non-specific plant-purified human IgGl (negative control; human

myeloma IgGl [Athens Research and Technology] was spiked into untreated N.

53

benthamiana plant extract and subsequently purified using the same scheme developed

for purification plant-produced trastuzumab), 2 ug/mL of plant-produced trastuzumab, or

2 [4g/mL of Herceptin®; untreated cells were also included as a control. Relative cell

proliferation was determined by viable cell counts using trypan blue stain (Invitrogen).

Cell counts were performed every two days for a total of eight days. Data are expressed

as a percentage of the untreated control.

3.4 Results

3.4.1 Accumulation of Trastuzumab in N. benthamiana Plants

Trastuzumab was expressed in N. benthamiana plants using the magnlCON®

viral-based transient expression system (Giritch et al., 2006). Six-week old N.

benthamiana plants were vacuum-infiltrated with A. tumefaciens clones transformed with

provectors containing the HC- and LC-coding sequences of trastuzumab. Results of

preliminary experiments determined that the murine signal peptide (SP) did not allow

much accumulation of trastuzumab; therefore, the murine SP was replaced by the

Arabidopsis basic Chitinase SP on both HC- and LC-expression constructs. The assembly

of trastuzumab with the Arabidopsis SP-containing constructs was examined 8 d.p.i on a

non-reducing western immunoblot treated with a mixture of anti-human IgG y- and K-

chain specific probes. As shown in Figure 3.2, the tetrameric form of the antibody (H2L2)

was the most prominent band. Trastuzumab expression was also determined by non-

reducing immunoblot and confirmed by quantitative ELISA through comparison with

known concentrations of a human IgGl standard. Plants expressed an average of 43.3 ±

4.7 mg of trastuzumab per kilogram of fresh leaf tissue (0.59 ± 0.08% total soluble

protein; TSP).

54

1 2 3 4 5 6 7 8 9 10 11 12

ml mtffg.**»**"*'"••- - l^fdtf ***

MW (kDa)

250 — 150— 1 0 0 — 75 —

50 —

3 7 —

25 _ 20

Figure 3.2 Quantification of trastuzumab expression in N. benthamiana. Trastuzumab was expressed in N. benthamiana using a viral vector-based transient expression system. Crude plant extracts were analyzed on a non-reducing immunoblot probed with a mixture of anti-human IgG y- and K-chain specific probes. Lane 1: protein molecular weight standard; Lane 2-8: human IgGl, 1000, 500, 250, 125, 62.5, 31.3, 15.1 ng, respectively + 10 |o,g total soluble protein (TSP) from untreated N. benthamiana; Lane 9: 10 ug TSP from untreated N. benthamiana; Lane 10-12: 10 \xg TSP from three replicate N. benthamiana plants expressing trastuzumab. Molecular weights of protein standards are indicated on the left. The tetrameric (H2L2) form of plant-produced trastuzumab is indicated by the arrow on the right.

3.4.2 Purification and Characterization of Plant-Produced Trastuzumab

A purification scheme was developed to facilitate the recovery of trastuzumab

from N. benthamiana plants. Primary plant extracts were treated with 20% ammonium

sulfate to remove high molecular weight contaminants, followed by 60% ammonium

sulfate to enrich antibody yield through precipitation. Trastuzumab was subsequently

purified by both Protein G and then Protein A affinity chromatography. Purified plant-

produced trastuzumab was compared with Herceptin using SDS-PAGE under reducing

conditions followed by staining with Coomassie blue. As seen in Figure 3.3A, the two

55

major bands observed at ca. 50 kDa and 25 kDa are the heavy and light chains of

trastuzumab, respectively. The heavy chain of plant-produced trastuzumab migrated

slightly faster than the heavy chain of Herceptin®, likely due to differences between plant

and mammalian post-translational glycosylation. As expected, there were no detectable

differences in the electrophoretic mobilities of the light chains of Herceptin and plant-

produced trastuzumab. In addition to the bands representing the heavy and light chains of

trastuzumab, two less prominent bands were observed that migrated between the 25 and

37 kDa markers; these bands were enhanced by immunoblotting (Figure 3.3B). A series

of non-reducing SDS-PAGE gels and immunoblots probed with y- or K-chain specific

probes revealed that these were heavy chain degradation products (not shown), likely

produced by protein degradation in planta since the addition of a protease inhibitor

cocktail to the extraction buffer had no effect on the antibody banding pattern in crude

plant extracts (not shown).

N-terminal sequencing by Edman degradation indicated 100% cleavage of the

Arabidopsis SP from both the heavy- and light-chains of the plant-produced trastuzumab.

The N-termini of both the HC and LC polypeptides of plant-produced trastuzumab (Glu-

Val-Gln-Leu-Val-Glu and Asp-Ile-Gln-Met-Thr-Gln, respectively) are identical to those

of Herceptin (Drug bank accession # BTD00098). This result, in combination with the

similarity in sizes between the HC and LC of plant-produced trastuzumab and

Herceptin®, strongly suggests that both mAbs have identical primary structures.

3.4.3 Specificity of Plant-Produced Trastuzumab

Qualitative binding analyses were performed to demonstrate the specificity of

plant-produced trastuzumab for HER2. MCF-7 and BT-474 cell lysates were resolved on

56

a western immunoblot that was subsequently probed with either plant-produced

trastuzumab or Herceptin . One major band was observed between 100 and 150 kDa on

both of the immunoblots probed with either mAb (Figure 3.4). The single band on both

immunoblots corresponds to HER2 from the BT-474 cell lysates. A large smear and a

less prominent band between 25 and 50 kDa were also observed on both immunoblots

and likely represent artifacts of the cell lysate preparation procedure. No bands were

observed in the lane containing the MCF-7 cells lysates, as this cell line does not

overexpress HER2.

MW MW

(kDa) 1 2 3 ( k°a)

250 , „ . 250 •

100 «—» 100 75 mmmm 75

50 mamk^gm^^^ 50

37 ssitil* 37

mi 2 5 ,___™ :^^^ 2 5

15 „ . _ . 15

Figure 3.3 Analysis of the purity of plant-produced trastuzumab. Reducing, Coomassie stained SDS-PAGE (A) and immunoblot (B). Lane 1: protein molecular weight standard; Lane 2: Herceptin®, 1.2 \xg; Lane 3: plant-produced trastuzumab, 1.2 ug. Immunoblot was probed with a mixture of anti-human IgG y- and K-chain specific probes. Molecular weights of protein standards are indicated on the left.

57

A B M W M W (kDa) 1 2 3 (kDa)

250 250 -150 150 -100 *"** 100 -75 75 -

50 50 -

37 37 -

25 25 -

20 20

Figure 3.4 Qualitative analysis of the binding of plant-produced trastuzumab to HER2 ligand. MCF-7 and BT-474 cell lysates were analyzed by non-reducing immunoblots probed with Herceptin® (A) or plant-produced trastuzumab (B). Lane 1: protein standard; Lane 2-3: 25 ug TSP of the MCF-7 and the BT-474 cell lysates, respectively.

3.4.4 Inhibition of Tumor Cell Proliferation

The effect of plant-produced trastuzumab on the growth of breast tumor cells that

overexpress HER2 was examined using a cell proliferation assay. Both HER2

overexpressing tumor cells (BT-474 and SK-BR-3) and normal HER2 expressing tumor

cells (MCF-7) were treated with Herceptin® or plant-produced trastuzumab. After 8 days,

both plant-produced trastuzumab and Herceptin® showed 52.5% and 48.8% inhibition of

BT-474 cell proliferation, respectively (Figure 3.5A). After 4 days, plant-produced

trastuzumab and Herceptin® showed 47.1% and 47.8% inhibition of SK-BR-3 cell

proliferation, respectively (Figure 3.5B). The growth of SK-BR-3 cells, but not BT-474

cells, rose after 6 days of treatment with both plant-purified trastuzumab and Herceptin®

(Figure 3.5B). This could be explained by the fact that SK-BR-3 cells have

58

A B

o

o

120

100

120

100

a so • "5 O H 60 • <a

o. 40 •

20 •

0

• • • • Human lgG1 from plant extract (-ve control)

• - A — Plant-produced trastuzumab

- B — Commercial trastuzumab

2 4 6 8

Time (days)

Figure 3.5 Effect of plant-produced trastuzumab on the proliferation of human breast tumor cells that overexpress HER2. BT-474 (A), SK-BR-3 (B) and MCF7 (C) cells were seeded into 6-well plates (5 x 104 cells/well) and treated with 2 p,g/mL of non-specific plant-purified human IgGl (negative control), 2 |j,g/mL of plant-produced trastuzumab or 2 (j.g/mL of Herceptin®. Cell counts were performed every two days to determine the relative cell proliferation. Data are expressed as a percentage of untreated control and are presented as means of triplicates ± SEM.

approximately two times more HER2 on their cell surfaces than BT-474 cells (Lewis et

al., 1993). As shown in Figure 3.5C, plant-produced trastuzumab and Herceptin® had no

anti-proliferative effect on MCF-7 cells. Thus, plant-produced trastuzumab selectively

59

inhibits the proliferation of both BT-474 and SK-BR-3 cells. As a negative control, all

breast tumor cell lines were also treated with a non-specific plant-purified human IgGl.

This non-specific plant-purified antibody had no effect on the breast tumor cell

proliferation, which demonstrates the absence of plant contaminants that could inhibit the

proliferation of breast tumor cells.

3.5 Discussion

Numerous researchers have shown that plant-produced mAbs retain biological

activities (i.e., specificity, cytotoxicity, and neutralization activity) that are similar to

parental mAbs produced in mammalian cell culture (reviewed in refs (De Muynck et al.,

2010; Fischer et al., 2009); however, no study has yet been conducted to characterize and

compare a plant-produced mAb to a clinically approved therapeutic antibody with the

identical primary structure. TheraCIM®, an anti-epidermal growth factor receptor (EGF-

R) antibody with conditional registry approval in Cuba, is a clinically approved mAb that

has also been produced in plants (Rodriguez et al , 2005). Although it was determined

that the plant-produced antibody and TheraCIM® have binding similar to A431 human-

tumor-culture cells, the plant-produced antibody was modified to remove glycosylation

sites and to add a KDEL ER-retention signal (Rodriguez et al., 2005). Another plant-

produced antibody, anti-HIV mAb 2G12, will soon enter human clinical trials (Gilbert,

2009; Ramessar et al., 2008), but its parent antibody has not yet had clinical approval.

Our research on the expression and purification of the anti-breast cancer antibody

trastuzumab contributes further evidence that plants can be used for the production of

biosimilar therapeutic mAbs, as we were able to produce trastuzumab in N. benthamiana

with identical primary structures to those of its mammalian cell-derived counterpart.

60

Although analysis of our primary plant extracts revealed that N. benthamiana plants

express an average of 43.3 ± 4.7 mg of trastuzumab per kg of fresh weight (0.59 ± 0.08%

TSP), optimization of this expression system should allow the production of 500 mg to 5

g per kg fresh weight (Bendandi et al., 2010; Giritch et al., 2006). Antibody expression

levels increase in a time-dependent manner, but antibody stability can also decrease over

time; heavy- and light-chain polypeptide expression levels increase until 10-11 d.p.i., and

antibody stability begins to decrease at 8-9 d.p.i (Giritch et al., 2006). We chose to

harvest plants 8 d.p.i. in order to obtain the most full-length IgG with the lowest amount

of antibody fragments/breakdown products.

A purification scheme was developed to facilitate the recovery of trastuzumab

from plants. Although analysis of plant-purified trastuzumab revealed the presence of

heavy chain degradation products, these products likely result from proteolytic

degradation inplanta (supported by the research in ref (Sharp and Doran, 2001)). These

impurities could thus be removed using an additional chromatography step such as gel

filtration or ion exchange chromatography. Most importantly, plant-produced

trastuzumab was found to have specificity similar to that of Herceptin for binding to

HER2 and was determined to be as effective as Herceptin® in inhibiting the growth of

cells overexpressing HER2.

One of the remaining major limitations of producing therapeutic mAbs in plants is

the addition plant-specific 7V-glycans, which may induce an immune response in human

treatment, especially with repeated immunotherapy (Cabanes-Macheteau et al., 1999;

Gomord et al., 2004; Sack et al., 2007). Antibody glycosylation is also essential for

structural stability, decreased protease sensitivity, complement activation, and effector

61

function (Cabanes-Macheteau et al., 1999; McLean et al., 2007). Several strategies are

currently being developed to generate genetically modified plants with humanized N-

glycan profiles (Strasser et al., 2008, 2009; Vezina et al., 2009). These strategies include

knocking out the endogenous glycosyltransferases responsible for the addition of the

plant-specific TV-glycans pl,2-xylose and al,3-fucose (Strasser et al., 2008, 2009) as well

as the expression of pi,4-galactosyltransferase (GalT) for the addition of terminal (31,4-

galactose (Bakker et al., 2006; Frey et al., 2009; Vezina et al., 2009). Future research on

plant-produced trastuzumab will require either the examination of the effect of plant-

specific ./V-glycans on its therapeutic efficacy in vivo as well as its potential

immunogenicity or the production of it in host plants modified to express proteins with

mammalian-like glycosylations.

Validation of plant-produced mAbs as biosimilar therapeutics will require that

they be shown to have biological properties (i.e., bioactivity and biosafety) similar to

clinically-approved parental mAbs. This paper clearly shows that a plant-expression and

purification system can produce a therapeutic mAb with identical primary structures and

similar in vitro bioactivities to those of its clinically approved parental mAb, indicating

that agricultural biopharming could be an effective alternative to mammalian cell systems

for the production of biosimilar therapeutics such as mAbs.

3.6 Acknowledgements

We thank ICON Genetics, GmbH for the use of the magnlCON® expression

system. We are also grateful to Fernando Olea-Popelka for designing the plant-optimized

HC and LC coding sequences and to Ashley J. Meyers for her help in the laboratory, for

the development of the quantitative ELISA, and for critical review of this manuscript.

62

This work was funded by grants to JCH from the Ontario Ministry of Agriculture, Food

and Rural Affairs (OMAFRA), the Natural Sciences and Engineering Research Council

(NSERC), the Canada Research Chairs (CRC) Program, the SENTINEL Bioactive Paper

Network and PlantForm Corporation.

63

4 RESEARCH CHAPTER 2: PURIFICATION OF A PLANT-PRODUCED ANTIBODY USING HOLLOW FIBER TANGENTIAL FLOW MICROFILTRATION

4.1 Abstract

Genetically modified plants can be used as an alternative to traditional

mammalian cell expression systems for the large-scale production of therapeutic

antibodies. Current purification of mAbs from plant bioreactors, however, remains far

from ideal due to inefficient post-harvest processing and purification procedures. A

scalable scheme that combines hollow fiber tangential flow microfiltration with Protein A

affinity chromatography and SP Sepharose cation exchange chromatography was thus

developed. Initial clarification of crude plant extracts by hollow fiber tangential flow

microfiltration was effective at removing extraneous plant proteins and compounds

without a significant loss of trastuzumab (< 4%). Through a single Protein A affinity

chromatography step, 67% of plant-produced trastuzumab was recovered. A high level of

antibody purity was also achieved with minimal column fouling. Plant-produced

trastuzumab was subsequently polished by SP Sepharose cation exchange

chromatography. The purity and antibody banding-pattern of plant-produced trastuzumab

was confirmed by immunoblotting and was determined to be comparable to Herceptin®.

Results confirm that hollow fiber tangential flow microfiltration, Protein A affinity

chromatography, and SP sepharose cation exchange chromatography can be used to

achieve purification of therapeutic antibodies from plants.

64

4.2 Introduction

Over the past 20 years, genetically modified plants have shown tremendous

potential for the large-scale production of therapeutic antibodies. Compared to traditional

mammalian cell-expression systems, genetically modified plants provide the advantages

of decreased upstream production costs, scalability of agricultural production, and

biological safety due to the absence of animal pathogens and toxins (Arntzen, 2008;

Huang et al., 2010). Transient-based expression systems have also allowed for higher

expression levels to be achieved faster than with any other biopharmaceutical production

system (Bendandi et al., 2010; Faye and Gomord, 2010; Giritch et al., 2006). However,

despite the successful expression of a wide variety of novel and biosimilar therapeutic

antibodies in plants (De Muynck et al., 2010), the agricultural production of antibodies

has yet to be adopted by the pharmaceutical industry (Faye and Gomord, 2010).

The large-scale production of antibodies in plants is currently limited by costly

post-harvest processing and purification procedures, which can account for more than

80% of the total cost of plant biopharming (Evangelista et al., 1998; Hassan et al., 2008;

Mison et al., 2000). Inefficiency arises from the numerous purification steps that are

required to separate plant-produced antibodies from the complex mixture of plant

proteins, alkaloids, pigments, polyphenols, and mucilages (Platis and Labrou, 2006;

Valdes et al., 2003). Using a series of low efficiency clarification and concentration steps

also leads to greater product loss, longer processing times, and higher costs (Aguilar and

Rito-Palomares, 2010; Platis and Labrou, 2006; Pujol et al., 2005). Thus, the overall

economics of large-scale agricultural production and purification of therapeutic

antibodies remains to be determined.

65

To improve the purification of antibodies from plants, the inefficiency of initial

post-harvest clarification and concentration procedures must be addressed. The fewer

clarification and purification steps that are required, the more economical and efficient

the overall manufacturing process will become. Clarification of crude plant extracts is an

essential part of any purification scheme that employs chromatographic separation. Direct

application of a crude plant extract to a chromatography column causes fouling due to

non-specific binding of plant proteins and other compounds to the resin (Hassan et al.,

2008). As a result, column life and binding capacity are reduced, while overall production

costs are increased (Hassan et al., 2008; Pujol et al., 2005). A clarification step reduces

the burden of plant extracts on a chromatography column by removing extraneous

particles and macromolecules. Many techniques can be used, alone or in combination, to

clarify plant extracts including acidic precipitation (Woodard et al., 2009), ammonium

sulfate precipitation (Grohs et al., 2010; Huang et al., 2010), centrifugation, filtration, and

two-phase aqueous partitioning (Platis and Labrou, 2006; Platis et al., 2008). However,

many of these techniques can be costly, time consuming and are not easily scalable.

Tangential flow (cross-flow) microfiltration has been used as an alternative to

centrifugation, depth filtration, and expanded-bed chromatography for the initial

purification of therapeutic proteins from mammalian cell culture (van Reis and Zydney,

2001). In tangential flow microfiltration, the feed stream passes parallel to the membrane,

which reduces the adsorption of proteins and compounds to the surface of the membrane

(Hill and Bender, 2007). Tangential flow microfiltration can thus be used to clarify hard-

to-filter solutions such as crude plant extracts, which quickly cause membrane fouling

when conventional dead-end filtration methods are used (Stoger et al., 2004).

66

Furthermore, clarification by tangential flow microfiltration does not require any phase,

pH or temperature change, which can affect antibody stability and/or structural integrity

(i.e. through protein denaturation, deactivation, and/or degradation) (Zeman and Zydney,

1996a). Several different membrane configurations can be used for tangential flow

filtration, including hollow fiber modules, flat plate modules, spiral wound modules, and

tubular devices (Zeman and Zydney, 1996b). Hollow fiber modules are composed of

narrow-bore cylindrical membranes (fibers) and thus offer higher packing densities. The

small diameter of the fibers also allows high mass transfer rates to be achieved at low

volumetric flow rates. Conversely, hollow fiber modules are typically more susceptible to

plugging and can only be manufactured from certain polymers (Zeman and Zydney,

1996b).

In our previous research, the anti-breast cancer antibody, trastuzumab (Herceptin®),

was successfully expressed in Nicotiana benthamiana (Grohs et al., 2010). Purification of

plant-produced trastuzumab was achieved using a combination of ammonium sulfate

precipitation, Protein G and then Protein A affinity chromatography. Plant-purified

trastuzumab was determined to be just as effective as the innovator drug Herceptin® at

inhibiting the proliferation of HER2-overexpressing breast cancer cells. However, the

purification of plant-produced trastuzumab was impeded by long processing times.

Analysis of plant-purified trastuzumab also revealed the presence of antibody fragments

and/or degradation products, which were not removed by Protein G and/or Protein A

chromatography. Therefore, the purpose of this study was to develop a more efficient

scheme to purify and polish plant-produced trastuzumab. By combining hollow fiber

tangential flow microfiltration with Protein A affinity chromatography and SP sepharose

67

cation-exchange chromatography, trastuzumab was successfully purified from N.

benthamiana plants. Through immunoblot analysis, plant-purified trastuzumab was

determined to have an antibody-banding pattern and purity comparable to Herceptin®.

4.3 Materials and Methods

4.3.1 Plant Material

N. benthamiana plants expressing trastuzumab, an anti-HER2/new humanized

murine IgGlK antibody, were previously developed by Grohs et al. (2010), using the

magnlCON® expression system (Giritch et al., 2006; Icon Genetics).

4.3.2 Extraction and Clarification

Infiltrated N. benthamiana leaf tissue was harvested and extracted as previously

described in Grohs et al. (2010). Briefly, 100 g of frozen leaf tissue was disrupted in a

food processor containing 200 mL of cold extraction buffer [40 mM phosphate buffer, 50

mM ascorbic acid, and 10 mM ethylenediaminetetraacetic acid (EDTA) disodium salt

dihydrate at pH 7.0] for three 30 s pulses and subsequently homogenized using a

Polytron® homogenizer for 4 min 30 s. Insoluble material was removed by vacuum

filtration through two layers of Miracloth (EMD Biosciences, La Jolla, CA), followed by

centrifugation in a Sorvall RC-5B centrifuge (Sorvall GSA rotor; Thermo Scientific,

Nepean, Canada) at 12,000 rpm for 30 min at 4°C. The supernatant was collected and

adjusted to 350 mL with cold extraction buffer.

Micro filtration of the plant extract was achieved by passing the extract through a

hollow fiber tangential flow filtration (TFF) module with a pore rating of 0.05 um

(MiniKros® sampler M10S-320-01P; Spectrum Laboratories Inc., Rancho Dominguez,

CA) connected to a Watson Marlow 520SN/R2 peristaltic pump (Watson Marlow,

68

Georgetown, Canada) (Figure 4.1). Inlet, retentate, and permeate pressure were

monitored using three pressure transducers and KrosFlo® Digital Pressure Monitor

(Spectrum Laboratories Inc.). Permeate flow rate was monitored using a KrosFlo

Permeate Scale (Spectrum Laboratories Inc.). KF Comm Data Collection Software

(Spectrum Laboratories Inc.) was used to collect real time process data. Specifications for

the TFF module and system parameters are outlined in Table 4.1. Following each use,

the TFF module was cleaned using 0.5 M NaOH and then stored in 0.1 M NaOH at 4°C.

Prior to each use, the 0.1 M NaOH was drained from the module and the system was

flushed with 40 mM sodium phosphate buffer, pH 7. The module was subsequently

treated with 70% isopropyl alcohol for 1 h and then flushed with ultra-pure water (18.2

flM-cm) from a Diamond NANOpure water purification unit (Barnstead International,

Dubuque, IA).

PUMP

M

3—(p?y

© PERMEATE

PROCESS RESERVOIR

Figure 4.1 Schematic of the hollow fiber tangential flow microfiltration system flow-path. Pi, inlet pressure transducer; PR, retentate pressure transducer; PP, permeate pressure transducer; M, hollow fiber module.

69

Table 4.1 Specifications for the hollow fiber tangential flow filtration module (M10S-320-01P) used in the purification of trastuzumab from N. benthamiana.

Module Parameters

Filter type Area Fiber I.D. Pore rating

Polysulphone (PS) 420 cm2

0.5 mm 0.05 urn

System Parameters

Recirculation rate Shear Transmembrane pressure Average permeate flux

(TMP)

1150 mL/min 12,000 s"1

5 psi 27.1 LMH1

LMH: liters per square meter per hour.

4.3.3 Chromatography

The clarified plant extract was applied to a Protein A affinity column (5 mL

HiTrap™ Protein A HP column, GE Healthcare) at a flow rate of 1 mL/min using an

AKTA-FPLC system (Amersham Pharmacia Biotech, Uppsala, Sweden) at ambient

temperature. Following a series of washings with 20 mM phosphate buffer, pH 7.0, the

antibody was eluted from the column with 0.1 M glycine, pH 2.2, and immediately

neutralized with 1 M TrisCl, pH 9.0. The Protein A eluate was dialyzed against 20 mM

sodium acetate buffer, pH 5.0 using a 50-mL Amicon ultrafiltration stirred cell

(Millipore, Billerica, MA) fitted with a molecular cutoff membrane of 100 kDa

(Millipore).

The dialyzed Protein A eluate was applied to a chromatography column (ID = 2.5

cm; Bio-Rad) containing 25 mL of SP Sepharose cation exchange media (GE Healthcare,

Baie d'Urfe, QC) at a flow rate of 2.0 mL/min at ambient temperature. After washing the

70

column with 20 mM sodium acetate, pH 5.0, a multi-step elution (0, 70, 80, 100, 125, 250

mM, and 1 M NaCl) was performed using 20 mM sodium acetate, pH 5.0 containing 1 M

NaCl. All fractions collected throughout the SP Sepharose purification were analyzed by

western immunoblot. Fractions containing tetramer (H2L2) were pooled and compared to

Herceptin® (Roche, Mississauga, ON).

4.3.4 SDS PAGE and Immunoblot Analyses

Coomassie-stained SDS-PAGE gels and western immunoblots were used to

analyze each step of the purification scheme and to determine the integrity and purity of

plant-produced trastuzumab. TSP concentrations were determined with the Bio-Rad

Protein Assay, using Bovine Serum Albumin (BSA) (Thermo Scientific) as the protein

standard. Western immunoblots were probed using a mixture of goat anti-human IgG y-

and K-chain specific probes conjugated to alkaline phosphatase (Sigma-Aldrich; 1.5 uL of

eachinlOmLofPBST).

4.3.5 Quantitative ELISA

Antibody recovery was quantified using a direct sandwich ELISA. The quantitative

ELISA was performed as outlined in Grohs et al. (2010), with several modifications.

Briefly, microtiter plates (Corning Inc. Life Sciences, Lowell, MA) were coated with

0.625 ug/mL of mouse anti-human IgG y-chain specific antibody (Sigma-Aldrich) for 16

h at 4°C. Plates were blocked with 4% (w/v) skim milk (EMD Biosciences) for 5.5 h at

37°C and subsequently washed with PBST. Serial dilutions of samples collected

throughout the purification procedure were added to the plate, which was incubated at

37°C for 1 h. Serial dilutions of human myeloma IgGl (Athens Research & Technology

71

Inc., Athens, GA) were used as the standard. Antibody samples were detected with 1.5

(0,g/mL of polyclonal rabbit anti-human IgG (H+L)-horseradish peroxidase (HRP)

conjugate (Abeam, Cambridge, MA) for 1 h at 37°C.

4.4 Results

4.4.1 Purification of Trastuzumab from N. benthamiana

The purification of plant-produced trastuzumab was previously achieved by

clarifying crude plant extracts using a two-step precipitation with 20% and then 60%

ammonium sulphate (Chapter 3; Grohs et al., 2010). Trastuzumab was subsequently

purified by Protein G and then Protein A affinity chromatography. The structural integrity

of plant-produced trastuzumab was analyzed on a non-reducing immunoblot treated with

a mixture of y- and K-chain specific probes (Figure 4.2). Plant-produced trastuzumab was

also compared to human IgGl, Herceptin® and human serum IgG. All antibody samples

contained bands with similar electrophoretic mobilities; however, plant-produced

trastuzumab had additional bands (marked by asterisks in Figure 4.2A). Further

examination of plant-produced trastuzumab on an immunoblot probed with a y-chain

specific probe revealed that the band at ca. 50 kDa represents unassembled heavy chains

and the two bands between 25 and 37 kDa represent heavy chain degradation products

(Figure 4.2B). Plant-produced trastuzumab was also examined on an immunoblot probed

with a K-chain specific probe. It appears that the bands at ca. 45 kDa and 25 kDa

represent Fab fragments and unassembled light chains, respectively (Figure 4.2C).

In an attempt to improve the purity of plant-produced trastuzumab and to reduce

processing time, trastuzumab was purified from N. benthamiana tissue according to the

scheme outlined in Figure 4.3A. Samples from each step of the purification scheme were

72

A B MW (kDa)

250 -150 -100 -75 -

50 -

37 -

MW (kDa)

250 -150 -100 -

75 -

50 -

37 -

25

20

15

25

20

15

M W (kDa)

250-150 -100 -75 -

50 -

37 -

25

20

15

ure 4.2 Structural integrity of plant-produced trastuzumab purified using a combination of ammonium sulfate precipitation, Protein G, and Protein A chromatography (described in Chapter 3; Grohs et al., 2010). Non-reducing immunoblot analyses of plant-produced trastuzumab, compared with human IgGl, Herceptin® and human serum IgG. Immunoblots were probed with a mixture of anti-human IgG y- and K-chain specific probes (A), y-chain specific probe (B), and K-chain specific probe (C). Lane 1 protein standard; Lane 2: blank; Lane 3: human IgGl, 250 ng; Lane 4 Herceptin®, 250 ng; Lane 5: plant-produced trastuzumab, 250 ng; Lane 6 human serum IgG, 250 ng. Molecular weights of protein standards are indicated on the left. Asterisks mark the additional bands of plant-produced trastuzumab.

73

B Harvest Leaf Tissue

Grind & Homogenize

-0-Centrifuge

Lane # 1

Microfiltration (TFF)

Permeate

Lane #2

Protein A Chromatography

Rctentate

Flow through Elution

Lane #3 Lane #4

41 Ultrafiltration

SP Sepharose Chromatography

250 150 100 75

50

37

25

20

M 1

^••' -<W # # # * • -S«$» *

M

250 — 150 — 100 — 75 —

50 —

37 —

25 20

Figure 4.3 Purification scheme for the recovery of trastuzumab from 100 g of frozen N. benthamiana tissue (A). TFF: tangential flow filtration; SP: sulphopropyl. Samples from each purification step were collected and analyzed by non-reducing western immunoblot (B) and Coomassie-stained gel (C). The immunoblot was probed with a mixture of anti-human IgG y-and K-chain specific probes. M: protein standard, molecular weights are indicated on the left. The tetrameric (H2L2) form of plant-produced trastuzumab is indicated by the arrow on the right. Asterisks mark the bands that resemble those shown in Figure 4.1, where plant-produced trastuzumab was purified using ammonium sulfate precipitation, Protein G, and Protein A chromatography.

74

collected and analyzed by non-reducing immunoblot and Coomassie-stained SDS-PAGE

gel (Figure 4.3B and 4.3C, respectively). The concentration of trastuzumab recovered at

each step was also quantified using a direct sandwich ELISA (Table 4.2).

The conditions required to efficiently extract trastuzumab from N. benthamiana

were previously optimized (Grohs et al., 2010). Briefly, frozen leaf tissue was ground in a

blender with two volumes of cold extraction buffer, followed by homogenization with a

benchtop Polytron® homogenizer. Crude plant extracts were subsequently filtered

through Miracloth and centrifuged to remove insoluble plant material. To improve the

clarification of crude plant extracts hollow fiber tangential flow microfiltration was used,

as opposed to ammonium sulphate precipitation (described in Chapter 3; Grohs et al.,

2010). Microfiltration of the plant extract reduced the amount of TSP by 3-fold, while

96% of the trastuzumab was recovered (Table 4.2). Consequently, this clarification

method significantly reduced processing time and increased recovery. Furthermore,

clarification of the plant extract by microfiltration did not allow antibody degradation to

occur (Figure 4.3B, lane 2), as there was no change in the antibody-banding pattern when

compared to the crude plant extract (Figure 4.3B, lane 1). This was also observed when

ammonium sulfate was used for clarification.

Table 4.2 Analysis of the recovery of trastuzumab from 100 g of TV. benthamiana.

Total Soluble Antibody Volume Total Antibody Protein Concentration (mL) Antibody Recovery

(mg/mL) (ng/mL) (mg) (%)

Crude plant extract 3.28 5.96 175 1.05 100

Microfiltration Permeate 1.05 5.64 179 1.01 96

Protein A Elution 0.02 34.36 20 0.70 67

75

Plant-produced trastuzumab was subsequently purified using Protein A affinity

chromatography (Figure 4.4). Antibody fragments and/or breakdown products were

detected in the column flow through (immunoblot; Figure 4.3B, lane 3) along with

unwanted plant proteins and other contamiants (gel; Figure 4.3C, lane 3). Plant-produced

trastuzumab was of high purity (Figure 4.3C, lane 4), since no plant proteins were

observed in the Protein A eluate (Figure 4.3C, lane 4); however, some antibody

fragments/degradation products still remained (i.e. between 37 and 50 kDa; marked by

asterisks in Figure 4.3 C).These antibody bands are similar to those found using the

previous purification scheme that combined ammonium sulfate precipitation, Protein G,

and Protein A chromatography (marked by asterisks in Figure 4.2A). Total antibody

recovery following Protein A chromatography was 67% (Table 4.2).

I £

o

< >

3500

3000

2500

2000

1500

1000

500

0

0 100 200 300 400

Effluent Volume (mL)

Figure 4.4 Purification of trastuzumab from clarified N. benthamiana extract using Protein A affinity chromatography (binding buffer: 20 mM sodium phosphate buffer, pH 7.0; elution buffer: 0.1 M glycine pH 2.2; flow rate: 1.0 mL/min).

76

4.4.2 Polishing of Plant-Purified Trastuzumab

SP Sepharose cation exchange chromatography was used to polish plant-produced

trastuzumab previously purified by Protein A chromatography (Figure 4.5). Selective

elution of the undesired lower molecular weight antibody fragments between 37 and 50

kDa was achieved using a multi-step elution with NaCl. Following analysis of each

fraction by non-reducing immunoblot, it was determined that antibody fragments were

selectively eluted from the column between 100 and 125 mM NaCl (Figure 4.6B-E).

Increasing the concentration of NaCl to 250 mM and then 1 M eluted the remainder of

the antibody (Figure 4.6E and F). Fractions containing the tetrameric form of

trastuzumab were pooled for analysis and comparison with Herceptin (dotted lines mark

pooled fractions in Figure 4.6D-F).

4.4.3 Characterization of Antibody Integrity and Purity

The integrity and purity of plant-produced trastuzumab was determined by non-

reducing and reducing immunoblot analysis (Figure 4.7). Under non-reducing

conditions, plant-produced trastuzumab had the same banding pattern as Herceptin®, with

the tetramer being the most prominent band in both samples (Figure 4.7A). Comparison

of plant-produced trastuzumab and Herceptin® under reducing conditions showed that

only bands corresponding to the heavy and light chains were present in both samples

(Figure 4.7B).

77

uv Conductivity Concentration of NaCl

^1 00

mAU

30.0

40.0

30.0

20.0

10.0

0.0

roS/cm

K0

•3.0

/L

•2.0

1.0

JO.O

200 400 600 ml

Figure 4.5 Polishing of plant-purified trastuzumab by SP Sepharose cation exchange chromatography (flow rate: 2.0 mL/min). 20 mM sodium acetate, pH 5.0 was used as the binding buffer. A multi-step elution (0, 70, 80, 100, 125, 250 mM, and 1 M NaCl - indicated by green line) was performed using 20 mM sodium acetate, pH 5.0 containing 1 M NaCl.

B H 70 m M 80 niM

250 150 100 75

50

37

25 20

100 mM

100 m M D

250 150 100 75

50

37

25

20

125 m M

»»»»««»

125 m M 250 m M

250 150 100 75

50

37

250 m M 100 m M f » « « 9 « » « » « 9 S

25

20

Non-reducing immunoblot analysis of the fractions collected throughout the SP Sepharose purification of plant-produced trastuzumab (A-F). Immunoblots were probed with a mixture of anti-human IgG y- and K-chain specific probes. The concentrations of NaCl used in the multi-step elution are indicated above the immunoblots. The fractions that were pooled for further analysis are indicated by the dotted lines above the immunoblots. M: protein standard, molecular weights are indicated on the left.

79

M 1 B

M 1

250 — 150 — 100 — 75 —

50 —

37 —

25 — 20 — 15 —

250 — 150 — 100 — 75 —

50 —

37 —

25 — 20 — 15 —

Figure 4.7 Non-reducing (A) and reducing (B) immunoblot analyses of the purity of plant-produced trastuzumab. M: protein standard, molecular weights are indicated on the left; Lane 1: blank; Lane 2: plant-produced trastuzumab; Lane 3: Herceptin®, 100 ng. Immunoblots were probed with a mixture of anti-human IgG y- and K-chain specific probes.

4.5 Discussion

Previous work by Grohs et al. (2010) showed that biosimilar trastuzumab was

successfully expressed in 7Y. benthamiana. However, the purification of plant-produced

trastuzumab was time consuming and antibody recovery was low. Furthermore,

comparison of plant-purified trastuzumab to Herceptin®, using a non-reducing

immunoblot, revealed that plant-produced trastuzumab had additional antibody fragments

and/or degradation products (marked by asterisks in Figure 4.2A). To improve the

purification of plant-produced trastuzumab, ammonium sulfate precipitation was replaced

by hollow fiber tangential flow microfiltration. Hollow fiber tangential flow

microfiltration successfully decreased the amount of TSP in crude plant extracts without

80

a significant loss of trastuzumab. The addition of this clarification step to the purification

scheme also allowed for antibody purification to be achieved with minimal column

fouling as there was no observable discoloration of the resin or decreased binding

capacity. As a result, the lifespan of the resin was extended, thus reducing the costs

associated with replacing fouled columns. Clarification of the plant extract was also

achieved three times faster than ammonium sulfate precipitation, thus decreasing the

exposure of the antibody to proteases and phenolics released during tissue disruption

(Fischer et al., 1999). Furthermore, the use of hollow fiber tangential flow microfiltration

reduced the number of chromatography steps (i.e. Protein A chromatography versus

Protein G and then Protein A chromatography) required to obtain antibody of comparable

purity Herceptin". An additional polishing step with the SP Sepharose cation exchange

resin was added to remove undesirable antibody fragments, which were not removed by

Protein A chromatography (Figure 4.3B) or by using the method of Grohs et al. (2010)

(Figure 4.2A). Immunoblot analysis revealed the purity and antibody-banding pattern of

plant-produced trastuzumab to be comparable to Herceptin , when the new purification

scheme was used.

Costly downstream processing and purification procedures currently hinder the

large-scale production of antibodies in plants. Numerous strategies have been developed

to address the absence of efficient clarification and concentration procedures, which

contribute to the long processing times and the high cost of current purification schemes

(Aguilar and Rito-Palomares, 2010; Fischer et al., 1999; Platis and Labrou, 2006). An

aqueous two-phase partitioning system (ATPS), for example, has been optimized for the

purification of antibodies from transgenic tobacco plants (Platis and Labrou, 2006; Platis

81

et al., 2008). However, similar to the purification scheme developed for this study,

clarification of the crude plant extract by centrifugation and filtration was required prior

to the aqueous two-phase partitioning step (Platis and Labrou, 2006; Platis et al., 2008).

Flat membrane tangential flow microfiltration has also been used as an initial

clarification step in the purification of antibodies from plants (Fischer et al., 1999; Yu et

al., 2008a). Compared to hollow fiber modules, flat plate modules have lower membrane

packing densities (i.e. ratio of membrane area to device volume) and mass transfer rates

(Zeman and Zydney, 1996b). In contrast, hollow fiber modules are typically more

susceptible to plugging (i.e. plugging of hollow fibers) and can only be manufactured

from certain polymers (Zeman and Zydney, 1996b). The hollow fiber module used in this

study, for example, was made of polysulphone, a material with high protein binding

properties. As a result, the permeation flux (LMH) decreased over time and the module

had to be thoroughly cleaned with NaOH after each use.

Validation of plant bioreactors for the production of therapeutic antibodies will

require a purification scheme that is identical or superior to those currently used for

mammalian cell culture systems (Woodard et al., 2009). Our initial clarification of the

plant extract through hollow fiber tangential flow microfiltration allowed for the use of a

purification scheme identical to the standard purification sequence of Genentech, which

is used to purify antibodies from animal cell-based bioreactors (i.e., centrifugation, depth

filtration, Protein A chromatography, and cation-exchange chromatography) (Fontes and

van Reis, 2009). Although improvements and alternatives to Protein A chromatography

are currently being investigated (Fontes and van Reis, 2009), any adaptations to the

82

standard purification and polishing sequence could also be applied to the scheme

developed in this study.

Future research on this purification scheme will involve testing the scalability of

the tangential flow system to ensure removal of unwanted plant proteins and

contaminants while still achieving the same level or improved antibody recovery. The

versatility of the scheme will also be tested through purification of other plant-produced

therapeutic antibodies. A more versatile purification strategy for plant-produced

antibodies would facilitate process development by decreasing the required time and

resources (Shukla et al., 2007).

This study clearly shows that a purification system that combines hollow fiber

tangential flow filtration, Protein A affinity chromatography, and SP Sepharose cation

exchange chromatography can be used to generate plant-produced trastuzumab with the

same antibody-banding pattern and purity as Herceptin®. This study provides further

evidence that plant expression systems are effective alternatives to mammalian cell

systems for the efficient production of therapeutic mAbs.

4.6 Acknowledgements

We would like to thank ICON Genetics, GmbH for the use of the magnlCON®

expression system. We also thank Drs. Dennis Yu, Mukesh Muyani, and Raja Ghosh of

McMaster University for helpful discussions. This project was funded by grants to JCH

from the Ontario Ministry of Agriculture, Food and Rural Affairs (OMAFRA), the

Natural Sciences and Engineering Research Council (NSERC) of Canada, the Canada

Research Chairs (CRC) Program, the SENTINEL Bioactive Paper Network, and

PlantForm Corporation.

83

5 GENERAL CONCLUSIONS AND FUTURE DIRECTIONS

The first research objective of this thesis was to express biosimilar trastuzumab in

N. benthamiana plants using the magnlCON® viral-based transient expression system.

Immunoblot analyses of crude plant extracts revealed that trastuzumab accumulates

within plants mostly in the fully assembled tetrameric form. Using a series of ELISAs, N.

benthamiana plants were determined to express 43.3 ± 4.7 mg of trastuzumab per kg of

fresh leaf tissue (0.59 ± 0.08 % TSP).

Although low expression of biosimilar trastuzumab inplanta was sufficient for

preliminary characterizations, future research will require improvement of this expression

system since high antibody concentrations are required for plant bioreactors to achieve an

economic advantage over mammalian cell cultures (Evangelista et al., 1998; Sharp et al.,

2001). Antibody expression levels as high as 500 mg to 5 g per kg fresh weight of N.

benthamiana have previously been reported using the magnlCON® system (Bendandi et

al., 2010; Giritch et al., 2006). Therefore, optimization of this system should allow for

increased bioaccumulation of trastuzumab. In this study for example, it was found that

switching from a murine signal sequence to the Arabidopsis basic Chitinase signal

sequence increased expression of trastuzumab (data not shown). Plant development and

age can also have a profound effect on antibody expression and stability in planta

(Stevens et al., 2000). Plant growth conditions that favor increased biomass production

(i.e. light and temperature) could thus be examined to ensure optimum antibody

expression is achieved (Stevens et al., 2000). Finally, although the tetrameric form of

trastuzumab was the most prominent band observed through immunoblotting, decreasing

the concentration of lower molecular weight fragments would be ideal. Co-expression of

84

protease inhibitors in the apoplast might improve the structural integrity of plant-

produced trastuzumab and thus facilitate the final polishing step of the purification

scheme (Komarnytsky et al., 2006).

The second research objective of this thesis was to characterize plant-produced

trastuzumab and compare to the innovator drug Herceptin®. The structural integrity and

specificity of plant-produced trastuzumab was confirmed by immunoblotting. Plant-

produced trastuzumab and Herceptin® were found to bind the same target in extracts of

HER2 overexpressing cells, and functional assays revealed that both mAbs elicit similar

anti-proliferative effects on HER2-overexpressing breast cancer cells.

In the future, more extensive biochemical and biological analyses need to be

conducted to validate plant-produced trastuzumab. Analytical methods such as IEF,

reverse-phase HPLC (RP-HPLC), and SPR should be conducted to confirm the identity,

consistency, and stability of plant-produced trastuzumab using Herceptin® as the standard

(Beck et al., 2005). An analysis of the glycosylation pattern of plant-produced

trastuzumab could also be performed. Trastuzumab has numerous proposed mechanisms

of action including both cytostatic (i.e. cell cycle arrest, receptor endocytosis, inhibition

of HER2 ECD cleavage, and reduced tyrosine phosphorylation) and cytolytic effects (i.e.

ADCC). However, preliminary research in this thesis only examined the anti-proliferative

effects of plant-produced trastuzumab on breast tumor cells that overexpress HER2.

Conducting cell-based assays to examine the biological effect induced by plant-produced

trastuzumab (i.e. downstream signalling effects and immune effector functions) would

thus confirm that plant-produced trastuzumab retains the exact biological activity of

85

Herceptin . Following completion of these assays, a pre-clinical mouse trial should be

conducted to test the efficacy of the plant-produced antibody in vivo.

One of the shortcomings of antibody production in plants is the high costs

associated with post-harvest processing and purification procedures for plant-produced

antibodies. In the first half of this thesis purification of plant-produced trastuzumab was

achieved by treating primary plant extracts with 20% ammonium sulfate to remove high

molecular weight proteins, followed by 60% ammonium sulfate to enrich antibody yield

through precipitation. Trastuzumab was subsequently purified by Protein G and then

Protein A affinity chromatography. This purification scheme was limited by long

processing times and low antibody recovery. Plant-purified trastuzumab also contained

antibody fragments and/or breakdown products that were not present in Herceptin®. The

final research objective of this thesis was thus to improve the purity of plant-produced

trastuzumab. Initial clarification of crude plant extracts by hollow fiber tangential flow

microfiltration was effective at removing extraneous plant proteins and compounds

without a significant loss of plant-produced trastuzumab (< 4%). Through a single

Protein A affinity chromatography step, 67% of plant-produced trastuzumab was

recovered. Similar to the first purification scheme, antibody purification was achieved

with minimal column fouling. Finally, SP Sepharose cation exchange chromatography

was used to polish plant-produced trastuzumab.

Although trastuzumab was successfully purified using hollow fiber tangential

flow microfiltration, Protein A affinity chromatography, and SP Sepharose cation

exchange chromatography, numerous challenges remain. First, the tangential flow

microfiltration step must be scalable. Hollow fiber tangential flow microfiltration can

86

easily be scaled by increasing the filtration surface area, which is accomplished by

increasing the number of hollow fibers and the diameter of the module. However,

experiments should be conducted to ensure the accuracy of this statement, and to

optimize the system on a larger scale. Different types of filtration (i.e. continuous and

discontinuous filtration or concentration) can also be conducted using hollow fiber TFF

modules. A second hollow fiber TFF step could thus be used on a large scale to reduce

the volume of the plant extract prior to the chromatography step.

Numerous purification schemes have been developed to address the inefficiency

of post-harvest processing and purification of plant-produced antibodies. However, a

scalable scheme that can be used to purify a broad range of full-length antibodies from

various plant tissues would thus be extremely valuable. The versatility of the purification

scheme outlined in this thesis should thus be examined by purifying another therapeutic

antibody from N. benthamiana plants or by purifying trastuzumab from N. tabacum

plants.

In conclusion, this thesis provides evidence that trastuzumab, an important

antibody used in human cancer therapy, can be both efficiently produced and purified

from N. benthamiana plants.

87

6 LITERATURE CITED

Aguilar, O., and Rito-Palomares, M. (2010). Aqueous two-phase systems strategies for the recovery and characterization of biological products from plants. J. Sci. Food Agric. 90, 1385-1392.

Akerstrom, B., and Bjorck, L. (1989). Protein L: an immunoglobulin light chain-binding bacterial protein. Characterization of binding and physicochemical properties. J. Biol. Chem. 264, 19740-19746.

Almquist, K. C , McLean,, M. D., Niu, Y., Byrne, G., Olea-Popelka, F. C, Murrant, C, Barclay, J., and Hall, J. C. (2006). Expression of an anti-botulinum toxin A neutralizing single-chain Fv recombinant antibody in transgenic tobacco. Vaccine 24, 2079-2086.

Alroy, I., and Yarden, Y. (1997). The ErbB signaling network in embryogenesis and oncogenesis: signal diversification through combinatorial ligand-receptor interactions. FEBS Lett. 410, 83-86.

Arnould, L., Gelly, M., Penault-Llorca, F., Benoit, L., Bonnetain, F., Migeon, C, Cabaret, V., Fermeaux, V., Bertheau, P., Gamier, J., et al. (2006). Trastuzumab-based treatment of HER2-positive breast cancer: an antibody-dependent cellular cytotoxicity mechanism? Br. J. Cancer 94, 259-267.

Arntzen, C. J. (2008). Plant science: using tobacco to treat cancer. Science 321, 1052-1053.

Bakker, H., Rouwendal, G. J. A., Karnoup, A. S., Florack, D. E. A., Stoopen, G. M., Helsper, J. P. F. G., van Ree, R., van Die, I., and Bosch, D. (2006). An antibody produced in tobacco expressing a hybrid (3-1,4-galactosyltransferase is essentially devoid of plant carbohydrate epitopes. Proc. Natl. Acad. Sci. U.S.A. 103, 7577-7582.

Balasubramaniam, D., Wilkinson, C, Van Cott, K., and Zhang, C. (2003). Tobacco protein separation by aqueous two-phase extraction. J. Chromatogr. A 989, 119-129.

Ballew, H. W., Clark, D., Crooks, G., Ho, S. V., Markee, C , Martinez, F. J., Milner, D., Partee, J., and Simmons, J. (2002). The ABCs of filtration and bioprocessing for the third millennium (Ballew, H.W., Martinez, F.J., Markee, C, and Eddleman, R.T. eds) Rancho Dominguez, CA: Spectrum Laboratories, Inc.

88

Baselga, J., Norton, L., Albanell, J., Kim, Y. M., and Mendelsohn, J. (1998). Recombinant humanized anti-HER2 antibody (Herceptin) enhances the antitumor activity of paclitaxel and doxorubicin against HER2/neu overexpressing human breast cancer xenografts. Cancer Res. 58, 2825-2831.

Baselga, J., Tripathy, D., Mendelsohn, J., Baughman, S., Benz, C. C, Dantis, L., Sklarin, N. T., Seidman, A. D., Hudis, C. A., Moore, J., et al. (1996). Phase II study of weekly intravenous recombinant humanized anti-pl85HER2 monoclonal antibody in patients with HER2/neu-overexpressing metastatic breast cancer. J. Clin. Oncol. 14, 737-744.

Baselga, J., and Albanell, J. (2001). Mechanism of action of anti-HER2 monoclonal antibodies. Ann. Oncol. 12, 35-41.

Baselga, J., and Swain, S. M. (2009). Novel anticancer targets: revisiting ERBB2 and discovering ERBB3. Nat. Rev. Cancer 9, 463-475.

Baulida, J., Kraus, M. H., Alimandi, M., Fiore, P. P. D., and Carpenter, G. (1996). All ErbB receptors other than the epidermal growth factor receptor are endocytosis impaired. J. Biol. Chem. 271, 5251 -5257.

Beano, A., Signorino, E., Evangelista, A., Brusa, D., Mistrangelo, M., Polimeni, M. A., Spadi, R., Donadio, M., Ciuffreda, L., and Matera, L. (2008). Correlation between NK function and response to trastuzumab in metastatic breast cancer patients. J. Transl. Med. 6, 1-10.

Beck, A., Bussat, M., Zorn, N., Robillard, V., Klinguer-Hamour, C, Chenu, S., Goetsch, L., Corvaia, N., Van Dorsselaer, A., and Haeuw, J. (2005). Characterization by liquid chromatography combined with mass spectrometry of monoclonal anti-IGF-1 receptor antibodies produced in CHO and NS0 cells. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 819, 203-218.

Belfort, G., Davis, R. H., andZydney, A. L. (1994). The behavior of suspensions and macromolecular solutions in crossflow microfiltration. J. Memb. Sci. 96, 1-58.

Ben-Kasus, T., Schechter, B., Lavi, S., Yarden, Y., and Sela, M. (2009). Persistent elimination of ErbB-2/HER2-overexpressing tumors using combinations of monoclonal antibodies: relevance of receptor endocytosis. Proc. Natl. Acad. Sci. U.S.A 106, 3294-3299.

Bendandi, M., Marillonnet, S., Kandzia, R., Thieme, F., Nickstadt, A., Herz, S., Frode, R., Inoges, S., Lopez-Diaz de Cerio, A., Soria, E., et al. (2010). Rapid, high-yield production in plants of individualized idiotype vaccines for non-Hodgkin's lymphoma. Ann. Oncol. DOI: 10.1093/annonc/mdq256.

89

Birch, J. R., and Racher, A. J. (2006). Antibody production. Adv. Drug Deliv. Rev. 58, 671-685.

Bjorck, L., and Kronvall, G. (1984). Purification and some properties of streptococcal Protein G, a novel IgG-binding reagent. J. Immunol. 133, 969-974.

Cabanes-Macheteau, M., Fitchette-Laine, A. C, Loutelier-Bourhis, C , Lange, C, Vine, N. D., Ma, J. K., Lerouge, P., and Faye, L. (1999). N-Glycosylation of a mouse IgG expressed in transgenic tobacco plants. Glycobiology 9, 365-372.

Carter, P., Presta, L., Gorman, CM. , Ridgway, J. B. B., Henner, D., Wong, W. L. T., Rowland, A. M., Kotts, C, Carver, M. E., and Shepard, H. M. (1992). Humanization of an anti-pl85HER2 antibody for human cancer therapy. Proc. Natl. Acad. Sci. U.S.A. 89, 4285-4289.

Chadd, H. E., and Chamow, S. M. (2001). Therapeutic antibody expression technology. Curr. Opin. Biotechnol. 12, 188-194.

Chames, P., Van Regenmortel, M., Weiss, E., and Baty, D. (2009). Therapeutic antibodies: successes, limitations and hopes for the future. Br. J. Pharmacol. 157, 220-233.

Chargelegue, D., Vine, N. D., van Dolleweerd, C. J., Drake, P. M., and Ma, J. K. (2000). A murine monoclonal antibody produced in transgenic plants with plant-specific glycans is not immunogenic in mice. Transgenic Res. 9, 187-194.

Churchill, R. L. T., Sheedy, C, Yau, K. Y. F., and Hall, J. C. (2002). Evolution of antibodies for environmental monitoring: from mice to plants. Anal. Chim. Acta 468, 185-197.

Citri, A., and Yarden, Y. (2006). EGF-ERBB signalling: towards the systems level. Nat. Rev. Mol. Cell Biol. 7, 505-516.

Clonis, Y. D. (2006). Affinity chromatography matures as bioinformatic and combinatorial tools develop. J. Chromatogr. A 1101, 1-24.

Cobleigh, M. A., Vogel, C. L., Tripathy, D., Robert, N. J., Scholl, S., Fehrenbacher, L., Wolter, J. M., Paton, V., Shak, S., Lieberman, G., et al. (1999). Multinational study of the efficacy and safety of humanized anti-HER2 monoclonal antibody in women who have HER2-overexpressing metastatic breast cancer that has progressed after chemotherapy for metastatic disease. J. Clin. Oncol. 17, 2639-2648.

90

Conley, A. J., Joensuu, J. J., Jevnikar, A. M., Menassa, R., and Brandle, J. E. (2009). Optimization of elastin-like polypeptide fusions for expression and purification of recombinant proteins in plants. Biotechnol. Bioeng. 103, 562-573.

Covic, A., and Kuhlmann, M. (2007). Biosimilars: recent developments. Int. Urol. Nephrol. 39, 261-266.

Cox, K. M., Regan, J. T., Sterling, J. D., Wingate, V. P., and Dickey, L. F. (2009). Production of antibodies in plants. In: Therapeutic Monoclonal Antibodies (An, Z., ed.), pp. 651-674. Hoboken, NJ: John Wiley & Sons, Inc.

Czermak, P., Nehring, D., and Wickramasinghe, R. (2007). Membrane filtration in animal cell culture. In: Animal Cell Biotechnology (Portner, R., ed.), pp. 397-420. Humana Press.

De Muynck, B., Navarre, C, and Boutry, M. (2010). Production of antibodies in plants: status after twenty years. Plant Biotechnol. J. 8, 529-563.

De Santes, K., Slamon, D., Anderson, S. K., Shepard, M., Fendly, B., Maneval, D., and Press, O. (1992). Radiolabeled antibody targeting of the HER-2/neu oncoprotein. Cancer Res. 52, 1916 -1923.

Deisenhofer, J. (1981). Crystallographic refinement and atomic models of a human Fc fragment and its complex with fragment B of Protein A from Staphylococcus aureus at 2.9- and 2.8-A resolution. Biochemistry 20, 2361-2370.

Dimitrov, D. S., and Marks, J. D. (2009). Therapeutic antibodies: current state and future trends - is a paradigm change coming soon? In: Therapeutic Antibodies (Dimitrov, A.S., ed), pp. 1-27. New York: Humana Press.

DrugBank (2009). Trastuzumab (DG00072). http://www.drugbank.ca/drugs/DB00072. Accessed November 4, 2010.

Evangelista, R. L., Kusnadi, A. R., Howard, J. A., andNikolov, Z. L. (1998). Process and economic evaluation of the extraction and purification of recombinant (3-glucuronidase from transgenic corn. Biotechnol. Prog. 14, 607-614.

Faye, L., and Gomord, V. (2010). Success stories in molecular farming-a brief overview. Plant Biotechnol. J. 8, 525-528.

Fendly, B. M., Winget, M., Hudziak, R. M., Lipari, M. T., Napier, M. A., and Ullrich, A. (1990). Characterization of murine monoclonal antibodies reactive to either the human epidermal growth factor receptor or HER2/neu gene product. Cancer Res. 50, 1550-1558.

91

Fischer, R., Liao, Y. C , and Drossard, J. (1999). Affinity-purification of a TMV-specific recombinant full-size antibody from a transgenic tobacco suspension culture. J. Immunol. Methods 226, 1-10.

Fischer, R., Schillberg, S., and Twyman, R. M. (2009). Molecular farming of antibodies in plants. In: Recent Advances in Plant Biotechnology (Kirakosyan, A. and Kaufman, P., eds.), pp 35-63. New York: Springer US.

Floss, D. M., Schallau, K., Rose-John, S., Conrad, U., and Scheller, J. (2010). Elastin-like polypeptides revolutionize recombinant protein expression and their biomedical application. Trends Biotechnol. 28, 37-45.

Fontes, N., and van Reis, R. (2009). Advances in technology and process development for industrial-scale monoclonal antibody purification. In: Process Scale Purification of Antibodies (Gottschalk, U. ed.), pp. 203-222. Hoboken, NJ: John Wiley & Sons, Inc.

Frey, A. D., Karg, S. R., and Kallio, P. T. (2009). Expression of rat (3(1,4)-N-acetylglucosaminyltransferase III in Nicotiana tabacum remodels the plant-specific N-glycosylation. Plant Biotechnol. J. 7, 33-48.

Fry, W. H., Kotelawala, L., Sweeney, C , and Carraway III, K. L. (2009). Mechanisms of ErbB receptor negative regulation and relevance in cancer. Exp. Cell Res. 315, 697-706.

Gilbert, N. (2009). Fees delay pharmed drug. Nature 458, 951.

Giritch, A., Marillonnet, S., Engler, C , Eldik, G. V., Botterman, J., Klimyuk, V., and Gleba, Y. (2006). Rapid high-yield expression of full-size IgG antibodies in plants coinfected with noncompeting viral vectors. Proc. Natl. Acad. Sci. U.S.A. 103, 14701-14706.

Gomord, V., Fitchette, A., Menu-Bouaouiche, L., Saint-Jore-Dupas, C , Plasson, C , Michaud, D., and Faye, L. (2010). Plant-specific glycosylation patterns in the context of therapeutic protein production. Plant Biotechnol. J. 8, 564-587.

Gomord, V., Sourrouille, C , Fitchette, A., Bardor, M., Pagny, S., Lerouge, P., and Faye, L. (2004). Production and glycosylation of plant-made pharmaceuticals: the antibodies as a challenge. Plant Biotechnol. J. 2, 83-100.

Gottlieb, S. (2008). Biosimilars: policy, clinical, and regulatory considerations. Am. J. Health Syst. Pharm. 65, S2-8.

Gouda, H., Torigoe, H., Saito, A., Sato, M., Arata, Y., and Shimada, I. (1992). Three-dimensional solution structure of the B domain of staphylococcal Protein A: comparisons of the solution and crystal structures. Biochemistry 31, 9665-9672.

92

Graille, M., Stura, E. A., Housden, N. G., Beckingham, J. A., Bottomley, S. P., Beale, D., Taussig, M. J., Sutton, B. J., Gore, M. G., and Charbonnier, J. B. (2001). Complex between Peptostreptococcus magnus Protein L and a human antibody reveals structural convergence in the interaction modes of Fab binding proteins. Structure 9, 679-687.

Graille, M., Stura, E. A., Corper, A. L., Sutton, B. J., Taussig, M. J., Charbonnier, J., and Silverman, G. J. (2000). Crystal structure of a Staphylococcus aureus Protein A domain complexed with the Fab fragment of a human IgM antibody: structural basis for recognition of B-cell receptors and superantigen activity. Proc. Natl. Acad. Sci. U.S.A. 97, 5399-5404.

Grohs, B. M., Niu, Y., Veldhuis, L. J., Trabelsi, S., Garabagi, F., Hassell, J. A., McLean, M. D., and Hall, J. C. (2010). Plant-produced trastuzumab inhibits the growth of HER2 positive cancer cells. J. Agric. Food Chem. 51, 10056-10063.

Guy, P. M., Platko, J. V., Cantley, L. C, Cerione, R. A., and Carraway, K. L. (1994). Insect cell-expressed pi80erbB3 possesses an impaired tyrosine kinase activity. Proc. Natl. Acad. Sci. U.S.A. 91, 8132 -8136.

Hage, D.S. (1998). Affinity chromatography. In: Handbook of HPLC (Katz, E., Eksteen, R., Shoenmakers, P., and Miller, N., eds.), pp. 483-498. New York: Marcel Dekker.

Hage, D.S., Bian, M., Burks, R., Karle, E., Ohnmact, C , and Wa, C. (2006a). Bioaffimty chromatography. In: Handbook of Affinity Chromatography (Hage, D.S., ed.), pp. 101-126. Boca Raton, FL: Taylor and Francis Group.

Hage, D.S., and Ruhn, P.F. (2006). An introduction to affinity chromatography. In: Handbook of Affinity Chromatography (Hage, D.S., ed.), pp 3-13. Boca Raton, FL: Taylor and Francis Group.

Hage, D.S., Xuan, H., and Nelson, M.A. (2006b). Application and elution in affinity chromatography. In: Handbook of Affinity Chromatography (Hage, D.S., ed.), pp. 79-97. Boca Raton, FL: Taylor and Francis Group.

Hassan, S., van Dolleweerd, C. J., Ioakeimidis, F., Keshavarz-Moore, E., and Ma, J. K. (2008). Considerations for extraction of monoclonal antibodies targeted to different subcellular compartments in transgenic tobacco plants. Plant Biotechnol. J. 6, 733-748.

Hefferon, K. L. (2010). Scale-up of plant-derived pharmaceuticals: prospects for commercial production and for global health. In: Biopharmaceuticals in Plants: Toward the Next Century of Medicine, pp. 117-146. Boca Raton, FL: Taylor and Francis Group.

93

Hiatt, A., Cafferkey, R., and Bowdish, K. (1989). Production of antibodies in transgenic plants. Nature 342, 76-78.

Hill, P., and Bender, J. (2007). Cell harvesting. In: Medicines from Animal Cell Culture (Stacey, G. and Davis, J., eds.), pp. 305-330. West Sussex, England: John Wiley & Sons, Inc.

Hober, S., Nord, K., and Linhult, M. (2007). Protein A chromatography for antibody purification. J. Chromatogr. B 848, 40-47.

Huang, Z., Phoolcharoen, W., Lai, H., Piensook, K., Cardineau, G., Zeitlin, L., Whaley, K. J., Arntzen, C. J., Mason, H. S., and Chen, Q. (2010). High-level rapid production of full-size monoclonal antibodies in plants by a single-vector DNA replicon system. Biotechnol. Bioeng. 106, 9-17.

Hudziak, R. M., Lewis, G. D., Shalaby, M. R., Eessalu, T. E., Aggarwal, B. B., Ullrich, A., and Shepard, H. M. (1988). Amplified expression of the HER2/ERBB2 oncogene induces resistance to tumor necrosis factor alpha in NIH 3T3 cells. Proc. Natl. Acad. Sci. U.S.A. 85, 5102-5106.

Hudziak, R. M., Lewis, G. D., Winget, M., Fendly, B. M., Shepard, H. M., and Ullrich, A. (1989). pl85HER2 monoclonal antibody has antiproliferative effects in vitro and sensitizes human breast tumor cells to tumor necrosis factor. Mol. Cell. Biol. 9, 1165-1172.

Huse, K., Bohme, H., and Scholz, G. H. (2002). Purification of antibodies by affinity chromatography. J. Biochem. Biophys. Methods 51, 217-231.

Hussack, G., Grohs, B. M., Almquist, K. C , McLean, M. D., Ghosh, R., and Hall, J. C. (2010). Purification of plant-derived antibodies through direct immobilization of affinity ligands on cellulose. J. Agric. Food Chem. 58, 3451-3459.

Hynes, N. E., and Stern, D. F. (1994). The biology of erbB-2/neu/HER-2 and its role in cancer. Biochim. Biophys. Acta 1198, 165-184.

Janson, J., and Jonsson, J. (1998). Introduction to chromatography. In: Protein Purification: Principles, High-Resolution Methods, and Applications (Janson, J. and Ryden, L., eds.), pp. 43-78. New York, NY: John Wiley & Sons, Inc.

Jansson, B., Uhlen, M., and Nygren, P. A. (1998). All individual domains of staphylococcal Protein A show Fab binding. FEMS Immunol. Med. Microbiol. 20, 69-78.

Jendeberg, L., Persson, B., Andersson, R., Karlsson, R., Uhlen, M., and Nilsson, B. (1995). Kinetic analysis of the interaction between Protein A domain variants and human Fc using plasmon resonance detection. J. Mol. Recognit. 8, 270-278.

94

Jones, R. B., Gordus, A., Krall, J. A., and MacBeath, G. (2006). A quantitative protein interaction network for the ErbB receptors using protein microarrays. Nature 439, 168-174.

Jurado, L., Oak, A., Gadgil, H., Moxley, R., and Jarret, H. (2006). Affinity chromatography in molecular biology. In: Handbook of Affinity Chromatography (Hage, D., ed.), pp. 547-570. Boca Raton, FL: Taylor and Francis Group.

Karg, S. R., and Kallio, P. T. (2009). The production of biopharmaceuticals in plant systems. Biotechnol. Adv. 27, 879-894.

Karimi, A., Matsumura, M., Wright, P. E., and Dyson, H. J. (1999). Characterization of monomeric and dimeric B domain of staphylococcal Protein A. JPept. Res. 54, 344-352.

Kindt, T., Goldsby, R., and Osbourne, B. (2007). Cell-mediated cytotoxic responses. In: Immunology (Ahr, K., Tymoczko, N., and Tascheck, K., eds.), pp. 351-370. New York, NY: Freeman and Company.

Klapper, L. N., Glathe, S., Vaisman, N., Hynes, N. E., Andrews, G. C, Sela, M., and Yarden, Y. (1999). The ErbB-2/HER2 oncoprotein of human carcinomas may function solely as a shared coreceptor for multiple stroma-derived growth factors. Proc. Natl. Acad. Sci. U.S.A. 96, 4995-5000.

Ko, K., Brodzik, R., and Steplewski, Z. (2009). Production of antibodies in plants: approaches and perspectives. In: Plant-Produced Microbial Vaccines (Karasev, A., ed.), pp.55-78. Berlin, Heidelberg: Springer-Verlag.

Komarnytsky, S., Borisjuk, N., Yakoby, N., Garvey, A., and Raskin, I. (2006). Cosecretion of protease inhibitor stabilizes antibodies produced by plant roots. Plant Physiol. 141, 1185-1193.

Kronvall, G., and Williams, R. C. (1969). Differences in anti-Protein A activity among IgG subgroups. J. Immunol. 103, 828-833.

Kumar, R., Shepard, H. M., and Mendelsohn, J. (1991). Regulation of phosphorylation of the c-erbB-2/HER2 gene product by a monoclonal antibody and serum growth factor(s) in human mammary carcinoma cells. Mol. Cell. Biol. 11, 979-986.

Lane, H., Motoyama, A., Beuvink, I., and Hynes, N. (2001). Modulation of p27/Cdk2 complex formation through 4D5-mediated inhibition of HER2 receptor signaling. Ann. Oncol. 12, S21-S22.

Larsson, P. (1993). Support materials for affinity chromatography. In: Handbook of Affinity Chromatography (Kline, T., ed.) pp. 61-75. New York, NY: Marcel Dekker Inc.

95

Lazzara, M.J., and Lauffenburger, D.A. (2009). Quantitative modeling perspectives on the ErbB system of cell regulatory processes. Exp. Cell Res. 315, 717-725.

Le, X., Claret, F., Lammayot, A., Tian, L., Deshpande, D., LaPushin, R., Tari, A. M., and Bast, R. C. (2003). The role of cyclin-dependent kinase inhibitor p27Kipl in anti-HER2 antibody-induced Gl cell cycle arrest and tumor growth inhibition. J. Biol. Chem. 278, 23441-23450.

Le, X., McWatters, A., Wiener, J., Wu, J., Mills, G. B., and Bast, R. C. (2000). Anti-HER2 antibody and heregulin suppress growth of HER2-overexpressing human breast cancer cells through different mechanisms. Clin. Cancer Res. 6, 260-270.

Lemmon, M. A. (2009). Ligand-induced ErbB receptor dimerization. Exp. Cell Res. 315, 638-648.

Lenferink, A. E., Pinkas-Kramarski, R., van de Poll, M. L., van Vugt, M. J., Klapper, L. N., Tzahar, E., Waterman, H., Sela, M., van Zoelen, E. J., and Yarden, Y. (1998). Differential endocytic routing of homo- and hetero-dimeric ErbB tyrosine kinases confers signaling superiority to receptor heterodimers. EMBO J. 17, 3385-3397.

Leong, S. S., and Chen, W. N. (2008). Preparing recombinant single chain antibodies. Chem. Eng. Sci. 63, 1401-1414.

Lewis, G. D., Figari, I., Fendly, B., Lee Wong, W., Carter, P., Gorman, C , and Shepard, H. M. (1993). Differential responses of human tumor cell lines to anti-pl85HER2 monoclonal antibodies. Cancer Immunol. Immunother. 37, 255-263.

Li, Y. M., Pan, Y., Wei, Y., Cheng, X., Zhou, B. P., Tan, M., Zhou, X., Xia, W., Hortobagyi, G. N., Yu, D., et al. (2004). Upregulation of CXCR4 is essential for HER2-mediated tumor metastasis. Cancer Cell 6, 459-469.

Linhult, M., Gulich, S., Graslund, T., Simon, A., Karlsson, M., Sjoberg, A., Nord, K., and Hober, S. (2004). Improving the tolerance of a Protein A analogue to repeated alkaline exposures using a bypass mutagenesis approach. Proteins 55, 407-416.

Liu, H., Gaza-Bulseco, G., Faldu, D., Chumsae, C , and Sun, J. (2008). Heterogeneity of monoclonal antibodies. J. Pharm. Sci. 97, 2426-2447.

Marillonnet, S., Thoeringer, C, Kandzia, R., Klimyuk, V., and Gleba, Y. (2005). Systemic Agrobacterium tumefaciens-mQdiatQd transfection of viral replicons for efficient transient expression in plants. Nat. Biotechnol. 23, 718-723.

96

Marty, M., Cognetti, F., Maraninchi, D., Snyder, R., Mauriac, L., Tubiana-Hulin, M., Chan, S., Grimes, D., Anton, A., Lluch, A., et al. (2005). Randomized phase II trial of the efficacy and safety of trastuzumab combined with docetaxel in patients with human epidermal growth factor receptor 2-positive metastatic breast cancer administered as first-line treatment: the M77001 study group. J. Clin. Oncol. 23, 4265-4274.

McLean, M. D., Almquist, K. C., Niu, Y., Kimmel, R., Lai, Z., Schreiber, J. R., and Hall, J. C. (2007). A human anti-Pseudomonas aeruginosa serotype 06ad immunoglobulin Gl expressed in transgenic tobacco is capable of recruiting immune system effector function in vitro. Antimicrob. Agents Chemother. 51, 3322-3328.

Menkhaus, T. J., Bai, Y., Zhang, C., Nikolov, Z. L., and Glatz, C. E. (2004). Considerations for the recovery of recombinant proteins from plants. Biotechnol. Prog. 20, 1001-1014.

Miele, L. (1997). Plants as bioreactors for biopharmaceuticals: regulatory considerations. Trends Biotechnol. 15, 45-50.

Mison, D., and Curling, J. (2000). The industrial production cost of recombinant therapeutic proteins expressed in transgenic corn. Biopharm. 13, 48-54.

Mohr, P., and Pommerening, K. (1985a). General considerations of the adsorption and elution step. In: Affinity Chromatography: Practical and Theoretical Aspects, pp. 85-119. New York: Marcel Dekker Inc.

Mohr, P., and Pommerening, K. (1985b). Solid matrix support. In: Affinity Chromatography: Practical and Theoretical Aspects, pp. 19-39. New York: Marcel Dekker Inc.

Mohsin, S. K., Weiss, H. L., Gutierrez, M. C , Chamness, G. C , Schiff, R., DiGiovanna, M. P., Wang, C , Hilsenbeck, S. G., Osborne, C. K., Allred, D. C, et al. (2005). Neoadjuvant trastuzumab induces apoptosis in primary breast cancers. J. Clin. Oncol. 23, 2460-2468.

Molina, M. A., Codony-Servat, J., Albanell, J., Rojo, F., Arribas, J., and Baselga, J. (2001). Trastuzumab (Herceptin), a humanized anti-HER2 receptor monoclonal antibody, inhibits basal and activated Her2 ectodomain cleavage in breast cancer cells. Cancer Res. 61,4744-4749.

Mori, K., Iida, S., Yamane-Ohnuki, N., Kanda, Y., Kuni-Kamochi, R., Nakano, R., Imai-Nishiya, H., Okazaki, A., Shinkawa, T., Natsume, A., et al. (2007). Non-fucosylated therapeutic antibodies: the next generation of therapeutic antibodies. Cytotechnology 55, 109-114.

97

Narayanan, S. R. (1994). Preparative affinity chromatography of proteins. J. Chromatogr. A 658, 237-258.

Nelson, D., and Cox, M. (2005). Protein function. In: Lehninger Principles of Biochemistry (Ahr, K., Ryan, M., O'Neill, J., and Wong, V., eds.), pp. 160-161. New York, NY: Freeman and Company.

Nikolov, Z. L., Regan, J. T., Dickey, L. F., and Woodard, S. L. (2009). Purification of antibodies from transgenic plants. In: Process Scale Purification of Antibodies (Gottschalk, U., ed.), pp. 387-406. Hoboken, NJ: John Wiley & Sons, Inc.

Nilson, B. H., Solomon, A., Bjorck, L., and Akerstrom, B. (1992). Protein L from Peptostreptococcus magnus binds to the kappa light chain variable domain. J. Biol. Chem. 267, 2234 -2239.

Nilsson, B., Moks, T., Jansson, B., Abrahmsen, L., Elmblad, A., Holmgren, E., Henrichson, C, Jones, T., and Uhlen, M. (1987). A synthetic IgG-binding domain based on staphylococcal Protein A. Protein Eng. 1, 107-113.

Nissim, A., and Chernajovsky, Y. (2008). Historical development of monoclonal antibody therapeutics. In: Therapeutic Antibodies (Chernajovsky, Y. and Nissim, A., eds.), pp. 3-18. Berlin, Heidelberg: Springer-Verlag.

Olea-Popelka, F., McLean, M. D., Horsman, J., Almquist, K., Brandle, J. E., and Hall, J. C. (2005). Increasing expression of an anti-picloram single-chain variable fragment (ScFv) antibody and resistance to picloram in transgenic tobacco (Nicotiana tabacum). J. Agric. Food Chem. 53, 6683-6690.

Ostrove, S. (1990). Affinity chromatography: general methods. Methods Enzymol. 182, 357-371.

Parmenter, D. L., Boothe, J. G., Rooijen, G. J. H., Yeung, E. C, and Moloney, M. M. (1995). Production of biologically active hirudin in plant seeds using oleosin partitioning. Plant Mol. Biol. 29, 1167-1180.

Pegram, M. D., Lipton, A., Hayes, D. F., Weber, B. L., Baselga, J. M., Tripathy, D., Baly, D., Baughman, S. A., Twaddell, T., Glaspy, J. A., et al. (1998). Phase II study of receptor-enhanced chemosensitivity using recombinant humanized anti-pi 85HER2/neu monoclonal antibody plus cisplatin in patients with HER2/neu-overexpressing metastatic breast cancer refractory to chemotherapy treatment. J. Clin. Oncol. 16,2659-2671.

Phillips, T.M. (2006). Affinity chromatography in antibody and antigen purification. In: Handbook of Affinity Chromatography (Hage, D.S., ed.), pp. 367-397. New York, NY: Taylor and Francis Group.

98

Phillips, T.M, and Dickens, B.F. (2000a). Affinity and immunoaffinity separations. In: Affinity and Immunoaffinity Purification Techniques (Weaver, S. and McAndrews, C , eds.), pp. 1-18. Natick, MA: Eaton Publishing.

Phillips, T.M., and Dickens, B.F. (2000b). Factors Influencing Separation. In: Affinity and Immunoaffinity Purification Techniques (Weaver, S. and McAndrews, C , eds.), pp. 55-84. Natick, MA: Eaton Publishing.

Phillips, T., Queen, W., More, N., and Thompson, A. (1985). Protein A-coated glass beads: universal support medium for high-performance immunoaffinity chromatography. J. Chromatogr. A 327, 213-219.

Platis, D., Drossard, J., Fischer, R., Ma, J., and Labrou, N. (2008). New downstream processing strategy for the purification of monoclonal antibodies from transgenic tobacco plants. J. Chromatogr. A 1211, 80-89.

Platis, D., and Labrou, N. E. (2006). Development of an aqueous two-phase partitioning system for fractionating therapeutic proteins from tobacco extract. /. Chromatogr. ,4 1128, 114-124.

Pujol, M., Ramirez, N. I., Ayala, M., Gavilondo, J. V., Valdes, R., Rodriguez, M., Brito, J., Padilla, S., Gomez, L., Reyes, B., et al. (2005). An integral approach towards a practical application for a plant-made monoclonal antibody in vaccine purification. Vaccine 23, 1833-1837.

Ramessar, K., Rademacher, T., Sack, M., Stadlmann, J., Platis, D., Stiegler, G., Labrou, N., Altmann, F., Ma, J., Stoger, E., et al. (2008). Cost-effective production of a vaginal protein microbicide to prevent HIV transmission. Proc. Natl. Acad. Sci. U.S.A 105, 3727-3732.

Reed, S. I. (2003). Ratchets and clocks: the cell cycle, ubiquitylation and protein turnover. Nat. Rev. Mol. Cell Biol. 4, 855-864.

van Reis, R., and Zydney, A. (2001). Membrane separations in biotechnology. Curr. Opin. Biotechnol. 12, 208-211.

Rodriguez, M., Ramirez, N. I., Ayala, M., Freyre, F., Perez, L., Triguero, A., Mateo, C, Selman-Housein, G., Gavilondo, J. V., and Pujol, M. (2005). Transient expression in tobacco leaves of an aglycosylated recombinant antibody against the epidermal growth factor receptor. Biotechnol Bioeng. 89, 188-194.

Rooijen, G. J. H. V., and Moloney, M. M. (1995). Plant seed oil-bodies as carriers for foreign proteins. Nat. Biotech. 13, 72-77.

Roque, A., Lowe, C , and Taipa, M. (2004). Antibodies and genetically engineered related molecules: production and purification. Biotechnol. Prog. 20, 639-654.

99

Roque, A. C , Silva, C. S., and Taipa, M. A. (2007). Affinity-based methodologies and ligands for antibody purification: advances and perspectives. J. Chromatogr. A 1160, 44-55.

Sack, M., Paetz, A., Kunert, R., Bomble, M., Hesse, F., Stiegler, G., Fischer, R., Katinger, H., Stoeger, E., and Rademacher, T. (2007). Functional analysis of the broadly neutralizing human anti-HIV-1 antibody 2F5 produced in transgenic BY-2 suspension cultures. FASEB J. 21, 1655-1664.

Sahin, O., and Wiemann, S. (2009). Functional genomics and proteomics approaches to study the ERBB network in cancer. FEBSLett. 583, 1766-1771.

Samac, D. A., Hironaka, C. M., Yallaly, P. E., and Shah, D. M. (1990). Isolation and characterization of the genes encoding basic and acidic chitinase in Arabidopsis thaliana. Plant Physiol. 93, 907-914.

Samanta, A., LeVea, C. M., Dougall, W. C, Qian, X., and Greene, M. I. (1994). Ligand and pl85c-neu density govern receptor interactions and tyrosine kinase activation. Proc. Natl. Acad. Sci. U.S.A. 91, 1711-1715.

Sanbe, H., and Haginaka, J. (2003). Uniformly sized molecularly imprinted polymers for bisphenol A and P-estradiol: retention and molecular recognition properties in hydro-organic mobile phases. J. Pharm. Biomed. Anal. 30, 1835-1844.

Sarup, J. C, Johnson, R. M., King, K. L., Fendly, B. M., Lipari, M. T., Napier, M. A., Ullrich, A., and Shepard, H. M. (1991). Characterization of an anti-pl85HER2 monoclonal antibody that stimulates receptor function and inhibits tumor cell growth. Growth Regul. 1, 72-82.

Sauer-Eriksson, A. E., Kleywegt, G. J., Uhlen, M., and Jones, T. A. (1995). Crystal structure of the C2 fragment of streptococcal Protein G in complex with the Fc domain of human IgG. Structure 3, 265-278.

Saxena, A., Tripathi, B. P., Kumar, M., and Shahi, V. K. (2009). Membrane-based techniques for the separation and purification of proteins: an overview. Adv. Colloid. Interface Sci. 145, 1-22.

Schillberg, S., Fischer, R., and Emans, N. (2003). Molecular farming of recombinant antibodies in plants. Cell Mol. Life. Sci. 60, 433-445.

Schmitz, K. R., and Ferguson, K. M. (2009). Interaction of antibodies with ErbB receptor extracellular regions. Exp. Cell Res. 315, 659-670.

Seon, J., Szarka, S., and Moloney, M. (2002). A unique strategy for recovering recombinant proteins from molecular farming: affinity capture on engineered oilbodies. J. Plant Biotechnol. 4, 95-101.

100

Sharp, J. M., and Doran, P. M. (2001). Characterization of monoclonal antibody fragments produced by plant cells. Biotechnol. Bioeng. 73, 338-346.

Shepard, H., and Lewis, G. (1988). Resistance of tumor cells to tumor necrosis factor. J. Clin. Immunol. 8, 333-341.

Shepard, H. M., Brdlik, C. M., and Schreiber, H. (2008). Signal integration: a framework for understanding the efficacy of therapeutics targeting the human EGFR family. J. Clin. Invest. 118, 3574-3581.

Sherr, C. J., and Roberts, J. M. (1999). CDK inhibitors: positive and negative regulators of Gl-phase progression. Genes Dev. 13, 1501-1512.

Shukla, A. A., Hubbard, B., Tressel, T., Guhan, S., and Low, D. (2007). Downstream processing of monoclonal antibodies-application of platform approaches. J. Chromatogr. B 848, 28-39.

Slamon, D. J., Clark, G. M., Wong, S. G., Levin, W. J., Ullrich, A., and McGuire, W. L. (1987). Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 235, 177-182.

Slamon, D. J., Godolphin, W., Jones, L. A., Holt, J. A., Wong, S. G., Keith, D. E., Levin, W. J., Stuart, S. G., Udove, J., and Ullrich, A. (1989). Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer. Science 244, 707-712.

Slamon, D. J., Leyland-Jones, B., Shak, S., Fuchs, H., Paton, V., Bajamonde, A., Fleming, T., Eiermann, W., Wolter, J., Pegram, M., et al. (2001). Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N. Engl. J. Med. 344, 783-792.

Sliwkowski, M. X., Lofgren, J. A., Lewis, G. D., Hotaling, T. E., Fendly, B. M., and Fox, J. A. (1999). Nonclinical studies addressing the mechanism of action of trastuzumab (Herceptin). Semin. Oncol. 26, 60-70.

Stahl, S., Nygren, P. A., Sjolander, A., and Uhlen, M. (1993). Engineered bacterial receptors in immunology. Curr. Opin. Immunol. 5, 272-277'.

Starovasnik, M. A., Skelton, N. J., O'Connell, M. P., Kelley, R. F., Reilly, D., and Fairbrother, W. J. (1996). Solution structure of the E-domain of staphylococcal Protein A. Biochemistry 35, 15558-15569.

Stevens, L. H., Stoopen, G. M., Elbers, I. J., Molthoff, J. W., Bakker, H. A., Lommen, A., Bosch, D., and Jordi, W. (2000). Effect of climate conditions and plant developmental stage on the stability of antibodies expressed in transgenic tobacco. Plant Physiol. 124, 173-182.

101

Stoger, E., Sack, M., Nicholson, L., Fischer, R., and Christou, P. (2005). Recent progress in plantibody technology. Curr. Pharm. Des. 11, 2439-2457.

Stoger, E., Sack, M., Fischer, R., and Christou, P. (2002). Plantibodies: applications, advantages and bottlenecks. Curr. Opin. Biotechnol. 13, 161-166.

Stoger, E., Schillberg, S., Twyman, R. M., Fischer, R., and Christou, P. (2004). Antibody production in transgenic plants. In: Antibody Engineering - Methods and Protocols (Lo, B. K., ed.), pp. 301-318. Totowa, NJ: Humana Press Inc.

Strasser, R., Castilho, A., Stadhnann, J., Kunert, R., Quendler, H., Gattinger, P., Jez, J., Rademacher, T., Altmann, F., Mach, L., et al. (2009). Improved virus neutralization by plant-produced anti-HIV antibodies with a homogeneous pi,4-galactosylated N-glycan profile. J. Biol. Chem. 284, 20479-20485.

Strasser, R., Stadlmann, J., Schahs, M., Stiegler, G., Quendler, H., Mach, L., Glossl, J., Weterings, K., Pabst, M., and Steinkellner, H. (2008). Generation of glyco-engineered Nicotiana benthamiana for the production of monoclonal antibodies with a homogeneous human-like N-glycan structure. Plant Biotechnol. J. 6, 392-402.

Suzuki, E., Niwa, R., Saji, S., Muta, M., Hirose, M., Iida, S., Shiotsu, Y., Satoh, M., Shitara, K., Kondo, M., et al. (2007). A nonfucosylated anti-HER2 antibody augments antibody-dependent cellular cytotoxicity in breast cancer patients. Clin. Cancer Res. 13, 1875-1882.

Tashiro, M., Tejero, R., Zimmerman, D. E., Celda, B., Nilsson, B., and Montelione, G. T. (1997). High-resolution solution NMR structure of the Z domain of staphylococcal Protein A. /. Mol. Biol. 272, 573-590.

Terman, D. S. (1985a). Protein A and staphylococcal products in neoplastic disease. Crit. Rev. Oncol. Hematol. 4, 103-124.

Terman, D. S. (1985b). Antitumor effects of immobilized Protein A and staphylococcal products: linkage between toxicity and efficacy, and identification of potential tumoricidal reagents. Eur. J. Cancer Clin. Oncol. 21, 1115-1122.

Torigoe, H., Shimada, I., Saito, A., Sato, M., and Arata, Y. (1990). Sequential 1H NMR assignments and secondary structure of the B domain of staphylococcal Protein A: structural changes between the free B domain in solution and the Fc-bound B domain in crystal. Biochemistry 29, 8787-8793.

Turkova, J. (1999). Oriented immobilization of biologically active proteins as a tool for revealing protein interactions and function. J. Chromatogr. B Biomed. Sci. Appl. 722, 11-31.

102

Twyman, R. M, Schillberg, S., and Fischer, R. (2005). Transgenic plants in the biopharmaceutical market. Expert Opin. Emerg. Drugs 10, 185-218.

Twyman, R. M., Schillberg, S., and Fischer, R. (2007). Molecular farming of antibodies in plants. In: Improvement of Crop Plants for Industrial End Uses (Ranalli, P., ed.), pp. 435-469. Dordrecht, The Netherlands: Springer Netherlands.

Uhlen, M., and Moks, T. (1990). Gene fusions for purpose of expression: an introduction. Methods Enzymol. 185, 129-143.

United States Food and Drug Administration (FDA) Center for Drug Evaluation and Research (2009). Therapeutic Biologic Applications (BLA) - Trastuzumab Product Approval Information - Licensing Action 9/25/98. http://www.fda.gov/Drugs/DevelopmentApprovalProcess/HowDrugsareDevelope dandApproved/ApprovalApplications/TherapeuticBiologicApplications/ucm0805 91.htm. Accessed November 4, 2010.

Valdes, R., Gomez, L., Padilla, S., Brito, J., Reyes, B., Alvarez, T., Mendoza, O., Herrera, O., Ferro, W., Pujol, M., et al. (2003). Large-scale purification of an antibody directed against hepatitis B surface antigen from transgenic tobacco plants. Biochem. Biophys. Res. Commun. 308, 94-100.

Vanhaesebroeck, B., and Waterfield, M. D. (1999). Signaling by distinct classes of phosphoinositide 3-kinases. Exp. Cell Res. 253,239-254.

Varchetta, S., Gibelli, N., Oliviero, B., Nardini, E., Gennari, R., Gatti, G., Silva, L. S., Villani, L., Tagliabue, E., Menard, S., et al. (2007). Elements related to heterogeneity of antibody-dependent cell cytotoxicity in patients under trastuzumab therapy for primary operable breast cancer overexpressing HER2. Cancer Res. 67, 11991-11999.

Vezina, L., Faye, L., Lerouge, P., D'Aoust, M., Marquet-Blouin, E., Burel, C, Lavoie, P., Bardor, M., and Gomord, V. (2009). Transient co-expression for fast and high-yield production of antibodies with human-like N-glycans in plants. Plant Biotechnol. J. 7, 442-455.

Vivanco, I., and Sawyers, C. L. (2002). The phosphatidylinositol 3-kinase-AKT pathway in human cancer. Nat. Rev. Cancer 2, 489-501.

Vogel, C. L., Cobleigh, M. A., Tripathy, D., Gutheil, J. C., Harris, L. N., Fehrenbacher, L., Slamon, D. J., Murphy, M., Novotny, W. F., Burchmore, M., et al. (2002). Efficacy and safety of trastuzumab as a single agent in first-line treatment of HER2-overexpressing metastatic breast cancer. J. Clin. Oncol. 20, 719-726.

103

Wen, X., Yang, G., Mao, W., Thornton, A., Liu, J., Bast, R. C., and Le, X. (2006). HER2 signaling modulates the equilibrium between pro- and antiangiogenic factors via distinct pathways: implications for HER2-targeted antibody therapy. Oncogene 25, 6986-6996.

Wideback, K., and Kronvall, G. (1982). Surface receptors for serum albumin in group C and G streptococci show three different types of albumin specificity. Infect. Immun. 38, 1154-1163.

Wideback, K., Seal, U. S., and Kronvall, G. (1982). Receptor in group C and G streptococci detects albumin structures present in mammalian species. Infect. Immun. 36, 469-475.

Wikstrom, M., Drakenberg, T., Forsen, S., Sjobring, U., and Bjorck, L. (1994). Three-dimensional solution structure of an immunoglobulin light chain-binding domain of Protein L. Comparison with the IgG-binding domains of Protein G. Biochemistry 33, 14011-14017.

Woodard, S. L., Wilken, L. R., Barros, G. O., White, S. G., andNikolov, Z. L. (2009). Evaluation of monoclonal antibody and phenolic extraction from transgenic Lemna for purification process development. Biotechnol. Bioeng. 104, 562-571.

Worthylake, R., Opresko, L. K., and Wiley, H. S. (1999). ErbB-2 amplification inhibits down-regulation and induces constitutive activation of both ErbB-2 and epidermal growth factor receptors. J. Biol. Chem. 21 A, 8865 -8874.

Yakes, F. M., Chinratanalab, W., Ritter, C. A., King, W., Seelig, S., and Arteaga, C. L. (2002). Herceptin-induced inhibition of phosphatidylinositol-3 kinase and Akt Is required for antibody-mediated effects on p27, cyclin Dl, and antitumor action. Cancer Res. 62,4132-4141.

Yang, H., Zhou, B. P., Hung, M., and Lee, M. (2000). Oncogenic signals of HER-2/neu in regulating the stability of the cyclin-dependent kinase inhibitor p27. J. Biol. Chem. 275, 24735-24739.

Yarden, Y., and Sliwkowski, M. X. (2001). Untangling the ErbB signalling network. Nat. Rev. Mol. Cell Biol. 2, 127-137.

Yu, D., McLean, M. D., Hall, J. C , and Ghosh, R. (2008a). Purification of a human immunoglobulin Gl monoclonal antibody from transgenic tobacco using membrane chromatographic processes. J. Chromatogr. A 1187, 128-137.

Yu, D., McLean, M. D., Hall, J. C., and Ghosh, R. (2008b). Purification of monoclonal antibody from tobacco extract using membrane-based bioseparation techniques. J. Memb. Sci. 323, 159-166.

104

Zeman, L., and Zydney, A. (1996a). Biotechnological applications. In: Microfiltration and Ultrafiltration - Principles and Applications, pp. 544-563. New York, NY: Marcel Dekker Inc.

Zeman, L., and Zydney, A. (1996b). Module Design and Membrane Configurations. In: Microfiltration and Ultrafiltration - Principles and Applications, pp. 327-349. New York, NY: Marcel Dekker Inc.

Zheng, D., Aramini, J., and Montelione, G. (2004). Validation of helical tilt angles in the solution NMR structure of the Z domain of staphylococcal Protein A by combined analysis of residual dipolar coupling and NOE data. Protein Set 13, 549-554.

Zhou, B. P., Liao, Y., Xia, W., Spohn, B., Lee, M. H., and Hung, M. C. (2001). Cytoplasmic localization of p21Cipl/WAFl by Akt-induced phosphorylation in HER-2/neu-overexpressing cells. Nat. Cell Biol. 3,245-252.

Zhou, B. P., Hu, M. C , Miller, S. A., Yu, Z., Xia, W., Lin, S., and Hung, M. (2000). HER-2/neu blocks tumor necrosis factor-induced apoptosis via the Akt/NF-icB pathway. J. Biol. Chem. 275, 8027-8031.

105