A presentation by KBI Scientist Shahid Rameez, Ph.D. at the American Chemical Society Annual Meeting– Biochemical Technology (BIOT) Division, New Orleans, LA

1. High throughput bioreactor mimetic in
l d l t t d l tearly and late stage process development
American Chemical Society – Biochemical Technology (BIOT) Division
245th – ACS National Meeting, New Orleans, LA
Shahid Rameez, Ph.D.Shahid Rameez, Ph.D.
Scientist I, Process Development
KBI Biopharma Inc, Durham, NCp

2. Overview
Ambr System
Bioreactor
Mimetic
In process
Mimetic
KBI Evaluation
p
Activities
&
Design Space
Evaluation
Experimental Design
in Early and Late
C t l St t i
y
Stage Process
Development
Control Strategies

3.  A key bottleneck in biopharmaceutical development has been they p p
rapid development of robust and scalable manufacturing processes that
can permit accelerated progress of products into clinical trials.
 Innovations in this area can have a very significant impact on the
overall economics of biopharmaceutical drug development by
d i h i i k h h li idecreasing the time it takes to reach the clinic.
 Mammalian cell culture processes typically have the longestp yp y g
experimental duration with 2-3 weeks being a typical duration for the
production bioreactor step with additional time spent on the seed
culturescultures.

4.  Shake flasks provide the capability to perform high throughput
experiments, but with an inability to control process parameters likeexperiments, but with an inability to control process parameters like
agitation rate, dissolved oxygen (DO) and pH.
 Th t l h l h d l i b t ffi i t These parameters play a huge role when developing a robust, efficient
cell culture process which dictates the product quality and yield.
 Moreover, reproducibility and scalability of process and culture
performance is a pre-requisite for successful and efficient process
development at a scale down level.development at a scale down level.
 At KBI we did case studies evaluating the ambr™ system (an
a tomated micro scale bioreactor s stem) to establish the roleautomated micro-scale bioreactor system) to establish the role
this system could play in accelerating biotech drug development.

5. Automated Miniaturized Bioreactors
1mL tips
ambrTM Technology
1mL
or
Liquid
Handler
4mL tips
Handler
Used Tips
Discard
Culture stations; each holding 12 bioreactors

6. ambrTM Deck Layout
Culture stations
1ml tips
4ml tips4ml tips
24 deep-well plate Tip box lid
Plate lid

7. The Vessel
Vessel Cap:
In-line filter
on gas supply
DO sensor
on gas supply
Impeller
pH sensorp

8. Culture Station Nomenclature
Culture station 1 (CS1) Culture station 2 (CS2)
CS1-1
CS1-2
CS1-3
CS1-4
CS1-5
CS1-6
Culture station 1 (CS1)
CS2-1
CS2-2
CS2-3
CS2-4
CS2-5
CS2-6
Culture station 2 (CS2)
7
8
9
10
11
12
7
8
9
10
11
12
CS1-7
CS1-8
CS1-9
CS1-1
CS1-1
CS1-1
CS2-7
CS2-8
CS2-9
CS2-1
CS2-1
CS2-1
Id l ki l i 13mL W ki l i 11 15mLIdeal working volume is 13mL; Working volume range is 11 – 15mL
Things to consider:
– Feed strategy and Sampling strategyFeed strategy and Sampling strategy
– Low volumes – gas entrainment / vortexing
– High volumes – Limits kLa.

9. PART 1:
Reproducibility for Results and Key Observations during Cell Culture Process Development
 The processes evaluated in ambrTM were previously developed from a rigorous
cell culture process development performed in classical bioreactors of various
scales. The process were successfully carried in bioreactors across various scales: 2L,
10L d 200L10L and 200L.
 This study aimed at studying the reproducibility of the key observations of the
processes in ambrTM. Thus a reverse engineering approach was adopted, where we
were cognizant about the outcomes from most of the experiments as far as
inducing process variations was concerned.
 The reproducibility of key historical results in ambrTM would corroborate towards
its capability as a high throughput bioreactor mimetic in cell culture process
development.

10. K Ob i f Hi i l D
Case Study 1: Reproducibility evaluation for the production of a monoclonal antibody in a
recombinant Chinese Hamster Ovary (CHO) cell line.
Key Observations from Historical Data:
• Temperature Shift during the cell culture process was found to be the most important
process factor to regulate the productivity of the antibody titer.
• The CHO cell line performed better at lower pH set point of 6.85 as compared to pH
set point of 7.00.
• Feeding intermittently had shown to regulate growth and productivity in the process.
Intermittent feeding had showed better results than just Day 0 additions for feed.
• A highly basic and a critical feed, referred here as FDX, had to be added without pre-
neutralization with acids to avoid osmolality increase in cultures. Thus, a better control
had to be established in the bioreactors to control the pH drift with addition of FDX.
h h d b b h d hThis was achieved in classical bioreactors by tuning the PID controllers and with
regulation in cascade feedback gassing of CO2 and Air.

11. Results: Both lower pH set-points and Temperature Shift showed higher cell growth, better
cell viabilities. DO as suspected at negligible effect on cell growth and viability.
Time courses for viable cell growth and viability for recombinant CHO cell line with changing
(A) Process pH (B) Temperature (C) Dissolved Oxygen (DO) levels and (D) Feeding Strategies. The
experimental data shows an average of 2-3 vessels in the ambr 24. The error bars show the standard
deviationdeviation.

12. Results: Both lower pH set-points and Temperature Shift showed higher cell titers. As
observed historically for this process, Temperature shift was found to be the most important
process factor to regulate the productivity of the antibody titer.
 The ambrTM system can be used as a high-throughput platform to make key
process decisions during the early process development phase of
bi h i l d lbiopharmaceutical development.

13. PART B:
Scalability Assessment in Cell Culture Process Development
Case Study 2: Comparison across scales for the production of a monoclonal antibody in a
recombinant CHO cell line.
Ambr (n = 3).
2L (n = 1).
10 ( 4)10L (n = 4).
200L (n = 1).
• Harvest Titers within 1 5 1 7 g/L• Harvest Titers within 1.5 -1.7 g/L
across all scales.
Comparison of time courses for viable cell growth and viability for recombinant CHO cell line in ambrTM and other
scales bioreactors: 2, 10L Glass bioreactors and 200L disposable bioreactor.p

14. Case Study 3: Comparison across scales for the production of a protein molecule in a
recombinant Chinese Hamster Ovary (CHO) cell line.
Results: The cell growth, cell viability and cell
viabilities were comparable between ambrTM,10 and
200L Bioreactors200L Bioreactors.

15. Case Study 3: This protein molecule had two Isodimers (A and B). The levels of Isodimers A
and B were a product quality attribute.
Results: Ratio of Isodimers A
and B were similar (± 5% of
mean values) across ambrTM , 10
and 200L Bioreactors.
 The process decisions and results from ambrTM were reproducible to the
results in other scales bioreactors.
Both the case studies (with antibody and a non antibody) demonstrate the
utility of the ambr™ system as a high throughput system for cell culture
process development.p p

16. PART C:
Control for process pH and DO during Cell Culture Process Development
 pH control in ambrTM is established using the
automated liquid handler based base additions
when pH drops below the pH set pointwhen pH drops below the pH set point.
 When the pH exceeds the pH set point, the
CO2 flow rate increases to establish control on
the pH drift.
CO2: 0 - 1.24 mL/min. Delivered on demand to control pH.
O2 : 0 - 1.24 mL/min. Delivered on demand to control dissolved oxygen.
N : 0 1 24 mL/min Flow rate is constantN2 : 0 - 1.24 mL/min. Flow rate is constant.

17.  Online profiles for process pH (top Online profiles for process pH (top
figure) and DO (bottom figure) levels
during the culture duration for CHO cell
line expressing a recombinant antibody
in ambrTM.
 The spikes in the DO profiles
corresponded to bioreactor samplingcorresponded to bioreactor sampling,
Liquid additions and Sampling.
 All these disturb the headspace and
l h ki l Th i falter the working volume. The time for
the DO traces to equilibrate to setpoint
after such manipulations would depend
on the controller setup.p

18. Case Study 4: Artificial perturbations in pH and DO (by adding a basic feed and changing DO
set points respectively) during production of an antibody molecule in a recombinant CHO cell
process. Through adjustments to the PID control loop and gas flow rates the capability of
b ™ t l t dambr™ system was evaluated.
Results: Tuning the gas flow limits and proportional gains in the PID loop of ambr™ system.
By changing the proportional gain by eight folds and CO2 gas limits by 1.25 and > 2folds as
opposed to default manufacturer values , the pH drifts were reduced by 23 and 47 % of initial
value, respectively.value, respectively.

19. Results: The DO set points were changed to 80% from 20 and 40%, respectively and
changed back to original values The level for DO was maintained at 80% for duration of 6changed back to original values. The level for DO was maintained at 80% for duration of 6
hours and returned to original set points 20 and 40% in ≈ 90 and 120 mins, respectively.
 Th p bilit f ind in d i ti n n h lp in d i nin r t The capability of inducing deviations can help in designing worst-case
experiments. It enables to test operating limits with respect to particular
key operational parameters (DO, pH) in a process.

20.  The combination of pH and DO control and an automated liquid handling system in
ambrTM system overcomes major limitations of conventional small-scale cultures vessels
Conclusions
ambr system overcomes major limitations of conventional small scale cultures vessels
especially shake flasks.
 The single-use, pre-calibrated, and instrumented vessels used in ambrTM system provides
a platform for high-throughput in cell culture process development while mimicking a
stirred-tank bioreactor environment.
 The reproducibility of key observations observed in historical process developmentT e ep od c b ty o ey obse vat o s obse ved sto ca p ocess deve op e t
demonstrated that ambr™ is capable of providing predictive results under bioreactor
relevant process conditions.
R d ibili l bili d h bili f h d b i h Reproducibility, scalability and the ability of the system to respond to perturbations show
ambrTM to be adequate to consider this system for early and late stages of cell culture
process development.
 The studies at KBI aimed to demonstrate the utility of the ambr™ system as a
high- throughput bioreactors that can offer the realistic possibility of decreasing the
process development time for investigational biopharmaceuticals to reach the clinic.

21. Process
Characterization
Commercial ProcessProcessCell LineDiscovery Manufacturing
Biopharmaceutical Development Process.
Characterization
and Validation
DevelopmentDevelopmentDevelopment
y
Stage
RAPID PRODUCT DEVELOPMENT AT KBI
Manufacturing
ambrTM
Process Development
(Design Space & Optimization)
Cell Line DevelopmentDiscovery Stage Manufacturing
• Platform Downstream Processes
• High-throughput Resin Screening
• Single - Use Technology
TM The combination of methodologies such as ambrTM, Platform Downstream
Processes, High-throughput Resin Screening and use of Single-use technology
can significantly shorten the window for process development and
f t imanufacturing.

22. Acknowledgements
• Joe McMahon President and CEO
• Abhinav Shukla, Ph.D. VP, Process Development and Manufacturing
• Sigma Mostafa Ph D Director Process DevelopmentSigma Mostafa, Ph.D. Director, Process Development
• Haiou Yang, Ph.D. Scientist II, Process Development
• Christopher Miller Scientist II, Process Development
• Anushya Mani Scientist I Process Development• Anushya Mani Scientist I, Process Development
• Joe Jirka Product Specialist, TAP Biosystems
 Process Development Team at KBI Process Development Team at KBI

23. ThanksThanks
Questions??Questions??