BAHAGIAN A – Pengesahan Kerjasama*

Adalah disahkan bahawa projek penyelidikan tesis ini telah dilaksanakan melalui

kerjasama antara _______________________ dengan _______________________

Disahkan oleh:

Tandatangan : Tarikh :

Nama :

Jawatan :

(Cop rasmi)

* Jika penyediaan tesis/projek melibatkan kerjasama.

BAHAGIAN B – Untuk Kegunaan Pejabat Sekolah Pengajian Siswazah

Tesis ini telah diperiksa dan diakui oleh:

Nama dan Alamat Pemeriksa Luar : Prof. Dr. Arbakariya Ariff

Department of Bioprocess Technology,

Faculty of Biotechnology & Biomolecular

Sciences, UPM 43400 Serdang, Selangor.

Nama dan Alamat Pemeriksa Dalam : Prof. Madya Dr. Firdausi Razali

Department of Bioprocess Engineering,

Faculty of Chemical & Natural Resources

Engineering (FKKKSA), UTM, 81310

UTM, Skudai, Johor.

Nama Penyelia Lain (jika ada) :

Disahkan oleh Timbalan Pendaftar di SPS:

Tandatangan : Tarikh :

Nama :

CHROMATOGRAPHIC PURIFICATION STRATEGIES FOR RECOMBINANT

HUMAN TRANSFERRIN FROM SPODOPTERA FRUGIPERDA

WEE CHEN CHEN

A thesis submitted in fulfilment of the

requirements for the award of the degree of

Master of Engineering (Bioprocess)

Faculty of Chemical and Natural Resources Engineering

Universiti Teknologi Malaysia

JUNE 2008

iii

To my beloved grandparents, parents and brothers

iv

ACKNOWLEDGEMENT

In this long research journey, I receive all kind of guidance and support,

technically, financially and also spiritually. Thanks to everyone and the institution

for making this work possible. Special thanks are due to my supervisor, PM. Dr

Azila Abdul Aziz and co-supervisor, Dr. Badarulhisam Abdul Rahman for the

opportunity to be involved in this interesting project, their professional advice and

their encouragement in the effort to complete this research. I appreciate the given

opportunity to have a closer insight into biomanufacturing industry. Thanks also to

Prof. Dr. Michael J. Betenbaugh of Johns Hopkins University, USA for providing

recombinant baculoviruses.

I would like to express gratitude to all Bioprocess Department laboratory staff

especially Puan Siti Zalita, Encik Muhammad, Encik Malek and Encik Yaakop. I

also would like to thank all of the staff of Research Manage Center and Faculty of

Chemical and Natural Resources Engineering especially Cik Yun and Pn Naza.

They have been very helpful. I am also feel gratitude to have a group of kind

labmates and friends. Thanks to Dr. Taher, Wei Ney, Clarence, Hafiz, Kamalesh and

Kian Mou for their knowledge sharing. “Fui Ling, Melissa, Lee Yu and Seat Yee,

thanks for your company and motivation through out this long run”. Not to forget,

thanks to all the teachers and lecturers who had taught me all the basic knowledge.

My deepest appreciation would be dedicated to my sweet family members.

Their patience, consideration, encouragement, consistent support, recognition and

invaluable love make me strong and proud. Thank you.

v

ABSTRACT

Insect cell-baculovirus system is an excellent artificial system for the

production of recombinant glycoprotein despite its glycosylation deficiencies. In this

study, laboratory scale production of recombinant human transferrin (rhTf) from

insect cell-BEVS was conducted and chromatographic purification strategies were

employed to obtain rhTf in high yield and high recovery. Research was started with

the amplification of recombinant baculovirus, using low multiplicity of infection

(MOI). Virus stock in a 1.2 x 109 pfu/ml infected suspension culture of Spodoptera

frugiperda (Sf9) at 15 MOI had produced 31µg/ml of rhTf. To purify the rhTf,

hydrophobic interaction chromatography, dialysis and ion exchange chromatography

were performed. For hydrophobic interaction chromatography, elution strategy,

flowrate and rhTf loading capacity of phenyl sepharose were optimized. By loading

38µg rhTf/ml of gel, employing step elution with 50% 1.2M (NH4)2SO4/0.4M

Na3C6H5O7, pH6 (buffer A) and 25% buffer A and flowrate at 1ml/min, 74.6% of

rhTf had been recovered from phenyl sepharose. For ion exchange chromatography,

batch purification in reduced size was used to select suitable anion exchange matrix,

suitable pH of equilibration buffer and concentration of equilibration buffer. 20mM

Tris/HCl buffer, pH8.5 and gradient elution with the increase of of 5mM NaCl/CV

succeeded in giving pure rhTf with 52.5% recovery from Q-sepharose. The overall

recovery of pure rhTf was 34% with 200 purification fold. A brief glycan

characterization of the recovered pure rhTf was performed for a better understanding

of the glycosylation feature of this protein expressed using optimized medium from

BEVS. The carbohydrate component of the purified rhTf was determined. The

purified rhTf was hydrolyzed and the release sugar was labeled with 1-Phenyl-3-

Methyl-5-Pyrazolone (PMP) before analysis with High performance Liquid

Chromatography (HPLC). The molar fractions of Man, GlcNAc and Gal of rhTf

were 3.78, 1.69 and 0.93, respectively.

vi

ABSTRAK

Sistem pengekspresan sel serangga-bakulovirus merupakan sistem pilihan

yang baik untuk menghasilkan rekombinan glikoprotein meskipun kekurangan

glikosilasi. Penghasilan produksi skala makmal mendapat rekombinan human

transferrin (rhTf) dari sistem sel serangga-bakulovirus dan strategi purifikasi jenis

kromatografi telah dijalankan untuk mendapatkan rhTf yang tulen dan perolehan

yang tinggi. Kajian bermula dengan peningkatan kuantiti rekombinan bakulovirus

dari gandaan jangkitan (MOI) yang rendah. Stok virus dalam 1.2 x 109 pfu/ml

menjangkiti kultur ampaian sel Spodoptera frugiperda (Sf9) dengan 15 MOI telah

menghasilkan 31µg/ml rhTf. Dalam proses purifikasi, kromatografi saling tindak

hidrofobik, dialisis dan kromatografi penukaran ion telah dijalankan. Bagi

kromatografi saling tindak hidrofobik, strategi elusi, kelajuan dan kapasiti muatan

rhTf ke atas phenyl sepharose telah dioptimumkan. Penggunaan muatan 38µg

rhTf/ml gel dengan elusi berperingkat menggunakan 50% 1.2M (NH4)2SO4/0.4M

Na3C6H5O7, pH6 (larutan penimbal A) and 25% larutan penimbal A dan kelajuan

pada 1ml/min berjaya memperoleh 74.6% rhTf daripada phenyl sepharose. Bagi

kromatografi penukaran ion, purifikasi dalam saiz kecil telah digunakan untuk

memilih matrik penukar ion, pH larutan penimbal pada fasa keseimbangan dan

kepekatan larutan penimbal pada fasa keseimbangan. 20mM Tris/HCl larutan

penimbal, pH8.5 and elusi cerun dengan peningkatan 5mM NaCl/CV berjaya

menghasilkan rhTf tulen dengan 52.5% perolehan daripada Q-sepharose. Perolehan

rhTf tulen secara keseluruhan ialah 34% dengan 200 lipat purifikasi. Pencirian

glikan secara kasar telah dijalankan ke atas rhTf tulen untuk mendapat pemahaman

tentang ciri-ciri glikosilasi bagi protein ini yang diekspresikan dengan sistem

pengekspresan sel serangga-bakulovirus dan media optimum. Komposisi

karbohidrat untuk rhTf tulen telah dikenalpasti. rhTf yang tulen telah dihidrolisis.

Gula telah dilepaskan, dan dilabelkan dengan 1-Phenyl-3-Methyl-5-Pyrazolone

(PMP) sebelum dianalisis dengan menggunakan kromatografi cecair prestasi tinggi

(HPLC). Nilai fraksi molar Man, GlcNAc and Gal daripada rhTf ialah 3.78, 1.69 and

0.93.

vii

TABLE OF CONTENTS

CHAPTER TITLE PAGE

DECLARATION ii

DEDICATION iii

ACKNOWLEDGEMENTS iv

ABSTRACT v

ABSTRAK vi

TABLE OF CONTENTS vii

LIST OF TABLES xii

LIST OF FIGURES xiii

LIST OF SYMBOLS/ ABBREVIATIONS xvi

1 INTRODUCTION 1

1.1 Preface 1

1.2 Objectives 6

1.2 Scopes of Research 6

viii

2 LITERATURE REVIEW 7

2.1 Recombinant Protein Expression System 7

2.2 Insect Cell Baculovirus Expression System 11

2.2.1 Insect cell 11

2.2.2 Baculoviruses 11

2.2.2.1 Invivo and Invitro Replication 13

2.2.2.2 Recombination 15

2.3 Glycosylation 17

2.3.1 N-Glycosylation and O-Glycosylation 17

2.3.2 Glyscosylation Pathway 20

2.3.2.1 Glycosylation Pathway in Insect Cell 21

2.3.3 Model Protein- Transferrin 23

2.3.3.1 Recombinant Human Transferrin 27

2.4 Analysis Method 28

2.4.1 Bicinchoninic Acid (BCA) Assay 28

2.4.2 Enzyme Linked Immunosorbent Assay (ELISA) 29

2.4.3 Sodium Dodecyl Sulfate -Polyacrylamide Gel

Electrophoresis (SDS-PAGE) 31

2.4.4 Western Blot 32

2.4.5 Glucose, Lactic Acid and Glutamine Analyzer 33

2.4.6 Carbohydrate Analysis Using High

Performance Liquid Chromatography (HPLC) 34

2.4.6.1 Hydrolysis 34

2.4.6.2 1-Phenyl-3-Methyl-5-Pyrazolone (PMP)

Derivative of Sugar 34

2.4.6.3 Reverse Phase-HPLC 35

2.5 Purification of Transferrin 36

2.5.1 Hydrophobic Interaction Chromatography

(HIC) 37

2.5.1.1 Factors Affecting HIC 39

2.5.2 Ion Exchange Chromatography 43

2.5.2.1 Factor Affecting IEX 44

2.5.3 Optimization Method in Process

Chromatography 48

ix

2.6 Summary of Literature Review 49

3 MATERIALS AND METHODS 51

3.1 Materials 51

3.1.1 Cell lines and Recombinant Baculovirus 51

3.1.2 Equipments 51

3.1.3 Chemicals 52

3.2 Spodoptera frugiperda (Sf-9) Cells Culture 53

3.2.1 Cells Thawing 53

3.2.2 Cells Count 54

3.2.3 Adapting Serum Contain Culture to Serum Free

Culture 55

3.2.4 Adapting Monolayer Cells to Suspension

Culture 55

3.2.5 Maintaining Suspension Culture 56

3.2.6 Preparation of Optimized Medium 56

3.2.7 Adapting Suspension Culture in SFM900II to

Optimized Medium 57

3.2.8 Cells Freezing 57

3.3 Recombinant Baculovirus 58

3.3.1 Generating Pure Recombinant Virus Stock 58

3.3.2 Amplification of Virus Stock 58

3.3.3 Optimization of rhTf Expression 59

3.3.4 Virus Titration (End-Point Dilution) 59

3.4 Recombinant Human Transferrin Detection 60

3.4.1 Enzyme Linked Immunosorbent Assay

(ELISA) 60

3.4.2 Sodium Dodecyl Sulfate -Polyacrylamide Gel

Electrophoresis 62

3.4.2.1 Silver Staining 63

A

x

3.4.2.2 Coomassie Blue Staining 63

3.4.3 Western Blot 64

3.5 Characterization of Nutrient Consumption and

Substances Release 65

3.5.1 Analysis of Glucose, Lactic Acid and

Glutamine 65

3.5.2 Ammonia Test 66

3.6 Protein Assay 67

3.6.1 Bicinchoninic Acid (BCA) Assay 67

3.7 Purification 68

3.7.1 Hydropbobic Interaction Chromatography 68

3.7.2 Dialysis 69

3.7.3 Initial Screening Step of IEX using Batch

Purification in Reduced Volume 70

3.7.4 Ion Exchange Chromatography 71

3.8 Monosaccharide Composition Analysis of rhTf by

HPLC 72

3.8.1 Preparation of Apotransferrin, rhTf, Standard

Monosccharides 72

3.8.2 Hydrolysis 73

3.8.3 Pre-column Derivatization 73

3.8.4 HPLC Analysis 74

4 RESULTS AND DISCUSSION 75

4.1 Expression of rhTf 75

4.1.1 Growth Profile of Infected Virus 75

4.1.2 Time Course Expression Profile of rhTf 80

4.2 Purification 83

4.2.1 Profile of Sample Elution from Hydrophobic

Interaction Chromatography 85

xi

4.2.2 Optimization of Hydrophobic Interaction

Chromatography 86

4.2.2.1 Optimization of Elution Method 86

4.2.2.2 Optimization of Elution Flowrate 89

4.2.2.3 Optimization of rhTf Loading Capacity 92

4.2.3 Initial Screening Step of IEX Using Batch

Purification in Reduced Volume 95

4.2.4 Anion Exchange Chromatography 98

4.2.4.1 Maximizing The Selectivity of Anion

Exchange Chromatography 98

4.2.5 Characterization of rhTf Purification 100

4.3 Characterization of The Carbohydrate Composition of

rhTf 104

5 CONCLUSIONS 108

5.1 Conclusions 108

5.2 Recommendations 110

REFERENCES 112

APENDICES 137

xii

LIST OF TABLES

TABLE NO. TITLE PAGE

1.1 Comparison of pharmaceutical expression system

(Elbehri, 2005) 3

2.1 Characterization of selected host systems for protein

production from recombinant DNA (Shuler and Kargi, 2002) 10

2.2 Posttranslational processing and yield of the protein product in

various expression systems (cited from Luckow and Summers,

1988) 10

2.3 Selected private company with the protein engineering

platform 26

2.4 Functional groups used on ion exchangers 46

2.5 Capacity data for sepharose fast flow ion exchangers 47

2.6 Characteristics of Q, SP, DEAE and CM Sepharose Fast Flow 47

3.1 Culture volume for different flask size 56

3.2 Specification of YSI calibrator 65

3.3 Applied Condition for different study factors 71

4.1 Summary of the characteristic of small scale production of rhTf 83

4.2 Optimization of step-wise elution method for achieving higher

recovery of rhTf 87

4.3 Optimization of elution flowrate 90

4.4 Optimization of rhTf loading capacity 93

4.5 Summary of the characteristic of purification of rhTf 101

4.6 Carbohydrate Composition Analysis of Glycoprotein 107

xiii

LIST OF FIGURES

FIGURE NO. TITLE PAGE

1.1 Worldwide sales forecast for protein drugs, 2006 and 2011

(Talukder, 2007) 2

1.2 Strength and weaknesses of various expression systems

(Cox, 2004) 4

2.1 Electron micrographs and schematic of baculoviruses 12

2.2 Structural compositions of the two baculovirus phenotypes,

budded virus (BV), and the polyhedron derived virus (PDV) 12

2.3 The baculovirus life cycle in vivo and in vitro 14

2.4 Construction of baculovirus expression vectors 16

2.5 Structure of the N-glycosidic bond and O-glycosidic bond

found in glycoproteins. 18

2.6 Structure of the different types of oligosccharidic chains of

N-glycoproteins 19

2.7 Pathway for generation of the dolichol-linked oligosaccharide donor

for protein N-glycosylation 21

2.8 Protein N-glycosylation pathways in insect and mammalian

cells 22

2.9 A ribbon diagram of a diferric rabbit serum transferrin

molecule 25

2.10 Reaction schematic for BCA assay 29

2.11 Schematic represents the a) Direct Sandwich ELISA; b)

Indirect ELISA; c) Sandwich ELISA; d) Competition ELISA 30

xiv

2.12 SDS-PAGE 31

2.13 Immobilized enzyme biosensor of YSI 33

2.14 Hydrolysis time course of bovine fetuin 34

2.15 Derivatization with pyrazolone derivatives 35

2.16 Different hydrophobic ligands coupled to cross-linked

agarose matrices 40

2.17 The Hofmeister series on the effect of some anions and

cations in precipitating proteins 41

2.18 Relative effects of some salts on the molal surface tension of

water 41

2.19 Effect of pH on protein at different net charge 44

2.20 Ion exchanger types 45

3.1 Schematic representative of the procedures employed for

virus titer-end point dilution 60

3.2 Schematic representative of the procedures used in ELISA

method 61

3.3 Schematic representation of the BCA protein assay 67

3.4 Schematic diagram of the dialysis procedure 70

3.5 Schematic diagram of the set up of the chromatography

equipment. 72

4.1 Photography of control and infected culture 76

4.2 Growth Characteristics of sf9 during rhTf virus propagation 78

4.3 Growth Characteristic of sf9 during rhTf production in

optimized suspension culture 79

4.4 The profile of glucose, glutamine consumption and lactate

formation in supernatant post infection 80

4.5 rhTf production profile in supernatant 81

4.6 Characterization of the rhTf production profile of infected

Sf9, using 9%, Coomassie blue staining, SDS-PAGE 82

4.7 Characterization of the rhTf production profile of infected

Sf9, using Western Blot 82

4.8 Steps and gradient elutions of rhTf from HIC column 85

4.9 HIC chromatograms for the optimization of elution method 88

4.10 HIC chromatograms for the optimization of elution flowrate 91

xv

4.11 The relationship between recovery percentage and loading

capacity 93

4.12 HIC chromatograms for the optimization of rhTf loading

capacity 94

4.13 SDS-PAGE characterizing the elution profile of rhTf 95

4.14 Binding capacity of two anion exchange matrix with Tris and

phosphate buffer used as equilibration buffer 96

4.15 Binding capacity of Q-Sepharose with equilibration buffer of

different pH 97

4.16 Binding capacity of Q-Sepharose with different concentration

of equilibration buffer 97

4.17 Anion exchange chromatograms for the optimization of

selectivity 99

4.18 SDS-PAGE characterizing the elution profile of rhTf 100

4.19 HIC chromatogram characterizing the separation and elution

profile of sample 102

4.20 SDS-PAGE characterizing the separated protein from phenyl

sepharose 6 fast flow column 102

4.21 Anion exchange chromatogram characterizing the separation

and elution profile of sample of after HIC and after dialysis 103

4.22 SDS-PAGE characterizing the separated protein from

Q-Sepharose column 103

4.23 SDS-PAGE characterizing the sample pooled from each

purification step 104

4.24 Chromatogram shows HPLC separation of PMP-labeled

transferrin 106

4.25 Standard calibration graph of monosaccharides 107

xvi

LIST OF SYMBOLS/ ABBREVIATIONS

% Percentage

α Alpha

β Beta

µm Micro meter

°C Degree Celsius

µg Micro gram

µg/ml Micro gram per milliliter

µl Microliter.

µm Micrometer

µmol/ml Micro mol per milliliter

AAGR Average annual growth rate

Ablank Absorbance for blank

AcMNPV Autographa californica multiple nuclear polyhedrosis virus

ACN Acetonitrile

AcNPV Autographa californica nuclear polyhedrolysis

Asample Absorbance for sample

Asn-X-Ser Asparagine-X-Serine

Asn-X-Thr Asparagine-X-Threonine

Astandard Absorbance for standard

ATCC American Tissue Culture Collection

BEVS Baculovirus expression vector system

BHK Baby hamster kidney cells

Bm Bombyx mori

BmNPV Bombyx mori nuclear polyhedrosis virus.

xvii

BV Budded virus

BmNPV Bombyx mori nuclear polyhedrosis virus.

BV Budded virus

BVs Budded viruses

cDNA Complementary deoxyribonucleic acid

cells/ml Cells per milliliter

CHO Chinese Hamster Ovary

CM Carboxymethyl

cm/hr Centimeter per hour

cm2 Centimeter square

CMP-NeuAc Cytidine-5’-monophospho N-acetylneuraminic acid

Cu1+

Cuprous ion

CuSO4•5H2O Copper (II) sulfate pentahydrate

CV Column Volume

DEAE Diethylaminoethyl

DMSO Dimethyl sulphoxide

DNA Deoxyribonucleic Acid

DO Dissolved oxygen

DPA Dipicolylamine

e- Electron

E.coli Escherichia coli

ELISA Enzyme linked immunorsorbent assay

ER Endoplasmic recticulum

FBS Fetal bovine serum

FDA Food and Drugs Administration

Fe3+

Ferric ion

Fuc Fucose

g Gravitational

g/l Gram per liter

Gal Galactose

GalNAc N-Acetylgalactosamine

GDP-mannose Guanosine diphoshate mannose

GlcN Glucosamine

GlcNAc N-Acetylglucosamine

xviii

GLDH Glutamate dehydrogenase

GLP-1 Glucagons-like peptide 1

GLP-1-R Glucagons-like peptide 1-receptor

GMP Good manufacturing practice

gp Glycoprotein

GV Granuloviruses (GV)

H+ Hydrogen cation

H2O2 Hydrogen peroxide

H3PO4 Phosphoric acid

HIC Hydrophobic interaction chromatography

His6 Hexahistidine

HPLC High performance Liquid Chromatografi

HRP Horseradish peroxidase

Hrs Hours

hTf Human transferrin

IEX Ion exchange chromatography

IgG Immunoglobulin G

IMAC Metal affinity chromatography

k constant

Kb/kbp Kilo base pair

kDa Kilo Dalton

M Molar

Man Mannose

Man3–1GlcNAc2 3(Mannose)-2(N-Acetyl Glucosamine)

Man3GlcNAc2 3(Mannose)-2(N-Acetylglucosamine)

Man8–GlcNAc2 8(Mannose)-2(N-Acetylglucosamine)

Man9GlcNAc2 9(Mannose)-2(N-Acetylglucosamine)

MeOH Methanol

mg Milligram

mg/ml Milligram per milliliter

min Minutes

ml/min Milliliter per minutes

mmol/L milli mol per liter

MOI Low multiplicity of infection

xix

MPa Mega Pascal

MW Molecular weight

MWCO Molecular Weight Cut Off

N Normal

N.D Not defined

NaCl Sodium Chloride

NADP+/NADPH Nicotinamide adenine dinucleotide phosphate

Na3C6H5O7 Sodium citrate

NaOH Sodium hydroxide

ng/ml Nanogram per milliliter

NH3 Ammonia

(NH4)2SO4 Ammonium Sulphate

Ni2+

Nickel ion

nm Nano meter

NPV Nucleopolyhedoviruses

O2 Oxygen

OB Occlusion bodies

ODS Octadecyl silica

ODV Occlusion derived virus

OV Occluded virus

p10 Phage-encoded protein-10

PBS Phosphate buffered saline

pfu/ml Plug performing unit per milliliter

pH Potential hydrogen

pI Isoelectric point

PIBs Polyhedral inclusion bodies

pmol Pico mol

PMP 1-Phenyl-3-Methyl-5-Pyrazolone

QAE Quaternary Aminoethyl

Q-sepharose Quaternary ammonium

rhTf Recombinant human transferrin

RP-HPLC Reversed phase HPLC

rpm Rotation per minutes

RT Retention time

xx

S Methyl sulphonate

S. cerevisiae Saccharomyces cerevisiae

SDS Sodium dodecyl sulfate

SDS-PAGE Sodium dodecyl sulfate-polyacrylamide gel electrophoresis

SFM Serum Free Medium

SP Sulphopropyl

T.ni Trichoplusia ni

TBS Tris buffered saline

TCID50 50 % Tissue Culture Infectious Dose

TCID50/ml 50 % Tissue Culture Infectious Dose per milliliter

TEMED N,N,N',N'-tetramethylethylenediamine

TFA Trifluoroacetic acid

TM Trademark

TMB 3,3’,5,5’-tetramethylbenzidene

TN5B1-4 High 5

TOI Time of Infection

Tris-HCl Tromethamine and Hydrochloric Acid

UDP Uridine-5’-diphophate

UDP-Gal Uridine-diphosphate galactose

UDP-Glc Uridine-diphosphate glucose

UDP-GlcNAc Uridine-diphophate N-acetylglucosamine

V Volts

W.R Working reagent

xxi

LIST OF APPENDICES

APPENDIX NO. TITLE PAGE

A-1 Stock Solution for SDS-PAGE 134

A-2 Working Solution for SDS-PAGE 135

A-3 Separating and Stacking Gel Preparation 136

B Coomassie Blue Staining 137

C Preparation of Optimized Medium 138

D Example of TCID50 Calculation (spreadsheet) 139

E Working Solution for ELISA 141

F Working Solution for Western Blot 143

G Mobile Phase for Purification 144

H Glycan Analysis 145

CHAPTER 1

INTRODUCTION

1.1 Preface

The biopharmaceutical industry has experienced a significant transformation

based on the development of recombinant DNA and hybridoma technologies in the

1970s. The industry has moved beyond simple replication of human proteins (such

as insulin or growth hormones) and played a key role in the development of large-

molecule drugs such as any protein, virus, therapeutic serum, vaccine, and blood

component. These genetically engineered therapeutic drugs are targeting some of the

major illnesses such as cancer, cardiovascular, and infectious diseases and they have

the full potential to tackle a whole array of new diseases effectively and safely.

By mid 2003, 148 biopharmaceuticals proteins were approved in the United

States and Europe compared to 84 in 2000 (Birch and Onakunle, 2005). The total

global market for protein drugs was $47.4 billion in 2006 and the market is presumed

to reach $55.7 billion by the end of 2011 with an average annual growth rate

(AAGR) of 3.3% (Figure 1.1). It is expected that current cell culture facilities are

unlikely to meet expected demand. The imbalance of supply-demand is

2

expected to get worse in the future, as more biotech therapeutics proteins are

approved. 20–50% of potential therapeutics could be delayed due to the lack of

manufacturing capacity (Fernandez et al., 2002). Hence, the ability in expanding the

existing capacity and producing a larger variety of products are crucial in order to

meet future demand. Drug companies and biotech firms are considering alternative

manufacturing platforms, besides increasing fermentation capacity (Table 1.1)

(Elbehri, 2005).

Figure 1.1: Worldwide sales forecast for protein drugs, 2006 and 2011 (Talukder,

2007).

Generally, recombinant therapeutic protein can be generated and produced in

various prokaryotic and eukaryotic expression systems. Until the early 1990s, the

majority of recombinant proteins were expressed in either microbial or mammalian

cell culture systems. The first approved recombinant therapeutic glycoproteins,

insulin is produced from Escherichia coli. Today, the manufacturing of

biotechnology products relies heavily on the use of mammalian cells, chiefly on

Chinese Hamster Ovary (CHO) cells. The well-known drugs Avonex (interferon

beta 1-a, Biogen, Inc) and EPOGEN/EPREX (epoetin alfa, Amgen Inc/ Ortho

Biotech) are produced in CHO. Insect, transgenic plant, transgenic animal and yeast

3

cells are also attractive as hosts for the production of recombinant proteins, as they

represent potentially inexpensive and versatile expression systems. Optimal

expression system can be varied, based on different critical parameters of the protein

of interest. Selecting an appropriate expression system for the protein of interest will

affect factors such as time to market, cost of goods, product characteristics,

regulatory hurdles, and intellectual property (Figure 1.2).

Table 1.1: Comparison of pharmaceutical expression system (Elbehri, 2005).

Expression System Advantages Disadvantages Applications

Cost per gram

Bacteria

Established regulatory track; well-understood genetics; cheap and easy to grow

Proteins not usually secreted; contain endotoxins; no posttranslational modifications

Insulin (E. coli; Eli Lilly); growth hormone (Genentech); growth factor; interferon

N.R

Yeast

Recognized as “safe;” long history of use; fast; inexpensive; posttranslational modifications

Overglycosylation can ruin bioactivity; safety; potency; clearance; contains immunogens/antigens

Beer fermentation; recombinant vaccines; hepatitis B viral vaccine; human insulin

$50-100

Insect cells

Posttranslational modifications; properly folded proteins; fairly high expression levels

Minimal regulatory track; slow growth; expensive media; baculovirus infection (extra step); mammalian virus can infect cells

Relatively new medium; Novavax produces virus-like particles

N.R

Mammalian cells

Usually fold proteins properly; correct posttranslation modifications; good regulatory track record; only choice for largest proteins

Expensive media; slow growth; may contain allergens/ contaminants; complicated purification

Tissue plasminogen activator; factor VIII (glycoprotein); monoclonal antibodies (Hercepin)

$500–5,000

Transgenic animals

Complex protein processing; very high expression levels; easy scale up; low-cost production

Little regulatory experience; potential for viral contamination; long time scales; isolation/GMPs on the farm

Lipase (sheep, rabbits; PPL Therapeutics); growth hormone (goats; Genzyme); factor VIII (cattle)

$20–50

Transgenic plants

Shorter development cycles; easy seed storage/scaling; good expression levels; no plant viruses known to infect humans

Potential for new contaminants (soil fungi, bacteria, pesticides); posttranslational modifications; contains possible allergens

Cholera vaccine (tobacco; Chlorogen, Inc.); gastric lipase (corn; Meristem); hepatitis B (potatoes; Boyce Thompson)

$10–20

N.R- Not Reported

4

The baculovirus expression vector system (BEVS) has a number of

significant advantages over other methods of recombinant protein production. It is

best known as providing quick access to biologically active proteins and used as a

research tool (Cox, 2004). The major advantages of BEVS over bacterial and

mammalian expression system is the very high expression of recombinant proteins

which in many cases are antigenically, immunogenically and functionally similar to

their native counterparts (Goosen, 1993). Lack of adventitious viral agents that

could replicate in mammalian cells (John Morrow, 2007), make BEVS a powerful

manufacturing platform for health care solutions to pandemic, biodefense, and

emergency scenarios (Cox, 2004). However, BEVS also has its limitation in

producing authentic mammalian proteins and glycoproteins. An absence of complex

sugars in BEVS-produced proteins may result in poor pharmacological activity in

vivo due to the rapid clearance from the circulatory system of glycoproteins with

non-human glycans (Betenbaugh et al., 2004)

Figure 1.2: Strength and weaknesses of various expression systems (Cox, 2004).

The deficiency of BEVS in producing mammalian like-glycoproteins of

potential therapeutic is a hot topic among researchers in this field. BEVS had been

reported to produce sialylated complex type N-glycan through the modification of its

metabolic engineering pathway (Betenbaugh et al., 2004; Viswanathan et al., 2005;

5

Yun et al., 2005). Protein Sciences Corporation (PSC) had developed technology for

large-scale (600 L) production of proteins in insect cells using the BEVS (Cox,

2004). Although currently there are no FDA-approved therapeutic proteins

expressed using BEVS, a number of products are in advanced clinical trials and

several are about to get acceptance. Among these, three vaccines that are close to

market are Provenge™, a prostate cancer immunotherapy from Dendreon

(www.dendreon.com); Ceravix™, a papilloma virus vaccine from GlaxoSmithKline

(www.gsk.com); and FluBIOk™ from Protein Sciences, a non-egg based flu vaccine

(John Marrow, 2007).

BEVS have tremendous potential to become the next therapeutic

manufacturing system. In this study, recombinant human transferrin was used as a

model protein. Transferrin was chosen because of the simplicity of its structure and

its recent important role in protein engineering. Non-glycosylated transferrin had

been used as a scaffold to extend the half life of peptide and proteins. Various

chromatographic methods for purification of transferrin have been reported. Among

these reports, Ali et al. (1996) and Ailor et al. (2000) had purified rhTf from sf9 and

Tn cells using phenyl sepharose and Q-Sepharose. In this study, hydrophobic

interaction chromatography utilizing phenyl sepharose was used as the capture step

and IEX chromatography utilizing Q-sepharose was used for further purification of

rhTf. To obtain pure rtTf, optimization of both chromatographic techniques had

been carried out. Basic characterization of the carbohydrate content of the pure rhTf

had also been carried out to get a better understanding of the glycan.

6

1.2 Objectives

The objective of this work was to optimize the chromatographic purification

process of recombinant human transferrin expressed using BEVS to obtain pure rhTf

in improved yield and recovery.

.

1.3 Scopes of Research

The following are the scopes of this work:

1) Propagation of baculovirus.

2) Small scale production of rhTf using optimized medium.

3) Characterization of productivity profiles using SDS-PAGE, ELISA and Western

Blot.

4) Optimization of purification process of rhTf using HIC and IEX.

5) Characterization of the monosaccharide composition of the expressed rhTf.

CHAPTER 2

LITERATURE REVIEW

2.1 Recombinant Protein Expression System

Procaryotic has been employed in the protein manufacturing system and one

of the dominant workhorses for commercial production is E. coli. (Lee, 1996;

Makrides, 1996). Prokaryotic expression systems offer high production yields at

reduced cost (Table 2.1). However, the expressed-recombinant protein is

aglycosylated and lost of biological effector functions (Wright and Morrison, 1997).

E. coli cannot produce some proteins containing complex disulfide bonds or

mammalian proteins that required posttranslational modification for activity. The

system maybe best suited to production of antibody fragments, rather than complete

immunoglobolins because of the complexity of the protein folding pathway. Product

of E. coli was primarily in the form of inclusion bodies, and thus biologically

inactive, misfolded and insoluble. Biologically active proteins can only be recovered

by complicated and costly denatured and refolding processes. Another disadvantage

of the system is release of endotoxins from inclusion bodies which affect the

recovery and purification.

8

Recently, the manufacturing of therapeutic compound relies heavily on the

use of mammalian cells. Recombinant protein production using mammalian cells

offers several advantages over microbial systems (Table 2.1). Mammalian cells are

able to secrete the protein product and perform post-translational modifications

which are necessary for human therapeutics protein. Chinese Hamster Ovary (CHO)

and murine myeloma (NSO) cells are favored because they efficiently assemble

complex multi proteins (such as immunoglobulins) and are believed to synthesize

glycans similar to those found in human glycoproteins (Chu and Robinson, 2001). In

mammalian cells, protein N-glycosylation is carried out by an elaborate, but well-

characterized metabolic pathway (Kornfeld and Kornfeld, 1985; Montreuil et al.,

1995; Varki et al., 1999) and closest to its natural counterpart. However, mammalian

cells have significantly slower growth rates, lower protein expression level and are

much more complex in their nutritional requirements compared to microbes.

Insect and yeast cells are attractive as hosts for the production of recombinant

proteins too, as they represent potentially inexpensive and versatile expression

systems (Table 2.1). Yeast is an attractive host for the expression of heterogous

protein (Reiser, 1990; Romanos et al., 1992; Muneo et al., 1992). It offers the

advantage of both bacterial and mammalian system. Saccharomyces cerevisiae was

the first to be used for the production of recombinant protein such as interferon

(Tuite et. al., 1982) and hepatitis surface antigen (Valenzuela et. al., 1982). The

advantages of yeast expression system are capability in processing authentic and

bioactive mammalian protein, high level of secretion into protein free medium, rapid

growth rate, ease of high density fermentation, scale up without loss of yield, ease of

genetic manipulation, lower cost compare to mammalian expression systems, lack of

endotoxins, lytic viruses and no know panthogenic relationship with man (Li et al.,

2001). However, yeasts sometimes form hypermannosyl glycans and add 50 or even

more Man residues to Man8–GlcNAc2 (Betenbaugh et al., 2004).

Hypermannosylation can hamper downstream processing of recombinant

glycoproteins and may complicate complete molecular characterization of the

molecules (Vervecken et al., 2004).

9

Another important technology that has gained much ground is insect cell

culture (Tulsi. 2004). Insect cells used in conjunction with the baculovirus

expression vector system (BEVS) have been widely used for the mass production of

heterologous proteins (Possee, 1997). The Insect-BEVS has significant advantages

over other methods of recombinant protein production, such as ease of culture, ideal

for suspension culture, ease of scale up, high product expression, high gene

expression, higher tolerance to osmolality and the absence of harmful factors that

could replicate in mammalian cells. The method required for generating and

maintaining baculovirus recombinants and stable insect cell are simple and cost

effective, requiring incubation without the support of carbon dioxide. Glycosylation

was found to be stable and rather insensitive to variations in ammonia concentration,

temperature and dissolved oxygen concentration (Donaldson et al. 1999). In addition

to high gene expression, BEVS allows synthesis of proteins varying in size and in

complexity, posttranslational proteolytic processing, cleavage of signal peptides,

expression of nonspliced genes, and adequate compartmentation of recombinant

proteins (Beljelarskaya, 2002). BEVS also allows simultaneous expression of

several genes and production of heterodimeric proteins in one infected cell (An et al,

1999).

Insect cells-BEVS provide for protein maturation and modification typical of

eukaryotic systems, including glycosylation, phosphorylation, palmitylation

(acylation with fatty acid residues), amidation, and carboxymethylation (table 2.2).

The heterologous proteins are posttranslationally modified in a similar pattern to

those observed in mammalian cells (Ailor and Betenbaugh, 1999). They form

disulfide bonds and assume native secondary and tertiary structures. However,

unlike the multiantennary, sialylated complex N-glycans produced in mammalian

cells, recombinant protein produced in insect cells are typically paucimannosidic

Man3–1GlcNAc2 N-glycans (Betenbaugh et al., 2004). Proteolysis is another

problem of the BEVS system due to its lytic nature. Several cell or baculovirus

proteases are involved in degradation events during protein production by insect cells

which affect both quality and quantity of the product. The problem is exacerbated in

serum free culture where there is lack of protection by serum proteins such as

albumin and macroglobulin.

10

Table2.1: Characterization of selected host systems for protein production from

recombinant DNA (Shuler and Kargi, 2002).

Organism

Characteristic E. coli

Yeast

(S. cerevisiae) Insect Mammalian

High growth rate E VG P-F P-F

Availability of genetic

systems E G F-G F-G

Expression levels E VG G-E P-G

Low-cost media

available E E P P

Protein folding F F-G VG-E E

Simple glycosylation No Yes Yes Yes

Complex glycosylation No No Yesa Yes

Low Levels of

proteolytic degradation F-G G VG VG

Excretion or secretion

P normally

VG in special

cases

VG VG E

Safety VG E E G

E, excellent; VG, very good; G, good; F, fair; P, poor.

aGlycosylation patterns differ from mammalian cells.

Table 2.2: Posttranslational processing and yield of the protein product in various

expression systems (cited from Luckow and Summers, 1988).

Expression System E. coli Yeast cells Mammalian

cells Insect cell

Proteolytic cleavage +/- +/- + +

Glycosylation - + + +

Secretion +/- + + +

Secondary structure formation +/- +/- + +

Phosphorylation - + + +

Acrylation - + + + Amidation - - + +

Protein yield, % dry weight 1-5% 1% <1% 30%

11

2.2 Insect Cell Baculovirus Expression System

2.2.1 Insect cell

The three most popular insect cell lines used in the BEVS are Sf9 and Sf21

from the fall armyworm, Spodoptera frugiperda (S.f), and TN5B1-4 (High 5) from

the cabbage looper, Trichoplusia ni (T.ni). Sf are the most frequently used cell line

and the popularity is due to the effectiveness in making proteins and being the best

cell line for producing viruses (Tulsi, 2004). T.ni is excellent for protein production,

especially secreted protein. However, the high metabolic activity of this cell line

results in a higher proportion of by-product accumulation (Rachel et al., 1995).

Besides that, T.ni has transposons that can inhibit the efficient production of insect-

derived virus-like particles (VLPs) (Tulsi, B 2004). Other cell lines, Bombyx mori

(Bm-N), Mamestra brassicae (e.g., MB0503), and Estigmene acrea are also notable

because of its glycosylation potential. In general, all the cell lines are obtained from

embryonic (Altmann et al., 1999).

2.2.2 Baculoviruses

Baculoviruses (family Baculoviridae) are viral pathogens, which cause fatal

disease in insects, mainly in members of the families Lepidoptera, Diptera,

Hymenoptera and Coleoptera. More than 600 baculoviruses have been identified,

categorized in two subfamilies: the nucleopolyhedoviruses (NPV) and granuloviruses

(GV) (Murphy et al., 1995). Baculoviruses are highly specific, not known to

propagate in any non-invertebrate host. They can reduce the size of insect pests in

agriculture and forestry as alternative to chemical insecticides (Granados and

Federici, 1986; Payne, 1998). Baculovirus genome is replicated and transcribed in

the nuclei of infected host cells. The large Baculovirus DNA (between 80 and 200

12

kb) is double stranded, circular, supercoiled DNA molecules that packaged into rod-

shaped nucleocapsids (Summers and Anderson, 1972; Burgess, 1977), that are

enveloped singly or in bundles by a unit membrane (Figure 2.1). Nucleocapsids exist

in distinctive virion phenotypes: 1) occluded virus (OV) and 2) budded virus (BV)

(Figure 2.2). Size of these nucleocapsids is flexible and large amounts of foreign

DNA can be accommodated by recombinant baculovirus.

Figure 2.1: Electron micrographs and schematic of baculoviruses A) Baculovirus

particles, or polyhedra; B) Cross-section of a polyhedron; C) Diagram of polyhedron

cross-section. Electron micrographs (A&B) by Jean Adams, graphics (C) by V.

D'Amico.

Figure 2.2: Structural compositions of the two baculovirus phenotypes, budded virus

(BV), and the polyhedron derived virus (PDV). Graphics by Kalmakoff & Ward.

13

BVs generally contain a single nucleocapsid and are enclosed in an envelope

obtained as the nucleocapsids bud out through the cell wall. Prior to the budding of

the virus, the cell wall is modified by the addition of the viral protein glycoprotein

(gp) 64. This protein has been shown to be required for effective spread of the virus

within the host. The occlusion derived virus (ODV) is the form of the virus which is

produced in the latter stages of viral infection and is enclosed in a proteinaceous

occlusion body. They allow for horizontal spread of the virus from insect to insect

and allow the virus to persist for long periods in the environment.

2.2.2.1 Invivo and Invitro Replication

Wild-type baculoviruses in both in vivo and invitro conditions exhibit both

lytic and occluded life cycles that develop independently throughout the three phases

of virus replication (Figure 2.3a). In the early phase which is also known as the virus

synthesis phase, the virus prepares the infected cell for viral DNA replication. Steps

of infection include attachment, penetration, uncoating, early viral gene expression,

and shut off of host gene expression. Actual initial viral synthesis occurs 0.5 to 6

hours (hrs) after infection. Late genes that code for replication of viral DNA and

assembly of virus are expressed in the late phase which is also known as the viral

structural phase. Between 6 and 12 h after infection, the cell starts to produce BV,

also called non-occluded virus (NOV) or extracellular virus (EV). The BV contains

the plasma membrane envelope and gp64 necessary for virus entry by endocytosis.

Peak release of extracellular virus occurs, 18 to 36 hrs after infection. The BV is

responsible for cell to cell transmission within an infected insect and cell culture.

In the very late phase, the viral occlusion protein phase, the polyhedrin and

p10 genes are expressed, OV—also called occlusion bodies (OB) or polyhedral

inclusion bodies (PIBs)—are formed between 24 and 96 h after infection. Particles

of OV assemble inside the nucleus, contain nuclear membrane envelopes and are

14

embedded in a homogenous matrix made predominantly of polyhedrin protein

(Rohrmann, 1986; Summers and Smith, 1978). The polyhedrin protein is not

essential for the life cycle in invitro cell culture, but essential in invivo replication for

its dissemination into the environment and allowing primary infection in susceptible

larva. Multiple virions like gp41 and gp74 are produced and surrounded by a

crystalline polyhedra matrix. OV are released when the infected cells lyses.

Occluded virions are protected from desiccation in the environment. Once

ingested, the occlusion body is solubilized in the gut, releasing virions which fuse

with midgut cells. The virion nucleocapsid migrates through the cytoplasm to the

nucleus. The core is uncoated from the capsid structure in the nucleus and

replicated. Secondary infection is mediated by the budded form of the virus entering

adjacent cells via adsorptive endocytosis. In vitro, a polyhedron gene modified to

express a recombinant gene product is used. Recombination takes place within the

insect cells between the homologous regions in the transfer vector and the

baculovirus DNA. Recombinant virus produces recombinant protein and also infects

additional insect cells thereby resulting in additional recombinant virus (Figure 2.3b).

Figure 2.3: The baculovirus life cycle (A) in vivo and (B) in vitro (adapted from

Pharmigen, 1999)

15

2.2.2.2 Recombination

Two of the most common isolates members used in foreign genes expression

are Autographa californica nuclear polyhedrolysis (AcNPV; also written AcNMPV)

and Bombyx mori (silkworm) nuclear polyhedrosis virus (BmNPV). Entire genome

of AcNPV has been mapped and fully sequenced (Ayres et al., 1994; Kool and Vlak,

1993; Harrap, 1972). For the recombinant in vitro infection (Figure 2.1b), the

naturally occurring polyhedrin gene within the wild-type baculovirus genome is

replaced with a recombinant gene or cDNA. Deletional or insertional inactivation of

the polyhedrin gene in AcNPV does not affect virus propagation but results in the

production of occlusion body-negative viruses. Promoters of varying strength and

differential expression during the virus-like cycle like polyhedrin and p10 promoters

can be used to control the expression of foreign gene. The promoter of the

polyhedrin gene has been widely used for directing the high level production of

heterogolous proteins. During the very late phase of infection, the inserted

heterologous genes are placed under the transcriptional control of the strong AcNPV

polyhedron promoter. Thus, recombinant product is expressed in place of the

naturally occurring polyhedrin protein.

The baculovirus genome is generally too large to easily insert foreign genes.

Several procedures were proposed for constructing recombinant baculoviruses,

including direct enzymic ligation of a foreign DNA fragment into the virus genome,

employment of large bacterial plasmids and use of shuttle vectors for insect cells

(Davies, 1994; Peakman et al., 1992; Luckow et al., 1993; Patel et al., 1992). The

most common method is based on homologous recombination between a transfer

vector and a wild-type virus (Matsuura et al., 1987). The transfer vector contains an

appreciable viral DNA fragment and the cDNA to be expressed, which is controlled

by the promoter of a baculovirus gene. It is constructed and amplified in E. coli.

Co-transfection of the transfer vector and AcMNPV DNA into Sf cells allows

recombination between homologous sites, transferring the heterologous gene from

the vector to the AcMNPV DNA (Figure 2.3b, 2.4). AcMNPV infection of Sf cells

results in the shut-off of host gene expression allowing for a high rate of recombinant

16

mRNA and protein production. Recombinant viruses can be easily identified and

purified because they produce occlusion body-negative viruses that formed distinctly

different plaques from wild type virus. Recombinant proteins can be produced at

levels ranging between 0.1% and 50% of the total insect cell protein.

As shown in Figure 2.4, the polyhedron gene (dashed area) is replaced by

foreign gene or the gene of interest (strippled area). Virus DNA and transfer vector

are co-transfected into the host insect cell and homologous recombination between

the flanking sequences common to both DNA molecules occurs. This causes the

insertion of the gene of interest into the viral genome at the polyhedrin locus,

resulting in the production of a recombinant virus genome. Plaque assay used to

screen the wild type and recombinant baculovirus. The genome then undergoes

replication within the host nucleus, generating recombinant baculovirus vector

containing the foreign gene under the control of the strong, late viral polyhedrin

promoter.

Figure 2.4: Construction of baculovirus expression vectors.

17

2.3 Glycosylation

Glycosylation is the process of addition of carbohydrate moiety to proteins or

lipids in covalent chemical linkage. Glycosylation is one of the principal post-

translational modification steps in the synthesis of membrane and secreted proteins.

It is a site specific, enzymatic process which involves a sequential series of trimming

and elongation reactions carried out by enzymes localized along the cellular

secretory pathway. The products of glycosylation are glycoproteins or glycolipids.

Many of the high-value therapeutic proteins in the market and in clinical

development today are glycoproteins. The carbohydrate components of

glycoproteins or the glycan are critical in biologic functions such as immunogenicity,

solubility, receptor recognition, inflammation, pathogenicity, metastasis, and other

cellular processes (Olden et al., 1982). Besides that, the specific glycan structures

are also essential for their structure, stability and functionality (Varki, 1993; Traving

and Schauer, 1998) and affect a number of physiological properties including in vivo

half-life, bioavailability, and tissue targeting.

2.3.1 N-Glycosylation and O-Glycosylation

The two main types of glycosylation are N-linked glycosylation and O linked

glycosylation (Figure 2.5). N-linked glycoprotein consists of glucose, mannose and

N- acetylglucosamine molecules. The glycosylation begins with the addition of 14-

sugar precursor to an asparagine amino acid via an amide bond in an Asn-X-Ser or

Asn-X-Thr motif and X can be any amino acid other than Proline. This entity is then

transferred to the endoplasmic recticulum (ER) lumen and the oligosaccharyl

transferase enzymes continue the glycosylation by attaching the oligosaccharide

chain to asparagine. The oligosaccharide attached protein sequence now folds

correctly and is now translocated to the Golgi body where the mannose residue is

18

removed. N-linked glycosylation is important for the folding of some of eukaryotic

proteins. It occurs widely in archaea, but very rarely in bacteria.

O-linked glycosylation begins with an enzyme mediated addition of N-acetyl-

galactosamine followed by other carbohydrates to hydroxyl group of serine or

threonine residues. O-linked glycosylation occurs at a later stage in protein

processing probably in the golgi apparatus. O-glycosidic chain or O-glycan is

smaller than N-glycan. Termination of O-linked glycans usually includes Gal,

GlcNAc, GalNAc, Fuc, or sialic acid. This linkage is found in mucinous

glycoproteins and fibrillar collagens (Carson, 1992). It is also important to form

components of the extracellular matrix, adhering one cell to another by interactions

between the large sugar complexes.

Figure 2.5: Structure of the N-glycosidic bond and O-glycosidic bond found in

glycoproteins. (Whitaker, 1977)

N-glycosylproteins can be categorized into three forms which are high

mannose type, complex type and hybrid type. High mannose or oligomannosidic

type glycoproteins are uniquely composed of mannose residues (Figure. 2.6). They

are basically the precursors to hybrid and complex type chains and have been

identified in plants, animals and yeast (Cummings et al.. 1989; Montreuil et al.,

1986; Kimura et al., 1992). The complex type glycoprotein contain almost any

number of the other types of saccharides, including more than the original two N-

19

acetylglucosamines. These chains can be still classified as biantennary, triantennary

or tetraantennary based on their branching pattern (Figure. 2.6). Additionally, the

glycans contain galactose, fucose and sialic acids. Hybrid type chains of have

structural features of both the high mannose and complex types (Figure. 2.6).

Figure 2.6: Structure of the different types of oligosaccharidic chains of N-

glycoproteins (adapted from Cummings et al., 1989).

20

2.3.2 Glyscosylation Pathway

In eukaryotic cells, the glycosylation takes place in the membrane cellular

compartments: endoplasmic reticulum (ER), golgi apparatus, lysosomes (Cumming,

1992). The biosynthesis of N-glycoproteins is initiated by the formation of a

precursor, consisting of a lipid, a dolichol, linked to an oligosaccharide by a

pyrophosphate bond (Lennarz, 1975; Waechter and Lennarz, 1976; Parodi and

Leloir, 1979) and followed by the transfer of the oligosaccharide from the precursor

to the protein. Oligosaccharide intermediates destined for protein incorporation are

synthesized by a series of transferases on the cytoplasmic side of the ER while linked

to the dolichol lipid. Following the addition of a specific number of mannose and

glucose molecules, the orientation of the dolichol precursor and its attached glycan

translocate to the lumen of the ER where further enzymatic modification occurs

(Figure 2.7). The completed oligosaccharide is then transferred from the dolichol

precursor to the Asn of the target glycoprotein which catalyzed by a high specific

enzyme, the dolichol pyrophosphoryl oligosaccharide polypeptide

oligosaccharyltransferase (Kaplan et al., 1987). Three glucose and one mannose

residues are then removed by two glucosidases (glucosidases I and glucosidases II)

and by a mannosidase located in the membrane of endoplasmic reticulum. Further

processing includes trimming of residues such as glucose and mannose, and addition

of new residues via transferases in the ER and, to a great extent, in the golgi (Figure

2.7). In the golgi, high mannose N-glycans can be converted to a variety of complex

and hydrid forms which are unique.

21

Figure 2.7: Pathway for generation of the dolichol-linked oligosaccharide donor for

protein N-glycosylation. The first reaction is the transfer of an N-acetylglucosamine-

phosphate from UDP-N-acetylglucosamine to a dolichol phosphate. Then, one N-

acetylglucosamine and five mannose residues are added to this product from UDP-N-

acetylglucosamine and GDP-mannose, respectively. Finally, the complete molecule

is obtained by the addition of four mannosyl three glucose from dolichol-phosphate-

mannose and three glucoses from dolichol-phosphate-glucose (Abeijon and

Hirschberg, 1992).

2.3.2.1 Glycosylation Pathway in Insect Cell

The nature of N-linked glycosylation is dependent on the protein expressed

and the host cell line. Insect cells, like other eukaryotic cells, modify many of their

proteins by N-glycosylation. At the early stage, N-glycosylation in insect cells is

similar to that in mammalian in ER and form Man9GlcNAc2 moiety. Then, this

moiety is usually trimmed to shorter oligosaccharide structures of Man3GlcNAc2 by

exoglycosidases and a glycosyltransferase. Man3GlcNAc2 is the common

intermediate to both mammalian cells and insect cells. In mammalian cells, terminal

glycosyltransferases can elongate this common intermediate to produce hybrid and

complex N-glycans with terminal sialic acids. However, N-glycan processing

22

machinery of insect cell generally does not produce complex, terminally sialylated

N-glycans. In contrast, they have insufficient expression of processing enzymes

including glycosyltransferases responsible for generating complex-type structures

and metabolic enzymes involved in generating appropriate sugar nucleotides. In

some cases, insect cells have a competing exoglycosidase that can remove the

terminal N-acetylglucosamine residue from GlcNAcMan3GlcNAc2-N-Asn. Hence,

the majority of processed N-glycan produced by insect cells is usually the one with

paucimannosidic structure, Man3GlcNAc2-N-Asn (Hollister et al., 2002; Betenbaugh

et al., 2004; Figure 2.8).

Figure 2.8: Protein N-glycosylation pathways in insect and mammalian cells.

Monosaccharides are indicated by their standard symbolic representations, as defined

in the key. The insect and mammalian N-glycan processing pathways share a

common intermediate, as shown. The major products derived from this intermediate

are paucimannose and complex N-glycans in insect and mammalian cells,

respectively (adapted from Jarvis, 2003).

23

A lot of efforts have been done to modify the glycosylation pathway in insect

cells. There are reports that mentioned the treatment of several established insect

cell lines with a β-N-acetylglucosaminidase inhibitor (Watanabe et. al., 2002) and

culture of the cells in the presence of the sialic acid precursor, N-acetylmannosamine

(Joshi et al., 2001) allowed the production of recombinant glycoproteins with

terminally sialylated N-glycans. Co-infection of recombinant baculovirus expressing

the mammalian β1,4-galactosyltransferase and α2,6- sialyltransferase genes (Jarvis

et al. 2001) and the genetically transform insect cell lines with the required-

glycosyltransferases (Breitbach and Jarvis, 2001; Hollister and Jarvis 2001; Joosten

and Shuler 2003; Aumiller et al., 2003; Yun et al., 2005) are able to express

recombinant glycoproteins containing sialic acid residues. CMP-NeuAc metabolic

pathway also has been engineered to produce CMP-NeuAc which is the crucial

substrate for sialylation of glycoproteins (Lawrence et al., 2001; Viswanathan et al.,

2005). In conclusion, engineered insect cell-BEVS is capable of producing complete

glycoprotein.

2.3.3 Model Protein- Transferrin

Transferrin is the major iron-carrier protein in human plasma and

extracellular space in tissues (von Bonsdorff, L et al., 2001). Transferrin is the most

important source of iron for red cells (Ponka, 1997) and erythroid progenitor cells in

the bone marrow. Transferrin can be divided into four main members: the serum

transferrins (STf) from blood stream, the lactoferrins, found in milk, tears and other

bodily secretions of numerous mammals, the ovotransferrins, found in avian egg

white, and the melanotransferrins, found on the surface of melanocytes (Bullen et al.,

1999). Serum transferrin has a role in iron transport around the body.

Ovotransferrin may help protect the developing embryo in the semi-permeable egg

by sequestering iron that microbes need to grow. Lactoferrin can act as a site-

specific DNA binding protein. Transferrin exists as an extracellular protein (He and

24

Furmanski, 1995) and all the members have similar polypeptide folding patterns

(Baker and Lindley, 1992).

Human serum transferrin is a single chain glycoprotein of 679 amino acid

residues, with 19 disulphide bridges, 2 homologous lobes, two asparagines linked

glycan chains and a glycosylation dependent molecular mass in the range of 76±81

kDa (MacGillivray et al., 1982; MacGillivray et al., 1983). The two homologous

lobes, N-lobe and C-lobe of about 330 amino acids which were linked by a short

flexible spacer peptide, contain two dissimilar domains divided by a cleft which is

the binding site for Fe3+

(Bailey et al., 1988; Wang et al., 1992; Figure 1). At the

iron binding site, four of the six Fe3+

co-ordination sites are occupied by the protein

ligands (2 tyrosine, 1 histidine and 1 aspartate residue) and two by the bidentate

carbonate anion (Bailey et al., 1988; Hirose, 2000). Two N-linked oligosaccharides

are found in the C-lobe at aspargine residues Asn413 and Asn611. The glycan

chains are mainly biantennary (85%) and triantennary (15%) complex-type glycans

(Fu and van Halbeek, 1992; Spik et al., 1985). There are 4-6 sialic acid residues per

transferrin molecule. Variation in microheterogeneity of transferrin occurs during

certain physiological and pathological conditions, such as pregnancy, rheumatoid

arthritis, malignancies, alcohol abuse and genetic polymorphism (van Eijk et al.,

1987; de Jong et al., 1990; de Jong et al., 1992; Léger et al., 1989; Yamashita et al.,

1989; Stibler et al., 1978). Anyway, this variation neither influences the secretion

rate of transferrin by hepatoma cells (Bauer et al., 1985) nor the binding of

transferrin to its receptor (Mason et al., 1993).

Transferrin binds iron avidly with a dissociation constant of approximately

1022

M-1

at pH 7.4 (Aisen and Listowsky, 1980). It also capable of binding several

other metals, but with a lower affinity (Harris and Aisen, 1989). Ferric iron couples

to transferrin only in the company of an anion (usually carbonate) that serves as a

bridging ligand between metal and protein (Aisen and Listowsky, 1980; Harris and

Aisen, 1989; Shongwe et al., 1992). Each molecule of transferrin can bind two Fe3+

ions. Upon binding of iron, the lobes undergo a conformational transition from the

apo-structure with an open interdomain cleft to a closed holo-structure (Hirose,

25

2000). Transferrin exists in four iron forms: iron-free apotransferrin, the monoferric

transferrins with iron in the C- or the N-lobe, respectively, and the diferric

holotransferrin (Harris and Aisen, 1989). Under normal condition, all circulating

plasma iron (0.1% of the body iron) is bound to transferrin, and only 20–35%

transferrin is saturated with iron. Transferrin-bound iron which is in redox-inactive

state does not catalyze hydroxyl radical formation (Baldwin et al., 1984).

Transferrin-bound iron is taken up by the cells by receptor mediated endocytosis

(Richardson and Ponka, 1997) whereafter apotransferrin is recycled back to

circulation (Huebers and Finch 1987). Decrease of pH or protonation of the iron

ligands release metal from transferrin. This can be accelerated by other chemical

compounds capable of complexing iron such as pyrophosphates (Morgan, 1979) and

citrate (Gumerov et al., 2003).

Figure 2.9: A ribbon diagram of a diferric rabbit serum transferrin molecule. The

arrow indicates the position of the Fe3+

molecule in the inter-domain cleft in the N-

lobe (Hall et al., 2002).

Transferrin acts as chelating agent, which renders iron soluble under

physiologic conditions and facilitates transport of iron into cells (Lee et al., 2006). It

also plays role as an antioxidant (Chauhan et al. 2004) and anti microbial protein

which prevents iron-mediated free radical toxicity by controlling the level of free

iron and keeps the iron inaccessible from most bacteria and fungi (Weinberg, 1984).

26

Iron binding capacity of transferrin of patients undergoing high dose chemotherapy

(Harrison et al., 1994; Beare and Steward, 1996), myeloablative therapy and bone

marrow stem cell transplantation (Bradley et al, 1997; Sahlstedt et al., 2001) is

exceeded. Patients with leukaemia and other malignancies typically have a low

serum transferrin concentration. Administration of iron-free apotransferrin would be

a better alternative (von Bonsdorff et al., 2001) over clinically used iron chelator,

deferoxamine which has limited efficacy in the binding of non-transferrin-bound iron

and displays dose-related toxicity (Porter et al., 1996).

Table 2.3: Selected private company with the protein engineering platform (Haan

and Maggos, 2004).

Company Plattform

Affibody Uses protein scaffold based on a domain in Protein A to develop antibody-like

molecules

Ambrx Adds non-encoded amino acids to proteins, enabling the synthesis of proteins with

chemical diversity

BioRexis Uses protein scaffold based on transferrin to develop antibody-like molecules, make

fusion proteins and receptor agonists

Borean Uses protein scaffold bases on a C-type lectin to develop antibody-like molecules,

protein trimerization technology

Catalyst Engineers proteases to degrade targeted molecules

Compound

Therapeutics

Uses a fibronectin domain to develop antibody-like molecules; creates bi- functional

proteins with target-binding domain linked to enzymatic domain

KaloBios Develops improved methods for antibody humanization

Pleris Uses protein scaffold based on lipocalin to develop antibody-like molecules

Scil Uses protein scaffold based on gamma-crystallin to develop antibody-like molecules

Selecore Uses protein scaffold based on cysteine knots to develop antibody-like molecules

Trubion Engineers desired effector function into its SMIP antibody-like proteins

Xencor Uses its PDA technology to engineer desired effector function into antibodies and

create dominant-negative proteins and proteins with enhanced properties,

Transferrin also plays an important role in protein engineering. Transferrin

molecules which have multiple surface loops have excellent stability profile and

show non immunogenic behaviour are suitable to be used as scaffold or carrier

protein. Non-glycosylated transferrin has a half life of 14-17 days. A transferrin

27

fusion protein will similarly have an extended half life, provides high bioavailability,

biodistribution and circulating stability (Haan and Maggos, 2004). An alternate

monoclonal antibody (MAbs) using transferrin as scaffold and was produced in a

yeast expression, Trans-bodyTM

had been developed by BioRexis. Glucagons-like

peptide 1 (GLP-1) was also fused to tranferrin as an alternative to Exenatide, the first

GLP-1-R agonist compound for treating diabetes. This product requires less

frequency of parenteral injection time.

2.3.3.1 Recombinant Human Transferrin

The expression of a wide range of human serum transferin (hSTf) variants:

recombinant full-length and the truncated protein are important for mutagenesis

studies (Ali et al., 1996) and for the study of factors affecting mechanism of

homeostasis (Mason et al. 2001). Different expression systems have been applied to

produce recombinant transferrin. The most common system, which uses baby

hamster kidney cells (BHK) (Mason et al. 2001), has been shown to be successful

although the yield is relatively low. The expressed recombinant hTf was comprised

of numerous glycoforms (Mason et al., 1993). Bacterial expression systems have

been reported (Ikeda et al., 1992; Steinlein and Ikeda, 1993; de Smit et al., 1995).

Escherichia coli-expressed hSTf is biologically inactive, largely due to incorrect

intramolecular disulphide bond formation (Ikeda et al. 1992; de Smit et al., 1995).

Functional hSTf N-lobe was efficiently produced using methylotrophic yeast,

Pischia pastoris with a satisfactory high yield but lack of the full length protein

(Steinlein et al. 1995).

Recombinant hSTf expressed from insect cell-BEVS has the same structural

conformation and biological activity as native hSTf (Ali et al., 1996). The

recombinant protein can bind two ferric ions in the presence of bicarbonate, and is

actively taken up by receptor-mediated endocytosis (Ali et al., 1996). Study of

28

Lopez, 1997 showed that, both human and bovine lactoferrin expressed in Mamestra

brassiase cells lack of complex or hybrid structures. The glycan structures of

recombinant hTf expressed in Tn-5B1-4 cells (Ailor et al. 2000) consists of 54%

paucimannosidic, 30.8% high-mannose and 13.9% hybrid glycans with over 50%

containing fucose.

2.4 Analysis Method

2.4.1 Bicinchoninic Acid (BCA) Assay

BCA is a protein quantitation method based on colorimetric detection. The

principles of total protein method can be divided into protein-dye binding chemistry

(coomassie/Bradford) and protein-copper chelation chemistry. A rapid method of

determining the existence of protein is absorbance at UV 280nm. A few assays like

Bradford assay, Lowry assay and Bicinchoninic acid (BCA) assay with different

specifications and sensitivities are the most popular protein methods

BCA assay is a protein-copper chelation assay. BCA Protein Assay

combines the reduction of Cu2+

to Cu1+

by protein in an alkaline medium with the

highly sensitive and selective colorimetric detection of the cuprous cation (Cu1+

) by

bicinchoninic acid (Figure 2.10). Protein chelates the copper in an alkaline

environment and forms a blue colored complex. This is also known as the biuret

reaction. The color development reaction is started when two molecules of BCA™

reagent chelate with one cuprous ion (Cu1+

) (Figure 2.10) and formed a purple

colored product. This water soluble BCA/Copper Complex exhibits a strong linear

absorbance at 562 nm with increasing protein concentrations. The presence of any of

four amino acid residues (cysteine or cystine, tyrosine, and tryptophan) in the amino

acid sequence of the protein strongly influenced the formation of color. The rate of

29

BCA Color Formation is dependent on the incubation temperature, the types of

protein present in the sample and the relative amounts of reactive amino acids

contained in the proteins. Linear working range for BCA assay at 37°C is 20µg/ml-

2000µg/ml and 60°C is 5µg/ml-250µg/ml.

Figure 2.10: Reaction schematic for BCA assay (Pierce Biotechnology, 2005).

2.4.2 Enzyme Linked Immunosorbent Assay (ELISA)

ELISA is a sensitive enzyme immunoassay which has been widely used for

diagnostic purpose. It is a simple and economic analytical method which allows the

use of small volume and avoided the troublesome of separation. ELISA can rapidly

analyze a large number of samples with high sensitivity and precise estimation of

biological parameters. It has been applied in detection and identification of disease

agents; discrimination of disease agents; quantification of agent to estimate parasite

or immunogenic protein loaded in vaccines. Basically, the mechanism of ELISA

involves the immunological reaction of antibodies and antigens, detection of enzyme

linked antibody/antigen and color change of soluble substrate by the enzyme activity.

ELISA can be classified as direct ELISA, indirect ELISA, sandwich ELISA and

competition ELISA (Figure 2.11). The result of ELISA is a color reaction that can

30

be observed visually and read rapidly by multichannel spectrophotometer or ELISA

plate reader.

Figure 2.11: Schematic represents the a) Direct ELISA; b) Indirect ELISA; c)

Sandwich ELISA; d) Competition ELISA. a) Antigen is attached to the solid phase

and detected by enzyme-labeled antibodies. After incubation period and washing,

the substrate system is added and the color is allowed to develop. b) Antibodies

from a particular species react with antigen attached to the solid phase. Any bound

antibodies are detected by addition of an antispecies antiserum labeled with enzyme.

This system is widely used in diagnosis. c) This system exploits the antibodies

attached to the solid phase to capture antigen. This is then detected using an

enzyme-labeled serum specific for the antigen. The detecting antibody can be the

same serum or from different sources. The antigen must have at least two different

antigenic sites. d) The test scheme involves the reaction of two antibodies with an

antigen attached to the solid phase. Competition implies simultaneous addition of

reagents. The degree of inhibition by binding of antibodies contained in sample for a

pretitrated enzyme labeled antibodies reaction is determined (Crowther, 1995).

31

2.4.3 Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis (SDS-

PAGE)

SDS-PAGE is a technique used to separate proteins according to their

electrophoresis mobility. During eletrophoresis, protein molecules will generally

migrate in a direction and at a speed that reflects their size and net charge. The

folding pattern of protein molecules would not affect the mobility because they have

been linearized and form a complex with negatively charged molecules of sodium

dodecyl sulfate (SDS) (Figure 2.12). Therefore, protein molecules migrate as a

negatively charged SDS-protein complex through the porous polyacrylamide gel. A

reducing agent (mercaptoethanol) would break any –S–S– linkages in or between

proteins. Under these conditions, proteins migrate at a rate that reflects their

molecular weight. The migration is proportional to the molecular weight with

formula as below:

log (MASS) = k (Migration Distance) (2.1)

The bands of the separated proteins can only be visualized after stained. Different

staining method with different sensitivity like Coomassie blue staining, silver

staining, zinc staining is available.

(a) (b)

Figure 2.12: SDS-PAGE. (a) Folded single unit protein or protein with 2 subunits

will be denatured, linearized and became single strand negative charge SDS-protein

molecules after heated with SDS and mercaptoethanol. (b) Treated sample (SDS-

32

protein molecules) which is loaded into the well of SDS-Polyacramide gel will

mobilize from upper part of gel to lower part of the gel according to the molecular

weight. Light molecules will move faster than the heavy one.

2.4.4 Western Blot

Western blot or immunoblot is a method to detect a specific protein in a given

sample of tissue homogenate or extract by means of antigenicity and molecular

weight. Proteins are first separated by mass in the SDS-PAGE, and then specifically

detected in the step of immunoassay. The proteins are transferred from

polyacrylamide to a membrane (typically nitrocellulose or PVDF) prior detection.

The immobilization of protein on membranes matrix is preferred than

polyacrylamide gel because the proteins are more accessible, easier to be handled,

smaller amounts of required reagents and shorter processing time (Gershoni and

Palade, 1982). Principle of the immunoassay of Western blot is similar to ELISA

which involves immunological reaction and detection of enzyme linked

antibody/antigen. Various probes are available for the detection of antibody binding,

for example: conjugated anti-immunoglobulins, conjugated staphylococcal Protein

A, which binds IgG of various species of animal and biotinylated primary antibodies.

The applied chromogenic substrate is different from what is used in ELISA method

in which it involves immunoprecipitation instead of showing color change in

soluable solution. Besides chemiluminescent substrates, other possibilities for

probing are fluorescent and radioisotope labels.

33

2.4.5 Glucose, Lactic Acid and Glutamine Analyzer

The main principle of the analyzers is the application of enzyme sensor

technology. The technology is fast and gives accurate measurement. Enzyme sensor

technology employs specific enzyme to catalyze reactions to produce hydrogen

peroxidase. Hydrogen peroxidase is electrochemically oxidized at the anode to

produce signal current which would convert to concentration value base on single

point calibration. The membrane of the enzyme sensor contains three layers (Figure

2.13). The first layer is porous polycarbonate which limits the diffusion of the

analyte to enzyme layer to prevent enzyme-limited reaction. Oxidization of the

analyte takes place once the analytes enters the enzyme layers. Different type of

enzymes is applied for detection of different substance. The specific enzyme

reaction is show as below:

Dextrose + O2 →OxidaseGlu

H2O2 + D-Glucono-δ-Lactone…………………….(2.2)

L-lactate + O2 →− OxidaseLacL

H2O2 + Pyruvate…………………………………(2.3)

L-Glutamine + O2 →aseGluta min

L-Glutamate + NH3……………………………(2.4)

L-Glutamate + O2 →OxidaseGlut

H2O2 + α-Ketoglutarate + NH3……………….(2.5)

H2O2 + O2 →anodePlatinum

2H+ + O2 + 2e

-………………………………………(2.6)

The third layer, cellulose acetate is used to eliminate many electrochemically-active

compounds that could interfere with the measurement and permits only small

molecules, such as hydrogen peroxide, to reach the electrode.

Figure 2.13: Immobilized enzyme biosensor of YSI (adapted from YSI, 2001).

34

2.4.6 Carbohydrate Analysis Using High Performance Liquid

Chromatography (HPLC)

2.4.6.1 Hydrolysis

Accuracy in monosaccharide composition analysis of oligosaccharide and

glycoprotein relies to a large extent on effective hydrolysis. Fu and O’Neill (1995)

have studied in detail the hydrolysis of free N-linked oligosaccharides and intact

glycoproteins at 121°C under various conditions. Figure 2.14 shows that hydrolysis

of fetuin at 121°C with 4N TFA was completed after 3hrs of hydrolysis.

Figure 2.14: Hydrolysis time course of bovine fetuin. Fetuin sample, 0.5mg in 4N

TFA (5ml) was hydrolyzed at 121°C (Fu and O’Neill, 1995).

2.4.6.2 1-Phenyl-3-Methyl-5-Pyrazolone (PMP) Derivative of Sugar

A number of reactions have been reported as pre-column derivatization. The

condensation between the active hydrogen of PMP or 1-(p-methoxy)-phenyl-3-

methyl-5-pyrazolon (PMPMP) with the carbonyl group of the reducing

carbohydrates under slightly basic conditions, resulting in bis-PMP and bis-PMPMP

35

derivatives. The bis-PMP-sugars which have no stereoisomers are used for

component sugar analysis (Honda et al., 1989). The procedure requires slightly

alkaline conditions (pH8.3). PMP reacts with reducing carbohydrates almost

quantitatively under mild reaction conditions without epimerization to yield strongly

UV-absorbing (245 nm) and electrochemically sensitive derivatives. This method is

attractive for the sialylated oligosaccharides because no loss of sialic acid occurs.

The detection limit is 1-0.1 pmol (Honda et al., 1980).

Figure 2.15: Derivatization with pyrazolone derivatives (Hase, 1996)

2.4.6.3 Reverse Phase-HPLC

High-performance liquid chromatography (HPLC) is a form of column

chromatography which is used to separate components of a mixture by using a

variety of chemical interactions between analyte and the chromatography column. It

is also sometimes referred to as high-pressure liquid chromatography. Different type

of HPLC which include normal phase chromatography, reverse phase

chromatography, size exclusion chromatography, ion-exchange chromatography and

bioaffinity chromatography are available. Reversed phase HPLC (RP-HPLC)

consists of a non-polar stationary phase and an aqueous, moderately polar mobile

phase. The stationary phase is a silica bonded with straight chain alkyl group such as

C18H37 or C8H17. RPC operates on the principle of hydrophobic interactions, which

result from repulsive forces between a polar eluent, the relatively non-polar analyte,

and the non-polar stationary phase. Molecules which are more non-polar in nature

36

are retained longer than polar molecules. Retention Time (RT) is increased by the

addition of polar solvent to the mobile phase and decreased by the addition of more

hydrophobic solvent. The retention can be decreased by adding less-polar solvent

(MeOH, ACN) into the mobile phase to reduce the surface tension of water. Mobile

phase modifiers like inorganic salts can causes a moderate linear increase in the

surface tension of aqueous solutions and increase the retention time of analyte.

Another important component influence the retention time is pH. pH can change the

hydrophobicity of the analyte. Most methods use a buffering agent, such as sodium

phosphate, to control the pH. The buffers serve multiple purposes: they control pH,

neutralize the charge on any residual exposed silica on the stationary phase and act as

ion pairing agents to neutralize charge on the analyte.

2.5 Purification of Transferrin

The purity of a protein is a pre-requisite for its structure and function studies

or its potential application. For structure studies or therapeutic applications, protein

of high degree is required. A wide variety of protein purification techniques like gel

filtration chromatography, ion-exchange chromatography, affinity chromatography

and hydrophobic interaction chromatography (HIC), are available. Every separation

technique is important and the application is dependent on target proteins which vary

in biological and physico-chemical properties: molecular size, net charge, biospecific

characteristics and hydrophobicity (Kennedy, 1990; Garcia and Pires, 1993).

Phenyl Sepharose chromatography has been widely used in the purification of

transferrin. Vieira and Schneider, 1993; Choudhury et al., 2002 had used phenyl-

Sepharose CL 4B to purify avian serotransferrin. Testicular transferrin from rat

sertoli cells, rat serum transferrin (Skinner et al., 1984), rhTf from Sf9 (Ali et al.,

1996) and rhTf from Trichopulsia ni cells (Ailor et al., 2000) was also purified using

phenyl sepharose. Ion-exchange chromatography is very popular too. SP and Q-

37

sepharose had been used to purify apotransferrin from human plasma (von Bonsdorff

et al., 2001). Two steps anion chromatography: Q sepharose fast flow and mono Q

had been use to purify ovotransferrin produced from Pichia pastoris (Mizutani et al.,

2004). Steinlein et al., 1995 used Whatman DE52 column to purify N-terminal half

human serum transferrin from Pichia pastoris. Ali et al., 1996 used Q Sepharose as

a second column to purify recombinant human transferrin from Sf9.

Improved recombinant hTf with histidine tagged has employed immobilized

metal affinity chromatography (IMAC) as the main purification method.

Hexahistidine (His)-6 epitope tag hTf from transfected Drosophila melanogaster S2

cells (Lim et al., 2004), His-tagged hTf secreted from transfected BHK (Mason et

al., 2001) were purified using metal chelate column. Transferrin isolated from

Manduca sexta which possesses a large number of histidine residues was purified by

high capacity and low capacity Ni2+

-dipicolylamine (DPA)-Novarose gel

(Winzerling et al., 1995). Affinity colun, anti-hTf-IgG immobilized Sepharose 4

Fast Flow, had been use to purify recombinant human serum transferrin expressed by

Lymantria dispar 652Y cells (Choi et al., 2003) and recombinant His-tagged hTf

from a transformed insect cell line (Tn5b4GalT) (Tomiya et al.,2003).

2.5.1 Hydrophobic Interaction Chromatography (HIC)

Hydrophobic interactions have a great importance in biological systems.

They are the dominant force in protein folding and structure stabilization (Privalov

and Gill, 1988; Dill, 1990a; Murphy et al., 1990; Makhatafze and Privalov, 1995)

and the maintenance of the lipid bilayer structure of biological membranes (Tanford,

1973). Proteins comprise of a number of hydrophobic amino acids, with different

distribution and hydrophobicity. Hence, a specific separation can be possible with

hydrophobic supports or matrices (Ochoa, 1978; Vogel et al., 1983; Lindahl and

Vogel, 1984). Although HIC exploits nonspecific affinities, it has been successfully

38

used for separation purposes as it displays binding characteristics complementary to

other protein chromatographic techniques (Janson and Rydén, 1993).

Many theories about the principle of HIC have been proposed. Porath, 1986,

proposed ‘‘salt-promoted adsorption’’ and suggested a salting-out effect in

hydrophobic adsorption (Porath et al. 1973), which extended the earlier observations

of Tiselius, 1948. Hofstee, 1973 and later Shaltiel and Er-El, 1973 believed that the

mode of interaction between proteins and the immobilized hydrophobic ligands was

similar to the self association of small aliphatic organic molecules in water.

Melander and Horvath, 1977 suggested that hydrophobic interaction accounted for

the by increase in the surface tension of water arising from the structure – forming

salts dissolved in it. Srinivasan and Ruckenstein (1980); Van Oss et al. (1986)

proposed that HIC is due to van der Waals attraction forces between proteins and

immobilized ligands caused by the increase of the ordered structure of water in the

presence of salting out salts.

The commercial availability of well-characterized HIC adsorbents opened

new possibilities for purifying a variety of biomolecules such as serum proteins

(Janson and Låås, 1978; Hrkal and Rejnkova, 1982), membrane-bound proteins

(McNair and Kenny, 1979), nuclear proteins (Comings et al., 1979), receptors

(Kuehn et al., 1980), cells (Hjertén, 1981), and recombinant proteins (Lefort and

Ferrara, 1986; Belew et al., 1991 in research and industrial laboratories. The

principle for protein adsorption to HIC media is complementary to ion exchange

chromatography and gel filtration. HIC can separate the pure native protein from

other forms (Fausnaugh et al., 1984; Regnier, 1987). HIC has also found use as an

analytical tool to detect protein conformational changes.

39

2.5.1.1 Factors Affecting HIC

The main factors affecting HIC are: 1) Ligand type and degree of

substitution, 2) Type of base matrix, 3) Type and concentration of salt, 4) pH, 5)

Temperature and 6) Additives (Amersham Biosciences, 2000).

The type of immobilized ligand determines primarily the protein adsorption

selectivity of the HIC absorbent. HIC contain alkyl or aryl chains of any size, and in

practice, most separation employ phenyl and butyl group. Figure 2.14 showed the

glycidyl ether coupling HIC media, which produces charge free gels and only have

hydrophobic interactions with proteins. At constant substitution, the protein binding

capacities of HIC absorbents, hydrophobicity and the strength of interaction would

increase, but the adsorption selectivity would decrease with increased alkyl chain

length. Increased degree of substitution of immobilized ligand would also increase

the protein binding capacities. At sufficient high degree of ligand substitution or n-

alkyl chain length, the strength of interaction would increase although the apparent

binding capacity of the adsorbent remains constant and the bound solutes are more

difficult to elute due to multi-point attachment (Jennissen and Heilmeyer, 1975;

Rosengren et al., 1975; Lăăs, 1975; Maisano, et al., 1985). The selectivity of a

copolymer support can change even with same type of ligand. The two most widely

used types of support are strongly hydrophilic carbohydrates, e.g. cross-linked

agarose, or synthetic copolymer materials.

40

Figure 2.16: Different hydrophobic ligands coupled to cross-linked agarose matrices

(Amersham Biosciences, 2000).

According to Melander, et al. (1984), the most important parameters that

determine the effect of salt on the retention in HIC are the salt molality and the molal

surface increment of the salt. The effects of salts in HIC can be accounted for

referring to the Hofmeister series for the precipitation of proteins or for their positive

influence in increasing the molal surface tension of water (Figure 2.15, Figure 2.16).

The salts at the beginning of the series promote hydrophobic interactions and protein

precipitation (salting-out or antichaotropic), are considered to be water structuring;

whereas salts at the end of the series (salting-in or chaotropic ions) randomize the

structure of the liquid water and thus tend to decrease the strength of hydrophobic

interactions (Porath, 1987). Salts such as sodium, potassium or ammonium sulfates

are the most effective to promote ligand protein interactions. Magnesium sulphate

and magnesium chloride do not enhance the protein retention despite the fact that

they increase the surface tension of water. Type of salt in the eluent not only altered

the overall retention of the proteins, but also affects selectivity of the separations

(Rippel and Szepesy, 1994).

41

Increasing precipitation (“salting- out”) effect

Anions: PO43-

, SO42-

, CH3·COO-, Cl

-, Br

-, NO3

-, CLO4

-, I

-, SCN

-

Cations: NH4+, Rb

+, K

+, Na

+, Cs

+, Li

+, Mg

2+, Ca

2+, Ba

2+

Increasing chaotropic (“salting-in”) effect

Figure 2.17: The Hofmeister series on the effect of some anions and cations in

precipitating proteins (Amersham Biosciences, 2000b).

Na2SO4>K2SO4>(NH4)2SO4>Na2HPO4>NaCl>LiCl…>KSCN

Figure 2.18: Relative effects of some salts on the molal surface tension of water

(Amersham Biosciences, 2000b).

The concentration of salt strongly influences the selectivity in protein

adsorption and the influence is different and dependent both on the stationary phase

and the buffer salts (Oscarsson and Kårsnås, 1998). In HIC, the use of high salt

concentration on the equilibration buffer and sample solution promotes the ligand–

protein interactions and consequently the protein retention. As the concentration of

such salts is increased, the amount of bound proteins also increases almost linearly

up to a specific salt concentration and continues to increase in an exponential manner

at still higher concentrations. The adsorbed proteins are eluted by step wise or

gradient elution at decreasing salt concentration in the eluent. The viscosity, UV

transparency and stability at alkaline pH values are other important factors in

choosing the neutral salts (Narhi et al., 1989).

In general, an increase in pH weakens hydrophobic interactions (Porath et al.,

1973; Hjertén, S., 1973); a decrease in pH results in an apparent increase in

hydrophobic interactions. This is probably due to changing of charged groups at

different pH and thereby leading to an increase in the hydrophilicity or

42

hydrophobicity of the proteins. Proteins which do not bind to a HIC adsorbent at

neutral pH bind at acidic pH (Halperin et al., 1981). Hjertén et al. (1986) found that

the retention of proteins changed more drastically at pH values above 8.5 and/or

below 5 than in the range pH 5–8.5. These findings suggest that pH is an important

separation parameter in the optimizing the selectivity of hydrophobic interaction

chromatography.

In HIC, increasing the temperature enhances protein retention and lowering

the temperature generally promotes the protein elution (Hjertén et al., 1974). Van

der Waals attraction forces, which operate in hydrophobic interactions (Srinivasan

and Ruckenstein, 1980) increase with increase in temperature (Parsegian and

Ninham, 1970). However, an opposite effect was reported by Visser and Strating

(1975). This apparent discrepancy is probably due to the differential effects exerted

by temperature on the conformational state of different proteins and their solubilities

in aqueous solutions (Amersham Bioscience, 2000).

Additives can be used in HIC, not only to improve protein solubility or to

modify protein conformation, but also to promote the elution of the bound proteins.

The most widely used are water-miscible alcohols (e.g. ethanol and ethylene glycol)

and detergents. Additives decrease the surface tension of water thus weakening the

hydrophobic interactions to give a subsequent dissociation of the ligand-solute

complex. The non-polar parts of alcohols and detergents compete for the adsorption

site to displace the bound proteins. The separation mode involving charged group of

detergent is a mixed ion-exchange hydrophobic interaction process (Janson and

Rydén, 1993). Elution using additive could lead to denaturation of protein. Hence, it

is only applied for cleaning up HIC columns and when other milder conditions do

not promote protein recovery.

43

2.5.2 Ion Exchange Chromatography

Ion exchange is probably the most frequently used chromatographic

technique for the separation and purification of proteins, polypeptides, nucleic acids,

polynucleotides, and other charged biomolecules (Bonnerjera et al., 1986). The

reasons for the success of ion exchange are its widespread applicability, its high

resolving power, its high capacity, and the simplicity and controllability of the

method. Separation in ion exchange chromatography depends upon the reversible

adsorption of charged solute molecules to immobilized ion exchange groups of

opposite charge. Separation is obtained since different substances have different

degrees of interaction with the ion exchanger due to differences in their charges,

charge densities and distribution of charge on their surfaces. These interactions can

be controlled by varying conditions such as ionic strength and pH.

The separation using ion exchange is based primarily on differences in the

ionic properties of surface amino acids. Thus, at a given pH, proteins possess an

overall net charge. The relationship of the protein and the net charge can be

visualized as a titration curve (Figure 2.17). This curve reflects how the overall net

charge of the protein changes according to the pH of the surroundings. The

isoelectric point (pI) of each protein is the pH at which the protein has zero surface

charge. The net charge will be more positive at a pH lower than pI protein; more

negative at a higher pH. Proteins with different pI can be separated by being passed

through chromatofocusing. Selected working pH is 1 unit away from the pI of

protein.

44

Figure 2.19: Effect of pH on protein at different net charge (Amersham Biosciences,

2000b).

2.5.2.1 Factor Affecting IEX

Matrix of IEX may be based on inorganic compounds, synthetic resins or

polysaccharides. The characteristics of the matrix determine its chromatographic

properties such as efficiency, capacity and recovery as well as its chemical stability,

mechanical strength and flow properties. The nature of the matrix will also affect its

behaviour towards biological substances and the maintenance of biological activity.

The first ion exchangers designed for use with biological substances were the

cellulose ion exchangers developed by Peterson and Sober (1956), then Ion

exchangers based on dextran (Sephadex), followed by those based on agarose

(Sepharose) and cross-linked cellulose (Sephacel). Hydrophilic nature of cellulose

has little tendency to denature protein, but it has low capacities and has poor flow

properties due to its irregular shape.

45

An ion exchanger consists of covalently bound charged group to an insoluble

matrix. The charged groups are associated with mobile counter ions which can be

reversibly exchanged with other ions of the same charge without altering the matrix.

Positively charged exchangers have negatively charged counter-ions (anions)

available for exchange and are called anion exchangers; negatively charged

exchangers have positively charged counter-ions (cations) and are termed cation

exchangers (Figure 2.18).

Figure 2.20: Ion exchanger types (Amersham Biosciences, 2000b).

The presence of charged groups is a fundamental property of an ion

exchanger. The type of group determines the type and strength of the ion exchanger;

their total number and availability determines the capacity. Table 2.3 show some

functional groups which have been chosen for use in ion exchangers. Sulphonic and

quaternary amino groups are used to form strong ion exchangers; the other groups

form weak ion exchangers. Strong ion exchangers like sulfo group and quaternary

ammonium (Q) group are completely ionized over a wide pH range whereas with

weak ion exchangers, the degree of dissociation and thus exchange capacity varies

much more markedly with pH. Carboxymethyl (CM) group begin to protonated at

pH below 5, diethylaminoethyl (DEAE) groups become uncharged at pH above

pH8.5. DEAE- and Q- groups are highly charged at low pH, so they also suitable to

purify low pI protein.

46

Table 2.4: Functional groups used on ion exchangers (Amersham Biosciences,

2000b).

Anion Exchangers Functional Group

Diethylaminoethyl (DEAE) -O-CH2-CH2-N+H(CH2CH3)2

Quaternary aminoethyl (QAE) -O-CH2-CH2-N+(C2H5)2-CH2-CHOH-CH3

Quaternary ammonium (Q) -O-CH2-CHOH-CH2-O-CH2-CHOH-CH2-N+(CH3)3

Cation Exchangers Functional Group

Carboxymethyl (CM) -O-CH2-COO-

Sulphopropyl (SP) -O-CH2-CHOH-CH2-O-CH2-CH2-CH2SO3-

Methyl sulphonate (S) -O-CH2-CHOH-CH2-O-CH2-CHOH-CH2SO3-

The pH in the micro environment of an ion exchanger is not exactly the same

as eluting buffer because Donnan effect can repel or attract protons within the

adsorbent matrix. In general, pH in the matrix is up to 1 unit higher than that in the

surrounding buffer in anion exchanger and 1 unit lower in cation exchanger. The

lower the ionic strength of the buffer, the larger the Donnna effect. This

phenomenon is very important considering the stability of enzymes as a function of

pH. The Donnan effect limits the operational pH range of ion exchangers, especially

in the mildly acid range.

The charges, the nature of the matrix particles in terms of bead size, flow rate

required, capacity also determine the choice of adsorbent. Table 2.4 and Table 2.5

show the capacity data and the characteristics of 4 common commercial ion

exchange matrices.

47

Table 2.5: Capacity data for sepharose fast flow ion exchangers (Amersham

Bioscience, 2000b).

Ion Exchanger Q Sepharose

Fast Flow

SP Sepharose

Fast Flow

DEAE

Sepharose Fast

Flow

CM Sepharose

fast Flow

Total ionic capacity

(µmol/ml gel)

180-250 180-250 110-160 90-130

Dynamic binding capacity*

(mg/ml gel)

Thyroglobulin (MW

669000)

HAS (MW 68000)

α-lactalbumin (MW 14300) IgG (MW 160000)

Bovine COHb (MW 69000)

Ribonuclease (MW 13700)

3

120

110

N.D.

N.D.

N.D.

N.D.

N.D.

N.D.

50

50

70

3.1

110

100

N.D.

N.D.

N.D.

N.D.

N.D.

N.D.

15

30

30

N.D. = Not determined

*For anion exchangers (DEAE and Q) the starting buffer was 0.05 M Tris, pH 8.3

and for cation exchangers (CM and S) 0.1 M acetate buffer, pH 5.0. Limit buffers

were the respective start buffers containing 2.0 M NaCl.

Table 2.6: Characteristics of Q, SP, DEAE and CM Sepharose Fast Flow

(Amersham Biosciences, 2000b).

Product Q Sepharose

Fast Flow

SP Sepharose

Fast Flow

DEAE

Sepharose Fast Flow

CM Sepharose

fast Flow

Type of gel Strong Anion Strong Cnion Weak Anion Weak Cation

Total ionic capacity

(µmol/ml gel)

180-250 180-250 110-160 90-130

Recommended working

flowrate range (cm/hr)

100-300 100-300 100-300 100-300

Approx. mean particle size

(µm)

90 90 90 90

Particle Size Range (µm) 45-165 45-165 45-165 45-165

Working pH Range* 2-12 4-13 2-9 6-10

pH Stability**

Short Term

Long Term

1-14

2-12

3-14

4-13

1-14

2-13

2-14

4.13

* working pH range refers to the pH range over which the ion exchange groups

remain charged and maintain consistently high capacity.

** pH stability, long term refers to the pH interval where the gel is stable over a long

period of time without adverse effects on its subsequent chromatographic

performance. pH stability, short term refers to the pH interval for regeneration and

cleaning procedures

48

2.5.3 Optimization Method in Process Chromatography

Method development work has to focus on purifying the product of interest to

the highest yield and the required purity as quickly, cheaply and easily as possible.

Process chromatography can be divided into capture, intermediate purification and

polishing, based on chromatography purpose. The objective of the capture step is to

adsorb and isolate the protein of interest quickly from the crude sample from other

critical contaminants. It is designed to maximize capacity or speed at the expense of

some resolution. Intermediate purification is used to remove most of the significant

impurities. In this step, achieving resolution of similar components and good

capacity are both important, optional balance should be decided. Polishing step is

focus on achieving the highest possible resolution and removes most impurities

except for some trace amounts. In any chromatography steps, resolution, speed,

capacity and recovery are the important performance properties which need to be

adjusted. The relative priority of these properties is similar for HIC, ion exchange,

gel filtration and affinity chromatography but different depending on the particular

purposes.

In capture step, the prime consideration when optimizing capture step is to

find the highest possible sample load over the shortest possible sample application

time with acceptable loss in yield. The applied media should offer high speed and

high capacity. Selectivity during sample absorption would affect loading capacity.

Typically, binding conditions are selected to avoid binding of contaminating

substances and increase the availability of the protein of interest. Step wise elution is

often applied. Optimization of pH and elution condition is necessary to remove

critical contaminants, provided loading capacity is not affected. For IEX

chromatography, pH buffer is adjusted as near to value isoelectric point (pI) of

protein of interest without losing the capacity and yield. Dynamic capacity of a

chromatography adsorbent for HIC and IEX is a function of the linear flow rate.

Sample loading capacity must be checked at different flow rates to reveal the

optimum level that gives highest productivity. Significant increase in flow rate

during sample application will always give a decrease in dynamic binding capacity.

49

Besides that, the salt concentration during sample application for HIC should not be

too low since this will have a negative effect on dynamic binding capacity.

If a column is used as an intermediate or polishing step, the resolution is

maximized by working on the eluting conditions such as gradient shape, gradient

slope or concentration and volume of steps in a step-wise elution procedure to meet

the purity requirement. Shallow gradients or even isocratic elution is applied in

polishing steps. In many intermediate steps and always in polishing step, the sample

loading capacity is limited by the required resolution since certain bed height is

required to achieve separation between closely related substances. High efficiency

media with small bead size is important in a typical polishing situation because

working on the selectivity alone may not achieved required resolution. Besides that,

the flow rate during elution will affect resolution between compounds to be separated

and also the concentration of these compounds. In HIC, as in ion exchange

chromatography, sample load, flow rate and gradient volume are interrelated. Flow

rate and sample load are optimized to find highest possible productivity where

resolution is still high enough to meet the predefined purity requirements. Increased

flow rate will give a decrease in resolution, but this decrease will be negligible at

high sample loadings. In laboratory separations especially in polishing step, the best

possible separation is frequently a major consideration and the flow rate is frequently

traded off against improved resolution.

2.6 Summary of Literature Review

Literature review gives a brief insight into some closely related topics of the

research, which are expression system, characteristics of model protein, analytical

methods and chromatographic purification. Advantages and disadvantages of

various recombinant protein expression systems which include E. Coli, mammalian

cell, yeast and insect cell are discussed and compared. Review about insect cell-

50

BEVS which include insect cell, baculovirus, in vivo and in vitro replication and

recombination and knowledge about glycosylation which comprises of type of

glycosylation and glycosylation pathway have given a better understanding of the

recombinant protein expression pathway and subsequently lead to a better

manipulation of the system. The model protein, transferrin, is reviewed in terms of

the functionality, structure and physiology. Principle of the analysis methods which

include BCA, ELISA, SDS-PAGE, Western Blot, biochemical analyzer and

carbohydrate analysis using HPLC are illustrated. Chromatographic purification

methods of transferrin are summarized. HIC and IEX are discussed in detail.

Information about optimization methods in process chromatography provides a clear

guideline for the optimization of chromatography, which is a highlighted study in

this research. In summary, the literature review has fully supported the research by

giving both overall picture and sufficient technical knowledge for the

‘chromatographic purification strategies for recombinant human transferrin from

Spodoptera frugiperda’.

CHAPTER 3

MATERIALS AND METHODS

3.1 Materials

3.1.1 Cell lines and Recombinant Baculovirus

Spodoptera frugiperda (Sf9) insect cell was purchased from ATCC

(Rockville, MD). The recombinant baculovirus carrying human transferrin gene was

provided by Prof Dr Michael J. Betenbaugh of Johns Hopkins University, USA.

3.1.2 Equipments

The electrophoresis system used was Mini-Protean II from Bio-Rad

(California, USA). Western blot analysis was done using Mini Trans-Blot

Electrophoretic Transfer Cell from Bio-Rad (Melville, NY). Shimadzu UV-160

spectrophotometer (Minnesota, USA) was used to measure absorbance at 450 nm

52

and 562 nm. The Minitec bench top shaker was purchased from The Tec family

(Switzerland). Class II type biological safety cabinet was from Muaire (Minnesota,

USA). Inverted phase contrast microscope and transmitted light microscope were

from Zeiss Instruments (Germany). Hemacytometer was purchased from Fortuna

(Wertheim, Germany). Incubator was purchased from Memmert (Germany). Fixed

rotor centrifuge and swinging bucket centrifuge were purchased from Selecta

(Germany). Biochemical analyzer YSI 2700 SELECT from YSI Life Sciences

(Ohio, USA) was used to analyze glucose, lactate and glutamine contents. Econo

Column 1.5 x 15 cm and Econo-Column® Flow Adaptor and fraction collector from

Waters (Japan) and peristaltic pump from Pharmacia Biotech (Sweden) were used for

chromatography. Micro 22R centrifuge from Hettich (Germany) and Amicon Ultra-

4 centrifugal filter from Milipore (Ireland) was used to concentrate purified rhTf.

The high performance liquid chromatography (HPLC) systems used was Series 200

from Perkin Elmer (USA)

3.1.3 Chemicals

Fetal bovine serum (FBS), penicillin-streptomycin and Sf-900 II serum free

media (SFM) were from GIBCO BRL (Gaithersburg, MD). Calibrator-human

reference serum, goat anti-human transferrin-affinity purified, goat anti-human

transferrin-horseradish peroxidase (HRP) conjugate, 0.02% peroxidase in citric acid

buffer and 3,3’,5,5’-tetramethylbenzidene (TMB) peroxidase substrate were obtained

from Bethyl Laboratories Inc (Texas). Broad range protein molecular marker and

TMB stabilized substrate for HRP (water insoluble) were purchased from Promega,

(Madison, WI). Acrylamide, bis-acrylamide, bovine serum albumin (BSA),

ammonium persulfate, citric acid, dimethyl sulphoxide (DMSO), glycine, glucose,

glutamine, lipid 1000x, 2-mercaptoethanol, N,N,N',N'-tetramethylethylenediamine

(TEMED), tris, triton X-100 and trypan blue, glucose Sigma Ultra and ammonium

acetate were purchased from Sigma (Missouri, USA). Other products purchased

from Sigma include galactose Sigma Ultra and phenyl-3-methyl-5-pyrazolone (PMP)

53

(Germany), D-glucosamine hydrochloride (China) and mannose (India). Acetic acid,

bromophenol blue, ethanol, dichloromethane sodium chloride, 38% formaldehyde,

hydrochloric acid, glycerol, methanol, phosphoric acid, potassium chloride,

potassium dihydrogen phosphate, potassium phosphate dibasic, skimmed milk,

sodium bicarbonate, sodium dodecyl sulfate (SDS), sodium hydroxide, tween 20,

were from Fluka (Missouri, USA). Ammonium hydroxide, ammonium sulphate and

trifluoroacetic acid were purchased from Merck (New Jersey, USA). Silver nitrate

was purchased from Unilab (Sydney, Australia). Ammonia enzymatic-UV test kit

which consists of Nicotinamide adenine dinucleotide phosphate (NADPH), α-

ketaglutarate, triethanolamine, glutamate dehydrogenase (GLDH), and ammonia

were from Randox Laboratories (Antrim, UK). BCA protein assay kit which

consists of CuSO4•5H2O, BCA™ and tartrate in an alkaline carbonate buffer and

albumin standard were from Pierce (Illinois, USA). D-Glucose, L-Lactate, L-

Glutamine and L-Glutamate calibrator were from YSI laboratory (Ohio. USA). Q-

Sepharose Fast Flow and Phenyl Sepharose 6 Fast Flow (high sub) were purchased

from Amersham Biosciences (Sweden)

3.2 Spodoptera frugiperda (Sf-9) Cells Culture

3.2.1 Cells Thawing

A vial of cells was taken out from the liquid nitrogen facility. The cells were

thawed rapidly, by holding the vial in a water bath at 37°C. The vial was thoroughly

sprayed with ethanol 70% and opened in laminar hood. A small amount of growth

medium, SFM900II/10% FBS was dropped into the vial. Then, the cells were

pipetted out and gently added to 10ml of 10% FBS containing growth medium in a

centrifuge tube. Immediately, the cells viability was checked. A new vial would be

thawed, if there were more dead cells than healthy one. Then, the cells were

centrifuged at 1000rpm, 5min and resuspended in 10ml, 10% FBS & 1% antibiotic

54

containing growth medium. The cell suspension was equally split into two, 25cm2 T-

flask and 5ml each. The cells were incubated at 27°C for about 1 to 2 hours until

they became attached to the surface. The DMSO containing medium was removed

and replaced with fresh growth medium consisting of 10% FBS and 1% antibiotic.

The cells were incubated until they formed a confluent monolayer.

3.2.2 Cells Count

The T-flask was knocked gently to dislodge the adherent and semi adherent

cells into suspension. 100-200 µl of cell suspension was removed under sterile

condition. Then, equal volume of Trypan Blue was added and mixed to the cells

gently. Next, the hemacytometer was cleaned with tissue paper. The cover-slip was

moistened with exhaled breath. Then, the cover-slip was then slid over the chamber

back and forth using slight pressure until Newton’s refraction rings appear (Newton’s

refraction rings are seen as rainbow-like rings under the cover-slip). Both sides of the

chamber (approx. 5-10 µl) were filled with cell suspension. Then, the

hemacytometer was viewed under a light microscope using 20X magnification.

Viable cells were seen as bright cells and the non viable cells were stained blue. The

numbers of viable and non-viable cells within 8 major squares of the chambers were

counted. The concentration of viable and non-viable cells and the percentage of

viable cells were calculated using the equations below.

Viable cells concentration (cells/ml)

= (total non stained cells within 8 major squares) x (dilution factor) x 104/8 3.1

Percentage Viability

= 100 x numbers of viable cells / total numbers of cells. 3.2

55

3.2.3 Adapting Serum Contain Culture to Serum Free Culture

Cells were adapted to serum free medium after a few passages. Confluent

cells were split 1:1 during each passage of subculture. At the next routine passage,

the cells were transferred into a 75% growth medium/10% FBS and 25% serum-free

medium. The cells were allowed to become confluent. At the following passage, the

cells were transferred into an equal mixture of serum-containing medium and serum

free medium. The cells were allowed to grow to confluence. This took 2 to 3 days.

The previous step was repeated if growth was slow. For the next passage, the cells

were transferred into a 75% serum-free medium and 25% serum-consisting medium.

Finally, the cells were split and transferred into 100% serum free medium after

reaching confluence. The cells would take another two to three passages to grow to

optimum densities.

3.2.4 Adapting Monolayer Cells to Suspension Culture

Insect cells were dislodged from the bottom of flasks. Confluent cells from

two units of 75cm2 T-flask would be sufficient to initiate a 50ml suspension culture.

After cell count, cell suspension was diluted to 5x105cells/ml in serum free growth

medium. Suspension culture was maintained in a shaker flask or a spinner flask.

Stirring rate for shaker flask and spinner cultures was started at 100rpm and 75rpm

respectively. Culture condition was maintained at 27°C±5°C. The cells were

subcultured when viable cell density reached 1-2x106cells/ml. Stirring rate was

increased by 5-10rpm with subsequent passage until constant stirring speed reached

130-150rpm for shaker culture and 90-100rpm for spinner culture. If the viability

dropped below 75%, stirring speed would be decreased by 5rpm for one passage till

the culture viability recover to 80% and above.

56

3.2.5 Maintaining Suspension Culture

Insect cell culture was incubated at 27°C in a non CO2 aerated incubator for

both adherent and suspension cultures. Generally, suspension culture was

subcultured twice weekly, centrifuged at 150g (1000rpm) for 5min, and resuspended

in fresh medium once every 3 weeks. For each subculture, confluent cells (2-

3x106cells/ml) were diluted to 5x10

5cells/ml in serum free medium. Stirring rate

maintained at 130rpm-150rpm for shaker and 90rpm-100rpm for spinner flask.

Suitable volume for respective flask size was shown in Table: 3.1. The caps of flask

were loosen to about ¼ to ½ of a turn to maintain the aeration of cultures.

Table 3.1: Culture volume for different flask size.

Flask Size (ml) Shaker Flask Culture

Volume (ml)

Spinner Flask Culture

Volume (ml)

125

250

500

1000

3000

25-50

50-125

125-200

200-400

400-800

50-100

150-200

200-300

300-1000

2000-3000

3.2.6 Preparation of Optimized Medium

Optimized medium was SFM900II added with 2211.2mg/ml of glutamate,

1291.95mg/ml of glucose and 0.64% (v/v) of lipid mixture 1000x. Powder of

glutamate and glucose were dissolved in SFM900II and filtered with nitrocellulose

membrane, 0.22µm. Original stock of glutamate and glucose was prepared in 25g/l

respectively. A defined volume of optimized medium was prepared by mixing the

calculated volume of glucose solution, glutamate solution, lipid mixture and

SFM900II.

57

3.2.7 Adapting Suspension Culture in SFM900II to Optimized Medium

When the suspension culture has reached more than 2x106cells/ml, the culture

was split and optimized medium in equal volume was added in. Then, the cells were

allowed to become confluent again. The culture was split again and equal volume of

optimized medium was added in. This step was repeated for few passages. Finally,

the suspension culture was centrifuged then transferred into 100% optimized

medium. The culture was always seeded at ≈1.0 x 106 cells/ml when optimized

medium was used as the growth medium. Lower seeding densities in optimized

medium would cause the denaturation of cells.

3.2.8 Cells Freezing

Cell density and viability were determined to ensure only cells with density

of at least 1x106cells/ml and 90% viable were used for freezing. The cells were

centrifuged at 1000rpm for 5 min at room temperature, and resuspended in fresh

growth medium /10%FBS. The cell concentration was adjusted to 1-2 x107cells/ml

with fresh growth medium/10%FBS. A freezing mixture which consisting of 80%

(v/v) growth medium/10% FBS and 20% (v/v) DMSO was prepared and chilled to

4°C. An equal volume of freezing mixture was added to the Sf9 cell suspension and

was quickly mixed until homogeneous. 1ml aliquots were placed into each

cryogenic vial. Immediately, the cryogenic vials were placed upright in a 4°C

refrigerator for 15 minutes, -20°C freezer for 1-2 hours and then -70°C to -80°C

freezer for 4-6 hours, or overnight. Finally, the vials were transferred promptly to

the liquid nitrogen storage facility. All frozen cell stocks were recorded.

58

3.3 Recombinant Baculovirus

3.3.1 Generating Pure Recombinant Virus Stock

Purified rhTf virus stock was obtained from Ongkudon (2006). Method of

end point dilution was applied to generate pure virus. The virus stock has to be

diluted until only 10% or less of the total cultures are infected. Cells were diluted to

a concentration of 1-2.5x105cells/ml with growth medium. Virus stock of 10

-6 and

10-7

dilution were prepared. 10µl of each dilution was mixed with 100µl of cell

suspension and seeded into each well of a 96 well plate. 40 replicates were tested for

each dilution of virus. 4 wells were kept uninfected as controls. Plates were

incubated at 27°C in humidified environment for 7 days. Supernatants of all infected

cells and control were tested for product expression using ELISA. Samples that gave

high level of hTf yield were then selected to undergo the purification process twice

more or until the hTf level reached a constant yield provided that other parameters

remained unchanged for every purification round

3.3.2 Amplification of Virus Stock

Purified virus which had been kept in 4°C for few months was amplified

before use. Suspension culture at 0.5-2x106cells/ml was centrifuged and

resuspended in fresh growth medium. Cell was infected with low MOI (0.1-0.2

pfu/cells) by simply adding the required volume of seed virus stock to the suspension

culture. The infected cells were then left for 6 days with stirring. When the cell was

well infected, the medium was harvested by centrifuging at 150g (1000rpm) for 5

minutes. The virus stock was titrated using end point dilution method.

Amplification of virus was repeated until a high titer working stock was obtained.

59

Volume of Inoculum (ml) =(pfu/ml) inoculum viraloftiter

(pfu) cells ofnumber x totalMOI Desired.............(3.3)

3.3.3 Optimization of rhTf Expression

Suspension culture which has adapted to optimized medium was seeded at

1x106cells/ml. When the suspension culture reached 1.6x10

6cells/ml, the culture was

resuspended in fresh optimized medium. The culture was infected with amplified

rhTf baculovirus at day 2, at MOI 15. The infected culture was harvested at day 8 or

day 6 post infection. The product was harvested by centrifuging at 1000g (2600rpm)

for 5 minutes.

3.3.4 Virus Titration (End-Point Dilution)

Virus titration by end point dilution involves the estimation of the dilution of

virus that would infect 50% of the cultures. The quantity of virus is known as 50%

tissue-culture infection dose or TCID50. Virus titers may be expressed as TCID50/ml

or converted to pfu/ml. A serial of dilution of virus stock was prepared from 10-1

to

10-8

. Cells were diluted to a concentration of 1-2.5x105cells/ml with growth

medium. 100µl aliquots of each virus dilution were mixed with 900µl aliquots of the

cell suspension in appendorf tubes. Cell-virus suspension of the same dilution was

added to ten wells of 96 wells flat shaped micro titer plate, in same row and 100µ l

each. This was repeated every row for cell-virus suspension of different dilution

from 10-1

to 10-8

. The last two wells in the row were seeded with 100µl of cells, as

uninfected controls. The plate was sealed in a plastic bag with damp paper towel and

60

incubated at 27°C for 5 to 7 days. Signs of infection in the lower dilutions were

monitored daily.

Figure 3.1: Schematic representative of the procedures employed for virus titer-end

point dilution.

3.4 Recombinant Human Transferrin Detection

3.4.1 Enzyme Linked Immunosorbent Assay (ELISA)

Direct Sandwich ELISA was applied to determine hTf activity. A mixture of

capture antibody (goat anti-human transferrin) and coating buffer with ratio 1:100

was coated on a flat bottom, 96 wells plate or a ELISA plate for 1 hour at room

temperature. After incubation, the capture antibody solution will be aspirated from

each well and washed with washing solution (TBS-Tween-20) 3 times. 200 µl of

blocking solution (1% BSA in TBS) per well was incubated for 30 minutes to block

the plate. Then, the plate was washed three times. The standards (human serum

61

transferrin) and samples were diluted in samples diluents (1% BSA in TBS-Tween

20). The samples were diluted in sample diluents based on the expected

concentration so that they will fall within the concentration range of the standards.

100µl of standards and samples were transferred to assigned wells and incubated for

1hour.

After incubation, samples and standards were removed and the washing steps

were repeated 5 times. HRP-detection antibody (Goat anti-human transferrin-HRP

conjugate) was diluted in conjugate diluents (1% BSA in TBS-Tween 20) with a

range of 1:50,000. 100µl of the HRP-detection antibody was transferred to each well

and incubated for 60 minutes. Washing steps were repeated 5 times after incubation.

Equal volumes of TMB substrate and solution B from ELISA Started Accessory

package were mixed. 100µl of the mixture was incubated in each well for 15

minutes. Blue color developed immediately. The TMB reaction was stopped using

100µl of 1M phosphoric acid (H3PO4). The absorbance was read at 450nm

wavelengths.

Figure 3.2: Schematic representative of the procedures used in ELISA method.

62

3.4.2 Sodium Dodecyl Sulfate -Polyacrylamide Gel Electrophoresis (SDS-

PAGE)

SDS-Page separates proteins based on their molecular weight. SDS-PAGE

was used to check the existence of the desired product. Operation of SDS-PAGE can

basically be divided into 3 steps, (1) preparation of gel (2) electrophoresis of the

sample and (3) staining. The gel can divided into two parts: stacking gel for loading

of samples and separating gel for protein separation. Working solution for separating

and stacking gel was prepared as the methods listed in Appendix A. Plastic gloves

should be worn throughout the preparation because acrylamide is a neurotoxin.

A solution of the separating gel was added with 50µl of ammonium persulfate

and 10µl of TEMED. The solution was mixed well and poured between 2 plates

until about 0.5cm below the level where the wells will be formed by comb. Then,

water was layered on top of the separating gel solution to keep the gel surface flat.

The gel was allowed to polymerize and this took about 30 minutes. After the

separating gel polymerized, water that covered the separating gel was poured off.

Ammonium persulfate and TEMED were added to a solution of the stacking gel.

Stacking gel solution was pipetted onto a separating gel until the top of the front

plate. A comb was inserted carefully into the gel sandwich until the bottom of the

teeth reached the top of the front plate. Stacking gel took less than 30 minutes to

polymerize. After that, the polymerized gel in the plates was attached to the

electrode assembly and then was inserted into the electrophoresis chamber.

Electrophoresis buffer was added to the chamber and covered until the top of the gel.

The comb was removed carefully.

Protein sample was combined with 5x sample buffer at 4:1 in an appendorf

tube. Then, the protein sample was heated at 100ºC for 5 minutes. 20µl sample

solution, 5µl standard hTf and protein molecular weight marker were introduced into

the wells using a sample loading tips. Electrode plugs was attached to proper

electrodes. Power supply was turned on to 100V at a constant voltage for 90 minutes.

63

3.4.2.1 Silver Staining

Silver staining occurs mainly at gel surfaces, so thin gel (0.5-0.75mm) is

suggested. The gel was soaked in 50% methanol/10% acetic acid in a clean

container for at least 1 hour with 2-3 changes of methanol/acetic acid. The gel was

rinsed with water, soaked in water for 30 minutes, with constant stirring and at least

3 changes of water. Solution C was prepared by adding solution A (0.8g silver

nitrate in 4ml distilled water) to solution B (21ml 0.36 % NaOH mixed together with

1.4ml of 14.8M ammonium hydroxide) with constant stirring and then water was

added to make a total volume of 100ml. The gel was removed to a clean container

and stained in Solution C for 15 minutes with gentle, constant agitation. The gel was

rinsed twice and then soaked in deionized water for 2 minutes with gentle agitation.

Gel was removed to a clean container and developed by washing the gel in solution

D (0.05% (v/v) formaldehyde/0.5% (v/v) citric acid 0.1%). Bands appeared in less

than 15 minutes. Reaction was stopped by rinsing in 1% (v/v) acetic acid. The gel

was washed in water for at least 1 hour with at least 3 changes of water

3.4.2.2 Coomassie Blue Staining

.

Coomassie blue staining and destaining solution was prepared according to

the formulation mentioned in Appendix B. The gel was soaked in a staining solution

enough to cover the whole gel and agitated on an orbital shaker for 15 minutes.

Longer staining is required if recycled staining solution was used. After that, the

solution was discarded, and the gel was rinsed with distilled water. The gel was

destained overnight using destaining solution. Frequent changing of the destaining

solution would accelerate destaining of gel.

64

3.4.3 Western Blot

A mini Trans-Blot electrophoretic transfer cell was used to transfer

polyacrylamide gel to cellulose membrane. Bio-Ice cooling system unit was filled

with water and stored at -20°C. Transfer buffer and PBS buffer were prepared

according to the method mentioned in Appendix C. Transfer buffer was chilled to

4°C. The cellulose membrane and filter paper were cut to the dimensions of the gel.

The gel was equilibrated and the membrane, filter paper and fiber pads were soaked

in transfer buffer for 15min to 1hr. Gel holder cassette with gray side down, one

presoaked fiber pad, a sheet of filter paper, an equilibrated gel, a pre-wetted

membrane, another sheet of filter paper and the another fiber pad were placed layer

by layer. The gel sandwich was rolled by a glass roll gently to remove any bubbles.

The cassette was closed firmly and locked with the white latch. Then, the cassette

was placed in an electrode module. After that, the electrode module, frozen Bio-Ice

unit and a stirring bar were placed in buffer tank. The tank was filled with transfer

buffer. Finally, the blot was run on a stirring platform at a fixed voltage of 100V for

1 hour.

After transfer of protein, the membrane was incubated in blocking solution

(5% skimmed milk in PBS) for 1hour at room temperature or overnight at 4°C with

gentle agitation. The membrane was rinsed briefly with washing solution. Washing

step was repeated another 2 times for 10 minutes per each wash and the washing

solution was aspirated after each washing steps. HRP-detection antibody (Goat anti-

human transferrin-HRP conjugate) was diluted in blocking solution at 1:25,000 and

incubated the membrane with gentle agitation for 1 hour. Washing steps was

repeated as mentioned above. Then, the membrane was incubated with TMB

precipitating substrate for 15-30 minutes. A light green color stain developed on

supports bearing horseradish peroxide labeled conjugates.

65

3.5 Characterization of Nutrient Consumption and Substances Release

3.5.1 Analysis of Glucose, Lactic Acid and Glutamine

Glucose, lactic acid and glutamine were analyzed using biochemical analyzer

from YSI utilizing YSI immobilized enzyme membrane. A packet of buffer

concentrate from YSI was reconstituted in 500ml distilled water. Electrical leads

from the sensor were assembly into buffer and the YSI calibrations standard

respectively. YSI immobilized enzyme membrane was gently assembled onto the

probe face. Type of the enzyme membranes that had been used includes glucose

oxidase, lactate oxidase, glutaminase, glutamate oxidase membranes. The instrument

initialized the baseline current and auto calibrated every 15 minutes when it was in

run mode. Sample of about 500µl in appendorf tube was started analyzed once a

stable calibration was established. The samples were diluted with distilled water if

the sample was out of the detection range. The detection range was varied with

different standards.

Table 3.2: Specification of YSI calibrator.

Standards Calibration Point Detection Range

D-Glucose 2.5g/L 0-9g/L

L-Lactate 0.5g/L 0-2.67g/L

L-Glutamate 5.00mmol/L 0-10mmol/L

L-Glutamine 5mmol/L 0-8mmol/L

66

3.5.2 Ammonia Test

Randox’s kit was used to check ammonia content. Enzymatic UV method

was applied. Ammonia combines with α-ketoglutarate and NADPH in the presence

of glutamate dehydrogenase (GLDH) to yield glutamate and NADP+. The

corresponding decrease in absorbance at 340nm is proportional to the plasma

ammonia concentration.

α-ketoglutarate + NH3+

NADPH →GLDH

glutamate + NADP+……..............(3.4)

Each vial of reagent 1 of the kits (0.26mM NADPH/3.88mM α-ketoglutarate)

was reconstituted with 5ml of 0.15M triethanolamine buffer, pH8.6. 0.1ml of water

as blank, standard and samples were pipetted into different cuvettes. Duplicate set

was prepared. Then, 1ml of reagent 1 was added to the cuvettes. The mixture was

mixed and allowed to stand for 5 minutes. The absorbance of the mixture was read

at 340nm. Then, 10µl of GLDH was added to each cuvette. The solution was mixed

and left to stand for 5 minutes. Finally, the absorbance at 340nm was read once

again.

Concentration of ammonia = 294tan

xAA

AA

blankdards

blanksample

−

− µmol/l..……………………(3.5)

blankA = Absorbance (1) for Blank – Absorbance (2) for Blank …………………(3.6)

dardsA tan = Absorbance (1) for Standard – Absorbance (2) for Standard……..…..(3.7)

sampleA = Absorbance (1) for Sample – Absorbance (2) for Sample.……………..(3.8)

67

3.6 Protein Assay

3.6.1 Bicinchoninic Acid (BCA) Assay

BCA protein assay kit was used to quantify total protein. BCA working

reagent was prepared by mixing reagent A with reagent B at a ratio of 50:1.

Sufficient volume of working reagent was prepared for duplicate set of standards and

samples. Serials dilution method was used to prepare the standards. 0.05ml of each

standard and unknown sample replicate was placed into an appropriately labeled test

tube. 1.0 ml of the working reagent was added to each tube and well mixed. For

working ranges between 20-2,000 µg/ml, test tubes were incubated in water bath at

37°C for 30 minutes. For working ranges between 5-250 µg/ml, test tubes were

incubated in water bath at 60°C for 30 minutes. After that, all tubes were cool to

room temperature. With the spectrophotometer set to 562 nm, the reading was auto-

zeroed with cuvettes filled only with water. Subsequently, the absorbances of all the

samples was measured within 10 minutes.

Figure3.3: Schematic representation of the BCA protein assay.

68

3.7 Purification

3.7.1 Hydropbobic interaction Chromatography

Slurry of Phenyl Sepharose 6 fast flow was prepared by decanting 20%

ethanol solution and replacing it with water or other low ionic strength buffer in a

ratio of 50–70% settled gel to 50–30% packing solution. The gel was de-gassed

using a vacuum pump filter system. Econo-column 1.5 x 20 cm and the flow adaptor

from Bio-Rad were used to pack the matrix. The column was flushed with distilled

water to eliminate air from the column dead spaces. Then, the column was closed,

with some water remaining in the column. The slurry was poured into the column in

one continuous motion using a glass rod held against the wall of the column. The

rest of column was filled with distilled water until an upward meniscus was formed

at the top.

The flow adaptor which was connected to the pump was flushed and fully

filled with distilled water. After removing all bubbles, the pump was stopped and the

adaptor was inserted into the top of the column at an angle until it reached the gel

slurry. The adaptor o-ring was kept tight to give a sliding seal on the column wall.

The bottom outlet of the column was opened and the pump was set at the desired

flow rate. Ideally, Phenyl Sepharose 6 Fast Flow matrix should be packed at a

constant pressure of 0.15 MPa (1.5 bar) or flow rate less than 400 cm/hr. The

packing flow rate was maintained for 3 bed volumes till a constant bed height was

reached. The pump was closed, the bottom outlet was closed and the adaptor was

repositioned and locked on the surface of the matrix. The column was ready for used

when the bed medium was stable.

The column was equilibrated with starting buffer (1.2M Ammonium

Sulphate/ 0.4M Sodium citrate buffer, pH6) for 3 column volumes. Sample was

filtered through 0.45 µm membrane, mixed with 2x starting buffer (2.4M

69

Ammonium Sulphate/ 0.8M Sodium citrate buffer, pH6), and loaded into column

using pump. Then, 3 column volumes of starting buffer were used to wash away

unbound proteins. Elution buffer is the mixture of starting buffer and deionized

water. Percentage of the elution buffer was the percentage of starting buffer of the

mixture. For each step elution, three to four column volumes of elution buffer was

applied. Gradient elution was monitored using 2 pumps which drew deionized water

to starting buffer and to the column after homogenously mixing the solution. All the

equilibrating and operating flowrates were the same. Various flowrates, steps elution

and loading capacity were applied and studied for column optimization.

The column was washed with three column volumes of deionized water at

flowrates of 4ml/minute, and re-equilibrate with starting buffer after each run. For

the cleaning in place, precipitated proteins was removed by washing the column with

4 column volume (CV) of 1M NaOH solution at a flow rate of 1.2-1.4ml/min,

followed immediately with 2 to 3 CV of deionized water and re-equilibrated with 5

CV of starting buffer. Strongly hydrophobically bound proteins was removed by

washing the column with 4 CV of 70% ethanol, followed by with water and re-

equilibrated with starting buffer. The column was stored in 20% ethanol in distilled

water at 4°C when not in used.

3.7.2 Dialysis

Dialysis was used for desalting, buffer exchange and removal of small

molecular weight contaminants in samples. Snake SkinTM

pleated dialysis tubing

with 10,000 molecular weight cut off was used. The already-open tubing was pulled

from the stick to the required length. The amount of tubing can be calculated using

3.7ml sample per cm of dry tubing. 2-3 inches of one end of the tubing was briefly

dipped into water and tied tightly in the wetted end of the tubing. Sample was added

into the open end of the tubing. Then, another knot was tied securely in the other

70

open end. Finally, the tubing was immersed in 2 liters 20mM Tris/HCl buffer, pH

8.5, with constant stirring for 24 hours. The buffer was changed every 12 hours.

stirrer

Dialysis

tubing

Schoot

bottle

Buffer

Stirring

Platform

4oC

Figure 3.4: Schematic diagram of the dialysis procedure

3.7.3 Initial Screening Step of IEX using Batch Purification in Reduced

Volume

300µl anion exchange matrix was transferred into appendorf tube. Appendorf

tube was centrifuged at 500g (3000rpm) for 3–5 min to sediment the matrix. The

supernatant was discarded carefully. The matrix was washed five times with 3

matrix volumes of equilibration buffer. For each time, the slurry was centrifuged at

500 × g for 3–5 min and the equilibration buffer was discarded carefully. 500µl of

sample in equilibration buffer was added to the matrix. It was estimated that 1ml of

matrix could bind approximately 30mg of protein. Sample after HIC and after

dialysis was incubated in the matrix and agitated gently on a shaker for 2 hours at

room temperature. After that, the appendorf tube was centrifuged at 500g (3000rpm)

for 3–5 min to sediment the matrix. The supernatant was collected, and the rhTf in

the supernatant was determined using ELISA. Binding capacity was calculated by

minusing the total rhTf in supernatant from the total loaded rhTf per volume of

matrix.

71

Type of matrix, type of equilibration buffer, a range of pH of the selected

equilibration buffer and varied concentration of the selected equilibration buffer was

screened through to look for optimum conditions which give best binding capacity.

All the screened through parameters was listed in Table 3.3.

Table 3.3: Applied condition for different study factors.

Studied Factors Types/Parameters

Matrices Q-Sepharose and DEAE Sephadex A-25

Equilibration buffers Phosphate buffer and Tris/HCl buffer

pH of the equilibration buffer pH 7, pH 7.5, pH 8, pH 8.5, pH 9.5

Concentration of the equilibration buffer 10mM, 20mM, 30mM, 40mM, 50mM,

100mM

3.7.4 Ion Exchange Chromatography

Matrix Q-Sepharose fast flow was settled in starting buffer (20mM Tris

Buffer, pH8.5) and packed as mentioned in 3.7.1. The column was equilibrated with

starting buffer for 3 column volumes. Sample after dialysis in starting buffer was

loaded into the column using a pump. Then, 2 column volumes (CV) of starting

buffer were used to wash away unbound protein. The elution method was a

combination of gradient and steps elution. It was started with a linear gradient

elution where the percentage of buffer B (0.5M NaCl/ Tris Buffer, pH8.5) was

increased from 10% to 20% in nine column volumes and followed by step elution

with 20% of buffer B and lastly 100% of buffer B. Regeneration of Q-sepharose fast

flow was performed by washing 1M NaCl and followed by reequilibrating in 100ml

of starting buffer at flow rates of 4–5 ml/min. Matrix was stored in 20% ethanol in

distilled water at 4°C.

72

Fraction

CollectorTest tube

column

matrix

adaptor

Peristaltic

pumps

Buffer A

stirrer

Stirring

platform

Buffer B

Figure 3.5: Schematic diagram of the set up of the chromatography equipment.

3.8 Monosaccharide Composition Analysis of rhTf by HPLC

3.8.1 Preparation of Apotransferrin, rhTf, Standard Monosccharides

After expression and purification, rhTf was almost ready for monosaccharide

composition analysis. Pure rhTf at 6.69µg/ml was concentrated using Amicon

ultracentrifuge tube, 10000 MWCO, by centrifuging at 5000rpm for 20 minutes,

twice. Standards apo transferrin was prepared in same concentration, 383.08µg/ml.

5mg/ml of stock solution of mannose (Man), glucosamine (GlcN), galactose (Gal)

and glucose (Glc) were prepared and diluted to 50µg/ml. Set of monosaccharides

standards, standard apo-transferrin and sample rhTf in total volume of 100µl were

mixed and prepared as in Appendix H.

73

3.8.2 Hydrolysis

1ml of 4.4M trifluoroacetid acid (TFA) was added to all monosaccharides

standards, apo transferrin standard and rhTf sample in hydrolysis tubes. The

hydrolysis tube was sealed and incubated at 1210C for 4hrs in oven. After being

cooled to room temperature, the hydrolysis tubes were opened. Each reaction

mixture was transferred to a microcentrifuge tube and then evaporated to dryness by

concentration under reduced pressure in a desiccator. The residue was then dissolved

in 0.5ml of 2-propanol and again evaporated to dryness to remove residue TFA.

3.8.3 Pre-column Derivatization

The dried hydrolyzed glycoprotein or oligosaccharides samples,

monosaccharides standards, and the neuraminidase-/ aldolase-treated samples were

directly labeled with PMP by adding 20µl of PMP solution (0.5M in methanol) and

20µl of sodium hydroxide solution (0.3M). Then, the microcentrifuges tubes were

vortexed, and incubated at 70°C for 2hr. After that, the mixture was neutralized by

adding 20µl hydrochloric acid solution (0.3M). Butyl ether (0.5m) was added and

mixed thoroughly by vortexing for at least 5 seconds. Phase separation was

enhanced by brief centrifugation. The organic phase (upper layer) was carefully

removed and discarded. This extraction process was repeated two additional times.

The resulting phase was mixed with 250µl water before being analyzed by HPLC.

74

3.8.4 HPLC Analysis

A Inertsil ODS-2 column (150x3mm) was used to separate PMP-labeled

carbohydrates. The flowrates was set to 200µl/min and the wavelength for UV

detection was 245nm. For neutral and amino sugar separation, buffer A and B were

100mM ammonium acetate (pH5.5) with 10% and 25% acentonitrile, respectively.

A combined gradient of 45% to 55% buffer B in 30 minutes and 25 minutes elution

of 55% buffer B was used for separation. Column was equilibrated to 45% buffer B

before the next run.

CHAPTER 4

RESULTS AND DISCUSSION

Purification is important in the production of human therapeutic protein and

also in structural study. The focus of this research work was to produce rhTf from

Sf9-BEVS and to optimize the purification of rhTf in high yield and recovery.

Hydrophobic interaction chromatography and ion exchange chromatography were

used to purify the rhTf. Optimizations of these two columns were done to improve

the recovery and selectivity of rhTf. Carbohydrate component of the expressed rhTf

was characterized by HPLC.

4.1 Expression of rhTf

4.1.1 Growth Profile of Infected Virus

Sf9 cells were infected with AcMNPV. Success of infection was identified

by visual checking using inverted microscope. Virus infectivity was determined

using the end point dilution method. The physical appearances of rhTf-

76

-AcMNPV-infected sf9 were different from normal healthy cells. As shown in

Figure 4.1, the infected cells were swollen, enlarged and showed rough surfaces

around the cells.

Figure 4.1: Photography of control and infected culture. (a) is image of normal sf9

and (b) is image of infected sf9, observed using a x40 objective on an inverted

microscope.

Purified rhTf-AcMNPV stock which showed very low infectivity response

(<<106 pfu/ml) was amplified to increase the infectivity and virus stock volume for

rhTf production. Amplification of virus was done twice, both at low multiplicity of

infection (MOI), <0.5 pfu/cells to minimize the reproduction of defective interfering

particles (DIPs). These defective particles have extensive mutations in their genome

which resulted in the reduction of infectious virus yield. Infection at high MOI will

lead to a rapid increase in proportion of defective particles (Kool et al., 1991).

For rhTf production, the culture was seeded at 1.6x106cells/ml and infected

with rhTf virus at day 2 and MOI 15. Optimized medium (Ongkudon, 206) was used

to grow the cells. Cell culture was infected at high MOI for the production of

recombinant proteins to ensure synchronous infection of the majority of cells. Cell

was adapted to culture environment before infection. The spent medium might

contain secreted growth promoting factors with a positive effect on protein

production (Jesionowski and Ataai, 1997). Supernatant was harvested at day 6 post

(b) (a)

77

infection, the optimum harvest time reported by Ongkudon (2006). The harvested

rhTf was 31µg/ml.

Figure 4.2 shows the growth characteristic of Sf9 during rhTf virus

propagation. After infection at low MOI, cell viability dropped but cell cultures

continued to grow post infection and reached a plateau at day 6 (day 4 post infection).

Infectivity of virus stock was amplified to 5.6x107pfu/ml after 1st amplification and

1.2x109pfu/ml at final amplification. Virus stock from the second amplification

which showed high infectivity was used for rhTf production. Figure 4.3 shows the

growth characteristics of Sf9 during rhTf production. Sign of infection during rhTf

production was more significant. Cell viability dropped and cells stopped growing at

day 4 (day 2 post infection). Net growth rate and doubling time of cells after

infection at high MOI were 0.006hr-1 and 115hrs respectively and maximum cell

density was 2.99x106cells/ml (Table 4.1).

78

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

45.0

50.0

0 2 4 6

Times (day)

Via

ble

cell

Den

sity

(cel

ls/m

l x105 )

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0

Via

bilit

y,%

Viable Cell Density (control)

Viable Cell Density (1st Amplification)Viable Cell Density (2nd Amplification)

Viability,% (control)Viability,% (1st Amplification)

Viability,% (2nd Amplification)

Figure 4.2: Growth Characteristics of sf9 during rhTf virus propagation. Opened

and closed circles refer to viability percentages (%) and viable cell density (cells/ml)

of noninfected Sf9 as control; square refers to growth profile during first

amplification; triangle refers to second amplification. Arrows shows the day of

infection. Baculovirus was amplified at low MOI and harvested at day 6. Non-

optimized medium (SFM900II) was used. Results are means ±SE for 2 replicates.

79

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

0 2 4 6 8Time (Day)

Via

ble

Cell

Den

sity

x 10

5 (ce

lls/m

l)

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0

Viability Percentage (%

)

Viable Cell DensityViability, %

Figure 4.3: Growth Characteristic of sf9 during rhTf production in optimized

suspension culture. Sf9 was seeded at 1.6x106cells/ml, infected with rhTf virus at

day 2 and MOI 15. Optimized medium was used. Results are means ±SE for 2

replicates.

Glucose was the most important single sources of organic C for insect cells in

all cultures (Bedard et al., 1993). Glutamine was the second most rapidly consumed

amino acid (Bedard et al, 1993). Lipid mixtures, glutamine and glucose were found

to have the most positive effect on rhTf production with more than 95% significance

(Ongkudon, 2006). Optimized medium with the addition of 2211.2mg/L of

glutamine, 1291.95mg/L of glucose and 0.64% (v/v) lipids mixtures 1000x in

SFM900II (Ongkudon, 2006) was used for rhTf production. Time course profiles of

glucose, glutamine consumption and lactate formation in supernatant post infection is

shown in Figure 4.4. Glucose and glutamine was consumed post infection and the

consumption stopped at day 5 and day 4 respectively. Concentration of glucose,

glutamine and lactate in harvested supernatant were 7.41g/L, 15.1mM, 0.45g/L

80

respectively. Concentration of ammonia remaining in the culture was lower than

2mM, which would not affect the growth of sf9 (Bedard and Tom, 1993). Lactate

content was oxidized to carbon dioxide (Chiou et al. 2000) and lactate concentration

drop post infection. Low lactate level maintained pH level and thus improves

productivity (Gorfien et al. 2003). Lactate started to accumulate at day 4 post

infection. This maybe due to the decrease in dissolved oxygen (DO) (Palomares and

Ramirez, 1996).

0

2

4

6

8

10

12

0 1 2 3 4 5 6Day Post infection

Conc

entra

tion

of G

luco

se (g

/L)

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

2.2

Conc

entra

tion

of L

acta

te (g

/L) &

Co

ncen

tratio

n of

Glu

tam

ine

(x10

mM

)

glucose lactate glutamine

Figure 4.4: The profile of glucose, glutamine consumption and lactate formation in

supernatant post infection. Results are means ±SE for 2 replicates.

4.1.2 Time Course Expression Profile of rhTf

Time course expression of rhTf post infection was quantified using ELISA.

As illustrated in Figure 4.5, concentration of rhTf in supernatant increased post

infection. Expression of rhTf started to decrease after day 5 post infection (Figure

81

4.5), however, the expression of rhTf, as reported by Ongkudon (2006), only started

to decrease after day 6. To ensure complete expression of rhTf, rhTf was harvested

at day 6 post infection. The extracellular protein from the expression system was

qualified using SDS-PAGE, stained in Coomassie blue. Protein expression from the

infected cell culture increased daily (Figure 4.6). The most significant protein band

was almost at the same row as standard transferrin and was assumed to be the

expressed rhTf. Expression of rhTf was reconfirmed using Western blot. Protein

was trans-blotted to nitrocellulose membrane from SDS-polyacrylamide gel and

detected by HRP-anti-transferrin antibody and TMB. As shown in figure 4.7, rhTf

was detected. The molecular weight of the expressed rhTf was slightly lower than

that of native human transferrin (76kDa). This might be due to the lack of complex

type oligosaccharides which were attached to the polypeptide (Ailor et al., 2000).

0

5

10

15

20

25

30

35

0 1 2 3 4 5 6 7

Day Post-Infection (Day)

Con

cent

ratio

n of

rhTf

( µg/

ml)

Figure 4.5: rhTf production profile in supernatant. Supernatant rhTf was collected

daily post infection and the rhTf content was checked using ELISA. Miniature graph

in the figure is the rhTf standard curve of ELISA which shows absorbance at 450nm

over logarithm concentration of hTf in ng/ml. Results are means ±SE for 2 replicates.

0 1 2 30.0

0.1

0.2

0.3

0.4

Log[concentration of hTf] ng/ml

Abs

at 4

50nm

82

Figure 4.6: Characterization of the rhTf production profile of infected Sf9, using 9%,

Coomassie blue staining, SDS-PAGE. Lanes d0 to d6 represent supernatants harvest

at day 0 to day 6 post infection. m, molecular weight standards and s, commercial

human transferrin.

Figure 4.7: Characterization of the rhTf production profile of infected Sf9, using

Western Blot. Protein in SDS-PAGE was trans-blotted to nitrocellulose membrane

and detected by HRP-anti-transferrin antibody and TMB. Lanes d0 to d6 represent

supernatants harvested at day 0 to day 6 post infection. m, molecular weight

standards and s, commercial human transferrin.

225kDa 150kDa 100kDa 75kDa 50kDa 35kDa 25kDa

s m d0 d1 d2 d3 d4 d5 d6

s d0 d1 d2 d3 d4 d5 d6

76kDa

Produced rhTf (73kDa)

Produced rhTf

83

Table 4.1: Summary of the characteristic of small scale production of rhTf.

Parameter Small Scale Production of rhTf

Type of culture Suspension

Type of Flask 250ml Shaker Flask

Speed 130rpm

Culture volume per flask 50ml/flask

Type of Medium

Optimized Medium

SFM900II with 2211.2mg/ml of Glutamate;

1291.95mg/ml of Glucose and 0.64% (v/v) of

lipid mixture

Seeding Density 1.6 x 106 cells/ml

MOI 15

TOI 48hrs

Length of cultivation Period 192hrs

Maximum Cell Density 2.99 x 106 cells/ml

Doubling Time (Post Infection) 115hrs

Net Growth Rate (Post Infection) 0.006hr-1

Protein Density 6200 µg/ml

rhTtf Density 31 µg/ml

4.2 Purification

Purification of all kinds of transferrin had been reported in a number of

papers. Hydrophobic interaction chromatography (Skinner et al., 1984; Ali et al.,

1996; Ailor et al., 2000), anion and cation exchange chromatography (Steinlein et al.,

1995; Mizutani et al., 2004), immunosorbent affinity chromatography (Tomiya et al.,

2003; Choi et al., 2003), metal ion affinity chromatography (Mason et al., 2001; Lim

et al., 2004) had been used to purify recombinant transferrin. Among these, Ali et al.

84

(1996) and Ailor et al. (2000) had purified rhTf from sf9 and Tn cells using phenyl

sepharose and Q-Sepharose and recovered rhTf up to 95% purity. Phenyl sepharose

together with affinity gel had been used to purify testicular transferrin from rat sertoli

cells and the 100% pure transferrin gave 28% overall recovery (Skinner et al., 1984).

Choi et al. (2003) purified rhTf from Lymantria dispar 652Y cells using column anti-

hTf-lgG immobilized Sepharose 4 Fast flow. Using immobilized metal affinity

chromatography (IMAC) Hexahistidine (His6) epitope tag hTf with 95.5% purity had

been obtained from transfected Drosophila melanogaster S2 cells with a recovery of

32% (Lim et al., 2004).

Recombinant transferrin from different expression systems needs different

extent of purifying. Some need simple purification, whereas others need more

complicated purification methods. Baculoviruses have lytic infection mode. Large

proportions of the host cells are lysed and degradative enzymes are released when the

products are harvested. Hence, the insect cell-baculovirus system is not considered a

“clean” secretion system (Altmann et al., 1999). It had been reported that neither

apo nor diferric hTf was bound to metal chelate matrix to any appreciable extent.

(Mason et al., 2001). Thus, metal affinity chromatography is not applicable to our

non-histidine tag rhTf. Expression of rhTf, as reported by Ali et al. (1996) and Ailor

et al. (2000), was similar to our expression system. Hence, the reported purification

methods which involving phenyl sepharose and Q-sepharose chromatography were

used as main references. Hydrophobic interaction chromatography utilizing phenyl

sepharose was used as the capture step and IEX chromatography utilizing Q-

sepharose was used for further purification of rhTf.

85

4.2.1 Profile of Sample Elution from hydrophobic Interaction

Chromatography

A screening step which involved a combination of step and gradient elutions

of transferrin from phenyl sepharose was done to understand the elution profile of

sample from HIC column. Sample was loaded at 15µg rhTf/ml of gel. Column was

equilibrated with 100% buffer 1.2M ammonium sulphate/0.4M sodium citrate (buffer

A), pH6.0. Proteins was eluted with 3 column volumes of 50% buffer A, followed

by gradient elution from 50% to 25% buffer A in 10 column volume (CV) and lastly

3 CV of water. Figure 4.8 shows that a large amount of unwanted protein was

hydrophilic and was washed out by the equilibration buffer. The third and second

large peaks in figure 4.8 showed that 50% buffer A eluted protein without rhTf and

0

0.2

0.4

0.6

0 20 40 60 80 100 120 140 160Fraction

Abs

at U

V 2

80nm

0

10

20

30

40

50

60

70

80

90

100

Perc

enta

ge o

f Buf

fer A

(%) &

Co

ncen

tratio

n of

r-hT

f in µg

/ml

AbsSteps & Gradient ElutionrhTf

Figure 4.8: Step and gradient elution of rhTf from HIC column. Sample was loaded

at 15µg rhTf /ml of gel. Unbound protein was washed out when the column was

equilibrated with 100% buffer A. Elution started with 3 CV of 50% buffer A and

followed by gradient elution using 50% to 25% buffer A in 10 CV. rhTf was pooled

from fraction 80 to fraction 111 and the recovery was 34%. Lane of SDS-PAGE, (a)

is sample loaded to column; (b) is sample recovered from fraction 80-111; (c) is

impurity eluted at 0% buffer A; m is marker, s is standard hTf.

a s b c m

hTf

86

0% buffer A or water (buffer B) eluted strong bound proteins which contained a

small amount of rhTf, respectively. Elution of a broad rhTf peak started when the

buffer contained less than 35% buffer A and stopped before 25% buffer A was

applied. rhTf pooled from fractions 80 to fraction 111 gave 34% recovery and did

not give satisfactory selectivity. Although fractions eluted using 0% buffer A

slightly contained rhTf, they were neglected because of the significant impurities

content (Figure 4.8).

4.2.2 Optimization of Hydrophobic Interaction Chromatography

HIC was used as the capture step, where recovery of target proteins is more

important compared to resolution. Type of gel, type of buffer and pH of buffer

which affects resolution was maintained as in 4.2.1 (Ali et al., 1996; Ailor et al.,

2000). Flowrates, loading capacity and elution method were optimized in order to

improve capacity, recovery, resolution and ease of use.

4.2.2.1 Optimization of Elution Method

Adsorption coefficient, α value of protein, in HIC changes only slowly with

changes in buffer condition. Protein-protein interaction brings similar proteins to

interact with each other and the adsorbent. Sharp separation is usually not achieved

in HIC (Scopes, 1994). Since gradient elution could not give satisfactory resolution

(Figure 4.8), step-wise elution which is technically simpler, reproducible and also

able to elute interest protein in a more concentrated form was applied in this study.

A few approaches using step-wise elutions had been done in order to increase the

resolution of the area where the peak of interest eluted without affecting the recovery

87

of rhTf. Strength of elution buffer was optimized to elute all less strongly bound

compounds, but must not exceed the level where peak of interest start to co-elute. As

mentioned in section 4.2.1, rhTf was eluted in between 35% to 25% of buffer A,

which mean 1st elution step can be optimized using buffer containing 50%-35%

buffer A. The second step elution was fixed at 25% buffer A because 25% buffer A

was expected to give complete elution of rhTf with minimum unwanted compound

(Figure 4.8).

HIC using 3 different step wise elutions involving 50% buffer A, 45% buffer

A and 35% buffer A was studied as the 1st elution buffer at a fixed flowrate of

0.5ml/min. 32±2µg rhTf per bed volume was loaded. Recovery of rhTf was the

highest when 50% buffer A was used as 1st elution buffer, 25% buffer A as 2nd

elution buffer (Table 4.2). Chromatograms and SDS-PAGE in Figure 4.9

characterizes the elution profile of the 3 different step wise elution methods.

Although 1st step elution with lower percentage of buffer A increased the elution of

unwanted compound (Figure 4.9) but the recovery of rhTf was low (Table 4.2).

Eluted rhTf did not show significant differences in resolution for the various step

elutions applied (Figure 4.9). Hence, step wise elution using 50% buffer A as 1st

elution buffer and 25% buffer A as 2nd elution buffer which gave 64% recovery was

chosen as the best elution method.

Table 4.2: Optimization of step-wise elution method for achieving higher recovery

of rhTf.

No Step elution Flowrates

(ml/min)

Recovery of

rhTf (%)

A 50% Buffer A and 25% Buffer A, (50/25) 0.5 64

B 45% Buffer A and 25% Buffer A, (45/25) 0.5 42

C 35% Buffer A and 25% Buffer A, (35/25) 0.5 29

Note: Different step elutions were optimized at a fixed flowrate, 0.5ml/min. Loaded

rhTf per bed volume was fixed at 32±2µg/ml. Recovery of rhTf was percentage of

pooled rhTf over total loaded rhTf.

88

0

0.1

0.2

0.3

0.4

0.5

0 20 40 60 80Fraction

Abs

at U

V 2

80nm

0

20

40

60

80

100

120

Step

s(Pe

rcen

tage

of B

uffe

r A(%

) &

Conc

netra

tion

of rh

Tf in

µg/

ml

AbsSteps ElutionrhTf

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 20 40 60 80 100 120 140Fraction

Abs

at U

V 2

80nm

0

20

40

60

80

100

120Pe

rcen

tage

of B

uffe

r A(%

) &

Conc

entra

tion

of r-

hTf i

n µg

/ml

AbsSteps ElutionrhTf

0

0.1

0.2

0.3

0.4

0.5

0.6

0 20 40 60 80 100 120Fraction

Abs

at U

V 2

80nm

0

20

40

60

80

100

120

Perc

enta

ge o

f Buf

fer A

(%)

&

Conc

entra

tion

of r-

hTf i

n µg

/ml

AbsSteps ElutionrhTf

*Volume for the first 24 fractions collected during equilibration was double compared to other

fractions

Figure 4.9: HIC chromatograms for the optimization of elution method. (a) 50/25

elution profile, (b) 45/25 elution profile and (c) 35/25 elution profile. Lane of SDS-

PAGE, (a) to (c) characterize the sample, which was pooled from experiment (a) to

(c); m is marker, s is standard hTf.

m a s b c *a

b c

89

4.2.2.2 Optimization of Elution Flowrate

Flow rate and sample load are interrelated. Flow rate and sample load are

optimized to find highest productivity where resolution is still high enough to meet

the predefined purity requirement. In this study, flowrate had not effect upon the

purity because step wise elution was applied. For the optimization of elution

flowrate, an average loading capacity of rhTf of 33±5µg/ml of gel and the optimized

step elution method as reported in section 4.2.2.1 were used. The prime

consideration when optimizing for highest possible productivity is to find the highest

possible sample load over the shortest possible sample application time with

acceptable loss in yield.

The elution flowrate which gave the highest recovery was 1ml/min and

followed by 0.5ml/min and 2ml/min (Table 4.3). Figure 4.10 shows the elution

profile of rhTf at different elution flowrates. High elution flowrate will always give

a decrease in dynamic binding capacity (Scopes, 1994) which will affect elution

profile and recovery of step elution. This may be the reason for the low recovery of

rhTf obtained at the elution flowrate of 2ml/min. Components of samples would

tend to ‘saturate’ non-specific binding sites of a column (Liljedah, 2000). A very

low flow rate will increase the sample holding time and thus increases the

opportunity of the nonspecific binding and result in loss of recovery. This may be

the reason the elution flowrate at 0.5ml/min showed lower recovery of rhTf than

1ml/min. In this study, optimized elution flowrate was 1ml/min which gave 74.6%

recovery.

90

Table 4.3: Optimization of elution flowrate.

No Flowrates Step Elution Recovery of

rhTf (%)

A 0.5ml/min 50/25 64

B 1ml/min 50/25 74.6

C 2ml/min 50/25 62.1

Note: Elution flowrate at 0.5ml/min, 1ml/min and 2 ml/min were studied at fix steps

elution. Loaded rhTf per bed volume was fixed at 33±5µg/ml. Recovery of rhTf

was percentage of pooled rhTf over total loaded rhTf.

91

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 20 40 60 80Fraction

Abs

at U

V 2

80nm

0

10

20

30

40

50

60

70

80

90

100

Perc

enta

ge o

f Buf

fer A

(%) &

Co

ncen

tratio

n of

rhTf

in µ

g/m

)

AbsSteps ElutionrhTf

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 20 40 60Fraction

Abs

at U

V 2

80nm

0

10

20

30

40

50

60

70

80

90

100

Perc

enta

ge o

f Buf

fer A

(%) &

Co

ncen

tratio

n of

rhTf

in µ

g/m

l)

AbsSteps ElutionrhTf

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 20 40 60Fraction

Abs

at U

V 2

80nm

0

10

20

30

40

50

60

70

80

90

100

110

Perc

enta

ge o

f Buf

fer A

(%) &

Co

ncen

tratio

n of

rhTf

in m

g/m

l

AbsSteps ElutionrhT f

All the fraction is 5ml/fraction except fraction 25-63 of figure 4.10a are 2.5ml/fraction.

Figure 4.10: HIC chromatograms for the optimization of elution flowrate. Elution

profiles at (a) 0.5ml/min flowrate, (b) 1ml/min flowrate and (c) 2ml/min flowrate.

The peaks characterizing the elution of rhTf at 25% Buffer A were pooled.

a

c

b

92

4.2.2.3 Optimization of rhTf Loading Capacity

After fixing the elution mode and the elution flowrates, rhTf loading capacity

was also optimized to get the optimum level that gives highest recovery. 15, 38, 55

and 74µg rhTf/ml of gel were loaded respectively. The optimized elution method and

flow rates were as mentioned in section 4.2.2.1 & section 4.2.2.2 respectively. The

loading capacity of rhTf at optimized flowrates and steps elution which gave best

recovery was 55µg rhTf/ml gel (Table 4.4, Figure 4.11). The relationship between

loading capacity and recovery percentage in figure 4.11 shows that loading of rhTf

between 38-58µg/ml gel was expected to result in the recovery of 75% and above.

Loading of 38-55µg rhTf/ml gel was selected as the optimal range of loading

capacity. Figures 4.12 and 4.13 show the elution profile of rhTf.

As mentioned in 4.2.2.2, we lost some amount of sample due to the

nonspecific binding of each matrix. The percentage of loss is more significant when

the loading is small. This is the reason why the lowest loading of rhTf gave lower

recovery (Figure 4.11) although the loading also not exceed the dynamic binding

capacity. Loading capacity could also affect the elution profile of rhTf. Binding

strength of rhTf became weaker and it was eluted earlier when using higher

percentage of buffer A. As shown in Figure 4.12d, rhTf was eluted when 50% buffer

was applied. Hence, the recovery of rhTf was decreased even though all the rhTf

was bound to the gel during equilibration stage (Figure 4.12d). Figure 4.13 shows

that the resolution was not significantly affected by the loading capacity. Higher

loading increased both the concentration of rhTf and impurities in collected fraction.

Anyhow, this conclusion is not applicable to the case of overloading since the

binding strength and the elution profile were affected. Figure 4.13d, shows more

unwanted impurity compare to the others. It was predicted that early elution of

unwanted protein, which supposed to be eluted at 0% buffer A, was due to high

loading.

93

Table 4.4: Optimization of rhTf loading capacity.

No Flowrates Step Elution Total rhTf/ml of gel Recovery

A 1ml/min 50/25 15 µg/ml 67.5%

B 1ml/min 50/25 38 µg/ml 74.6%

C 1ml/min 50/25 55 µg/ml 79.7%

D 1ml/min 50/25 74 µg/ml 40%

Note: Loading capacity of rhTf was studied at optimized flowrates and steps elution.

Recovery of rhTf was percentage of pooled rhTf over total loaded rhTf.

30

40

50

60

70

80

90

10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0

Loaded rhTf per bed volume (µg/ml)

Rec

over

y pe

rcen

tage

(%

)

Figure 4.11: The relationship between recovery percentage and loading capacity.

94

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 20 40Fraction

Abs

at U

V 2

80nm

0

10

20

30

40

50

60

70

80

90

100

110

Perc

enta

ge o

f Buf

fer A

(%) &

Co

ncen

tratio

n of

rhTf

in µ

g/m

l)

AbsSteps ElutionrhTf

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 20 40 60Fraction

Abs

at U

V 2

80nm

0

10

20

30

40

50

60

70

80

90

100

Perc

enta

ge o

f Buf

fer A

(%) &

Co

ncen

tratio

n of

rhTf

in µ

g/m

l)

AbsSteps ElutionrhTf

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 20 40 60Fraction

Abs

at U

V 2

80nm

0

10

20

30

40

50

60

70

80

90

100

110

Perc

enta

ge o

f Buf

fer A

(%) &

Co

ncen

tratio

n of

rhTf

in µ

g/m

l)

AbsSteps ElutionrhTf

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 20 40 60 80Fraction

Abs

at U

V 2

80nm

0

10

20

30

40

50

60

70

80

90

100

Perc

enta

ge o

f Buf

fer A

(%) &

Co

ncen

tratio

n of

rhTf

in µ

g/m

l)

AbsSteps ElutionrhTf

Figure 4.12: HIC chromatograms for the optimization of rhTf loading capacity.

Flowrate was fixed at 1ml/min and the step elution method was fixed at 50/25.

Loading capacity at (a) 15µg rhTf/ml of gel, (b) 38µg rhTf/ml of gel, (c) 55µg

rhTf/ml of gel and (d) 74µg rhTf/ml of gel. The peak characterizing the elution of

rhTf at 25% buffer A was pooled.

a b

c d

95

(a) (b)

(c) (d)

Figure 4.13: SDS-PAGE characterizing the elution profile of rhTf. (a), (b), (c) & (d)

show the eluted fractions at 25% buffer A of chromatograms (a) to (d) of Figure 4.12;

m is protein marker.

4.2.3 Initial Screening Step of IEX using Batch Purification in Reduced

Volume

Batch purification was carried out using appendorf tube to screen the rhTf

binding capacity when using different anion exchange matrices, different types of

equilibration buffer, different pH of the equilibration buffer and different

concentration of equilibration buffer. Binding capacity was calculated by

centrifuging the appendorf tube that had been loaded with rhTf and subtracting the

m m

m m

rhTfrhTf

rhTf rhTf

96

amount of rhTf in the supernatant from the total rhTf loading. This step was taken as

a preliminary study of the ion exchange column to gain insight into the relationship

between buffer, matrix and binding capacity. Tris buffer gave higher binding

strength to anion exchange matrix compared to phosphate buffer (Figure 4.14).

Binding Strength of Q-Sepharose was stronger than DEAE Sephadex-25. Binding

capacity of Q-Sepharose was optimun when pH was increased to at least 8.5 and

concentration of Tris buffer was less than 30mM (Figure 4.15, Figure 4.16). The

choice of pH for anion exchanger should be the lowest pH that gives high binding

capacity. Ionic strength of the counter ion should not be too low to avoid pH

fluctuation and dilution of sample. Q-Sepharose with 20mM Tris/HCl buffer, pH8.5

which gave the highest binding capacity (figure 4.14, 4.15, 4.16) was identified as a

suitable parameter for the second stage separation.

0.0

0.5

1.0

1.5

2.0

2.5

3.0

DEAE Sephadex A-25 Q-Sepharose

Type of Buffer

Bind

ing

Capa

city

of A

nion

Exc

hang

e M

atrix

/( µg

hTf/m

l) 20mM Tris-HCl buffer, pH7.5

20mM Sodium Phosphate

Figure 4.14: Binding capacity of two anion exchange matrix with Tris and

phosphate buffer used as equilibration buffer. Matrices in the study were DEAE

Sephadex A-25 and Q-Sepharose; buffers were 20mM Tris-HCl buffer, pH7.5

and 20mM sodium phosphate buffer, pH7.5. Results are means ±SE for 2

replicates.

97

2.40

2.60

2.80

3.00

3.20

3.40

3.60

6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0pH Buffer

Bind

ing

Capa

city

( µg

hTf/m

l of m

atrix

)

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

Conc

entra

tion

of h

Tf in

Sup

erna

tant

(µg

/ml)

Binding capacity/(ug ofbound hTf/ml of matrix)Concentration of hTf inSupernatant/(ug/ml)

Figure 4.15: Binding capacity of Q-Sepharose with equilibration buffer of

different pH. Optimization buffer was carried out to achieve higher recovery of

rhTf. 20mM Tris-HCl buffer was used as the equilibration buffer and Q-

Sepharose as the matrix. Results are means ±SE for 2 replicates.

2.60

2.70

2.80

2.90

3.00

3.10

3.20

0 20 40 60 80 100Concentration of Buffer/mM

Bind

ing

Capa

city

( µg

hTf/m

l of m

atrix

)

-0.05

0.00

0.05

0.10

0.15

0.20

Conc

entra

tion

of h

Tf in

Sup

erna

tant

(µg

/ml)

Binding capacity/(ug ofbound hTf/ml of matrix)

Concentration of hTf inSupernatant/(ug/ml)

Figure 4.16: Binding capacity of Q-Sepharose with different concentration of

equilibration buffer. Buffer Tris-HCl buffer, pH8.5 was used as the equilibration

buffer and Q- sepharose as the binding matrix. Results are means ±SE for 2

replicates.

98

4.2.4 Anion Exchange Chromatography

According to Ali et al. (1996), anion exchange chromatography with Q

sepharose was used as the polishing step in rhTf purification in which 20mM

Tris/HCl buffer, pH8.0 was used as the equilibration buffer and a gradient elution of

0-100% KCl was employed. 95% purity of rhTf was obtained. In this work, the

relationship between matrix and buffer with rhTf binding capacity was obtained

(Section 4.2.3). Q-sepharose matrix and 20mM Tris buffer, pH8.5 would be used for

subsequent anion exchange chromatography. pH 8.5 which are slightly different

from the pH mentioned in Ali et al. (1996) but gave good binding capacity was

considered. The purity of the rhTf was the prime concern as anion exchange

chromatography was taken as the final polishing step. To improve selectivity, the

focus was on the gradient elution profile.

4.2.4.1 Maximizing the Selectivity of Anion Exchange Chromatography

Flowrates, pH and concentration of elution buffer would affect the resolution

of chromatography. In maximizing the selectivity of anion exchange column,

gradient elution was optimized by varying the shallowness of the gradient. Flowrate

of elution was not optimized but was fixed at a very slow, 0.5ml/min to maximize the

selectivity. Step elution was combined to simplify the elution where 20mM Tris-HCl,

pH8.5 was used to wash out unbound protein during equilibration and 100% 0.5M

NaCl/20mM Tris-HCl, pH8.5 (buffer B) was used to remove strongly bound protein.

rhTf was predicted to be eluted at around 15% of buffer B, so gradient elution was

started from 10% buffer B to 20% or 30% of Buffer B (Figure 4.17).

Three experiments using different gradient elutions were carried out with

different percentage of buffer B per CV at 3% per CV, 2% per CV and 1% per CV.

99

Figure 4.17 and 4.18 show that decreasing of the gradient slope from 3% buffer

B/CV (10%-30% Buffer B in 65 ml) (a) to 2% buffer B/CV (10-20% Buffer B in

53ml) (b) did not improved selectivity. Decreasing the slope of the gradient elution

to 1% buffer B/CV (10%-20% buffer B in 147 ml) (c) gave 100% pure rhTf. This

showed that by increasing the total gradient volume (decreasing gradient slope) of a

linear gradient, resolution could be improved (Figure 4.17).

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0 40 80 120 160 200 240 280 320time/minutes

Abs

at U

V 2

80nm

0

5

10

15

20

25

30

35

40

45

50

Perc

enta

ge o

f Buf

fer B

/% &

C

once

ntra

tion

of rh

Tf in

µg/m

l

AbsSteps & Gradient ElutionrhTf

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0 40 80 120 160 200 240 280time/minutes

Abs

at U

V 2

80nm

0

5

10

15

20

25

30

35

40

45

50

Perc

enta

ge o

f Buf

fer B

/% &

C

once

ntra

tion

of rh

Tf in

µg/m

l

AbsSteps & Gradient ElutionrhTf

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.1

0 50 100 150 200 250 300 350 400 450 500 550 600 650 700Time/Minutes

Abs

at 2

80nm

and

% o

f Buf

fer B

0

5

10

15

20

25

30

35

40

45

50Pe

rcen

tage

of B

uffe

r B/%

&

Con

cent

ratio

n of

rhTf

in µg

/ml

AbsSteps & Gradient ElutionrhTf

Figure 4.17: Anion exchange chromatograms for the optimization of selectivity.

Each chromatogram, (a) to (c), shows the elution profile of rhTf, with different

gradient elution. Gradient elutions of (a) to (c) were elutions with increasing of

buffer B, 3%per CV, 2%perCV and 1%perCV. Column volumes of (a) and (b) were

fixed at 10.6ml; column volume (c) was 16.8ml.

a b

c

100

(a) (b)

(c)

Figure 4.18: SDS-PAGE characterizing the elution profile of rhTf. (a), (b) & (c)

show the eluted fractions consisted of rhTf as shown in chromatogram (a), (b) & (c)

of Figure 4.17, during gradient elution; (m) is protein marker. (a) is fractions eluted

in between 156 to 192 minutes, (b) is fractions eluted in between 136-172 minutes

and (c) is fractions eluted in between 310 to 390 minutes.

4.2.5 Characterization of rhTf Purification

After optimization, a compete sequence of rhTf purification was carried out

using the optimized parameters. Crude sample with 0.5% yield of rhTf, was

harvested at day 6 post infection. Sample was loaded to column Phenyl Sepharose 6

fast flow at 38µg rhTf/ml of gel and at a flowate of 1ml/min. The column was

equilibrated with 100 buffer A. The sample was eluted with a step wise sequence

profile of 50% buffer A, 25% buffer A and water. 74.56% of rhTf was recovered.

After dialysis, the sample was loaded to Q-Sepharose fast flow. 20mM Tris-HCl, pH

m m

m

rhTfrhTf

rhTf

101

8.5 was used as the equilibration buffer. Elution flowrate was fixed at 0.5ml/min.

Gradient elution was initiated by increasing 50mM NaCl to100mM NaCl in 5

column volumes. 100% pure rhTf with 34% overall recovery was achieved (Table

4.5). Due to the difference in the characteristics of the tools in analyzing rhTf and

total protein, the purity percentage was actually slightly more than 100%, but was

taken to be 100%. Figure 4.19 and Figure 4.20 characterize the elution profile of

rhTf from Phenyl Sepharose column. Figures 4.21 and 4.22 characterize the elution

profile of rhTf from Q-Sepharose. Figure 4.23 qualifies the purity of the sample

from crude to final purification step. The total purification fold is 200.

Table 4.5: Summary of the characteristic of purification of rhTf.

Parameter Sample Phenyl

SepharoseDialysis Q Sepharose

Volume (ml) 23.00 59.00 79.50 36.20

rhTf Concentration (µg/ml) 31.00 9.01 5.81 6.69

Protein Concentration (µg/ml) 6200.00 - 24.47 6.62

Total hTf (µg) 713.00 531.63 461.50 242.08

Total Protein (µg) 142600.00 - 1945.66 239.70

Purity (%): 0.5 23.72 100.00

Recovery (%) 74.56 86.81 52.46

Overall Recovery (%) 74.56 64.73 33.95

102

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 20 40 60Fraction

Abs

at U

V 2

80nm

0

10

20

30

40

50

60

70

80

90

100

Perc

enta

ge o

f Buf

fer A

(%) &

Co

ncen

tratio

n of

rhTf

in µ

g/m

l)

AbsSteps ElutionrhTf

Figure 4.19: HIC chromatogram characterizing the separation and elution profile of

sample. Chromatography was carried out using optimized flowrate, step elution and

loading.

m. marker

a. Fraction-40

b. Fraction-42

c. Fraction-43

d. Fraction-45

e. Fraction-47

f. Fraction-49

g. Fraction-51

h. Fraction-53

i. Fraction-55

Figure 4.20: SDS-PAGE characterizing the separated protein from phenyl sepharose

6 fast flow column. Protein in fractions 40, 42, 43, 45, 47, 49, 51, 53, 55 were shown

in 9%, silver staining, SDS-PAGE. m is molecular weight standards.

a b c m d e f g h i

rhTf

103

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.1

0 20 40 60 80 100 120 140Fraction

Abs

at U

V 2

80nm

0

10

20

30

40

50

60

Perc

enta

ge o

f Buf

fer B

/% &

Con

cent

ratio

n of

rhTf

in µ

g/m

l

AbsSteps & Gradient ElutionrhTf

Figure 4.21: Anion exchange chromatogram characterizing the separation and elution

profile of sample of after HIC and after dialysis. Q-Sepharose Chromatography was

carried out with 20mM Tris HCl, pH 8.5 as the equilibration buffer.

m- marker

a. Fraction-69

b. Fraction-71

c. Fraction-73

d. Fraction-75

e. Fraction-77

f. Fraction-79

g. Fraction-81

h. Fraction-83

i. Fraction-85

Figure 4.22: SDS-PAGE characterizing the separated protein from Q-Sepharose

column. Protein in fractions 69, 71, 73, 75, 77, 79, 81, 83, 85 were shown in 9%,

silver staining, SDS-PAGE. m is molecular weight standards.

a m b c d e f g h i

rhTf

104

Figure 4.23: SDS-PAGE characterizing the sample pooled from each purification

step. m is molecular weight standards, (a) is supernatant sample harvest at day 6 post

infection, (b) is sample after hydrophobic interaction chromatography and after

dialysis and (c) is pure rhTf after anion exchange chromatography.

4.3 Characterization of The Carbohydrate Composition of rhTf

Characterization of the carbohydrate composition of purified rhTf had been

carried out. 36.5µg standard apo-transferrin (76kDa) and purified rhTf (73kDa) were

hydrolyzed, labeled with PMP and analysed using HPLC. Separation profile of the

standard human transferrin and rhTf from insect cells are shown in Figure 4.24.

Mannose was eluted at t= 20.5±0.5min; Glucosamine was eluted at t= 28.5±0.5min;

Glucose was eluted at t= 51.5±0.5min; Galactose was eluted at t= 53.0±0.5min. The

carbohydrate content was quantified using the standard calibration curve of each

monosaccharide (Figure 4.25). During acid hydrolysis, N-acetylglucosamine would

loss the acetyl group to yield glucosamine. So, glucosamine was used as standard to

quantify N-acetylglucosamine.

m a b c

rhTf

225kDa150kDa 100kDa75kDa 50kDa 35kDa 25kDa

105

Results of the standard apo-transferrin analysis were compared with reported

values. The ratios of Man:GlcNAc:Gal of the apo-transferrin were 3:4.9:2. These

values were similar to the ratio reported by Spik, 1985 and Fu, 1995 (Table 4.6). For

the expressed rhTf, molar fractions of Man, GlcNAc and Gal over rhTf were 3.78,

1.69 and 0.93; the determined ratios of Man:GlcNAc:Gal were 3:1.34:0.74 (Table

4.6). These values were similar to the ratios reported by Salmon (1997) (Table 4.6).

Glucose was detected in both standard apo transferrin and the purified rhTf

(Figure 4.24). Glucose attached structure which was found early in oligosaccharide

processing pathway, was identified by Ailor et al. (2000). Alonzi et al. (2007)

mentioned that inhibition of endoplasmic reticulum (ER) alpha-glucosidases I and II

by imino sugars caused the retention of glucose residues on N-linked

oligosaccharides. Re-glucosylation by insect glucosyltransferase had been

recognized (Parker et al., 1995).

According to van Die (1996), galactosyltransferase activity was detected in

Spodoptera frugiperda. UDP-GlcNAc, UDP-Gal, UDP-Glc also existed in cultured

Sf9 (Tomiya, 2001). However, galactosylated or sialylated complex-type glycans are

rarely found in glycoprotein produced by insect cells (Jarvis and Summers, 1989;

Wathen et al., 1991; Grabenhorst et al., 1993; Yeh et al., 1996; Ogonah et al., 1996).

Despite above, minor galactose over rhTf as reported by Salmon et al., 1997 was

found in this study.

Total weight percent of Man, GlcNAc and Gal over standard human

transferrin in this study was 3.5%, which was similar to 3.9% reported by Zdebska

and Kościelak (1999). The total weight percent of Man, GlcNAc and Gal was only

1.61 % for the transferrin expressed from Sf9, which was about half of standard

human transferrin. This result was similar to that obtained by Salmon et al. (1997)

which was 1.88%. The difference with the native protein is related to the nature of

the expression system. Glycoproteins expressed in insect cells often have incomplete

glycan structure compared to those expressed in mammalian cells (Kuroda, 1990).

106

There are also other factors that can alter the glycosylation pathway. Castro et al.

(1995) found that supplementation of lipid affected the glycosylation pattern.

Gawlitzek et al. (1995), Hayte et al. (1992) and Hayte et al. (1993), also found that

glycosylation was dependent on media components. The optimized culture medium

used in this work which consisted of high glucose, glutamine and lipid content was

believed to affect the glycosylation pattern.

a)

b)

Figure 4.24: Chromatograms show HPLC separation of PMP-labeled transferrin. (a)

36.5µg of standard apo-transferrin and (b) purified sample rhTf which were

hydrolyzed, labeled and analyzed by HPLC as described under material and methods.

Mannose eluted at t= 20.5±0.5min; Glucosamine eluted at t= 28.5±0.5min; Glucose

eluted at t= 51.5±0.5min; Galactose eluted at t= 53.0±0.5min.

Man

nose

Glu

cosa

min

e

Glu

cose

Gal

acto

se

0

0

55

55

Retention time (min)

Retention time (min)

Man

nose

Glu

cosa

min

e

Glu

cose

Gal

acto

se

107

Standard for Monosaccharide Analysis using HPLC

R2 = 0.9916

R2 = 0.9933

R2 = 0.9907R2 = 0.9973

0.00E+00

2.00E+05

4.00E+05

6.00E+05

8.00E+05

1.00E+06

1.20E+06

1.40E+06

1.60E+06

1.80E+06

0.00 20.00 40.00 60.00 80.00 100.00 120.00

Concentration, µM

Are

a

Mannose Glucose Galactose Glucosamine

Figure 4.25: Standard calibration graph of monosaccharides. Hydrolyzed and

labeled monosaccharides standards in different concentration were analyzed using

HPLC. The graph shows the linearity of UV response for PMP labeled Glucosamine,

Mannose, Galactose and Glucose.

Table 4.6: Carbohydrate Composition Analysis of Glycoprotein.

Sample Man GlcNAc Gal

Literature: Human Serotransferrin (Fu, 1995) Determined ratio Human Serotransferrin (Spik, 1985) Determined ratio Recombinant Human Lactoferrin from Sf9 (Salmon, 1997) Determined ratio

3 3 3

4.44 4.7 1.5

1.88 2.4 0.5

Apotransferrin (Sigma T4382) Amount, mol/ mol protein Determined ratio

4.06 3

6.68 4.9

2.73 2

Sample, rhTf from Sf9 Amount, mol/ mol protein Determined ratio

3.78 3

1.69 1.34

0.93 0.74

CHAPTER 5

CONCLUSIONS

5.1 Conclusions

Using the optimized expression method reported by Ongkudon (2006),

31.1µg/ml of rhTf had been produced from Sf9. Expression of transferrin was

initiated by infecting Sf9 suspension culture in optimized medium at the seeding

density of 1.6x106 cells/ml with 15 MOI recombinant transferrin baculovirus at day 2

post culture. Maximum cell density, 2.99x106 cells/ml was achieved at day 2 post

infection. rhTf in 6200µg/ml of protein was harvested at day 6 post infection. Cells

growth profile, nutrient consumption and rhTf expression after infection were studied

to give more insight into the behavior of this insect cells-BEVS culture which were

done in suspension and using optimized medium. Concentration of glucose,

glutamine and lactate in harvested supernatant were 7.41g/L, 15.1mM and 0.45g/L

respectively. Concentration of ammonia remaining in the culture was lower than

2mM.

Purification was performed after rhTf expression. Few purification strategies

were studied to improve the purification yield and recovery. As reported in Ali et al.

(1996) and Ailor et al. (2000), phenyl sepharose and Q-sepharose chromatography

109

were used as the main tools in this separation. A screening step involving the

combination of step and gradient elution of transferrin from phenyl sepharose

provides a better understanding of the elution profile of the sample from HIC

column. For the column optimization, step wise elution which was technically

simpler, reproducible and also able to elute interest protein in a more concentrated

forms was used instead of gradient elution. For the optimization of elution method,

step elution at 50% buffer A and 25% buffer A gave rhTf in good recovery within a

predefined resolution.

The prime consideration when optimizing for highest possible productivity is

to find the highest possible sample load over the shortest possible sample application

time with acceptable loss in yield. Flow rates and loading capacity were optimized

in order to improve capacity, recovery and ease of use. The optimized flow rate was

1ml/min. Maximum loading capacity of rhTf at the optimal elution method and flow

rate was 55µg/ml. The selected range of loading of rhTf from 38-58µg/ml gel was

expected to result in the recovery of 75% and above.

Preliminary study of the second column, the ion exchange column, gave

better knowledge about the relationship between buffer, matrix and the binding

capacity. Batch purification method in reduced volume was applied to screen the

rhTf binding capacity of different anion exchange matrices, different types of

equilibration buffer, different concentrations of equilibration buffer and different pH

of the equilibration buffer. Q-Sepharose with 20mM Tris/HCl buffer, pH8.5 gave

the highest binding capacity. These parameters were applied for rhTf purification in

the second column. In maximizing the selectivity of the anion exchange column,

gradient elution was optimized by varying the slope of the gradient. Gradient elution

affects selectivity significantly. Gradient elution with 1% buffer B/CV (10%-20%

buffer B in 147 ml) succeeded in giving pure rhTf.

Crude sample with 0.5% yield of rhTf was loaded to Phenyl Sepharose 6 fast

flow column at 38µg rhTf/ml of gel using the optimized parameters gave 74.56%

110

recovery. The partially purified sample was loaded to Q-Sepharose fast flow column

at 0.5ml/min and pure rhTf was obtained. 34% overall recovery with 200

purification fold were achieved.

Lastly, a brief glycan characterization of the recovered pure rhTf was done

for a better understanding of the glycosylation feature of the expressed protein. The

carbohydrate component of the purified rhTf was determined. The ratios of

Man:GlcNAc:Gal for rhTf were 3:1.34:0.74. Minor galactose was found in this

study. Glucose attached structure was also obtained. Re-glucosylation had probably

occurred. The total weight percent of Man, GlcNAc and Gal was only half of the

standard apo-transferrin. Applied optimized culture condition had probably affected

the glycosylation pathway.

5.2 Recommendations

In this research, the whole process train for the production of rhTf, which

include expression, purification and basic characterization of the carbohydrate

content had been completed. There are a lot more improvement and interesting study

that should be performed in the future. A few recommendations are given to

improve the production, purification and the glycosylation of the system.

(a) Detail glycan characterization using nuclear magnetic resonance (NMR) or

HPLC mass spectrometry (HPLC-MS) to have a better understanding of the

glycan alteration.

(b) Detail study about the relationship between culture condition and

glycosylation in order to get optimized expression with improved

glycosylation.

(c) Improvement of glycosylation using in-vitro galactosylation and in-vitro

sialylation.

111

(d) Apply statistical analysis instrument such as Response Surface Method

(RSM) in the optimization of chromatographic process.

(e) Scaling up of the expression of rhTf to a large scale bioreactor for

commercial purpose. Study the alteration of glycosylation pathway after

scaling up and create a model to adjust the optimized culture condition and

minimize the deviation.

(f) Scaling up of the purification of rhTf to pilot scale.

REFERENCES

Abeijon, C. and Hirschberg, C. (1992) Topography of glycosylation reactions in the

endoplasmic reticulum. Trends in Biochemical Sciences. 17, 33-36.

Ailor, E., Betenbaugh, M.J. (1999). Modifying Secretion and Post-Translational

Processing in Insect Cells. Current Opinion in Biotechnology. 10, 142–145.

Ailor, E., Takahashi, N., Tsukamoto, Y., Masuda, K., Rahman, B.A., Jarvis, D.L.,

Lee, Y.C. and Betenbaugh, M.J. (2000). N-glycan patterns of human

transferrin produced in Trichoplusia ni insect cells: effects of mammalian

galactosyltransferase. Glycobiology. 10, 837-847.

Aisen, P., and Listowsky, I. (1980). Iron transport and storage proteins. Annual

Reviews of Biochemistry 49, 357-93.

Ali, S.A., Joao, H.C., Csonga, R., Hammerschmid, F. and Steinkasserer, A. (1996).

High-yield production of functionally active human serum transferrin using a

baculovirus expression system, and its structural characterization.

Biochemical Journal. 319, 191-195.

Alonzi, D., Neville, D., Lachmann, R., Dwek, R. and Butters, T. (2007).

Glucosylated Free Oligosaccharides Are Biomarkers of Endoplasmic

Reticulum Alpha-glucosidase Inhibition. Biochemical Journal. 409, 571-580.

113

Altmann, F., Staudacher, E., Wilson, I.B.H. and März, L. (1999). Insect Cells as

Hosts for The Expression of Recombinant Glycoproteins. Glycoconjugate

Journal. 16, 109–123.

Amersham Biosciences (2000b). Ion Exchange Chromatography: Principles and

Methods. (Edition AA). Amersham Biosciences.

Amersham Biosciences. (2000). Hydrophobic Interaction Chromatography:

Principles and Methods. (Edition AB). Amersham Biosciences.

An Ke, Gillock, E.T., Sweat, J.A., Reeves, W.M. and Consigli, R.A. (1999). Use of

The Baculovirus System to Assemble Polyomavirus Capsid-like Particles

with Different Polyomavirus Structural Proteins: Analysis of The

Recombinant Assembled Capsid-like Particles. The Journal of General

Virology. 80, 1009–1016.

Aumiller, J.J., Hollister, J.R. and Jarvis, D.L. (2003). A Transgenic Insect Cell Line

Engineered to Produce CMP-sialic Acid and Sialylated Glycoproteins.

Glycobiology. 13, 497-507.

Ayres, M.D., Howard, S.C., Kuzio, J., Lopez-Ferber, M. and Possee, R.D. (1994).

The Complete DNA Sequence of Autographa Californica Nuclear

Polyhedrosis Virus. Virology. 202, 586-605.

Bailey, S., Evans, R.W., Garratt, R.C., Gorinsky, B., Hasnain, S., Horsburgh, C.,

Jhoti, H., Lindley, P.F., Mydin, A. and Sarra, R. (1988). Molecular Structure

of Serum Transferrin at 3.3-A Resolution. Biochemistry. 27, 5804-5812.

Baker, E. N. and Lindley, P. F. (1992). New Perspectives on The Structure and

Function of Transferrins. Journal of Inorganic Biochemistry. 47, 147-160.

Baldwin, D.A., Jenny, E.R. and Aisen, P. (1984). The Effect of Human Serum

Transferrin and Milk Lactoferrin on Hydroxyl Radical Formation from

114

Superoxide and Hydrogen Peroxide. The Journal of Biological Chemistry.

259, 13391-13394.

Bauer, H. C., Parent, J. B. and Olden, K. (1985). Role of Carbohydrate in

Glycoprotein Secretion by Human Hepatoma Cells. Biochemical and

Biophysical Research Communications. 128, 368-375.

Beare, S. and Steward, W.P. (1996). Plasma Free Iron and Chemotherapy Toxicity.

The Lancet. 347, 342–343.

Bedard, C., Tom, R. and Kamen, A. (1993). Growth, Nutrient Consumption and End-

Product ccumulation in Sf9 and BTI-EAA Insect Cell Cultures: Insights into

Growth Limitation and Metabolism. Biotechnology Progress. 9, 615-624.

Belew, M., Yafang, M., Bin, L., Berglöf, J. and Janson, J-C. (1991). Purification of

Recombinant Hepatitis B Surface Antigen Produced by Transformed Chinese

Hamster Ovary (CHO) Cell Line Grown in Culture. Bioseparation. 1, 397–

408.

Beljelarskaya, S.N. (2002). A Baculovirus Expression System for Insect Cells.

Molecular Biology. 36(3), 281–292.

Betenbaugh, M.J, Tomiya, N., Narang, S., Hsu, J.T.A. and Lee, Y.C. (2004).

Biosynthesis of Human-Type N-Glycans in Heterologous Systems. Current

Opinion in Structural Biology. 14,601–606.

Birch, J.R. and Onakunle, Y. (2005). Biopharmaceutical Proteins: Opportunities and

Challenges. In: Mark Smales, C. and James, D.C. Therapeutic Proteins:

Methods and Protocols. New Jersey: Humana Press.

Bonnerjera, J., Oh, S., Hoare, M. and Dunhill, P. (1986). The Right Step at The Right

Time. Bio/Technology. 4, 954-958.

115

Bradley, S.J., Gosriwatana, I., Srichairantankool, S., Hider, R.C. and Porter, J.B.

(1997). Non-transferrin-bound Iron Induced by Myeloablative Chemotherapy.

British Journal of Haematology. 99, 337–343.

Breitbach, K., Jarvis, D.L. (2001). Improved Glycosylation of A Foreign Protein by

Tn-5B1-4 Cells Engineered to Express Mammalian Glycosyltransferases.

Biotechnology And Bioengineering. 74(3), 230-239.

Bullen, J.J., Griffiths, E. (1999). Iron and Infection, Molecular, Physiological and

Clinical Aspects. (2nd Edition). Chichester: John Wiley and Sons.

Burgess, S. (1977). Molecular Weights of Lepidopteran Baculovirus DNAs:

Derivation by Electron Microscopy. The Journal of General Virology. 37,

501-510.

Carson, D.D. (1992) Proteoglycans in Development. In: Fukuda, M. (Ed.). Cell

Surface Carbohydrates and Cell Development. (pp. 257-274). London: CRC

Press.

Castro, P.M.L., Ison, A.P., Hayte, P.M. and Bull, A.T. (1995). The

Macroheterogeneity of Recombinant Human Interferon-γ Produced by

Chinese Hamster Ovary Cells is Affected by The Protein and Lipid Content

of The Culture Medium. Biotechnology and Applied Biochemistry. 21, 87-

100.

Chauhan, A., Chauhan, V., Brown, W.T. and Cohen, I. (2004). Oxidative Stress in

Autism: Increased Lipid Peroxidation and Reduced Serum Levels of

Ceruloplasmin and Transferrin – The Antioxidant Proteins. Life Sciences. 75,

2539–2549.

Chiou, T.W., Hsieh, Y.C. and Ho, C.S. (2000). High Density Culture of Insect Cells

using Rational Medium Design and Feeding Strategy. Bioprocess

Engineering. 22, 483-491.

116

Choi, O., Tomiya, N., Kim, J.H., Slavicek, J.M., Betenbaugh, M.J. and Lee, Y.C.

(2003). N-Glycan Structures of Human Transferrin Produced by Lymantria

dispar (Gypsy moth) Cells using The LdMNPV Expression System.

Glycobiology. 13(7), 539-548.

Choudhury, D., Thakurta, P.G., Dasgupta, R., Sen, U., Biswas, S., Chakrabarti, C.

and Dattagupta, J.K. (2002). Purification and Preliminary X-ray Studies on

Hen Serotransferrin in Apo- and Holo-Forms. Biochemical and Biophysical

Research Communications. 295,125–128.

Chu, L. and Robinson, D.K. (2001). Industrial Choices for Protein Production by

Large Scale Cell Culture. Current Opinion in Biotechnology. 12: 180-187

Comings, D.E., Miguel, A.G. and Lesser, H.H. (1979). Nuclear proteins. VI.

Fractionation of Chromosomal Non-Histone Proteins using Hydrophobic

Chromatography. Biochimica et Biophysica. Acta. 563, 253–260.

Cox, M.M.J (2004). Chapter 3: Commercial Production in Insect Cells-One

Company’s Perspective. Bioprocess International. 2(2), 34-38.

Crowther, J.R. (1995). Elisa: Theory and Practice. Totowa, New Jersey: Human

Press.

Cumming, D.A. (1992). Physiological Relevance of Protein Glycosylation.

Development in Biological. Standardization. 76, 83-94

Cummings, R.D., Merkle, R.K. and Stults, N.L. (1989). Separation and analysis of

glycoprotein oligosaccharides. Methods in Cell Biology. 32, 141-183.

Davies, A.H. (1994). Current Methods for Manipulating Baculoviruses.

Bio/Technology. 12, 47–50.

117

de Jong G., van Noort W. L., Feelders R. A., de Jeu-Jaspars C. M. H. and van Eijk H.

G. (1992). Adaptation of Transferrin Protein and Glycan Synthesis. Clinical

Chimica Acta. 212, 27-45.

de Jong, G., van Dijk, J.P. and van Eijk, H.G. (1990) The Biology of Transferrin.

Clinical Chimica Acta. 190, l-46.

de Smit, M.H., Hoefkens, P., de Jong, G., van Duin, J., van Knippenberg, P.H. and

van Eijk, H.G. (1995). Optimized Bacterial Production of Nonglycosylated

Human Transferrin and Its Half-Molecules. The International Journal of

Biochemistry and Cell Biology. 27, 839–850.

Dill, K.A. (1990a). Dominant Forces in Protein Folding. Biochemistry. 29, 7133–

7155.

Donaldson, M., Wood, H.A., Kulakosky, P.C., Shuler, M.L. (1999). Glycosylation of

A Recombinant Protein in The Tn5B1-4 Insect Cell Line: Influence of

Ammonia, Time of Harvest, Temperature, and Dissolved Oxygen.

Biotechnology and Bioengineering. 63, 255–262.

Elbehri, A. (2005). Biopharming and The Food System: Examining The Potential

Benefits and Risks. AgBioForum. 8(1), 18-25.

Fausnaugh, J.L., Kennedy, L.A. and Regnier, F.E. (1984). Comparison of

Hydrophobic Interaction and Reversed-Phase Chromatography of Proteins.

Journal of Chromatography. 317, 141–155.

Fernandez, M., Crawford, L. and Hefferan, C. (2002). Pharming the Field: A Look at

The Benefits and Risks of Bioengineering Plants to Produce Pharmaceuticals.

Proceedings from A Workshop Sponsored by The Pew Initiative on Food and

Biotechnology, FDA, and USDA. July. Washington D.C., 4-12.

118

Fu, D. and O’Neill, R.A. (1995). Monosaccharide Composition Analysis of

Oligosaccharides and Glycoprotein by High-Performance Liquid

Chromatography. Analytical Biochemistry. 227, 377-384.

Fu, D. and O’Neill, R.A. (1995). Monosaccharide Composition Analysis of

Oligosaccharides and Glycoprotein by High-Performance Liquid

Chromatography. Analytical Biochemistry. 227, 377-384.

Fu, D., van Halbeek H. (1992). N-glycosylation Site Mapping of Human

Serotransferrin by Serial Lectin Affinity Chromatography, Fast Atom

Bombardment-Mass Spectrometry, and 1H Nuclear Magnetic Resonance

Spectroscopy. Analytical Biochemistry. 206, 53-63

Garcia, F.A.P. and Pires, E.M.V. (1993). Chromatography. In: Kennedy, J.F. and

Cabral, J.M.S. (Eds.) Recovery Processes for Biological Materials. (pp. 415-

451). London: Wiley.

Gawlitzek, M., Condradt, H.S. and Wagner, R. (1995). Effect of Different Cell

Culture Conditions on The Polypeptide Integrity and N-Glycosylation of A

Recombinant Model Glycoprotein. Biotechnology and Bioengineering. 46,

536-544.

Gershoni, J.M. and Palade, G.E. (1982). Electrophoretic Transfer of Proteins from

Sodium Dodecyl Sulfate-Polyacrylamide Gels to a Positively Charged

Membrane Filter. Analytical Biochemistry. 124, 396-405.

Goosen, M.F.A. (1993). Insect Cell Culture Engineering: An Overview. In: Goosen,

M.F.A., Gaugulis, A.J. and Faulkner, P. (Eds.) Insect Cell Culture

Engineering. New York: Marcel Dekker.

Gorfien, S.F., Paul, W., Judd,D., Tescione, L. and Jayme, D.W. (2003). Optimized

Nutrient Additives for Fed-Batch Cultures. Biopharm. International. 34-40.

119

Grabenhorst, E., Hofer, B., Nimtz, M., Jager, V. and Conradt, H.S. (1993).

Biosynthesis and Secretion of Human Interleukin-2 Glycoprotein Variants

from Baculovirus Infected Sf21 Cells: Characterization of Polypeptides and

Post-Translational Modifications. European Journal of Biochemistry. 215,

189−197.

Granados, R.R. and Federici, B.A. (Eds.). (1986). The Biology of Baculoviruses.

(2nd volume). Boca Raton, Florida: CRC Press.

Gumerov, D.R., Mason, A.B. and Kaltashov, I.A. (2003). Interlobe Communication

in Human Serum Transferrin: Metal Binding and Conformational Dynamics

Investigated by Electrospray Ionization Mass Spectrometry. Biochemistry. 42,

5421-5428.

Haan, K. and Maggos, C. (2004). The Protein Engineer. BioCentury.12(5), A1-A7.

Hall, D.R., Hadden, J.M., Leonard, G.A., Bailey, S., Neu, M., Winn, M. and Lindley,

P.F. (2002). The Crystal and Molecular Structures of Diferric Porcine and

Rabbit Serum Transferrins at Resolutions of 2.15 and 2.60 A, Respectively.

Acta Crystallografica. Section D, Biological Crystallography. 58, 70-80.

Halperin, G., Breitenbach, M., Tauber-Finkelstein, M. and Shaltiel, S. (1981).

Hydrophobic Chromatography on Homologous Series of Alkyl Agaroses. A

comparison of Charged and Electrically Neutral Column Materials. Journal

of Chromatography. 215, 211–228.

Harrap, K.A. (1972). The Structure of Nuclear Polyhedrosis Viruses. The Inclusion

Body. Virology. 50, 114-123.

Harris, D.C., and Aisen, P. (1989). Physical Biochemistry of The Transferrins. In

Loehr, T.M., Gray, H.B. and Lever, A.B.P. (Eds.). Iron Carriers and Iron

Proteins. (pp. 239-351). Weinheim: VCH Publishers.

120

Harrison, P., Marwah, S.S., Hughes, R.T. and Bareford, D. (1994). Nontransferrin

Bound Iron and Neutropenia after Cytotoxic Chemotherapy. Journal of

Clinical Pathology. 47, 350–352.

Hase, S. (1996). Precolumn Derivatization for Chromatographic and Electrophoretic

Analyses of Carbohydrates. Journal of Chromatography A. 720, 173-182.

Hayter, P.M., Curling, E.M.A., Baines, A.J., Jenkins, N., Salmon, I., Strange, P.G.,

Tong, M.M. and Bull, A.T., (1992). Glucose-Limited Chemostat Culture of

Chinese Hamster Ovary Cells Producing Recombinant Human Interferon-γ.

Biotechnology and Bioengineering. 39, 327-335.

Hayter, P.M., Curling, E.M.A., Gould, M.L., Baines, A.J., Jenkins, N., Salmon, I.,

Strange, P.G. and Bull, A.T., (1993). The Effect of The Dilution Rate on

CHO Cell Physiology and Recombinant Interferon-γ Production in Glucose-

Limited Chemistat Culture. Biotechnology and Bioengienering. 42, 1007-

1085.

He, J. and Furmanski, P. (1995). Sequence Specificity and Transcriptional Activation

in the Binding of Lactoferrin to DNA. Nature. 373, 721-724.

Hirose M. (2000). The Structural Mechanism for Iron Uptake and Release by

Transferrins. Bioscience Biotechnology and Biochemistry. 64, 1328-1336.

Hjertén, S. (1973). Some General Aspects of Hydrophobic Interaction

Chromatography. Journal of Chromatography. 87, 325–331.

Hjertén, S. (1981). Hydrophobic Interaction Chromatography of Proteins, Nucleic

Acids, Viruses and Cells on Non-Charged Amphiphilic Gels. In: Glick, D.

(Ed.). Methods of Biochemical Analysis. (pp. 89-108). Canada: John Wiley

and Sons.

121

Hjertén, S., Rosengren, J. and Påhlman, S. (1974). Hydrophobic Interaction

Chromatography. The Synthesis and The Use of Some Alkyl and Aryl

Derivatives of Agarose. Journal of Chromatography. 101, 281–288.

Hjertén, S., Yao, K., Eriksson, K.-O. and Johansson, B. (1986). Gradient and

Isocratic High Performance Hydrophobic Interaction Chromatography of

Proteins on Agarose Columns. Journal of Chromatography. 359, 99–109.

Hofstee, B.H.J. (1973). Hydrophobic Affinity Chromatography of Proteins.

Analytical Biochemistry. 52, 430–448.

Hollister, J., Grabenhorst, E., Nimtz, M., Conradt, H., and Jarvis, D.L. (2002).

Engineering the Protein N-Glycosylation Pathway in Insect Cells for

Production of Biantennary, Complex N-Glycans. Biochemistry. 15093-15104.

Hollister, J.R. and Jarvis, D.L. (2001). Engineering Lepidopteran Insect cell Cells for

Sialoglycoprotein Production by Genetic Transformation with Mammalian

β1,4-Galactosyltransferase and α2,6-Sialytransferase Genes. Glycobiology.

11(1), 1-9.

Honda, S., Akao, E., Suzuki, S., Okuda, M., Kakehi, K. and Nakamura, J. (1989).

High-Performance Liquid Chromatography of Reducing Carbohydrates as

Strongly Ultraviolet-Absorbing and Electrochemically Sensitive 1-Phenyl-3-

Methyl-5-Pyrazolone Derivatives. Analytical Biochemistry. 180, 351-357.

Hrkal, Z. and Rejnkova, J. (1982). Hydrophobic Interaction Chromatography of

Serum Proteins on Phenyl-Sepharose CL-4B. Journal of Chromatography.

242, 385–388.

Huebers, H.A. and Finch, C.A. (1987). The physiology of Transferrin and

Transferrin Receptors. Physiological Reviews. 67, 520–582.

Ikeda, R.A., Bowman, B.H., Yang, F.M. and Lokey, L.K. (1992) Production of

Human Serum Transferrin in Escherichia coli. Gene. 117(2), 265-269.

122

Janson, J.-C. and Rydén, L. (1993). Protein Separation and Purification. In: Rehm,

H.-J. and Reed, G. (Eds.). Biotechnology, Vol. 3. (pp. 617-642). Weinheim:

VCH.

Janson, J-C. and Låås, T. (1978). Hydrophobic Interaction Chromatography on

Phenyl- and Octyl-Sepharose CL-4B. In: Roger, E. (Ed.). Chromatography of

synthetic and biological macromolecules. Chichester, England: Ellis

Horwood.

Jarvis, D.L. (2003). Developing Baculovirus-Insect Cell Expression Systems for

Humanized Recombinant Glycoprotein Production. Virology. 310, 1–7.

Jarvis, D.L. and Summers, M.D. (1989). Glycosylation and Secretion of Human

Tissue Plasminogen Activator in Recombinant Baculovirus-Infected Insect

Cells. Molecular Cell Biology. 9, 214–223.

Jarvis, D.L., Howe, D. and Aumiller, J.J. (2001) Novel Baculovirus Expression

Vectors That Provide Sialylation of Recombinant Glycoproteins in

Lepidopteran Insect Cells. Journal of Virology. 75, 6223–6227.

Jennissen, H.P. and Heilmeyer, I.M.G. (1975). General Aspects of Hydrophobic

Chromatography. Adsorption and Elution Characteristics of Some Skeletal

Muscle Enzymes. Biochemistry. 14, 754–760.

Jesionowski, G.A. and Ataai, M.M. (1997). An Efficient Medium for High Protein

Production in The Insect Cell/Baculovirus Expression System. Biotechnology

Progress. 13, 355-360.

John Morrow, K. (2007, March 1). Strategic Protein Production: Revamping and

Reengineering Existing Technology Can Increase Protein Yield and Efficacy.

Genetic Engineering and Biotechnology News. Retrieved September 6, 2007,

from http://www.genengnews.com

123

Joosten, C.E., Shuler, M.L. (2003). Production of A Sialylated N-Linked

Glycoprotein in Insect Cells: Role of Glycosidases and Effect of Harvest

Time on Glycosylation. Biotechnology Progress. 19, 193–201.

Joshi, L., Shuler, M.L. and Wood, H.A. (2001). Production of A Sialylated N-Linked

Glycoprotein in Insect Cells. Biotechnology Progress. 17, 822–827.

Kaplan, H.A., Welply, J.K. and Lennarz, W.L. (1987). Oligosacchary1 Transferase:

The Central Enzyme in The Pathway of Glycoprote in Asssembly.

Biochimica et Biophysica. Acta. 906, 161-173.

Kennedy, R.M. (1990). Hydrophobic Chromatography. In: Deutscher, M.P. (Ed.)

Guide to Protein Purification, Methods Enzymology, Vol. 182. (pp. 339-343).

London: Academic Press Inc.

Kimura, Y., Nakagawa, Y., Tokuda, T., Yamai, M.. Nakajima, S., Higashide. E. and

Takagi, S. (1992) Structures of N-linked Oligosaccharides of Microsomal

Glycoproteins from Developing Castor Bean Endosperms. Bioscience,

Biotechnology, and Biochemistry. 56, 215-222.

Kool, M. and Vlak, J.M. (1993). The Structural and Functional Organization of The

Autographa Californica Nuclear Polyhedrosis Virus Genome. Archives of

Virology. 130, 1-16.

Kool, M., Voncken, J.W., Vanlier, F.L.J., Tramper, J. and Vlak, J.M. (1991).

Detection and Analysis of Autographa Californica Polyhedrosis Virus

Mutants with Defective Interfering Properties. Virology. 183, 739-746.

Kornfeld, R. and Kornfeld, S. (1985). Assembly of Asparagine-Linked

Oligosaccharides. Annual Review of Biochemstry. 54, 631–66.

Kuehn, L., Meyer, H. and Reinauer, H. (1980). Hydrophobic Interaction

Chromatography as A Tool in Insulin Receptor Study In: Brandenburg, D.

124

and Wollmer, A. (Eds.) Proceedings of the 2nd. International Insulin

Symposium. Aachen, 243–250.

Kuroda, K., Geyer, H., Geyer, R., Doerfler, W. and Klenk, H.D. (1990). The

Oligosaccharides of Influenza Virus Hemagglutinin Expressed in Insect Cells

by A Baculovirus Vector. Virology. 174, 418–429.

Lăăs, T. (1975). Agar Derivatives for Chromatography, Electrophoresis & Gel-

Bound Enzymes. IV. Benzylated Dibromopropanol Cross-Linked Sepharose

as An Amphophilic Gel for Hydrophobic Salting Out Chromatography of

Enzymes with Special Emphasis on Denaturing Risks. Journal of

Chromatography. 111, 373–387.

Lawrence, S.M., Huddleston, K.A., Tomiya, N., Nguyen, N., Lee, Y.C., Vann, W.F.,

Coleman, T.A. and Betenbaugh, M.J. (2001). Cloning and Expression of

Human Sialic Acid Pathway Genes to Generate CMP-Sialic Acids in Insect

Cells. Glycoconjugate Journal. 18, 205-213.

Lee, K.S., Kim, B.Y., Kim, H.J., Seo, S.J., Yoon, H.J., Choi, Y.S., Kim, I., Han, Y.S.,

Je, Y.H., Lee, S.M., Kim, D.H., Sohn, H.D. and Jin, B.R. (2006). Transferrin

Inhibits Stress-Induced Apoptosis in A Beetle. Free Radical Biology and

Medicine. 41, 1151–1161.

Lee, S.Y. (1996). High Cell Density Cultivation of Escherical coli. Trends

Biotechnology. 14, 98-105.

Lefort, S. and Ferrara, P. (1986). Hydrophobic Adsorbants for The Isolation and

Purification of Biosynthetic Human Growth Hormone from Crude

Fermentation Mixtures. Journal of Chromatography. 361, 209–216.

Léger D., Campion, B., Decottignies, J.-P., Montreuil, J. and Spik, G. (1989).

Physiological Significance of The Marked Increased Branching of The

Glycans of Human Serotransferrin during Pregnancy. Biochemical Journal.

257, 231-238.

125

Lennarz, W.J. (1975). Lipid Linked Sugars in Glycoprotein Synthesis. Science. 188,

986-991.

Li, P.Z., Gao, X.G., Arellano, R.O. and Renugopalakrishnan, V. (2001).

Glycosylated and Phosphorylated Proteins—Expression in Yeast and Oocytes

of Xenopus: Prospects and Challenges—Relevance to Expression of

Thermostable Protein. Protein Expression and Purification. 22, 369–380.

Lim, H.J., Kim, Y,K., Hwang, D.S. and Cha, H.J. (2004). Expression of Functional

Human Transferrin in Stably Transfected Drosophila S2 Cells. Biotechnology

Progress. 20, 1192-1197.

Lindah, L. and Vogel, H.J. (1984). Metal-Ion-Dependent Hydrophobic Interaction

Chromatography of α-Lactalbumins. Analytical Biochemistry. 140, 394–402.

Luckow, V. L., and Summers, M.D. (1988). Trends in The Development of

Baculovirus Expression Vectors. Bio/Technology. 6, 47-55.

Luckow, V.A., Lee, S.C., Barry, G.F. and Olins, P.O. (1993). Efficient Generation of

Infectious Recombinant Baculviruses by Site-Specific Transposon-Mediated

Insertion of Foreign Genes into A Baculovirus Genome Propagated in

Escherischia coli. Journal of Virology. 67, 4566−4579.

MacGillivray, R. T. A., Mendez, E., Sinha, S. K., Sutton, M. R., Lineback-Zins, J.

and Brew, K. (1982). The Complete Sequence of Human Serum Transferrin.

Proceedings of The National Academic of Sciences of the United States of

America.79(8), 2504-2508.

MacGillivray, R. T. A., Mendez, E., Shewale, G., Sinha, S. K., Lineback-Zins, J. and

Brew, K. (1983). The Primary Structure of HumanSerum Transferrin. The

Structures of Seven Cyanogen Bromide Fragments and The Assembly of The

Complete Structure. The Journal of Biological Chemistry. 258, 3543-3553.

126

Maisano, F., Belew, M. and Porath, J. (1985). Synthesis of New Hydrophobic

Adsorbents Based on Homologous Series of Uncharged Alkyl Sulphide

Agarose Derivatives. Journal of Chromatography. 321, 305–317.

Makhatafze, G.I. and Privalov, P.L. (1995). Energetics of Protein Structure.

Advances in Protein Chemistry. 47, 307–325.

Makrides, S.C. (1996). Strategies for Achieving High Level Expression of Gene in

Escherichia coli. Microbiological Reviews. 60, 512-538.

Mason, A.B., He, Q-Y., Adams, T.E., Gumerov, D.R., Kaltashov, I.A., Nguyen, V

and MacGillivray, R.T.A. (2001). Expression, Purification, and

Characterisation of Recombinant Non-Glycosylated Human Serum

Transferrin Containing A C-Terminal Hexahistidine Tag. Protein Expression

and Purification. 23, 142–150.

Mason, A.B., Miller, K.M., Funk, W.D., Banfield, D.K., Savage, K.J., Oliver,

R.W.A., Green, B.N., MacGillivray, R.T.A. and Woodworth, R.C. (1993).

Expression of Glycosylated and Nonglycosylated Human Transferrin in

Mammalian Cells. Characterization of the recombinant proteins with

comparison to three commercially available transferrins. Biochemistry. 32,

5472-5479.

Matsuura, Y., Possee, R.D., Overton, H.A. and Bishop, D.H.L. (1987). Baculovirus

Expression Vectors: the Requirements for High Level Expression of Proteins,

Including Glycoproteins. Journal of General Virology. 68, 1233–1250.

McNair, R.D. and Kenny, A.J. (1979). Proteins of The Kidney Microvillar

Membrane. The Amphipathic Form of Dipeptidyl Peptidase IV. Biochemical

Journal. 179, 379-395.

Melander, W. and Horvath, C. (1977). Salt Effects on Hydrophobic Interactions in

Precipitation and Chromatography of Proteins: An Interpretation of The

Lyotropic Series. Archives of Biochemistry and Biophysics. 183, 200–215.

127

Melander, W., Corradini, D. and Horváth, C. (1984). Salt Mediated Retention of

Proteins in Hydrophobic Interaction Chromatography. Application of

Solvophobic Theory. Journal of Chromatography. 317, 67–85.

Mizutani, K., Okamoto, I., Fujita, K., Yamamoto,K. and Hirose, M. (2004).

Structural and Functional Characterization of Ovotransferrin Produced by

Pischia pastoris, Bioscience Biotechnology and Bichemistry. 68(2), 376-383.

Montreuil, J., Vliegenthart, J.F.G. and Schachter, H. (1995). Glycoproteins. In:

Neuberger, A. and Van Deenen, L.L.M. (Eds.). New Comprehensive

Biochemistry. (pp. 29a). Amsterdam: Elsevier.

Montreuil. J., Bouquelet. S., Dcbray, H., Foumet. B., Spik. G. and Strecker, G.

(1986). Glycoproteins. In Chaplin, M.F. and Kennedy, J.F. (Eds.)

Carbohydrate Analysis. (pp. 143-204). Britain: Oxford University Press.

Morgan, E.H. (1979). Studies on The Mechanism of Iron Release from Transferrin.

Biochimica et Biophysica Acta. 580, 312-326.

Muneo, T., Ken, O., Massanori, M. And Toshizumi, T. (1996). Secretion of A

Variant of Human Single-Chain-Urokinase-Type Plasminogen Activator

Without An N-Glycosylation Site in The Methyiotrophic Yeast, Pischia

pastoris and Characterization of The Secreted Product. Yeast. 12, 541-553.

Murphy, F. A., Fauquet, C. M., Bishop, D. H. L., Ghabrial, S. A., Jarvis, A. W.,

Martelli, G. P., Mayo, M. A. and Summers, M. D. (Eds.). (1995). Virus

Taxonomy. Sixth Report of the International Committee on Taxonomy of

Viruses. Berlin: Springer Verlag.

Murphy, K.P., Privalov, P.L. and Gill, S.J. (1990). Common Features of Protein

Unfolding and Dissolution of Hydrophobic Compounds. Science. 247, 559–

561.

128

Narhi, L.O., Kita, Y. and Arakawa, T. (1989). Hydrophobic Interaction

Chromatography in Alkaline pH. Analytical Biochemistry.182, 266–270.

Ochoa, J.L. (1978). Hydrophobic Interaction Chromatography. Biochimie. 60, 1–15.

Ogonah, O.W., Freedman, R.B., Jenkins, N., Patel, K., and Rooney, B.C. (1996).

Isolation and Characterization of An Insect-Cell Line Able to Perform

Complex N-Linked Glycosylation on Recombinant Proteins. Bio/Technology.

14, 197−202.

Olden, K., Parent, J.B. and White, S.L. (1982). Carbohydrate Moieties of

Glycoproteins A Reevaluation of Their Function. Biochimica. et Biophysica.

Act. 650, 209-232.

Ongkudon, C.M. (2006). Optimization of Recombinant Human Transferin

Expression in Insect Cells Baculovirus System. Master, Universiti Teknologi

Malaysia, Skudai

Oscarsson, S. and Kårsnås, P. (1998). Salt-Promoted Adsorption of Proteins Onto

Amphiphilic Agarose-Base Adsorbents II. Effects of Salt and Salt

Concentration. Journal of Chromatography A. 803, 83–93.

Parker, C.G., Fessler, L.L., Nelson, R.E. and Fessler, J.H. (1995). Drosophila UDP-

Glucose: Glycoprotein Glycosyltransferase: Sequence and Characterization

of The Enzyme That Distinguishes Between Denatured and Native Proteins.

The EMBO Journal. 14, 1294-1303.

Parodi, A.J. and Leloir, L.L.. (1979) The Role of Lipid Intermediates in The

Glycosylation of Proteins in The Eukaryotic Cell. Biochimica et Biophysica

Acta. 559, l-37.

Parsegian, V.A. and Ninham, B.W. (1970). Temperature-Dependent van der Waals

Forces. Biophysical Journal. 10, 664–674.

129

Patel, G., Nasmyth, K., and Jones, N. (1992). A New Method for The Isolation of

Recombinant Baculovirus. Nucleic Acids Research. 20, 97–104.

Patterson, R.M., Selkirk, J.K. and Merrick, B.A. (1995). Baculovirus and Insect Cell

Gene Expression: Review of Baculovirus Biotechnology. Enviromental

Health Perspectives. 103, 7-8.

Payne, S.L., Qi, X.M., Dwyer, A. and Fuller, F. (1998). Disease Induction by Virus

Derived from Molecular Clones of Equine Infectious Anemia Virus. Journal

of Virology. 72, 483-487.

Peakman, T.C., Harris, R.A. and Gewert, D.R. (1992). Highly Efficient Generation

of Recombinant Baculoviruses by Enzymatically Medicated Site-Specific in

Vitro Recombination. Nucleic Acids Research. 20, 495–500.

Peterson, E.A. and Sober, H.A. (1956). Chromatography of Proteins I. Cellulose Ion

Exchange Adsorbents. Journal of The American Chemical Society. 78, 751-

755.

Pharmigen (1999). Baculovirus Expression Vector System Manual. (6th Edition). San

Diego, USA: Pharmigen.

Pierce Biotechnology. (2005). Protein Assay: Technical Handbook. United States:

Pierce Biotechnology.

Ponka, P. (1997). Tissue-Specific Regulation of Iron Metabolism and Heme

Synthesis: Distinct Control Mechanisms in Erythroid Cells. Blood. 89(1), 1-

25.

Porath, J. (1986). Salt-Promoted Adsorption: Recent Developments. Journal of

Chromatography. 376, 331–341.

130

Porath, J. (1987). Metal Ion-Hydrophobic, Thiophilic and II-Electron Governed

Interactions and Their Application to Salt-Promoted Protein Adsorption

Chromatography. Biotechnology Progress. 3, 14–21.

Porath, J., Sundberg, L., Fornstedt, N. and Olson, I. (1973). Salting-Out in

Amphiphilic Gels as A New Approach to Hydrophobic Adsorption. Nature.

245, 465–466.

Porter, J.B., Abeysinghe, R.D., Marshall, L., Hider, R.C. and Singh, S. (1996).

Kinetics of Removal and Reappearance of Non-Transferrin Bound Plasma

Iron with Deferoxamine Therapy. Blood. 88, 705–713.

Possee, R.D. (1997). Baculoviruses as Expression Vectors. Current Opinion in

Biotechnology. 8, 569–572.

Privalov, P.L., Gill, S.J. (1988). Stability of Protein Structure and Hydrophobic

Interaction. Advances in Protein Chemistry. 39, 191–234.

Regnier, F.E. (1987). The Role of Protein Structure in Chromatographic Behavior.

Science. 238, 319–323.

Reiser, J. (1990). Advance in Biochemical Engineering/Biotechnology. (pp. 75-102).

Berlin, Germany: Spriner-Verlag.

Richardson, D.R. and Ponka, P. (1997). The Molecular Mechanisms of The

Metabolism and Transport of Iron in Normal and Neoplastic Cells.

Biochimica et Biophysica Acta. 1331, 1-40.

Rippel, G. and Szepesy, L. (1994). Hydrophobic Interaction Chromatography of

Proteins on An Alkyl-Superose Column. Journal of Chromatography A. 664,

27–32.

Rohrmann, G.F. (1986). Polyhedrin Structure. Journal of General Virology. 67,

1499-1513.

131

Rosengren, J., Påhlman, S., Glad, M. and Hjertén, S. (1975). Hydrophobic

Interaction Chromatography on Non-Charged Sepharose Derivatives. Binding

of A Model Protein, Related to Ionic Strength, Hydrophobicity of The

Substituent, and Degree of Substitution (Determined by NMR). Biochimica.

et Biophysica. Acta. 412, 51–61.

Sahlstedt, L., Ebeling, F., von Bonsdor, L., Parkkinen, J., Ruutu, N. (2001). Non-

Transferrin-Bound Iron During Allergenic stem cell transplantation. British

Journal of Haematology. 113, 836–838.

Salmon, V., Legrand, D., Georges, B., Slomianny, M-C., Coddeville, B. and Spik, G.

(1997). Characterization of Human Lactoferrin Produced in the Baculovirus

Expression System. Protein Expression and Purification. 9, 203-210.

Shaltiel, S. and Er-El, Z. (1973). Hydrophobic chromatography: Use for Purification

of Glycogen Synthetase. Proceedings of The National Academic of Sciences

of the United States of America. 70, 778–781.

Scopes, R.K. (1994). Protein Purification. (3rd Edition). New York: Springer-Verlag.

Shongwe, M.S., Smith, C.A., Ainscough, E.W., Baker, H.M., Brodie, A.M., and

Baker, E.N. (1992). Anion Binding by Human Lactoferrin: Results from

Crystallographic and Physicochemical Studies. Biochemistry. 31(18), 4451-

4458.

Shuler, M.L and Kargi, F. (2002). Bioprocess Engineering. (2nd edition). Upper

Saddle River, N.J.: Prentice Hall PTR.

Skinner, M.K., Cosand, W.L. and Griswold, M.D. (1984). Purification and

Characterization of Testicular Transferrin Secreted by Rat Sertoli Cells.

Biochemical Journal. 218, 313-320.

132

Spik, G., Debruyne, V., Montreuil, J., van Halbeek, H. and Vliegenthart J.F.G.

(1985). Primary Structure of Two Sialylated Triantennary Glycans from

Human Serotransferrin. FEBS Letters. 183(1), 65-69.

Srinivasan, R. and Ruckenstein, E. (1980). Role of Physical Forces in Hydrophobic

Interaction Chromatography. Separation and Purification Methods. 9, 267–

370.

Steinlein, L.M. and Ikeda, R.A. (1993). Production of N-Terminal and C-Terminal

Human Serum Transferrin in Escherichia coli. Enzyme and Microbial

Technology. 15, 193–199.

Steinlein, L.M., Graf, T.N. and Ikeda, R.A. (1995). Production and Purification of N-

Terminal Half-Transferrin in Pichia pastoris. Protein Expression and

Purification. 6, 619-624.

Stibler, H., Allgulander, C., Borg, S. and Kjellin, K.G. (1978). Abnormal

Microheterogeneity of Transferrin in Serum and Cerebrospinal Fluid in

Alcoholism. Acta Medica Scandinavica. 204(1-2), 49-56.

Summers, M.D. and Smith, G.E. (1978). Baculovirus Structural Polypeptides.

Journal of Virology. 84, 390-402.

Summers, M.D., Anderson, D.L. (1972). Granulosis Virus Deoxyribonucleic Acid: A

Closed, Double-Stranded Molecule. Journal of Virology. 9(4), 710-713.

Talukder, G. (2007, January). The Market for Bioengineered Protein Drugs. BCC

Research. Retrieved September 6, 2007, from http://www.bccresearch.com

Tanford, C. (1973). The Hydrophobic Effect: Formation of Micelles and Biological

Membranes. New York: Wiley.

Tiselius, A. (1948). Adsorption separation by salting out. Arkiv för Kemi, Mineralogi

Geologi. 26B, 1–5.

133

Tomiya, N., Howe, D., Aumiller, J.J., Pathak, M., Park, J., Palter, K.B., Jarvis, D.L.,

Betenbaugh, M.J. and Lee, Y.C. (2003). Complex-Type Biantennary N-

Glycans of Recombinant Human Transferrin from Trichoplusia ni Insect

Cells Expressing Mammalian β-1,4-Galactosyltransferase and β-1,2-N

Acetylglucosaminyltransferase II. Glycobiology. 13(1), 23-34.

Traving, C. and Schauer, R. (1998). Structure, Function and Metabolism of Sialic

Acids. Cellular and Molecular Life Sciences. 54, 1330-1349.

Tuite, M.F., Dobson, M.J., Roberts, N.A., King, R.M., Burke, D.C., Kingsman, S.M.,

and Kingsman, A.J. (1982). Regulated High Efficiency Expression of Human

Interferon-γ in Saccharomycas cerevisiae. The EMBO Journal. 1, 603-608.

Tulsi, B. (2004). Bugs Punch the Clock as Next Protein Manufacturers: Interest in

The Capabilities of Competing Insect Cell-Based Technologies Heightens as

Demand for Therapeutic Protein Production Capacity Escalates. Drug

Discovery & Development. 7, 56-62.

Valenzuela, P., Medina, A., Rutter, W.J., Ammerer, G. and Hall, B.D. (1982).

Synthesis and Assembly of Hepatitis-B Virus Surface Antigen Particles in

Yeast. Nature. 298, 347-350.

van Die, I., van Tetering, A., Bakker, H., van den Eijnden, D.H. and HJoziasse, D.

(1996). Glycosylation in Lepidopteran Insect Cells: Identification of A β1-4-

N-Acetylgalactosaminyltransferase Involved in The Synthesis of Complex-

Type Oligosaccharide Chains. Glycobiology. 6(2), 157-164.

van Eijk, H.G., van Noort, W.L., de Jong, G. and Koster, J.F. (1987) Human Serum

Sialo Transfertins in Diseases. Clinical Chimica Acta. 165, 141-145.

Van Oss, C.J., Good, R.J. and Chaudhury, M.K. (1986). Nature of The Antigen-

Antibody Interaction. Primary and Secondary Bonds: Optimal Conditions for

Association and Dissociation. Journal of Chromatography. 376, 111–119.

134

Varki, A. (1993). Biological Roles of Oligosaccharides: All of The Theories are

Correct. Glycobiology. 3, 97-130.

Varki, A., Cummings, R., Esko, J., Freeze, H., Hart, G. and Marth, J. (1999).

Essentials of Glycobiology. Cold Spring Harbor, NY: Cold Spring Harbor

Press.

Vervecken, W., Kaigorodov, V., Callewaert, N., Geysens, S., De Vusser, K., and

Contreras, R. (2004). In Vivo Synthesis of Mammalian-Like, Hybrid-Type N-

Glycans in Pichia pastoris. Applied And Environmental Microbiology. 70(5),

2639–2646.

Vieira, A.V. and Schneider, W.J. (1993). One-step chromatographic method for the

puification of avian serotransferrin. Protein Expression and Purification. 4,

110–113.

Visser, J. and Strating, M. (1975). Separation of Lipoamide Dehydrogenase

Isoenzymes by Affinity Chromatography. Biochimica et Biophysica Acta.

384, 69–80.

Viswanathan, K., Narang, S., Hinderlich, S., Lee, Y.C.| and Betenbaugh, M.J. (2005).

Engineering Intracellular CMP-Sialic Acid Metabolism into Insect Cells and

Methods to Enhance Its Generation. Biochemistry. 44, 7526-7534.

Vogel, H.J., Lindahl, L. and Thulin, E. (1983). Calcium-Dependent Hydrophobic

Interaction Chromatography of Calmodulin, Troponin C and Their

Proteolytic Fragments. FEBS Letters. 157, 241–246.

von Bonsdorff, L., Tölö, H., Lindeberg, E., Nyman, T., Harju, A. and Parkkinen, J.

(2001). Development of a Pharmaceutical Apotransferrin Product for Iron

Binding Therapy. Biologicals. 29, 27–37.

Waechter, C.J. and Lennarz, W.J. (1976) The Role of Polyprenol Linked Sugars in

Glycoprotein Synthesis? Annual Review of Biochemistry. 45, 95- 113.

135

Wang, Y., Chen, J., Luo, Y., Funk, W.D., Mason, A.B., Woodworth, R.C.,

MacGillivray, R.T. and Brayer, G.D. (1992). Preliminary Crystallographic

Analyses of The N-Terminal Lobe of Recombinant Human Serum

Transferrin. Journal of Molecular Biology. 227, 575-576.

Watanabe, S., Kokuho, T., Takahashi, H., Takahashi, M., Kubota, T. and Inumaru, S.

(2002). Sialylation of N-Glycans on The Recombinant Proteins Expressed by

A Baculovirus-Insect Cell System Under β-N-Acetylglucosaminidase

Inhibition. The Journal Of Biological Chemistry. 277(7), 5090–5093.

Wathen, M.W., Aeed, P.A., and Elhammer, A.P. (1991). Characterization of

Oligosaccharide Structures on A Chimeric Respiratory Syncytial Virus

Protein Expressed in Insect Cell Line Sf9. Biochemistry. 30, 2863−2868.

Weinberg, E. (1984). Iron Withholding: A Defense Against Infection and Neoplasia.

Physiological Reviews. 64, 65–102.

Whitaker, J.R. (I 977). Enzymatic Modification of Proteins Applicable to Foods. In:

Feeney, R.E. and Whitaker, J.R. (Eds.). Food Proteins. (pp. 126-134).

Advance in Chemistry Series 160. Washington D.C: American Chemical

Society.

Winzerling, J.J., Pham, D.Q.-D., Kunz, S., Samaraweera, P., Law, J.H. and Porath, J.

(1996). Purification of Recombinant Insect Transferrin from Large Volumes

of Cell Culture Medium Using High Capacity Ni 2+ -Dipicolylamine Gel

Protein Expression and Purification. 7, 137–142.

Wright, A. and Morrison, S. (1997). Effect of Glycosylation on The Antibody

Function: Implications for Genetic Engineering. Trends biotechnology. 15,

26-32.

Yamashita, K., Koide, N., Endo, T., Iwaki, Y. and Kobata, A. (1989). Altered

Glycosylation of Serum Transferrin of Patients with Hepatocellular

Carcinoma. The Journal of Biological Chemistry. 264, 2415-2423.

136

Yap, W.N. (2006). Development of Novel Galactosylation Method for the

Expression of Recombinant Human Transferrin in Insect Cell Culture. Master,

Universiti Teknologi Malaysia, Skudai.

Yeh, J.C., Seals, J.R, Murphy, C.I., van Halbeek, H. and Cummings, R.D. (1996).

Site-Specific N-glycosylation and Oligosaccharide Structures of

Recombinant HIV-1 gp120 Derived from A Baculovirus Expression System.

Biochemistry. 32(41), 11087–11099.

YSI. (2001). Immobilized Enzyme Biosensor. YSI. Retrieved October 13, 2007, from

http://www.ysilifesciences.com.

Yun, E.Y., Goo, T.W., Kim, S.W., Choi, K.H, Hwang, J.S., Kang, S.W and Kwon,

O.Y (2005). Galatosylation and Sialylation of Mammalian Glycoproteins

Produced by Baculovirus-Mediated Gene Expression in Insect Cells.

Biotechnology Letters. 27, 1035-1039.

Zdebska, E. and Kościelak, J. (1999). A Single-Sample Method for Determination of

Carbohydrate and Protein Contents Glycoprotein Bands Separated by Sodium

Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis. Analytical

Biochemistry. 275, 171-179.

APPENDIX A

APPENDIX A-1 Stock Solution for SDS-PAGE

1. 2M Tris-HCl (pH8.8), 100ml

Weight out 24.2g Tris-base and add to 80ml distilled water.

Add 1M HCl slowly to adjust pH buffer to pH 8.8.

Add distilled water to total volume 100ml.

2. 1M Tris-HCl (pH 6.8), 100ml

Weight out 12.1g Tris base and add to 80ml distilled water.

Add 1M HCl slowly to adjust pH buffer to pH 6.8.

Add distilled water to total volume 100ml.

3. 10% SDS(w/v), 50ml

Weight out 5g SDS.

Dissolve the SDS with a total volume of 50ml distilled water.

4. 50% glycerol (v/v), 100ml

Add 50ml distilled water to 50ml 100% glycerol.

5. 1% bromophenol blue (w/v),10ml

Weight out 100mg bromophenol blue.

Dissolve the powder with a total volume of 10ml distilled water.

138

APPENDIX A-2 Working Solution for SDS-PAGE

1. Solution A (Acrylamide Stock Solution), 100ml

30% (w,v) acrylamide, 0.8% (w/v) bis-acrylamide

Weight out 29.2g acrylamide and 0.8g bis-acrylamide and make total volume to

100ml.

2. Solution B (4x separating gel buffer), 100ml

75ml 2M Tris-HCl (pH8.8)

4ml 10% SDS

21ml distilled water

3. Solution C (4x stacking gel buffer), 100ml

50ml 1M Tris-HCl (pH6.8)

4ml 10% SDS

46ml distilled water

4. 10% ammonium persulfate

0.05 g in 0.5ml distilled water

5. Electrophoresis buffer, 1L

3g Tris

14.4g glycine

1g SDS

Add distilled water to make 1L.

6. 5x sample buffer, 10ml

0.6ml 1M Tris-HCl (pH6.8)

5ml 50% glycerol

2ml 10% SDS

0.5ml 2-mercaptoethanol

1ml 1% bromophenol blue

0.9ml distilled water

139

APPENDIX A-3 Separating and Stacking Gel Preparation

1. Separating Gel X% Preparation

Chemical Volume

Solution A x/3 ml

Solution B 2.5 ml

H2O (7.5-x/3) ml

10% ammonium persulfate 50ul

TEMED 10ul

2. Stacking Gel Preparation

Chemical Volume

Solution A 0.67ml

Solution C 1.0ml

H2O 2.3ml

10% ammonium persulfate 30ul

TEMED 10ul

140

APPENDIX-B

Coomassie Blue Staining

1. Coomassie Gel Stain, 1 liter

1g Coomassie Blue R-250

450ml methanol

450ml distilled water

100ml glacial acetic acid

2. Coomassie Gel Destain, 1 liter

100ml methanol

100ml glacial acetic acid

800ml distilled water

141

APPENDIX C

Preparation of Optimized Medium

1) Stock Solution

Note: *Stock solution was purchased from Sigma

**Powder glutamine dissolved in medium SFM 900 II

2) Working Solutions:

According to this the equation: M1V1 = M2V2

Note: All the stock solutions, V1 were mixed and diluted to a total volume of 500ml

medium SFM900II

M1 V1 M2 V2

Chemicals Stock Solutions

(mg/L)

Volume

(ml)

Working Solutions

(mg/L)

Volume

(ml)

Glucose 100000 6.46 1291.95 500

Glutamine 25000 44.22 2211.2 500

Lipids 1000xc 100% 3.2 0.64 500

Chemicals Concentration g/L/v or v/v%

*Glucose 100

**Glutamine 25

Lipids 1000xc 100%

142

Example of TCID50 Calculation (spreadsheet)

(1)

(2)

Number Number Total Total % % % Log of of Number Number Total Above Above Below Dilution Infected Uninfected Infected Uninfected Infected 0.5 0.5 0.5 Above

Dilution

Wells Wells 0.5 0.0001 7.000 3.000 22.000 3.000 0.880 TRUE 0.000 0.000 0.000

0.00001 7.000 3.000 15.000 6.000 0.714 TRUE 0.714 0.000 -5.0000.000001 4.000 6.000 8.000 12.000 0.400 FALSE 0.000 0.400 0.000

0.0000001 4.000 6.000 4.000 18.000 0.182 FALSE 0.000 0.000 0.000 71.429 40.000 -5.000

Number Number Total Total % % % Log of of Number Number Total Above Above Below Dilution Infected Uninfected Infected Uninfected Infected 0.5 0.5 0.5 Above

Dilution

Wells Wells 0.5 0.0001 9.000 1.000 26.000 1.000 0.963 TRUE 0.000 0.000 0.000 0.00001 7.000 3.000 17.000 4.000 0.810 TRUE 0.000 0.000 0.000 0.000001 5.000 5.000 10.000 9.000 0.526 TRUE 0.526 0.000 -6.000 0.0000001 5.000 5.000 5.000 58.000 0.079 FALSE 0.000 0.079 0.000 52.632 7.937 -6.000

APPE

ND

IX D

142

143

APENDIX D: (Continued)

Spread sheet (1) Spread Sheet (2)

Number of Wells 10.000 10.000

ml/well 0.010 0.010

Prop. Dist. 0.682 0.059

Log TCID -5.682 -6.059

TCID50 2.081E-06 8.732E-07

1/TCID50 4.806E+05 1.145E+06

TCID50/ml 4.806E+07 1.145E+08

pfu/ml 3.32E+07 7.90E+07

Average pfu/ml: (3.32E+07)+(7.9E+07)/2= 5.61E+07

144

APPENDIX E

Working Solution for ELISA

1. Coating Buffer-Carbonate-bicarbonate buffer (50ml)

0.05M carbonate sodium carbonate, pH 9.6.

Dissolve 1 sodium carbonate-bicarbonate capsule (Sigma) in 100ml distilled

water.

Add 1M HCl to adjust pH to 9.6

Add distilled water to a total volume of 50ml

2. Tris Buffered Saline (TBS)

50mM Tris/0.14M NaCl, pH 8.0

Dissolve 12.11g Tris, 58.44g NaCl in 1800 ml distilled water.

Add 1M HCl to adjust pH buffer to pH8.0.

Add distilled water to a total volume of 2000ml

3. Wash Solution

50mM Tris buffered saline, pH 8.0; 0.14M NaCl,; 0.05% Tween 20

Add 0.5ml Tween 20 to 1000ml TBS

4. Blocking Solution (freshly prepared)

50mM Tris, 0.14 M NaCl, 1%BSA, pH8.0.

Dissolve 0.5g BSA with 50ml TBS.

5. Sample/Conjugate Diluent (freshly prepared)

50mM Tris, 0.14 M NaCl, 1%BSA, 0.05% Tween 20, pH 8.0

Dissolve 1g BSA with 100ml washing solution

145

6. Stopping Solution

1M Phosphoric Acid

Diluted 125ml 4M phosphoric acid to a total volume of 500ml 1M phosphoric

acid

7. Preparation of Standard

Step Ng/ml Calibrator Sample diluents

0 1000 5ul 1ml

1 250 1ml from step 0 1ml

2 125 1ml from step 1 1ml

3 62.5 1ml from step 2 1ml

4 31.25 1ml from step 3 1ml

5 15.625 1ml from step 4 1ml

6 7.8 1ml from step 5 1ml

7 3.9 1ml from step 6 1ml

146

APPENDIX F

Working Solution for Western Blot

1. Towbin Buffer

24 mM Tris, 192 mM glycine and 10% methanol

2. 1X Phosphate Buffer Saline (PBS), pH 7.4

8g NaCl, 0.2g K2HPO4 and 0.24g KH2PO4

Adjust to pH7.4 and total up volume to 1L.

Sterilize by autoclaving

3. Blocking Solution

5% skimmed milk in 1X PBS buffer

4. Washing Solution

0.05% Tween 20 in 1X PBS buffer

5. TMB (3,3’,5,5’-tetramethylbenzidene) Stabilized Substrate for HRP

(Promega, Madison, WI)

147

APPENDIX G

Mobile Phase for Purification

Buffer for HIC

(1) 2.4M Ammonium Sulphate/0.8M Sodium Citrate Buffer, pH6.0 (2x Buffer A)

Dissolved 79.28g ammonium sulphate, 50.5g sodium citrate and 5.38g citric acid

in 250ml distilled water.

(2) 1.2M Ammonium Sulphate/0.4M Sodium Citrate Buffer, pH6.0 (Buffer A)

Dissolved 158.568g ammonium sulphate, 107.052g sodium citrate and 6.917g

citric acid in 1000ml distilled water

Buffer for IEX

(1) 20mM Tris-HCl, pH8.5

Dissolved 4.846g Tris in 1800ml distilled water.

Adjust pH with 1M HCl to pH8.5

Total up volume to 2000ml with distilled water.

(2) 20mM Tris-HCl/ 0.5M NaCl, pH8.5

Dissolved 2.4223g Tris and 29.22g NaCl in 800ml distilled water

Adjust pH with 1M HCl to pH8.5

Total up volume to 1000ml with distilled water.

148

APPENDIX H

Glycan Analysis

Stock Solution

Chemical Molecular Weight

(g/mol) Concentration

Mannose 180.16 5mg/ml

Glucosamine 215.64 5mg/ml

D-Galactose 180.16 5mg/ml

D-Glucose 180.16 5mg/ml

Arabinose 150.13 5mg/ml

TFA 114.02 4.4M

NaOH 40.00 0.3M

HCl 36.45 0.3M

Apo-Transferrin 76000 383.8µg/ml

Sample 73000 383.8µg/ml

149

Example for the Preparation for sample and standards for carbohydrate hydrolysis

Volume

Apo Transferrin

(365µg/ml)

Sample/rhTf

(365µg/ml)

Std 1

(10µg/ml )

Std 2

(5µg/ml)

Std 3

(2.5µg/ml)

Std 4

(1.25µg/ml)

Apo-Transferrin

(383.8µg/ml)

95µl - - - - -

Sample

(383.8µg/ml) - 95µl - - - -

Mannose

(50µg/ml) - - 20µl 10µl 5µl 2.5µl

Glucosamine

(50µg/ml) - - 20µl 10µl 5µl 2.5µl

Galactose

(50µg/ml) - - 20µl 10µl 5µl 2.5µl

Glucose

(50µg/ml) - - 20µl 10µl 5µl 2.5µl

Water 5µl 5µl 20µl 60µl 80µl 90µl

Total Volume 100µl 100µl 100µl 100µl 100µl 100µl

*Each sample/standard was mixed in hydrolysis tube and added with 1ml TFA.

APPE

ND

IX H

: Continued

M2 M1

149

150

APPENDIX H: Continued

Mobile phase of HPLC

(1) Mobile Phase A

0.1M Ammonium Acetate/ 10% Acetonitrile, pH5.5

Dissolved 7.708g Ammonium Acetate in 250ml acetonitrile and about 600ml

deionized water

Adjust pH with 1M HCl to pH5.5

Total up volume to 1000ml with deionized water.

(3) Mobile Phase B

0.1M Ammonium Acetate/ 25% Acetonitrile, pH5.5

Dissolved 7.708g Ammonium Acetate in 250ml acetonitrile and about 600ml

deionized water

Adjust pH with 1M HCl to pH5.5

Total up volume to 1000ml with deionized water.