Evaluation of a Novel Micro-Bioreactor System For Cell Culture Optimisation

Giles Wilson
1,

1
Novo Nordisk A/S, Hagedornsvej 1, Gentofte 2820, Kbenhavn, Denmark.

GCW@novonordisk.com

Conclusions
The original objectives of this experiment were as follows:-
1. Assess if the SimCells can mimic a larger scale fermenter.
2. Identify an optimum pH and temperature.
In the first case the SimCells have been extremely successful.
An optimum pH and temperature has been identified and
the data that has been generated is statistically valid. As
stated previously the experiment performed using the
SimCells evaluated over 50 different combinations of pH and
temperature. This experiment would have taken us 1 year
to perform in ten 5 litre fermenters. The advantage of the
SimCell system is that it allows us to examine a wide area of
experimental space in a short amount time. In particular pH
has been shown to have had a very large effect upon culture
performance. The data generated by the SimCells is comparable
to that seen at 5 litre scale in high cell density fermentations
such as suspension cell perfusion and microcarrier culture. The
absolute titres of product is generally lower in the SimCells but
the behaviour of individual cells in terms of specific production
rates is very similar. In conclusion, the SimCell system has
generated useful data and pH has been shown to have a very
large effect upon culture performance. A small difference in pH
can have a wide ranging effect on product titre and this may be
particularly relevant at very large scales of operation where pH
gradients will form in the vessel during pH control.
Abstract
BioProcessors SimCell high throughput cell culture process development system is a roboti-
cally controlled micro-bioreactor system capable of conducting over 1000 parallel bioreactor
experiments.
The animal cell cultures are grown in approximately 600l volumes in a format capable of simu-
lating environmental control and stresses that are present in conventional fermenters.
Each SimCell Micro Bioreactor Array contains 6 x 600 l mini-fermenters and allows the operator
to rapidly screen many different parameters using a robotic control system.
In this case we operated the dual chamber SimCell micro-bioreactors in a perfusion mode with a
CHO cell line in serum free conditions for over 10 days and examined the effects of temperature
and pH in 50 different combinations. With 9 replicate data points for statistical analysis this study,
in the normal course of events, would have utilised ten 5 litre tanks for 1 year.
In comparison to 5 litre perfusion fermentations (both suspension cell perfusion and microcarrier)
the microbioreactors gave equivalent levels of growth and production, demonstrating that the
micro-scale system can mimic a larger fermenter.
An optimisation of pH and temperature showed that culture pH has a very strong effect on pro-
ductivity and that small changes in pH can have a large effect upon culture performance.
The scalability of these observations made in the SimCell MicroBioreactor Arrays were confirmed
in our validated 5L bench top fermenter model.
Evaluation Project Background
Purpose:
Evaluation of a high throughput system for cell culture process development capable of the
monitoring and control associated more with bench-top bioreactors than shake flasks.
Rationale and Experimental Approach:
The speed with which any production process can be optimised is limited by the number of
experiments that can be performed and the availability of equipment. In some cases, such as
medium screening, simple technologies such as shake flasks are more than adequate to meet
this task. Shake flasks are simple to operate, cheap, and many experiments can be performed
at one time.
However, certain parameters such as pH, DOT, temperature, feeding strategy and high cell
densities can only be evaluated in a fermenter. In this study we wished to evaluate the
effects of shifting temperature and pH in a high cell density process. It was proposed that a
combination of 10 different temperature and 5 different pH set points was a sufficiently broad
range of parameters to allow us to accurately identify an optimum set of conditions. However,
to study all of these parameters would take at least 50 fermentations.
The SimCell system from BioProcessors Corp is a micro scale bioreactor designed to fit a 96 well
plate format. Each plate MicroBioreactor Array contains 6 micro-bioreactors with a 300-800l
working volume each and it is possible to control temperature, DOT, pH, feed rates, medium
exchange rates and perfusion rate. The system can also monitor the pH, temperature and
biomass concentration of each bioreactor and can sample the reactors for off-line parameters
such as substrate and product concentrations.
In this study, the SimCell system will be used to evaluate the effects of a shift in temperature
and pH on cell culture performance. This represents 50 different combinations of pH and
temperature. The cell line under evaluation is a CHO cell line expressing a recombinant protein
in a perfusion fermentation.
Characterization of the Process
Response to Temperature and pH
Resource Requirements
Time & Resources needed to complete this study in con-
ventional fermenter systems.
(450 experiments / 10 fermenters) x 1 week = min 45 weeks
(max 68 weeks)
37 fermenters or 6 x SixFors = 3 months
Significantly Less Resources needed to complete this
study in MicroBioreactors.
Assuming a fully automated robotic system with 5 incubators
10 temperatures requires, 2 runs (5 temperatures per run)
Elapsed time 1 month (2x2weeks)
Requiring a total of 1 person and an estimated 50-60 fte hours
Surface Response Characterization Results Summary
Process Response to Temperature and pH
All of the results presented are after the shift in pH and
temperature.
As expected a decrease in temperature caused a reduction in
doubling time.
A change in pH had a minimal effect at high temperatures.
At low temperatures however, a low pH severely depressed cell
growth.
An increase in temperature and pH caused a high lactic acid
concentration.
A high pH in particular caused a high lactic acid concentration at
mid to high temperatures.
A calculation of yield coefficient for lactate from glucose gave a
similar trend. The plateau is not caused by substrate limitation.
A shift in glucose metabolism at high temperature and pH is
implicated.
A high temperature had a negative effect upon production.
A high pH had a positive effect upon production.
Small changes in pH had a very large effect upon production.
Characterization of the Process Response to
Temperature and pH Experimental Outline
Process to Be Characterized:
Perfusion process for production of a therapeutic protein from
CHO cells. Cells are suspension adapted and grown in an animal
component free medium.
Experimental Summary:
Cells cultivated in SimCells in a batch mode (at the same pH and
temperature) until the cell number is greater than 2x10
6
cells ml
-1
A shift in temperature and pH is initiated
10 different temperature and 5 different pH values were
studied.
50 conditions repeated with n = 9, or 450 experiments
SimCell micro-bioreactor operated in perfusion mode after the
shift in pH and temperature.
5 day process
reactor volume exchanged twice per day
Samples retained for metabolite and product concentration.
SimCell

MicroBioreactor Array Design

6 SimCell micro
bioreacIors per array
Workihg volumes rahge
!rom 300 - 800l
Micro!luidic chahhels !or ihoculaIioh
!eeds, pH ad|usImehI & samplihg
Mammaliah Cell CulIures achieve higher Ihah
5x10 Viable Cells/ml
7
SimulaIes all sIahdard producIioh modes:
8aIch, Fed 8aIch ahd Per!usioh
ProprieIary gas-permeable maIerials
!aciliIaIe gas IrahsporI wiIhouI Ihe use o!
IradiIiohal sIirrers

CulIure mohiIorihg via exIerhal
opIical ihIerrogaIioh o! ih-chamber
sehsors

Automation Management System
IhcubaIors
OpIical MohiIorihg
Fluidics Samplihg
Fluidics
Feeds
Characterization of the Process Response to
Temperature and pH Resource Requirements
0ROCESS2ESPONSE%XPERIMENTS
1emp vs pH Crid = 50 CohdiIiohs
9 replicaIes = 450 8ioreacIor experimehIs
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
10
9
8
7
6
5
4
3
2
5 4 3 2 1
1
e
m
p
e
r
a
I
u
r
e

S
e
I

P
o
i
h
I
pH SeI PoihIs
!NALYTICAL$ATA0ER%XPERIMENT

Daily:
Cell CouhI
Clucose
CluaImihe
LacIaIe
ProIeih
1wice Daily:
pH
>22500 daIa poihIs !rom 450 experimehIs

Mean Doubling Time

Peak Lactic Acid Production
low
high
lo
w
0
0,1
0,2
0,3
0,4
0,5
Yc
pH
Temperature
Mean Yield Coefficient (Lactate)
low
high
low
high
0
5000
10000
15000
20000
25000
Y
i
e
l
d

(
A
U
/
l
/
1
2

h
o
u
r
s
)
pH
Temperature
Peak Production
SimCell

Evaluation Fermenter Comparison Results

Summary
All of the results are presented in the table. All of the data is
taken at the same pH and after 5 days in perfusion conditions
at the respective temperature. The product concentration for
each fermenter type as been adjusted to be in arbitrary units of
concentration per day.
The data between all of the systems is comparable particularly
with regards to specific productivities at different temperatures.
A larger difference is present in the product concentration
between the SimCells and the 5 liter scale studies.
However, conditions which increase productivity are the same in
both systems.

SimCell

Evaluation Fermenter

Robotic management of experiments
- SimCells are automatically moved
from station to station
1260 simultaneous cell culture
experiments
Automatic measurement, control
and adjustment of experiment
parameters
Automatically inoculate, build media,
feed and base adjust cultures
Automated downstream sample
handling for nutrient, metabolite
and product profiles.
Automated, integrated data
management
Surface Response Characterization Results
lo
w
h
ig
h
lo
w
0
1
2
3
4
5
6
7
Days
pH
Temperature
low
high
low
high
0
0,5
1
1,5
2
2,5
3
3,5
g/l
pH
Temperature
Parameter
Fermentation Method
SimCell
Perfusion:
5 litre
Microcarrier:
5 litre
High
Temperature
Low
Temperature
High
Temperature
Low
Temperature
High
Temperature
Biomass
(x10
6
cells ml
-1
)
6.08 9.3 5.75 7.8 9.47
Product
(AU l
-1
d
-1
)
16600 31150 23400 41500 24500
Specic Production
(acd)
2600 6250 3100 4150 6230 - 7270 2080 - 3100

Integrated SimCell System Monitoring

and Control Capability
BioMass
Microbioreactors support cell growth to over 20million cell/ml
Batch, Fed Batch and Perfusion culture simulation
Accurate count of total cells by forward Light Scattering OD measurement
Temperature Temperature of cultures controlled to 0.1oC
pH Culture pH monitored and controlled to 0.05 and 0.1pH units respectively
Oxygen
Oxygenation capability in the range of conventional bioreactors 1-10hr-1 KLA
DO measured and controlled
Carbon Dioxide
CO2 off gas capability equivalent to a 33hr-1 KLA
CO2 measured and controlled
Agitation
Culture agitated by an orbital rotary motion.
Rapid mixing, peak and average shear control
Off-line analytics Viability, nutrients, metabolites and product