application note usd 3385 · the bioreactor – but also provides the ability to image multiple...
TRANSCRIPT
www.pall.com/biotech
Application Note USD 3385
The Pall Xpansion® Bioreactor Supports Human
Embryonic Stem Cell Derived Progenitor Cell
Growth to >1 Million Cells/cm2 and Proper
Differentiation to a Mature Cell Fate
2
Contents
1 Background Information ........................................................................................................................... 3
1.1 Pall Xpansion Bioreactor ............................................................................................................. 3
1.2 Progenitor Cell Project Overview................................................................................................. 4
2 Materials and Methods ............................................................................................................................. 5
2.1 Xpansion Bioreactor Harvest Condition Optimization ................................................................. 5
2.2 Xpansion Bioreactor Plate Coating ............................................................................................. 6
2.3 Cell Culture, Differentiation, Harvest, and Washing .................................................................... 6
3 Results and Discussion ............................................................................................................................ 7
3.1 Xpansion Bioreactor Harvest Condition Optimization ................................................................. 7
3.2 Cell Recovery Post-Harvest and Post-Washing .......................................................................... 9
3.3 Cell Growth, Cell Morphology, and Process Control ................................................................. 11
3.4 Cell Differentiation to a More Mature Cell Fate ......................................................................... 16
4 Conclusions ............................................................................................................................................ 17
www.pall.com/biotech 3
1 Background Information
1.1 Pall Xpansion Bioreactor
The Xpansion bioreactor is a closed, single-use, multi-plate bioreactor system designed to overcome
challenges associated with manufacturing clinical-grade cell therapies in traditional flatware formats, such as
t-flasks, multi-plate flasks, or roller bottles. These challenges include high facility overhead costs due to large
incubator and biological safety cabinet (BSC) footprints, high labor costs due to the number of manual
processing steps required, risks associated with open processing, and limited ability to monitor and control
culture conditions. The Xpansion bioreactor addresses these challenges by making a cell growth surface
area of up to 122,400 cm2 available in a footprint of 48 cm (L) x 40 cm (W) x 90 cm (H), reducing the number
of manual processing steps, and incorporating weldable tubing to eliminate the need for open processing.
Finally, the bioreactor is integrated with single-use pH and dissolved oxygen (DO) sensing technology to
monitor culture conditions. The Xpansion bioreactor is paired with the Pall mPath™ control tower, which
uses a state-of-the-art programmable logic controller (PLC) and mass flow controllers (MFC) to deliver a
custom 4-gas blend – and optional secondary gas blend – to the bioreactor to control pH and DO at the
required set points.
The Xpansion bioreactor is available in 4 different sizes: 10-layer (XPN10), 50-layer (XPN50), 100-layer
(XPN100), and 200-layer (XPN200). The surface area of each layer is 612 cm2, approximately equal to the
surface area of one layer in a traditional cell factory. Each layer is manufactured using crystal, medical-grade
polystyrene by an injection molding process. After manufacturing, the plates are “tissue culture treated” via a
vacuum plasma hydrophilization process, ensuring that the Xpansion provides a cell attachment surface
comparable to traditional flatware. Each plate is circular in shape, with a center cut-out and radial slits
uniformly spaced to allow for media flow between plates (Figure 1a). When the vessel is assembled, these
plates are stacked 1 mm apart such that the center cut out creates a central column, which contains a
silicone tubing coil (Figure 1b). The slits are offset from the plate above and plate below, forcing cell culture
media to cover the entire surface of a plate before rising to the plate above (Figure 1c). This enhances the
mixing properties. The 1 mm spacing not only reduces the overall height – and therefore space occupied by
the bioreactor – but also provides the ability to image multiple plates using a specialized microscope.
Figure 1
Schematic diagrams of the Xpansion bioreactor
4
A custom 4-gas blend is supplied through the silicone coil, which is used to control pH and dissolved oxygen
(DO) via bubble-free oxygen diffusion from the coil into the cell culture media and carbon dioxide diffusion
from the media into the coil or vice versa. Beneath the bottom plate is a stir bar chamber (Figure 1d). When
bioreactor stirring is activated, this stir bar begins to recirculate the media over the plates from bottom to top
(Figure 1e), following the path of the staggered radial slits. This linear media speed can be precisely
controlled to ensure good bioreactor mixing, while also limiting shear for fragile or loosely adhered cells.
When the media reaches the top of the vessel, it encounters the single-use pH and DO sensors, before
returning to the bottom via the central column, where the media comes into contact with the gas blend in the
silicone coil. Finally, the bioreactor is integrated with several weldable tubing assemblies to help facilitate
liquid transfer to and from the vessel, including aseptic sampling.
1.2 Progenitor Cell Project Overview
Pall partnered with one of its cell therapy customers1 to evaluate the use of Xpansion bioreactor to amplify
and differentiate a progenitor cell for clinical manufacturing. This process begins with embryonic stem cells
derived from a single donor. The cells undergo an initial amplification and differentiation phase to produce
progenitor cells. Following the initial amplification, the progenitor cells are differentiated to a more mature cell
fate. The mature cell fate is indicated by the positive expression of cell surface markers, which are also
thought to correlate with the therapeutic efficacy of the cells. The mature cells are then washed, sorted, and
cryopreserved before administration to patients. The final amplification and differentiation phase was the
focus of the Xpansion bioreactor evaluation.
Historically, this final differentiation phase was performed in T-150 flasks using one of two different protocols.
The protocols are distinguished by the media exchange frequency and the composition of the mixture of
small molecules administered with each media exchange to facilitate proper cell differentiation. The first
protocol, referred to as the “22-day” protocol, generally results in greater cell amplification, but less overall
differentiation to the more mature cell fate. The second protocol, referred to as the “16-day” protocol, by
comparison, results in less cell amplification, but more differentiation to the mature cell fate, as measured by
the percentage of cells expressing the cell surface markers.
Implementation of the Xpansion bioreactor for the final phase of this process initially had some challenges.
The first challenge was adapting the legacy flatware coating protocol to ensure the Xpansion plates were
properly coated with a proprietary protein solution. This solution plays a critical role in facilitating enhanced
cell binding and later, proper cell differentiation. Once coated, the Xpansion bioreactor was seeded with
progenitor cells at densities greater than 1 million cells/cm2. Cells were cultured for 3 – 4 weeks to provide
enough time for amplification and differentiation to occur. This cell density and cell culture duration placed
high demands on the bioreactor to maintain favorable cell culture conditions, most notably the oxygen flow
rate needed to maintain DO. Once cell culture is complete, cells would need to be recovered from the
bioreactor by adapting the legacy flatware harvest protocol to the Xpansion bioreactor. Since the media
volume above each plate is different in the Xpansion bioreactor compared to legacy flatware, the abundance
per surface area of any added dissociation enzyme would also be different, potentially impacting cell
detachment from the bioreactor plates. The final challenge was to develop a process to handle, wash, and
sort the large volume of cell suspension harvested from the Xpansion bioreactor.
1 This partner has asked that their company name, cell type, disease indication, and some process conditions not be disclosed publicly
www.pall.com/biotech 5
To measure the success of the Xpansion bioreactor evaluation, several criteria were established. These
criteria were evaluated for eight Xpansion bioreactor batches, with Pall personnel providing on-site support
for the first two batches.
1. Differentiation: Surface marker expression in greater than 90% of the cell population
2. Amplification: Minimum 2-fold, put preferably 4-fold. The best case, or “gold standard,” was a
6-fold amplification using the 22-day differentiation protocol (total recovery of 5 million cells/cm2).
3. Viability: >95% post-wash
4. Morphology: Cells in the Xpansion bioreactor self-assemble into a monolayer with characteristic
cell colony formation
2 Materials and Methods
2.1 Xpansion Bioreactor Harvest Condition Optimization
To determine the optimal conditions to harvest cells from the Xpansion bioreactor, a series of screening
studies were performed in T-150 flasks utilizing a Pall-provided protocol. In short, cells were seeded on the T-
150 flasks using parameters reflective of the most current cell culture process. Cells were cultured for the
normal duration, with daily sampling to measure pH. These pH measurements were used to determine the
pH set point to use in the Xpansion bioreactor. After culture, flasks were harvested using different conditions.
In the positive control, the existing harvest process was performed without alteration. For the other
experimental conditions, cells were exposed to various concentrations of TrypLE♦ (Gibco) at ambient
temperature for 50 minutes, as indicated in Table 1. After harvest, recovered cells were counted and imaged
to determine the extent of aggregation.
Table 1
Design parameters of the first harvest optimization series in T-150 flasks
Condition TrypLE Concentration (%) Incubation Temperature (°C) Incubation Time (min)
Positive control 100 37 35
Test A 100 21 50
Test B 50 21 50
Test C 25 21 50
After results from the first optimization series were obtained, a second set of experiments was performed to
optimize the TrypLE incubation time and temperature. Cells were again seeded on the T-150 flasks using the
parameters reflective of the most representative cell culture process. Cells were cultured for the normal
duration, and were harvested using a panel of different conditions, as indicated in Table 2. Since the 100%
TrypLE condition performed best during the first series of optimization studies (results below), all harvest
conditions for this optimization series were performed in 100% TrypLE. After harvest, recovered cells were
counted and imaged to determine the extent of aggregation.
6
Table 2
Design parameters of the second harvest optimization series in T-150 flasks
Condition Name Incubation Time (min) TrypLE Pre-Warm Temperature (°C)
Incubation Temperature (°C)
Test A1 55 37 21
Test A2 55 37 37
Test A3 55 40 21
Test A4 55 40 37
Test A5 80 37 21
Test A6 80 37 37
Test A7 80 40 21
Test A8 80 40 37
2.2 Xpansion Bioreactor Plate Coating
For the T-150 flasks, 50 mL of 1x protein coating solution was prepared per flask, added to the flask, and
incubated overnight. Following incubation, the protein solution was decanted, and the flask was rinsed three
times with phosphate-buffered saline solution (PBS). This resulted in a protein coating solution volume to
surface area ratio of 0.33 mL/cm2 in the flask. In the Xpansion bioreactor, the volume to surface area ratio
above each plate is fixed at 0.16 mL/cm2. To ensure an equal abundance of protein molecules per surface
area, the protein concentration in the solution used to coat the Xpansion plates was increased to 2x. Like the
T-150 flasks, the Xpansion bioreactor was incubated overnight with the protein solution and then triple rinsed
with PBS following the incubation.
2.3 Cell Culture, Differentiation, Harvest, and Washing
For the first Xpansion bioreactor batch (Run 1), the 22-day differentiation protocol was tested. Cells were
seeded at 0.75 million/cm2 in 1.6 L of media. Media exchanges were performed every 2 – 3 days, with each
media exchange consisting of a different regimen of small molecules aiding in cell differentiation. The media
exchange frequency was increased to once daily later in the culture to better maintain dissolved oxygen. The
media pH was maintained at a set point of 7.35 with one-sided (CO2) control only, while the DO was
maintained at a set point of 50% air saturation via O2 addition and increased bioreactor agitation.
Temperature was maintained by placing the vessel in a 37 °C incubator. To harvest, spent media was
drained from the vessel. Next, the bioreactor was incubated in 1x TrypLE for 55 minutes at 37 °C. Following
incubation, the resulting cell suspension was decanted and collected for cell washing and sorting. No
mechanical agitation was performed. After collection, the bioreactor was refilled with fresh 1x TrypLE and
incubated for 10 minutes at 37 °C. This cell suspension was also collected without agitation. The resulting
TrypLE cell suspension collections were separately aliquoted into eight 250 mL centrifuge tubes and
quenched with cell culture media. Cells were pelleted, resuspended in fluorescence-activated cell sorting
FACS buffer, and collected through a 100 μm cell strainer such that the two initial collection pools were re-
established. Cells in each collection were diluted to 500 mL before counting.
For the second Xpansion bioreactor batch (Run 2), the 16-day differentiation protocol was tested. Cells were
seeded at 1.0 million/cm2 in 1.6 L of media. Media exchanges were performed every 2 – 3 days, with each
media exchange consisting of a different regimen of small molecules aiding in cell differentiation. The media
pH was maintained at a set point of 7.35 with one-sided (CO2) control only, while the DO was maintained at a
set point of 50% air saturation via O2 addition and increased bioreactor agitation. Temperature was
maintained by placing the vessel in a 37 °C incubator. To harvest, spent media was decanted from the
vessel. Attached cells were briefly and gently washed with fresh PBS, which was decanted and discarded.
Next, the bioreactor was incubated in 1x TrypLE + DNAse for 55 minutes at 37 °C. Following incubation, the
www.pall.com/biotech 7
bioreactor was mechanically agitated by implementing ~60 “strikes” with the Xpansion harvest station.
The resulting cell suspension was decanted and collected for cell washing and sorting. After collection, the
bioreactor was washed with fresh media to collect any biomass hold up. This media wash was pooled with
the previous cell suspension and an additional 3 L of fresh media, diluting the collected cell suspension to a
total volume of ~6 liters. This 6 L cell suspension was processed with an automated cell washer to
concentrate and exchange cells into buffer in preparation for the CliniMacs♦ assay (Miltenyi Biotech).
For both batches, a series of T-150 control flasks were seeded in 39 mL media at a cell density equivalent to
the Xpansion bioreactor. After the first media exchange, the working volume in one set of flasks was
increased to 90 mL in each flask while in a second set, the volume was maintained constant at 39 mL
throughout the culture to mimic the media volume to surface area ratio used in the Xpansion bioreactor.
These control flask sets were respectively referred to as the high-volume and low-volume controls.
Subsequent media exchanges were performed at a frequency equivalent to the Xpansion bioreactor,
consisting of the same regimen of differentiation-inducing small molecules. pH and DO control was not
possible in the flasks, although the flasks were incubated in a 5% CO2 incubator at 37 °C. To harvest, media
was decanted, and cells were incubated in 1x TrypLE for 35 minutes at 37 °C. The TrypLE was quenched
with equal volume media and centrifuged to pellet the cells. Cells were then resuspended and collected
through a 100 μm cell strainer before counting.
3 Results and Discussion
3.1 Xpansion Bioreactor Harvest Condition Optimization
Cells from this study were counted to determine total recovery and imaged to determine the extent of
aggregation. Cell recovery results are displayed in Table 3 and images of cell aggregates are displayed in
Figure 2.
Table 3
Cell recovery results from Series I screening studies in T-150 flasks
Condition Name 1x TrypLE Concentration (%)
Incubation Temperature (°C)
Incubation Time (min)
Cell Recovery (millions)
Positive control 100 37 35 370
Test A 100 21 50 158
Test B 50 21 50 120
Test C 25 21 50 86
None of the three test cases resulted in as high of a cell recovery or as uniform of a single cell suspension as
the baseline control, indicating that there was room for improvement by optimizing other harvest parameters.
Of the three test cases, Test A (100% TrypLE) resulted in the highest cell recovery and greatest number of
single cells, so 100% TrypLE was maintained for all future testing.
8
Figure 2
Cell aggregation results from Series I screening studies in T-150 flasks. Cell clumping observed for the baseline harvest process (left), 50% TrypLE (center), and 100% TrypLE (right)
A second series of harvest conditions were screened, this time incorporating additional parameters such as
incubation time and temperature. Cells from this study were counted to determine total recovery. Results are
displayed in Table 4.
Table 4
Cell recovery results from Series II screening studies in T-150 flasks
Condition Name
Incubation Time (min)
TrypLE Pre-Warm Temperature (°C)
Incubation Temperature (°C)
Cell Recovery (millions)
Cells in Aggregation (%)
Test A1 55 37 21 141 15
Test A2 55 37 37 270 22
Test A3 55 40 21 90 14
Test A4 55 40 37 135 2
Test A5 80 37 21 121 57
Test A6 80 37 37 252 11
Test A7 80 40 21 99 47
Test A8 80 40 37 252 21
These results clearly suggest that incubation at 37 °C results in both higher cell recovery and less cell
aggregation than incubation at ambient temperature. Therefore, the optimized conditions for the first
Xpansion bioreactor harvest were chosen to be a 37 °C in 100% 1x TrypLE. The incubation time and TrypLE
prewarm temperature did not appear to have a major impact on harvest results, so 55 minutes was selected,
as it was the shorter of two conditions, and 40 °C prewarm temperature, as this was thought to lead to more
robust temperature control, assuming some temperature drop during the process of transferring TrypLE from
the prewarmed container to the Xpansion bioreactor.
www.pall.com/biotech 9
3.2 Cell Recovery Post-Harvest and Post-Washing
Table 5
Run 1 Xpansion bioreactor cell harvest post-washing results
Collection
Total Cell Recovery
(billions)
Viable Cell Density (million/cm2)
Viability
(%)
Fraction of
Total Yield (%)
Xpansion spent media 1.8 0.3 80 8
Xpansion TrypLE collection #1 20 3.2 96 86
Xpansion TrypLE collection #2 1.4 0.2 92 6
Xpansion Total 23.2 3.7 NA1 100
Low volume control 1.3 2.9 93 100
High volume control 1.6 3.5 99 100 1Viability for the pooled cells was not measured
The results in Table 5 reflect Run 1 cell numbers obtained after each cell collection was concentrated and
washed using protocols previously described. In total, the Xpansion bioreactor yielded 23.2 billion cells, or 3.7
million cells per cm2 - more than either of the control flasks. However, not all these cells were recovered in
the primary cell collection, TrypLE collection #1. About 8% of cells were lost when spent cell culture media
was decanted prior to harvest. Although undesirable, this result was not completely unexpected, given that
cells had been observed detaching from the Xpansion plates on Day 20 of culture. In addition, 6% of the cell
total was collected following the second TrypLE incubation, which indicated that the first collection did not
extract all the cells attached to the bioreactor. Optimizing this first TrypLE collection is important, as the small
number of cells recovered in the second collection does not justify the material cost and labor associated with
a second harvest.
Post centrifugation, the cells harvested from the Xpansion bioreactor were more aggregated than cells from
the control flasks. This could be due to the longer time required to process the larger harvest volumes via
centrifugation. Therefore, increased cell yield loss also occurred during cell straining performed just before
counting, as the cell aggregates were too large to pass through the strainer.
All the cell counts reported in Table 5 were performed with a Nucleocounter♦ NC200. It is worth noting that
multiple counts were performed on the TrypLE collection #1, which are detailed below:
1. The first NC200 count was 25 billion total cells.
2. Afterward, a manual Trypan Blue count was performed and yielded 35 billion total cells.
3. Due to this discrepancy between count #1 and #2, a second sample was prepared so these counts
could be repeated. During this time, cells started aggregating, and became more difficult to handle and
process. The second NC200 count was 15 billion total cells.
4. The second Trypan Blue count yielded 17 billion total cells.
The final cell count for this collection was reported at 20 billion cells, an average of the two NC200 counts,
but this was likely an under-count due to the cell aggregation that occurred between subsequent cell counts.
With faster processing times, or strategies to mitigate cell aggregation, 25 billion cells may have been
obtained. Viability for all counts from TrypLE collection #1 was greater than 95% and equivalent between the
Xpansion bioreactor and control flasks.
10
If the TrypLE collection #1 is considered as the primary cell product stream, then the 20 billion total cells, or
3.2 million cells/cm2, represents a 4-fold cell amplification over the original seeding density, meeting the
previously stated success criteria. With improvements targeted to address the aforementioned challenges of
premature cell detachment, bioreactor harvest, cell clumping, and large volume cell handling, this batch could
have resulted in a total recovery of 28.2 billion cells (25 billion from TrypLE collection #1), or 4.5 million
cells/cm2, a 6-fold amplification that would have exceed the gold standard.
Table 6
Run 2 Xpansion bioreactor cell harvest results pre- and post-washing
Product Stream
Total Cells
(billions)
Cell Density (million/cm2)
Viability
(%)
Recovery
(%)
Xpansion harvest (pre-wash) 23.5 3.8 98 NA
Xpansion washed cell product (automated) 23.3 3.8 99 99
Xpansion washed cell product (centrifuge) 16.2 2.6 98 69
Low volume control (post-wash) 0.6 3.9 98 NA
High volume control (post-wash) 0.5 3.6 98 NA
All Run 2 Xpansion bioreactor results reported in Table 6Table 5 reflect cells obtained from the first TrypLE
collection, also referred to as primary product collection. Cells from the spent media or subsequent bioreactor
washes were not processed. As with Run 1, the total cells recovered per cm2 from the Xpansion bioreactor
was equivalent to or higher than the control flasks. The 23.5 billion total cells, or 3.8 million cells/cm2,
represents a 3-fold cell amplification over the original seeding density, meeting the desired minimum
objective of a 2-fold cell amplification. Viability for all cells was greater than 95% and equivalent between the
Xpansion bioreactor and control flasks.
Notably, 99% of the cells harvested from the Xpansion bioreactor were recovered following an improved cell
washing protocol implemented for this batch, utilizing a proprietary automated cell washer. This yield was
extremely favorable in comparison to the 69% cell recovery obtained when these cells were processed with
the legacy centrifugation and cell straining protocol previously described.
A summary of all eight Xpansion bioreactor batch yields executed for this project can be found in Figure 3. In
this figure, the cell recovery from the Xpansion bioreactor was calculated from the primary TrypLE collection
only, and compared to the cell recovery from the control flasks on a surface area basis. The number above
the bar indicates the fold amplification of cells in the Xpansion bioreactor. In summary, cells cultured from the
Xpansion bioreactor reached densities similar to the control flasks in a reproducible manner. When
interpreting this data, it is worth noting that these experiments were conducted within the scope of process
development exercise, so no two experiments were identical. For instance, the initial amplification and
differentiation protocol used to produce progenitor cells was modified for Run 3. In addition, despite the
success of the automated cell washer, different cell washing strategies were evaluated after Run 2. These
strategies were not as successful in reducing cell aggregation, which accounts for the slightly lower Xpansion
bioreactor harvest cell density observed for some batches.
www.pall.com/biotech 11
Figure 3
Summary of eight Xpansion bioreactor batch cell recoveries
3.3 Cell Growth, Cell Morphology, and Process Control
The Xpansion bioreactor is an adherent cell bioreactor, so daily cell counts to assess growth are not possible
without terminating the culture. It was also difficult to use metabolites as a surrogate indicator of cell growth,
due to the frequency of media exchanges. Instead, periodic images were obtained from both the Xpansion
bioreactor and control flasks to assess the cell growth qualitatively. Images from one representative batch
(Run 1) are displayed in Figure 4.
Figure 4
Images of cell growth and morphology in the Xpansion bioreactor and control flasks
Xpansion Bioreactor Control
Day 0
N/A
4x
3.2x
1.8x
2.1x
2.8x3x
1.6x
2.6x
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
1 2 3 4 5 6 7 8
Har
vest
cel
l des
nit
y (m
illio
n/c
m2)
Xpansion High Volume Control Low Volume Control
12
Day 1
Day 4
Day 12
Day 18
www.pall.com/biotech 13
Day 22
These images illustrate similar and expected cell growth in both the Xpansion bioreactor and control flasks.
On the day of cell seeding, after attachment, cells appear uniformly distributed on the plates. By 24 hours
post-seeding (Day 1), cells begin to migrate into small colonies. Over the next 12 days of growth, these
colonies began to grow denser, and a new cell monolayer was generated between the colonies. By Day 22,
the cell colonies also formed dense cell bridges between them, atop a robust cell monolayer. This
morphology indicated proper cell differentiation and assembly into appropriate 3D structures suitable for
harvest.
During the course of all bioreactor batches, pH and DO were both monitored and controlled to maintain a
favorable growth environment. During Run 1, the control tower struggled to maintain DO at the desired set
point, as indicated by the drop in DO measurement on Day 7 displayed in Figure 5.
Figure 5
Xpansion bioreactor Run 1 DO and oxygen flow trend (Day 0 – 11)
0
50
100
150
200
250
300
350
400
450
500
0
20
40
60
80
100
120
0 2 4 6 8 10
O2
flow
(m
L/m
in)
DO
(%
), S
tirr
ing (
rpm
)
Days
DO Process Value DO Set Point O2 Process Value
14
To improve DO control, we evaluated several possible countermeasures and the consequences of each:
1. Do nothing: Since the effect of different DO set points on the cells had not been previously studied,
it is possible that the cells were well suited to the hypoxic environment seen post-Day 7. However, if
they were not, then the resulting hypoxic environment could have a deleterious effect on growth and
differentiation.
2. Increase oxygen flow: This would be the best way to increase oxygen transfer to the cell culture
media without increasing shear via faster agitation. However, the maximum flow rate of oxygen
through the Xpansion bioreactor silicone coil had not been previously evaluated, and there were
concerns that high pressure through the narrow tubing could cause it to burst.
3. Increase the stirring speed: Past work has demonstrated stirring speed to have the largest impact
on increasing oxygen transfer to the media; however, it also would have the largest impact on
increasing shear.
4. Increase media exchange frequency: This would be the overall least risky method to increase
oxygen transfer, as it would not result in increased shear or the compromising of the silicone coil.
However, it would also be the least effective method, as the DO would temporarily rise because of
the highly aerated media but immediately fall to pre-media exchange levels in a couple of hours.
Media exchange is also the most labor intensive and most costly method to raise DO.
For Run 1, each of the above strategies was evaluated in a step-wise fashion, as indicated by Figure 6.
Figure 6
Xpansion bioreactor Run 1 DO, oxygen flow, stirrer rate trends (Day 11 – 22)
0
50
100
150
200
250
300
350
400
450
500
0
20
40
60
80
100
120
10 12 14 16 18 20 22
O2
flow
(m
L/m
in)
DO
(%
), S
tirr
ing (
rpm
)
DaysDO Process Value DO Set Point Stirrer Process Value O2 Process Value
1 2
3
4
5
www.pall.com/biotech 15
A summary of the actions taken is below:
1. Day 11: The stirring speed (purple line) was increased, resulting in a temporary, but not sustained,
improvement in oxygen transfer and DO maintenance.
2. Day 12: The media exchange frequency was increased to daily. This also resulted in a temporary,
but not sustained, improvement in oxygen transfer and DO maintenance.
3. Day 13: The maximum oxygen flow rate (green line) was increased to 90 mL/min. This did not seem
to have an appreciable. impact.
4. Day 20: Some cells were observed detaching from the Xpansion plates, a phenomenon not totally
understood but initially attributed to the higher stirring speed. As a result, media exchanges were not
performed the remaining two days of culture, and the bioreactor harvest protocol was modified to
exclude the first PBS wash, for fear this wash would result in cell loss.
5. Day 21: The DO measurement increased. At the time, this observation was interpreted to be due to
cell death or reduced metabolic activity of the existing cells, possibly due to sustained hypoxic
conditions.
During Run 2, the bioreactor maintained DO at the desired set point for the duration of the batch, as indicated
in Figure 7. The cells cultured in Run 2 did not demand as much oxygen – and as a result, did not challenge
the Xpansion bioreactor’s oxygen supply capabilities like Run 1 – because they were seeded at a lower
density, cultured for a shorter duration, and cultured with a protocol believed to result in better cell
differentiation at the expense of proliferation. It is worth noting that the deviations between DO set point and
process value observed on Day 10 and Day 12 resulted from operator error, as oxygen supply had not been
restored following a media exchange.
Figure 7
Xpansion bioreactor Run 2 DO, stirring speed, and oxygen flow trends
0
50
100
150
200
250
300
350
400
0
20
40
60
80
100
120
0 2 4 6 8 10 12 14 16
O2
flow
(m
L/m
in)
DO
(%
), S
tirr
ing s
peed (
rpm
)
Days
DO Process Value DO Set Point Stirring Speed O2 Process Value
16
3.4 Cell Differentiation to a More Mature Cell Fate
Cell differentiation was assessed by measuring the expression of positively expressed on-target surface
markers (SM1 & SM3) and transcription factors (TF1 – TF4), while also measuring the absence of off-target
surface markers (SM2) and transcription factors (TF5 – TF9). The percent of the cell population expressing
surface markers was measured via florescence activated cell sorting (FACS), while the percent of the cell
population expressing transcription factors was measured via an immuno-cytochemistry (ICC) assay. Four
populations of cells were measured: unsorted (us) cells harvested from the high volume control flasks (T-150
reference-us), unsorted cells harvested from the low volume control flasks (T-150 XP1 control-us), unsorted
cells harvested from the Xpansion bioreactor (XP10-us), and cells harvested from the Xpansion bioreactor
positively (pos) sorted for SM1 (XP10-pos). Differentiation results from four randomly selected batches are
displayed in Figure 8.
Figure 8
Cell surface marker and transcription factor expression in washed cell product
In summary, these results suggest that the desired cell differentiation was achieved. Arguably the most
important surface marker, SM1, was expressed in greater than 90% of the cell population harvested from
Xpansion bioreactors, a fraction greater than or equal to the cells harvested from the control flasks. This
same observation was made for the other on-target surface marker (SM3) and all of the on-target
transcription factors. In addition, off-target surface markers and transcription factors were either not
expressed in cells harvested from the Xpansion bioreactor, or if they were, they were at levels equal to or
below those levels expressed by cells harvested from the control flasks.
Finally, although it was not possible to directly asses the effectiveness of the Xpansion plate coating protocol,
the high cell recovery and effective cell differentiation indirectly suggests that plate coating was successful,
otherwise cell differentiation to this extent would not have occurred.
www.pall.com/biotech 17
4 Conclusions
The evaluation of the Xpansion bioreactor was a success, as all of the criteria were met or exceeded:
1. Differentiation: On-target SM1 expression exceeded the stated goal of expression in greater than
90% of the cell population, equal to or greater than the expression in the control flasks, indicating that
cells cultured in the Xpansion bioreactor differentiated into the desired mature cell fate.
2. Amplification: Cells harvested from the Xpansion bioreactor routinely met the minimum desired 2-
fold amplification. The best case, or “gold standard,” 6-fold amplification is also possible with
improvements to the cell harvest and washing processes.
3. Viability: The goal of >95% cell viability was met or exceeded for each Xpansion bioreactor harvest.
4. Morphology: In the Xpansion bioreactor, cells self-assembled into a characteristic monolayer with
cell colony formation, equivalent to the flatware controls.
In addition to the stated objectives, the following milestones were also achieved:
1. The Xpansion bioreactor plates were successfully coated with a proprietary protein critical to, as
inferred by the successful cell attachment, high density cell harvest and cell differentiation into a
more mature cell fate.
2. The Xpansion bioreactor can support some adherent human cell culture at densities of >1 million
cells/cm2 for extended culture durations.
3. T-flasks can be used to optimize several harvest parameters and the optimized parameters translate
successfully when transferred to the Xpansion bioreactor.
4. Compared to centrifugation, the implementation of an automated cell washing process was more
effective at concentrating and washing larger volumes of cell suspension, reducing yield loss due to
the formation of cell aggregates.
In summary, the closed, single-use, multi-plate Xpansion bioreactor enabled this Pall customer to overcome
challenges associated with the clinical manufacturing of a sensitive progenitor cell type in traditional flatware.
18
Filtration. Separation. Solution.SM
Visit us on the Web at www.pall.com/biotech
Contact us at www.pall.com/contact
Corporate Headquarters
Port Washington, NY, USA
+1.800.717.7255 toll free (USA)
+1.516.484.5400 phone
European Headquarters
Fribourg, Switzerland
+41 (0)26 350 53 00 phone
Asia-Pacific Headquarters
Singapore
+65 6389 6500 phone
International Offices
Pall Corporation has offices and plants throughout the world in locations such as: Argentina, Australia, Austria,
Belgium, Brazil, Canada, China, France, Germany, India, Indonesia, Ireland, Italy, Japan, Korea, Malaysia, Mexico,
the Netherlands, New Zealand, Norway, Poland, Puerto Rico, Russia, Singapore, South Africa, Spain, Sweden,
Switzerland, Taiwan, Thailand, the United Kingdom, the United States, and Venezuela. Distributors in all major
industrial areas of the world. To locate the Pall office or distributor nearest you, visit www. Pall.com/contact.
The information provided in this literature was reviewed for accuracy at the time of publication. Product data may
be subject to change without notice. For current information consult your local Pall distributor or contact Pall directly.
© 2020 Pall Corporation. The Pall logo, Pall, mPath and Xpansion are trademarks of Pall Corporation. ® indicates a trademark registered in the USA and TM indicates a common law trademark. Filtration. Separation. Solution is a service mark of Pall Corporation.
♦ CliniMacs is a trademark of Miltenyi Biotec B.V. & Co. KG, Nucleocounter is a trademark of ChemoMetec A/S Corporation, and TrypLE is a trademark of Life Technologies Corporation