UCL - flutcore.eu

Report
Optimisation of the Primary Purification
Steps of a VLP production process
Olotu Ogonah
Benjamin Blaha
Tarit Mukhopadhyay
VLP Process Flow Chart – early phases
GSK Stevenage
1
Fermentation and
RNA removal
2
Cell harvesting
Centrifugation
3
Homogenisation
4
5
Clarification
Centrifugation/Normal flow filtration
0.45, and 0.2 m
Concentration/Diafiltration
HF TFF: NMWC 100kDA
Identification of the Critical Process Parameters
(CPP): parameters that impact process
performance.
Homogenisation
• The Lab40 is a validated scale down model to the
Manufacturing scale
– Use process knowledge and expertise to suggest critical
process parameters (CPP): parameter which impact process
behaviour (response) or critical product quality attributes
(CQA).
– Will confirm through model building.
• In these experiments the responses evaluated will be selected for
their potential impact on process efficiency and scalability; i.e.
soluble vlp recovered and DNA concentration.
Homogenisation screening experiment
design – model building.
# Homog. pass
•
•
•
MODDE software used to enhance the design of
screening experiments.
• These experiments will identify parameters
which have a statistically significant impact
on process efficiency.
Model VLP → GFP insert.
• GFP concentration will be used for a
surrogate for VLP/core concentration.
Parameter range
Name
Abbr.
Homogenisation
pass#
pass
pH
pH
•
•
Units
Settings
2, 3, 4
Precision
0.05
7.5, 8, 8.5
0.1
pressure
bar
bar
300 to 1000
10
detergent
% solids
Tri
%_sol
%
0.1 to 1
3 to 15
0.1
0.2
Responses: GFP concentration and DNA
concentration.
Create a model which describes process
behaviour.
4
4
3
3
3
4
2
4
3
3
3
4
4
2
2
3
2
3
3
2
2
2
4
4
3
3
2
2
2
4
2
4
2
3
GFP yield
flur.
Units
3.0 4159.9
15.0 1586.6
9.0 3993.3
9.0 4453.3
3.0 14426.6
15.0 3973.3
15.0 3499.9
3.0 12013.3
15.0 6186.6
15.0 2066.6
15.0 1586.6
15.0 3853.3
3.0 9199.9
15.0
959.9
15.0 2753.3
9.0 3726.6
3.0 14213.3
3.0 7106.6
3.0
953.3
3.0
226.6
3.0 4979.9
3.0 2208.4
3.0 4568.4
15.0
938.9
9.0 2815.7
9.0 2959.1
3.0 4907.7
15.0 5951.7
15.0
758.7
15.0 9099.1
15.0 2049.7
3.0 8303.1
3.0
948.8
9.0 3078.4
Pressure Detergent % solids
bar
%
(w/w)
pH
7.5
8.0
8.0
8.0
8.5
8.5
8.5
8.0
8.0
8.5
7.5
7.5
8.5
8.0
7.5
8.0
8.5
8.0
7.5
7.5
8.0
8.5
8.5
7.5
8.0
8.0
7.5
8.5
7.5
8.5
8.5
7.5
7.5
8.0
650.0
300.0
650.0
650.0
1000.0
650.0
650.0
1000.0
1000.0
300.0
650.0
1000.0
300.0
300.0
1000.0
650.0
1000.0
650.0
300.0
300.0
650.0
300.0
300.0
300.0
650.0
650.0
1000.0
1000.0
300.0
1000.0
300.0
1000.0
300.0
650.0
0.1
1.0
0.6
0.6
0.1
0.1
0.6
0.6
0.1
0.6
1.0
0.6
1.0
0.1
1.0
0.6
1.0
1.0
0.6
0.1
0.6
0.1
0.1
0.1
0.6
0.6
0.1
0.1
0.1
1.0
1.0
1.0
1.0
0.6
DNA
mg/l
24.1
29.0
37.0
46.1
43.4
59.6
52.2
46.9
107.4
23.5
31.5
122.4
31.3
16.4
102.9
54.6
43.2
24.5
7.7
5.3
18.0
Actual vs predicted plots (DNA and GFP)
Graphical presentation of model terms: The scale and centred
coefficients can be used to evaluate the significance of the model
terms.
Impact of homogenisation pass # on GFP yield
3 passes
•
3% load →1.74 X (V/V) than 15% loading.
•
3% load →8.75 X (g/g) than 15% loading.
• Re-evaluation of initial fermentation data
reveal low vlp yield was an artefact of the
homogenisation conditions used.
• Reducing loading from 15%→ 6% solids
increases soluble VLP → ~ 55%)
4 passes
Inter-fermentation analysis: VLPs
MeOH
30°C
(Lab40)
Mixed
20°C
(Lab40)
MeOH
20°C
(Lab40)
MeOH
20°C
(Soni.)
Fluorescence
6% load
48
S I
S I
S I
S I
21
Western blots (AB – HepB core protein)
Impact of # of homogenisation passes on DNA levels
3 passes
•
•
•
4 passes
High DNA
contamination.
Allowable DNA content
for final formulated
product ~ <10 ng/dose.
Data suggests need for
a DNA removal step.
The story so far……..
•
Homogenisation:
–
–
–
–
•
↑pH, ↑pass#, and ↑pressure increases vlp release.
Low % solids (3%) gives better vlp release (8 fold (wt./wt.) increase) compared to current protocol.
↑ % Load, ↑homogenisation pass #, and ↑pressure increase total DNA released.
High genomic DNA contamination indicate DNA removal step will be necessary.
• Floculation
• Precipitation
• AIEX
Fermentation
–
–
–
3 batches run under different conditions.
Yields and quality (multiple bands on western) impacted by fermentation conditions.
• Repeating assays with 6% load during homogenisation ( instead of 15%) reinforces impact of
homogenisation conditions on yield.
GFP assay appears to overestimates soluble fraction compared to western.
• Assay measure total (free and vlp attached) GFP, hence amplified signal ?
• Need assays to distinguish between free and attached antigen (Octet ?). Use as measure of stability?
Next….. TFF
Akta Crossflow
What is Tangental Flow Filtration (TFF)?
•TFF - Fluid continuously sweeps the
membrane surface, recirculating the feed
stream across the membrane.
Advantages
•Minimizes clogging membrane pores (fouling) and promotes consistent,
long-term productivity.
•Units can be cleaned, stored, and re-used as needed; i.e. it is very cost
effective.
• Suitable for scale-up
•Two types
•Cassettes: High shear, High flux → Higher ∆P
•Hollow Fibre: Low shear, lower flux than cassettes. Suitable for
fouling feed streams.
Hollow Fibre vs Cassettes: Initial Flux vs TMP
Flux vs . T M P
Ru n 3 - 2 5 ml/m in
Ru n 3 - 3 5 ml/m in
Ru n 3 - 5 0 ml/m in
70
60
50
F l u x [ LM H ]
40
30
20
10
0
0 .2
0 .3
0 .4
0 .5
0 .6
0 .7
0 .8
TMP [ b a r]
0 .9
1 .0
1 .1
1 .2
Flux vs. T M P
Run 16 - 250 ml/min (8000/s)
60
Run 16 - 125 ml/min (4000/s)
Run 16 - 63 ml/min (2000/s)
55
50
45
40
F lux [ LM H ]
35
30
25
20
15
10
0.1
0.2
0.3
0.4
0.5
0.6
0.7
TMP [ bar]
0.8
0.9
1.0
1.1
1.2
1 .3
At large scale systems are run in constant flux
mode in linear (non fouling) region of the flux vs
TMP curve.
Cassettes
•
Steep Flux vs TMP curve
•
System difficult to control at lower TMP
because:
– Flux sensitive to perturbations in TMP.
– Requires high ∆P and permeate
pressures to maintain low TMP.
Hollow fibres:
•
Very flat flux vs TMP curve.
•
Easier to maintain steady state.
•
Lower  pressure required to achieve
acceptable permeate flow. Will tolerate
fouling streams.
Summary
•
Homogenisation:
–
–
–
–
•
↑pH, ↑pass#, and ↑pressure increases GFP (and by implication vlp) release.
Low % solids (3%) gives better vlp release: 8 fold (wt./wt.) increase over current protocol.
↑ % Load, ↑homogenisation pass #, and ↑pressure increase total DNA released.
High genomic DNA contamination indicate DNA removal step will be necessary.
TFF
–
Cassettes: Sensitivity to perturbations in critical process parameters suggests not suitable for scale up.
–
Early data suggests Hollow Fibre preferred route.
•
Next Steps
•
•
•
Begin fermentation optimisation
Complete parameter screening/optimisation of TFF step.
Evaluate additional dead end filtration step as a replacement for centrifugation during primary clarification of
homogenate.
Look at DNA removal step
Include Quality assays in evaluation of CPPs.
• OCTET – biacore without the flow! Kinetic and concentration data.
• TEM of vlp.
•
•
•
Define Critical Attributes (CQA) ?
Revised VLP Process Flow Chart – What the
future may look like.
GSK Stevenage
1
Fermentation and
RNA removal
2
Cell harvesting
Centrifugation
3
Homogenisation
4
5
6
Clarification
Normal flow filtration
1.2, 0.45, and 0.2 m
Concentration/Diafiltration
HF TFF: NMWC 100kDA
DNA capture
Anion exchange Chromatography
END
Identification of the Critical Process Parameters
(CPP): parameters that impact process
performance.
Summary of model statistics
Impact of Filter Type on ∆P at initial conditions
D e lta P vs . T M P
Ca sse tt Flo w 1
Ca sse tt Flo w 2
Ca sse tt Flo w 3
Ho llo w Fib re Ru n 1
Ho llo w Fib re Ru n 2
Ho llo w fib re ru n 3
1 .5 0
1 .2 5
D elt aP [ bar]
1 .0 0
0 .7 5
0 .5 0
0 .2 5
0 .0
0 .1
0 .2
0 .3
0 .4
0 .5
0 .6
0 .7
TM P [ b a r]
0 .8
0 .9
1 .0
1 .1
Note: as concentration factor increases P will also increase
1 .2
1 .3

similar documents