Cell line development: Reducing timelines and increasing titres by identification of host cell lines with improved characteristics Dr Alison Porter, Fujifilm Diosynth Biotechnologies, UK Cell Line Development and Engineering, 09 September 2014, Berkeley CA 1
41
Embed
Reducing Timelines & Increasing Titres by Host Cell Lines with Improved Characteristics
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Cell line development: Reducing timelines
and increasing titres by identification of
host cell lines with improved characteristics
Dr Alison Porter, Fujifilm Diosynth Biotechnologies, UK
Cell Line Development and Engineering,
09 September 2014, Berkeley CA1
FUJIFILM Diosynth Biotechnologies
EU and USA operations
Track record
~950 staff
Over 1000 cGMP batches manufactured
5 Commercial Products
FDA & MHRA Inspection history
Extensive LSS application
Contract Manufacturing &
Process Development
Billingham (UK)
RTP (NC, USA)
Scope
►Aims and Objectives
►Host cell line development (CLD)
• Directed evolution approaches
• Results
►CLD run-through
• Experimental design
• Results
►Summary
3
Aims and Objectives
►To develop a mammalian expression platform
which rapidly leads to efficient, robust and high
quality biomanufacturing processes
►New host cell line
►New vector
►Optimised cell line development process
►Complementary medium and feed platform
4
5
Host cell line development
Desirable features/attributes for a host cell line
Adapted to CD media
Adapted to chemically
defined (CD) media
Short doubling time
1 to 5 g/L titre for a model
mAb
Suitable product quality
Rapid adaptation to
production media
CHO-based
6
How do we improve host cell functional capability?
CELL LINE
ENGINEERING?
DIRECTED
EVOLUTION?
• Despite decades of research
and a variety of different
strategies to engineer CHO
cells, outcomes remained
mixed
• May be a need to engineer
multiple targets
• May require a systems-based
analysis for a tailor-made
response
• Utilise functional heterogeneity
within CHO cell populations to
identify better host cell lines
• We already utilise this
approach
• Identify high producing
recombinant clones during
CLD
• Adaptation to serum-free
chemically defined media
• Potential ability to manipulate
multiple phenotypic outputs
7
Isolate a cell line
with more desirable
attributes from a
heterogeneous cell
population
Approaches taken to obtain a host cell line
with improved features/attributes
Adaptation
to
chemically
defined
conditions
ChemostatSubculture
regimeCloningFACS
Multiple rounds of
sorting for cells with
extended viability
during batch shake
flask screening
Allows for evolution
of rare mutants with
desirable growth
properties within a
large population
Continuous
subculture (different
regimes) and ability
to grow in low [Gln]
for lower [Amm]
production
8
Derivation of multiple DG44 variants able to grow
in chemically defined (CD) culture medium
Static culture
100% MEM-α
(+10% FBS)
Suspension
100% MEM-α
(+10% FBS)
50% CD-DG44
50% MEM-α
(+10% FBS)
75% CD-DG44
25% MEM-α
(+10% FBS)
90% CD-DG44
10% MEM-α
(+10% FBS)
C1
100% CD-DG44
Suspension
75% OPTI-CHO
25% MEM-α
(+10% FBS)
80% OPTI-CHO
20% MEM-α
(+10% FBS)
90% OPTI-CHO
10% MEM-α
(+10% FBS)
C2
100% CD-DG44
50% CD-DG44
50% OPTI-CHO
75% CD-DG44
25% OPTI-CHO
C3
100% CD-DG44
9
0.00
0.25
0.50
0.75
1.00
0 50 100 150 200
Gro
wth
rate
(d
ay
-1)
Days in culture
The long and winding road to CD media
adaptation: The C2 story
120 rpm110 rpm 140 rpmStatic
100% MEMα
+10%FBS
75% OptiCHO
80% OptiCHO
85% OptiCHO
90% OptiCHO 100% CD DG44
Clumping issues
3-day sub3/4-day sub
0.2 x 106 seeding0.3 x 106 seeding0.2 x 106 seeding
1:100 anti-clump 1:700 anti-clump No anti-clump
10
Desirable growth rate
for a recombinant cell
line assuming:
• Seed = 0.2 x 106/mL
• VCD obtained = 2.0 x
106/mL
• Culture duration = 4
days
• Allows a 1:10 split
ratio during seed
train
► Assess growth and metabolite profiles during an 8-day batch shake-
flask screen (at which point viability < 50%)
• Indication of a cell line’s potential ability to grow during extended
culture
• Ideally have a limited accumulation of waste metabolites (e.g.
ammonium and lactate)
Batch shake-flask screening of CD variants
11
0
5
10
15
C1 C2 C3 2B-1 2B-2
IVC
(x
10
6cell∙day/m
L)
0
10
20
30
C1 C2 C3 2B-12B-2
Meta
bo
lite
C
on
cen
trati
on
(m
M)
Amm
Lac
0
20
40
60
80
C1 C2 C3 2B-1 2B-2
Tra
ns
fec
tio
n
eff
icie
nc
y (
%)
► Assess transfectability (using GFP) and overall expression/secretion capacity
(using a model mAb)
• Comparison of percentage GFP positive cells 24 hours after transfection
• Comparison of secreted mAb concentration 5 days post-transfection
► Higher transfection efficiency and transient titre obtained with cell line C2
C2 progressed to directed evolution stages due to superior
performance during evaluation
Higher transfection efficiency and transient
titre obtained with cell line C2
12
0
1
2
3
4
C1 C2 C3 2B-1 2B-2
Tra
nsie
nt
mA
b
exp
ressio
n (
mg
/L)DG44-2B in ACF
Comparison of host cell lines obtained by
directed evolution
Adapted to CD media
Growth rate during
subculture
IVC obtained in batch shake-flask screen
Colony survival in semi-solid
medium
Waste metabolite
(Amm and Lac) production in batch shake-flask screen
Transient expression of
mAb 1 and mAb2
13
0
5
10
15
20
25
30D
G44
-2B
C2
Sort
-1
Sort
-2
Sort
-3
Sort
-4
Sort
-5
Sort
-6
Clo
ne
-7
Clo
ne
-11
Clo
ne
-26
Clo
ne
-27
Clo
ne
-36
Clo
ne
-46
Clo
ne
-56
Clo
ne
-59
Sta
t-1
Sta
t-2
Sta
t-3
Sta
t-4
Sta
t-5
Sta
t-6
8 m
M
6 m
M
4 m
M
2 m
M
1 m
M
Do
ub
lin
g tim
e (
h)
Doubling time during subculture
14
► Improved doubling time with C2 compared to DG44-2B: ~27hrs to ~21hrs
► Directed evolution approaches had little effect on doubling time
0
2
4
6
8
10
12
14
DG
44
-2B
C2
Sort
-1
Sort
-2
Sort
-3
Sort
-4
Sort
-5
Sort
-6
Clo
ne
-7
Clo
ne
-11
Clo
ne
-26
Clo
ne
-27
Clo
ne
-36
Clo
ne
-46
Clo
ne
-56
Clo
ne
-59
Sta
t-1
Sta
t-2
Sta
t-3
Sta
t-4
Sta
t-5
Sta
t-6
8 m
M
6 m
M
4 m
M
2 m
M
1 m
M
Batc
h s
hake-f
lask s
cre
en
harv
est
IVC
(x
10
6cell∙day/mL)
IVC obtained during batch shake-flask screening
15
► Potential to increase IVC from directed evolution approaches
► Higher IVC for DG44-2B may be media dependent (presence of hydrolysate)
0
1
2
3
DG
44
-2B
C2
Sort
-1
Sort
-2
Sort
-3
Sort
-4
Sort
-5
Sort
-6
Clo
ne
-7
Clo
ne
-11
Clo
ne
-26
Clo
ne
-27
Clo
ne
-36
Clo
ne
-46
Clo
ne
-56
Clo
ne
-59
Sta
t-1
Sta
t-2
Sta
t-3
Sta
t-4
Sta
t-5
Sta
t-6
8 m
M
6 m
M
4 m
M
2 m
M
1 m
M
mA
b 2
tra
ns
ien
t e
xp
res
sio
n (
mg
/L)
0
2
4
6
8
10
12D
G44
-2B
C2
Sort
-1S
ort
-2S
ort
-3S
ort
-4S
ort
-5S
ort
-6C
lone
-7C
lone
-11
Clo
ne
-26
Clo
ne
-27
Clo
ne
-36
Clo
ne
-46
Clo
ne
-56
Clo
ne
-59
Sta
t-1
Sta
t-2
Sta
t-3
Sta
t-4
Sta
t-5
Sta
t-6
8 m
M6 m
M4 m
M2 m
M1 m
M
mA
b 1
tra
ns
ien
t ex
pre
ssio
n (
mg
/L)
Transient expression
16
► Improved titre obtained from C2 compared to DG44-2B
► Directed evolution had
variable effect: up to 3-fold
increase compared to
DG44-2B
0
20
40
60
80
100
120D
G44
-2B
C2
Sort
-4
Sort
-5
Sort
-6
Clo
ne
-7
Clo
ne
-11
Clo
ne
-26
Clo
ne
-27
Clo
ne
-36
Clo
ne
-46
Clo
ne
-56
Clo
ne
-59
Sta
t-1
Sta
t-2
Sta
t-3
Sta
t-4
Sta
t-5
Sta
t-6
8 m
M
6 m
M
4 m
M
2 m
M
1 m
M
% C
olo
ny s
urv
ival
in s
em
i-so
lid
med
iaGrowth in semi-solid media
17
► Improved colony survival obtained from C2 compared to DG44-2B
► Directed evolution approaches have a variable effect on colony survival
0
2
4
6
8
10
12
14
16
18
Clo
ne
11
Sort
5
Sort
4
1 m
M
Clo
ne
27
2 m
M
Sta
t 6
Clo
ne
56
Sort
6
6 m
M
Clo
ne
26
Sta
t 2
Sta
t 5
Sta
t 1
Clo
ne
59
DG
44
-2B
4 m
M
Clo
ne
7
8 m
M C2
Sta
t 4
Sta
t 3
Clo
ne
46
Clo
ne
36
Avera
ge R
an
kin
gAverage ranking used to select host cell lines
to progress
Top 5 cell
lines
progressed
18
Simultaneous comparison of host cell lines:
Summary
► For each screen, cell lines were ranked according to their performance
• The average ranking from all experiments gave an overall score for each
host cell line
► Top ranked host cell line isolated from cloning of parental C2 host
• Clone 11
► Not all cell lines identified performed better than parental C2 host
• 2 cell lines from the chemostat and cloning approaches ranked lower
► Performance of cell lines varied between the screens:
• The chemostat derived cell lines ranked towards the top when measuring
growth properties in shake-flasks
• In contrast, these cell lines performed poorly when assessing growth in
semi-solid medium, with only low numbers of colonies being produced
• Demonstrates importance of assessing multiple characteristics
► Top 5 host cell lines, Clone-11, Sort-5, Sort-4, 1mM and Clone-27,
were chosen to evaluate further in a ‘mini-pool’ assessment
(representative of early stage stable cell line generation)19