physiological studies of a murine monoclonal ...epubs.surrey.ac.uk/856461/1/27721069.pdf1.2...
TRANSCRIPT
PHYSIOLOGICAL STUDIES OF A MURINE
HYBRIDOMA: GROWTH AND
MONOCLONAL ANTIBODY PRODUCTION
A thesis submitted to the University of Surrey
for the degree of
Doctor of Philosophy
by
MICHELLE FRANCES SCOTT B.Sc.
Department of Microbiology
University of Surrey
Guildford
September
1990
ProQuest Number: 27721069
All rights reserved
INFORMATION TO ALL USERS The qua lity of this reproduction is d e p e n d e n t upon the qua lity of the copy subm itted.
In the unlikely e ve n t that the au tho r did not send a co m p le te m anuscrip t and there are missing pages, these will be no ted . Also, if m ateria l had to be rem oved,
a no te will ind ica te the de le tion .
uestProQuest 27721069
Published by ProQuest LLC (2019). C opyrigh t of the Dissertation is held by the Author.
All rights reserved.This work is protected aga inst unauthorized copying under Title 17, United States C o de
M icroform Edition © ProQuest LLC.
ProQuest LLC.789 East Eisenhower Parkway
P.O. Box 1346 Ann Arbor, Ml 48106 - 1346
ABSTRACT
The physiology of growth and monoclonal antibody (Mab) production by
a murine hybridoma, PQX Bl/2, which secretes an antibody to paraquat, has
been investigated.
A triphasic sequence characterising antibody production was revealed by
partial cubic spline interpolation of measurements of culture biomass, Mab
titre, glutamine, ammonia and amino acids concentrations.
During phase I peaks in the utilisation of amino acids, primarily GLN,
and ammonia production, were revealed together with apparent inhibition of
growth. Specific growth rate peaked in phase II, during which ammonia was
apparently assimilated. This assimilation was investigated, and the mechanism
proposed to be via a cyclic coupling of glutamate dehydrogenase to alanine
aminotransferase. Further peaks in the assimilation of amino acids were observed
in phase III during which Mab production rate reached a maximum. These
observations are discussed in relation to possible cell-cycle events.
The processes of growth and product formation were further investigated
by determination of materials balance equations. The nutritional requirements for
each phase were found to be different.
The significance of amino acids for hybridoma culture has been
highlighted, and a fractional factorial method employed to investigate the
importance of individual amino acids for growth and production. The
requirements for each process were found to be different: some amino acids
were important for growth but not critical for Mab production (eg lysine)
whereas others (eg serine) could be omitted with little effect on growth but a
profound effect on product formation.
ACKNOWLEDGMENTS
I would like to thank my supervisor Dr Mike Bushell for his advice,
support and help during this project. I am very grateful to Dr Noel Warded
for his technical assistance with the bioreactor experiments and his endless
encouragement and sense of humour- especially when results were not
forthcoming, and reports were due. Thanks also to Sarah Bell for her help,
encouragement, and croissants on the "dawn patrol"; to Dr Pete Sanders, Dr
Keith Snell, Tina O’Shaunessy and Niall Higbee (the other members of
"Team Surrey") for their technical assistance and advice.
I am grateful to Eric Ah-Sing from the department of Biochemistry at
the University of Surrey, and to SmithKline Beecham Pharmaceuticals for
their advice and technical assistance with amino acid analysis.
Thanks also to the many members of the Department of Microbiology,
past and present, for their friendship and help, especially Jane Thorpe, Jackie
McDermott, Ming Lo, Peter Angell, Sue Wall, Jane Newcombe, Maureen
Hodson and Di Simpson.
Finally, I would like to thank my family, for their endless help,
support and encouragement over the years, which made this work possible.
I gratefully acknowledge the financial support of the SERC
Biotechnology Directorate (Grant No.GR/D/96019).
CONTENTS
Abstract i
Acknowledgments ii
Contents iii
List of figures vü
List of tables xi
List of abbreviations xii
CHAPTER 1. INTRODUCTION
1.1 Animal Cells In Culture 1
1.1.1 The History 1
1.1.2 The Nutrition of Animal Cells in Culture 2
1.1.3 Production and Application of Animal Cell Products 6
1.2 Hybridoma Technology 8
1.2.1 The Humoral Immune Response 9
1.2.2 B-Cell hybridomas 14
1.2.3 Applications of Monoclonal Antibodies 17
1.2.4 Monoclonal Antibody Production Systems 22
1.2.5 Factors Affecting Mab Production 26
1.3 The Aims of the Project 29
CHAPTER 2 MATERIALS AND METHODS
2.1 Cell Culture Methods
2.1.1 The Cell Line 31
2.1.2 Sterilization Procedures 31
2.1.3 Cell Counting 32
2.1.4 Dry Weight Determination 34
i i i
2.1.5 Culture media 35
2.1.6 Cell passaging 35
2.1.7 Large scale culture 36
2.1.8 Sterility Checking 37
2.1.9 Cryopreservation 37
2.2 Biochemical Methods
2.2.1 Polyacrylamide gel electrophoresis 38
2.2.2 Lowry Protein Assay 39
2.2.3 Affinity purification of anti-paraquat monoclonal antibody 39
2.2.4 Enzyme linked Immunosorbent Assay (ELISA) 40
2.2.5 Glutamine / Ammonia Assay 43
2.2.6 Lactate determination 43
2.2.7 Glucose determination 44
2.2.8 Amino Acid Analysis 45
2.2.9 Ornithine carbamyl transferase assay 48
2.2.10 Glutamate dehydrogenase assay 48
CHAPTER 3. GROWTH AND MONOCLONAL ANTIBODY
PRODUCTION
3.1 Introduction
3.1.1 Parameters of Growth 49
3.1.2 Relationship of Growth Rate to Product Formation Rate 53
3.1.3 Growth, monoclonal antibody production and the cell cycle 56
3.1.4 Process Analysis Techniques 60
3.2 Investigation of Growth and Monoclonal Antibody Production
3.2.1 Growth and monoclonal antibody production By PQX Bl/2 in 62
serum-supplemented medium
3.2.2 Nutrient assimilation by PQX Bl/2 in batch culture 73
3.2.3 Growth and Mab Production in Serum-Free 83
CHAPTER 4 PROCESS DESCRIPTION USING MATERIALS
BALANCING
4.1 Introduction
4.1.1 The use of materials balancing 94
4.1.2 The concept of material balancing 95
4.2 Materials Balance Description of Growth and Monoclonal Antibody
Production by PQX Bl/2
4.2.1 Gas Analysis of PQX Batch Culture 97
4.2.2 Derivation of a Materials Balance Equation 103
CHAPTER 5. NITROGEN METABOLISM
5.1 Introduction
5.1.1 Amino Acids and Animal Cells in Culture 109
5.1.2 Transport of Amino Acids across Membranes 111
5.1.3 The Role of Vit B6 in Amino Acid Metabolism 113
5.1.4 Amino Acid Catabolism 113
5.1.5 The Urea Cycle 114
5.1.6 Carbon Skeleton of Amino Acids 116
5.1.7 Anabolism of Amino Acids 117
5.1.8 Formation of Amino Acid Amides 117
5.1.9 Physiological Effects of Ammonia on Animal Cells. 118
5.2 Assimilation of Ammonia by PQX Bl/2
5.2.1 The Role of The Urea Cycle 120
5.2.2 15N-NMR Studies of Ammonia Assimilation 121
5.2.3 Determination of glutamate dehydrogenase activity in PQX Bl/2 126
v
5.3 Amino Acid Utilisation 128
5.3.1 Factorial Determination of Important Amino Acids 129
5.3.2 Elimination of Individual Amino Acids 139
5.3.3 Elimination of "Essential" Amino Acids 146
CHAPTER 6 . FINAL DISCUSSION 151
CHAPTER 7. CONCLUSIONS 158
CHAPTER 8 . FUTURE EXPERIMENTS 160
APPENDICES 1-4 xiv
APPENDIX 5 Backcover
REFERENCES xxxvi
v i
FIGURES Pg
1.1 Basic structure of an Immunoglobulin molecule 10
1.2 Gene rearrangements of the regions of the mouse 12
immunoglobulin heavy chain
1.3 The production of murine hybridomas from a B cell/ 16
myeloma fusion
2.1 A modified Fuch-Rosenthal Haemocytometer slide 33
2.2 Dilution series for the determination of Mab 42
titre of standard and samples by ELISA
2.3 An example of a chromatogram for the determination 47
of amino acid concentration of a standard solution
3.1 A typical growth curve for a microbial batch culture 52
3.2 The Leudeking-Piret Model 56
3.3 The cell cycle of cultured cells 57
3.4a The growth curve of PQX Bl/2 in batch culture 63
3.4b The determination of specific growth rate from the 64
change in biomass.
3.4c Determination of specific growth rate from the viable 64
cell count.
3.4d Volumetric production of Mab by PQX Bl/2 in batch 66
culture, and the major peak in product formation rate.
3.4e The relationship between specific growth rate and 67
product formation rate.
3.4f A Leudeking-Piret plot to show the association of 68
specific growth rate and product formation rate.
3.4g Complex kinetics of production of penicillin 69
3.4h The volumetric change in glutamine concentration 70
and the rate of utilisation
3.4i The major peaks ammonia production and ammonia 71
assimilation, compared to the major growth period
v i i
3.5 The phasic nature of nutrient assimilation
3.6a Growth curve for PQX Bl/2, showing the increase
in viable cell count, and the specific growth rate
3.6b The displacement in time of the peak in specific
growth rate and product formation rate
3.6c The major peaks in the utilisation of glutamine,
the major peaks in ammonia production rate and
apparent assimilation
3.7a The data from Bree et al (1988) and the determination
of a period of apparent ammonia assimilation
3.8a The volumetric change in Isoleucine concentration
and the rate of utilisation
3.8b The change in volumetric concentration of alanine,
and the rate of production
3.9a The change in glucose concentration and the
major peak in utilisation rate
3.9b The increase in lactate concentration as glucose
concentration decreased
3.10a The growth curve of PQX Bl/2 hybridomas adapted
to serum-free medium.
3.10b The volumetric production of Mab by PQX Bl/2 cells
adapted to serum-free medium, and the rate of
—product formation
3.10c The relationship between specific growth rate and
product formation rate in PQX Bl/2 hybridomas
adapted to serum-free medium.
3.10d The change in ATP production
3.10e The change in adenylate charge
3.10f A trajectory plot to show the relationship
between specific growth rate, energy charge
and Mab production.
72
73
74
75
76
77
78
80
81
85
86
87
90
91
91
v i i i
4.1
4.2a
4.2b
4.2c
4.2d
4.2c
4.2f
5.1
5.2
5.3a
5.3b
5.3c
5.3d
5.4
5.5
5.6a
5.6b
Components and uses of materials balancing
The typical growth curve for the PQX Bl/2 hybridomas
showing the increase in viable cell count and the
specific growth rate
The production of Mab, and major peaks in product
formation rate
The relationship between specific growth rate and
product formation rate
The change in respiratory quotient
The change in dissolved oxygen tension
The change in air flow rate
The urea cycle: intermediates and enzymes
the fate of the carbon skeleton of amino acids
The growth curve of PQX Bl/2 hybridomas in the
presence of ImM NH4C1
The production, of Mab by cells grown in the
presence of ImM NH4C1, and the major peak in
product formation rate
The assimilation peaks of ammonia during a batch
culture of hybridomas grown in the presence of
ImM NH4C1.
The cyclic coupling of glutamate dehydrogenase to
alanine aminotransferase
GDH activity was at a minimum between the peaks in
NH3 assimilation
The Plackett-Burman matrix
The growth curve for the control flask containing
all the amino acids
The growth curve of the control with all the amino
acids omitted
96
99
99
100
101
101
102
115
116
122
123
124
125
127
130
141
141
i x
5.7a Volumetric Mab production for the control flask 142
containing all the amino acids
5.7b Volumetric Mab production for flask 6 (minus SER) 143
5.7c Volumetric Mab production for flask 7 (minus PRO) 143
5.7d Volumetric Mab production for flask 13 (minus OHP) 144
5.7e Volumetric Mab production for flask 15 (minus GLY) 144
5.7f Volumetric Mab production for flask 17 (minus GLU) 144
5.7g Volumetric Mab production for flask 19 (minus ASP) 145
5.7h Volumetric Mab production for flask 20 (minus ASM) 145
5.8a Growth Curves of Flasks with and without the 147
"non-essential" amino acids
5.8b Mab production in flasks with and without the 148
"non-essential" amino acids
5.8c The glucose utilisation of flasks with and without 149
the "non-essential" amino acids.
5.8d The production of lactate from flask and without 150
"non-essential" amino acids.
x
TABLES Pg
1.1 Basic medium components and their function 3
1.2 Problems associated with the use of 4
serum-supplementation in cell culture medium
1.3 Advantage and disadvantages of serum-free 6
media for hybridomas
1.4 Examples of products from animal cell cultures 7
3.1 The difference in amino acid assimilation between 89
serum-free and serum-dependent cells.
4.1a Consumption and appearance of major reaction 105
components between 20 and 23 hours.
4.1b Consumption and appearance of major reaction 107
components between 45 and 45 hours.
4.1c Proposed feed formulation for a fed-batch culture 108
based on the materials balance equations
5.1a Amino acid transport systems 112
5.2 Appearance of 15N labelling in a PQX Bl/2 bioreactor 125
batch culture supplemented with ImM 15NH4C1
5.3a Viable counts for flasks 1-24 134
5.3b Ranking of amino acids:growth 135
5.3c Mab titre for flask 1-24 137
5.3d Ranking of amino acids: Mab production 138
5.4a Amino acid deletions 140
5.4b Labelling of flasks and amino acid omissions 142
6 A summary of proposed events of the phasic culture 155
in relation to the cell cycle.
xi
ABBREVTATTONS
ADP Adenosine diphosphateALA AlanineALT Alanine aminotransferaseAMP Adenosine monophosphateARG ArginineASN AsparagineASP AsparatateATP Adenosine triphosphateerr CitrullineCYS Cyst(e)inedl Decilitre(s)DNA Deoxyribonucleic acidEC Energy chargeELISA Enzyme linked immunosorbent assayg Gravitational force (when used in terms of centrifugation)g Gram(s)G6PDH Glucose-6-phosphate dehydrogenaseGDH Glutamate dehydrogenaseGLN GlutamineGLU GlutamateGLY GlycineGS Glutamine synthetaseh'1 per hourHK HexokinaseHIS HistidineHPRT Hypoxanthine phosphoribosyl transferaseIg Immunoglobulin1 Litre(s)ILE Isoleucinea-KG alpha-ketoglutarateKi Inhibition constantKs Saturation constantLEU LeucineLYS LysineM MolarMET MethionineMab Monoclonal antibodymg Milligram(s)mM MillimolarNAD Nicotinamide adenine dinucleotideNADH Nicotinamide adenine dinucleotide (reduced form)NMR Nuclear magnetic resonanceml Millilitre(s)
OCT Ornithine carbamyl transferaseOHP HydroxyprolineORN OrnithinePRO ProlineRPMI Roswell Park Memorial Institute MediumSDS Sodium dodecyl sulphateSER Serinerpm Revolutions per minuteTRY TryptophanTYR TyrosineVAL ValineIjl Specific growth rateHg Microgram(s)
1. INTRODUCTION
1.1 ANIMAL CELLS IN CULTURE
1.1.1 The History
A mammalian body is composed of millions of cells, each of which
contains one to three million genes. In differentiated tissues only a limited sub
set of the phenotypic potential is expressed.
Using a whole animal therefore, makes determination of the multiplicity
of metabolic reactions occurring in a particular cell very difficult. Developments
in cell culture techniques however, provide a means of investigating cellular
metabolism more fully. Culturing a particular cell type enables its’ physiology
and response to stimuli to be studied.
The earliest reports of maintenance of explanted tissue in vitro came in
the late nineteenth century, when chick embryo tissue was maintained in warm
saline for a few days (Roux, 1885). Work by Harrison (1907) stimulated
interest, with a study on growth of nerve cells in lymph clots. Subsequent work
(Burrows, 1910; Carrel and Burrows, 1911; Lewis and Lewis, 1912), led to
refinement of amino acid and salt composition in culture media containing
clotted lymph or plasma as growth promoting substances.
The first demonstration of serial propagation was by Rous and Jones
(1916). They dispersed chick embryo cells from tissue, grew them in a plasma
clot, redispersed using trypsin and showed subsequent regrowth of the cells.
This early work demonstrated that tissue physiology could be investigated by in
vitro culture techniques.
Work during the following three decades revealed that the cellular energy
requirements of cells in culture could be met by glucose (Lewis, 1922;
Warburg, 1930) and/or deamination of amino acids (Holmes and Watchom,
1927; Warburg and Kubowitz, 1927). Development of media with defined
1
components (Earle et al, 1951; Fischer et al, 1948; Morgan et al, 1950; White,
1949) culminated in the determination of essential amino acid and vitamin
requirements of several mammalian cell lines (Eagle, 1955a; 1955b). Indeed,
modifications of these early media are still used extensively today.
1.1.2 The Nutrition of Animal Cells in Culture
Basic Nutritional Requirements
The functions of cell growth medium may be defined as:
a) to maintain suitable physiochemical conditions (eg pH and osmolarity)
and
b) to provide nutrients for synthesis of cell biomass and products and to
meet energy requirements.
A nutrient may be defined as "a substance which enters the cells and is
utilised as a metabolic substrate or co-factor" (Lambert et al, 1985). Anything
else needed for cellular proliferation is normally classified as a supplement,
including all undefined additives such as serum. Indeed, a complete medium can
be considered theoretically, to be composed of two distinct parts:-
i) a basal nutrient medium that satisfies all cellular (maintenance)
requirements for nutrients and
ii) a set of supplements that satisfy other types of cellular growth
requirements and make the growth of cells in basal medium possible
(Ham, 1984).
The term nutrient is usually assumed to exclude substances whose roles
are primarily regulatory, such as hormones and growth factors, despite the fact
that such substances often enter cells and undergo catabolism. Attachment
factors, extracellular matrix-and carrier proteins are also generally not classified
as nutrients, although the carrier proteins often promote cellular multiplication
by making nutrients such as iron and lipids more readily available to the cells.
2
Inorganic ions however play major roles that fall outwith the definition of a
nutrient, such as maintenance of membrane potentials and osmotic balance. The
basic medium components and their functions are shown in Table 1.1.
Table 1.1
COMPONENT
Inorganic ions
Trace Elements
Amino Acids
Vitamins
Carbohydrates
Lipids
Basic Medium Components and Their Function
EXAMPLES FUNCTION
Na, K, Ca, Mg,
PO/, HCO3-, Cl,
Maintenance of osmotic
potentials, buffering, cofactors
for enzyme reactions.
Co, Cu, I, Fe,
Mn, Mo, Zn, Se,
Cr, Ni, V, As, Si
F, Sn
Elimination of free radicals,
cofactors.
Arg, Gin, Met, He
Leu ...Energy sources, precursors for
many intermediates in metabolism
Protein synthesis.
A, D, C, E, K,
B group
Cofactors
Glucose, fructose
galactoseEnergy sources, glycolytic
intermediates
Linoleic acid
Cholesterol,
Oleic acid
Membrane integrity
Energy sources
3
Serum
Serum is required for the growth of most diploid cells in culture, and
established cell-lines are usually cultured in a basal nutrient medium
supplemented with serum. It provides hormones, growth factors, nutrients and
binding proteins, most of which are needed for the survival and proliferation
of many cell- lines. In the absence of serum, cells become arrested in the
G0/G1 phase of the cell cycle. Upon re-addition of serum, DNA synthesis and
cell growth are stimulated (Yoshikura et al, 1967) and the final cell density will
depend on the serum concentration used (Holley and Kieman, 1971). The most
commonly used serum supplements are commercially prepared foetal bovine
serum, new bom calf serum or horse serum, at concentrations of 0.5-30%.
Serum is a complex mixture of diverse factors, many of which can have
multiple effects; being growth stimulating for some cell types and growth
restrictive for others (Chiang et al, 1984). Furthermore, these factors and the
pathways through which they function can act synergistically or antagonistically
in a cell-type dependent manner, to influence control of differentiation or
proliferation. There are many disadvantages associated with the use of serum
supplementation (Table 1.2) (Glassy et al, 1988).
Table 1.2 Problems Associated With Use of Serum Supplementation
in Cell Culture Medium
1. Variability in composition (important in GMP processes)
2. High cost and periodic scarcity
3. Source of contamination (mycoplasma, bacteriophage)
4. Source of toxins
5. Modification of cell surfaces by adsorption/incorporation of serum proteins
6 . Prevents definition of nutritional environment
7. Interference in product purification
8 . Complicates interpretation of experiments.
4
Investigating the effects of growth factors on proliferation and
differentiation of specific cell types, and on dissection of signal transduction
pathways involved in this regulation further illustrates the increasing necessity
to control of the culture environment more precisely. Towards this many
laboratories have developed serum-free media that support proliferation of various
cell types.
Serum-Free Media
Most serum-free media contain many of the ingredients of serum-
supplemented medium. Serum is substituted by combinations of trace elements,
lipids, hormones, growth factors, and purified proteins. Many commercially
available serum-free media are only partially defined for two reasons:-
1. for commercial purposes (patents)
2 . because the supplements are not completely defined
eg transferrin and albumin.
Most serum-free media are applicable to only a few cell-lines due to
differences in nutritional requirements. It is becoming increasingly important,
especially from an industrial perspective, to have a completely defined medium
which can be optimised for each particular cell line; preferably protein-free, but
this is not possible for many cell lines. It is possible for serum-dependent cell
lines to be weaned into serum-free medium, but this process can take from
several days to several months before "normal" growth is achieved. This
variability in the time taken to achieve serum-free adapted cells makes it very
diffic ult to adhere to good manufacturing practice (GMP). Which in turn
complicates the licensing of products, and obtaining FDA approval.
The advantages of serum-free media are outlined in Table 1.3.
5
Table 1.3 Advantages and Disadvantages of Serum-Free Media for
Hybridomas
ADVANTAGES
1. Consistent and chemically defined composition
2. Improves reproducibility of cell culture growth and product yield
3. Decreases potential of contamination
4. Reduces difficulty and cost of product purification
5. May provide cost savings with a low protein serum-free formulation
6 . In research context, enables elucidation of nutrient exchange profiles.
DISADVANTAGES
1. Requires optimisation for each hybridoma cell line
2. Applicable serum-free media have not been developed for all cell lines
3. May require serum hormones and growth factors which are difficult to
isolate or purify
4. Frequently results in longer lag phase
5. Cell growth rate, maximum cell density and cell viability are often lower
6 . Protease inhibitors (present in serum) not present to help protect cells
from enzymes such as trypsin
1.1.3 Production and Application of Animal Cell Products
Original tissue culture work focused on the growth of tissues and the
study of their physiology. Enders (1949) showed that poliomyelitis virus could
be grown in primary cell culture. This not only developed the use of cell
culture but also facilitated developments in virology. Large scale use of cell
culture techniques has since provided a means of viral vaccine production. The
6
importance of mammalian cell culture increased with studies of cellular and
structural function of cells in tissue culture. These studies led to the production
of an array of medical, veterinary and agricultural products which would not
be possible by other means. Major developments facilitated this production, such
as the advent of monoclonal antibody (Mab) production from hybridomas,
constitutive expression of useful therapeutic proteins from continuous cell lines,
and improvements in culture product purification (Kluft et al, 1983; Lazar et
al, 1987; Phillips et al, 1985). The potential for production of proteins from
mammalian cells has been greatly increased by development of techniques for
expression of recombinant proteins in mammalian cell lines (Bebbington and
Hentschel, 1985; Lubiniecki, 1987; Ringold et al, 1981), indeed many more
products are presently undergoing registration and development (Galante, 1989).
Examples of the products from animal cell culture are presented in Table 1.4.
Table 1.4 Examples of Products From Animal Cell Cultures.
1 VIRAL VACCINES
2 RECOMBINANT VIRAL ANTIGENS
3 IMMUNOREGULTAORY PROTEINS
4 MONOCLONAL ANTIBODIES
5 POLYPEPTIDE GROWTH FACTORS
6 ENZYMES
7 HORMONES
8 VIRAL INSECTICIDES
9 TUMOUR SPECIFIC ANTIGENS
10 ANIMAL CELLS AS PRODUCTS
The market for Mabs, cytokines, enzymes, immunoprophylactics and
hormones derived from animal cells is set to increase considerably (Spier, 1990).
There is considerable competition between animal cells, engineered
7
prokaryotes and lower eukaryotes, as production systems for recombinant
proteins. Engineered bacteria and yeasts do not carry out post-translational
modification of animal proteins in the same way as animal cells. Such proteins
can be unfolded and modified synthetically, but the cost advantage of using
bacteria is lost by the necessity to use large quantities of chemicals (Cartwright
and Crespo, 1990). These chemicals may pose a major disposal problem. A
potential competitor to animal cells, as a method of recombinant protein
production may come from transgenic animals. Foreign genes expressed in these
animals leads to secretion of recombinant proteins into their milk or blood
stream. The first large transgenic animals to achieve the status of ’quasi
bioreactor’ were sheep. These transgenics secrete human factor IX (a-l-
antitrypsin) into their milk. Although expression levels are low, factor IX does
seem to be active, and the trait heritable (Van Brunt, 1988). However there are
many hurdles to negotiate before transgenic animals could replace animal cell
cultures for commercial large-scale recombinant protein production (Spier, 1990).
Indeed, the commercial success of current products from mammalian cells in
culture, and the promise of future developments suggests that animal cells will
be the recombinant protein production method of choice for the foreseeable
future.
1.2 HYBRIDOMA TECHNOLOGY
To facilitate an understanding of the problems and processes associated
with production, use and potential (economic and clinical) of monoclonal
antibodies (Mab), comprehension of the basic structure and function of the
humoral branch of the immune system is useful. A brief account only will be
presented as there are many very detailed texts (Male et al, 1989; Roitt, 1988).
8
1.2.1 The Humoral Immune Response
Substantial knowledge of the functioning of the mammalian immune
system has developed over the past century. The first evidence of the ability
of the body to defend itself against microbial invasion came in the nineteenth
and early twentieth centuries (McCullough and Spier, 1990). During the past
twenty to thirty years this knowledge has expanded to give a detailed
biochemical and immunochemical insight into the structure and function of the
antibody molecule, the division of the immune system response to foreign
material (antigen) into acquired and innate mechanisms, and the genetic control
of antibody diversity.
Antibodies or (Immunoglobulins) are secreted by plasma cells
(differentiated B lymphocytes) in response to the presence of an antigen (Ag).
The specific affinity of an antibody (Ab) may not be for the entire
macromolecular Ag but for a particular site on it called the antigenic
determinant or epitope. Most small foreign molecules do not stimulate antibody
formation alone but are able to, if attached to macromolecules such as proteins,
polysaccharides or nucleic acids. The macromolecule is then the carrier of the
attached chemical group, which is called the haptenic determinant.
The Antibody Molecule
The basic structure of an antibody (immunoglobulin) was revealed by
Porter in the 1960’s using enzymatic degradation of globulin fraction isolated
from serum.
The antibody molecule is made up of two identical heavy and two
identical light chains held together by interchain disulphide bonds. The exposed
hinge region in some Ig’s is extended in structure due to high proline content
and is therefore vulnerable to proteolytic attack. The molecule can be split by
papain to yield two identical fragments, each with a single combining site for
antigen (Fab; fragment binding antigen), and a third fragment which lacks ability
9
to bind antigen and is termed Fc (fragment crystallisable). Pepsin acts at a
different point, and cleaves the Fc moiety from the remainder of the molecule
to leave a large fragment which is formulated as F(ab^ since it is still divalent
with respect to antigen recognition sites (paratopes) like the parent molecule
(Figure 1.1).
\ HEAVY, . C H A IN
Vh
'hi
papain
•pepsin
Ch.
Fc
Figure 1.1 Basic structure of an immunoglobulin molecule, showing the
sites of cleavage with papain and pepsin into F(ab)2/Fc and
Fab/Fc regions respectively. Also shown are the variable (V)
and constant (C) domains of the light (1) and heavy (h) chains.
10
Studies of myeloma proteins allowed further elucidation of
immunoglobulin structure. Sequencing a number of myeloma proteins revealed
that N-terminal portions of both heavy and light chains shows considerable
variation, whereas the remaining parts of the chain are relatively constant. These
regions are referred to as variable (V) and constant (C) (Figure 1.1). Within
V domains are hypervariable regions within which the major idiotopes of the
antibody molecule reside. The idiotopes, which are most important in Ab-Ag,
reactions are located within the paratope, or combining site. It is the uniqueness
of the idiotopes which determines the uniqueness of each antibody molecule: the
idiotype.
Based on the structure of the heavy chain constant regions (C„),
immunoglobulins can be classified into classes or isotypes, for example IgG,
IgA, IgM, IgD, IgE .
The idiotype differentiates each antibody molecule, whereas an isotype
differentiates an antibody class (and sub-class). Antibody structure is relevant to
hybridoma technology and monoclonal antibody production since it is under
genetic control.
In any one antibody molecule, only one of two types of light chain is
found; either K or L, coded for by the Vk, V„ Ck, and C, genes (Roitt, 1988).
Isotypic determinants for the class of antibody are found on CH and are coded
by genes Qz, Cô, Ct, Ce or Co? (McCullough and Spier, 1990). The genes
responsible for these C region sequences reside in specific locations on the
chromosome, such that gene rearrangement will result in defined class switching
(McCullough and Spier, 1990). One gene from each of the two chromosomes
is spliced to give the synthesis of one Ig. The arrangement of isotype gene loci
and which class switches occur are shown in Figure 1.2. Only these switches
occur, because they result from gene deletions and hence cannot be reversed.
The deletions (switching) occurs only in one of the chromosomes.
11
VHi — VH < 300 Di—►0<2o Ji — c5 Cy3 Cy1 Cy2b Cy2a C£ C0_ q . _ — Q - -Q -— -Q— -CHHH] C Z H Z Z H Z Z IH Z Z h D H Z Z H Z H Z
gene reassortmentIone one one
J c c cVh 0 6 S3 |~| S i
transcription, inter-gene splicing
translation
gene splicing
translation
gene splicing
IgM
=Zh igE
Permissible isotype switches:
IgM — -► IgD; igG3; IgG,; ig G *
IgD — ■— igG3; igG,; ig G * ; ig G *
lgG3 igG1( igG2ü; igG2a; igE;
igGi igG2b. igG *; igE; IgA
ig G a — IgG *; IgE; IgA
igG2a - — IgE; IgA
IgE — — IgA
lgG2a; IgE; IgA
IgE; IgA
IgA:
Figure 1.2 Gene arrangements of the regions of the mouse
immunoglobulin heavy chain. The figure shows how
through re-assortment and splicing, the
immunoglobulin isotype is selected, and which
switching is permissible (From Hood et al, 1984).
12
The Antibody Response
B-lymphocytes develop in bone marrow and foetal liver, and have the
genetic potential to produce Ab to an antigenic determinant.
Antibodies are synthesised according to the clonal selection theory
(Burnett, 1957) and the network theory (Jeme, 1974). Each B-lymphocyte is
programmed to make only one antibody which it places on its outer surface to
act as a receptor. When an Ag enters the body it is confronted with an array
of lymphocytes all bearing different antibodies, each with its own individual
recognition site. Ag will only bind to those receptors with which it makes a
good fit.
The majority of B-cells are quiescent, that is they are in GO phase of
the cell cycle. Activation of B-cells involves a shift into G1 phase, prior to
DNA synthesis (S phase) and then proceeds through G2 phase towards cell
division. Three phases of B-cell activation can be distinguished, each requiring
separate signals:-
i) a signal which moves the cell from GO to G1
ii) a second signal which initiates DNA synthesis. Once this is complete the
cell is normally committed to division and
iii) a group of factors causing the proliferating cells to mature into antibody
forming cells.
When Ag binds, which cross-links the immunoglobulin receptors, the
receptors migrate to a small area or pole on the cell surface (’capping’). This
initial signal supplied by Ag moves the cell from GO to Gl. The Ag-stimulated
(capped) B-lymphocyte, will, depending on conditions of stimulation, follow one
of two differentiation pathways which results in the production of either memory
cells (morphologically indistinguishable from the original B-cell), or major Ab-
secreting cells (plasma cells). Since lymphocytes are programmed to make only
one antibody, that secreted by the plasma cell will be identical with the original
13
lymphocyte receptor. The constant region of the antibody may alter during
differentiation of the lymphocyte clone, but the variable region retains its
singular specificity. The origin of this specificity is complex and involves not
only extensive rearrangement of DNA sequences in the chromosome which codes
for the Ab chain but also some degree of somatic mutation during development
(Tonegawa, 1983).
To prevent simultaneous synthesis of too many different antibody
molecules, lymphocytes which are triggered by contact with Ag undergo
successive waves of proliferation to build up a large clone of plasma cells
producing Ab. Clones tend to persist after the disappearance of Ag and retain
the capacity to be stimulated if the Ag reappears. This provides for
immunological memory and quick response to infection.
1.2.2 B-Cell Hvbridomas
History
The technique of cell fusion, or hybridisation has many practical
applications. It has proved to be a major tool in determining which human
chromosome codes for specific gene function by means of the analysis of
isoenzyme patterns produced by a panel of clones from human-rodent fusions
(Ruddle and Kuchelaral, 1974). In addition it has been used to characterise the
dominant and recessive nature of malignancy by means of fusions of normal
cells with malignant cells (Croce et al, 1974).
The first hybridoma production was reported by Sikovirs et al (1970) in
which virus-specific lymphocytes were fused with tumour cells. Several other
reports were subsequently published (Bloom and Nakamura, 1974; Schwaber
and Cohen, 1974).
Kohler and Milstein (1975) developed somatic cell fusion techniques to
produce continuous cultures of specific antibody-producing cells. They fused
splenic B-cells which secrete antibody and have a finite life span, with immortal
14
myeloma cells (tumorous B-cells). The progeny hybrid-myeloma, or "hybridoma'',
has the unlimited proliferative capacity of one parent together with the specific
antibody producing capacity of the other. The technique they developed has
become widely used. Only a brief outline will be given since the literature
provides a plethora of detailed descriptions (Epstein and Epstein, 1986; Coding,
19831
The Technique
An animal (usually a mouse) is immunised with the Ag of interest,
although in vitro immunisation techniques are also used (James and Bell, 1987).
Spleen cells are removed and gently disrupted to form a cell suspension.
In the body, the major Ab secreting cells are plasma cells (1.2.1),
however plasma cells are of little use in the generation of hybridomas. The
plasma cell is too differentiated, and its genes which code for proliferation
repressed, and thus would probably form unstable hybridomas when fused with
myelomas (McCullough and Spier, 1990). Precursors of the plasma cell which
can also secrete antibody (at a lower level) would appear to be a better fusion
partner for production of stable Ab secreting hybridoma cells. These plasma cell
precursors are committed to Ab production, unlike the resting B-lymphocyte
parent in which much of the Ab synthesising genes are under repression. Stahli
et al (1980) have shown that blast-cells, two to three times the size of the
resting B-lymphocyte, are the most efficient at yielding Ab-secreting hybridomas.
These Ab-producing cells are fused with a cultured myeloma cell line.
The original myeloma used by Kohler and Milstein (1975), MOPC-21 secreted
IgGi, with kappa light chains. This meant that mixed molecules of IgG
containing light and heavy chains from both MOPC-21 and spleen cells were
found.
Since Kohler and Milstein’s work several myeloma cell lines have been
produced which do not secrete antibody. These cell-lines include SP2/0-Agl4
(Shulman et al, 1978), X63-Ag8-653 (Kemey et al, 1979), and NS0/1 (Galfre
15
et al, 1981). Of these the X63-Ag8-653 is probably the cell line of choice
because it is widely available, has a high fusion frequency and is easy to grow.
The fusion of myeloma and B-cell takes place in the presence of a
"fusogen", a substance which increases cell membrane permeability and promotes
membrane fusion. Kohler and Milstein used Sendai virus as a fusogen, but this
has largely been replaced by polyethylene glycol (Davidson et al, 1976, Galfre
et al, 1977). More recently electrofusion has been used (Zimmerman, 1982), but
there is doubt whether any real advantage, such as increased fusion efficiency,
is obtained (Ah-Sing, 1990).
The result of the fusion is a mixture of parent cells; hybrids of each
parent to itself, and hybrids between one parent and the other. It is therefore
necessary to transfer and grow the cells in conditions that allow only the B-
cell/myeloma hybrid to grow (Figure 1.3).M O US E IMMUNISED
G U U T U B E O M Y E L O M A
H Y Bm O O M A S SE LECTED
WITH HAT MEDIUM
ASSAY FOB AIUTIBOASSAY FOB A N TIBO DYICLONE BBOOUCEBS
Figure 1.3 The Production of Murine Hybridomas From a B-Cell/
Myeloma Fusion
16
The cells are transferred to HAT medium (Littlefield, 1964) which is
supplemented with hypoxanthine, aminopterin and thymidine. Aminopterin is a
folate antagonist which blocks de novo synthesis of purines and pyrimidines.
Cells which possess the enzyme hypoxanthine phosphoribosyl transferase (HPRT)
can synthesize purine and pyrimidines by a salvage pathway which utilises
hypoxanthine and thymidine, and therefore bypasses aminopterin inhibition.
Myeloma cells are selected so that they are HPRT deficient, and therefore
die if de novo purine and pyrimidine synthesis is blocked by aminopterin.
Spleen cells can use the salvage pathway, but are mortal and can only survive
in culture for a few days. Therefore, under this selection pressure, only
successful B-cell/myeloma fusions (hybridomas) will survive.
Hybridoma cells can be screened for Ab secretion by radioimmunoassay
(RIA) or enzyme linked immunosorbent assay (ELISA). Any positive cells can
be cloned by means of limiting dilution such that one cell is present in each
well of a microtitre plate. These single cells usually require the addition of
macrophage feeder layers to provide growth factors for cellular division.
Selective cloning of hybridoma cells for enhanced immunoglobulin production can
also be performed using flow cytometric cell sorting and automated laser
nephelometry (Marder et al, 1990).
1.2.3 Applications of Monoclonal Antibodies
One of the most striking features of the immune response is the very
wide range of specificities that can be elicited by different Ags. This has not
only made it a fascinating biological system to study, but has also allowed the
development of reagents with precise specificities for a wide range of
applications in biology and medicine. The diversity of the response means that
even small amounts of impurities often produce unwanted antibodies. These are
difficult to remove and cause problems, for example when the anti-serum is
used with complex Ag mixtures. The techniques for making Mab developed by
Kohler and Milstein allows the preparation of pure Ab from impure Ag
17
mixtures. This has facilitated serological analysis of many biological systems,
previously inaccessible due to the extreme difficulty of preparing pure Ag in
sufficient quantities.
Mab have therefore found a wide range of uses in veterinary and clinical
fields, including diagnosis, tumour imaging, therapeutics, immunopurification, and
basic research.
Despite antibiotic therapy and other advances in patient management,
infection still remains a major cause of human morbidity and mortality. The
adaptive ability of microorganisms to develop resistance to antibiotics, to
undergo antigenic change and to exhibit unexplained changes in virulence, poses
challenging problems to those involved in management of bacterial and viral
infections. The potential of Mab technology in meeting these challenges has
resulted in devotion of considerable effort to development of Mab which can be
used in diagnosis and therapy of infectious diseases.
Diagnosis
Mab produced in mice are most commonly used in diagnosis. The
advantage of Mab over conventional sera is the availability of Mab of standard
titre for an indefinite period of time. This simplifies the direct comparison of
results between different laboratories. The speed and accuracy of diagnosis has
been increased because of the high specification of Mab. Thus Mab can be used
to assay common serum analytes such as growth hormone (Bundersen et al,
1980; Stuart et al, 1983), prolactin (Stuart et al, 1982), progesterone (Fantl et
al, 1982) and testosterone (Kohen et al, 1982). Other serum proteins which can
be assayed include interferon (Secher, 1981), alkaline phosphatase (Mayer et al,
1982), and insulin (Storch et al, 1987).
A very large number of Mab have been produced to a wide range of
viruses such as influenza (Gerhard et al, 1981), hepatitus A (Coulepis et al,
1985), hepatitus B (Fujita et al, 1986), polio (Ferguson et al, 1982), Epstein
Barr virus (Hoffman et al, 1986) and rabies (Wiktor and Koprowski, 1979).
18
The high specificity of Mab has led to accurate differentiation between similar
strains of virus such as Herpes simplex types I and II (Pereira et al, 1980;
and development of antigenic fingerprinting for diagnosis of foot and mouth
disease virus (McCullough and Spier, 1990). Mab have recently been described
which exhibit in vitro neutralisation of HIV-1 Ag (Fung et al, 1987; Matsushita
et al, 1988; Skinner et al, 1988; Thomas et al, 1988). This may have obvious
important implications in the diagnosis and therapy of HIV infected patients. In
addition to serving as probes for viral Ag, Mab can be used to purify viruses
or viral Ag for use as diagnostic reagents or as viral vaccines.
In bacteriology uses for Mab have centred on:
i) identification of specific Ag’s and grouping, typing and sub-typing for
epidemiological, taxonomic and evolutionary studies.
ii) elucidation of an antigenic role in virulence and other aspects of
antigenicity.
iii) studies of the relationship of Ag to transport of material into and out of
the bacterial cell and other biochemical functions.
iv) diagnostic and therapeutic aspects of bacterial infection.
Mab have been used for the diagnosis of Streptococcal infections (Pollin,
1980), leprosy and gonorrhoea (Buchanan, 1984). In addition Mab to bacterial
toxins have been generated (Kozbor, 1982; Remmers, 1982), but only anti
"gram-negative endotoxins" are in clinical trial (James, 1990).
The considerable potential of Mab in the study of parasitic diseases has
already been widely exploited in those most common, such as malaria (Yoshida
et al, 1980), Schistosomiasis (Taylor and Butterworth, 1982), Leishmania
(McMahaon-Pratt, 1981) and Chargas’s disease (Sher et al, 1982).
The advantages are, not only the obvious diagnostic ones but also, that
the study of the disease itself may be undertaken in greater depth, since many
different surface Ag’s may be expressed at varying stages of the life cycles of
the parasite (Rowe, 1980).
Mab can also be used in tissue typing (Broski et al, 1979).
19
Tumour Imaging
Radiolabelled polyclonal antibodies have been used to locate human
tumours by immunoscintigraphy. Polyclonal antibodies have inherent batch to
batch variation and can recognise several epitopes on complex molecules. Some
of these epitopes are likely to be shared with other related molecules found in
normal tissues. The defined specificity of Mab may provide the precision
required for effective tumour imaging.
Labelled mouse and rat Mab have been used for to locate tumours in
patients with breast cancer (Rainsbury et al, 1983), colorectal cancer (Farrands,
1982) and ovarian cancer (Eperietos, 1982). Mab can be used not only for
detection of primary tumours, but also for monitoring and management of the
disease and therapy. These techniques for detection of primary and secondary
tumour growth can be applied in the study of other disease states where
abnormal tissue or serum components are a characteristic feature. Examples of
this include cardiology (Haber et al, 1982), and autoimmune disease
(Schoenfield, 1983).
Therapeutics
Most of the Mab with therapeutic application have been raised in mice.
Murine Mab have the disadvantage of being foreign to the human immune
system and likely to lose efficacy on continual application as a host response
is mounted. This host response itself may be harmful to the patient leading to
"serum-sickness". Human Mab’s have obvious advantages over murine Mab’s
but the huge potential will not be realised until the considerable problems in
production are overcome. These problems include poor secretory capacity and
relative instability of immortalised human B cells compared with murine
hybridomas.
20
However murine hybridomas have been shown to be of value in the area
of bone marrow allogenic or autologous transplants. Mab can be used to reduce
graft-versus-host disease (Prentice et al, 1984), or to remove leukaemic cells
from autografts (Kemshead et al, 1986). There are other means by which Mab
can be applied in the treatment of neoplastic disease. These include passive
therapy, either cytotoxic or regulatory in nature; and use of cytotoxic Mab
immunoconjugates (including conjugates) to radioisotopes, chemotherapeutic
agents, and natural toxins (Hillman and Royston, 1984). There are a number
of additional therapeutic approaches under development, which include use of
bispecific Mab to selective concentrated drugs and effector cells at tumour sites
(James, 1990), and antibody directed enzyme therapy (Bagshawe, 1987). The
success of Mab in therapy thus far has been limited, but will undoubtedly
increase with production of human Mab.
Another area in which Mab therapy may prove increasingly useful is in
the management of drug overdoses and poisoning whether inadvertent or
intentional. Availability of Mab to paraquat, individual snake venom, insect
toxins or mycotoxins may also prove clinically useful. Such antibodies could
either be administered intravenously or, the poison removed by extracorporeal
circulation employing a solid-phase column to which the Mab had been linked
(James, 1990).
Immunopurification
In principle it should be possible to immobilise any Mab on an affinity
column and use it obtain large quantities of the required Ag from crude
mixture. The Mab needs to survive repeated cycles of binding, and acid (or
base) treatment to allow elution, and re-use of a given column. High affinity
antibodies obtained after hyperimmunisation regimes, cannot be used since they
frequently bind Ag with such firmness that the complexes cannot be dissociated.
Mabs have found wide application in immunopurification of clinically
important molecules such as tissue plasminogen activator, factor VIII, factor IX,
21
human growth hormone, follicle stimulating hormone, and various interferons
(Jack et al, 1987).
There are two main areas in which immunopurification can be improved
and expanded :
i) preparation of high capacity antibody columns
ii) development of antibody-based separation technique for handling large cell
numbers.
Basic Research
Mab can in principle resolve a single protein from a complex mixture,
indeed a single epitope responsible for a specific function of a complex
macromolecule. They have already been widely used in basic enzymology, in
nucleic acid structural studies, and in analysis of hormone receptors. Mab may
prove particularly valuable in studying chromosomal proteins and more
specifically those responsible for determination of cell phenotype.
1.2.4 Monoclonal Antibody Production Systems
The impact of Mabs on the field of in vitro diagnosis is such that the
U.S. market of $59 million in 1983 (Reuveny and Lazar, 1989) is expected to
increase to $500 million by the end of 1990 (Karkare et al, 1985). It is
expected that Mab for use in affinity purification, in vivo imaging and
immunotherapy will be required in kilogram quantities (Birch et a/,1985; Karkare
et al, 1985). Indeed the total U.S. market for Mab’s is estimated to reach
$1.9 billion by late 1990 (Karkare et al, 1985). This huge economic potential
has resulted in considerable effort being made to develop efficient methods for
large scale production of Mabs.
There are two basic alternatives for large scale production of Mabs. One
approach is an in vivo method in which hybridoma cells are injected into the
22
peritoneal cavity of immunodeficient mice. Within ten to twenty days injected
cells form tumours which secrete Mab into ascites fluid. This fluid contains
Mab which can be collected periodically from the peritoneal cavity. Each mouse
produces 5-10ml of ascites fluid containing l-10mg/ml of Mab. The main
advantage is that small quantities of highly concentrated Ab can be made easily
without special equipment and technology. However the product is contaminated
with pyrogens and foreign mouse IgG. This makes purification and quality
control more difficult and expensive. Scale-up involves thousands of mice which
becomes labour-intensive and very expensive. The ethical considerations in
subjecting 10,000-40,000 mice to the discomfort of an ascites tumour for
production of 1 kilogram of Mab, become increasingly important.
In the second approach, cells are propagated in vitro in an artificial
medium, and Ab is harvested from culture supernatant. Whist the in vitro
method produces Mab at a lower concentration (10-200/tg/ml in batch cultures)
than in vivo method (Reuveny and Lazar, 1989), the product is pyrogen-free,
contains only small quantities of contaminating proteins, and can be produced
in large volumes. Using bioreactors also opens up a wide range of possibilities
for culture manipulations to increase efficiency.
Hybridomas can be propagated in homogenous culture in several modes:
batch/modified batch, continuous and perfusion.
Batch and Modified-Batch Culture
In simple batch culture operation temperature, pH and dissolved oxygen
tension (d02) are usually controlled, but nutrients (eg glucose and glutamine)
are depleted and inhibitory waste products (eg lactic acid and ammonia)
accumulate. Thus propagation slows and eventually stops, and cell yield is
limited (Reuveny and Lazar, 1989). Batch culture yields the lowest cell densities
in a given growth medium compared to perfusion systems. Specific Mab
production tends to vary during the course of the batch culture. Both cell
density and Mab production are affected by depletion of essential nutrients due
23
to consumption and accumulation of toxic metabolites (Reuveny and Lazar,
1989). The growth of hybridoma populations in batch culture can be studied in
a similar manner to bacterial growth. After inoculation of a small population
of hybridoma cells into a suitable medium, there is a lag phase, followed by
an exponential increase in population density after which growth ceases
(stationary phase). The cells then enter a death and lysis phase. A more detailed
description of the growth and production kinetics of hybridomas will be given
in chapter 3.
A more sophisticated approach to cultivating hybridomas is by fed-batch.
In this mode of cell propagation, cells usually reach a higher concentration and
viability is maintained for longer periods than in simple batch culture, resulting
in extended Mab production. In hybridoma fed-batch culture, growth is not
limited by nutrient depletion, but rather by waste product accumulation. Bacterial
cultures in contrast are ultimately limited by oxygen limitation, which has led
to the development of cyclic fed-batch culture.
Continuous Culture
Continuous culture differs from batch and modified batch in that it
provides an opportunity for maintaining steady-state culture conditions by a
continuous medium feed. The culture is generally started as a batch system
and medium feed is initiated during exponential phase of growth. Steady state
conditions, in which accumulation rate equals washout rate, results in a constant
biomass concentration.
In a chemostat, the medium flow rate is maintained so that the dilution
rate (D) is less than the maximum specific growth rate (in order to prevent cell
wash-out).
A continuous culture offers a number of advantages for large scale Mab
production. Once established, a continuous culture can be maintained at high cell
density and high Mab production for several months compared with a batch
culture which only lasts a few days (Birch et al, 1985). An important pre
24
requisite for successful continuous production systems is that the hybridoma cell-
line be stable in terms of cell growth and antibody secretion. The exponential
cultivation period that characterises continuous culture may exert a selective
pressure against cells which synthesise a product such as Mab, which is
basically non-essential and may be considered as an unnecessary metabolic
burden. One would expect a non-producing variant to appear in the culture and
become predominant. Another disadvantage of continuous culture is vulnerability
of the system to contamination after long periods of operation, requiring the use
of antibiotics, although this can be overcome by system design.
Perfusion Systems
In contrast to continuous culture, in which a steady-state is maintained
by continuous dilution of the culture, in a perfusion system the cells are
physically retained in the container, such that continuous fresh medium is added
while spent medium is removed, without dilution of the cell culture.
Unlike the chemostat, the cell concentration in this system constantly
increases until it becomes limited at high cell densities by contact inhibition.
Perfusion systems have several advantages; the cells are grown in relatively
small volumes of culture enabling high cell densities and concentrated product
to be achieved. In many systems, cells can be perfused with culture medium
in a vessel separate from the cell growth chamber, where coarse environmental
adjustments can be made (such as a high rate of sparging), before returning the
medium to the culture vessel. Other advantages reported are; improved
productivity (per cell), better utilisation of growth medium, lower concentration
of serum in growth medium, and reduced exposure of Mab to proteases
(Reuveny et ti/,1986).
The major problems encountered are:-
i) difficulties in operating special devices for continuous separation of cells
from growth medium during culture. This is true mainly of homogenous
systems in which filters, used for this purpose, tend to block.
25
ii) difficulties in supplying nutrients, especially oxygen to the cells, in high
density perfusion culture. This problem is more serious in non-
homogenous systems in which nutrient gradients are generated in the cell
growth chamber.
iii) the system is rather complex; thus there are increased chances for
contamination or mechanical failure.
1.2.5 Factors Affecting Mab Production
There are numerous factors which have an effect on the production of
Mab in vitro. These include Mab production kinetics, growth medium,
environmental factors, stability of Mab, and shear sensitivity of cells.
Kinetics of Mab Production
Several types of production kinetics for Mab production by hybridomas
in batch culture have been described (Merten et al, 1985). A description of
these kinetics, their significance and application will be discussed in chapter 3.
Growth Medium
The nutritional requirements of animal cells in culture is described in
1. 1.2 .
Generally the composition of the growth medium does not affect the rate
of Mab production directly but rather indirectly by affecting cell growth
(Tharakan et al, 1986). However, there are several publications showing a direct
effect of growth medium composition on Mab production rate (Cleveland, 1983;
Lebhertz, 1987; Roberts, 1976; Rupp, 1985).
A modified medium for Mab production containing a 2-fold concentration
of amino acids was used by Rupp (1985). He found that cell density was not
affected by this medium. However Mab production increased from 200/xg/ml to
26
about 1,100 jitg/ml, indicating that some amino acids limit production.
Environmental Factors
Environmental factors such as pH, d02, and temperature drastically affect
hybridoma cell growth. The optimal range for these parameters are:-
In general those levels which are optimal for growth are also optimal for
Mab production (Boraston, 1984; Emery et al, 1987; Fleishaker, 1986; Rudge,
1987; Wergeland, 1987). However in several instances, it has been found that
optimal concentrations for cell growth are not always necessarily optimal for
Mab production. Reuveny et al (1986) found that the highest cell density of a
certain hybridoma propagated in a stirred reactor is obtained at d02 equivalent
to 60% air saturation, while the highest titre of Mab was obtained at 25% air
saturation. A similar phenomenon was reported with other hybridomas propagated
in a split-flow reactor (Phillips et al, 1987).
It was reported that pH in the range of 6.8 to 7.2 did not affect Mab
production (Reuveny et al, 1986), while other work (White et al, 1987) showed
that 7.3 was optimal for Mab production. Miller et al (1986) showed that
specific Mab production is higher at pH 6.8 and 7.7 than from 7.1 to 7.4.
Merten et al (1985) found that cells maintained at a temperature below
37°C led to increased genetic instability of human hybridomas, but did not affect
the rate of Mab production. At 34°C however, Mab production is drastically
decreased compared to 37°C (Reuveny et al, 1986).
pH
d02
Temperature
6.9 - 7.5
40%
37°C
27
Antibody Stability
There is some evidence that over a period of days, Mab produced by
hybridomas in stirred fermenters, undergoes degradation; probably as a result of
proteolytic enzymes released from dead cells, or present in some of the culture
medium components (Jeme, 1963; Tharakan et al, 1986). Rudge et al (1987)
have shown that, in their system, acid protease is present in continuous culture
but not present in batch culture. Macmillan et al (1987) showed that IgG
produced by certain hybridoma cell lines may be degraded during large scale
production in bioreactors. The Ab concentration measured by ELISA during
production peaked at 4 days and then declined, following, in general, the cell
growth curve. If however, the Ab was assayed by radialimmunodiffusion (RID),
productivity appeared to rise continuously throughout the 8 day fermentation
period. The difference can be explained if fragmented Ab could not bind Ag
in the ELISA but could react with Ag in the less sensitive RID reaction, which
because of the nature of the reaction is not dependent on the integrity of single
reaction sites.
"Shear" Sensitivity
Compared to microbial cells, animal cells, (which lack rigid cell walls),
are more sensitive to, and can be damaged by, hydrodynamic forces. The
hydrodynamic sensitivity is often referred to as "shear" sensitivity.
Hybridoma cell lines vary in their sensitivity to shear forces, a
characteristic that becomes an important factor when cells are grown in agitated
cultures for Ab production. Dodge and Hu (1986) tested the effect of
mechanical agitation on the growth and Mab production of a particular cell-line.
They reported that the growth of a specific hybridoma was reduced at agitation
rates of 240rpm indicating hybridomas are only sensitive to excessive mechanical
agitation. Bâcher et al (1986) tested cell damage in stirred reactors with marine
impellors and found no noticeable effect on cell viability at 170rpm. Fazekas
28
de St Groth (1983) found that in serum-free medium impellor speed had to be
kept below 60rpm to prevent damage to hybridoma cells.
The sensitivity of hybridomas to shear and the classification of cells as
"fragile" or "shear sensitive" is contested as a myth by some workers (Handa-
Conigan, 1990). Work by Handa-Corrigan et al (1989) has shown that the
damage apparently caused to cells by sparging is due to formation of unstable
foams and the disengagement of bubbles from the surface of the culture. Two
mechanisms of cell damage have been proposed:
i) damage due to rapid oscillations caused by rapidly bursting bubbles (eg
in media containing antifoam)
ii) damage due to entrainment of cells in draining liquid film as found in
unstable foams (eg media with low serum or protein concentration).
They proposed that this damage could be minimised by addition of
surface active agents such as pluronic F-68 or high serum and protein
concentrations. This may act as a cell protecting agent by reducing film
drainage, and bubbles bursting in the vicinity of the cells.
Recent work by Al-Rubeai et al (1990), has shown that DNA synthesis
is inhibited under intense dynamic stress, while cellular metabolic activity
increases. This increase is thought to feed repair mechanism; indeed sub-optimal
temperature and nutrient deprivation may inhibit these mechanisms, causing
increased susceptibility to hydrodynamic stress.
1.3 THE AIMS OF THE WORK
The economic and clinical importance and potential of Mab production
by hybridomas have been illustrated. The systems employed to produce them are
becoming increasingly complex through attempts at process intensification.
Bioreactors have been increasing in size with concomitant increases in production
and costs.
There is obviously a limit to the size of production reactors, so
alternatives must be explored.
29
One way of increasing production with a corresponding increase in
productivity and profit, is to modify the cell to produce more Mab. In order
to do this there must be a greater understanding of basic physiological processes
involved in cell growth and product formation. This would enable key metabolic
pathways to be targeted and also possibly key enzymes, which may then be
amplified or inducibly regulated.
Considering the commercial and social importance of Mab’s, there has
been comparatively little work in hybridoma physiology. The aim of this work
therefore, is to investigate and develop an understanding of physiological
processes underlying hybridoma growth and Mab production. This will hopefully
facilitate improvement of large scale Mab production processes.
30
2 . MATERIAL AND METHODS
2.1 CELL CULTURE
2.1.1 The Cell-Line
The cell line used in this study was PQX Bl/2, a gift from ICI
Pharmaceuticals. It is a hybridoma derived from a polyethylene glycol fusion of
B ALB/C mouse spleen cells with X63-Ag-8.65.3 myeloma. The B ALB/C mouse
was immunized with paraquat conjugated to ovalbumin with Freund’s complete
adjuvant (used to stimulate cell-mediated immunity).
The myeloma used does not synthesise light or heavy chains and thus
the Mab secreted by the hybridoma is a true Mab.
Paraquat is a bipyiiylium herbicide produced by ICI. It is used in many
countries throughout the world and is safe when used properly. There have,
however, been several hundred fatalities attributed to paraquat poisoning. This
has been mainly due to ingestion of the concentrated commercial product for
suicidal purposes .
ICI have developed hybridoma cell-lines, which produce anti-paraquat
antibodies in an attempt to detoxify the herbicide’s effect in mammalian systems.
The idea came from published work showing successful Ab therapy in digoxin
poisoning (digoxin is a cardiac glycoside). Anti-paraquat antibodies have been
shown to bind and neutralise paraquat in vitro (Wright et al 1987), but not, as
yet, in vivo.
The PQX Bl/2 hybridoma was used in this study as a model system for
investigating physiological aspects of Mab production in hybridomas.
2.1.2 Sterilisation Procedures
All heat-stable solutions, and empty glassware were sterilised before use
for cell culture purposes by autoclaving at 121°C/15psi for a minimum of 15
31
minutes depending on the presence and volume of liquids. Heat-labile solutions
were filtered through 0.22jwm pore-size filters. Volumes of less than 50ml were
filtered, using disposable Flowpore-D 0.22^m (Flow Laboratories) filters attached
to syringes (Sabre). Volumes from 50-500ml were filtered using disposable filter
units (Flow laboratories). The basal medium for fermenters was filter-sterilised
using Sartobran Capsule 0.45pm, 0.2/zm filters (Sartorius).
The tissue culture flasks, pipettes, centrifuge tubes (upto 50ml) were
bought as sterile disposable units.
2.1.3 Cell Counting
Viability and cell concentration of cultures were determined by means of
the dye exclusion principle. This technique is based on the observation that the
surface membrane of living cells is able to selectively exclude certain substances
whereas dead cells allow them to enter freely, enabling viable and non-viable
cells to be distinguished. The dye used was Erythrocin B (Sigma) and was
kept as a stock solution of 0.04% v/v in 0.15M NaCl.
A 0.2ml aliquot of cells was diluted by addition of 0.4ml of Erythrocin
B solution. Cells were visualised microscopically on a modified Fuchs-Rosenthal
haemocytometer. Live cells have a transparent/blue appearance whereas dead cells
retain the dye and are red in colour. The total number of cells of each type
contained within the five squares indicated in Figure 2.1 were counted. Both the
grids on the haemocytometer were counted, and the average used. Viability and
viable cell number were determined using the following equations:
Vc = V x d x 103 Where: Vc = Viable cell number per
Tc = (V + NV) x d x 10
% VIABILITY = Vç x 100 Tc
ml of culture Tc = Total number of cells per
ml of culture V = Number of viable cells per
gridNV = Number of non-viable
cells per grid d = Dilution factor
32
,v Figure 2.1 A modified Fuch-Rosenthal Haemocytometer Slide, with
the squares used for counting shaded. This is an
example of one of several types of haemocytometer
currently used in animal cell culture
33
2.1.4 Dry Weight Analysis
Filter discs, 47mm in diameter, GF/C (Whatman), were dried by heating
in a microwave oven for 3 minutes at full power (600 watts). The filter discs
were removed using plastic forceps, placed in a dessicator to cool and then
weighed on a 5-figure balance. Individual filters were washed by immersion in
a solution of 1% Tween 80 (Sigma) in MQ, then carefully placed in the filter
holder (Millipore). Two mis of MQ water were added and a vacuum applied.
Each filter was washed twice with 2x1 Omis of MQ. For analysis, 30mls of cell
suspension was carefully added to the filter and a gentle vaccuum applied (to
prevent biomass being forced through the membrane or causing tears and hence
an underestimation of the weight). Filters were then washed three times by
addition of 3xl0mls of 0.9% w/v NaCl in MQ. The test filter was then
carefully removed, transferred to a petri-dish lined with filter paper (No 1,
Whatman),(to prevent the filter sticking to the petri-dish when dry), and dried
by heating in the microwave for two, three minute heating cycles on full
power. The filters were allowed to cool in a dessicator for 10 minutes and re
weighed. The analysis was performed in duplicate, and a control using sterile,
cell-free medium was employed. A 500ml bottle containing cold water was
included in each heating cycle to act as a heat sink. The dry weight was
determined as follows.
Dry Weight (mg/ml) = (Wf - WO - (Control)
Volume of Sample
Where Wf = Final weight of sample
Wi = Initial weight of sample
34
2.1.5 Culture Media
The cell-line was routinely cultured in the RPMI 1640 (Gibco BRL))
supplemented with 2mM L-glutamine (Gibco BRL), 0.15% sodium hydrogen
carbonate (sodium bicarbonate) and 10% heat-treated equine serum (Imperial
Laboratories) (Appendix 1A), unless otherwise stated. Bulk RPMI 1640 medium
for large scale work was prepared in 10 litre volumes from RPMI 1640 powder
supplied without glutamine or sodium bicarbonate (Gibco BRL). The RPMI
powder (101.2g) was added to 5 litres of MQ Water containing 4.5ml of
concentrated HC1, to facilitate the dissolution of the powder. Sodium bicarbonate
(15g) was added and the pH increased to 7.1 by addition of NaOH (5M). One
litre of phosphate buffer (50mM, pH 7.1, Appendix IB) was added, the volume
increased to 6.9 litres with MQ and the solution mixed thoroughly using a
magnetic stirring bar. This basal medium was then used for bioreactor
experiments (2.1.7).
2.1.6 Cell Passaging
Cells were sub-cultured (passaged) at late log-phase or every 2-3 days
depending on the growth and viability. The cell suspension was gently agitated
and the flask tapped lightly on the bench to mix the suspension, and dislodge
any cells which may have adhered to the bottom or sides of the flask. Viable
cell number and viability were determined as described (2.1.3). Cell suspension
was pipetted into a sterile conical bottomed centrifuge tube (Falcon Plastics) and
centrifuged at 200g for 5 minutes. The supernatant was discarded and the cells
gently resuspended by aspirating, in a small volume of pre-warmed (37°C) fresh
medium. This suspension was then transferred to a tissue culture flask
containing an appropriate volume of pre-warmed medium, such that the cells
were at a final viable cell concentration of 2 x 107ml of medium. Flasks were
then incubated at 37°C in a 5% carbon dioxide, humidified atmosphere.
35
2.1.7 Large Scale Culture
In a number of experiments, 101 batch cultures were carried out in a
bioreactor. In each case, the experiments were conducted in a similar way.
Individual modifications are discussed in the relevant chapter.
An LH 2000 series 141 bioreactor, 101 working volume, (LH
Fermentation) was employed.
The culture vessel, including air filters, associated ’addition’ lines and
sampling device, was sterilized by autoclaving (30 minutes at 15 psi/121°C).
Basal medium (6.91, Appendix 1C) was filter-sterilised into the vessel and 11
of heat-treated equine serum (Imperial Laboratories), together with 100ml of L-
glutamine (200mM, Gibco BRL) added separately, aseptically.
An inoculum of 2 litres of cells of high percentage viability, at a viable
cell density of IxlOVml was introduced into the bioreactor. The culture was
continuously stirred using a pitched-blade impellor, mechanically driven at
200rpm. Temperature was maintained at 37°C by a thermocirculator. The culture
was oxygenated by sparging with air, controlled by a mass-flow meter (Hi-Tec)
to maintain the d02 at 40% air saturation, measured using an Ingold d02 probe.
pH was measured using an electrode (Ingold) and maintained at 7.1 by addition
of acid (1M HC1) or alkali (1M NaOH). pH stability was enhanced by the
presence of 5mM phosphate buffer in the medium. Oxygen and carbon dioxide
were measured on-line using paramagnetic and infra-red meters respectively.
Foaming was prevented by manual addition of a sterile 10% v/v solution of
Silicone Antifoam C (Sigma) in MQ.
The first sample was taken 1 hour after inoculation to ensure thorough
mixing of the culture. Samples (100ml) were then taken thereafter every 4-6
hours as required, and the viable, total-cell densities and viability determined.
In addition, samples of culture (containing cells), and cell-free supernatant, after
centrifugation at 400g for 5 minutes, were prepared and frozen at -20°C for
future biochemical analysis and determination of Mab titre.
36
2.1.8 Sterility Testing
Bacteria and Fungi
All cultures were checked for bacterial and fungal contamination on
revival and regularly during routine passaging. Media components and complete
media were also checked.
An aliquot of medium or cell supernatant was centrifuged at 1630g for
10 minutes. The pellet was resuspended in 1ml of supernatant and 0.1ml
volumes of this used to inoculate the following in duplicate: nutrient agar
(Unipath), Sabouraud’s dextrose agar (Unipath), tryptone soya broth (Unipath),
and thioglycollate broth (Unipath). One set was incubated at 37°C and the other
at 25°C, for 14 days. Sterility media were examined at intervals for any
bacterial or fungal growth.
All sterility media were prepared according to standard recipes and
manufacturers recommendations.
Mycoplasma
Periodically cultures were sent to the European Collection for Animal
Cell Culture (ECACC), at CAMR, Porton Down, to be checked for mycoplasma
contamination by Hoersct staining and specialist culture techniques.
2.1.8 Crvopreservation
Cell Freezing
Cells were harvested by centrifugation at 80g for 5 minutes. The pellet
was resuspended in freezing mix (Appendix ID) at 10 x 105 viable cells/ml.
This suspension was then aliquoted in 1ml volumes into sterile, labelled,
cryotubes (Nunc). The tubes were sealed, and placed in a polystyrene box at
-70°C for 12-18 hours. (This gives the required slow freezing rate of about
rc/minute). The ampoules were then stored in the vapour phase of a liquid
nitrogen vat (Union Carbide). After 48 hours one vial was revived to check
viability.
37
Cell Revival
A cryotube was carefully thawed as quickly as possible in a 37°C water
bath. The contents were pipetted into a sterile 30ml plastic universal (Sterilin),
and 10ml of pre-warmed growth medium (Appendix 1A) added dropwise. The
cells were harvested by centrifugation at 100g for 5 minutes, and resuspended
in lOmls of medium in a 25cm2 tissue culture (TC) flask (Flow). This was
incubated at 37°C in a humidified, 5% C02 atmosphere for 24 hours, after
which the cells were passaged at 3 x 105 cells/ml until the viability exceeded
80%. Cells were then passaged routinely as described previously (2.1.5).
2.2 BIOCHEMICAL METHODS
2.2.1 SDS Polyacrylamide Gel Electrophoresis (PAGE)
PAGE is a technique whereby proteins may be separated from each other
upon the basis of their molecular weight. Proteins are first dissociated into
peptide sub-units by sodium dodecyl sulphate (SDS), and take on a negative
charge due to SDS binding (Homes, 1981). All the polypeptides therefore have
the same charge density and thus migration through the gel and subsequent
separation will depend only upon size.
Polyacrylamide gel is the result of polymerization of acrylamide
monomers into long chains which are in turn cross-linked by N,N’ Methylene
bisacrylamide. This polymerisation is initiated by the formation of free radicals
from ammonium persulphate by N,N,N, ’N’-tetramethylenediamine (TEMED). The
polymerisation is inhibited by oxygen and so all solutions must be degassed
prior to use.
Polyacrylamide gel electrophoresis of purified Mab was carried out based
on the method of Laemmli (1970) (Appendix 2A).
38
2.2.2 Lowrv Protein Assay
In this assay, based on the method of Lowry et al (1951), protein reacts
with the Folin-Ciocalteau reagent to give a coloured complex. The colour, so
formed, is due to the reaction of alkaline copper with protein and reduction of
phosphomolybdate by tyrosine and tryptophan in the presence of protein. The
intensity of the colour depends on the amount of the aromatic amino acids
present and will thus vary for different proteins (Appendix 2B).
2.2.3 Affinity Purification of Anti-Paraquat Mab
The pH was adjusted to 2.5 by addition of HC1 (1M), and the volume
increased to 11 with MQ.
Cell-free supernatant (100ml) from a 3-day culture was circulated through
10ml of paraquat-AH-Sepharose (ICI) packed in a 20ml syringe (Sabre), at 0.5ml
min'1 at 4°C, over a period of approximately 2>l/i hours. The column was then
washed with PBS containing NaCl (1M) at a flow rate of 1ml min'1 until no
more protein could be detected in the eluent, and then washed quickly with PBS
containing NaCl at 0.15M. Finally the Mab was eluted with glycine buffer at
a flow rate of 0.5ml min'1. The protein concentration of each 0.5ml fraction
was estimated by determination of the optical density at 280nm. Fractions
containing significant concentrations of protein were pooled and the pH adjusted
to 7.0 with tris buffer. Benzamidine (Sigma), a protease inhibitor, and
Reagents
Glvcine/HCl buffer (0.5M, pH 2.5)
Glycine (BDH)
MQ
75.07g
to 800ml
Protocol
39
thimerosal (BDH) a bacteria inhibitor, were added to final concentrations of
lO^cM and 0.1% (w/v) respectively. The mouse IgG content of the pooled
fraction was assessed by ELISA (2.2.4), as below but using Sheep anti-mouse
IgG (Sigma) to coat the plates and Mouse IgG (Sigma) as the standard. The
protein concentration of the fraction was determined by the Lowry method
(2.2.2) and the purity by SDS-PAGE (2.2.1). Under non-reducing conditions a
single band at 150kD was observed, whilst under reducing conditions single
bands at 50kD and 25kD molecular weight were observed representing heavy
and light chains.
Anti-paraquat Mab was stored in 10jA aliqouts at -20°C.
2.2.4 Enzvme-Linked Immunosorbent Assay (ELISA)
Reagents
Phosphate Buffered Saline. PBS (xlO concentrate!
NaH2P04
Na2HP04
NaCl
MQ to 11
2.97g
11.5g
85.Og
Wash Buffer
PBS (X10)
Tween 20 (Sigma)
MQ to 11
1ml
100ml
40
Coupling Buffer
NaCl 0.9g
NaHC03 2.93g
NagCOg 1.06g
MQ to 11
Substrate Buffer
Glycine solution 0.5M pH 10.4 20.0ml
ZnCl2 (Sigma), 0.1M 1.0ml
MgCl2 0.1M 1.0ml
MQ to 100ml
Substrate
Alkaline Phosphatase (Sigma) 1 mg/ml in substrate buffer
Paraquat-BSA Antigen
Paraquat-BSA (Kent University) diluted to 5/zg/ml in coupling buffer.
IgG conjugated to alkaline phosphatase
Anti-mouse IgG conjugated to alkaline phosphatase (Sigma) diluted 1:1000-
4000 (Depending on batch) in wash buffer.
Murine Anti-Paraquat Antibody
Affinity purified Murine Anti-Paraquat antibody (2.2.3)
diluted to 1/xg/ml in wash buffer.
Protocol
Paraquat-BSA antigen (50 jul) was added to each well of a 96-well
microtitre plate (Nunc Immuno I), which was covered in aluminium foil and left
41
at 4°C overnight. The plate was first washed by immersing 3 times in wash
buffer,, patted dry and. then 50^1 of wash buffer was added to each well,.,
followed, by 50/d. of sample or standard added (in duplicate) to wells in column
’A’. A double dilution series was carried out across the plate (Figure 2.2), after
which the plate was incubated at 37°C with gentle agitation for 1 hour.
1 2 3 4 5 s % s s na 11 12
(§) O O o 0 o o o o o o oO o o 0 o o o o o o 0
© o o o 0 o o o o o o o© o o 0 o o o o o o o 0© o o o 0 o o o o o o o© o o O' o o o o o o o o© o o o 0 0 o o o o o o© o o' o 0 o o o o o o o
SERIAL D ILU TIO N i------------ >-
S-standard
T- samples0,2,3,etc
Figure 2.2 The dilution series for the determintion of of Mab
concentration of standard and sample by ELISA
The plate was then washed 4 times and dried as before. 50^1 of anti
mouse IgG conjugate was added to each well, the plate covered and incubated
as before.
42
Excess conjugate was removed by washing 5 times and drying. 50/d of
substrate was added to each well and the plate incubated for a further 30-60
minutes and then "read" at 410nm using a plate reader (Dynatech) with the
reference beam set at 630nm. A calibration curve was constructed from the
standard results and the sample Mab concentrations determined.
2.2.5 Glutamine and Ammonia Determination
In order to measure the concentration of glutamine in cell culture
supernatants it is necessary to convert the glutamine to glutamate and ammonia,
via glutaminase (EC 3.5.1.2) action. The pH (optimum for glutaminase is 4.9)
is increased to alkaline on completion of the reaction thus converting ammonium
ions into free ammonia which can be measured by an ammonia electrode. The
method is based on that of Thorpe (1989) (Appendix 2C).
2.2.6 Lactate Assay
Lactate is a waste product of metabolism in mammalian cell culture. The
classic method for determination of lactate was first applied to blood by Wolf
(1914), and subsequently assays have been developed involving titration and
photometric measurement. It was not until the advent of enzymatic assays that
the sensitivity and specificity could be greatly enhanced.
The assay protocol used routinely for the measurement of lactate was the
Sigma procedure No 826-UV (Sigma Chemical Co.), based on a technique first
used by Henry (1968) (Appendix 2D).
This method utilises the enzyme lactate dehydrogenase (LDH, EC
1.1.1.27) which reversibly catalyses the conversion of pyruvate to lactate, in the
presence of nicotinamide adenine dinucleotide (NAD+) and its reduced form
NADH. To measure lactate production excess NAD+ is required to convert
lactate to pyruvate.
43
Thus:-
NADH + PYRUVATE— LACTATE + NAD+
To force the reaction to completion the pyruvate formed is trapped by
the presence of hydrazine. The assay detects the increase in absorbance at
340nm due to the accumulation of NADH. This increased absorbance is directly
proportional to the concentration of lactate in the sample.
2.2.7 Glucose Determination
Glucose concentrations in cell-free supernatants were determined using a
Cobas-Bio glucose analyser (Roche) in conjunction with UNI-KIT III (Roche)
reagents.
The concentration of NADH formed according to the equations below,
is directly proportional to the glucose concentration and is determined by
measuring the absorbance at 340nm.
D-glucose + ATP — D-glucose-6-phosphate + ADP
D-glucose-6-phosphate + NAD+ - G61>DH»d-gIuconate-6-phosphate + NADH
+ H+
Protocol
Glucose standard solutions of SOOmgdl'1, 300mgdU and lOOmgdl"1 were
prepared and aliquoted into the reagent tray. The reaction vial was reconstituted
in 20ml of MQ and 5ml added to the reagent tray. The reagent tray was
inserted into the analyser and glucose concentrations determined. If the standard
concentrations were within 5% of their prepared concentration, the results were
recorded. Results outside this were rejected and the samples repeated.
44
2.2.8 Amino Acid Analysis
Introduction
The amino acid concentrations in culture supernatants were analysed using
the Waters PICO.TAG Amino Acid Analyser system. It is an automated gradient
liquid chromatography designed for analysis of amino acids.
The method involved 2 steps:
1. pre-column derivatisation of the sample
2. analysis by reverse phase high pressure liquid chromatography (HPLC).
In the first stage the samples are derivatised with phenylisothiocyanate
(PITC) to produce phenylcabamyl (PTC) amino acids.
R I
+ NH2— C H —COO*
Amino Acid
▼
S RI I
- NH— C — NH—CH— COO
PTC Amino Acid
In the second stage the amino acids are separated using a CIS column
and their elution detected at 440nm.
Reagents
Eluent A
Sodium Acetate trihydrate (BDH) 19g
MQ 1000ml
Triethylamine, T.E.A., (Aldrich) 500/d
— N = C = S
PITC
45
The pH of the solution was adjusted to 6.40 with glacial aceitic acid (BDH)
and filtered through a 0.22/xM Millex-GS filter (Millipore). To 940ml of filtered
solution 60ml of Acetonitrile (HPLC grade, FSA) was added.
Eluent B
Acetonitrile (HPLC grade, FSA) 600ml
MQ 400ml
The solution was degassed by sonicating under vacuum for 20 seconds.
Sample diluent
Na^HPCL (BDH) 710mg
MQ 1000ml
The pH was titrated to 7.4 with 10% phosphoric acid (BDH) and acetonitrile
added to give a final concentration of 5% by volume.
Redrying Mix
Methanol (FSA)
MQ
T.E.A. (Aldrich)
Derivatisation Reagent
Phenylisothiocyanate (Pierce)
MQ
T.E.A. (Aldrich)
Methanol (FSA)
Protocol
Derivatisation
25^1 of sample or standard (Food Hydrolysate Standard, Sigma) was
pipetted into a labelled glass culture tube (Bibby-Sterilin). A "blank" of 0.1M
50^1
50^1
50/d
350/d
100/d
100/d
50/d
46
HC1 was included. The samples were loaded into a vacuum vial and a vacuum
pulled gently. The samples were allowed to dry for 30-45 minutes. 10/d of
redrying mix was added to each tube. The solutions were vortexed to mix and
returned to the work station for drying (approximately 30 minutes).
20 /d of derivatisation solution was added to each redried sample, the
contents mixed by vortexing, and incubated at room temperature for 20 minutes.
The samples were dried as before (approximately 45 minutes) and resuspended
in 100/d of sample diluent. The samples were loaded into "WISP" sampling
vials and analysed.
Analysis
The samples were analysed using a "Nova-Pak" CIS (Waters) HPLC
column and the concentration of amino acids determined by comparison of peak
area and retention times with standards (Figure 2.3).
id.
G.
U _ J y v
15O
R E T E N T I O N T I M E CminJ
Figure 2.3 An example of a chromotogram for the determination of
amino acid concentration of a standard solution
47
Each peak of the chromatogram corresponds to an individual amino acid,
and has a specific retention time, which is used in the identification of the
peaks. The area under the peaks is proportional to the concnetration of that
amino acid. The retention times and peak areas of samples can therefore be
compared to standards and the cocnetrations of the amino acids determined.
2.2.9 Ornithine Carbamvl Transferase Assay (EC 2.1.3.3)
The method used is based on that of Ceriotti (1985) (Appenidx 2E). The
ornithine carbamyl transferase (OCT) is measured by determination of citrulline
formed per time unit according to the following equation:
Carbamoyl-P + Ornithine — Citrulline + Pi
The equilibrium of the reaction lies on the side of citrulline. Citrulline
is determined by the highly sensitive diacetylmonoxime-antipyrine reaction which
results in a chromophore, the optical density of which is measured at 460nm.
Interference by urea is eliminated by inclusion of urease in the reaction mix.
Each sample has its own "blank" to correct for citrulline already present.
2.2.10 Glutamate Dehydrogenase Assay (EC 1.4.1.3)
Glutamate dehydrogenase (GIDH) is measured by following the oxidation of
NADH according to the following equation:
2-Oxoglutarate + NADH + NH4+ —DH» L-Glutamate + NAD+ + H20
The equilibrium of the GIDH reaction is greatly in favour of glutamate
formation and thus measurements of catalytic activity are performed with 2 -
oxoglutarate and N H / as substrates, and NADH as a coenzyme. The decrease
in absorbance due to NADH oxidation is a measure of GIDH catalytic activity.
The method is based on that of Schmidt and Schmidt (1985) (Appendix 2F)
48
3. GROWTH AND MONOCLONAL ANTIBODY PRODUCTION
3.1 INTRODUCTION
3.1.1 Parameters of Growth
Population and Specific Growth Rates
The growth of animal cells in culture can be considered in a similar
way to bacterial growth. Indeed, the principles used to describe mammalian
cell culture systems are based upon microbial culture theory. Two sets of
factors influence the growth of cells in culture: intracellular factors and
extracellular factors. Intracellular factors include structure, metabolic mechanisms
and the genetic material of the cell, whilst the extracellular factors include
dissolved oxygen tension, pH, osmolality, temperature, efficiency of mixing,
shear forces and nutritional environment (1.1.2).
The pre-requisites for growth of biomass in a culture are:
i) a viable inoculum
ii) energy sources
iii) nutrients to provide essential material from which biomass can be
synthesised
iv) absence of growth inhibitors
v) suitable physiochemical conditions
If all these requirements are satisfied then in a batch culture, during an
infinitesimally small time interval (dt), the increase in biomass (dx), will be
proportional to the amount (x) present initially and to the time interval.
dx/dt = /ix (1)
dx = /tx.dt (2)
49
The differential expression dx/dt is the rate of biomass increase. The
parameter fi, defined as the rate of increase of population per unit of population
(dx/dt) (1/x), is termed the specific growth rate and has the dimension of
reciprocal time (1/t).
Thus:
fi — dx.l (3)
dt x
Metabolic Quotient
The rate of consumption of a substrate in a culture at a particular
moment is given by:-
-ds/dt = qx (4)
Where x is the biomass and the coefficient q is known as a metabolic quotient
or specific metabolic rate (unit of substrate/unit of biomass/unit time).
Effect of Substrate Concentration on Growth Rate
The substrate concentration has virtually no effect on growth rate if its
concentration is large compared to the value of the saturation constant. Thus
Michaelis-Menten enzyme kinetics can be used to describe the substrate
consumption, if s is the substrate consumption and q is the metabolic quotient,
then:-
q = Qm ( s \ (5)\ K + s)
Where k, is the saturation constant and is equivalent to the Michaelis-Menten
constant, and q^ is the maximum value of q obtained when s > k* .
50
If q = ju/Y and Qm = /^/Y (where Y is the growth yield, dx/ds),
are substituted into equation (5) then:-
t* An I s( s + K
This equation is referred to as the Monod equation, since Monod (1942)
first showed mathematically that the expression described the relationship between
bacterial growth rate and substrate utilisation.
The Monod equation forms the basis of many models used to describe
and predict the behaviour of hybridomas in culture. The difficulty with using
this equation in this situation, is that mammalian cell requirements and media
are very much more complex than than those of bacteria and the identification
of a growth rate limiting substrate is difficult. Often more than one substrate
may be limiting. This situation is additionally complicated by the inhibitory
effects of culture metabolites and models have been developed using an extended
form of the equation, which include an appropriate inhibition constant (Ki).
Growth Phases of a Simple Batch Culture
A simple homogenous batch culture progresses through different phases
which reflect changes in the cell population and environment. These phases are
generally known as the lag, log, stationary and death phases (Figure 3.1).
During the lag phase there is apparently little growth, but significant
metabolic activity. The length of this phase may depend on one or a number
of factors including; nutritional and/or physical environment, state of the
inoculum or the presence of inhibitors. Cells in lag phase may induce synthesis
of enzymes in order to exploit a nutritionally favourable environment, reduce
an inhibitory substrate concentration, or adapt in some other way to a changed
physical environment. If the inoculum is taken from a culture in stationary-
51
phase then re-organisation of the cell, necessary to reverse the changes caused
by cessation of growth, may contribute to the length of the lag phase.
I ag stationaryexponent ia l death
tnin<5o5ooj
TIME
Figure 3.1 A typical growth curve for a microbial batch culture.
After the lag-period, growth occurs at the maximum rate for the given
physiochemical conditions, and finally ceases; either through the lack of a
nutrient, accumulation of an inhibitory product, or, some changes in the physical
environment. Once the biomass reaches a maximum value, there may be a
stationary phase where the amount of biomass remains constant. Eventually
however the biomass declines, probably as a result of autolysis.
52
3.1.2 Relationship of Growth Rate to Product Formation
a) Growth Rate-Associated
In general, the rate of product formation is given by the following
equation
dp/dt = QpX (7) (Pirt, 1985)
Where p is the product concentration, x is the biomass concentration, and qp
is the specific product formation rate.
When the product is growth-rate associated the amount of product formed
is directly proportional to the biomass formed.
Hence
dp = Yp/Xdx (8)
Where Yp/X is the product yield referred to biomass formation.
Thus:-
dp/dt = Yp/X dx/dt = Yp/x/xx (9)
If the product yield is expressed in terms of the substrate used then:-
dp = Yp/Sds (10)
Where Yp/S is the product yield referred to substrate utilisation
Hence:-
dp/dt = Yp/sds/dt = Yp/^x/Y^ (11)
Where Y^ is the biomass or growth yield referred to the substrate utilised.
53
Ypfc/Y* (12)
Thus from equations (10) and (12):
qp = dp/dt = Yp/X ii (13)
b) Growth Rate-Dîssociated
Non-growth-linked product formation can be of two types;
i) qp is independent of growth rate
ii) qp varies with specific growth rate in a complex way
One example of type i) is the production of penicillin from Pénicillium
which is independent of growth rates greater than 0.015 h"1 (Pirt and Righelato,
1967).
For a non-growth linked product, qp can be a complex function of the
specific growth rate. An example of this is melanin formation by A. niger
(Rowley and Pirt 1972) which is represented by qp = qp™ -kfi where qp™
and k are constants.
The product formation relationship with growth can be determined using
a Leudeking-Piret model (Figure 3.2).
From this, the following equation describes the line of the graph (Sinclair
and Kristianson, 1987)
rp = a rx + I5xv (14)
Where rx is the rate of cell growth
xv is the viable cell concentration
rp is the product formation rate
a is the growth related product formation coefficient
fi is the non-growth related product formation coefficient.
54
Thus from the equation, if fi (the y-axis intercept) = 0 and a (the gradient
of the line) > 0, then the product formation is growth associated. However,
if a — O and fi > 0 then there is non-growth associated product formation.
There is an intermediate or complex relationship between n and qp if a and
f i> 0.
XV
rxxv
a °<>° /3=o b e<= O j& > 0 c o i &. a
Figure 3.2 The Leudeking-Piret Model
For hybridomas, several types of production kinetics and have been
described for Mab by various investigators (Merten et al 1981). The three types
discussed are:
I. The highest specific Mab production rate is found during the lag j t e
of cell growth and the beginning of exponential. This type of
production may be growth-associated in a complex manner.
II. The highest specific Mab production rate occurs at the beginning of
exponential growth and during the stationary phase. This description
suggests growth-dissociated kinetics.
III. Production kinetics are growth-associated and expressed whilst cells are
experiencing constant specific-rate only in the logarithmic growth phase,
with no production during stationary and decline phase.
The production kinetics of the Mab will therefore largely determine the
production system employed.
3.1.3 Growth. Monoclonal Antibody Production and The Cell Cycle
Assuming that hybridoma cell kinetics are similar to those of other
mammalian cells; a hybridoma cell in culture passes through, or is arrested at
one of 4 phases of the cell cycle (Darzinkiewicz, 1987; Lloyd, 1982; Mitchison,
1971): the gap 1 phase (Gl-phase), the DNA synthesising phase (S-phase), the
gap 2 phase (G2-phase) and the mitotic phase (M-phase) as shown in Figure
3.3. There may or may not be a GO phase preceding Gl, which is a senescent
phase.
Progression through the cell cycle is driven by numerous factors:
(i) A sequence of gene activations
(ii) brief surges of regulatory metabolites such as Ca2+, Ca2+-calmodulin,
cAMP
56
(iii) transient unmasking or insertion of receptors for external growth
factors into the plasma membrane.
(iv) bursts in enzyme activity
(v) cytoskeletal restructuring
(Boynton and Whitfield, 1983; Hochhauser et al 1981; Leffert et al 1979).
Addition o#Ca2> or serum or POGFor p p 6 0 v ’ -y/’tractlvation or high pH shock
Ca2> external and / or Ca2* internal
SHUTDOWNPROCESSES
M /D-o
cu
REPL1CAT10N SWITCH ON
SWITCH
C hrom osom e replication
Figure 3.3 The Cell Cycle of Cultured Cells
(Whitfield et al, 1985)
57
Consequently, cellular morphology, the types, densities, binding, and
regulatory characteristics of surface receptors, and the cell’s enzyme profile,
change during progression through the cycle. The progressive changing of
structural and functional composition means that regulators such as cAMP will
promote cell-cycle transit when generated at the right times in the cycle, but
will block transit if made to appear in abnormal amounts at other times
(Boynton and Whitfield, 1983).
The Mab production rate of a single hybridoma will depend on the
phase of the cell cycle. Many of investigators who measured the Ab production
rate in each phase reported that maximum Ab synthesis rate occurs in Gl or
early S phase in myelomas or lymphoids (Buell, 1969; Garatun-Tjeldsto et al,
1976; Takahaski, 1969). When considering the work of these investigators, due
consideration should be giving to the method of cell synchronisation used. Many
methods involve harsh treatment with drugs and the effect of these chemicals
on the cell may be overlooked in the conclusions drawn. Indeed, certain
treatments can radically affect the results (MacMillan and Wheatly, 1981).
The time for a cell to traverse the Gl phase increases as specific growth
rate (p) decreases, while the periods required for a cell to traverse the S-G2
and M-phase are approximately constant, and independent of p (Alberts, 1983).
Thus, the overall cell cycle time increases as n decreases. The processes of
replication will normally stop when a cell reaches a point late in Gl-phase
called the restriction point, unless signalled to continue through another cycle
(Alberts, 1983). A cell stopping at late Gl is called an "arrested cell". In order
for the cell to progress to Gl, the correct stimuli must be received. Further
stimulation progresses the cell from Gl to S phase. If the stimulation is absent
or too low, cells may remain in Gl or slip back into GO. During Gl phase
there are gene activations, cascades of enzyme activation and structural changes
which lead to the production of active DNA-replicating enzymes, initiator
proteins, chromosomal proteins and other components needed for a round of
chromosome replication (Boynton and Whitfield, 1983; Das, 1980;, Floras et al,
58
1978; Gutowski and Cohen, 1983; Hochhauser et al 1981; Jazwinski et a/,1976;
Mercer et al 1982; Rao et al, 1982; Whitson, 1980).
The role of Ca2+ in the progression from Gl is very interesting.
Extracellular Ca2+ (Ca2+ ext) deprivation results in reduction of DNA synthetic
activity and thus prevents cells from entering S phase. If the regulator proteins
are partially suppressed to sub-threshold level without allowing complete
suppression or degradation, the cell will remain in Gl. The level of Ca2+ ^
may suppress the regulator proteins or may prevent replication enzymes, their
subunits or replication-initiator proteins, from entering the nucleus to start
chromosome replication. There may be two kinds of Ca2+-dependent processes
operating during the Gl phase; one of which is the permissive requirement for
Ca2+-dependent protein kinase C activity, and the other is triggering by a
transient surge of internal Ca2+-calmodulin (a Ca2+ binding protein) complexes
(Whitfield et al, 1982). cAMP and cAMP-dependent protein kinases also trigger
critical events in late Gl. In a wide variety of continuously cycling cells,
including Chinese hamster ovary cells (CHO), there is a short burst of adenylate
cyclase activity, a transient reduction of cyclic nucleotide phosphodiesterase
activation and a resulting transient cAMP surge (Boynton and Whitfield, 1982;
Sullivan and Wedlicani, 1981). Exposure of external cAMP to an adenylate
cyclase stimulator (eg Ca2+) triggers DNA synthesis of cell trapped in Gl by
Ca2+ deprivation (Boynton and Whitfield, 1982).
The viable-cell fraction arrested at late Gl-phase, (where maximum Ig
production occurs) has been reported for myelomas (Byars and Kidson, 1970;
Garatun-Tjeldsto et al, 1976;) of lymphoids (Buell and Fahey, 1969; Takahasthi
et al, 1969), increases when the growth-rate decreases. At jumax, cells traverse
the cell cycle with a minimum time and there are few cells (if any) arrested
in Gl-phase. As the culture environment deviates a little from the optimum for
growth, the cell-cycle time increases and the fraction of arrested cells increases
but is still small. However, as the environment becomes less favourable to
growth this fraction increases further.
59
3.1.4 Process Analysis Techniques
There are a number of ways in which the data from a culture can be
analysed. These include the use of curve-fitting techniques to facilitate derivation
of important rate quantities, and the use of models to predict events which may
occur.
Mathematical Modelling of Bioreactor Processes
Empirical investigations of the behaviour of a bioreactor system under
all possible conditions, and at all possible scales are often not possible.
Mathematical modelling is particularly useful in this situation for predicting the
effects of change of scale or conditions. These predictions are essential if an
operating process is being designed on the basis of laboratory investigations, but
they are also useful in handling laboratory data and in elaborating hypotheses
to explain observations. In addition, the actual construction and checking of a
kinetic model of a process is often the first step, and a very powerful aid; in
designing further experiments, in exposing areas of unsuspected ignorance, and
in furthering general understanding of a problem.
In the literature, there are an increasing number of mathematical models
describing hybridoma growth, Mab production, substrate utilisation, and
inhibition (Batt and Kompala, 1989; Bree and Dhmjat, 1988; Miller et al,
1986). Most are modified from the Monod equation.
Spline Functions
The important rate variables in a batch or fed-batch culture, such as
specific growth rate and specific product formation rate, are derived from
experimental data. It is, for example, necessary to differentiate the batch biomass
and product concentration to estimate the specific growth rate (/*), and the
specific product formation rate (qp). Since differentiation of measured
60
concentrations amplifies the experimental error, one needs to use the best
available methods for estimating fi and qp. Spline functions allow differentiation
of data and give approximation for data interpolation and/or smoothing (Reinsch,
1967). A cubic spline is a piecewise continuous function, consisting of cubic
polynomial segments that join at points called knots. Thus a cubic spline is
formed by fitting a polynomial to three data points, then taking the next three
points and fitting another polynomial, and finally joining the two polynomials
together. This is then repeated until all the data points have been included. An
interpolating spline forces the polynomials through the points, whereas a
smoothing or partial cubic spline does not go through all the points and uses
a fit factor which determines how close the points are to the curve. An
interpolating spline should only be used if the data are known to completely
free of "noise", otherwise some degree of smoothing is called for (Thornhill,
1989). Cubic splines are able to give good fits to data which show abrupt
changes, and they can also mimic a wide range of analytical functions.
In this study partial (or smoothing) cubic splines have been applied
where appropriate, to data from the bioreactor batch cultures and the curves
used to derive the required rate quantities.
3.2 INVESTIGATION OF GROWTH AND MONOCLONAL
ANTIBODY PRODUCTION
In this series of experiments, 101 bioreactor batch cultures of the PQX
Bl/2 hybridomas were carried out as described previously (2.1.7). Samples
were taken at regular intervals (6-8 hours), and the viable and total cell
densities determined (2.1.4). The following analyses were performed depending
on the experiment: dry weight (2.1.5), Mab (2.2.4), GLN/NH3 (2.2.5), amino
acids (2.2.8) and adenylate charge (Bell, 1990).
61
Where:
Adenylate Energy Charge (EC) = [ATP] + %[ADP]
[ATP] + [ADP] + [AMP]
3.2.1 Growth and Mab Production Bv POX Bl/2 in Serum
Supplemented Medium
Results and Discussion
Cell Growth
The viable cell growth curve for PQX Bl/2 was typical for a batch cell
cultivation (Bailey and Ollis, 1986). There was a short lag phase of 10 hours
during which the cells adapted to the new nutritional environment and
synthesised the hormones, growth factors and enzymes required for growth.
This was followed by the exponential growth phase of the culture, which
yielded 8 x 105 viable cells/ml (Figure 3.4a).
The short lag phase is indicative of a viable inoculum at a high
concentration and a high concentration of growth factors. This reduces the time
required for the cells to synthesise their own growth factors, and can therefore
devote more energy and metabolites to growth processes.
After 30 hours of exponential growth the culture entered a decline phase.
That the total cell number did not peak with viable cell number as
would have been expected for a population of high viability, may be due to
maximum viable and total cell number being attained between sampling points.
The experimental data presented in the graphs of chapter 3 is included as i
Appendix 5 on the back cover of this thesis.
62
12.00 0.18
9 .6 0 0 .1 4in
<o
12.0 - 0.11
4 .8 0 ^ 0 .0 7
8 2 .4 0 0 .0 4
0.00 0.004 00 20 6 0 8 0 1 0 0
TIME (hours)
Figure 3.4a The growth curve of PQX Bl/2 in batch culture. The figure
shows the change in viable cell count ( • ) , the increase in total
cell number ( a .) and the increase in biomass ( v ) .
The smooth curve fitted to the data also suggested a maximum viable cell
number in excess of that recorded.
The determination of specific growth rate in batch cultures is difficult
as conditions change continuously. In order to determine fi from the data in
Figure 3.4a, a partial cubic spline was fitted to the data. This smoothing curve
was then differentiated to give dx/dt, and multiplied by 1/x to give p. The data
manipulations were facilitated by a BBC microcomputer and appropriate software
(Appendix 3).
Maximum specific growth rate Qxmax) was determined from dry weight
data (ie x = biomass), to be 0.049b"1 (Figure 3.4b). As a comparison the
maximum specific growth rate was also determined using viable cell
concentration (x = viable cell number) and was 0.046h'1 (Figure 3.4c).
63
DRY
WEI
GHT
(mg/
ml)
o.i a 0.60
0.14 - 0.48
ck
0.11 - 0.36
0.07 k b - 0J24
0.04 - 0.12
0.00 0.00o 20 40 1 0 060 80
0S;1tnh
IoLLâo5
TlfvE (hours)
Figure 3.4b The determination of specific growth rate (/t) from the change
in biomass ( • ) c0.6010
£in4o - 0.48
x!~
8ILU
ai<>
- 0.24
- 0 .1 2
0.0080 1 0 0604020O
O*
LUH-<crxh
BoLL
S%
TIME (hours)
Figure 3.4c The major peak in specific growth rate (fi) determined from
the viable cell count ( • )
64
The maximum specific growth rate determined from biomass occurred
at the same time as the rate determined from viable cell number. The growth
rates were similar, with the dry weight method producing a marginally higher
value. It is probably, in the case of animal cells, more accurate to determine
growth rate from the viable cell number rather than biomass, because of
variability in the size of cells. The size of a hybridoma can vary according to
the nutritional environment and stage of the cell cycle. For example in mitosis
(M-phase), there will be two daughter cells smaller than the parent or cells in
S-phase. In a batch culture, the cells are all in different stages of the cell cycle
and the distribution of cells through the cycle will cause variations in the
average cell dry weight, and thus may cause discrepancies in the calculated
biomass and growth rate. These errors may also occur if the cells change size
and composition due to a change in nutritional environment. It is also common
for cells to adhere to glass and metal surfaces in a bioreactor, then slough off
into the medium as clumps, which may cause errors in dry weight determination
and the growth rate.
Moreover, the volume of cell suspension required for dry weight analysis
is much higher than the requirement for viable cell determination. It is
obviously desirable to minimise the total volume removed from the bioreactor
as sample, in order to maintain the surface area to volume ratio and other
physical parameters constant. Viable cell counts are more economical in terms
of sample volume, and therefore can be made more frequently, and a more
accurate value for growth rate determined. In small scale experiments where the
total volume of cell suspension may only be 50ml, dry weight analysis is not
possible.
These considerations together with the extended processing time for
dry-weight analysis has meant that all specific growth-rates in future experiments
were determined from viable cell number.
65
Monoclonal Antibody Production
Mab titre was plotted against time using partial cubic spline curve fitting
as described, and showed an increase from 7.6/ig/ml at inoculation to 44.2/zg/ml
at 65 hours. Thé product formation rates (dP/dt) determined by differentiation
of the cubic spline (Figure 3.4d) showed a peak at 65 hours.
2.508 50iLUcr 40P>8 30 m
2.00
1.50
1.0020
o
- 0.50
0.001 0 0806040200
i-3*LU
$(J
I
TIME (hours)
Figure 3.4d The Volumetric production of Mab ( • ) by PQX Bl/2 in batch
culture, and the major peak in production formation rate (dP/dt)
There was however some production of Mab during the exponential growth
period, suggesting that there is not a simple relationship between growth and
Mab production. The kinetics appear to be similar to the type H kinetics, as
described by Merten (1989), in that the highest Mab production rate is at the
; end I of exponential growth. ' ~ This pattern
of Mab production is in agreement with the observations of other workers
66
(Boraston et al, 1984; Fazekas de St Groth, 1983; Merten et al, 1985; Seaver
et al 1984; Vêtez et al, 1986). To examine the relationship between growth
and Mab production, the specific growth rate and the product formation rates
were plotted on the same graph with respect to time (Figure 3.4e). Specific
growth rate peaked at 35 hours while dP/dt peaked 30 hours later at 65 hours.
oX
kHiH
F!5yLL9§5
0.60
0.48
d P / d t
0.36 - 1.50
0.24 • 1.00
0.12 - 0.50
0.00 0.0040 6020 1 0 00 80
TIME (hours)
Figure 3.4e The relationship between specific growth rate (fi) and product
formation rate (dP/dt)
To investigate this relationship further the growth and product formation data
were fitted to a Leudeking-Piret model (Figure 3.9f).
The values of rx and rp were determined from the differential curves from
Figures 3.4c and 3.4d respectively. The plot shows a correlation coefficient of
0.827 determined by linear regression. The equation of the line was Y =
1.371xl0"2 X + 2.4xl0'3, and therefore:
67
Mab
PPRO
DU
CTIO
N
RATE
jp
g/m
l/h)
r/Xy = L371xlO'2 r/Xy + 2.4 xl0‘3xv
Therefore the growth related product formation coefficient (a) = 1.371
x 10*2 (g product/g cells), and the non-growth related product formation
coefficient (B) = 2.4 x 10'3 (g product/g cells/h).
0J20 -
0.15
0.10
0.05
0.002.001.601.200.800.400.00
rx/xv
Figure 3.4f A Leudeking-Piret plot to show the association of specific
growth rate and product formation rate.
This suggests that there are growth-associated and growth-dissociated functions
in the relationship, thus supporting. the proposal that the relationship is not
simply growth rate associated or dissociated. Merten (1980) used specific
product formation rates in his work, whereas the work presented here uses
volumetric rates which simplifies investigation of trends and relationships,
without relating them to growth. This is also illustrated by Figure 3.4g which
shows the difference between specific and volumetric rates.
68
Using volumetric rates allowed the delineation of the different phases of
penicillin production, which are masked by including in the equation the
expression which relates then to growth.
u
0.20
0.15
0.10
0.05
25 1.25
20 1.00
U 15 - 0.75
10 - 0.50
5 0.25
0
0.5 0.020 1.0
0.4 0.015 - 0.8
0.3 0.60.010 -
0.2 _ :q Xoa
0.005 -
0.1 0.2
0 0 0
-0.1
M Mycelial wt (dry) g/L B Penicillin, g/L C Sugar used, g/L
_ 1 l ......1— i i i i _. A Growth, g/g, h
- \ X XB Penicillin
Z \ synthesis, g/(jL-h)
- A fv Sugar utilization,
V g/cvh)
c
\ VV ^ Oxygen _ \ uptake,"
D g/(L*h)
- A1
0.0012
0.0010
0.0008
0.0006
0.0004
0.0002 -
T 1------1-----rA Growth, g/(g*h) - B Penicillin
synthesis, g/(g*h)- C Sugar utilization,
g/(g'h)D Oxygen uptake,
g/(geh)B'
0 20 40 60 80 100 120 140Fermentation time, h
Figure 3.4g Complex kinetics for production of penicillin (Leudeking, 1967)
a) Time course; b) volumetric rates; c) specific rates
69
Glutamine and Ammonia Assimilation
The rate of GLN uptake (-dGLN/dt) was determined using partial cubic
splines as described, and showed peaks at 1*8 hours, and again later at 45 hours
(Figure 3.4h).
0.502.50
0.40
0.30
0.201.00 -
0.100.50
0.000.001 0 08 0604 0200
TIME (hours)
Figure 3.4h The volumetric change in glutamine concentration (#) , and the
rate of utilisation (-dGLN/dt) determined from a 100% cubic
spline fit of the volumetric data
The production of ammonia was investigated in a similar manner and found to
peak at 25, 45 and 65-hours (Figure 3.4i). Between 30 and 40 hours there was
no ammonia production, but apparent assimilation (Figure 3,4i). GLN utilisation
and ammonia production rate peaks appeared to be phasic in relation to growth
and production. The first GLN peak appeared before the maximum growth peak
while the second coincided with maximum Mab production. This suggests that
70
GLN
UTIL
ISAT
ION
RATE
(m
M/h
)
k
GLN is important for the preparation for growth, and for Mab production.
Ammonia disappearance occurred in between these peaks, when the cells were
growing at their maximum.
0.150.15
ié 0.12111 4 .
5% © P 9
I kï Ls I
I.O B
A
0.12
0.09
0.06
0.03
0.001 0 08 06 04 020
TIME (hours)
Figure 3.4i The major peaks in NJti3 production (A) and NH3 assimilation
(B). These compare to the major growth period (-----)
These results suggest multi-phasic growth and Mab formation (Figure
3.5).During the first phase there was a lag period with apparently little growth but
utilisation of GLN with concomitant production of ammonia. During the second
phase (Phase II), growth was at a maximum and ammonia was assimilated. The
main period of Mab production occurred during the third phase (Phase IE)
together with further GLN utilisation and ammonia production.
71
NH3
UTIL
ISAT
ION
RATE
(m
M/h
)
A
TIME
Figure 3.5 The phasic nature of nutrient assimlation. The figure shows the
major peaks in GLN assimlation rate (B), the major peak in
specific growth rate (A) which coincides with a major period
of ammonia assimilation (C). During the final phase there is
the major peak in Mab production (D).
72
3.2.2 Nutrient Assimilation Bv POX Bl/2 in Batch Culture
In order to further investigate the phasic pattern the experiment was
repeated with addition nutrients assayed.
Results and Discussion
Cell Growth
The results from this experiment confirm the phasic nature of growth
by PQX Bl/2 in batch culture. Viable cell number peaked at 8.72 x 105/ml
(45 hours), with a maximum specific growth rate of O.OSSh*1 (24 hours) (Figure
3.6a).
0.6010
- 0.48Ein<01$IB
0.36
- 0.24
m<> - 0. 1,2
0.001 0 08 040 60O 20
TIME (hovrs )
Figure 3.6a The growth curve for PQX Bl/2, showing the increase in
viable cell count ( • ), and the specific growth rate (/*).
73
SPEC
IFIC
GR
OWTH
RA
TE
(h-1
) (x
lti"1
)
There was a decline in cell viability during the final phase of the culture
which may be due to depletion of nutrients, a build up of toxic metabolites or
to an inability of the cells to maintain the membrane integrity. There is an
energy requirement to maintain ionic gradients across membranes; a fundamental
characteristic of all living cells. Na+, K+-ATP'ase utilises the high-energy
phosphate bond of ATP to pump ions across the membrane against concentration
gradients. In addition to maintaining a K+-rich intracellular environment, the
ionic gradients generated by this enzyme are coupled to transport processes,
intracellular Ca2+ level and distribution, oxidative phosphorylation and glycolysis.
Moreover, the Na+, K+-ATPase activity is coupled with cell growth cycles and
DNA synthesis (Kennedy et al, 1984; Leister et al, ;1985; Rozengurt et al
1975). As cellular energy decreases, the ability to exclude vital stains (eg trypan
blue and erythrosin B) diminishes. Cells are therefore more likely to retain the
dye, and appear nonwiable. Weakening of cell membranes may also be
accelerated by shear forces imposed in bioreactors, (for example though agitation
by the impellor or sparging). Such forces may only be significant to cells with
compromised membrane integrity.
Monoclonal Antibody Production
0 .6 0
5 0.48
1I 0.36
I5 024ULLÜ 0.12
60.00
0 2 0 4 0 6 0 - 8 0 100
TIME (hours)
Figure 3.6b The displacement in time of the peak in specific growth rate (^)
and product formation rate (dP/dt)
74
Mab production rate was determined as before and peaked at 40 hours,
16 hours after the peak in growth rate (Figure 3.6b), corroborating the results
from the previous experiment.
Ammonia Assimilation
There were two peaks -in ammonia production which coincided with the
peaks in glutamine utilisation (Figure 3.6c).
0.15 0.20sz
0.12 0.16
0.09 - 0.12
(5< 0 .06 0.08
Ecn A0.03 - A 0.04
Br IB0.00 0.00
0 20 4 0 6 0 1 0 08 0
TIME (hours)
Figure 3.6c The major peaks in the utilisation rate of GLN (A), and the
major peaks in NH3 production rate (B), with apparent
assimilation (C)
75
GLN
UTIL
ISAT
ION
RATE
(m
M/h
)
The apparent assimilation of ammonia during maximum growth was also evident
again; an observation interesting for two main reasons:
(i) there is no similar report in the literature
(ii) ammonia is considered to be a toxic metabolite and it therefore seems
strange that a cell would assimilate a toxin.
Evidence of ammonia assimilation can be found in the literature, although
the observation has not received attention. For example Bree et al (1988)
studied the use of models of hybridoma cultures, and compared them to
experimental data. Application of partial cubic spline curve fitting, to their data
(Figure 3.7a), reveals a similar apparent assimilation of ammonia (Figure 3.7b).
2 .3 -
s2 . 0 -
.3 -
. 0 .3
0 2 J 5 3 3rim* (Ocv*)
3.00 0.25
2.40
5 :< 1.80
U 1:20
89 0.60
0 .15
0.10
0.05
0.00 0.000 1 2 3 54 6 7 a
T l^€ (days)
Figure 3.7 These figure show the data presented by Bree et al (1988), A)
and the determination of a period of apparent ammonia
assimilation(-dNH3/dt)(Fig B) by differentiation of a 100% cubic
spline.
It is however difficult to extrapolate the data since no experimental variation is
indicated, and the interpretation may be misleading.76
In this case, although sample points are at 24 hour intervals, the drop in
ammonia concentration seems to be significant (0.35mM).
Further investigations into ammonia assimilation are discussed in Chapter
5.
Amino Acid Assimilation
Data from amino acid analysis were fitted with partial cubic splines and
the utilisation rates determined (Figures 3.8a and 3.8b).
50 2.50
dlLE/dt■6k 4 0 - 2.00
1.50
20 1.00
10 0.50
0.00O 2 0 4 0 6 0 1 0 08 0
TIME (hours)
Figure 3.8a The volumetric change in isoleucine concentration ( • ) , (a 100%
cubic spline fit), and the rate of utilisation (-dDLE/dt). The
profile is a typical example of the utilisation pattern of the
following amino acids: ASN, GLN, HIS, SER, ARG, THR,
VA1, MET, CYS, ILE, LEU, LYS, PHE, TYR, and CIT.
77
ILE
UTIL
ISAT
ION
BATE
jU
g/m
l/h)
The utilisation rates of ASN, GLN, HIS, SER, ARG, THR, VAL, MET,
CYS, ILE, LEU, LYS, PHE, TYR and GIT peaked during the lag phase of
the culture, and again at the end of exponential growth. ALA, ASP and ORN
were assimilated during mid-exponential growth, between the other amino acid
assimilation peaks’
4.001 0 0
E05 3.208 0
2.406 0
1.604 0
85
dAL A /d t 0.802 0
0.001 0 04 0 60 8 0O 2 0
TIME (hours)
Figure 3.8b The change in volumetric concentration of alanine ( • ) (100%
cubic spline fit) and the rate of production (dALA/dt). This
is a very similar nrofile to that os ASP and ORN.
The utilisation of ARG was interesting in that the volumetric curve mirrored the
ammonia volumetric curve, such that, as ammonia was assimilated, ARG was
produced.
78
ALA
PROD
UCTI
ON
RATE
6u
g/ml/h
>
The utilisation rates of most of the amino acids appear to correlate with
the phases of growth, showing distinctive peaks. ASN, GLU, GLN, HIS, SER,
ARG, THR, VAL, MET, CYS, ILE, LEU, LYS, CIT, PHE and TYR all
peaked during phase I of the culture. The assimilation of so many amino acids
suggests significant metabolic activity such as synthesis of enzymes and structural
proteins, in preparation for exponential growth. Indeed, it was proposed that
cells require a threshold intracellular concentration of amino acids for proteins
synthesis (Eagle et al, 1961; Griffiths, 1972). During maximum growth, in
phase II, there is assimilation of ALA, ASP, and ORN. The energy requirement
during maximum growth, may exceed that of other stages in the culture. This
energy may be derived from the amino acids assimilated during the lag phase,
or by assimilation of ALA and ASP during the log phase. ALA and ASP can
be metabolised via aminotransferase activity to form a-ketoglutarate, a key
component in the tricarboxylic acid (TCA) cycle.
During the final phase of the culture during which Mab was produced
at a maximum rate, there was a significant period of amino acid assimilation
(Figure 3.8a). This uptake was probably to meet the requirements of protein
synthesis both in terms of precursors and energy.
Glucose Assimilation. Lactate Production and Energy Synthesis
A change in glucose concentration was observed and the rate of
utilisation determined. Glucose was assimilated during growth of the PQX Bl/2
hybridoma, however maximum glucose utilisation occurred after maximum growth
(Figure 3.9a).
A major source of energy would be expected to be catabolism of
glucose. The carbon flux of glucose in glycolysis may be directed via five
different pathways: pentose phosphate, lipid synthesis, lactate production, amino
acid synthesis and the TCA cycle.
79
3 .5 0200 -
11 6 0 =
O
<OC 1 2 0 -
- 2 .8 0
- 2.10"dGLUC/dt
y§ 8 0 •
O
œO 4 0 -
3- i
- 1 .40
; 0 .7 0
0.001 0 08 06 02 0 4 0O
TIME (hours)
Figure 3.9a The change in glucose concentration (#') and the major peak
in utilisation rate (-dGLUC/dt). The figure shows that the
maximum utilisation of glucose occurred after the peak in
specific growth rate ( f i )
The rate of glycolysis is usually much greater than the maximum rate
of utilisation of glycolytic intermediates (Hume et al, 1979), and therefore most
of the glucose is metabolised to lactate. That the lactate concentration increased
as the glucose concentration decreased (Figure 3.9b), suggests a significant
conversion of glucose to lactate.
80
GLUC
OSE
UTIL
ISAT
ION
RATE
(m
g/dl
/h)
12.00 0.50
I9.60 - 0.40
7.20
4.80 0.20
2.40 0.10O L /d t
0.00 0.00o 2 0 40 60 80 1 0 0
TIME (hours)
Figure 3.9b The increase in lactate concentration ( # ), as glucose
concentration decreased ( A )
The net lactate production exceeds the net glucose utilisation and
therefore suggests that lactate is also produced by some other means. Although
glycolytic activity is generally high in proliferating cells (Pederson, 1978; Hume
et al, 1979), it has been shown that glucose is not essential for normal or
tumour growth (Hume et al, 1978; Poussegurt et al, 1980; Romano, 1982) or
for growth of human fibroblasts and HeLa cells (Wice et û/,1981; Zielke et al,
1976).
It has been proposed (Hume et al, 1978), that the glucose requirement
for cell growth is to supply precursors needed in nucleotide synthesis, together
with amino acid and lipid formation, rather than for energy production. The
results from this experiment suggest that there is a net synthesis of amino acids
during the maximum growth phase. It seems likely therefore that glucose is
indeed being used primarily for nucleotide, lipid and amino acid synthesis during
growth. The production of lactate does however suggest that there is some
81
LACT
ATE
UTIL
ISAT
ION
RATE
(m
M/h
)
energy production from glucose. Immediately before the maximum Mab
production, glucose utilisation rate peaks which implies that amino acids and
energy required for Mab synthesis may be derived from glucose utilisation. It
seems therefore, that glucose performs different roles during the progression of
the culture and to different extents. The flux of carbon from glucose may
change in accordance with cellular metabolic requirements, and perhaps in
response to environmental changes.
Glutamine Metabolism and Energy Production
If glucose does not produce all the energy required by the cell then it
must be derived from some other source. When both glucose and glutamine are
available, it has been estimated that between 30 and 65% of all cell energy
requirements are derived from glutamine (Reitzer, 1979; Zielke et al, 1984).
Carbon from glutamine can be directed through a variety of branched pathways,
each yielding different amounts of ATP. For example glutamine can be
completely oxidised to C02 to provide 24 to 27 moles of ATP (Glacken 1988).
However, if Glutamine is catabolised to lactate each mole of glutamine forms
6 to 9 moles of ATP assuming NADH produces 3 moles of ATP (which may
also account for the extra lactate unattributable to glucose metabolism). Carbon
from partial oxidation of glutamine to glutamate and subsequently to pyruvate,
(after passing through the TCA cycle), can contribute to the formation of lipids
or lactate.
Incomplete oxidation of GLN to ALA or ASP would result in a
maximum of 9 moles of ATP per mole of GLN. This mechanism of GLN
catabolism is known as glutaminolysis (McKeehan et al, 1982), because of the
similarity to the partial oxidation of glucose. Along this pathway GLN
contributes precursors for the formation, of major intracellular building blocks:
amino acids, nucleotides, proteins and lipids. Glutaminolysis together with the
oxidation of serum-derived ketone bodies and fatty acids could provide most of
the energy required by the cell for growth (Ardawi et al, 1982). It seems likely
82
therefore that the main source of energy is glutamine utilisation, with a
contribution from the catabolism of other amino acids and glucose. The
proportion of cellular ATP derived from GLN oxidation (as opposed to glucose
catabolism) varies according to the cell type, the environmental conditions
(Glacken, 1988) and the stage of the cell in the cell cycle.
3.2.3 Growth and Mab Production in Serum-Free Medium
Serum-dependent PQX Bl/2 hybridomas were weaned into serum-free
medium (Appendix IF) over a period of about 12 weeks, by reduction of
serum-supplementation. These cells grown in serum-free medium were used in
this experiment.
Results and Discussion
Cell Growth
The growth curve of the PQX Bl/2 hybridomas adapted to serum-free
medium was displaced with time due to an extended lag phase. This lag phase
was probably due to either a low inoculum density or a characteristic of the
serum-free medium. If the inoculum density is low, then there are fewer cells
to synthesize growth factors and hormones to enable the population to multiply.
Hybridomas do secrete non-immunoglobulin factors and these factors are
speculated to have a growth stimulatory function (Glassy et al 1988). With a
serum-supplemented medium it is probable that certain serum components provide
the stimulatory activity necessary for growth. However, since these factors are
not present in a serum-free formulation, promotion of cell growth probably
depends on self-secreted autocrine factors, and the accumulation of these factors
reaches a critical concentration at a fairly high cell density.
Another cause of an extended lag may be the time taken for the
detoxification of cellular environment: the higher the initial cell concentration the
83
greater the level of detoxification (Glacken et al 1989). For example, it has
been shown that hydrogen peroxide (H20 2) is spontaneously formed in cell
culture medium when it is exposed to light. Catalase alleviates the toxic effects
of H20 2, if it is either added to serum-free medium (Darfler and Insel, 1982),
or synthesised by the cells. Since the serum-free formulation used contains no
catalase, then any toxic effect of H20 2 must be neutralised by synthesis of
catalase by the cells.
It is also possible that hybridomas re-activate some spontaneously
inactivated growth factors. For example cysteine, a free thiol, is spontaneously
oxidised to cystine, a disulphide, in the presence of oxygen and/or iron
(Toohey, 1975). Cystine has been shown to be transported very slowly into
lymphocytes, whereas cysteine is transported, rapidly (Ohmori et al, 1982).
Mammalian cells have a strict requirement for either form (Glacken et al,
1989). Thiol compounds in serum are known to engage in thiol exchange
reactions with cystine (Ohmori et al, 1983a). The mixed disulphides produced
from these reactions are transported more rapidly than cystine into lymphocytes
(Ohmori et al 1983b). There is some evidence that lymphocytes have the
capability to reduce cystine to the more active form (Ohmori et al 1983a). It
is possible therefore that in serum-containing medium the reduction of cystine
to the more "active" cysteine is mediated by serum components. In serum-free
medium this reduction may be performed by the hybridoma cells. This later
reduction method may take longer, and therefore play a role in the
determination of the length of the lag phase.
The general growth curve was comparable with cells grown in serum-
supplemented medium, with a maximum cell count of 7.2 x lOVml (Figure
3.10a).
84
E 0.40in<oT—X 0.30
0.20ni
i>
- 0.10
0.001 0 08 06 0402 00
TIME (hours)Figure 3.10a The growth curve of PQX Bl/2 hybridomas adapted to serum-
free medium, showing the increase in viable cell number (# ),
and the specific growth rate (^)
The maximum specific growth rate of serum-free cells (0.042h*1) was slightly
lower than the ^max for serum-dependent cells of O.CHÇh’1 (3.2.1). This lower
rate may be due to the cells devoting energy to synthesis of components for
maintenance; components normally present in serum.
The serum-free cells maintained à lower viability than the serum-
dependent cells. It has been reported that serum affords physical protection to
cells in culture (Handa et al 1989), and provides fatty acids, essential for the
synthesis of prostaglandins and membranes, and hence maintenance of membrane
integrity. By omitting serum and not replacing the proteins and lipids,
membranes of the cells may be more prone to "stress'*. If the membrane
integrity is compromised, then the cells may be more permeable to the dyes
used in the determination of cell viability, which will appear lower. This
observation pattern of a reduction in viability and extended lag phase are
characteristic of many serum-free media (Glassy et al, 1988).
85
SPEC
IFIC
GR
OWTH
RA
TE
(h-D
fxl
Monoclonal Antibody Production
The rate of Mab production peaked at 63 hours (Figure 3.10b), and the
maximum Mab titre was higher in serum-free medium (64^ig/ml) than in serum-
supplemented medium (44^ig/ml)..
75
dP/dt
60 -
45
30
1 0 08060402 0O
§Ea
3 ,HI
OC
ob-
oLL
a
oDCCL
TIME (hours)
Figure 3.10bThe volumetric production of Mab ( a ), by PQX Bl/2 cell
adapted to serum-free medium, and the rate of product
formation (dP/dt)
This may be due to the absence of components found in serum which
■’* may inhibit secretion of the Mab. Proliferation and differentiation of lymphocytes
and subsequent secretion of IgG are mediated by chemical signals which interact
with cell surface receptors. Semm-proteins complicate the biology of these
86
effector molecules (Glassy et al, 1988), and thus a similar effect may occur in
Mab synthesis by hybridomas in serum-supplemented medium.
The peaks in fi and dP/dt in the serum-free cells were separated in
time, with some overlap (Figure 3.10c).
ytT -
k
111
sI
1
oLL
Ü &
Figure 3.10c The relationship between specific growth rate (ji) and product
formation rate (dP/dt) in PQX Bl/2 hybridomas adapted to
serum-free medium.
This pattern is very similar to the serum-dependent cell growth characteristics
discussed in 3.2.1 and 3.2.2, and suggests that Mab production by PQX Bl/2
' in serum-free medium, has a "complex association" with growth rate as is the
case with serum-dependent cells, but this may not necessarily imply the same
relationship.
0 .5 0
0 .4 0
dP/dt
0 .3 0
0.20
0 .1 0
0.001 0 04 0 6 0 8 02 0
TIME (hours)
87
PROD
UCT
FORM
ATIO
N RA
TE
jpg/
ml/h
)
The Mab titre levelled off towards the end of the culture, and this may
be due to one of two things. Even though antibody is not a primary metabolite
(Glassy et al 1988), manufacturing a large protein does require energy. It is
therefore possible that due to energy shortage, Mab synthesised and transported,
may not be completed efficiently, resulting in the release of incomplete, and
hence non-functional antibody. Also, the death and lysis of hybridomas would
invariably lead to increased proteolytic activity in the culture medium, which
could result in the generation of non-functional antibody. This non-functional or
incomplete Mab may not be detected by ELISA.
Amino Acid Utilisation
Patterns of amino acid utilisation were similar in both cell lines, in that
there were peaks in utilisation of most amino acids in phase I of the culture
and again during mid-late exponential growth (at the beginning of the production
phase - Phase III). It has been suggested (Luan et al, 1987), that the exhaustion
of essential amino acids, is the primary cause of the onset of the death phase.
The results presented so far indicate that essential amino acids (apart from
GLN), are not depleted, there may however be a critical level, below which
death ensues.
The utilisation of TYR, VAL, and LYS was greater in serum-free
medium than in serum-supplemented medium, in which there was greater
utilisation of SER, GLY, ARG, PRO, MET, ILE, LEU, and PHE (Table 3.1).
The difference may be a reflection of the different jwmax value, or it may be
due to transport differences, ie, some amino acids may be taken into the serum-
free cells more easily because of fewer inhibitor proteins (normally found in
serum, and thought for example to inhibit Mab secretion, Glassy et al, 1985).
88
Table 3.1 The difference in amino acid assimilation between serum-free
and serum-dependent cells
AMINO ACID % CHANGE
SERUM-FREE SERUM-DEPENDENT
ASP + 18.3 -0.8
GLU +93.1 +90.7
SER -17.5 -32.45
GLY -6.2 -9.91
ARG -4.1 -19.5
THR -38.9 -64.7
ALA +259.3 +360.2
PRO -3.5 -10.2
TYR -55.6 -46.1
VAL -75.0 -69.8
MET -43.7 -50.7
ILE -52.7 -60.5
LEU -56.8 -68.8
PHE -36.7 -37.7
TRY +40.2 +262.2
LY -64.8 -62.3
- UTILISATION
+ PRODUCTION
89
Adenylate Charge and ATP Levels
The ATP concentration peaked at 60 hours (Figure 3.10d).
5 0 0
4 0 0
H 3 0 0
200
1 0 0
1 0 08 06 04 02 0O
TIME (hoirs)
Figure 3.10d The change in ATP production
Adenylate (E.C) peaked at 35 hours corresponding with the peak in
growth rate, and then again at 65 hours coinciding with the peak in Mab
production (Figure 3.10e). This suggests that, as may be expected, an increased
amount of energy is required for growth and production. This may have a role
1 in feedback control, in that as the demand for energy increases so does the*
supply.
90
1.00
LU 0.80(5
0 0.60 -
I 0.40 -
0.20
0.00 '-' ' L-----------------------------------------10 20 40 60 80 100
TIN/E (hours)
Figure 3.10e The change in adenylate charge
A trajectory plot of specific growth rate against energy charge shows the
relationship between growth rate, B.C., and Mab production (Figure 3.10f).
EC
1.0
0.6 •
0.4
0.2
0-040 0-01 0-02
SPECIFIC GROWTH RATE ( h"1)
Figure 3.10f A trajectory plot for serum-free (A) and serum-dependent (B)
hybridoma cells. The period of Mab production is also
shown(— )
91
There is particular region of the graph during which Mab is produced.
The E.C. in this region has a value between 0.8 and 0.6. As the E.C. declines
further so the synthesis of Mab stops. A similar trajectory plot for serum-
dependent medium indicated a similar relationship, suggesting that the complex
relationship between growth and Mab production may involve changes in E.C.
and may be independent of the medium.
During growth and Mab production, amino acids are utilised. These have
roles as precursors for protein synthesis and regulation of RNA synthesis, which
means that an increase in uptake may lead to an increase in the rate of
macromolecular synthesis. The resulting, more rapid, rate of utilisation of
nucleotide triphosphates will tend to decrease the concentration of the nucleotide
pool. Through the action of end-product negative feed-back control, the rate of
synthesis of nucleotides will increase to bring them into balance with the rates
of utilisation (Atkinson, 1977). Thus peaks in amino acid assimilation will occur
in tandem with periods of high macromolecular synthesis (Pre-growth and Mab
production), and peaks in energy charge.
The numerical value of E.C. in metabolising cells is thought to be very
similar for all cell types, and also to change very little with changes in
metabolic conditions (Atkinson, 1977). The results of this experiment do not
bear this out. One way of causing a decrease in the value of the energy charge
is to increase the utilisation of ATP by addition of a substrate for a kinase. If
the phosphorylated product can be metabolised to lead to the regeneration of
ATP, the E.C., can be expected to recover promptly. If the product is not
metabolised then there may be a longer period over which the energy charge
falls. This may, in part, account for fluctuation in the PQX Bl/2 B.C.. There
may also be increased production of AMP which will also cause the energy
charge to fall.
During steady-state growth of cells when all other regulatory parameters
are also at steady state, the E.C. is very close to 0.9 (Atkinson, 1977). In a
batch culture, conditions are constantly changing and so are the cells, and thus
92
there is no steady-state. Such fluctuations in the cells and their environment may
make a significant contribution to fluctuations in E.C.
The rates of many or most ATP-requiring processes must be sharply
reduced when the E.C. falls below it’s "normal" range. It seems likely that
there is a hierarchy of such processes in terms of their response to the value
of E.C.. Energy-storing pathways, such as synthesis of polysaccharide or fats,
should be the most sensitive to a decrease in E.C.. Biosynthesis of structural
macromolecules should be next, however, activities which are essential for
maintenance of the cell should be able to function at lower values of energy
charge. The concentration of metabolic intermediates and metabolic starting
materials decreases during a batch culture in parallel with a decrease in the
ability to generate ATP. It is therefore difficult to estimate to what extent
hierarchical metabolic pathways depend on the adenylate system. Hence, it is
difficult to propose a simple overall mechanism of control for cell growth,
Mab production and E.C.. It may be that ’switching-on’ of Mab synthesis
results from a lowering of E.C. causing ATP utilising pathways to be reduced.
For instance, the reduction in activity of an enzyme or metabolite concentration
in this pathway may activate the genes necessary for Mab production.
A more complex explanation may involve regulation of the cell-cycle,
and this will be discussed in more detail in Chapter 6.
93
4. PROCESS DESCRIPTION USING MATERIALS BALANCING
4.1 INTRODUCTION
4.1.1 The Uses of Materials Balancing
Many industrial fermentations and bioreactor processes are carried out
in fed-batch mode. This has many advantages over simple batch culture and
may be effective overcoming substrate inhibition, catabolite repression, and
product inhibition.
The main object of fed-batch culture is to extend or enhance product
formation. This requires a knowledge of production kinetics to establish when
product formation occurs, and knowledge of nutrient utilisation profile to identify
components necessary for production. Thus one needs to know what to feed,
when to feed it and at what rate.
Materials balancing can provide the following:
(i) A real time indication of the progress of the process
(ii) Real time indication of process improvements
(ii) Feed-back parameters for on-line optimisation.
Thus material balancing could be used to facilitate definition of the fed-
batch parameters of composition and time of feed, and give a more detailed
understanding of the processes of growth and product formation.
Materials balancing has been used to characterise sequential secondary
metabolite formation (Bushell et al, 1983) and has proved useful in gathering
information on substrate utilisation, product accumulation, yield and efficiencies
in bacterial cultures (Acevado, 1987).
94
4.1.2 The Concept of Materials Balancing
Growth of a hybridoma can be represented by a general equation such
as:-
Glucose + Amino Acidl + 0 2 + Factors ------ ► Cells + COz + H20 +
Amino Acid2 Metabolite + NH3 +
Lactate
This can be converted into an empirical equation (Equation 2) thus:-
aC6H120 6 + bC5H10N2O3 + cOz —-----------► CHL6O0 N0.2 + dC02 + eH20
fC6H14N20 2 C2-5H2Oo-2No.i8 + gNH3 +
iC3 H6 0 3
This equation balances according to the law of conservation of matter.
Thus the equation can be balanced by the 0 2 and C02 values. Therefore in
this case the respiratory quotient (RQ) is d/c.
Where RQ = C02 Produced
0 2 Consumed
These equations are valid when the culture is growing but not producing
Mab. When Mab is being produced then there is an extra component on the
right hand side of the equation (Mab), and thus the equation must ’re-balance’.
Thus:
Glucose + Amino Acidl + 0 2 ► Cells + C 02 + H20 +
Amino Acid2 Metabolite + NH3 +
Lactate + Mab
95
The equation may of the following form:-
aC6H120 6 + bC5H10N2O3 + x 0 2
fC6H14N20 2
> CH1.6Oo.5No.2 + yC02 + eH20
C2sH2O0i2N0.i8 + gNH3 +
iC3H603 + j C H12.2N1802.5
-2 1X2V,0.21'I0.18etc.
In this case the RQ changes to x/y. There may theoretically therefore
be a change in RQ value as the culture changes from growth to product
formation. This may only be true if growth and Mab production are separated
in time. If there was an overlap then there may be an RQ value associated
with this; providing an on-line indicator of the process progress. This therefore
enables a real-time estimation of the on-set of product formation.
The equations for growth and Mab production can be compared to
determine the utilisation of various components and assess their importance in
production. This enables a feed to be formulated, and a fed-batch culture used
on a more informed basis. Fed-batch cultures can be formulated without
materials balance equations but require numerous trila and error determination.
In order to derive a materials balance there are a number of parameters
which need to be investigated. These include nutrient assimilation profiles,
microanalysis of ’unknowns’ (eg Biomass), and gas analysis (Figure 4.1).
96
GAS ANALYSIS SAMPLES FROM
GROWTH/PRODUCTION
PHASE
Figure
C 02, 0 2, RQ
EXCHANGE PROFILE
Biomass
Supematant-
MICRO ANALYSIS
(C,H,0,N)
RQ v dP/dt
RELATIONSHIP MATERIALS BALANCE
EQUATIONS
PROCESS DESCRIPTION
TIME OF FEEDi _______
FEED FORMULATION
FED-BATCH
ANALYSES
Glucose, Mab,
Lactate, NH*+
Amino Acids
UPTAKE/PROD
RATES
FEED-RATE>
4.1 Components and Uses of Material Balancing 97
4.2 MATERIALS BALANCE DESCRIPTION OF GROWTH AND
MONOCLONAL ANTIBODY PRODUCTION BY POXB1/2
A series of 101 bioreactor batch cultures of the PQX Bl/2 hybridomas
were carried out as previously described (2.1.7). Samples were taken at regular
intervals (6-8 hours), and the viable and total cell densities determined (2.1.4).
The following analyses were performed depending on the experiment: Mab
(2.2.4), amino acids (2.2.8), glucose (2.2.7), lactate (2.2.6) and GLN/NH4
(2.2.5). On-line 0 2 and C02 gas analysers were interfaced to a BBC
microcomputer, and RQ values calculated and logged.
4.2.1 Gas Analysis of POX Bl/2 Cells In Batch Culture
Results and Discussion
Cell Growth
After a lag phase of 40 hours, the hybridomas followed a typical growth
curved, and reached a maximum specific growth rate of 0.053h*1 (65 hours,
Figure 4.2a). The long lag phase was probably due to the low inoculum
density, as discussed previously.
Monoclonal Antibody Production
Mab was produced throughout the culture; the peak in production rate
occurred at 90 hours (Figure 4.2b); 30 hours after the peak in growth rate
(Figure 4.2c), corroborating the results of previous experiments. The complex
relationship between growth and Mab production was again eluded to by the
overlap of Mab production and growth.
98
I4O
o
Uj
>
12.00 6.00
9.60
7J20 3.60
.4.80 2.40
2.40 1.20
0.00 0.000 25 50 75, 100 125
TIMEÛtolts)
Figure 4.2a The typical growth curve for PQX Bl/2, showing the
increase in viable cell count ( • ) , and the specific growth rate
t o
18.00 0.35
dP/dtE 14.40
HIŒ 10.80 H
!
0.28
0.21
0.14
3.60 0.07
0.00 0.00O 25 50 75 100 125
§£
F
81
TIME (hours)
Figure 4.2b The production of Mab ( • ) , and the major peak in
product formation rate (dP/dt)
99
SPEC
IFIC
GR
OWTH
RA
TE
6c 10
-2
h-1)
gX 6.00 035
4.80
3.60 0.21
g 2.40
0LLO 1.20LU&
0.14
0.07
0.00 0.000 25 50 75 100 125
ü
!
T1NÆ (hoirs)
Figure 4.2c The relationship between specific growth rate ( t) and
product formation rate (dP/dt)
Gas Analysis
Due to C02 being released from the HC03" in the medium as the culture
was sparged, the ratio of C02 produced to 0 2 consumed was high, giving an
artificially high value. The subsequent immediate decline in apparent RQ was
as a result of all the C02 being vented due to sparging, and growth rate being
very low. There then followed a slow decline in RQ, before a peak at about
39 hours, which corresponded with d02 tension reaching 40% air saturation
(Figure 4.2d and 4.2e).
This RQ value also suggests that there, was 0 2 consumption without
much C02 evolution, this was corroborated by the fall in d02 tension. This
implies that there was efficient aerobic biosynthetic activity
100
3.00
2.40
Ia 1.80
Ih 1.20
0.00 125100755025
Figure 4.2d The change in respiratory quotient (RQ)
100
80</)
H60
0 40
1y so
Û
75 12510O50250
TIME (hours)
Figure 4.2e The change in dissolved oxygen tension
101
Co*
PROO
UC
Btt
cy.>
During the growth phase there was little change in the RQ value, even
though the air flow rate was increasing to meet the demands for 0 2 (Figure
4.2f). At the onset of the maximum Mab production phase, RQ levelled out at
about 1.0. This then fell later in the culture because there was no demand for
air, and hence no gas flow through' the system to the gas analysers. The air
flow was switched to a manual bleed of O.Ollmin"1 which is indicated by a dip
at 116 hours (Figure 4.2f). The RQ then returned to a value of about 1.0.
12!Ü<:o
çr<
o #’ 0 25 f 12550 75 100
TlfvE (hours)
Figure 4.2f The change in air flow rate, the flow rate was changed to a
manual bleed at 166 hours ( t ) .
The results from this experiment indicate that there is an RQ value
associated with Mab production of approximately 1.0, which enables further
investigation of the derivation and use of material balancing with the PQX
Bl/2 hybridomas.
102
4.2.2 Derivation of a Materials Balance Equation
The gas analysis experiment (3.3.2) was repeated. At two times points
(determined by viable cell density and RQ value), before and during maximum
Mab production, 2 samples (200ml) were taken three hours apart. 100ml of cell
suspension was centrifuged at 200g for 5 minutes and the supernatant decanted
into a sterile 250ml duran bottle (Schott). The supernatant was freeze-dried using
an acetone/dry-ice bath with a vacuum dryer. The sample was then stored
desiccated at 4°C. The pellet of cells was washed by resuspension in 10ml of
sterile 0.85% (w/v) NaCl in MQ. The cells were harvested by centrifugation
at 200g for 5 minutes, the supernatant poured to waste and the washing
repeated twice more. The cell biomass was then freeze-dried as described, and
together with the freeze-dried supernatant was analysed for elemental carbon (C),
hydrogen (H), oxygen (O), and nitrogen (N) composition by the Chemistry
Department, University of Surrey. In order to determine the C, H, O, N,
content of pure anti-paraquat Mab for use in the materials balance equation
several milligrams of purified Mab are required. Purification of such a quantity
of Mab free from a C H O N- containing solution is difficult. There is a high
percentage of homology of C H O N content between different murine IgG and
thus a commercially purified murine IgG (Sigma) was analysed. Analysis of
supernatant metabolites was as before. The equations were then derived.
Results and Discussion
Cell Growth
The culture followed the phasic pattern as previously described.
A materials balance equation for the growth period of the culture was
derived from the assimilation and production profiles of nutrients and biomass
for the period 20-23 hours (Table 4.1a). The figures in the Tables 4.1a and
b represent the milli-molar changes in concentration of each component during
103
the three hours between the samples. In order to represent the significant
changes observed, experimental data of some of the components, it is necessary
to show 4 decimal figure places. The total figures were obtained by summing
the full eight figure value, thus avoiding rounding errors which can be
significant for some of the parameters used.
The equation for growth (Table 4.1a) accounted for 87% of the carbon
exchange, but only 51% of the nitrogen. Comparison of the microanalysis of
supernatants revealed no apparent difference (Table 4.1a).
The discrepancy in the balance of nitrogen during growth phase may be
due to the production of a nitrogen containing compound which was not assayed
individually. This proposal is supported in that the percentage nitrogen in the
supernatant at 20 hours is very similar to the percentage nitrogen at 23 hours.
104
Table 4.1a Consumption and appearence of maj or reaction
components between 20 and 23 hours (RQ = 1.9)
IN m M O U T m M0 2 0.2445 C 02 0.4646
Q H A 0.5023 CH1.74N0.2S0o.450.5026(Glucose) (Biomass)
NH, 0.3750 CHi.7NojO„j 0.0656(Antibody)
C5H,0O,N2 0.0910 CA O , 0.9750(Glutamine) (Lactate)
CAiOjN 0.0077 C3H70 2N 0.0890(PHE) (ALA)
CA .O sN 0.0069 C„H120 2N2 0.0133(TYR) (TRY)
C9H1103N 0.0064 C5H ,04N 0.0156(VAL) (GLU)
C A A N 0.0040 C3H70 3N 0.0103(LYS) (SER)
CAOjN 0.0057 C A 0 4N 0.0071(GLY) (ASP)
C A A N S 0.0001 C A 0 2N 0.0068(MET) (PRO)
C6H130 2N 0.0061 C A 0 3N 0.0038(LEU/ILE) (THR)
TOTALSC 3.68519951 3.19267470N 0.59404405 0.30446757
Culture Supernatant:20b C(46.27%): H(7.02%): N(11.10%): 0(35.61%)23b C(46.58%): H(7.25%): N(11.42%): 0(34.75%)
105
Monoclonal Antibody Production
During the production phase (45-48 hours), a second materials balance
equation was derived (Table 4.1b). Less carbon assimilated could be accounted
for in the carbon produced. Similarly, the apparent difference in nitrogen
balance is not reflected in the supernatant microanalysis (Table 4.1b). The Toss’
of carbon and nitrogen in the production phase may be due to conversion into
cell debris or lysis products since no net increase in biomass products was
observed at this point.
Amino Acid Assimilation
The assimilation rates of amino acids were higher in the production
phase, with GLN, ARG, and ALA being utilised most, and are consistent with
other experimental results. This suggests the increased uptake was to meet the
macromolecular biosynthetic demand imposed on the cell during Mab production.
ARG was incorporated at a rate second only to GLN during Mab production,
suggesting that it may be important.
The results confirm that there is a significant difference in RQ between
growth and Mab production, coupled to this is a difference in nutrient
assimilation. The materials balance equation give an accurate assessment of the
processes of growth and Mab production, and alude to important components
for each. From these balanced equations a feed can be formulated for a fed-
batch culture, and the feed-rate set to the utilisation rates of the nutrients. The
RQ profile enables the time at which the feed should be switched on to be
determined (ie when the RQ reaches 1).
When the RQ value falls to 1 (after peaking at 2.40), then this is indicative of
the onset of the maximum Mab production phase and therefore when the feed
should be initiated.
106
Table 4.1b Consumption and appearence of major reaction
components between 45 and 48 hours (RQ = 0.91)
IN m Mo 2 1.2368CfrHyO,; 0.3885(Glucose)n h 3 0 .2200
c 5h 10o 3n 2 0.1140(Glutamine)
0.0187(PHE)c 6h 14o 2N4 0.0793(ARG)C A 4Q2N 0.0063(LYS)c 4h 9o 3n 0.0214(THR)c 3h 9o 2n 0.0216(PRO)c 3H7o 2N 0.0765(ALA)C Æ A N 0.0469(LEU/ILE)c 4h 7o 4n 0.0216(ASP)c 5h 9o 4n 0.0214(GLU)c 3h 7o 3n 0.0314(SER)
TOTALSCN
4.383126770.99663455
Culture Supernatant:
OUTco2
c h 17n mo 0.5(Antibody)c 3h 603(Lactate)
2.770400000.45000000
45b C(48.59%): H(6.94%): N(11.64%): 0(32.83%)48b C(45.52%): H(6 .8 6 %): N(13.76%): 0(33.86%)
mM1.1255
0.1506
0.4550
107
Based on the results presented above, the feed could be employed (Table
4.1c) at a rate of 20ml/hf1.
Table 4.1c Proposed Feed Formulation For A Fed-Batch Culture
Based On The Materials Balance Equation
COMPONENT FEED RATE (mM/l/hl
Glucose 0.1295
NH3 0.0733
GLN 0.038
PHE 0.0264
ARG 0.0264
LYS 0.0021
THR 0.0071
PRO 0.0072
ALA 0.0255
LEU/ILE 0.0156
ASP 0.0072
SER 0.0105
GLU 0.0080
108
5. NITROGEN METABOLISM
5.1 INTRODUCTION
Antoine Lavoisier gave the name ’azote’ to the gas nitrogen when he
discovered it in 1787, meaning without life, because of both it’s lack of
chemical reactivity and its inability to support life. Nevertheless, the metabolism
of nitrogenous compounds is central to the metabolic processes of all living
organisms. This is highlighted in the previous chapter by the high percentage
of nitrogen in the biomass of the hybridomas, and the presence of many
nitrogen containing components in the proposed feed for fed-batch culture.
Microorganisms and plants are capable of using simple inorganic nitrogen
compounds such as nitrogen gas, ammonia, nitrites and nitrates, and
incorporating these into such organic compounds as amino acids, purine and
pyrimidine nucleotides, amino sugars and a variety of co-factors and coenzymes
that are vital for animals. Animals are not able to make use of inorganic
nitrogen compounds in the diet. These are taken largely as proteins and amino
acids, although dietary nucleic acids and other nitrogenous compounds can also
be used to a greater or lesser extent. The waste products of nitrogen
metabolism excreted by animals are generally relatively simple organic
compounds (urea and purines), together with moderate amounts of ammonia and
other inorganic salts.
5.1.1 Amino Acids and Animal Cells in Culture
Amino acids play a diverse role in the metabolism of cells. Their direct
participation in the synthesis of the structural proteins and enzymes, relate to
all aspects of metabolism. Their transport, catabolism, anabolism, and ultimate
utilisation potentially have an effect at all levels of growth and metabolism.
All cell culture media contain amino acids, but there is wide variation
in the number and concentration. The early tissue culture media, derived from
109
body fluids and tissue extracts, were rich in amino acids, and thus the
importance of the amino acids was not realised at first. Investigations by
Burrows and Neyman (1917) led to ideas that amino acids were toxic to cells
and inhibited growth. These conclusions were logical at the time; the results
being due to the additions of inordinately high concentrations of amino acids.
This toxicity enigma persisted for some time amongst tissue culture workers. It
was not until the work of Rose (1938) that the importance of amino acids in
the nitrogen balance of animals was realised. The first analytical approach was
taken by Fischer (1941), who dialysed serum in Ringer’s solution containing
glucose, and found that dialysed culture medium would not support cell growth.
Using this medium as a base and systematically adding amino acids, he
investigated their effect on cell growth. He discovered that more amino acids
were required by cells in culture than cells in vivo, and that cysteine was the
most important amino acid. This work was extended by Eagle who studies the
specific amino acids requirements of L and HeLa cells (Eagle, 1955a; 1955b)
using simple basal medium supplemented with dialysed serum. Eagle
distinguished between thirteen "essential" amino acids: arginine (ARG), cysteine
(CYS), glutamine (GLN), histidine (HIS), isoleucine (ILE), lysine (LYS),
methionine (MET), phenylalanine (PHE), threonine (THR), tryptophan (TRY),
tyrosine (TYR), and Valine (VAL), and seven "non-essential" amino acids:
alanine (ALA), aspartate (ASP), glycine (GLY), glutamic acid (GLU),
hydroxyproline (OHP), proline (PRO), and serine (SER). The results for HeLa
and L cells have been confirmed by other investigators but there are exceptions;
Jensen sarcoma cells ( McCoy et al, 1959a), Walker sarcoma 253 cells (McCoy
et al, 1956), both require asparagine (ASN) in addition to the thirteen
"essential" amino acids. Novikoff hepatoma cells (McCoy et al, 1959b) required
glycine, and rabbit fibroblasts RM356 (Haff and Swim, 1957-58) required SER.
Morgan and Morten (1957) found that chick embryonic heart fibroblasts
did not require ILE or GLN. The work of Griffiths and Pirt (1967) with LS
cells showed that the cell yield could be increased by the addition of a mixture
of "non-essential" amino acids, and substitution of GLN by GLU could also
110
increase cell yields. More recently work on BHK cells (Arathoon and Telling,
1982) showed that the amino acids utilised in the greatest quantity were; GLN,
GLU, LEU, ASN, SER, and ILE. GLN was also seen to be primarily utilised
by Adamson et al, (1987), but was followed by ARG, LEU, ILE, LYS, VAL,
THR. These observations are in good agreement with other workers (Birch et
al, 1984; Butler and Thilly, 1984). Not only do the requirements of different
cells vary, but also the ability to act upon medium components. Thus, some
amino acids exert sparing effects on others by active interconversion,
transamination, and deamination of keto acids (Patterson, 1972).
5.1.2 The Transport of Amino Acids Across Membranes
Amino acids are charged molecules under physiological conditions and
therefore are not freely permeable across lipid membranes. They are accumulated
in cells by both active transport mechanisms, in which there is a net
accumulation of amino acid to a greater concentration than in the extracellular
fluid, and also by facilitated diffusion, when the intracellular concentration is
equal to that outside. Both of these processes are carrier mediated, as is the
efflux of amino acids from cells. A number of different carriers of amino acids
have been identified in various mammalian cells, differing largely in their
specificity for the amino acids transported, and also in whether or not transport
or linked to Na+ ions. The principal amino acid transport systems are listed in
Table 5.1a. Mammalian amino acid transport has been studied in three systems:
the intestinal mucosa, the renal tubule epithelium and across the blood-brain
barrier. In general, the same carrier systems are present in all three tissues, and
it is assumed that results obtained from these tissues are also applicable to the
transport of amino acids in other tissues.
Ill
Table 5.1a Amino Acid Transport Systems In Mammalian Tissues
(Bender, 1985)
Systems for neutral amino acids (mav also carry basic or dicarboxvlic
amino acids at favourable pH)
Na+-dependent
System Gly
System A
System ASC
System N
Gly and sarcosine
Most neutral amino acids
Most amino acids
HIS, GLN, ASN; hepatocytes and hepatoma
cells
Na+-independent
System L Preference for Branched non-polar side-chain
System L, and L2 Branched non-polar side-chain; hepatocytes
Systems for Cationic Amino Acids (Na+ -dependent)
ARG; liver mitochondria
System Ly, and Ly2 LYS; intestinal mucosa
System y+ Diamino acids
Systems for Anionic Amino Acids
Na+-dependent
System xA*
System xG'
Na+-independent
System xc‘
ASP; renal tubule
GLU; renal tubule
GLU and CYS
112
5.1.3 The Role of Vitamin B< in Amino Acid Metabolism
The metabolically active vitamer of vitamin B6, pyridoxal 5’- phosphate
(PLP), is the coenzyme for enzymes catalysing a number of different reactions
of amino acids including decarboxylation to amines, racemization, transamination
to a-keto-acids, and various elimination and replacement reactions of the side
chain.
5.1.4 Amino Acid Catabolism
For the most part, cells in culture utilise amino acids through established
metabolic pathways (Meister, 1965). The a-amino group of many amino acids
is transferred to a-ketoglutarate to form glutamate.
H HI I
H3N - C— R ------------► H3N-C— CH2— COOI ICOO coo
Amino Acid Glutamate
The transfer of an a-amino group from an a-amino acid to an a-keto
acid is catalysed by aminotransferases or transaminases. One of the most
important of these is aspartate aminotransferase which catalyses the transfer of
the amino group of aspartate to a-ketoglutarate (a-KG).
ASP + a-KG v Oxaloacetate +GLU
In a similar manner alanine aminotransferase catalyses the transfer of the
amino group of alanine to a-KG.
ALA + a-KG ^ Pyruvate + GLU
113
Thus GLU is catabolized by way of a-KG; it can undergo either
transamination or oxidative deamination, depending largely on the carrier by
which it enters the mitochondrion. When GLU enters on the Glutamate/Aspartate
counter-transport mechanism, it will largely be transaminated so as to maintain
a supply of ASP to permit continued uptake. This does not represent a net
catabolism of GLU, since an equivalent amount of oxaloacetate leaves the pool
of citric acid cycle intermediates. By contrast, when GLU enters the
mitochondrion in exchange for OH* ions it is largely deaminated by glutamate
dehydrogenase (GDH), and in this case there is net oxidative catabolism of the
resultant a-KG.
GLU a-KG+ NH3
The equilibrium of GDH is determined by the relative concentration of
GLU, a-KG, N H /: when the concentration of NH4+ is high then a-KG will
be withdrawn from the citric acid cycle pool to synthesize GLU, whereas when
the NH4+ concentration is low and there is a relatively large amount of GLU
available, it will be deaminated.
GDH is allosterically regulated, with GTP and ATP being allosteric
inhibitors, and GDP and ADP being allosteric activators. Hence, a lowering of
the energy charge (B.C.) accelerates the oxidation of amino acids.
Where E.G. = [ATP]+i/2[ADP]
[ATP] + [ADP] + [AMP]
In liver some of the ammonia is assimilated into the urea cycle.
5.1.5 The Urea Cvcle
The formation of carbamylphosphate from NH3, HC03* and ATP is
generally considered the first step in the urea cycle. The carbamyl group is then
114
transferred to ornithine forming citrulline which reacts with aspartate to form
argininosuccinic acid. This is then cleaved to form fumarate and arginine, from
which one molecule of urea is removed, regenerating ornithine for the next turn
of the cycle. The urea cycle, the intermediates and enzymes involved are shown
in Figure 5.1.
NH,+
*CarbamylPhosphate Ornithine Carbamyl Synthetase Transferase
*CARBAMYLPHOSPHATE
CITRULLINE ASPARTATE
Arginino Succinate Synthetase
ORNITHINE ARGIMNOSUCCINATE
Arginase
Argininosuccinase
ARGININE FUMARATE
Figure 5.1 The Urea Cycle: Intermediates and Enzymes
115
5.1.6 The Carbon Skeleton of Amino Acids
Once the a-amino group is removed the resulting carbon skeleton is
converted into major metabolic intermediates. These intermediates are pyruvate,
acetyl Co-A, fumarate and oxaloacetate (Figure 5.2).
AlaGlyCysS e rT hrT rp Leu
Lys.PheT yrT r p
G L U C O S E
HeLeuTrp
p y r u v a t e
PHOSPHOEIM O L P Y R U V A T E
A C E T Y L Co A
AC ETOAC ETYL
Co A
AsnA sp O X A L O A C E T A T E
T y rP heA sp
CITRATEF U M A R A T E
GinGluHisProArg
HeMetVal
S U C C IFNYL Co A
Figure 5.2 Fate of the Carbon Skeleton of Amino Acids
116
5.1.7 The Anabolism of Amino Acids
Amino acids are classically grouped into those which can and cannot be
synthesised by humans: the "non-essential" and "essential" amino acids
respectively. This classification differs from that of Eagle in that some amino
acids (GLN, and ARG) designated as "essential" by Eagle, are classically termed
"non-essential".
The a-amino group of most amino acids comes from the a-amino group
of GLU by transamination. GLU is synthesised from NH4+ and a-KG by the
action of GDH. This is a major pathway for the incorporation of NH4+. There
are a great number of aminotransferases in all organisms that can transfer the
amino group from an amino acid onto an oxo-acid and many use the
glutamate/a-ketoglutarate pair as the amino donor. This means that ammonia
incorporated into glutamate can be used for the synthesis of any amino acid
for which the corresponding a-oxo acid can be synthesised, and for which the
organism has a suitable aminotransferase (Bender, 1985).
The synthesis of the "non-essential" amino acids is relatively simple:
ALA and ASP are synthesised by transamination of pyruvate and oxaloacetate
respectively. PRO and ARG are derived from GLU, while SER, formed from
3-phosphoglycerate, is the precursor of GLY and CYS. TYR is synthesised by
the hydroxylation of PHE.
The pathways for the biosynthesis of the "essential" amino acids are
much more complex. Most of these pathways are regulated by feed-back
inhibition, in which the committed step is allosterically inhibited by the final
product.
5.1.8 Formation of Amino Acid Amides
A second major pathway by which ammonium ions can be incorporated
into cellular metabolites is by the formation of amino acid amides. Both the
dicarboxylic acids, glutamate and asparagine respectively can form w-amides:
117
glutamine and asparagine respectively. Both GLN and ASN synthetases are
widely distributed, both are ATP-requiring reactions.
The synthesis of GLN is especially important as a means of reducing the
circulating concentration of NH4+ ions in the mammalian blood stream. The
ammonium ions are highly toxic; the glutamine is a soluble relatively non-toxic
intermediate carrier of nitrogen, as well as being required for protein synthesis.
It also a primary source of nitrogen in cultured mammalian cells, donating
nitrogen atoms in a number of biosynthetic reactions, including the synthesis
of purines, pyrimidines, amino sugars, TRY and HIS. In addition GLN provides
an additional carbon and energy source, and may provide upto 65% of the cell
energy requirements (Zielke et al, 1984).
5.1.8 Physiological Effects of Ammonia on Animal Cells
As described above catabolism of amino acid yields ammonia. The
importance of GLN is such that 2 to 4mM is added to cells in culture. The
major portion of ammonia which appears in cell culture media is as a result
of catabolism of GLN. In general there is molar conversion of GLN to
ammonia.
Ammonia has been proposed as potential "toxic" waste product (Butler,
1985; Dean et al, 1984; Jessup et al, 1983; Reuveny et al, 1986)
Ammonia accumulation has be shown to increase the susceptibility of
suspension cultures to hydrodynamic stress (Petersen et al, 1988). Ammonium
has also been shown to inhibit the growth of mouse L cells (Ryan and Cardin,
1966), canine kidney (MDCK) cells (Glacken et al, 1986), production of a
number of viruses (Furasawa and Cutting, 1962; Jensen et al, 1961) as well as
interferon (Commoy-Chevalier, 1978; Ito and McLimans, 1981).
The mechanisms for ammonium inhibition are not known. It was
proposed by Glacken et al (1988), that ammonia’s inhibitory effect may be the
result of the destruction of electrochemical gradients that may occur due to
ammonia’s basic nature. Implicit in this hypothesis is the premise that unionized
118
ammonia is the actual inhibiting moiety, and not the ionized ammonium ion.
Gaseous ammonia can diffuse rapidly through the cytoplasmic membrane,
whereas the ionised form cannot (Benjamin et al, 1978). Once the unionised
molecule diffuses across the membrane, it may pick up a hydrogen ion to form
the ionized ammonium. In this way, the pH of the cytosol and sub-cellular
vesicles may be increased. It has been shown that NH4C1 inhibits the
spontaneous secretion of lysozyme by macrophages and may also interfere with
endocytosis (Jessup et al 1983). Seglen and Reith (1977), found that the output
of proteins by cultured hepatocytes was also inhibited by ammonium chloride.
Weak bases, such as ammonia, causes vacuolation in cells due to their
accumulation within the cell in regions of low pH. Lysosomes are the major
site of accumulation, however there may also be accumulation in endosomes and
phagosomes (DeDuve, 1974). The rate of accumulation depends upon the Pk and
Pka of the amines involved. The pH of several of the intracellular compartments
is raised by the uptake of amines. This cause vacuolation and rapid loss of
cellular integrity and viability. It has also been shown that enhanced or retarded
incorporation of precursors into macromolecules depends upon the ammonium
concentration and the cell type (Zimber et al 1969; 1970). It may also be that
NH4+ interferes with a-KG utilisation. This may lead to an decreased citric acid
cycle and thus decreased oxidative metabolism. High level of NH4+ would lead
to increased formation of GLU effectively removing a-KG.
5.2 ASSIMILATION OF AMMONIA BY POX BI/2
The results from chapter 3 suggested that during maximum growth of
PQX Bl/2, ammonia was assimilated. The mechanism for this assimilation is
not clear as there are several potential pathways through which cells can
assimilate ammonia. These include the urea cycle, and via the action of
glutamate dehydrogenase or glutamine synthetase.
The following series of experiments aimed to clarify the pathway(s)
responsible for ammonia assimilation.
119
The experiments were carried out as 101 bioreactor batch cultures as
described previously. In addition to viable and total cell counts, GLN/NH4 and
amino acids were analysed, depending on the experiment.
5.2.1 The Role of The Urea Cvcle
The amino acid analysis from the experiments in chapter 3 suggested
that arginine concentration mirrored ammonia concentration. This together with
the appearance of citrulline and ornithine suggested that perhaps the ammonia
was being assimilated via a partial or complete urea cycle. The evidence was
circumstantial, and there are no reports in the literature of a urea cycle in
hybridomas. Lavery et al (Unpublished) deduced from their amino acid data that
there may be urea cycle involvement but no firm evidence has been presented.
This lack of evidence may be because the concept has not been investigated.
Urea cycle enzymes have been found in tissues other than liver, but not
in lymphocytes, myelomas, or hybridomas. It seems possible that during the
fusion of the original myeloma and spleen cells, genes which would otherwise
be repressed were activated. The activity of ornithine carbamyl transferase,
which catalyses the addition of carbamyl-P to ornithine: the point of entry of
ammonia into the urea cycle (Figure 5.1), would if present confirm the presence
of a partial or complete urea cycle.
During this particular experiment, samples of 1 x 107 viable cells were
taken at intervals and the cells harvested by centrifugation at 200g for 5
minutes. The cells were resuspended in 1ml of tris buffer and sonicated for 3
x 10 seconds (Soniprobe), with 30 second intervals during which the cells were
kept on ice. After sonication, the cells were then kept on ice for a further 30
minutes. This suspension was then centrifuged for 5 minutes at 5000rpm
(microfuge), and the supernatant was assayed for ornithine carbamyl transferase
activity (2.2.9).
120
Results and Discussion
The bioreactor run exhibited the typical phasic nature shown in previous
experiments. No activity of ornithine carbamyl transferase (OCT) was detected,
suggesting that it was either at a level too low to be detected, or was absent.
The latter seems the more likely, because the assay was selected because of its
sensitivity and specificity. The absence of OCT activity suggests that there is
not a urea cycle involvement in ammonia assimilation in PQX. The presence of
ornithine and citrulline therefore may be due to metabolism of arginine (Eagle,
1959; Levintow et al 1957; Piez et al 1958), or may be from another
metabolic pathway.
5.2.2 15N-NMR Studies of Ammonia Assimilation
In order to determine how ammonia was assimilated by the PQX
hybridomas a labelling study was carried out, in which assimilation of 15N
labelled NH4C1 was followed using 15N Nuclear Magnetic Resonance (NMR)
spectroscopy.
At intervals during the bioreactor culture, samples (100ml) were taken
and the cells removed by centrifugation at 200g for 5 minutes. The cells were
discarded, the supernatant freeze-dried (4.3), and sent to Oxford University for
15N NMR analysis.
Results and Discussion
Cell Growth
The increased initial concentration of ammonium chloride seemed to affect
the progress of the culture, by extending the lag phase (Figure 5.3a). Similar
observations have been reported by Brindle et al (1990), using HeLa cells
grown in equine serum-supplemented medium. The effect may not be due to
121
NH4+ ions alone but to the Cl* ions. Increasing extracellular Cl* concentration
may cause an imbalance in the intracellular/extracellular gradient. This in turn
may cause the efflux of amino acids and/or inorganic salts as an attempt to
restore the balance. The fall in intracellular amino acids concentration would
inhibit protein synthesis and cellular growth (Levintow and Eagle, 1961),
effectively holding the cells in G1 phase of their cell cycle.
0.65
- 0.52I 10.00 in4ox 7.50 ■ 0 3 9
d0.265.00
LU
i 0.132.50
0.000.001006020 8 04 0O
TIME (hours)
Figure 5.3a The growth curve of PQX Bl/2 hybridomas in the presence of
ImM NH4CI showing an extended lag phase
Monoclonal Antibody Production
The Mab titre reached a maximum of 23^g/ml at 70 hours with the peak
in production rate occurring at 60 hours (Figure 5.3b). This titre was lower
than in previous experiments, and may be a reflection of the NH4C1 addition.
122
OD
prr'
ipm
n
;nn
\A/T
U
na
tp
Zh
—-t\
Z
vin
1.0025
0.8020
I-3n 15
rF 10
1
0.60
0.40
0.20
0.00100806040200
TIME (hours)Figure 5.3b The production of Mab ( • ) of cells grown in the presence of
ImM NH4CI, and the major peak in product formation rate
(dP/dt)
Ammonia Assimilation
The period of "growth inhibition", coincided with a peak in ammonia
assimilation (Figure 5.3c).
This may be associated with the extracellular NH4+ concentration, and
may be an attempt by the cells to parry the inhibition. It may also be that the
cells are using ammonia as a nitrogen source for amino acids synthesis, perhaps
to increase the intracellular pool depleted by the Cl* gradient.
The second period of ammonia assimilation occurred at the end of the
maximum growth period, slightly after the "normal" ammonia assimilation peak.
123
PROD
UCT
FORM
ATIO
N RA
TE
(pg/
ml/h
)
3.50
IZ 2.80OH<H 2 .1 0
0.20
0.16
0.12
OO 1.40
<z0 0.70
0.08
0.04
<0.00 0.00
o 20 40 6 0 8 0 100
TIME (hours)
Figure 5.3c The assimilation pçaks of ammonia during a batch culture of
hybridomas grown in the presence of ImM NH4C1
15N Incorporation
15N NMR analysis revealed incorporation of 13NH4+ into the cellular
metabolites, confirming the observations in previous experiments. In particular
15N label appeared in ALA, and GLU towards the end of the culture (Table
5.2). The incorporation of 15N label was low, suggesting that much of the N
in ALA and GLU must come from other amino acids, probably GLN. This low
incorporation has also been observed in similar experiments with CHO and
HeLa cells (Brindle et al 1990).
124
NH3
ASSI
MIL
ATIO
N RA
TE
(mM
/h)
Table 5.2 Appearance of Labelling in a PQX Bl/2 Bioreactor Batch
Culture supplemented with ImM 15NH4C1 (Brindle et al, 1990)
TIME (Hours) % ALA Labelling %GLU Labelling
18 1.8 '
30 1.8
42 2.0
54 3.1
66 1.7
76 2.6
90 5.3 1.9
The appearance of 15N into ALA and GLU, suggests a possible coupling
of alanine aminotransferase to glutamate dehydrogenase in a cycle manner
(Figure 5.3d).
NH
GDH
glutamateALT
alanine pyruvate
Figure 5.3d The Cyclic Coupling of Glutamate Dehydrogenase (GDH)
to Alanine Aminotransferase (ALT)
125
The appearance of label in GLU at the end of the culture may indicate
that the ALA concentration (~ LOmM) is such that it is causing end-product
feed-back inhibition of the cycle. The similar textbook liver-cell pathway is
usually represented as the reverse reaction to the one suggested here, in that
ammonia is liberated. In muscle tissues however, the ammonia assimilation
direction for the pathway occurs (Lehninger, 1975). It is feasible that during the
initial gene rearrangements after the hybridoma was first produced, such a
pathway could be activated. Thus this experiment has confirmed the assimilation
of ammonia by the PQX Bl/2 hybridomas, and suggests that the incorporation
is via glutamate dehydrogenase coupled to alanine aminotransferase, with a-KG
recycled.
5.2.3 Determination of Glutamate Dehydrogenase Activity in POX
Bl/2 in Batch Culture
During this experiment, samples (IxlO7 viable cells/ml) were taken from
the bioreactor culture, the cells harvested by centrifugation (200g for 5 minutes),
and resuspended in 1ml of 0.85% w/v NaCl in MQ. The cells were sonicated
as previously described, and kept on ice for 60 minutes, after which the
sample was centrifuged (5000rpm, for 5 minutes), and the supernatant assayed
for Glutamate Dehydrogenase activity (GDH), (2.2.10).
Results and Discussion
Cell Growth
The culture again followed the phasic pattern, with the viable count
peaking at 9.66 x lOVml.
126
Glutamate Dehydrogenase and Ammonia Assimilation
There was evidence of GDH activity throughout the culture. The peaks
in activity occurred on either side of the peak in ammonia assimilation rate
(Figure 5.4), although there was significant GDH activity during this period.
0.15 100
IJ 0.121115£ 0.09
O
< 0 .0 6
80
6 0
4 0_J
È50.03 20
2 A
0.0020 4 0 100O 6 0 8 0
TIME (hours)
Figure 5.4 GDH activity ( v ) was at a minimum between the peaks in
ammonia assimilation ra te (A )
It is likely that this activity may account for the ammonia assimilation,
but precise calculations are difficult for two reasons. Firstly, GDH is bound to
the mitochondrial matrix in the cell, and the location is such that it is very
difficult to extract all the enzyme. Thus the total activity of GDH may in fact
be higher than that detected here. Secondly, ammonia is seen as a net uptake
at this time, however, there is probably an equilibrium between assimilation and
127
GDH
ACTI
VITY
j^
M/m
in/l)
production. As the equilibrium shifts, perhaps as a reflection of metabolic or
environmental conditions, at other times during the culture, a net production will
be seen. This will therefore make determination of the exact available substrate
concentration difficult.
These results do confirm that there is significant GDH activity which
probably accounts for the ammonia assimilated.
The other main mechanism by which ammonia may be assimilated is via
glutamine synthetase in the following reaction.
GLUTAMATE + N H / GLUTAMINE
The essential role of GLN in mammalian cell culture means that a cell
able to synthesise its own GLN, would be at an advantage towards the end of
the culture when the GLN is depleted. The activity of GS in a PQX Bl/2
bioreactor batch culture was investigated by Bushell et al, (1988). The activity
of GS was found to peak at the beginning, and at the end of the culture, with
no activity during the period of net ammonia assimilation. The initial activity
is thought to be due to "carry-over" from the inoculum.
The regulation of GS was found to be at the transcriptional level, and
repressed by the presence of high concentrations of GLN. Towards the end of
the culture when GLN concentration is very low, this repression may be lifted,
and the activity at the end of the culture, may reflect the importance of GLN
in Mab synthesis, suggested by previous experiments.
5.3 AMINO ACID UTILISATION
The triphasic culture profile of PQX Bl/2 hybridomas proposed in
chapter 3 suggested that in preparation for growth and during Mab production,
particular amino acids were assimilated. The pattern of utilisation suggested that
amino acids play an important role in growth and production. The materials
balance equations derived in chapter 4, confirmed that some amino acids were
128
assimilated more than others, and at particular times during the culture. This
suggests that some amino acids may be more important than others for growth
and/or production.
A greater understanding of these roles of amino acids in hybridoma
metabolism may facilitate process optimisation, leading to more efficient growth
and Mab production.
It is difficult to detach the role of one amino acid, and assess it’s
importance while other amino acids are present and at varying concentrations.
It is probable that some amino acids inter-convert and substitute function with
others, thus causing further complication.
5.3.1 Factorial Determination of Important Amino Acids
The aim of this experiment was to screen efficiently 20 amino acids, and
determine the relative importance of each for growth and Mab production.
The classical method is to study one variable at a time, while holding
all others constant. This is a very inefficient method requiring an inordinately
high number of trials for few components. Using a full factorial method, in
which every possible combination of independent variables at appropriate
concentrations are investigated, for X variables at Y levels, Yx trials are
required. Therefore 20 variables at 2 concentrations requires 400, (202), trials.
This is obviously very expensive and time consuming.
The approach adopted in this investigation was developed by Plackett and
Burman (1946). They proposed a series of basic designs for upto 100
experiments based on a rationale known as balanced incomplete blocks: a
fractional factorial method. This allow the evaluation of N-l variables by N
experiments.
The experiments were carried out according to a design matrix (Figure
5.5).
129
Arg Asn Asp Cys DUM Glu Gin Gly His OHp H e Leu Lys Met DUM Phe Pro Ser Thr Try Tyr DUM Va]
1 + + + + + — + - + + - - + + - - + — + — - - -2 + + + + - + _ + + — - + + - - + — + - - - — +3 + + + - + + + - — + + — - + - + - - - - + +4 + + — + — + + - - + + - - + - + - - - - + + +5 + - + — + + — — + + — - + - + - — - — + + + +6 - + - + + — - + + - — + . - + - - - — + + + + +7 + - + + - - + + - — + ■ — + - - - — + + + + + -8 - + + - — + + — + — + - - — — • + + + + + - +9 + + - — + + - - + — + - - - - + + + + + - + -10 + — - + + - - + — + - - — — + + + + + — + — +11 - + + - - + - + - - - - ' + + + + + - + - + +12 - + + — — + - + - - - — + + + + + - + — ■ + + -13 + + — - + - + - - — + + + + + — + - + + - -14 + - - + - + — — - - + + + + + - + — + + - - +15 - — + — + - - - ~ + + + + + - + - + + - . - + +16 - + + - — - - + + + + + - + - + + - - + + -17 + - + - — — - + + + + + - + - + + - - +• + - -18 - + - - - - + + + + + - + + + - - + + - . - +19 + — — — — + + + + + ■ — + - + + - - + + - — + -20 - - ' - — + + + + + — + - + + - - + + - - + - +21 - — - + + + + + - + — + + — - + + - - + — + -22 - - + + + + + - + - + + — - + + - - + - + - -
23 - + + + + + - + - + + - - + + - - + - + - - -24 - — - - — - - — - - - _ — — - — — - — ■ - — -
Figure 5.5 The Plackett-Burman Matrix
Each amino acid (component) was tried at 2 levels; a high concentration
(+) and a low concentration (-). The experimental error was estimated by
including 3 "dummy variables" (replicate controls). Each amino acid was tried
at it’s high level 12 times and at it’s low concentration 12 times. The
130
contribution of each factor to the response was determined as the difference
between the average values of the response for the 12 experiments at the high
level and the average response at the low level. The effect of one variable was
distinguished from all the other variables which were changing at the same time.
When one variable was at its high level any other variable was at it’s low
level 6 times, and at it’s high level 6 times. Similarly when the first variables
was at a low level. The net effect of changing a second variable was therefore
cancelled out when calculating the effect of the first.
No changes were made corresponding to the effects of the + and - on
the dummy column (used to estimate the variance of the effect).
From the variance the standard error was calculated, and from this the
significance level determines using the student t-test. The t-test for any
individual effect allows an evaluation of the probability of finding the observed
effect purely by chance (ie the effect is due simply to experimental variation).
If this probability is sufficiently small, the idea that the effect was cause by
varying the level of the parameter under test is accepted.
The significance level given by the t-test was used to rank the amino
acids. Confidence levels greater than 70% were accepted (Stow and Mayer 1966)
Experimental Details
The basal medium was prepared as follows
RPMI 1640 without amino acids (Appendix IE) 250ml
MQ water 1425ml
NaHCO-,' (7.5% Solution, Gibco BRL) 50ml
NaOH (1M) 25 ml
Heat-treated equine serum (Imperial) 250ml
Basal medium (70ml) was added to each of 24 flasks labelled 1-24.
Amino acids were added according to Figure 5.5. The ’ + ’ denotes a high level
addition, which was 1ml, and gives a final concentration equivalent to the
131
concentration of that amino acid in complete RPMI 1640 (Appendix If). The
low level denoted by was half of this (0.5ml).
The hybridomas were grown in standard growth medium (Appendix la),
harvested by centrifugation at 200g for 5 minutes, and washed by resuspension
in 20ml of basal medium. The cells were then harvested by centrifugation at
80g for 5 minutes, and resuspended at a viable cell density of 2 x 106 cell/ml
in basal medium. 10ml of this suspension was then added to each flask to give
a viable cell density of 2 x 105 cells/ml. The flasks were agitated to mix the
contents and a 2ml sample taken. From this 0.5ml was removed for a cell
count (2.1.4), the rest was centrifuged at 1640g for 10 minutes, and stored at -
20°C in 0.5ml aliquots for ELISA (2.2.4).
A 2ml sample was taken every 24 hours from each flask and treated as
described.
The highest viable count, minus the initial inoculum was used as the
response for growth, and the highest Mab titre as the response for production.
The t-statistic for each amino acid was calculated for growth and production
using the following equations.
Ea R at (+) R at (-)
1212
Veff S(Ed)n
S.E. eff VŸëff
t Ea
S.E. eff
132
Where:-
R
S.E. eff
Ea
Veff
d
= the effect of an amino acid
= the response
= variance of the effect
= effect shown by the dummy variables
= standard error of an effect
These calculations were facilitated by a BBC microcomputer and
appropriate software (Appendix 4). The t-values were then interpreted using
Cambridge Tables (Lindley and Scott 1984).
Results and Discussion
Cell Growth
Most of the highest viable cell counts were recorded after 48 hours
(Table 5,3a). The t-statistics allowed a ranking of amino acid in order of
importance (Table 5.3b).
For growth, the most important amino acids were LYS, GLN, HIS,
SER, TRY and CYS. The concentrations of these amino acids appears to be
critical, and reducing the concentration to half that in RPMI has a significant
detrimental affect on the growth of the hybridoma. Indeed these amino acids
may be growth limiting at this concentration. The other amino acids which do
not appear significant may be important for growth but have a lower critical
concentration than that used in the experiment. This does suggest that they may
be less important than the significant ones because they are required in lower
concentrations. At the bottom of Table 5.3b LEU has a significant negative t-
statistic, which may imply that at these concentrations LEU may have an
inhibitory effect on growth.
133
Table 5.3a Viable Cell Counts (xlOVml) For Flasks 1-24
(Values used for ranking calculation are highlighted)
TIME (hours)
FLASK 0 24 48 72 96 120
1 0.93 5.39 8.45 6.23 4.35 2.03
2 1.07 3.18 5.00 1.10 0.38 0.23
3 1.40 2.75 2.95 2.03 2.05 1.35
4 1.22 2.28 3.98 3.88 2.83 1.73
5 0.92 3.18 4.28 1.70 0.43 0.35
6 1.22 3.59 4.38 2.18 0.90 0.38
7 1.20 3.72 6.85 5.50 3.03 0.98
8 1.40 3.03 3.98 2.98 1.85 0.98
9 1.26 3.65 4.60 1.98 0.70 0.23
10 1.13 3.00 4.40 2.13 0.28 0.08
11 1.47 3.18 5.08 4.23 1.93 0.83
12 1.20 3.20 3.10 1.85 0.75 0.28
13 1.08 3.05 8.13 3.48 1.00 1.23
14 1.26 3.17 4.50 2.03 0.38 0 .00
15 1.47 3.54 4.53 1.80 0.28 0 .00
16 1.01 3.75 4.55 1.48 0.58 0.45
17 0.99 2.57 2.55 1.88 0.38 0 .0 0
18 1.20 3.30 7.40 5.68 1.65 0.43
19 1.34 2.81 3.73 2.25 1.15 0.78
20 1.22 3.24 8.82 5.08 5.53 2.43
21 1.20 3.17 6.28 6.33 1.20 0.23
22 1.22 2.82 4.60 4.58 2.38 0.35
23 1.25 3.77 4.70 1.78 0.08 0 .00
24 1.29 3.17 2.73 0.80 0.23 0 .0 0
134
Table 5.3b Ranking of Amino Acids: Growth
(t values are for 3 degrees of freedom)
AMINO ACID t VALUE SIGNIFICANCE
LYS 2.5760 95%
GLN 1.9248 90%
HIS 1.0283 80%
SER 0.8206 75%
TRY 0.7327 70%
CYS 0.6559 70%
MET 0.4651
ASN 0.4052
TYR 0.2704
ARG 0.2349
HYP 0.1263
GLY 0.1226
THR 0.1132
PHE 0.0608
ILE 0.0459
VAL 0.0421
PRO 0.0047
GLU 0.2030
ASP 0.5268
LEU 0.8038 75%
135
Production
The highest Mab titres were recorded after 96 or 120 hours (Table 5.3c).
The most important amino acids for Mab production were GLN, ARG,
GLY, THR, GLU, LEU, SER, and HYP (Table 5.3d). Reduction in the
concentration of these amino acids may cause a reduction in Mab production.
Again the other amino acids may be important for Mab production but the
concentration below which they become significant, appears to be lower than
those used in the experiment. TRY and LYS have negative t-statistics, and may
suggest an inhibitory effect on Mab production.
The results from this experiment have demonstrated that the concentrations
of amino acids are important for growth and Mab production. Specific amino
acid have been highlighted as particularly important at particular concentrations.
GLN and SER seem to be important for growth and Mab production by PQX
El/2. The importance of GLN may have been predicted from earlier chapters.
LEU however, appears to be inhibitory for growth but important in Mab
production, whereas LYS seems to be the opposite: stimulatory to growth and
inhibitory to Mab production. This experiment corroborates the feed formulated
in the materials balance experiments in which GLN and ARG were particularly
important for production by virtue of their assimilation rates. It is at variance
however with the material balance results for LYS which implied that LYS
assimilation was important for Mab production. The Plackett-Burman study
indicates a weakness in the argument that assimilation of a component is
necessary for a process simply because it is observed during that process.
This fractional factorial method efficiently screened 20 amino acids and
identified the potentially important ones, and forms a good foundation for future
experiments.
136
Table 5.3c Mab Titre fyig/ml) For Flasks 1-24
(highlighted values were used in the ranking calculation)
TIME (hours)
FLASK 0 24 48 72 96 120
1 0 1.5 3.4 3.6 5.9 5.12 0 1.4 3.6 2.9 4.1 4.03 0 1.2 3.8 4.7 7.3 11.84 0 1.3 3.8 4.4 6.8 1035 0 1.4 2.7 2.3 4.1 2.86 0 1.7 3.1 2.4 4.0 3.77 0 1.7 4.5 3.5 7.9 4.18 0 1.7 2.3 4.6 6.8 5.39 0 1.7 3.4 3.0 6.3 5.4
10 0 1.7 4.4 2.9 6.6 4.611 0 1.7 4.8 5.5 8.0 7.912 0 1.8 3.8 2.7 2.8 3.813 0 1.6 3.2 4.4 5.8 5.914 0 1.6 4.3 3.4 2.5 4.215 0 1.6 3.4 3.1 2.7 1.316 0 1.0 2.9 2.2 3.6 3.617 0 0.9 3.7 3.9 2.7 4.218 0 1.0 3.3 4.8 3.8 5.719 0 1.0 3.9 5.9 8.320 0 0.8 4.0 4.2 4.3 4.621 0 1.3 3.3 3.2 3.6 6.022 0 1.3 4.5 3.0 7.4 7.423 0 1.6 3.0 4.2 4.1 4.024 0 1.3 2.7 2.4 4.6 4 .2
137
Table 5.3d Ranking of Amino Acids: Mab Production
(t values are for 3 degrees of freedom)
AMINO ACID t VALUE SIGNIFICANCE
GLN 2.3265 90%
ARG 1.5410 85%
GLY 0.7908 75%
THR 0.7112 70%
GLU 0.6988 70%
LEU 0.6626 70%
SER 0.6493 70%
HYP 0.6148 70%
MET 0.4069
HIS 0.3449
ILE 0.1035
VAL 0.1132
ASN 0.2158
CYS * 0.3327
PRO 0.2672
ASP 0.2840
PHE 0.2875
TYR 0.5095
TRY 0.6642 70%
LYS 1.3835 85%
138
5.3.2 Elimination of Individual Amino Acids
The Plackett-Burman experiment (5.3.1) identified amino acids important
for growth and production. The aim of this experiment was to extend this
amino acids work, in serum-free medium, and look at the effects of omitting
individual amino acids. This will extend the knowledge of amino acid
metabolism in hybridomas, and because of the importance of amino acids, may
facilitate optimisation of Mab production.
Experimental Details
Into each of 22 flasks basal medium (70ml) was added.
Basal Medium
RPMI 1640 without amino acids (Appendix IE) 250ml
H20 1372ml
NaOH (1M) 25ml
NaHC03 (7.5% solution, Gibco BRL) 50ml
Bovine Serum Albumin Fraction V (7.5%, Gibco) 23ml
Insulin (2mg/ml, Sigma) 10ml
Transferrin (Img/ml, Sigma) 10ml
Composite (Appendix Ig) 10ml
Phosphate Buffer (50mM, pH7.1) 250ml
Amino acids were added to each flask according to Table 5.4a. PQX
Bl/2 cells adapted to serum-free medium were used. The cells were harvested
from serum-free medium by centrifugation at 200g for 5 minutes, and
resuspended at 2 x 106 viable cells/ml in basal medium. 10ml of this suspension
was then added to each flask. The contents were mixed, a sample taken (5ml),
and the cells counted. The 5ml sample was centrifuged at 200g for 5 minutes,
and the supernatant stored at -20°C for future amino acid and Mab analysis.
139
Table 5.4a Amino Acid Additions
(values represent ml of stock amino acid solutions)
FLASK
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
ARG 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0
ASN 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 0
ASP 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 0
CYS 1 1 1 1 0
GLN 1 1 1 1 0
GLU1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 0
GLY 1 1 1 1 0
HI S 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 0
OHP1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 0
I LE 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 0
LEU 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 0
LYS 1 1 1 1 0
MET1 1 1 1 0
PHE 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 0
PRO 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0
SER 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0
THR 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0
TRY 1 1 1 1 0
TYR 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0
VAL1 0 1 i i i i i i i i i i i i i i i i i 1 0
H, 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 20
TOT 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
Results and Discussion
Cell Growth
Growth curves for typical flasks are shown in Figures 5.6a and 5.6b.
Cells in flask 2,3,4,5,8,9,10,11,12,14,16,18,21 and 22 (Table 5b), began to die
soon after inoculation, and by 24 hours the viabilities had fallen below 30%. After 48
hours the viabilities of the cells in these flasks was 11% or less.
140
In contrast the cells in flasks 1,6,7,13,15,17,19 and 20 (Table 5.4b) grew well,
and viability only dropped below 30% after 166 hours in culture. These results suggest
that individual omission any one of the following amino acids has little effect on
growth of PQX Bl/2: ASP, ASN, GLY, GLU, OHP, PRO and SER. Results are in
agreement with Eagle (1955a;b). There are two exceptions in that he makes no mention
of ASN and includes ALA as a "non-essential" amino acids. ALA is not added to the
culture medium because it is produced in significant quantities throughout the culture.
It is therefore also considered non-essential for growth.
10
md5
UJ
>
170136102683 40TIME (hoirs)
Figure 5.6a The growth curve for the CONTROL flask containing all the
amino acids. This profile is typical of flasks 1, 6, 7, 13, 15, 17, 19,
and 20 which had amino acids deleted according
to Table 5b
md
tn
ÎUJ
>
1701361023 4 68OTIME (hours)
Figure 5.6b The growth curve of the negative control with all the amino acids
omitted. This was a typical profile for flasks 2, 3, 4, 5, 8, 9, 10,
11, 12, 14, 16, 18, 21 and 22
141
Table 5.4b Labelling of Flasks, and Amino Acid Omissions
FLASK NUMBER AMINO ACID OMITTED
1 NONE: +VE CONTROL
2 VAL
3 TYR
4 TRY
5 THR
6 SER
7 PRO
8 PHE
9 MET
10 LYS
11 LEU
12 ILE
13 OHP
14 HIS
15 GLY
16 GLN
17 GLU
18 CYS
19 ASP
20 ASN
21 ARG
22 ALL: -VE CONTROL
Monoclonal Antibody Production
Mab production curves for flasks 1 (Control), 6 (-SER), 7 (-PRO), 13 (-OHP),
15 (-GLY), 17(-GLU), 19 (-ASP), and 20 (-ASN) are shown in Figures 5.7(a-
h). Flasks in which the cells did not grow were not assayed.
142
175
140
105
7 0
0 3 4 102 136 170
TI&/S (hocra)
Figure 5.7a Volumetric Mab titre for the control flask containing all the
amino acids
175
140
105
<r.70
35
O 34 58 102 136 170
TIME (hocrs)
Figure 5.7b Volumetric Mab production for flask 6 (minus SER)
175
140
105
II-I
70
35
0 34 58 102 136 170
TIV6 (hours)
Figure 5.7c Volumetric Mab production for flask 7 (minus PRO)
143
175
140
105
70
35
0170102 1366334O
TIV€ (htxrs)
Figure 5.7d Volumetric Mab production for flask 13 (minus OHP)
175
140
105
7 0
35
O170136102sa3 4O
T iv e (bexrs)
Figure 5,7e Volumetric Mab production for flask (minus GLY)
175
140
105
70
35
0 136 170sa 10234OTIME (hou-s)
Figure 5.7f Volumetric Mab production for flask 17 (minus GLU)
144
175
140
'3' 105
7 0
35
0 34 58 102 136 170
TIME (hotrs)
Figure 5.7g Volumetric Mab production for flask 19 (minus ASP)
175
140
7 0
3 5
34 . 68O 102 170136
TIME (hotrs)
Figure 5.7h Volumetric Mab production for flask 20 (minus ASN)
The results suggest that omission of an amino acid classified as "non-
essential” for growth does make some difference to the Mab titre. This may be
as a result of energy being diverted to the synthesis of that non-essential amino
acids, energy otherwise used for Mab production.
145
5.3.3 Elimination of "Essential" Amino Acids
The aim of the experiment was to determine the effect of removing all
amino acids classified as non-essential in the previous experiment.
Experimental Details
Basal medium (83ml) was added to each of 4 flasks.
Basal Medium
RPMI 1640 - amino acids 100ml
H20 (MQ) 553ml
NaOH (1M) 10ml
NaHC03 (7.5% solution, Gibco BRL) 20ml
BSA (7.5%, Gibco BRL) 9ml
Insulin (2mg/ml) 2ml
Transferrin (lmg/ml) 2ml
Phosphate Buffer (50mM, pH7.1) 100ml
ARG (20mg/ml) 10ml
CYS (5mg/ml) 10ml
GLN (30mg/ml) 10ml
HIS (1.5mg/ml) 10ml
ILE (5mg/ml) 10ml
LEU (5mg/ml) 10ml
LYS (4mg/ml) 10ml
MET (1.5mg/ml) 10ml
PHE (1.5mg/ml) 10ml
THR (2mg/ml) 10ml
TRY (0.5mg/ml) 10ml
TYR (2mg/ml) 10ml
VAL (2mg/ml) 10mlcomposite 2m|
146
The 150cm2 TC flasks were labelled ‘AV, ‘A2‘, (CONTROLS) and ‘BV,
‘B2‘ (TESTS). To the ‘A‘ flasks 1ml of each of SER (3mg/ml), GLY
(lmg/ml), OHP (2mg/ml), PRO (2mg/ml), ASP (2mg/ml), ASN (5.68mg/ml) and
GLU (2mg/ml) were added. To the ‘B‘ flasks 7ml of sterile MQ water were
added.
The serum-free cells were harvested by centrifugation at 200g for 5 minutes,
and resuspended in basal medium at 2 x 106 viable cells/ml. 10ml was then
used to inoculate each of the four flasks. The contents were mixed, and a 5ml
sample taken. The cells were counted and the sample supernatant aliquoted and
stored at -20°C.
Stored samples were assayed for Mab, glucose and lactate.
Results and Discussion
Cell Growth
There was little difference between the growth curves of the cells grown
with or without the "non-essential" amino acids (Figure 5.8a).
10
too
CONTROLX
Ü
Ï>TEST
35O 70 105 140 175TIME (hotrs)
Figure 5.8a Growth curves of flasks with (CONTROL) and without (TEST)
the "non-essential" amino acids.
147
The control flasks reached a slightly higher maximum viable count, and
maintained a marginally higher viability for longer. This suggests that the
maintenance of viability (membrane integrity), requires energy. There may be
more energy available for this purpose in the ’control’ cells than in the ’test’.
Again implying that energy is expended synthesising the omitted amino acids.
Monoclonal Antibody Production
The Mab titres of the cells grown without non-essential amino acids were
very much lower than those growth with these amino acids (Figure 5.8b).
120
CONTROL
96
48TEST
24
o 35 70 105 175140TIME (hours)
Figure 5.8b Mab production in flask with (CONTROL) and without (TEST)
"non-essential" amino acids
This may again imply that energy expended on amino acid synthesis is
diverted from Mab production. It may also be however that although enough
non-essential amino acids can be synthesised to support growth, the demand by
Mab production is only met in part. Any "error" in the synthesis of Mab such
148
as incomplete formation, or conformational change due the absence or change
of an amino acid, will be excluded by ELISA. Thus even though Mab is being
synthesised, because it is not active Ab is remains undetected by ELISA.
Glucose and Lactate
The synthesis of GLY and SER usually occurs from glucose. The ability
of the cells to grow in the absence of these amino acids, suggest an ability to
efficiently synthesis them from glucose. The utilisation of glucose therefore may
be expected to increase, to meet this demand. There was very little difference
between the utilisation of glucose between the two sets of cultures (Figure
5.8c). The same was also true of lactate production (Figure 5.8d).
12.00
i0
19.60
7.20
TEST
4.80
CONTROL
2.40
0.001751401 0 57 035O
TIME (hours)
Figure 5.8c The glucose utilisation of CONTROL and TEST flasks with
and without the "non-essential" amino acids
149
15
0
1 TEST
CONTROL
17570 105 14035O
TIME (hours)
Figure 5.8d The production of lactate from flasks with (CONTROL) and
without (TEST) the "non-essential" amino acids
There was a slight decrease in lactate concentration towards the end of the
cultures suggesting that lactate dehydrogenase was being released from dead or
dying cells.
150
6. FINAL DISCUSSION
The results presented in this thesis show that the PQX Bl/2
hybridoma, grown in batch culture, exhibits a tri-phasic pattern of growth,
Mab production and nutrient assimilation. The events occurring in these
phases can be interpreted in terms of the cell-cycle.
PHASE 1
During the first phase, after inoculation, the cells show apparently
little growth: comparable with the lag phase of a bacterial batch culture. The
results are consistent with the idea, that the cells, at different stages in the
cell-cycle, begin to arrest in Gl. During Gl, there are gene activations,
cascades of enzyme activations, and structural changes, which lead to the
production of active DNA-replicating enzymes, initiator proteins, chromosomal
proteins, and other components needed for a round of chromosome
replication. There is massive biosynthetic activity which is reflected by the
peak in assimilation rate of many of the amino acids; primarily GLN. GLN
can not only provide energy, probably by glutaminolysis, but also precursors
for nucleotide synthesis. The energy requirement for this biosynthesis is high,
and so ATP is synthesised at a rate to meet this demand; seen as a high
B.C. value of 0.9. In order for these hybridoma cells, "arrested": in Gl, to
progress through the cycle, the correct "signals" must be received for the
cell to traverse the restriction point. This "signal", may be one of a number
of factors, some of which are aluded to in the results presented in chapter
3. The length of the lag phase seemed variable, and was extended in cells
grown in serum-free medium. The factors involved were suggested to be one
or a combination of the following: low inoculum density, low concentration
of stimulatory autocrine or medium-supplied growth factors, presence of
growth inhibitors (detoxicification of H20 2 for example), and activation of
repressed growth promoters (for example reduction of cystine). In serum-
151
to result from either a low inoculum density, or in the case of the 15N NMR
experiment (5.2.2), the presence of NH/Cl". In the later case both NH4+ and
Cl' can cause growth inhibition. The NH4+ may interfere with a-KG utilisation,
(implied by a net ALA production), decrease the TCA cycle flux, and therefore
oxidative metabolism. Obviously during the lag period the TCA cycle plays a
crucial role in synthesising ATP to drive biosynthetic processes, and needs
therefore to be operating at maximum efficiency. Hence interference in the flux
through the TCA cycle will decrease energy production, retarding the
progression of the cells to exponential growth. An increase in extracellular
concentration of Cl may cause a change in the electrochemical status of the
cells and a change in the flux of such components as Na+ and K+. This may
then have a cascade effect, causing the efflux of amino acids, thus reducing the
intracellular concentration in an attempt to redress the electrical and chemical
balance. If the amino acid concentration inside the cell falls below a critical
level, then protein synthesis may be inhibited (Levintow and Eagle, 1961),
effectively holding the cells in Gl. Gradually there may be enough growth
precursors synthesised to allow progression into S-phase if signalled to do so.
This signal may be one of a number of factors, the most likely of which is
a surge in cAMP (Boynton and Whitfield, 1982). Indeed, cells trapped in Gl
by Ca2+ deprivation, can be stimulated to synthesise DNA by external cAMP.
The role of Ca2+ in the progression of cells from Gl (3.3.1) is interesting, and
probably crucial in PQX Bl/2 growth. It is proposed that a change in Ca2+
uptake stimulates the cells, and releases the inhibition. The cells begin to
progress to S phase, and the culture towards phase II. The "trigger" for Ca2+
uptake may be a change in intracellular pH caused by an increase in the NH4+
concentration. The growth inhibitory effect of NH4+ was proposed by Glacken
et al (1986), to be due to a change in electrochemical gradient. It may be
therefore that the increasing extracellular NH4+ concentration causes an imbalance
with the intracellular concentration, and NH4+ is assimilated. By changing
intracellular pH, this assimilation may cause a change in the electrochemical
potential of the cell, triggering an influx of Ca2+ into the cell which may
152
initiate a short burst of adenylate cyclase activity, followed by a transient
reduction in cyclic nucleotide phosphodiesterase activity, and a resultant transient
cAMP surge (Boynton and Whitfield, 1982). As net ammonia assimilation
continues, the cells continue to be stimulated to follow the cell-cycle. As
specific growth rate increases to a maximum (~0.04-0.05h*1), then the cells
traverse the cell-cycle at their maximum rate.
PHASE n
The nutritional environment during phase II seems favourable for growth.
Demand for energy is very high and the reducing ability to synthesise ATP at
the required rate, coupled to the increased synthesis of cAMP, contributes to
the observed decline in E.C. value. Although maximum glucose utilisation occurs
during this period, the major energy contribution is thought to be from GLN
and other amino acids; glucose is thought to function mainly as a precursor for
nucleotides and amino acids (Hume et al, 1978). Indeed there is net amino acid
production during this phase. Assimilated of NH4+ during this period of
maximum growth was thought to be via glutamate dehydrogenase coupled to
alanine aminotransferase, in a cyclic manner (5.2.2). This mechanism produces
ALA, the major metabolite in which 15N appeared. This latter evidence
apparently contradicts the observation, from the amino acid uptake studies, and
the materials balance equations, that ALA is utilised during this phase.
However, the measured uptake at this point is a net rate, and it is suggested
that there is production, but this is masked by the utilisation profile. Moreover,
ALA is not supplied in the medium, and in order for there to be assimilation
of ALA there needs to be synthesis by the cells.
As the culture progresses further, the nutritional environment changes
such that the concentration of toxic metabolites increases, and that of nutrients
decreases. These factors may cause more and more cells to begin to arrest in
Gl as the requirements for progression to S are not met or are inhibited. The
153
ability of the cells to generate ATP falls, and the culture begins to enter the
third phase.
PHASE m
During this phase the viable cell concentration peaks, and Mab is
synthesised at its maximum rate. Maximum Ab synthesis is reported to occur
in Gl (Buell, 1969; Garatun-Tjeldsto et al 1976), so as more cells arrest in Gl
more Mab is synthesised. This agrees with the suggestions that maintenance of
viable cells without growth increases the Mab production. Energy previously
utilised for growth processes can be diverted to Mab production and
maintenance. From the trajectory plots of energy charge against growth rate, the
E.C. value seem to fall to a critical level, implying that perhaps the cells
ability to synthesise energy diminishes because of the inability to cross the
restriction point due to "energy deprivation", and this retains the arrested cells
in Gl .
There is obviously significant protein synthesis during this time, fuelled
by amino acids, shown by a peak in assimilation. GLN was suggested to be
important for Mab production both in the Plackett-Burman factorial experiment
(5.3), and in the materials balance equations. Its role may again be primarily
in energy production for such macromolecular synthesis. The demonstrated
activity of glutamine synthetase at the time reinforces the importance of GLN:
the cell is expressing an enzyme which for the most part of the culture is
repressed. In addition ARG was proposed to be important for the Mab
production phase of the culture in both the Plackett-Burman experiment and the
materials balance work, however the specific role is unclear.
The E.C. value continued to fall through during this phase, and, provided
the value remained in a particular domain, Mab synthesis continued. This
suggests that there may be sufficient energy available to keep the cells in Gl,
allow Mab production and cell maintenance. As the EC value falls further,
154
priority for energy shifts to maintenance processes only, the cells revert to GO
and die.
The events occurring in the three phases described above are summarised in
Table 6.
Table 6 A summary of proposed events of the phasic
culture in relation to the cell cycle
EVENT PHASE I PHASE II PHASE III
GROWTH Apparent inhibition Maximum growth Viable count peaks
then cells die
MAB Very little production Some production Maximum production
AMINO Net utilisation of most
ACIDS primarily GLN
Net production of
most amino acids
Net utilisation of
most, primarily
ARG, GLN
CELL Cells inoculated at Cells at different Cells begin to arrest in
CYCLE different stages stages. Cell cycle Gl then revert to GO and
begin to arrest in Gl progresses at its die
Later in phase I pass R point maximum rate
into S phase
155
Mab was proposed to be related to growth rate in a complex way.
There was some Mab production while the cells were in the growth phase ,
but because they spent progressively less time in Gl, as fi increased only a
small amount of Mab was produced. As jn decreased, so the length of Gl
increased and hence Mab production was stepped up. This explains the inverse
relationship of dP/dt to fi. Very little Mab synthesis during the lag phase,
probably because the nutritional environment was not optimal, and perhaps
because the energy level was too high, in this way perhaps analogous to
secondary metabolite production in bacteria.
On a final note as the work presented here progressed, the maximum
Mab titre achieved seemed to decline. Originally this was thought to be due
to instability of the cell-line. Instability may be exhibited as a complete loss
of production by a proportion of the cells, resulting in selection for "non
producers." The second possible cause of reduced production results from gene
deletions causing a change in the Ab sub-class, as discussed in chapter 1. The
gene arrangement of mouse Ig is such that a single deletion will cause a change
of IgGi to IgGa, a second to IgG^, a third to IgE and final to IgA. Any of
these single non-reversible deletions will result in a Ab subclass that will not
be recognised by the ELISA used and therefore, will simulate non-production.
There may be both a complete loss of Mab synthesis ability and a change in
Ab class, further investigation with different ELISA systems, specific for
particular Ig classes, or fluorescent activated cell sorting using fluorescnet Ab
would aid in determining the cause
The work presented in this thesis has shown physiological changes
occur during growth and Mab production by the PQX Bl/2 hybridoma, and
has suggested explanations for these. The Plackett-Burman experiments screened
the amino acids whose role of has been constantly highlighted as very important
and suggested that while LYS and GLN were important for growth, and ARG
and GLN for production, some amino acids notably LEU and LYS, may be
156
inhibitory to one phase of culture and stimulatory to another. Amino acid
omission experiments confirmed the results of Eagle (1955a, 1955b), highlighting
amino acids which could be omitted without significant effect on growth but
which were essential for Mab production. Results from the amino acid work
all suggest that optimisation of amino acids for growth and production, followed
by use of a fed-batch culture could significantly improve Mab production. A
clearer understanding of the metabolic pathways involved in catabolism and
anabolism of amino acids, may also identify key areas for further improvement
perhaps by controlled, amplified expression of key enzymes.
157
7. CONCLUSIONS
1. Growth, monoclonal antibody production and nutrient assimilation by the
PQX Bl/2 hybridoma was triphasic. This was inferred by use of partial
cubic spline interpolation of experimental data, and subsequent derived
variable computation. It is reproducible, and consistent with data from
the activities of glutamine synthetase, and glutamate dehydrogenase, and
energy charge measurements (and northern blot data obtained in a parallel
study).
2. During the second phase of the culture, ammonia was apparently
assimilated. Evidence from the amino acid utilisation profile of the culture
had suggested that ammonia assimilation may be due the activity of a
partial or complete urea cycle. However, no activity of the enzyme
ornithine carbamyl transferase was detected and this hypothesis was
therefore disproved.
3. Uptake of ammonia was confirmed by following the incorporation of
15N in cellular metabolites using NMR spectroscopy. The 15N was
assimilated into alanine, and towards the end of the culture, glutamate.
This is proposed to be via a cyclic coupling of glutamate dehydrogenase
to alanine aminotransferase. Analysis of glutamate dehydrogenase activity
provides conclusive evidence for this proposal.
4. Addition of ImM NH4C1 had no adverse affect on the growth of
PQX Bl/2.
5. Growth and monoclonal antibody production by PQX Bl/2 is shown to
be related in a complex manner. The major peak in production rate
occurred after the major peak in growth rate. This time separation of
158
growth and production rates allowed materials balance equations to be
determined for each phase.
These equations showed that the nutritional requirements for growth and
production were different. These differences were then exploited to
determine feed and feed-rate for use in a fed-batch culture.
6. On-line analysis of the exchange of 0 2 and C02 showed that the
respiratory quotient could provided an on-line indicator of Mab production
and therefore the time at which a feed in a fed-batch should be initiated.
7. The use of the Plackett-Burman factorial matrix was an efficient method
of ranking amino acids in terms of importance for growth and
monoclonal antibody production. The most significant for growth were
glutamine and lysine, whilst arginine and glutamine were the most
significant for Mab production.
8. Thirteen amino acids were found to be essential for growth of the PQX
Bl/2 hybridomas in serum-free medium, whilst seven others could be
omitted without a detrimental effect to growth. However, omission of
these amino acids drastically affected Mab production.
9. There was little difference between the growth of PQX Bl/2 in serum-
free and serum-dependent medium. There was however a marked increase
in Mab production in the serum-free medium.
159
8. FURTHER EXPERIMENTS
1. Measure the activity of alanine aminotransfease, and radioactively label
glutamate, glutamine, alanine and a- ketoglutarate, to further investigate
the ammonia assimilation mechanism.
2. Addition of serum to serum-free medium to investigate the reason for the
improved Mab production in serum-free medium and assess the possibility
of inhibition of Mab secretion by serum proteins.
3. Further investigate the energy status of the culture looking at the effect
of different substrates with varying energy-producing and energy-reducing
potential, and the effect on the energy charge.
4. Use the feed formulation, and feed-rate determined by the materials
balance equations in a series of fed-batch cultures, with the aim of
optimising Mab production, utilising qC02 for continuous on-line
optimisation.
5. Use flow cytometry to investigate the cell-cycle of the PQX Bl/2
hybridoma and to determine the effect of altering extracellular Ca2+ and
cAMP levels on the cycle. This will allow a more informed prediction
of the relationship between growth, Mab production and energy charge,
with possible cell cycle events.
6. To try unadapted cells in serum-free medium and compare to the results
presented in this thesis.
160
APPENDIX 1
A. Growth Medium
RPMI 1650, xlO concentrated liquid (Gibco BRL)
Milli Q Water
NaHCO,, 7.59% solution (Gibco BRL)
NaOH, 1M
Heat Treated Equine Serum (Imperial Laboratories)
Glutamine, 200mM (Gibco BRL)
B. Phosphate Buffer (50mM. pH7.1)
Solution A:
KH.PO,
Solution B:
KzHPO,
Working Solution: Solution A
Solution B
C. Large Scale Medium
RPMI 1640 powder (Imperial Laboratories)
Milli Q Water
NaHCO, (AnalaR, BDH)
HC1 (AnalaR, BDH)
K+P04- buffer (50mM, pH 7.1)
Heat Treated Equine Serum (Imperial laboratories)
Glutamine CXMndd, Gibco BRL)
ml/1
100
760
20
10
100
10
S/I
27.218
34.836
ml170
330
101.15g
6.91
15.0G
4 ml
11
11
100ml
D. Freezing Mix (For serum-dependent cell only)
ml/dl
RPMI 1640, xlO concentrated liquid (Gibco BRL) 10
Milli Q Water 70
Foetal Bovine Serum (Gibco BRL) 20
Dimethyl Sulphoxide, DMSO, (AnalaR, BDH) 10
E. RPMI 1640 (- amino acids)
Inorganic Salts mg/1
Ca(N03)2.4H20 100
KC1 400
MgS04.7H20 100
NaCl 6000
NaHP04.7H20 1512
Vitamins
Biotin 0.2
D-Ca Pantothenate 0.25
Choline chloride 3.0
Folic Acid 1.0
i-Inositol 35.0
Nicotinamide 1.0
p-Amino benzoic acid 1.0
Pyridoxine HC1 1.0
Riboflavin 0.2
Thiamine HC1 1.0
Vitamin B12 0.005
Other Components
Glucose 2000
Glutathione (reduced) 1
Phenol red 5
(all reagents were purchased from Sigma unless otherwise stated)
F. Serum-Free Medium
fi/1RPMI 1640 101.12
NaHC03 1.5
Glutamine 0.1462
Bovine serum albumin (fraction V) 0.7
Human transferrin 0.05
Bovine Insulin 0.01
■ Pyruvic acid 0.1101
■Ascorbic Acid 0.03
■Progesterone 0.000003
■ Hydrocortisone 0.000002
■Ethanolamine 0.001227
■Putrescene 0.000164
■Linoleic Acid 0.0000841
■ CuS04. 5H20 0.0000025
■FeS04. 7H20 0.000834
■MnS04.5H20 0.0000003
■(NH4),Mo20 24.4H20 0.0000124
■ ZnS04.7H20 0.0008626
■NH4V 03 0.0000012
■ NiS04.6H20 0.0000007
■ SnCl2,H20 0.0000007
■ NaHSe03 0.0000008
■ P04 buffer 5mM
■ COMPOSITE
APPENDIX 2
A. SDS Polyacrylamide Gel Electrophoresis (PAGE)
Reagents
Acrylamide Stock Solution (30% w/v Acrvlamide. 2.5% w/v cross linker)
Acrylamide (BDH) 29.25g
bis-acrylamide (BDH) 0.75g
MQ (Milli Q Water, Millipore) to 100ml
Stored dark at 4°C
Resolving Gel Buffer (Tris 1.5M, SDS 0.4%)
Trizma base (Sigma) 181.5g
SDS (Sigma) 4.0g
MQ 800ml
pH adjusted to 8.9 with 0.1M HC1, and the volume adjusted to 11 with MQ.
Stacking Gel Buffer (Tris 0.5M, SDS 0.4%)
Trizma base (Sigma) 59.9g
SDS (Sigma) 4.0g
MQ 800ml
pH adjusted to 6.6 with 0.1M HC1, and the volume adjusted to 11 with MQ
Gel tank Buffer (Tris 0.05M, SDS 0.1%, Glycine 0.05M)
Trizma base (Sigma) 6.32g
Glycine (BDH) 4.0g
SDS (Sigma) l.Og
MQ to 11
Ammonium Persulphate (10% w/v in HoOl
Ammonium persulphate (BDH) 1.0g
MQ 10.0ml
Made fresh just prior to use.
Stain Solution
MQ 467ml
Aceitic Acid (BDH) 467ml
Methanol (BDH) 66ml
Coomassie Brilliant Blue (BDH) 2.0g
The solution was filtered using Whatman Nol paper prior to use.
Fix
MQ 467ml
Aceitic Acid (BDH) 467ml
Methanol (BDH) 66ml
Destain
MQ 800ml
Glacial Aceitic Acid (BDH) 100ml
Methanol (BDH) 100ml
Sample Buffer
SDS (BDH) 0.025g
Glycerol (BDH) 3.0ml
Bromophenol Blue (BDH) 0.15g
Tris buffer (150mM, pH 6.8) to 10.0ml
Protocol
Glass plates used to form the gel were cleaned in a 1% (v/v) solution
of "7X" detergent (FLOW Laboratories), followed by 0.1% (v/v) HC1, then
rinsed with MQ and finally washed with 70% (v/v) ethanol. The plates were
dried with non-fluff tissues and assembled using 1mm spacers and 50mm bull
dog clips. Acrylamide, resolving-gel buffer and water were mixed, as in Table
2.1a, and the solution degassed under vacuum for 10 minutes. Ammonium
persulphate solution and TEMED were added quickly, the solution mixed (by
swirling) and poured between the glass plates. To eliminate air, the upper
surface was overlayed with water-saturated butan-l-ol and the gel left to polymerise for 30 minutes.
Stacking gel was prepared as in Table 2.1b, and degassed prior to the
addition of ammonium persulphate and TEMED. The butan-l-ol was removed
from the resolving-gel and the surface rinsed with 5 ml of resolving-gel buffer.
Stacking gel was then poured onto the resolving gel, a comb inserted
carefully to avoid formation of air-bubbles, and the gel allowed to polymerise
for a further 30 minutes. The bottom spacer was removed and the gel, and
plates, assembled onto a vertical electrophoresis apparatus. Gel-tank buffer was
poured into the top reservoir of the apparatus and the comb removed. Buffer
was also poured into the lower reservoir and any air-bubbles accumulating along
the base were removed, and the wells washed, by application of a gentle stream of buffer using a syringe.
10/d of sample was added to 10/d of sample buffer and the solution
boiled for 2 minutes. Molecular weight markers (Sigma) were treated in a
similar way. Samples were loaded into the wells and the apparatus connected
to a power pack. The gel was developed at either 8mA constant current for 16
hours, or 3-4 hours at 40mA constant current. The gel was recovered by gently
levering the plates apart, placing in a tray containing stain and agitating gently
for 30 minutes. After which the stain was removed and the gel placed in fix
for 30 minutes. The fix was decanted, replaced by destain solution, which was
changed as soon as it became heavily coloured. The gel destained until blue
bands, due to the presence of protein could be determined. The gel was stored
in 3.5% (v/v) glacial aceitic acid for subsequent drying down and photography.
A slab-gel dryer (Biorad Lab. ltd.) was used to dry the gels; industrial
methylated spirit/dry-ice water traps being used to prevent liquid entering the
vacuum pump. When the gel was completely dry (approximately 2 hours), the
vacuum was released and the gel sandwiched between two glass plates to keep
it flat for photography.
Approximate molecular weights of the sample proteins could be
determined by comparison with the molecular-weight marker proteins as follows:
The relative mobility value (Rf) for the protein band was determined:-
Rf = distance moved bv protein
distance migrated by dye in loading buffer
When denatured by heating in the presence of a thiol agent such as B-
mercaptoethanol, most polypeptides bind SDS with the result that they have
essentially identical charge densities and migrate in the gel according to size.
A plot of Log10 polypeptide molecular weight against Rf should reveal a straight
line relationship (Shapiro et al, 1967; Weber, 1969), enabling the sample protein
band molecular weight to be determined.
B. Lowrv Protein Assay
Reagents
Lowry A (NagCOg 0.24M, Na/K Tartrate 0.0018M, NaOH 0.1M)
NagCOg
Na/K tartrate
20. Og
0.5g
4.0gNaOH
MQ to 11
Lowry B (CuS04 0.004M)
CuS04
MQ
l.Og
to 11
Lowry C
Lowry A
Lowry B
90.0ml
10.0ml
Prepared immediately before use.
Folin Ciolcalteau Reagent (50% v/v in MOl
Folin Ciolcalteau Reagent (BDH)
MQ 10.0ml
10.0ml
Prepared immediately before use
Bovine Serum Albumin Standard Solution. lmg/ml fSigmal
Diluted in MQ to give solution in the range 0 - 100 /zg/ml
Protocol
Sample or standard (200/d) was pipetted into sterile tubes in duplicate.
Lowry C (3ml) was added to each, mixed well and incubated at room
temperature for 15 minutes. 300/d of diluted Folin’s reagent was added to each,
mixed well and incubated for a further 30 minutes. The contents of each tube
were transferred to a disposable macro-cuvette and the optical density at 750nm
determined. Standards were used to construct a calibration curve from which the
concentration of sample was determined.
C. Glutamine and Ammonia Assay
Reagents
Sodium Acetate Buffer (0.1m. pH4.91
CH3COONa.3H20
MQ
1.36g
80ml
The pH was reduced to 4.9 with glacial aceitic acid, and the volume adjusted
to 100ml with MQ
Glutaminase (5U/mD
Glutaminase (Sigma) diluted to 5U/ml in sodium acetate buffer, 0.1M, pH 4.9
Sodium hydroxide fO.IMl
NaOH 4g
MQ to 100ml
Ammonium Chloride Standard (0.1M1
NH4C1 0.534g
MQ to 100ml
Diluted with Milli Q water to give a range of 10'1 - 10'6M
Samples of cell culture supernatant were collected by centrifugation at
400g for 5 minutes and 500/d placed in each of 4 sterile plastic bijoux (Flow).
Two samples were used as replicate ’blanks’ and two as replicate ’tests’. 450/d
of pre-warmed (37°C) sodium acetate buffer was added to each tube. A further
50/d was added to the ’blank’ tubes and 50/d of glutaminase to the ’test’
samples. The tubes were mixed thoroughly and incubated at 37°C for 30
minutes. The samples were removed and placed on ice to stop the reaction.
Just before monitoring the ammonia emission with the probe, 100/d of 0.1M
Protocol
sodium hydroxide was added. The reading for the ’blanks’ and ’test’ samples
were recorded. A standard curve using NH4C1 solutions (101 to 10'6M) was
constructed and used to determine the ammonium content of both the ’blank’
and ’test’ samples. A dilution factor of 2 was taken into account when the NH3
concentrations were determined. The ammonia observed in the glutaminase-free
’blanks’ was that already present in the cell culture medium. The NH3 in the
’test’ samples treated with glutaminase was therefore due to both the residual
ammonia present and also to the glutaminase. By subtraction of the blank value
from the test for each sample the ammonia due to the breakdown of glutamine
could be determined. As this is directly proportional, the glutamine concentration
can therefore be determined.
D. Lactate Assay
Reagents
Perchloric Acid (8%)
Perchloric Acid, 60% (BDH) 13.3ml
MQ to 100ml
Glycine (0.5M1 / Hydrazine Buffer (D.4M1. pH9.0
Glycine (BDH) 3.75g
Hydrazine Hydrate (BDH) 2ml
MQ 85ml
pH adjusted to 9.0 with HC1 (1M), and volume diluted to 100ml with MQ.
Lactate Standard
Lactic Acid Standard, 4.4mM (Sigma)
Nicotinamide Adenine Dinucleotide (NAD+1 Solution
NAD+ (Sigma), diluted to 30mg/ml in MQ.
Lactate Dehydrogenase (LDH)
LDH (Sigma) 550U/ml in ammonium sulphate suspension
Protocol
Sample Preparation
To remove any residual lactate dehydrogenase activity the samples and
standard (STD) were deproteinised. 500/d of sample (or STD) were pipetted into
a 1.5ml microfiige tube (Sarstedt) containing 1ml of ice-cold perchloric acid
(8 %). The tubes were agitated to ensure thorough mixing, placed on ice for 10
minutes followed by centrifugation for 10 minutes at 2000rpm.
Lactate Determination
200/d of deproteinised supernatant from each sample (or STD) were
pipetted into separate 7ml sterile bijoux (Sterilin) in duplicate. A "blank" of
200/d of perchloric acid was also included. 2.5ml of pre-warmed (37°C)
glycine/hydrazine buffer was added to each tube together with 200/d of NAD+.
The tubes were thoroughly mixed and the initial optical density determined at
340nm against air, using an Ultrospec II spectrophotometer (LKB-Pharmacia).
A 50/d aliquot of LDH suspension was added to each tube, the contents
mixed thoroughly, incubated at 37°C for 30 minutes and the final absorbance
of all the samples recorded. The lactate concentrations were determined using
the following equation:
[LACTATE] mM = (S - B) x V
e d v
Where:-
S = change in absorbance at 340nm for sample
B = change in absorbance for blank
V = Final volume of sample in cuvette
(= 2.95ml)
v = volume of sample in cuvette (= 0.67ml)
e = Molar extinction coefficient of NAD+ at 340nm (=6.22)
d = path length of the cuvette (= 10mm)
E. Ornithine Carbamvl Transferase Assay (EC 2.1.3.31
Reagents
Barbiturate/Acetate Buffer (70mM. pH 7.01 - solution 1
Solution a:
Sodium Acetate Trihydrate (BDH) 9.715g
Sodium Barbiturate (BDH) 14.715g
MQ to 11
Solution b: (HC1 50mMl
HC1 (BDH) 0.43ml
MQ to 100ml
Working Solution:
Solution a 20ml
Solution b 24ml
MQ to 100ml
pH checked at 7.0 and adjusted as necessary.
Urease (2500U/H - Solution 2
Urease (5U/mg,3(fC) 5mg
Solution 1 10ml
Substrate/Urease solution (carbamoyl-P. 23mM: Ornithine. 2.5mM: Urease.
2500u/H - Solution 3
Carbamoyl-P dilithium salt (Sigma) 17.5mg
Ornithine hydrochloride (Sigma) 4.0mg
Solution 2 5ml
Citrulline Standard Solution (TOmM) - Solution 4
Citrulline 17.5mg
MQ 10ml
Diluted 1ml to 10ml with MQ (= ImM)
Antipyrine/Ferric Sulphate Solution - Solution 5
100ml of concentrated H2S04 (BDH) was added carefully with cooling to 500ml
of MQ.
To this the following was added:
Fe2(S04)3 . 9H20 (Sigma) 50mg
Antipyrine (Sigma) 4g
Cooled and diluted upto 1000ml with MQ.
Aceitic Acid (5%v/vl - Solution 6
Acetic Acid (BDH) 50ml
MQ to 1000ml
Diacetylmonoxime (0.5%w/v) - Solution 7
Diacetylmonoxime (Sigma) 5.0g
Solution 6 1000ml
Chromogenic Reagent - Solution 8
Immediately before use, mix equal volumes of solutions 5 and 7
Protocol
The following were pipetted into labelled test tubes:
Total Blank Basal Blank CIT BlankCIT CIT Basal STD CIT
CIT STDSample 25/xl — 25/d --- -— —
Soln 3 50/d — 50/d --- --- —
Soln 2 — — 50/d 50/d — —
Soln 4 — — — 25/d ---
Water — 25/d --- 25/d — 25/dSoln 1 — — --- — — 50/d
The tubes were mixed carefully by gentle rotation, and incubated in a
covered water bath at 37°C for exactly 30 minutes. To each tube, 3.0ml of
chromogenic reagent (Solution 8) was added. The tubes were mixed by vortex,
incubated in a boiling water bath for 15 minutes followed by cooling under
running water and the optical density read at 460nm. The activity of OCT was
determined using the following equations:-
b = A x C = 33.33 x A U/l
(A^ x Rt) A*,
Where:-
A — A tobl - A baa,
A total = Absorbance of total citrulline - Absorbance blank citrulline
Afoasa! = Absorbance of basal citrulline - Absorbance blank basal
citrulline
Astd = Absorbance of standard - Absorbance of standard blank
Rt = reaction time ( = 30 minutes)
C = concentration of standard ( = lOOO^M)
F. Glutamate Dehydrogenase Assay (EC 1.4.1.3)
Reagents
TEA/EDTA/CELCOONHL (TEA. 64.5mM:EDTA. 3.22mM:CKCOONH,129mMV
Solution 1
Triethanolamine-HCL (TEA) (Sigma) 600mg
EDTA-Na2H . 2H20 (Sigma) 60mg
Ammonium Acetate (CH3COONH4) (BDH) 500mg
MQ 45ml
The pH was adjusted to 7.6 with NaOH (2M) and the volume increased to
50ml with MQ.
NADH/ADP/LDH fNADH. 6.45mM:ADP. 32.3mM:LDH 65kU/B - Solution 2
NADH di-Sodium salt (Sigma) 9.2mg
ADP monopotassium salt dihydrate (Sigma) 33mg
Lactate dehydrogenase (LDH) 0.2ml
(ammonium sulphate suspension, 3.2M; Sigma) 2mg/ml, 300U/mg)
2-Oxoglutarate (217 mM. pH 6 .8-7.0 1 - Solution 3
2-Oxoglutarate di-sodium salt (Sigma) lOOmg
MQ 2ml
50ml
2ml
Reaction Mix
Solution 1
Solution 2
Prepared immediately before use.
Protocol
Pre-warmed (37°C) reaction mix (2.5ml) was pipetted into a cuvette followed
by 500/d of pre-warmed (37°C) sample. The contents were mixed thoroughly
with a plastic spatula and incubated at 37°C for exactly 5 minutes. The optical
absorbance was measured at 339nm each minute for five minutes (the initial
incubation) and was used to determine AJT, the consumption of endogenous
pyruvate.
100/d of pre-warmed (37°C) 2-oxoglutarate was added, the solution mixed
thoroughly and the absorbance read each minute for 5 minutes to determine
AJT, the consumption of 2-oxoglutarate.
The activity of GIDH was calculated from the following equation.
b (U/l) = Af x V x 1000
e x d x Tf x v
Where:
b = the activity concentration of GIDH (U/l)
Af /7f = A JT - A JT
V = Assay volume, 1
v = Volume of sample used in assay, 1
d = light path, mm
e = linear millimolar absorbtion coefficient of NADH
APPENDIX 3
PROGRAMME FOR PARTIAL CUBIC SPLINE CURVE FITTING
(M.E. BUSHELL)
5MODEO 6 * C AT
10DIMX(60),Y(60),DY(60),A(60),B (60),C (60),D (60),slope(300) 20DIM R(60),R1(60),R 2 (60),T (60),T1 (60),U(60),V(300),J(300) 30CLOSE£040INPUT"NAME OF DATA FILE," FILE$50D=OPENIN FILE$60INPUT£D,N2 7OPRINT"ENTER X,Ys"80FOR R= 1 TO N2 90 INPUT ' £D , X ( R )•, Y ( R )
100IF Y ( R ) =0 TH-ENY(R) =0.0001110IFR=1 THEN 140120IFX(R)>X(R-l) THEN 14013OPRINT"NOT ALLOWED": GOTO 9 0140 NEXT R , -150CLOSE£0160N1 = 1 :S = 10017 OPROCFIT17 2PROCMEAN175 FOR 1 = 1 TO N 2 :D Y (I )=mean/fit : NEXT 180M1=N1-1 190M2=N2+12 0 0 R ( M l )=0 i .210R(N1 ) =0 ‘220R1 (N2 ) =0 230R2(N2)=0 240R2(M2)=0 250U(M l )=0 2 60U(N l )=0 2 70U(N 2 )=0280U(M2)=0 ; \2 90P = 0 300H1=N1+1 310M2=N2-13 20H=X(M l )- X (N l )330F=(Y(M1)-Y(N1))/H 340FOR 1= Ml TO M23 50G=H36 0H=X(I+1)-X(I)3 70E = F380F=(Y(I+1)-Y(I))/H3 90A(I )=F-E
,400T(I)=2*(G+H)/3 410T1(I )=H/34 20R2(I )=DY(1-1)/G 4 3 O R (I )=DY(I + 1 )/H4 4 0R l (I )=-D Y (I )/G-D Y (I ) /H 4 5 0NEXT460FOR 1= Ml TO M2470B(I)=R(I)*R(I)+Rl(I)*Rl(l)+R2(I)*R2(I)
480C(I)=R(I)*R1(I+1)+R1(I)*R2 (i + 1 )4 9 0 D ( I ) = R ( I ) * R 2 ( I + 2 )500NEXT510F2=-S520FOR 1= Ml TO M2 530R1(I-1)=F*R(I-1)
i 540R2(1-2)=G*R(1-2)! 5 50R(I)=1/(P*B(I)+T(I)-F*Rl(1-1)-G*R2(1-2))
5 6 0 U ( I ) = A ( I ) - R l ( 1 - 1 ) * U ( 1 - 1 ) - R 2 ( 1 - 2 ) * U ( 1 - 2 )570F=P*C(I)+T1(I)-H*Rl(1-1)580G=H 590H=D(I )*P 600NEXT610FOR I=M2 TO Ml STEP -1620U(I)=R(I)*U(I)-R1(I)*U(I.+ 1 )-R2(I)*U(I + 2)630NEXT640E=0650H=0660FOR I=N1 TO M2 670G=H680H=(U(I+1)-U(I))/(X(I+l)-X(I))6 90V(I)=(H-G)*DY(I)*DY(I)700E=E+V(I)*(H-G)710NEXT7 20V(N2)=-H*DY(N2)*DY(N2)7 30G=V(N 2 )740E=E-G*H750G=F2760F2=E*P*P770IF F2> = S OR F2<=G THEN 910 780F=0790H=(V(M1)-V(N1))/(X(Ml)-X(Nl) )800FOR 1= Ml TO M2 810 G = II820H=(V( 1+1)-V(I))/(X(I+l)-X(I))830G=H-G-R1 (T-l)*R(I-l)-R2(I-2)*R(I-2)840F=F+G*R(I)*G850R(I)=G860NEXT8 7 0H = E-P*F880IF H <= 0 THEN 910 890P=P+(S-F2)/((SQR(S/E)+P)*H)900GOTO 520 910FOR I=N1 TO N29 20A(I)=Y (I )-P*V(I )930C(I )= U ( I )940NEXT950FOR I=N1 TO M2 .960H=X(I+1)-X(I)970D(I)=(C(I+1)-C(I))/(3*H)980B(I)=(A(T+1)-A(I))/H-(H*D(I)+C(I))*H 990NEXT
1000CLS1010J=010 20FOR I = 1 TO N21030FOR XX=X(I) TO X( 1 + 1 )-X(N2 )/300. STEP X(N2 ) /300 1035J = J + 1 : IF J> 300 THEN 10601050 V ( J ) = (X( I )+H) : J'( J) = ( ( (D( I ) *H + C( I ) )*H+B( I ) )*H+A( I ) ).. 1060NEXT,2010MODE12020VDU19,3,6,0,0,0
.2030 MOVEIOO,100 2040PROCMAX 20 50YMAX=02060FORI=2 T03 0 0 :IFJ(I )>YMAX THEN 2 300207 0 NEXT20 7 5PROCYMAX2078PROCSLOPES2 080PROCaxes(XMAX,YMAX)2160FORI=2TO J-l21700=2 .2180PROCplot(V(I),J(I),C ,I ,YMAX)2200IF slope(I )<0 THEN 22202210 PROCplot(V ( I ),slope(I ),C - l ,I ,slopemax)2220NEXT 2 2 2 5PROCZMAX 2230GCOL0,12 2 50FORR=1TON2: MOVE(82 + X ( R )/XMAX*1087 ),( 108+Y(R)*720/YMAX) 2260PRINT"*"2 2 7 0NEXT 2 280PROCend 2290END2 3 00YMAX=INT(J(I)+l):GOTp20702310DEF PROCaxes ( XMAX , YMAX ) „TTT,Tri c m tmtt"232OCOLOUR1: PRINT"VALUES USING " ;lab;"% PARTIAL CUBIC SPLINE2330 PRINT"FILE "; FILES 2340PRINT"INPUT P FOR PRINTOUT"235OPRINT" N FOR DIFFERENT 0 OF POLY"2360PRINT" M TO EXIT TO MENU"2 3 70MOVE100,820 :DRAW100,100 : DRAW 127 9,100 2380VDU28,0,5,39,0 2 3 90VDU5: FORI = 0TO102400X=100 + 110*I:MOVEX,100 : DRAW X, 90 2410Y=100 + 70*I: MOVE 100,Y : DRAW 90,Y 2 4 20MOVE X - 2 0 ,30 : PRINT ;I*XMAX/10 24 3 0MOVE 0,Y + 10 : PRINT ; I*YMAX/10 : NEXT I 2440MOVE100,100 ,2450 Y=0 24 60ENDPROC2470 DEF PROCplot(X ,Y ,C ,I ,M A X )24 90 PLOT69,(100+X/XMAX*1087),(100+Y*7 20/MAX)2500 ENDPROC
2 51ODEFPROCend 2520CLOSE£0 2530INPUT X$2540IF X$="P" THEN *GIMAGE EPC T H&170 V&170 2550IF X$ = "N" T(HEN ENDPROC 2 5 60ENDPROC 2 5 7 ODEFPROCMAX 2580 XMAX=INT(X (N 2))2 5 90REPEAT2 6 0 OXMAX=XMAX+1 * ,261OUNTIL XMAX/10 = INT(XMAX/10)2 620ENDPROC 3000DEFPROCYMAX 300 5YMAX=YMAX-1 3010REPEAT3 0 2 0YMAX=YMAX+1 3030UNTIL YMAX =,I NT (YMAX)304 OENDPROC4000DEFPROCMEAN4005total=04010FOR R=1 TO N2 4020total=total+Y(R) ,4030NEXT4 040raean=total/N2 4050ENDPROC4 999DEFPROCFIT 5000INPUT "%FIT",X5 0 5 0 Y = - 2 4 0 . 6 0 2 2 5 9 + ( X * 1 2 . 0 9 7 0 1 2 ) - ( X * X * 0 . 2 2 2 7 5 1 9 3 5 ) + ( X * X * X * 2 . 0 4 0 7 7 5 4 1
X* 7.5 3 1 3 9 1 0 5 E - 6 ) - ( X * X * X * X * X * l.2 2210512E-8)-(X*X*X*X* E-3)-(X*X*X*X * X * 4 . 4 4 6 1 4 3 4 4 E - 1 0 ) + ( X * X * X * X * X * X * X * 6 . 8 7 6 0 8 9 6 5 E - 1 2 )5060f it=Y 50701ab=X 5080ENDPROC6000DEFPROCZMAX /;6050MOVE 1200 ,100
, 6065DRAW 1200,820 6070FOR I%= 1 TO 106080Z = 100 + 70*I%: MOVE 1200,Z:DRAW 1190,Z 6090MOVE 1030,Z +10 : PRINT ;I%*slopemax/10 6100NEXT7000DEFPROCSLOPES 7005slopemax=0 7010FOR 1 % = 2 TO 3007015 IF (V( 1%)- V (I%-1))=0 THEN slope(I%)=0 7017 IF (V(I%)-V(I%-1))=0 THEN 70507020slope(I%)=(J(I%)-J(I%-1))/(V(I%)-V(I%-l)):IF J(I%)=0 THEN 7050 7030IF slope(1%)<0 THEN 70507040 IF slope(1% ) > slopemax THEN slopemax = slope(1%)7050NEXT7052 slopemax=slopemax*1000 7054 slopemax=INT(slopemax j 7056 slopemax=siopemax/ÎOOO 7060ENDPROC
APPENDIX 4
BASIC PROGRAMME FOR PROCESSING PLACKETT-BURMAN DATA
(M.E. BUSHELL)
10DIM A$(24) ,B$(24 ,24) ,data(24) ,e ffect(24) , t (24) ,LABEL$(24)12INPUT"NEW OR EXISTING DATA (N OR E)",N$13 INPUT"NAME OF FILE",NAMES14 IF N$="E" THEN 16 15PROCINPUT16 IF N$="E" PROOREAD 17PROCEDIT 20PROCMATRIX 30PROCEFFECTS 40PROCDUMMY 50PROCSIGNIFICANCE 60PROCORDER
800END999DEFPROCMATRIX
1000REM VDU2
1020 PRINT" A rg Asn Asp Cys DUM Glu Gin Gly His OHp He LeuSer T h r T ry T y r DUM Val
1030FOR I%=1 TO 23 1040READ A$(I%)1050NEXT1060I%=11070FOR J%= 1 TO 23 1080FOR I%=1 TO 23 1090B$(I%,J%)=A$(I%)1100NEXT 1110A$(24)=A$(1 )1120FOR I%=1 TO 23 1130 A$(I%)=A$(I%+1)1140NEXT 1150NEXT1160FOR I%=1 TO 23 1170READ B$(I%,24)1180NEXT1190FOR J%= 1 TO 24 1200 PRINT J%;"1210FOR I%=1 TO 24 1220PRINT B$(I%,J%);"1230NEXT1240PRINT1250NEXT1280DATA1290DATA1300ENDPROC
2000DE F PROCIN P U T 2002CLOSE£0 2006Y=OPENOUT NAME$2010FORI%= 1 TO 242030PRINT"RESULT FOR TREATMENT ";I%;2040INPUT data(I% )2045PRINT£Y,data(I%)2050NEXT 2055CLOSE£0 2060ENDPROC 3000DEFPROCEFFECTS 3005FOR J%=1 TO 23 3007 PLUS=0:MINUS=0 3010FOR 1% - 1 TO 243020 IF B$(I%,J%)="+" THEN PLUS=PLUS + data(I% ) 3030 IF B$(I% ,J%)="-" THEN MINUS=MINUS + data(I% ) 3040 NEXT3045 effect(J%)=(PLUS/12)- (M INUS/12)3050NEXT3060ENDPROC4000DEFPROCDUMMY4010 Vef f= ( (e f fec t (5 ) '" 2)+(ef fec t (15)A2 ) + ( e f f e c t ( 2 2 ) /x2 ) ) / 3 4020 PRINT "VARIANCE IN THIS EXPERIMENT =";Veff 4030 PRINT"( 0 IS THE IDEAL RESULT)" 4035SE=SQR(Veff)4040 PRINT "STAN DARD ERROR =";SE 40501F SE=0 THEN SE=.00000000000000001 4055ENDPROC5000DEFPROCSIGNIFICANCE 5O1OFOR I%=1 TO 23 5020t ( 1% )=ef f ect( 1% )/S E 5025PRINT" VARIAS LE ";I% ," ",t(I% )5030NEXT 5040ENDPROC 6000DEFPROCEDIT 6001 CLOSE£06005INPUT"DO YOU WISH TO EDIT",N$6007 IF N$="N" ENDPROC 6020N=24:I=06030INPUT " NAME OF FILE ",NAMES6040Y=OPENIN NAMES6080REPEAT60901=1+16100INPUT£Y,data(I)6 1 10PRINT I" VALUE ";data(I)6120INPUT"CHANGE",QS 6130IF Q$="N" THEN 6200 6140PRINT"TYPE CHANGE X,Y"6150INPUT d a ta (I)6200UNTIL I=N 6240CLOSE£0 6250Y=OPENOUT NAMES
6280 FOR 1=1 TO N 6290 PRINT£Y,data(I)6310NEXT6320CLOSE£06900ENDPROC7000DEFPROCREAD7001 CLOSE£07010Y=OPENIN NAMES7015I=0:N=247020REPEAT70301=1+17040INPUT£Y,data(I)7050PRINT I " VALUE ";d a ta (I)7060UNTIL I=N 7070CLOSE£0 7080ENDPROC 8000DEFPROCORDER 8005 num%=238010LOCAL temp,swap,count%8020REPEAT8030swap=FALSE8040FOR count%=1 TO num%-180501F t(count% )>=t(count% +1 ) THEN 81008060temp=t( cou nt% )8070t(count%)=t(count%+1 )8080t(count%+1 )=temp8090swap=TRUE8100NEXT8110UNTIL NOT swap 812OFOR l%= 1 TO 23 8130PRINT LABELS(I%);t(I%)8140NEXT8150ENDPROC
>*SPOOL
REFERENCES
Adamson S.R., Behie L.A., Gaucher G.M., Lesser B.H. (1987) Metabolism of
hybridoma cells in suspension cultures. Evaluation of three commercially
available growth media. In "Commercial production of monoclonal
antibodies: A guide to scale-up" ppl7 (Seaver S.S., Ed) Marcel-Dekker,
N.Y.
Ah-Sing E.(1990). Personnel Communication
Alberts B., Bray D., Lewis L, Raff M., Roberts K., Watson J.D. (1983)
"Molecular biology of the cell" Garland, NY
Al-Rubeai M., Oh S.K.W., Musaheb R., Emery A.N. (1990) Modified cellular
metabolism in hybridomas subjected to hydrodynamic stresses. Biotech.
Lett. 12(5). 323-328
Arathoon W.R., Telling R.C. (1982) Uptake of amino acids and glucose by
BHK 21 Clone 13 suspension cells during cell growth. Dev. Biol. Stand
50, 145-154
Ardawi M.S.M., Newsholme E.A. (1982) Maximum activities of some enzymes
of glycolysis, the TCA, and ketone body and glutamine utilisation
pathways in lymphocytes of rat. Biochem. J. 208, 743-748
Atkinson D.E. (1977) "Cellular energy metabolism and its regulation". Academic
Press, New York
Bâcher M.P., Metgler L.S., Slaber P.L., Nevitt K.L., Boded G.B.(1986) Large
scale production of monoclonal antibodies in suspension culture. Presented
at the 1986 ACC Annual Meeting Division of Microbiology and Biotech.
Bagshawe K.D.(1987) Antibody directed enzymes revive anti-cancer pro-drugs
concept. Br. J. Cane. 56, 531-532
Bailey I.E., Ollis D.F. (1986) "Biochemical engineering fundamentals" (Second
edition). McGraw-Hill, New York
Batt B.C., Kompala D.S.(1989) A structured kinetic modelling framework for
the dynamics of hybridoma growth and monoclonal antibody production
in continuous suspension cultures. Biotech. Bioeng. 34, 515-531
Bebbington C., Hentschel C. (1985) The expression of recombinant DNA
products in mammalian cell. Trends in Biotechnology 3, 314-317
Bell S.L. (1990). Genetic modification of glutamine metabolism in hybridoma:
effect on cell physiology. PhD thesis, University of Surrey
Bender D.A.(1985) "Amino Acid Metabolism" (second edition). John Wiley and
Sons, Chichester.
Benjamin A.M., Verjee Z.H., Quastel J.H. (1978) effects of ammonium ions
on spontaneous action potentials and on the contents of sodium,
potassium, ammonium anc chloride ions in brain in vitro. J. Neurochem
30, 131-143
Birch J.R., Thompson P.W., Lambert K., Boraston R. (1985) The large scale
cultivation of hybridoma cells producing monoclonal antibodies. In "Large
-Scale Mammalian Cell Culture" (Feder J., and Tolbert W.R., eds)ppl
-18 Academic Press,
Birch J.R. (1984) Presented at the 188th meeting of the American Chemical
Society, Philadelphia, P.A. August 26-31
Bloom A.D., Nakamura F.T. (1974) Establishment of tetraploid immunoglobulin
producing cell line from hybridisation of two human lymphocyte lines.
Proc. Natl. Acad. Sci.(USA) 71, 2689-2692
Boraston R., Thompson P.W., Garland S., Birch J.R. (1984) Growth and
oxygen requirements of antibody producing mouse hybridoma cells in
suspension culture. Dev. Biol. Stand. 55, 103-111
Boyton A.L., Whitfield J.F.(1983) The role of cyclic AMP in cell proliferation :
A critical assessment of the evidence. Adv. Cyclic Nucleotides Res. 15, 193-294
Bree M.A., Dhuijai P., Geoghegan R.F., Robnett B. (1988) Kinetic modelling
of hybridoma cell growth and immunoglobulin production in large scale
suspension culture. Biotech. Bioeng. 32, 1067-1072
Brindle K.M., Braddock P., Sri-Pathmanathan R.M. (1990) Fourth Scientific
Report to SERC Biotechnology Directorate.
Brodski P.M., Parham P., Barnstable C.J., Crumpton M.J., Bodmer W.F.
(1979) Monoclonal antibodies for analysis of the HLA system. Imm.Rev. 47, 3
Buchanan T.W. (1984) immunochemical and diagnostic uses of monoclonal
antibodies in leprosy and gonorrhea. In " hybridoma technology in
agricultural and veterinary research" (Stem N.J., Gamble H.R.,) pplOl
-110. Rowman and Allahead, New Jersey.
Buell D.N. Fahey J.L. (1969) Limited periods of gene expression in
immunoglobulin-synthesizing cells. Science 164, 1524-1525
Bundersen G., Drake R.G., Kelley k., Worsley I.G., Friesen H.G., Sehon A.H.
(1980) Radioimmunoassay of human growth hormone using monoclonal
antibodies. J. Clin. Endocrin. 51, 1477-1479
Burnett F.M. (1957) "The clonal selection Theory Of Acquired Immunity".
Cambridge University Press
Burrows M.T. (1910) The cultivation of tissues of the chick embryo outside the
body. J. Amer. Med. Assoc. 55, 2057-2058
Burrows M.T., Neyman C.A. (1917) Studies in the metabolism of cells in vitro
I. The toxicity of a-amino acids for embryonic chicken cells. J. Exp. Med. 25, 93
Bushel! M.E., Fryday A. (1983) The application of materials balancing to the
characterisation of sequential secondary metabolite formation in
Streptomyces cattelya NRRL 8075. J. Gen. Micro. 129, 1733-1741
Bushell M.E., Bell S.L., Sanders P.G., Scott M.F., Spier R.E., Wardell J.N.
(1988) Third Scientific Report to the SERC Animal Cell Biotechnology Directorate.
Butler M.(1985) Growth inhibition in high density microcarrier cultures. Dev. Biol. Stand. 60, 269-280
Butler M., Thilly W.G.(1984) seeding density effect and amino acid utilisation. in vitro 18, 213
Byars N., Kidson C. (1970) Programmed synthesis and export of
immunoglobulin by synchronised myeloma cells. Nature 226, 648-650
Carrel A., Burrows M.T. (1911) On the physicochemical regulation of the
growth of tissues. The effects of the dilution of the medium on the
growth of the spleen. J.Exp.Med. 13, 562-570
Cartwright T., Crespo A. (1990) Production of a pharmaceutical enzyme: Animal
cells or E.coli. Presented at the 10th meeting of ES ACT, Avignon, May
7-11.
Ceriotti G.(1985) Ornithine carbamoyl transferase. In : Methods of enzymatic
analysis" (Bergmeyer H.U.,ed)pp319 Springer-Verlag, Munich
Chiang L.E., Silnutzer J., Pipas J.M., Barnes D.W. (1984) "Methods for
serum-free culture of epithelial and fibroblastic cells".p265 A.R. Liss.
New York.
Cleveland W.L., Wood I., Erlanger B.F. (1983) Routine large-scale production
of monoclonal antibodies in a protein-free culture medium. J Immunol,
Meth. 56, 221-234
Commoy-Chevalier M.J., Robert-Galliot B., Chany C. (1978) Effect of
ammonium salts on the interaction of the interferon-induced antiviral state
of mouse L cells. J.Gen. Virol. 41, 541-547
Coulepis A.G., Veale M.F., McGregor A., Komitsher M., Gust I.D. (1985)
Detection of hepatitus A virus with monoclonal antibodies. J. Clin.
Microb. 22, 119
Croce C.M., Kopowski K. (1974) The dominant and recessive nature of
malignancy. J.Exp. Med., 139, 1350
Darfler F J . Insel P. A. (1982) Serum-free culture of resting PHA-stimulated and
transformed lymphoid cell lines including hybridomas. Exp.Cell Res. 138. 287-279
Darzinliewicz Z., Traganos F. (1982) RNA content and chromatin structure
cycling in cycling and non-cycling cell populations studied by flow
cytometry. In "Genetic Expression in the Cell Cycle" (Padilla G.M.,
McCarty K.S., eds)ppl03-128. Academic press, NY.
Das M. (1980) mitogenic hormone-induced intracellular message: Assay and
partial characterisation of an activator of DNA replication induced by
epidermal growth factor. Proc. Natl. Acad. Sci.. USA. 77, 112-116
Davidson R.L., O’Malley K.A., Wheeler T.B. (1976) Polyethylene glycol-induced
mammalian cell hybridisation: effect of polyethylene glycol molecular
weight and concentration. Somat. Cell Genet. 2, 271-279
Dean R.T., Jessup W., Roberts C.R. (1984) The effects of exogenous amines
on mammalian cells with particular reference to membrane flow. Biochem.J. 217, 27
DeDuve C., deBarsy T., Poole B., Trouvet P., Tulkens P., VanHoof F. (1974)
Lysosomotrophic agents. Biochem. Pharmacol. 23, 2495
Dillman R.O., Royston I. (1984) Application of monoclonal antibodies in cancer
therapy. Br. Med. Bull. 40 (31. 240
Dodge T.C., and Hu W-S (1986) Growth of hybridoma cells under different
agitation conditions. Biotech. Lett. 8 , 683-686
Eagle H. (1955a) Nutrition needs of mammalian cells in tissue culture. Science
122. 501-504
Eagle H. (1955b) The specific amino acid requirements of a mammalian cell
(strain L) in tissue culture. J. Biol.Chem 214, 839-852
Eagle H.(1959) Amino acid metabolism in mammalian cell cultures. Science 130.
432-437
Eagle H., Piez K., Levy M. (1961) the intracellular amino acid concentration
required for protein synthesis in cultured human cells. J. Biol. Chem.
2039-2042
Eagle H., Piez K. (1962) The population-dependent requirement by cultured
mammalian cells for metabolites which they can synthesize. J. Exp. Med.
116. 29-43
Earle W.R., Sanford K.K., Evans V.J., Waltz H.K., Shannon J.E. (1951) The
influence of inoculum size on proliferation in tissue culture. J. Natl.
Cane. Inst. 12, 133
Emery A.N., Lavery M., Williams B., Handa A. (1987) Large scale hybridoma
culture. In "Plant and Animal Cells: Process Possibilities" (Webb C.,
Mavituna F. eds) pp 137-148. Ellis Norwood Ltd., Chichester.
Enders J.R., Weler T.H., Robbins F.C. (1949) Cultivation of the lansing strain
of poliomyelitus virus in cultures of various human embryonic tissues.
Science 109, 85-86
Eperietos A., Britton K.E., Mather S., Shepherd J., Granowska M., Taylor
-Papadimitriou J., Nimmon C.C., Durbin H., Hawkins L.R., Malpas J.S.,
Bodmer W.F. (1982) Targeting of I125 labelled tumours associated
monoclonal antibodies to ovarian, breast, and gastrointestinal tumours.
Lancet, 2 999
Epstein N., Epstein M. (1986) The hybridoma technology: I Production of
monoclonal antibodies. In "Advances in Biotechnological Processes Vol
6 " (Mizrahi, A., Ed) ppl79-218, Alan R. Liss Inc.NY.
Fantl V.E., Wang D.Y. (1982) Production and characterisation of a mammalian
antibody to progesterone. J. Steroid Biochem. 17, 125
Farrands P., Pimm M.V., Embleton M.J. (1982) Radioimmunodetection of
human colorectal cancers by an antitumour monoclonal antibody. Lancet
2, 397
Fazekas de St Groth S. (1983) Automated production of monoclonal antibodies
in a cytostat. J. Immunol. Method. 57, 121-136
Ferguson M., Schild G.C., Minor P.D., Yates B.J., Spitz M. (1982) A
hybridoma line secreting antibody to polio virus type 3-D antigen:
detection in virus harvest of two D- antigen population. J. Gen Virol.
54, 437-442
Fleishaker R.J., Sinskey A.I., (1983) Oxygen demand and supply in cell culture.
Eur. J. Appl. Microbiol. Biotech. 12, 193-197
Floras J., Chang H., Baserga R. (1978) Stimulated DNA synthesis in frog
nuclei by cytoplasmic extracts of temperature-sensitive mammalian cells.
Science 201, 651-652
Fischer A. (1941) Die bedeuntung der cuminosauren fur die Gewebezellen in
vitro. ACTA physiol. Scand. 2, 143
Fischer A. (1948) growth of animal cells in artificial medium. Proc.Soc. Exp
Biol. Med. 67, 40-46
Fung M.S.C., Sun N-C., Chang N.T., Chang T.W. (1987) Monoclonal
antibodies that neutralise HIV-1 virions and inhibit syncytium formation
by infected cell. Bio/technology 5, 940-946
Furasawa E., Cutting W. (1962) Inhibitory effects of ammonium sulphate on
Columbia SK virus propagation in mouse ascites tumour cells in vitro
Pro. Soc. Exp. Biol. Med. H i , 71
Galante Y.M. (1989) Genetically engineered human therapeutic drugs: a survey
of registered approved products undergoing registration and development.
Chimiacaogg 7, 89-93
Galfre G., Howe S.C., Milstein C., Butcher G.W., Howard J.C. (1977)
Antibodies to major histocompatibility antigens produced by hybrdi cell
lines. Nature 266, 550-555
Galfre G., Milstein C. (1981) Preparation of monoclonal antibodies: strategies
and procedures, methods in Enzymol. 73, 3-46
Garatun-Tjelsto O., Pryme I.F., Weltman J.K., Dowben R.M. (1976) Synthesis
and secretion of light-chain immunoglobulin in two successive cycles of
synchronised plastocytoma cells. J.Cell.Biol. 6 8 , 232-239
Gerhard W., Yewdell J., Frankel M.E., Webster R. (1981) Antigenic structure
of influenza virus haemagglutinin defined by hybridoma antibodies. Nature
290. 713-717
Glacken M.W., Fleischaker R.J., Sinskey A.J. (1986) Reduction of waste
product excretion via nutrient control : possible strategies for maximising
product and cell yields on serum on cultures of mammalian cells.
Biotech. Bioeng. 28, 1376-1389
Glacken M.W. (1988) Catabolic control of mammalian cell culture.
Bio/Technology 6 , 1041-1050
Glacken M.W., Adema E., Sinskey A.J. (1989) Mathematical descriptions of
hybridoma culture kinetics: II The relationship between thiol chemistry
and the degradation of serum activity
Glassy M.C., Tharakan J.P., Chau P.C. (1988) Serum-free media in hybridoma
culture and monoclonal antibody production. Biotechnol. Bioeng. 32,
1015-1028
Coding J.W. (1983) "Monoclonal antibodies: principles and practice". Academic
Press. New York.
Griffiths J.B. (1972) The effect of cell population density on nutrient uptake and
cell metabolism: a comparative study of human diploid and heteroploid
cell lines. J. Cell Sci 10, 515-524
Griffiths J.B., Pirt S.J. (1967) The uptake of amino acids by mouse cells (strain
LS) during growth in batch culture and chemostat culture: the influence
of cell growth rate. Proc. Roy. soc. B. 168, 421-438
Gutowski J.K., Cohen S. (1983) Induction of DNA synthesis in isolated nuclei
by cytoplasmic factors from spontaneously proliferating and mitogen
-activated lymphoid cells, cell Immunol. 75, 300-311
Ham R.G.(1984) Formulation of basal nutrient media. In "Methods for
preparation of media supplements, and substrata for serum-free animal
cell cultures (Barnes., Sirbasku., Sato., eds) pp3-21. A.R. Liss N.Y.
Ham R.F., Swim H.E. (1957-58). The amino acid requirements of rabbit
fibroblasts strain RM 3-56. J. Gen. Physiol. 41,91
Hames D.B. (1981) An introduction to polyacrylamide gel electrophoresis. In
"Gel Electrophoresis of Proteins: a practical approach" (Eds Hames B.D.,
Rickwood D.) ppl-86 IRL press, Oxford.
Handa-Corrigan A.(1990) Oxygenating animal cell cultures: the remaining
problems. In "Animal Cell Biotechnology Vol 4". (Spier R.E., Griffiths
J.B. Eds). ppl23-131. Academic Press London.
Handa-Corrigan A., Emery A.N., Spier R.E. (1989). Effect of gas-liquid
interfaces on the growth of suspended mammalian cells: mechanisms of
cell damage by bubbles. Enzyme and Microbial Technol. H , 230-235
Harrison R.G. (1907) Observations on the developing living nerve fibre. Proc.
Soc. Exp. Biol. Med. 4, 140-143
Hochhauser S.J., Stein J., Stein G.S. (1981) Gene expression and cell cycle
regulation. Int. Rev. Cytol. 71, 95-243
Hoffman G.J., Sondra G., Lazarowitz H., Hayward S.D.(1986) Monoclonal
antibodies against a 250,000 dalton glycoprotein of Epstein-Barr virus
identified as a membrane antigen and a neutralising antibody. Proc. Natl.
Acad. Sci 77, 2979-2983
Holley R.W., Kieman J.A. (1971) Studies on serum factors required by 3T3
cells and 3U3T3 cells. In "Growth control in cultures" (Wolsenholme
G.G., Knight J., eds) pp3-10 Churchill Livingstone, London
Holmes B.E., Watchom E. (1927) Studies in the metabolism of tissues growing
in vitro I. Ammonia and urea production by the kidney. Biochem J. 21,
327-334
Hood L.E.., Weismann I.L., Wood W.B., Wilson J.H. (1984) "Immunology"
(second edition). The Benjamin Cummings Publishing Co, California.
Howard J.C., Butcher G.W. (1979) Antibodies as tools to analyse serological
and genetic complexities of major transplantation antigens. Immunol. Rev.
47, 139r -
Hume D.A., Radik J.L., Berber J.L., Weideman M.J. (1978) Aerobic glycolysis
and lymphocyte transformation. Biochem J. 174, 703-709
Hume D.A., Weidemnan M.J.(1979) J. Natl Cancer Inst. 62, 3
Ito M., McLimans W.F. (1981) Ammonia inhibition of interferon synthesis.
Cell Biol Int. Rep. 5, 661-666
Jack G.W., Wade H.E. (1987) Immunoaffinity chromatography of clinical
products. Trends in Biotech. 5, 91-95
James K. (1990) therapeutic monoclonal antibodies-their production and
applications. In "Animal cell biotechnology volume 4" (Spier R.E.,
Griffiths J.B., eds) pp205-258. Academic Press., London
James K., Bell G.T. (1987) Human monoclonal antibody production. Current and
future prospects. J. Immunol. Method. 100, 5-40
Jazwinsld S.M., Wang J.L., Edelamn G.M. (1976) initiation of replication in
chromosomal DNA induced by extracts from proliferating cells. Proc.
Natl Acad. Sci. USA. 73, 2231-2235
Jeme N.K., Nordin A.A. (1963) Plaque formation in agar by single antibody
producing cells. Science 140. 405
Jensen E.M., Liu O.C. (1961) Studies of the inhibitory effects of ammonium
ions in several virus-tissue culture systems. Proc. Soc. Exp. Biol. Med.
107. 834-838
Jeme N.K. (1974) Towards a network theory of the immune system. Ann.
Immunol. (Paris) 125, 373
Jessup W., Shirazi M.F., Dean R.T. (1983) Inhibition of some spontaneously
secretory processes in macrophages and fibroblasts by ammonium chloride.
Biochem. Pharmacol. 32.(181 2703
Karkare S.B., Phillips P.G., Burke D.H., Dean R.C. (1985) Continuous
production of monoclonal antibodies by chemostatic and immobilised
hybridoma culture. In "Large scale mammalian cell culture" (Feder J.,
Tolbert W.R., eds) ppl28-150. Academic Press, New York
Kemshead J.T., Heath L., Gibson E.M., Katz F., Richmond F., Treaven J.,
Vgelstad J.(1986) Magnetic micrspheres and monoclonal antibodies for
the depletion of neuroblastoma cells fro bone marrow: experience,
improvements and observations. Br. J. Cane. 54, 771-778
Kemey J.F., Radburch A., Lyesegang B., Rajewsky K. (1979) A new mouse
myeloma cell line that has lost immunoglobulin expression but permits
the construction of antibody secreting hybrid cell lines. J. immunol.
meth. 123, 1548-1550
Kluft C., van Wezel A.L., van der Velden C.A.M., Emeis J.J., Verheijen J.H.,
Wijngards G.(1983) Large-scale production of extrinsic (tissue-type)
plasminogen activator from human melanoma cells. In "Advances in
Biotechnological Processes Vol2:. (Mizrahi A., and van Weizel A.L.,
eds) pp 97-110, Alan R.Liss, new York.
Kohler G., Milstein C. (1975) Continuous cultures of fused cells secreting
antibody of predefined specificity. Nature 256, 495-497
Kozbor D., Lagarde A.E., Roder J.C. (1982) Human hybridomas constructed
with antigen specific EPV-transformed cell lines. Proc. Natl Acad. Sci.
79, 6651-6655
Laemmli U.K. (1970). Cleavage of structural proteins during the assembly of
the head of bacteriophage T4. Nature 227, 680-685
Lambert K., Birch J.R. (1985) Cell growth media. In "Animal Cell
Biotechnology" (Spier R.E., Griffiths J.B., eds) pp85-122, Academic
Press, London.
Lavery M., Emery A.N., Jefferies R., keams M.J., Price D.G. Hybridoma
growth and metabolism in serum supplemented and defined ’serum-free’
medium (unpublished)
Lazar A. Reuveny S., Geva J., Marcus D., Silberstien L., Ariel N., Epstein
N., Altbaum Z., Sinai J., Mizrahi A. (1987) Production of
carcinoembryonic antigen from a human colon adenocarcinoma cell line.
I Large scale cultivation of carcinoembryonic antigen producing cells on
cylindric cellulose-based microcarriers. Dev. Biol. Stand. 66, 423-428
Lebhertz W.B. (1987) Batch cultivation of monoclonal antibody by large scale
suspension culture. In "Commercial production of monoclonal antibodies:
A guide to scale-up" (Seaver S.S., ed) pp93, Marcel-Dekker, NY
Leffert H.L., Kock K.S., Moran T., Rubalcava B. (1979) Hormonal control of
rat liver regeneration. Gastroenterology 76, 1470-1482
Lehninger A.L. (1975) "Biochemistry" Worth Publishers, NY
Leudeking R.(1967) Fermentation process kinetics. In "Biochemical and
Biological Engineering Vol 1" (Blakebrough N., ed) pp205 Academic
Press London.
Levintow L., Eagle H., Piez K.A. (1957) The role of glutamine in protein
biosynthesis in tissue culture. J. Biol. Chem 227, 929-941
Levintow L., Eagle H. (1961) Biochemistry of cultured mammalian cells. Ann.
Rev. Biochem. 30, 605-640
Lewis M.R.(1922) Importance of dextrose in the medium for tissue culture.
J.Exp.Med. 53, 317-322
Lewis M.R., Lewis W.H. (1912) the cultivation of chick tissues in media of
known composition. Anat. Rec. 6, 207-211
Lowry O.H., Rosebrough N.J., Farr A.L., Randall A.J. (1951) Protein
measurement with the Folin-phenol reagent. J. Biol. Chem. 193, 265-270
Littlefield J.W. (1964) Selection of hybrids from matings of fibroblasts in vitro
and their presumed recombinants. Science 145, 709-710
Lloyd D., Poole R.K., Edwards S.W. (1982) "The cell division cycle: temporal
organization and control of cellular growth and reproduction". Academic
Press, Irandon
Luan Y.T., Mutharasan R., Magee W.E. (1987) Factors governing lactic acid
formation in long term cultivation of hybridoma cells. Biotech lett. 9,
752-756
Lubiniecki A.S. (1987) Pharmaceutical applications of recombinant DNA-modified
mammalian cells. Developments in Industrial Microbiology 28, 133-138
Macmillan J.C., Wheatly D.N. (1981) Amino acid pools in mitotic and
interphase HeLa cells. Exp Cell Res. 133, 470-475
Male D., Champion B., Cooke A. (1989) "Advanced immunology" Gower
medical, London.
Marder P., Maciak R.S., Fonts R.L., Baker R.S., Starling J. (1990) selective
cloning of hybridoma cells for enhanced immunoglobulin production
using flow cytometric cell sorting and automated laser nephelometry
Cytometry 11(4). 498-505
Matsushita S., Robert-Giroff M., Rusche J., Koito A., Hatton T., Hoshino H.,
Javaherian K., Takiatsuti K., Potney S. (1988) Characterisation of a
human immunodeficiency virus neutralising monoclonal antibody, and
mapping of the epitope. J. Gen. Virol. 62, 2107-2214
McCoy T.A., Maxwell M., Neuman R.E. (1956) The amino acid requirements
of the Walker carcinoma 256 in vitro. Cane Res. 16, 979
McCoy T.A., Maxwell M., Kruse P.P. (1959a) the amino acid requirements of
the Jensen sarcoma in vitro. Cane. Res. 19, 591
McCoy T.A., Maxwell M., Kruss P.P. (1959b) Amino acid requirements of the
Novikoff hepatoma in vitro .Proc. Roy. Soc. Exp. Biol. 100, 115
McCullough K.C., Spier R.E. (1990) "Monoclonal Antibodies in Biotechnology:
Theoretical and Practical Aspects". Cambridge University Press.
McKeehan W.L. (1982) Glycolysis, glutaminolysis and cell proliferation. Cell
Biol Int. Rep. 6, 635
Meister A. (1965) "Biochemistry of the amino acids" second edition ,volume 2.
Academic Pres., NY
Mercer W.E., Nelson D., Deleo A.B., Old L.J., and Baserga R. (1982)
microinjection of monoclonal antibody to protein p53 inhibits serum
induced DNA synthesis in 3T3 cells. Proc. Natl. Acad. Sci. 79,630962
Merten O-W., Reiter S., Himmler G., Scheire W., Katinger H. (1985)
Production kinetics of monoclonal antibodies. Dev. Biol. Stand. 60, 219
-227
Merten O-W. (1987) Concentrating mammalian cells. I Large Scale animal cell
culture. Trends in Biotechnology 5, 230-237
Meyer L.J., Lafferty M.A., Raducha M.G., Foster G.I., Goglin K.J., Harris
H. (1982) Production of a monoclonal antibody to human alkaline
phosphatase. Clin. Chem ACTA. 126, 109
Miller W.M., Blanch H., Wilke C. (1986) Kinetic analysis of hybridoma growth
in continuous suspension culture. Presented at ACS National Meeting,
September.
Mitchison J.M. (1971) "The Biology of The Cell Cycle". Cambridge University
Press.
Monod J. (1942) "Recherches sur la croissance des cultures bactériennes".
Hermann, Paris
Morgan J.F., Morton H.J., Parker R.C. (1950) Nutrition of animal cells in
tissue culture. I. initial studies on a synthetic medium. Proc. Soc. Exp.
Biol. Med. 73, 1-8
Morgan J.F., Morten H. (1957) The nutrition of animal tissues cultivated in
vitro. IV Amino acid requirements of chick embryonic heart fibroblasts.
J. Biophys. Biochem. Cytol. 3, 141
Ohmori H., Yamamoto I. (1982) Mechanisms of augmentation of the antibody
response in vitro by 2-mercatoethanol in murine lymphocytes. J. Exp.
Med. 155, 1277-1290
Ohmori H., Yamamoto I. (1983a) Mechanism of augmentation of the antibody
response in vitro by 2-mercaptoethanol in murine lymphocyes II. A
major role of mixed disulfide between 2-mercaptoethanol and cysteine.
Cell Immunol 79, 173-185
Ohmori H., Yamamoto I. (1983a) Mechanism of augmentation of the antibody
response in vitro by 2-mercaptoethanol in murine lymphocyes HI serum
bound and oxidised 2-mercaptoethanol are available for the augmentation.
Cell Immunol 79, 186-196
Patterson M.K.(1972) Uptake and utilisation of amino acids in cell cultures. In
" Growth, nutrition and metabolism of cells on culture" (Rothblat G.H.,
Cristofalo V.J., eds) pl71. Academic press, NY
Pederson P.L. (1978) Prog. Exp. Tumour Res. 22, 190
Petersen I.E., Mclntire L.V., Papoutsakis E.T. (1988) Biotechnology 7, 229
Phillips A.W., Ball G.D., Fantes K.H., Finter N.B., Johnston M.D. (1985)
Experience in the cultivation of mammalian cells on the 80001 scale. In
"Large-scale mammalian cell culture" (Feder J., and Tolbert W.R., eds)
pp87-96 Academic press. New York.
Phillips H.A., Scharer J.M., Bols N.C., Moo-Young M. (1987) Effect of
oxygen on antibody productivity in hybridoma culture. Biotech. Lett. 9,
745-750
Piez K.A., Eagle H. (1958) The free amino acid pool of cultured human cells.
J Biol Chem. 231, 533-545
Pirt S J . (1985) "Principles of Microbe and Cell Cultivation". Blackwell Scientific
Publications, Oxford.
Pirt S.J., Righelato R.C. (1967) Effect of growth rate on the synthesis of
Penicillin by Pénicillium chrysogenum in batch and chemostat culure.
Appl. Microbiol 15, 1284-1290
Plackett R.L., Burman J.P.(1946) The design of optimum multifactorial
experiments. Biometrika, 33, 305
Pollin G.A. (1980) Monoclonal antibodies against streptococcal antigens. In
"Monoclonal antibodies and hybridomas: a new dimension in biological
analysis" (Kennet R.H., McKeam T.T., Bechtol K.B., eds) pp353-360,
Plenum press, NY
Prentice H.G., Janossy G., Price-Jones L., Trejdosiewicz L.K., Panjwani D.,
Graphakos S., Ivory K., Blacklock H.A., Gilmore M.J., Tidman N.,
Skeggs D.B., Bill S., Patterson J., Hoffbrand A. (1984) Depletion of
T-lymphocytes in donor marrow prevents significant graft-versus-host
disease in matched allogenic leukaemic marrow transplant recipients.
Lancet i, 472
Rao P.N., Johnson R.T., Sperming K. (1982) "Premature chromosome
condensation: application in basic clinical and mutation research"
Academic press, NY.
Reitzer L.J., Wice B.M., Kennel D. (1979) Evidence that glutamine, not sugar,
is the major energy source for cultured HeLa cells. J. Biol. Chem. 254.
2669-2676
Remmers J. (1982) production and characterisation of monoclonal antibody to
cholera toxin. Infection and Immunity. 37, 70-76
Reuveny S., Velez D., MacMillan J.D., Miller L. (1986) Factors affecting cell
growth and monoclonal antibody production in stirred reactors. J.
Immunol. Meth. 86, 53-59
Reuveny S. Lazar A. (1989) Equipment and procedures for production of
monoclonal antibodies in culture. In "Monoclonal antibodies: products
and applications. Adv. in Biotech. Proc. vol 11" (Mizrahi A., ed). A.R.
Liss, NY
Ringold G., Dieckmann B., lee F.J.(1981) Co-expression and amplification of
dihydrofolate reductase cDNA and Escherichia coli xGPRT gene in
Chinese Hamster Ovary cells. Mol Appl. Genet, i , 165-175
Roberts R.S., Hsu H.W., Lin K.D., Yang T.J. (1976) Amino acid metabolism
of myeloma cells in culture. J cell Sci. 21, 609-615
Roitt I. (1988) "Essential immunology"
Rose W.C.(1938) the nutritive significance of the amino acids. Physiol. Rev.
18, 109
Rous P., Jones F.S.(1916) A method for obtaining suspensions of living cells
from the fixed tissues, and for plating out of individual cells. J. Exp.
Med. 23, 549-555
Roux W. (1885) Beitrage zur entwicklungsmechanik des embryo. Z.
Biol. (Munich) 21, 411-524
Rowe D.A (1980) Immunology Today 1, 30.
Rowley B., Pirt S.J. (1972) Melatonin production by Apergillus nidulans in
batch and chemostat culture. J. Gen Micro. 72, 553-563
Ruddle F.H., Kucheleral R.S (1974) Sci Amer. 231, 39
Rudge J., Desai M.A., Shojodsadaty S.A., Hyddiatt A. (1987) Continuous
culture of murine hybridomas with integrated recovery of monoclonal
antibody. In " Modem approaches to animal cell technology" (Spier
R.E., Griffiths J.B., eds) pp556 Butterworths, Guildford.
Rupp R.P. (1985) Use of cellular microencapsulation in large-scale production
of monoclonal antibodies. In "Large Scale Mammalian Cell Culture"
(Feder J., Tolbert W.R.,eds) ppl9-38, Academic press, New York
Ryan W.L., Cardin C. (1966) Amino acids and ammonia of fetal calf serum
during storage, proc. Soc. Exp. Biol. Med. 123, 27
Schmidt E., Schmidt F.W. (1985) Glutamate dehydrogenase. In "Methods of
enzymatic analysis" (Bermeyer H.U., ed) pp217. Springer-Verlar, Munich
Scwaber J., Cohen E.P. (1974) pattern of immunoglobulin synthesis and
assembly in a human-mouse somatic cell hybrid clone. Proc. Natl Acad. Sci. 71, 2203
Seaver S.S., Rudolph J.L ., Gabriels J.E. (1984) A rapid HPLC technique for
monitoring amino acid utilisation in cell culture. BioTechniques 2, 254 260
Secher D.S., Burke D.C. (1981) A membrane-based system for large-scale
purification of human leukocyte interferon. Nature 285, 446-450
Seglen P.O., Reith A. (1977) Ammonia inhibits protein secretion in isolated rat
hepatocytes. Biochim. Biophys. ACTA.496, 29
Shapiro A.L., Vinuela E., Maizel J.V. (1967) Biochem. Biophs. Res. Commun.
28, 815
Sher A., Snary D. (1985) Specific inhibition of morphogenesis of Trypanosoma
cruzi by a monoclonal antobody. Nature 300, 639-640
Shulman M., Wilde C.D., Kohler G. (1978) A better cell-line for making
hybridomas secreting specific antibodies. Nature 276, 269-270
Sikovirs J.G., Shirato J.G., Gyroky F., Cabiness J.R., Howe C.D. (1970)
"Leukaemia lymphoma yearbook". Medical Publishers, Chicago.
Sinclair C.G., Kristianson B. (1987) "Fermentation kinetics and modelling"
(Bu’Lock J.D., ed). Open University press, Milton Keynes.
Skinner M.A., Ting R., Langlois A.L, Weinhole K.J., Lyerly H.K., lavanevian
K., Mathews T.J. (1988) Characterisation of a neutralising monoclonal
antibody to HIV enveloped proteins. AIDS Res. Human retroviruses 4,
187-197
Spier R.E. (1990) Contemporary issues in animal cell biotechnology. In "Animal
Cell Biotechnology Volume 4." (Spier R.E., Griffiths J.B., eds) ppl-15
Academic Press, London.
Stahli G., Staehelin T., Miggianov V., Schmidt J., and Haring P. (1980) High
frequencies of Ag-specific hybridoma: dependence on immunisation
parameters and production by spleen cell analysis. J. Immunol Methods
32, 297-304
Storch M.J., Licht T., Petersen K.G., Obermeier R., Kerp L. (1987) Specificity
of monoclonal anti-human insulin antibodies. Diabetes 36, 1005
Stuart M.C., Walinchowski C.M., Underwood P.A., Hussain S., Harman D.F.,
Rathjen D.A., Van Stumer S.P. (1983) monoclonal antibodies to human
growth hormone. J Immunol Meth. 61, 33
Sullivan C.W., Volcani B.E. (1981) Silicon in the metabolism of diatoms. In
"Silicon and siliconaceous structures in biological systems" (Simpson
T.L., Volcani B., eds) pp 15-42. Springer-Verlag, Berlin
Takahaski M., Yagi Y., Moore G.E., Pressman D. (1969) Immunoglobulin
production in synchronised cultures of human haematopoietic cell lines.I
variation of cellular immunoglobulin level with the generation cycle. J
Immunol. 103, 834-843
Tharakan J.P., Chau P.C. (1986) serum-free fed-batch production of IgM.
Biotech. Lett. 8, 457
Tharakan J.P., lucas A., Chau P.C.(1986) Hybridoma growth and antibody
secretion in serum-supplemented and low protein serum-free media. J.
Immunol. Meth. 94, 225-235
Thomas E.K., Weber J.N., McClure J., Claphamm P.R., Singhal M.C., Shriver
M.K., Weiss R.A. (1988) Neutralising monoclonal antibodies to the AIDS
virus. AIDS 2, 25-29
Thornhill P. (1989) Personal communication
Thorpe J.S. (1989) An examination of the growth inhibitory metabolites of a
murine hybridoma cell line. PhD Thesis, University of Surrey.
Tonegawa R. (1983) Nature 276, 269-270
Toohey J.L (1975) Sulphydryl dependence in primary explant haemopoietic cells
inhibition of grwoth. Proc. Natl. Acad. Sci 72, 73-77
Van Brunt J. (1988) Molecular farming: transgenic animals as bioreactors.
Bio/technology 6, 1149-1154
Velez D., Reuveny S., Miller L., MacMillan J.D. (1986) Kinetics of
monoclonal antibody production in low serum growth medium. J
Immunol Meth. 86, 45-52
Warburg O. (1930) "The metabolism of tumours". Constable, London
Warburg O., Kubowitz F. (1927) Stoffwechsel wachsender zellen. Biochem Z.
189. 242-248
Weber K., Osborn M. (1969) The reliability of molecular weight determination
by dodecyl suplphate-polyacrylamide gel electrophoreisis.J. Biol. Chem.
244. 4406
Wergeland I., Walker C., Ericcson R. (1987) Hybridoma culture for continuous
production of monoclonal antibody: a fractional factorial design experiment
for process development. In "Modem approaches to animal cell
technology" (Spier R.E., Griffiths J.B., eds) pp795. Butterworths,
Guildford.
White P.R.(1949) Prolonged survival of excised animal tissue in nutrients of
precisely known constitution. J Cell. Comp. Physiol. 34, 7812-7819
Whitfield J.F., MacManus J.P., Boynton A.L., Durkin J., Jones A. (1982)
Futures of calcium, calcium-binding protein, cyclic AMP and protein
kinases in the quest for an understanding of cell proliferation and cancer.
In "Functional regulation at the cellular and molecular levels: (Corradino
R.A., ed) pp 61-87. Elsevier, north Holland.
Whitfield J.F., Boynton A.L., Rixon R.H., Toudale T. (1985) The control of
cell proliferation by calcium, Ca2+-calmodulin, and cyclic AMP. In
"Control of animal cell proliferation" pp331-365, Acad. Press, NY.
Whitson G.L. (1980)" Nuclear-cytoplasmic interactions in the cell cycle"
Academic press, NY.
Wice B.M., Reitzer L.J., Kennell D. (1981) the continuous growth of vertebrate
cells in the absence of sugar. J Biol. Chem. 256, 7812-7819
Wiktor T.J., Koprowski H. (1979) monoclonal antibodies against rabies virus
produced by somatic cell hybridisation: detection of antigenic variants.
Proc. Natl Aca. sci. USA 75, 3938-3942
Wolf C.G.L. (1914) The determination of lactic acid. J. Physiol 48, 341-3
Wright A.F., Green T. P., Smith L.L. (1987) The development of mouse
monoclonal antibodies that bind and neutralise paraquat. Dev. Biol. Stand.
66, 495-501
Yoshida N., Nussenzweig P., Potocnjak V., Nussenzweig V., Aikawa M. (1980)
Hybridoma produces protective antibodies directed against sporozooite
stages of malaria. Science 207, 71
Yoshikura H., Hirokawa J.T., Yamada M. (1967) Syncronised cell division
induced by medium changes. Exp. Cell. Res. 48, 226-228
Zielke H.R., Ozand P.T., Tildon J.T., Seudalian D.A., Comblath N. (1976)
Reciprocal regulation of glucose and glutamine utilisation by cultured
human diploid fibroblasts. J. Cell. Physiol. 95, 41-48
Zielke H.R., Zeilke C.L., Ozand P.T. (1984) Glutamine a major energy source
for cultured mammalian cells. Fed. Proc 43, 121
Zimber A., Visek W.J. (1969) An ammonia generating system and 3H
thymidine incorporation in vivo and in vitro. (Abstract) Fed. Proc.28,
362
Zimber A., Topping D.C. (1970) Metabolic effects of ammonia in ascites
tumour cells and mouse spleen cells in vitro Fed. Proc. 19, 428
Zimmerman U., Vienken J., (1982) Electric field induced cell to cell fusion.
J. Membrane Biol. 67, 165-182
APPENDIX 5
The following Appendix contains the original "raw" data from the graphs
included in the thesis.
flCfU&£ 2'Aq 13' ic 2' fM 3 -4 h ' l-4£b i l btf 63 63 /6 $ U To 1!
r --- —
Time ViMte cc-ab 7t> rat. cetts dm weiGfltr Mb 7/7% CO/3C3Chi) plf/ml Xfof/Ml isettftirm/*~i vree,***S9/nt E&g.O*i+/~*11'"H C<yrsc- e&zx>#-//- oonfc•V) A? •3BÜWV-*!')
0 t - w & -a 2'0Ç) 0-24 O-oio 0.0034 7-62 o-sl H i 0/77 /2 0 0086
i t s 041 3-03 0 4 3 0 0 % 6 0034 4-s'6 0-98 l-w 0/40 /-66 0 4%
2 f o 3-61 0-51 & 74 0 - f 2r 0 0 % Û.O034 /2-04 Oil /•/? 0-080 2-00 O/43
20-0 4V? esq 4 4 2 0-63 0-00 0.0039 /9-74 (11 0f( 0-0&4 2-96 0-/9?
i-2-S l-2b l-oi, 768 1*01 0 # 0.0073 22-82 (S3 / • J 2. 0-09f 2-oo O'/fa
4-P0 t - o i I # ?-y/ |-ll O-ifO 0 0074 26-81 1-EO 0 '6 4 O046 2-76 0(47
5 f 0 i -w 0-OO77 2/-75" I-4G 0-56 OOfO 2-76 0-/48
66 f s - n o-u tO-fb 144 075b ooo?4 3/-3i' 2-(0 0-00 0 2/7 0-228
12-0 H t 048 f-03 1-25 0-/66 0.0081 44-09 1-45 0 -OÎ o-oot &31 O
78-0 i - 3 ? O f8 9-22. 1-31 O l H 0. OO60 44-84 246 0 - ^ 0-0/1 2-32 0 -^
fo-o 0 80 0-/I lb-to 3-34 o m 0.0068 (tf-Ol 3-02. ,0-00 0 3-40 0-244
BsyssEÊCeE 7 3
3-fob 7 4-
3 -6c . 7 S
3 6 c 75"
flME(hours)
V'flSLE CEUsfïlt .l) Total Cells Viability MabtlTEE VOLUMETRICGLuTAMWE
V O L U rt en?/C' A n r l o w f i t
Co u n t(VlO$/»V«j)
EMo« V-( X l 0 s / m l )
CO UN TC * 1 0 * f a d )
E A n o R + t -
(xios/m)) (%) CONC.. f/Vd/ml)(
weo«v-/uglml) » ERRO «■*'/-
(*>M)C è n e Erroè
0 2 /3 0 3 0 2*26 0*32 94 1 4 2 — 2 2 4 O 1-00 0*0?
1 9 3 0 2 6 2 06 0 24 84 3 6 4 0-80 1-80 0*13 1*15 0 08
22-5 4 22 0-59 4 '2 S 0 54 44 1*64 — 2 2 7 0-16 1-41 013
38-5 7 2 0 0 62 7-25 0*6t 1 47 3*12 0*21 2 6 6 0*19 1*32 0*04
45-S 8*72 0*25 4 02 0*26 47 S* 15 0*68 1-80 0*13 2*19 0*15
6 2 5 7*20 1-00 4 / 9 1*28 78 — —» / — —
69'5 3 9 8 0*54 7 5 6 1*05 51 7-Si 0 2 4 2 4 2 0*18 2 5 2 019
37-5 1*47 0 23 7'40 116 2 0 6*57 0*50 0 — 2 6 + 019
AGy<?E 3-8 a -------------------------------------------------------------------------------------------------------------------------------------------------------—4PAûE 7? ----------------------- ------------------------------------------------ ------ ---------------- *tTimeChrs)
ASPARTATE 6LUTAMATE ASPARAûKVB SERINE. HISTIOiNE ARCriNiruECONC/vgrml Error v-uq/Ml CONC ERRoRV- /vqlivil /uflrMl CPNC ERROC*/- Ai /rwl a»<M
CONC ERRoe.*ÀCONC ERROR •+/- A311 /qiml CONC ercooa A9ImI vV3'Ml
0 25*4 183 72-3 3*54 394 2*44 27*5 1*37 *2-4 0*42 187*8 11*46
14*5 30*0 216 64*5 3-lb 36-4 2 26 24*6 1*22 16*8 0*64 167-5 10*21
225 32 4 2-37 60 8 299 36-0 2-23 2o-5 1*02 12-8 0 -S3 155*8 95»
38*5 32 + 2-33 69-8 3*42 42*0 2-6o 24*6 122 18*6 0 7? Î76-7 ro-76
46*5 23 0 147 55*3 241 24-7 1*84 19 *5 0*4? ll*7 0*48 46 5 994
6 2 5 2*0 2-02 50*4 2-47 310 192 20-1 100 10*9 0 45 42*9 8*71
G9'f 32.0 2*30 64 2 3*15 22-20 13* 14 2 0*95 10*3 0*42 102*3 6*24
97*5 24-4 2-16 54*6 2*66 19-9 1*23 219 3*08 10*3 0*42 75-7 2*99
Fiûuce 3 8 ^ 3 - 8 b 3 * 8 c t S »8o 3 8 * 3*8*»mcfE 7 8 77 7 7 7 7 7 ?
TIMÉ THREONINE alanine PRO U N E TV ROSINE V A LIN e METMIONiiveChfs) tone BAftOft +/- ju.qlmi CONC >0<ji Irwl >ULq)Ml CONCyUqlfvvl xy//Hl CONCvveiMi e<*o/c +/- CONC
jitol**! E#mc+/- CONCX*|/#W|ecDV
0 22-3 /•4S 1 8 -4 0 3 2 2 2 0 0 * 6 6 1 4 9 0 6 9 19-6 1*06 8 5 0*34
1 4 - ^ Z02 13) 2 8 * 5 M l 22-9 0 6 9 11*5 O SS 19*4 Ml 7*4 0*30
R2-5 H Z 0 9 3 6 9 6 2 * 7 1 1 2 3 0 0*69 7*9 0*36 8 6 0 * 4 9 4*7 0 19
38-5 ISO 1 1 7 40-7 IS 6 2 5 * 0 0 * 7 5 II q 0*55 14*5 0 * 3 7*2 O 29
44»S' 13S 0 3 8 6 6 3 2 59 196 0 * 5 9 6*5 0*3o 3 7 0 *50 , 8 3 0*21
6>2S 9 * 4 0 0 6 ( 82-3 3-21 210 0 63 5*0 0*23 16*6 0 4 5 3*8 045
M S 7 0 0*4fe 9 2 3 3-62 2 3 0 0 69 4 ' 5 0*2) 6 3 0*36 2 3 0*09
7 * 0-46 4 i-S* 3 5 7 22 5 0*68 2 6 04% 5*7 0 * 3 3 2 0 0 * 0 8
F iG u se 3 ' S q 3*89 3 8 4 3 8 9 3 . 8 4 3 8 a 3 8 5
PAGE 7 7 7 7 7 9 7*7 , 7 > 7 7 7 8II CYSTEi n E i s o u e v o N e UEv Ca n G AlENHAtANiNE LYSINE ClTRUUUNE 06NI THiNE
EAtoC*y-«UI'At 1OBNC E^eoev- CONC.
yU /nMeceoe.+/.S J tfln A
CO/MC/x |/H ©«0 <£■♦/-
A c i fn * jCOMOvU /iwf
& U O C Co (VC & U 0 it% ^ \
CONC e w e .
O 31*6 4 3 * 0 2 8 8 4 3 * 6 2*66 11*4 o*9t> 2 5 * 8 2 61 1*4* * 22*1
f 4 5 2 6 * 9 ¥ • 3 9 0 2 * 4 8 3 6 8 2 -2 5 10*3 6 * ? o 23*9 2-41 0*9 * 2 4 - 3 >fc
2 2 * 5 2 6 2 22*0 1*42 2 o*3 1 * 2 4 8*3 0-3% 11*0 l*H 0*9 3 2 ) t f -
3 8 8 2 9 1 3 3 4 2 * 2 4 3 3 9 2 * 0 6 8 * 9 0 -5 3 Z D S 2* OS 1 2 * 3 4 ?
4 6 * 5 2 S * 5 T ” 21*4 1*43 1 9 5 1*19 4 * 6 0*28 14*9 * 5 1 0 * 7 3 9 6*
6 2 5 2 3 9 * 1 9 6 1*18 14*8 0 * 9 o 3*0 0*18 10*0 1*01 o*9 * 4 3 * 9
6 9 5 2 1 0 * 10*9 0 * 9 2 8*9 0 5 3 2 * 3 0 / 4 5 2 0 * 5 3 M * 4 5 2
9 7 * 5 12*5 7 9 0 * S 2 6 7 0*41 2 0 0*1% 3*5 0 * 3 5 0 * 6 * 29*8
ERROR NOT KNOWN - ASSAN PERPOR.wej> BV SW tT H K U N ë GeECMAAl PHAlCWACeo-pCAtS.
FUkvee 34* 346 iPtokS 80 81 iTirtEIkn)
6LUC0SE LACTATECfewcl e-Knon+f-MqldL Cor*cfa”) ERROfi?/-M
q *68 12 13-28 2-39 0-IS
/4*S IS4-U I2'l9 S S 3 0-39
)42lSl \\2<o 4-ss 0 29
SB'S 103-8? 8-21 — » —
45S qi?e 7^ , *89 o-st 1fees 39 6% 248 10-20 o*fcs
Ws- 3$?4s 2-8Z 1041 0-69
1
FlQUfiE 3'IOq - - 3 106séauM-Rce*
'm & L E 3-1s&tu/yi-Ftzecr ' TABLE 3-1
PAQE s s / - / 8 6 3 9
TlrlE(hrs)
VIABLE cen s TOTAL c e u s VlASIUTN° b
rlob TITRE ASPAETATE Gt-LufArt/ireEftietX"*/- K IO$/m | CÔWCXIoS/lvlJ eeeoe-»/- sin* E K Z O d U -XloS"/i»i| 'C O N C-vg /nM BR.ieoK.t7-/LWi»vM O O b J C A# /fW eieioztV,«)/Avf
i II? OHO 1610 0-230 74 8-81 O-S'o 1468 1-42 2516 1-23
1-20 O .H i 148 0-251 6 9 10-04 ISO 18-80 I'ZS 2S4ST 125“
fgG 0-2W 2-SO 0-3S81 7 4 12-30 3 88 I8-OI 1-30 2939 136
41 3 (8 O-AsC 4-38 0-626 73 18-59 O 21-23 # 5 3 3^-89 I'Ti
4 9 3-84 0-54% 4-98 0 .1 (2 79 18-59 O 2o-3S * 49 . 3S3D 1*93
S 4 4*92. 0.104 6 3 9 0.1 H 7? 29 94 O IE-8S 1 21 3M3 1-S3
foS I'Ob f-o/4 9 18 1-213 79 5 1 5 6 4 4 2 1349 A 9 9 2 9 ? i l-3t
?*2| 1-030 9-Î8 131% 74 5 2 86 S ID 20 4* I-SI 43-81 a ir
73 6*41 OW q-is |-30% 7o S 6 4 9 18-99 — ——
99 4-oz 0- s # 8-2S l l %0 49 2 6 -4 0 2-60 —— — ——
«1 2 3 d 0-321 761 ( 0%8 3o 2S-I8 2-2S 22-28 169 49*5«J 2*38
89 1 6 4 0-23f 9 59 1*369 19 31-44 3 6 6 — — — -
AGI/SE S€ieuAI-Reé* TA6L6 3-1
S-FTA6L6r 31
SFTA6tfc 3 /
S FTA6L6 3-1
SFTAStfc 3-1
SF 'TASUZI
pfiQe 29 99 89 8 9 84 89Timef ^ J
SE^IMÊ GrUNOtNE A&CriNlWE ÎH2£0fU/N6 ALAM//V6 P«DUN£Co m c Ezeoei/- CONC>y/AM ERCoe- A
> c g M ,CONC>U,q W Eaeoc+/- CXTNC-yvy/M ■e*fiog.+t- Cer/C■*jglr*4 ERjeoet/- CONC
yU ;rwt&UL01L.
1 3 2 4 4 14% 2°9-60 12-99 146% 0 4 5 24*13 0 -94 2 2 #I
0-68
3? 3 6 5 4 1 9.6 19-85 0-93 190-54 11-62 15-56 IOI 28-39 I I I 2b 01 0 60
21 34*19 |-9o 16 4% O-fcA* 187-08 II 4» 14 42- 0 94 38-01 14» 1951 OS9
4-1 39 ST 1-89 18-34 0 95 213-06 13-00 14 58 o 9 5 5 5 5 6 2-19 212* 0 6 4
49. 36 0% 1?9- , (6 99 0 * 9 0 199-20 12-15 11 91 O 99 6 0 4 4 2-36 ! 20-14 0-60
54- 29-39 13? 12-85 O S 3 154-19 9-4» 8-9t> 0*5? 56-25 2 19 IS?? 0 * f ?
feS 20-41 l o i 11-59 0 4 9 12488 9-62 3-81 0 25 S0-6l 19? 11-93 0 3 5
fcl 30-92- 153 14*99 0 61 195-94 II 9 4 4 96 0-31 78-62 3*0? 19-03 0-5?
81 31-90 15» /6-4b 0 69 200-94 12-26 S 4 9 0-58 86-91 3 3 8 21-91 0-65
RQuee sekom- Ffiee 171616 3*1
seeuto-peee •7A8tâ 3-1
sûewn-fleesTkêieZ'i
seeuM-fleeerA6£63-l
11 'free*3-|
wç7X6tC.3-|
PA4G 89 89 89 89 e? 89
TlflflG TY&OSii*€ VAUN6 niETHlOMiN6 isouevONfc UÊUQN6 PH ÊNVLAt/WiwS.Civs) t»NCA«lM GUtoa+b CONCyVSZvwl eaece+/-yvgM <OMC
yV /<wf jjq ImA CtiNCyVS W taetofc*/-yVqlM CONC ersaoe+A•Vq /wv#CONCA q irvJ
EM-Of) -V4 /N)
1 2008 042 19*24 1-04 Î659 0 66 3-oi 40-99 24? this 048
19 18-14 093 1562 0»4 1491 0 59 38-00 2-5S* 39*39 2 10 74*66 081
24 0*9? 13-98 099 14*01 0-56 34-59 2-32 3* 31 f 91 7356 0-23
41 16-?^ 0*9? 12-99 0*?3 14*2? 0*59 3S-90 2-41 33*66 2 05 7564 045
4 ? 13*92 0 64 b*41 0 60 12*69 o-si 3204 2-15 29-87 1-82 4-32 0-8?
S4- 10-90 0*49 G *99 0*39 935 0*3? 23-64. 150 21-44 1*31 71» 093
65 6 12 0*29 4*51 0*26 5*90 0-24 15-61 1 05 13-63 083 10*02 061
64 141 0-06 3*94 022 8*52 0*34 22-1» 1-53 20-20 1*23 11-40 0-9D
g ! 8*11 0 41 4-5? 026 934 0-3? 21-38 1*43 19*49 1-07 70-21 062
AGtvee SaCUM-ftCÊS. 7A6te 3 1
Sfc^tm -AetSr TA&LÇ 3 1 :
PAGrB B9 V} - f•nme(hrs)
"mPToPHAhl LNSiwCowe- ewoK-A»4/«v1 C f i N L
AXjlrrfEXAiO&.vOfllwvl
1 S-1S 0-4$r SS60 2 4 6V
# 5-5S 0*44 3)96 2 4 9
29 $ r ? 0 4 6 269t> 2*22 | -
44 8-24 06* 2>ii z %
4 9 8 0 3 0 6? , 22^T if*29 ' 1
54- 6 SI) 0 ^ 4 4*43 4*24
b £ S4S 0 4 4 944 06 6
69 «•22 0 6% 11-30 0*44
8i 8 0 6 0 69 1249 I-04
FWKe 4-2b 4 loi 4 2 4 4-2*{ 4 - 4 4"2ePAQ6 19 1D| IOI IOI 102 ;oi1
3 VIABLE COUS XIOs /nvl
Mab Tfree f/yfml)
609 PZOOocTlbit (%)
O9 UflUSATlOM(V
a ç AlAao«o 0.0. T /%)
/ 0 1 2 , O .o^ l- s s . 0 - 9 2-21 0 104/B 0 4 < I' 32 0-OI 0 4 0 02 0 71-2
21 0 % 1 3 ^ O-oi- 0 -51 0 -0 2 0 63-2
42 t - lo 3-02 0 01 0 5 ? 0 - 0 1 0 4 4 1
49 i t s 313 4 4 4 2-23 2 -0 ? ; o-oi 4o-<
u 2 ^ 3 313 2-6% 2-20 1-22 0-ol 4 0 0
1 4 3 1 2 l - 2(> 1-64 I-SÔ 1-09 0-03 40-1
ID 9S 3 s - n O i l I-10 0-?4- 0-13 4 0 0
IS %-%r 3-0 | 131 O i l û - % 0-02 400
/It g-30 1 3 - 4 3 O 0 4 0 - 6 o 0 - 0 ? 0 4 2 2
/23 4-69 i f - 1? 1 2 ? 1-40 , 0-11 00 1 K-6
' * f o m m U o f R . ç f f o F u & s
W l £ R<P T tn iE <eo r i m e (L0 r - -
1 0 .0 1 1 o - 6 o 1 l -u
1 1 0 0 % 6> o - 0 4 . G 0 0 11
2 3 0 - 0 2 I ? i - 6,& 2 1 1 0 3
2 1 2 , 3 1 2 0 1 - I l 2 0 3 - 1 9 .
4 - 1 | - ( D 2 3 i - n Zo 1 - 2 4
4 1*
Û- 8 S 2 1 H o 3 4 1 ( 0 ,
5 4 Ô ' t o H 1 - 3 4 4 5 0 -8 6 , *
GS 0 -4 L 1 - 2 9 3 0 O-IT-
G? 0 - 4 2 4 2 0 1 % * 54 1-3? *6)1 0 -4 ? 4 4 o - l i 6)0 0 - 2 5
? 3 0 - 5 4 4 % 0 - 8 2 6 % (0 4 - 4
9 ? 0 0 1 * 5 4. Û - 4 6 . ^ !
S I 0 5 8 6 6 0 - T 8
8 1 0 - 8 0 ? Z 0 - 1 -6,
9 9 0 - 0 3
-
!
*• U M iT S o f n e Peftn m ffob fHooucxioH P a ts