Microcarrier

Cell Culture
Scale-Up Methods

18-1020-61
Edition AA
 Handbooks
from Amersham Pharmacia Biotech

Antibody Purification
Handbook
18-1037-46

The Recombinant Protein Handbook

Protein Amplification and Simple Purification
18-1142-75

Protein Purification
Handbook
18-1132-29

Ion Exchange Chromatography Reversed Phase Chromatography

Principles and Methods Principles and Methods
18-1114-21 18-1134-16

Affinity Chromatography Expanded Bed Adsorption

Principles and Methods Principles and Methods
18-1022-29 18-1124-26

Hydrophobic Interaction Chromatography Chromatofocusing

Principles and Methods with Polybuffer and PBE
18-1020-90 50-01-022PB

Gel Filtration Microcarrier cell culture

Principles and Methods Principles and Methods
18-1022-18 18-1140-62
Microcarrier Cell Culture
Scale-Up Methods

1
 Contents

Chapter 1
Microcarrier cell culture ................................................... 3
Introduction ..................................................................................... 3
Applications ..................................................................................... 3
Cell attachment ................................................................................ 4
Requirements for an optimum microcarrier .......................................... 6
Properties and characteristics ............................................................ 6
Instructions for use ........................................................................... 8
Preparation ........................................................................................................... 8
Inoculum .............................................................................................................. 8
Culture Procedure .................................................................................................. 8
Monitoring cell growth ............................................................................................ 8
Harvesting cells ..................................................................................................... 9
Scaling up by serial subcultivation .......................................................................... 9
Equipment recommendations ................................................................................ 10
Industrial applications .......................................................................................... 11

Trouble Shooting: ............................................................................ 12

Chapter 2
An Analysis of Alternatives and Process Development ........ 15
Introduction ................................................................................... 15
Process overview ............................................................................. 16
Provisions ........................................................................................................... 16
Process themes to be evaluated ............................................................................ 17
Principal process flow .......................................................................................... 17
Equipment used .................................................................................................. 17

Process outline and suggested experiments ....................................... 18

Process 1a .......................................................................................................... 18
Process 1b .......................................................................................................... 22
Process 2a .......................................................................................................... 24
Process 2b .......................................................................................................... 27

References .................................................................... 28

2
Chapter 1
Microcarrier cell culture

Introduction
Animal cell culture, vital to the study of structure, function, and differentiation, also provides
important biologicals for the pharmaceutical industry, including viral vaccines, enzymes, hormones
and antibodies. Microcarrier cell culture technology means that anchorage dependent animal cells are
grown on the surface of small (approximately 150 micron diameter) spheres which are maintained
in stirred suspension cultures. This technology is replacing conventional monolayer cell culture
methods as the extremely high surface area to volume ratio afforded by microcarriers is also used for:
• Routine cell culture
• New research applications
• High density culture
• Production culture volumes greater than 1000 litres
Some of the advantages of the microcarrier system are:
• Efficient monitoring and culture control
• Reduced costs and contamination

Applications
Cytodex™ enables all anchorage dependent animal cells to grow in suspension cultures, and is used
in both batch and perfusion culture systems, as well as to increase the surface area of traditional
monolayer cultures. In stirred suspension cultures, cells grow in a homogeneous environment where
the culture parameters are easily monitored and controlled. Cultures may be sampled, at will, to
examine cell morphology and to determine cell viability. Microcarrier techniques are therefore the
natural choice where cells are used for the production of biologicals. At the research level, applications
with Cytodex include:
• Studies of cell structure, function and differentiation
• Enzyme free subcultivation
• Implantation studies
• Harvesting of mitotic cells
• Light and electron microscopy
• Transportation and storage
Industrial scale applications utilizing Cytodex include the production of cells, cell products and viral
vaccines (see Industrial applications, Page 11).

3
Cell attachment
In order to culture anchorage dependent cells in vitro it is essential to know how the cells interact
with their environment. The cell membrane is surrounded by a "coat" called the glycocalyx
(Figure 1). The glycocalyx consist of glycolipids, glycoproteins and transmembrane proteoglycans.
A large number of these molecules contain sugars with a negative charge. So the net charge on the
glycocalyx is negative. That means that if a culture surface is positively charged there will be an
electrostatic attraction between the cell and the culture surface. It has been shown that the attachment
to gelatin microcarriers compared to charged carriers, is slower because the surface charge is
not optimal (1).

transmembrane adsorbed transmembrane

glycoprotein glycoprotein proteoglycan

= sugar residue

cell coat
(glycocalyx)
Fig. 1. Diagram of the cell coat (glycocalyx) which
glycolipid
is made up of the oligosaccharide side chains of
glycolipids and integral membrane glycoproteins,
lipid and polysaccharide chains on integral
bilayer proteoglycans. In addition, adsorbed glycoproteins
and absorbed proteoglycans (not shown) contribute
to the glycocalyx in some cells.
CYTOPLASM
© Garland Publishing Inc.

Among the glycoproteins in the membrane, there is fibronectin binding

a family of heterodimeric glycoprotein receptors H2N
called integrins (Figure 2). These integrins help bind NH2

cells to mainly RDG sequences in extracellular

matrix (ECM) proteins. The integrin receptors
require divalent ions such as Ca2+, Mg2+, Mn2+ for
binding to their ligands (2). This is the reason why A chain B chain

EDTA and Ca2+, Mg2+ free buffers are used when

harvesting cells in culture. The internal part of the S
integrin receptor, interacts with the cytoskeleton, S
membrane

which makes it possible for the cell to spread out

plasma

after attachment. The membrane integral

proteoglycan molecules also bind to extracellular
cytosol

matrix components. (Recently it has been shown HOOC COOH

that membrane bound proteoglycans bind to talin binding

growth factors and thereby modulate their activity. 10 µm

Syndecan, a heparan sulfate proteoglycan binds to

Fig. 2. The subunit structure of a cell-surface
basic FGF and Decorin and Betaglycan binds TGF- fibronectin receptor. By binding to fibronectin
b (3)). There are also unrelated transmembrane outside the cell and to the cytoskeleton (via the
attachment protein talin) inside the cell, the protein
glycoproteins that bind to collagen. serves as a transmembrane linker. The fibronectin
receptor belongs to a large superfamily of homologous
matrix receptors called integrins, most of which
recognize RGD sequences in the extracellular
proteins they bind.

© Garland Publishing Inc.

4
The ECM consist mainly of fibrous proteins embedded in a polysaccharide gel, which are secreted
locally. The main ECM components are collagen, elastin, fibronectin, laminin, and the
glycosaminoglycans. Glycosaminoglycans consists of long unbranched polysaccharide chains
composed of repeating disaccharide units. A large number of the sugars are sulfated or contain
carboxyl groups. This gives the GAG's high negative charges. The fact that the cells in vivo, are
surrounded by hydrated sugar gels and fibrous proteins is interesting when choosing materials for
microcarriers. The fact that cells also interact with collagen have been utilized when designing
Cytodex 3.
The attachment process in vitro is summarized in Figure 3.

adsorption contact attachment spreading

Ca2+
Mg 2+ 2+
Mn
cell
cell cell cell
– Fn
MHS –
MHS –
– Fn – Fn

+ + + + + + + + +
substrate substrate substrate substrate

Fig. 3. Simplified outline of steps involved in adhesion of animal cells to culture surfaces. The whole involves divalent cations and
glycoproteins (mainly fibronectin (Fn) and vitronectin) adsorbed to the culture surface. Under culture conditions the attachment
glycoprotein originates from the serum supplement in the medium. MHS is synthesized by the cells. With anchorage-dependent
cells proliferation occurs only after the spreading step. MHS – multivalent heparin sulphate.

5
Requirements for an optimum microcarrier

In order for a microcarrier to be suitable for animal cell culture at all scales, it must
fulfil certain basic criteria.
• Surface properties must be such that the cells can adhere with a degree of
spreading which permits proliferation. For homogeneous growth of cells the
surface of the microcarriers must have an even continuous contour. The surfaces
of all microcarriers in the culture should have consistent properties.
• Density of the microcarriers should be slightly greater than that of the surrounding
medium, thus facilitating easy separation of cells and the medium. The density
should also be sufficiently low to allow complete suspension of the microcarriers
with only gentle stirring. Under standard culture conditions the optimum density
for microcarriers is 1.030–1.045 g/ml.
• Size distribution should be narrow so that even suspension of all microcarriers is
achieved and that confluence is reached at approximately the same stage on
each microcarrier. Best growth of cells occurs when microcarriers have a size
distribution which lies within the limits of diameter in culture of 100–230 µm.
• Optical properties should be such that routine observation of cells on microcarriers
can be achieved using standard microscopy techniques. The microcarriers should
also permit use of routine cytology procedures.
• Non-toxic microcarriers are required not only for survival and growth of the
cells but also when cell culture products are used for veterinary or clinical
purposes.
• Non-rigid microcarrier matrices are required for good growth when the culture
is stirred. Collisions between microcarriers occur in stirred cultures and a
compressable matrix reduces the risk of damage to the microcarriers and the cells.
• Autoclavable for easier large scale handling.
• Quality of the microcarriers should be of the standard required for current Good
Manufacturing Practices (GMP).
• Large lot sizes (>100 kg) be available for industrial customers.
• A regulatory support file should preferably exist for the product.

Properties and characteristics

Two types of Cytodex are available to support the growth of all anchorage dependent animal cells
for a multitude of applications, and the microcarriers are designed to meet the special requirements
for this technology. The optimized bead size and density support maximum cell growth rate and cell
yield. A biologically inert matrix provides a stable - but non ridgid - substrate for stirred cell

6
cultures (80% of the bead volume is liquid i.e. the beads are small gel particles), from which cells
are easily harvested. Cytodex is transparent, allowing microscopic examination of attached cells.
The physical characteristics of Cytodex are shown in Table 1.

Table 1. Physical characteristics.

Cytodex 1 Cytodex 3
Density* (g/ml) 1.03 1.04
Size* d 50 (µm) 190 175
d5–95 (µm) 147–248 141–211
Approx. area* (cm2/g dry weight) 4 400 2 700
Approx. no micro-carriers/g
dry weight 4.3 x 106 3.0 x 106
Swelling factor* (ml/g dry weight) 20 15
*in 0.9% NaCl.

Size is based on diameter at 50% of the volume of a sample of

microcarriers (d50), or the range between the diameter at 5% and 95%
of the volume of a sample of microcarriers (d5-95). Thus size is
calculated from cumulative volume distributions.
Fig. 4. Scanning electron micrograph of BS-C1 cells
grown on Cytodex 3. The cells were treated with
trypsin for 3 minutes prior to preparation for SEM.

Cytodex 1, formed by substituting a cross-linked dextran matrix with positively charged DEAE*
(N,N-dimethylaminoethyl) groups distributed throughout the matrix.
Cytodex 3, formed by chemically coupling a thin layer of denatured collagen (Mr 60.000–200.000,
pig skin type 1) to the cross linked dextran matrix, is the microcarrier of choice for cells known to
be difficult to culture in vitro and particularly for cells with an epithelial-like morphology. This
collagen surface layer is susceptible to digestion by a variety of proteolytic enzymes and provides
unique opportunities for harvesting cells from the microcarriers while maintaining maximum cell
viability and membrane integrity. These issues may be critical in obtaining successful serial
subcultivation protocols required for scaling up culture volumes.
Figure 5 is a schematic representation of the two types of Cytodex. Pharmacia produces Cytodex in
accordance with current Good Manufacturing Practice (cGMP) and every batch must conform to
stringent specifications. Quality control tests are performed to satisfy both physiochemical and
functional (i.e. cell growth) properties (Figure 6). A certificate of Analysis is available on request.

Cytodex 1
+ ++
charges +++ ++++ CH2CH3
throughout + + + + + cross-linked dextran CH2CH2 N .HCL
matrix + + + ++ CH2CH3
++ +

Cytodex 3
OH
collagen layer
coupled to cross-linked dextran O CH2 CH CH2 NH (e-LYS.COLLAGEN)
surface

Fig. 6. Each batch of Cytodex is fully

quality controlled, including
Fig. 5. Schematic representation of the two types of Cytodex. monitoring of cell growth.

7
Instructions for use

Preparation
Cytodex, supplied as a dry powder, must be hydrated and sterilized before use. Autoclaving is the
simplest method for sterilization; Cytodex is extremely stable, and if necessary can be autoclaved
repeatedly (at least 5 cycles of 15 min, 115 °C, 15 psi) or extensively (12 h, 130 °C, 27 psi) without
affecting its performance. Prior to initiating a microcarrier culture, the pH and temperature of the
culture medium containing the prepared microcarriers should be adjusted for optimal initial
attachment of the cells to Cytodex. In order to achieve a good attachment the culture surface has to
be coated with fibronectin (15 ng/cm2 for BHK, 100 ng/cm2 for CHO) or vitronectin, which are
present in serum containing medium (10% v/v foetal calf serum contains 2–3 µg fibronectin/ml).
For some epithelial, endothelial cells it might be necessary to coat the culture surface with laminin.
If serum free medium is used these proteins must be added, unless the cell line secretes them itself.
In addition, the gas mixture in the culture vessel headspace should be equilibrated with the culture
medium.

Inoculum
The state of the inoculum is the most critical factor affecting microcarrier cultures. The cells should
be in logarithmic growth phase and in a good nutritional state, upon harvesting, and definitely not
in the resting G0 stage. This ascertains a short and efficient growth phase in the cultures. It is therefore
essential to optimize the harvesting procedure, and the timing for harvesting the inoculum culture,
to maximize viability and yield.
The cell inoculum should be introduced to the culture vessel only after equilibrium has been
established. Good initial cell attachment is essential in order to obtain maximum final cell yields
from microcarrier cultures. It is therefore of importance to experiment with the initial conditions.

Culture Procedure
The exact culture requirements i.e., medium, serum and supplements depend on the cell type.
The size of the final culture volume and the design of the culture vessel determine the optimal
culture procedure. Batch type microcarrier cultures usually contain up to 5 grams of Cytodex per
litre of culture volume. Fed batch and perfusion cultures can sustain higher cell densities (4) and
Cytodex concentrations may be greater than 20 grams per litre.

Monitoring cell growth

Representative samples of the microcarrier culture may be withdrawn at any time during the culture
period. The cells on the microcarriers can then be visualized directly by phase contrast microscopy
or by light microscopy after the cells have been stained with, for example haematoxylin. The
simplest and most accurate method for determining cell number is to use nucleus extrusion method
where cells on microcarriers are lysed in 0.1 M citric acid. The released nuclei are stained in crystal
violet and then counted in a haemacytometer (5). Another alternative is to measure intracellular
metabolites released upon lysis (enzymes, ATP, for example).

8
Harvesting cells
The most common methods for harvesting cells
from Cytodex involve standard procedures
employing proteolytic enzymes such as trypsin and
collagenase. Pre-washing the confluent microcarriers
with a solution of EDTA-PBS* (Ethylenediamine
tetraacetic acid-phosphate buffered saline), prior
to trypsinization, may improve the harvested yield
of strongly adherent cells, by binding divalent ions
and thereby destabilizing the binding strength
between integrins and ECM proteins (see above).
Especially when serum containing medium is used,
it is necessary to efficiently wash away the protease
inhibitors which are present in the serum. Other-
wise the enzyme activity goes down drastically!
Special attention should be paid to optimizing and Fig. 7. The modified Vibromixer vessel used by van Wezel
standardizing the harvesting procedure with respect et al., for concentrating, washing and trypsinizing the cells,
separating the cells from the used microcarriers.
to both the activity of and the time which the cells
are exposed to, the enzyme solution (6, 7, 8).
Other procedures include the use of chelating
agents, exposure to hypotonic conditions, cold
conditions, sonication or alteration of the surface
tension of the culture medium (Figure 7).

Scaling up by serial subcultivation

Serial subcultivation of cells on microcarriers is the
most cost-effective means for establishing animal
cell culture production facilities, achieved using a
series of culture vessels of increasing volume and
capacity. The cells cultured in one vessel are
subsequently used to inoculate the following larger
vessel. This approach to scale up eliminates large
numbers of culture vessels, such as roller bottles or
multi-trays, otherwise required to obtain sufficient
cells for inoculating the larger production scale
fermenters. The protocol for serial subcultivation
of cells from one culture of Cytodex to the next,
should be firmly established at the small scale prior
to scaling up to production culture volumes. Some
cell types are capable of undergoing bead-to-bead
migration with Cytodex, i.e. subcultivation can be
performed without a harvesting step that employs
proteolytic enzymes. In this case, scale up may be Fig. 8. Europe’s latest facilities for the production of polio
virus vaccine using Cytodex microcarriers. Courtesy of
achieved simply by increasing the culture volume Rijksinstituut voor Volksgezondheid en Milieuhygiëne
and adding more Cytodex. (RIVM), The Netherlands.

9
Suggestions to increase or facilitate the migration includes the stirring speed, which can be increased
(cells detach during cell division and re attatches to fresh beads), the medium composition can be
manipulated, increase phosphate and decrease Ca2+, Mg2+, Mn2+ concentrations, less fibronectin can
be used for coating the culture surface, RGD peptides can be added to decrease the cellular binding
to the surface or motility factors can be added to the culture medium. There are now three different
migration stimulating factors described, scatter factor, migration stimulating factor and autocrine
motility factor (9). Scatter factor has an effect on epithelial cells, both SMF and AMF act mainly on
fibroblasts.
Cells which are unable to undergo bead-to-bead migration must first be harvested from the confluent
culture, as outlined above for subsequent use as inocula. Routine procedures for harvesting these
cells as a single cell suspension - as opposed to clumps of cells - with maximum viability, retained
membrane integrity and high attachment efficiency, will greatly facilitate the successful inoculation
of the following culture and thereby the entire scaling up process.

Equipment recommendations
Microcarrier cultures can be contained in virtually any type of cell culture vessel which can be kept
sterile. However, maximum cell yield and productivity are obtained from general purpose
microcarrirer cultures using equipment with efficient mixing systems which provide a homogeneous
culture environment, an even suspension of microcarriers and adequate supply of oxygen. High
shear forces must be avoided due to the stress sensitivity of animal cells. For routine small scale
microcarrier cultures (up to a few litres working volume), some commercial suppliers have modified
traditional glass spinner vessels containing magnetically driven stirring rod or impeller, specifically
for improved use with microcarriers.
N.B. all glassware with which Cytodex comes into contact, should be siliconized before use.
Laboratory scale fermenters designed for animal cell cultures are also available from various
commercial suppliers. These fermenters, fitted with probes and control systems for pH, temperature
and dO2 values, facilitate the optimization of these most essential culture parameters at small scale,
for future reference at production scale. Also the availability of perfusion systems allows for
simplified long-term productivity studies. The process control and automation which can be
supplied with such systems, lead to economical and reproducible production facilities, many of
which are now in routine operation for culture volumes from tens to thousands of litres.

Fig. 9. Typical process development facility at

Amersham Pharmacia Biotech, Uppsala, Sweden.

10
Industrial applications
Industrial scale cell culture, incorporating Cytodex, has proven to be reliable and cost effective for
the manufacture of both human and animal health care products. Significant savings include greater
than 50% reduction in culture medium and serum costs, reduced labour and decreased risks of
contamination. A unit culture fermentation system of 100 litres containing 4 grams of Cytodex per
litre can be handled by one operator and offers the equivalent surface area for cell growth as a facility
containing 3000 x 800 cm2 roller bottles or 400 x 6000 cm2 multi tray units. Rapid developments
in recombinant DNA technology for the expression of foreign proteins in animal cells are resulting
in increased use of transformed (and normal) mammalian cells such as CHO (especially), C127,
Vero, COS, CV-1 etc. due to more complete posttranslational modifications (10).
Microcarrier cell culture technology is the means by which sufficiently large numbers of anchorage
dependent animal cells can be cultured at production scale to meet the needs of the pharmaceutical
industry. Virus vaccines, interferons, plasminogen activators and urokinases, cytokines, hormones
such as animal and human growth hormones and a variety of factors such as platelet derived growth
factor (PDGF), epithelial growth factor (EGF), tumor necrosis factor (TNF), erythropoetin (EPO),
colony stimulating factor (CSF), interferons and others are produced from such microcarrier facilities.
Some of these products are already available on the market as human and animal diagnostics and
therapeutics. Many are undergoing evaluation and clinical trials and even more are in the develop-
mental process.

11
Trouble Shooting:
When working with stirred microcarrier cultures for the first time some difficulties may be encoun-
tered. The following points list the occasional areas of difficulty and the most likely solutions. These
points can also serve as a check list when culturing each new type of cell.

1. Medium turns acidic upon addition of microcarriers.

• Check that the microcarriers have been properly prepared and hydrated.

2. Medium is alkaline upon addition of microcarriers.

• Gas the culture vessel and equilibrate with 95% air, 5% CO2.

3. Loss of microcarriers on surface of culture vessel.

• Check that the culture vessel has been properly siliconized.

4. Poor attachment of cells and slow initial growth.

• Ensure that the culture vessel is non-toxic and well washed after siliconization.
• Dilution of culture by PBS remaining after sterilization, rinse microcarriers in growth medium.
• Modify initial culture conditions, increase length of static attachment period, reduce initial
culture volume or increase the size of the inoculum.
• Control condition of inoculum, and that it has been harvested at the optimum time of culture
with an optimized procedure.
• Eliminate vibration transmitted from the stirring unit.
• Change to a more enriched medium for the initial culture phase.
• Check quality of serum supplement.
• If serum free medium is used, an increase in attachment protein concentration might be
necessary (fibronectin, laminin).
• Check for contamination by mycoplasma.

5. Microcarriers with no cells attached.

• Modify initial culture conditions, increase length of static attachment period, reduce initial
culture volume.
• Improve circulation of the microcarriers, to keep the beads in suspension during stirring.
• Control the condition of the inoculum. Especially that it is a single cell suspension.
• Check that the inoculation density is correct (number of cells/bead) (8).

6. Aggregation of cells and microcarriers.

• Modify initial culture conditions, reduce time that the culture is allowed to remain static.
• Increase stirring speed during growth phase, improve circulation of microcarriers.
• Reduce concentration of serum supplement as the culture approaches confluence.
• Reduce concentration of Ca2+ and Mg2+ in the medium.
• Prevent collagen production by the cells by adding proline analogues to the culture medium (11).

7. Rounded morphology of cells and poor flattening during growth phase.

• Replenish medium.
• Check pH and osmolality of culture medium.
• Reduce concentration of antibiotics if low concentrations of serum are being used.
• Check for contamination of mycoplasma.

12
8. Rounding of cells when culture medium is changed.
• Check temperature, pH and osmolality of replenishment medium.
• Reduce serum concentration.

9. Cessation of growth during culture cycle.

• Replenish medium or change to a different medium.
• Check that pH is optimal for growth.
• Re-gas culture vessel or improve supply of gas.
• Reduce stirring speed.
• Check for contamination by mycoplasma.

10. Difficulties in controlling pH.

• Check that the buffer system is appropriate.
• Improve supply of gas to culture vessel, lower concentration of CO2 in headspace or increase
supply of oxygen.
• Improve the supply of glutamine, supplement the medium with biotin or use and alternative
carbon source, e.g. galactose.

11. Difficulties in maintaining confluent monolayers.

• Check that pH and osmolality are optimal.
• Reduce the concentration of serum supplement.
• Improve the schedule for medium replenishment.
• Reduce the concentration of antibiotics.
• When culturing cell lines that produce proteases, in a serum free medium it might be
necessary to add protease inhibitors to prevent the cells from detaching (CHO cells have been
shown to secrete proteases!) (12).

12. Broken microcarriers.

• Ensure that the dry microcarriers are handled carefully.
• Check the design of culture vessel/impeller and ensure that the bearing is not immersed in the
culture.

13. Difficulty in harvesting cells from microcarriers.

• Ensure that the carriers have been washed extensively together with mixing.
• Check that roughly the same amount of protease (U/cell) is used as when harvesting cells
from flasks.

14. Floatation of microcarriers in foam during sparging.

• Reduce the serum/protein concentration as much as possible.
• If possible use Pluronic F68 or other detergents (13).
• Add polymers to increase viscosity.
• Aerate via:
- Silicon tubing (Diesel, bubble free aeration) (14).
- A wire cage/spin filter (New Brunswick, Celligen) (15).
- Vibromixer/wire cage (Chemap, Chemcell) (16).
- An external/internal hollow fiber loop (Microgon) (17).

13
14
Chapter 2
An Analysis of Alternatives and Process Development

Introduction
This analysis identifies four alternative scale-up methods, presented in the matrix below:

Culture system
Batch Perfusion
Subcultivation method 2–3 g/l 15–20 g/l

Trypsinize Process 1a Process 2a

Migrate Process 1b Process 2b

The ultimate choice between the alternatives will always depend on the particular characteristics of
the cell line/end product theme under investigation. The critical experiments, part of the process
development work, that need to be performed are presented below, together with an estimate of the
time that must be spent. Looking at the end result of the process development; a protocol of
procedures and equipment that brings the process economically and efficiently from the Master Cell
bank to whatever production-scale needed, the four processes we discuss offer principally different
economies.
Most processes can be regarded as two step procedures where the first step is the generation of cell-
mass for production and the second step is the production phase per se. The linkage of product
formation to cell growth varies considerably from product to product, but nevertheless this formal-
ized two step view is useful in discussing the typical process flow in the scaled up routine produc-
tion. This, because the major process costs are related to labor in generating enough cells for the
final production step. Ideally this high initial cost is "written off" over a prolonged, continuous
product harvest cycle.
The table below shows the cell mass generation efficiency for the processes discussed, together with
the typical process times involved.

Cell mass generated per time unit.

Process time
Process no. Cells/h (hours)

1a 1 x 108 624
8
1b 2 x 10 528
8
2a 9 x 10 720

2b 1.6 x 109 624

15
Process 2b is clearly an attractive alternative, particularly if one also consider savings in investment
costs. This however with the provision that the particular cell substrate spontaneously migrates or
can be manipulated to do so. In reality, the rewards of the scale up alternatives offer must also be
balanced against the necessary process development time. The table below summarizes our estimates:

Process development time

Process no (working days)

1a 148*

1b 140

2a 206

2b 202

* Excludes time spent on generating the cell mass needed for the experiments.

The data are offered to indicate the magnitude of work involved, with the inherent assumptions that:
- all four alternatives are investigated as part of the experimental program.
- that the work starts from a situation where no prior knowledge is available.
- but where the cell line is selected, checked and entered the Master Cell Bank.
Calendar time that must be spent on the process development themes will of course depend on the
number of people available to do the work, and how the work task is subdivided. Particularly in an
industrial production environment this is a heavy drain of available resources. Guiding experience
and documentation to help shorten the time for R/D is therefore of extreme value. Fortunately, such
guiding experience is available for many important microcarrier processes. For certain other cell
lines of interest to todays Bioindustry, rapid progress is being made, in conjunction with an emerging
new microcarrier system technology. This documentation illustrates both the "new" and "old"
segments of the Cytodex technology, as seen from the process development end.

Process overview

Provisions
1. Cell line selected:
- High and stable productivity during 800 hours (five subcultivations)
- A suitable medium selected in small scale (this might have to be changed during scale up!)
2. Cell line tested for:
- Mycoplasma
- Fungi
- Bacteria
- Viral contaminants
3. A master cell bank generated

16
Process themes to be evaluated
1. Batch culture with 2–5 g/l of Cytodex. When the final cell mass is generated selected products
may be harvested via fed batch/perfusion under prolonged periods of time.
2. High cell density cultures with 15–20 g/l of Cytodex and with media perfusion of 1–2 volumes/day.
These two can be done either via:
a) Trypsinization or
b) Migration
Note! b) can be used provided the cell used do migrate or can be stimulated to migrate. In the
described process we have based our calculations on C127 cells. This particular cell line is an
extremely good migratory cell.

Culture system
Batch Perfusion
Subcultivation method 2–3 g/l 15–20 g/l

Trypsinize Process 1a Process 2a

Migrate Process 1b Process 2b

Principal process flow

The principal process flow is as follows:
N2 - T-flask - T-flasks - spinner - spinners - fermentor - fermentor
- Scale up within the same fermentor

Equipment used
The equipment used in the described processes:
- 150 cm2 T-flasks
- 1 l Techne spinners
- 1 l LH fermentors (equipped with spin filters and perfusion)
- 150 l Electrolux fermentor (spin filter and perfusion)

17
 Process outline and suggested experiments

Process 1a

Step 1:
Thaw ampoules of frozen cells. Seed into a 150 cm2 flask in 50 ml of
medium. A normal value is 1 x 106 cells/vial. If we assume a plating
efficiency of 60% and a doubling time of 24 hours, confluence
(2 x 107 cells) will be reached in
5 days (120 hours)
x1
Experiments to be performed:
- Check the viability of the thawed cells and the plating
efficiency on plastic surfaces. 1 day
- Check productivity of the cells (final product titer and
productivity on a per cell/time basis). 1 day
- Set up a number of flasks to check growth kinetics
(daily crystal violet counts). 5 days

Step 2:
Harvest the cells from a culture flask using 0.02% trypsin (after rinsing
with EDTA-PBS). We normally find a recovery of 60% - 1.2 x 107 cells/
x1 flask. This is enough cells to seed 12 flasks.
These will be confluent after 5 days (120 hours)

Experiments to be performed:
- Check viability, recovery and plating efficiency of cells
harvested from flasks at different days after seeding to
x 12 find the time where it is possible to harvest the maximum
amount of viable cells, in a single cell suspension. 5 days
- Check the productivity of the cells (final product titer
and productivity on a per cell/time basis). 1 day

18
Step 3:
Harvest the cells from flasks using 0.02% trypsin.
12 x 1.2 x 107 = 1.4 x 108 cells. If these cells are seeded into a 1 l
microcarrier culture and we have an attachment efficiency to the
microcarriers of 70% we will start at 1 x 105 cells/ml.
This culture will be confluent in 4 days (96 hours)

Experiments to be performed: x 12
- Check the viability and % recovery of the cells. 1 day
- Check the attachment efficiency to different
microcarriers in stationary culture (Cytodex 1, 3). 2 days
- Do growth curves on microcarriers in spinners. 5 days
1x1l
- Check the appropriate microcarrier concentration
and optimal inoculation concentration. 10 days
- Check the starting conditions (initial culture volume,
static attachment period, stirring speed). 5 days
- Check medium pH. 5 days
- Check the optimal medium replacement strategy
(no change from start to end or replace a percentage
at intervals). 5 days
- Check the productivity of the cells. 1 day
Note: photo document carrier samples to check homogeneity of the
culture during attachment, to check the utilization of the available
surface area, to detect multi layering and bridging between beads or
detachment of cells from the carriers. Also eventual crushing of the
beads will be documented.

19
 Step 4:
Harvest the cells from the microcarriers using 0.2% trypsin.
From a 1 l culture at 1 x 106 cells/ml and an expected recovery
of 6 x 108 viable cells will be harvested. With an attachment
x 12 efficiency of 70% a total of 4.2 x 108 cells will be available
for the subsequent culture. This is enough cells to start
4 l at a seeding density of 1 x 105 cells/ml.
These will be confluent in 4 days (96 hours)

1x1l Experiments to be performed:

- Optimize the harvesting procedure
(washing procedure, trypsin concentration,
incubation time, pH). 1 day
- Check the viability, recovery, and attachment
efficiency at different times after inoculation of
the spinner culture to optimize the scale up process.
Should be done on different types of microcarriers. 5 days
- Check the growth rate. 5 days
- Check the productivity of the cells. 1 day

Step 5:
Harvest the four spinners. From four spinners
we can start 16 l. This is done in a stirred well
controlled fermentor.

4x1l The culture will be confluent in 4 days (96 hours)

Experiments to be performed:
- Check the viability, recovery, and attachment efficiency. 1 day
- Check the growth rate, medium replenishment rate. 5 days
- Check the productivity of the cells. 1 day
1 x 16 l

20
Step 6:
Harvest the 16 l culture. This can either be done within the fermentor or
by using a trypsinization vessel (Vibromixer E1, Chemap) (6). From 16 l
we can start 64 l. Given a vessel that allows a linear scale up (e.g. with a
conical shape) it is possible to do the scale up in the same vessel.
This culture will be confluent in 4 days (96 hours)
16 l
Experiments to perform:
- Check the viability, recovery, and attachment efficiency. 1 day
- Check the growth rate. 5 days
- Check the productivity of the cells. 1 day
- Develop and check reproducibility of standard
operating protocols. 64 l
- Optimize economy with respect to downstream steps,
media costs and labor.
- Check the karyology of the cells. 1 day
Note: At this stage volume dependent effects, e.g. in the
timing of routine operations, optimal stirring speed
as balanced against oxygen limitation effects etc.
are likely to be seen.

Comments:
The final culture will either be a pure closed batch or will be perfused to collect the product.
Perfusion is facilitated if the vessel is equipped with a spin filter or a sedimentation chamber. In
reality this step by step process development/scale up procedure will be more cyclic, where the
obtained experimental data in one step may require a repeat (or several) of prior steps e.g. a fall
back on stationary flask cultures to isolate mechanical shear effects detected or suspected in the
stirred cultures where the optimum stirring rate or reactor geometry remains to be determined.
In the case that the medium formulation must be changed, all experiments should be redone!

In conclusion:
• Total time spent on this scale up process from start to end; i.e. scale up time = 624 hours
(26 days).
• For experimental time and repeats add 3550 hours (148 days).
• It is essential to always have back up cultures running, to restart experiments. Especially
during the later stages of the process, a lot of time is otherwise wasted to restart an experiment.
• Number of cells generated: 6.4 x 1010 cells.
• Cell mass generation 1 x 108 cells/h.

21
 Process 1b

Steps 1-3:

The same as for 1a. 14 days (336 hours) – 47 days

Step 4:
Dilution of the spinner culture. Instead of harvesting
the microcarriers the culture is simply diluted with new
microcarriers and medium. In some cases it is possible to
1l do a 1:10 dilution by volume (c127) without slowing
down the growth. In such a case a 10 l culture (105/ml)
may be started from a 1 l culture (106/ml). (Another
alternative is to increase the microcarrier concentration
in steps from 2–3 g/l up to 15–20 g/l in the same vessel.
This is preferably done in one and the same fermentor.
See process 2b).

10 l This culture will be confluent in 4 days (96 hours)

Experiments to be performed:
- A series or experiments with different
dilution rates/ratios. 10 days
- Check growth rate. 5 days
- Check the productivity of the cells. 1 day
- Compare these data with those from process 1a
to see which process that is most efficient.

22
Step 5:
Dilution of the fermentor 10 l - 100 l.
This culture will be confluent in 4 days (96 hours)

Experiments to be performed:
- Check the productivity of the cells. 1 day
10 l
- Check the growth rate. 5 days
- Check the karyology of the cells. 1 day

100 l

Comments:
Compare this process to process 1a; Four full working days should be saved, with about a
twofold increase in total cell mass obtained. (add the possibilities for maintaining a high
density of cells in a productive "milking" stage, possibly at a reduced serum concentrations,
for products weakly linked to cell growth per se. See process 2b).

In conclusion:
• Scale up time: 528 hours (22 days).
• For experimental time and repeats add: 3360 hours (140 days).
• Number of cells generated: 1 x 1011.
• Cell mass generation 2 x 108 cells/h.

23
 Process 2a

Step 1-3:
Is the same as in 1a. 14 days (336 hours) – 47 days

Step 4:
Harvest the cells from 2 x 1 l spinner cultures. We use
1 x 109 cells to seed a 1 l culture with 15–20 g/l of
Cytodex. These cultures are perfused with 1–2 volumes
2 x 1 l LCD of medium/day after 15–20 hours.
The culture will be confluent (1 x 107 cells/ml) in 4 days (96 hours)

Experiments to be performed:
1 l HCD - Optimization of the harvesting procedure (see process 1a). 1 day
- Check the viability, recovery and attachment efficiency. 1 day
- The optimal bead concentration has to be determined
also the optimal perfusion rate (four reactors used). 15 days
- Check the growth rate. 5 days
- Check the productivity of the cells. 1 day
- A different medium composition might be necessary for
high density cultures (serum dependence should
be checked). 10 days
- Separate feeds to supplement the base medium perfusion
could be investigated (amino acid, glutamine, glucose, serum).

24
Step 5:
Harvest the high cell density culture. With an expected
recovery of 60% and a subsequent attachment efficiency
of 70% we get 4.2 x 109 cells out of a 1 l culture. This can
be used to start 4 x 1 l fermentors. 1 l HCD
These will be confluent in 4 days (96 hours)

Experiments to be performed:
- A new harvesting protocol has to be developed for the 4 x 1 l HCD
high density culture. 1 day
- Check viability, recovery and the attachment efficiency. 1 day
- Check growth rate. 5 days
- Check the productivity of the cells. 1 day

Step 6:
Harvest the four fermentors. From these four we can start
a 16 l culture in the large fermentor.
This culture will be confluent in 4 days (96 hours)
4 x 1 l HCD

Experiments to be performed:
- Check the viability, recovery and attachment efficiency. 1 day
- Check the growth rate. 5 days
- Check the productivity of the cells. 1 day

16 l HCD

25
 Step 7:
Harvesting of the 16 l culture. From this we can start
a 64 l culture. This scale up is done within the fermentor.
The culture will be confluent in 4 days (96 hours)

16 l HCD Experiments to be performed:

- Check the viability, recovery and attachment efficiency. 1 day
- Check growth rate. 5 days
- Check the productivity of the cells. 1 day
- Check the karyology of the cells. 1 day
Note: At the 1 l HCD level oxygen supply is a problem.
64 l HCD For possible solutions see Ref. 13-17.

In conclusion:
• Scale up time: 720 hours (30 days).
• For experimental time and repeats add 4940 hours (206 days).
• Number of cells generated: 6.4 x 1011.
• Cell generation time 9 x 108 cells/h.

26
Process 2b

Steps 1-4:
The same as in 2a. But just transfer the
microcarriers and allow the cells
to migrate. 18 days (432 hours) – 78 days

1 l HCD

Step 5:
Dilute the 1 l culture up to 10 l into the fermentor.
The culture will be confluent in 4 days (96 hours)
(An alternative is to start 100 ml at 3 g/l and then add more 1 l HCD
microcarriers and gradually increasing the perfusion).

Experiments to be performed:
- The dilution rate/ratio. 10 days
- Check the growth rate. 5 days
10 l HCD
- Check the productivity of the cells. 1 day

Step 6:
Dilute the 10 l culture to 100 l. The culture
will be confluent after 4 days (96 hours)
100 l HCD
Experiments to be performed:
- check the growth rate. 5 days
- check the productivity of the cells. 1 day
- check the karyology of the cells. 1 day

In conclusion:
• Scale up time: 624 hours (26 days).
• For experimental time and repeats add: 4850 hours (202 days).
• Number of cells generated: 1 x 1012.
• Cell generation time: 1.6 x 109/h.

27
References
1. Pure gelatin microcarriers: Synthesis and use in cell attachment and growth of fibroblasts and
epihtelial cells. Wisseman, K. W. and Jacobsson, B. S. In Vitro 21, 391–401 (1985).
2. Receptor functions for the integrin VLA-3: Fibronectin, Collagen, and Laminin binding are
differentially influenced by ARG-GLY-ASP peptide and divalent cations. Elices, M. J. et al.,
J. Cell Biology 112, 169–181 (1991).
3. Proteoglycans as modulators of growth factor activities. Rouslahti, E. and Yamaguchi, Y.
Cell 64, 867–869 (1991).
4. Large scale cell culture technology. Tolbert, W. R. and Feder, J. Annual Reports on
Fermentation Processes 6, 35–73 (1984).
5. Micricarrier cultures of animal cells. van Wezel, A. L. Tissue Culture: Methods and
Applications (Kruse, P. F., Patterson, M. K., eds.) Academic Press, New York, 372–377 (1973).
6. The production of inactivated poliovaccine on serially cultivated kidney cells from captive-bred
monkeys. van Wezel, A. L. et al., Develop. Biol. Standard 46, 151–158 (1980).
7. Subpassaging cells on microcarriers: the importance for scaling up to production.
Lindner, E. et al., Develop. Biol. Standard 66, 299–305 (1987).
8. Serial propargation of mammalian cells on microcarriers. Hu, W. S. et al., Biotechnol.
Bioeng. 27, 1466–1476 (1985).
9. Motility factors on the march. Warn, R. M. and Dowrick, P. Nature 340, 186–187 (1989).
10. Genetic analysis of protein folding pathways. King, J. Bio/Technology 4, 297-303 (1968).
11. Microsphere-induced aggregate culture of animal cells. Goetghebeur, S. and Hu, W. S.
Production of Biologicals from animal cells in culture 423–428 (1991).
12. Chinese Hamster Ovary cells continuously secrete a cystein endopeptidase. Satoh, M. et al.,
In vitro 26, 1101–1104 (1990).
13. Engineering developement in homogeneous culture of animal cells: Oxugenation of reactors
and scale-up. Aunins, J. G. et al., Biotech. and Bioeng. Symposium 17, 699–723 (1986).
14. Mammalian cell culture: Engineering principles and scale-up. Glachen, M. W. et al.,
Trends in Biotech. 1, 102–108 (1983).
15. The use of caged aeration for the growth of animal cells on microcarriers. Whiteside, J. P. et al.,
Develop. Biol. Standard 60, 283–290 (1985).
16. Chemcell brochure, supplied by Chemap.
17. Bubble-free cell culture aeration with porous moving membranes. Lehman, J. et al.,
Develop. Biol. Standard 66, 227–240 (1987).

28
Cytodex is a trademark of Amersham Pharmacia Biotech Limited or its subsidiaries.

Amersham is a trademark of Nycomed Amersham plc

Pharmacia and Drop Design are trademarks of Pharmacia & Upjohn

All goods and services are sold subject to the terms and conditions of sale of the company within
the Amersham Pharmacia Biotech group that supplies them.
A copy of these terms and conditions is available on request.
© Amersham Pharmacia Biotech AB 2000 – All rights reserved.

Amersham Pharmacia Biotech AB Björkgatan 30, SE-751 84 Uppsala, Sweden

Amersham Pharmacia Biotech UK Limited Amersham Place, Little Chalfont, Buckinghamshire HP7 9NA, England
Amersham Pharmacia Biotech Inc 800 Centennial Avenue, PO Box 1327, Piscataway, NJ 08855 USA
Amersham Pharmacia Biotech Europe GmbH Munzinger Strasse 9, D-79111 Freiburg, Germany
Amersham Pharmacia Biotech KK, Sanken Bldg. 3-25-1, Hyakunincho Shinjuku-ku, Tokyo 169-0073 Japan
www.apbiotech.com