APPLICATION NOTE No.

472

Increasing iPSC Numbers through Systematic Culture

Process Optimization
Felix Manstein1,2, Kevin Ullmann1,2, Robert Zweigerdt1,2
1
Leibniz Research Laboratories for Biotechnology and Artificial Organs (LEBAO), Germany
2
REBIRTH Cluster of Excellence, Hannover Medical School, Hannover, Germany

Contact: [email protected]

Abstract

Human induced pluripotent stem cells (hiPSCs) are a maintain optimal growth conditions for the cell of choice.
powerful tool for innovative approaches, such as drug In this study, DASbox® Mini Bioreactors and the DASbox
discovery, in vitro disease modelling, or regenerative Mini Bioreactor System were utilized to systematically
therapies. However, such procedures require high cell optimize process parameters of a hiPSC culture in a
numbers to be sufficiently applicable, a criterion that is step-by-step process. This approach led to a more than
hard to satisfy with traditional 2D culture methods. 10× increase in cell density (almost 35 × 106 cells/mL)
Stirred-tank bioreactors on the other hand offer a 3D compared to uncontrolled conditions while stem cell
culture environment suitable to provide, control, and features and viability were retained.

Introduction

Together with embryonic stem cells (ESC), induced for neurodegenerative diseases or heart failure [5].
pluripotent stem cells (iPSC) constitute the group of In that context, one of the most important aspects of
pluripotent stem cells (PSC). Both cell types have the ability stem cell therapy is achieving high, clinically relevant cell
to grow indefinitely and to develop into any cell lineage of numbers. Estimates for PSC therapies suggest cell doses of
the body [1, 2]. However, while ESCs are naturally occurring, 109 cells for treatment of a single patient and production of
iPSCs originate from reprogramming somatic cells in vitro by 1011 to 1014 cells per year for a single product [6].
delivering specific factors [3]. Therefore, they can be created Such numbers cannot be sufficiently produced by
as needed. This dynamic availability paired with their stem conventional 2D monolayer culture methods in shake
cell properties make iPSCs a powerful tool for innovative flasks or dishes [7]. The limited growth surface combined
approaches, such as drug discovery and in vitro disease with the possibility to grow in only two directions can
modelling [4]. lead to undesired cell proliferation behavior, poor cell
Furthermore, clinical trials are under way, opening the differentiation, improper gene and protein expression
doors towards regenerative therapies, such as tissue repair patterns, and an overall inaccurate representation of the
APPLICATION NOTE I No. 472 I Page 2

Disclaimer
All methods and results in this application note are published
A
work by the group of Dr. Robert Zweigerdt, Leibniz Research
Laboratories for Biotechnology and Artificial Organs, Hannover
Medical School, Germany (doi: 10.1002/sctm.20-0453) [9],
© 2021 The Authors. Stem Cells Translational Medicine
published by Oxford University Press. This application note
was produced with permission of Oxford University Press
in accordance with the CC-BY-NC License of the original
publication.
We thank Dr. Zweigerdt for the permission to share these data.

3D environment the cells originate from [8]. On top of that,

individual growth parameters, such as pH, dissolved oxygen B
(DO), culture feeding, or medium perfusion cannot be
controlled in 2D cultures.
Bioreactors on the other hand offer an efficient way to
mimic the natural habitat of stem cells more closely by
providing a 3D setting and precise growth parameter control
through bioprocess control units. Furthermore, scalability,
the process of increasing the bioprocess dimensions, enables
increased yields by transitioning to pilot and production
scale once the optimal growth conditions are determined in
small scale experiments.
The study described in this application note details such
an optimization process for establishing human iPSC (hiPSC)
culture at small scale level. Using DASbox Mini Bioreactors
and the DASbox Mini Bioreactor System, multiple important
growth parameters were systematically adapted in a step- Figure 1: (A) The DASbox Mini Bioreactor is equipped
by-step process in order to increase cell numbers during a with an 8-blade impeller (60° pitch) specifically designed
7-day-incubation period , as recently published [9,10]. The for the culture of shear-sensitive cells in suspension or as
precise control of pH, nutrient feeding, DO, and agitation by aggregates. (B) The DASbox Mini Bioreactor System is
the DASbox Mini Bioreactor System in combination with the designed as a 4-fold system able to handle up to 24 parallel
perfusion operation mode and pre-culture optimization led operating bioreactors. It was developed for both cell culture
to an increase from initially ~3 × 106 cells/mL after 7 days of and microbial fermentation applications and can be used with
incubation to ~18 × 106 cells/mL. Additional implementation standard glass or single-use bioreactors.
of in silico-supported culture modelling increased the cell
numbers even further to an unmatched cell density of
almost 35 × 106 cells/mL. During this optimization process, Learn more about of the DASbox Mini Bioreactor
the 60° pitched 8-blade impeller of the DASbox Mini System please visit: www.eppendorf.com/dasbox
Bioreactors, which is specifically designed for the culture of
shear-sensitive cells in suspension or as aggregates, ensured
excellent viability of the stem cells throughout the runs [11].
hiPSC cultured this way retained their stem cell properties,
as well as the ability to develop into tissues of the endo-,
meso-, and ectoderm.
APPLICATION NOTE I No. 472 I Page 3

Material and Methods

A B
I Uncontrolled conditions

Process optimization stages

II + pH control

III + Glucose feeding

IV + DO control

V + (Pre-)culture/DO optimization

VI + Perfusion/Feeding optimization

VII/VIII + in silico modelling

Figure 2: (A) Experimental procedure for iPSC culture parameter optimization in DASbox Mini Bioreactors under control of the
DASbox Mini Bioreactor System (Created with BioRender.com). (B) Overview of the process optimization steps and nomenclature.

The DASbox Mini Bioreactors and DASbox Mini

Bioreactor System
Stem cell culturing and culture optimization was performed Phoenix (HSC_ADCF_SeV-iPS2; MHHi001-A) and GMPDU_8
in DASbox Mini Bioreactors equipped with an 8-blade (CD34+hPBHSC_GMPDU_SeV-iPS8; MHHi008-A) cells, both
impeller (60° pitch) optimized for stem cell expansion [11], derived from CD34+ human cord blood hematopoietic stem
an overhead drive for agitation, pH and DO sensors, as well cells [16], [17]. GMPDU_8 cells were derived under GMP-
as temperature control in order to ensure precise regulation compatible conditions.
of critical process parameters (Figure 1A).
Perfusion operation mode was enabled by an outflow filter hiPSC pre-culture
device, allowing medium to flow in and out of the bioreactor Prior to the bioreactor experiments, the cells were expanded
while retaining the cells inside. By aligning the influx rate of as a feeder-free monolayer culture in flasks at 37°C with
fresh medium (feed) through a head plate port and the efflux 5% CO2, using Essential 8 (E8) medium (Thermo Fisher
rate of depleted medium (waste) through the outflow filter, Scientific®) supplemented with ROCK inhibitor (RI) Y-27623
the bioreactor volume was kept constant throughout the (10 μM) (Figure 2A). The medium was changed after 48
runs [12]. Medium flow was regulated via the pumps of the hours for a passaging interval of 3 days and after 48 and 72
DASbox Mini Bioreactor System (Figure 1B). hours for a passaging interval of 4 days. See Manstein et al.,
The system was also used to monitor and adjust process 2021 for the exact passaging protocol [9].
parameters, such as pH, feeding rate, or DO. Pump and
sensor calibration was performed according to the protocols Culturing hiPSCs in DASbox Mini Bioreactors under
by Kempf et al. [13] and Ackermann et al. [14]. control of the DASbox Mini Bioreactor System
All main experiments were performed as stem cell aggregate
Human induced pluripotent stem cell (hiPSC) lines culture without the use of scaffolds for cell attachment. In
The cell culture experiments were conducted using three order to inoculate the DASbox Mini Bioreactors, single-cell
different hiPSC lines: hHSC_1285i_iPS2 (MHHi006-A) suspension was achieved by detaching the cell monolayer
cells derived from hematopoietic stem cells [15], as well as via StemPro™ Accutase™ (Thermo Fisher Scientific) for
APPLICATION NOTE I No. 472 I Page 4

the hHSC_1285i_iPS2 and Phoenix cell lines, or by EDTA- Table 1: Parameters controlled by the DASbox Mini Bioreactor System
and the equivalent abbreviations used in this application note.
containing versene solution in case of the GMPDU_8 cells.
Process optimization steps Abbreviations used in text
Inoculation was performed with a viable cell density of
None I
5 × 105 cells/mL E8 medium + RI to a final volume of 150 mL
pH 7 II
(75 × 106 cells/bioreactor) (Figure 2A). pH 7, glucose feeding III
Bioreactor cultivation conditions were 37°C with agitation pH 7, glucose feeding, DO control IV
speeds of either 60, 80, 100, or 120 revolutions per minute pH 7, glucose feeding, optimized pre- V
(rpm), and headspace gassing with 3 standard liters (sL) per culture, optimized DO control, 80 rpm
hour with 21% O2 and 5% CO2. pH 7, optimized pre-culture, optimized VI
If applied, pH control was initiated once the pH within the DO control, 80 rpm, optimized
bioreactor reached ≤7.0. Hereafter, it was maintained at pH perfusion, optimized glucose feeding
7.0 first by reducing CO2 in the gas stream and afterwards by pH 7, optimized pre-culture, optimized VII
DO control, 80 rpm, optimized
addition of 1 M NaHCO3.
perfusion, optimized glucose feeding,
Medium was not changed during the first 24 hours after in silico modelling 1
inoculation. If not stated otherwise, the medium was then pH 7, optimized pre-culture, optimized VIII
constantly replaced with E8 without RI at a flow rate of DO control, 80 rpm, optimized
150 mL/day (one culture volume/day or 6.25 mL/hour) while perfusion, optimized glucose feeding,
the cells were retained via the cell retention device. in silico modelling 2
During the optimization process, different compounds,
such as glucose (either 3.15 - 6.15 g/L from day 3 or 6.15- 100 ng/mL of the transcription factor activin A and 3 μM
7.65 g/L from day 4 onwards) or the shear force-reducing of the GSK3beta inhibitor CHIR99021 in Erlenmeyer flasks
surfactant Gibco™ Pluronic™ F-68 (Thermo Fisher (20 mL working volume).
Scientific) were added to the medium. Precisely 24 hours later the medium was replaced
The exact parameters of pH control, perfusion rate, or by 100 ng/mL activin A and 0.8% KnockOut Serum
compound addition will be detailed within the respective Replacement (Thermo Fisher Scientific) in RPMI 1640
results section. All the described processes were conducted (Thermo Fisher Scientific).
for 7 days in the bioreactor. An overview of the optimization Again 24 hours later the medium was replaced by
steps is given in Figure 2B and Table 1. 100 ng/mL activin A and 8% KnockOut Serum Replacement
in RPMI 1640. On day 3 of differentiation, aggregates were
Cell sampling, aggregate analysis, and cell counting dissociated with Accutase for 3 minutes in a water bath at
Cell samples were received from the bioreactor without 37 °C, analyzed for definitve endoderm (DE) markers,
interrupting the stirring. Aggregate formation was monitored and further differentiated into intestinal cells by plating
by taking up to five independent light microscopic images down cells at 2 × 105 cells/cm2 in intestinal medium
per sample. Aggregate size was determined by the IamgeJ consisting of DMEM/F12 (Thermo Fisher Scientific),
software (Wayne Rasband, National Institute of Health). 2% fetal bovine serum, 500 ng/mL Fibroblast Growth Factor
Mean diameters represent arithmetic averages of 45 to 1096 (FGF) 4, 3 μM CHIR99021 and 1% penicillin/streptomycin.
single aggregates. Medium was changed every other day until day 7, when cells
Cell counting was performed after dissociating the were analyzed.
aggregates into single cells via accutase-treatment [13].
Cell supernatants were stored at -20 °C for subsequent Statistical analyses
metabolite analyses (glucose and lactate). Statistical analysis for three to four independent bioreactor
runs per culture condition were performed by the GraphPad
Definitive endoderm and intestinal differentiation Prism 6 software (GraphPad Software, Inc). One-way ANOVA
Definitive Endoderm and Intestinal Differentiation was or two-way ANOVA followed by Bonferroni’s post-test
performed as described previously [18]. were used to determine statistical significance. P-values of
In brief, after bioreactor expansion cell density <0.05, <0.01, <0.001 were considered significant. Results are
was determined and seeded at 10 × 105 cells/mL reported as mean and standard error of mean (SEM).
for differentiation in RPMI 1640 supplemented with
APPLICATION NOTE I No. 472 I Page 5

Results and Discussion

Influence of pH, glucose feeding, and DO control on hiPSC

bioreactor culture
In order to increase cellular yields, the DASbox Mini control (III, see Table 1) which resulted in higher culture
Bioreactor System was employed to introduce precise glucose concentrations (Figure 3B) but also increased levels
monitoring and control over pH, glucose feeding, and of lactate, a growth-inhibiting by-product of anaerobic
DO to an otherwise uncontrolled setting (I, see Table 1). glycolysis (Figure 3D). To provide a steadier oxygen supply,
With the integrated DASware software, the bioprocess DO-control was introduced with a setpoint of 40% starting
controller was able to automatically adjust and maintain a from the first day of culture (IV, see Table 1). This resulted
given growth condition. First, pH control with a setpoint in more stable culture oxygen levels, especially during later
of pH 7.0 (II, see Table 1) was initiated once a certain pH timepoints. However, initial DO levels (day 0-2) were lower
threshold was undercut, resulting in a steady culture pH compared to the other conditions (Figure 3C).
value of 7.0 throughout the run while higher fluctuations The resulting cell densities obtained from the
and a generally lower pH were present in the uncontrolled different growth parameter settings imply an
setting (Figure 3A). The pH-controlled culture displayed influence of the initial lowered oxygen levels under
a much higher glucose consumption over time compared condition IV (see Table 1). Each other optimization step
to the uncontrolled setting where glucose levels stagnated led to a significant increase in cell density compared to the
1 day post-inoculation (Figure 3B). 3.04 × 106 cells/mL from the uncontrolled situation at 7 days
To compensate for the decrease of this important post-inoculation. Introduction of pH control achieved about
carbohydrate source, glucose feeding (3.15 - 6.15 g/L) 4.03 × 106 cells/mL and additional glucose supplementation
was initiated from day 3 onwards in addition to the pH even 6.13 × 106 cells/mL (Figure 3E). Additional maintenance

A B C G

D E F

Figure 3: Effects of (A) pH (II, see Table 1), as well as additional (B) glucose feeding (III, see Table 1) and (C) DO control (IV, see
Table 1) on (D) lactate levels, (E) cell density, and (F/G) cell aggregate size of hiPSCs in stirred-tank bioreactor culture in comparison
to uncontrolled culture conditions (I, see Table 1).
APPLICATION NOTE I No. 472 I Page 6

of 40 % DO from the first day of culture however yielded (ii) A DO-level cascade was introduced, with an initial
only about the same levels as the glucose-supplemented DO level of 100 % that was automatically stabilized at
culture alone (6 × 106 cells/mL) (Figure 3E). This might 40 % once this value was undercut within the medium.
be explained by lower cell counts during the first days of (iii) The shear protectant Gibco Pluronic F-68 (Thermo
culture, coinciding with the reduced DO levels compared to Fisher Scientific) was added during inoculation in order
the other culture modes (Figure 3C/E). to (iv) provide a better cell environment for aggregate size
Also, the employed DO-strategy resulted in larger cell control achieved by increasing the agitation speed from the
aggregates by the end of the run (Figure 3F/G), possibly initial 60 to either 80, 100, or 120 rpm.
promoted by cell debris derived from a larger number of While all three agitation speed adjustments successfully
dead cells [19]. As oversized aggregates (>300 µm) can lead reduced the initial cell loss (now resulting in a typical lag
to detrimental oxygen and nutrient gradients, the size needs phase during the first 24 hours of incubation) (Figure 4B),
to be decreased. 80 rpm were already sufficient to keep the aggregate size
Nevertheless, these experiments show already the positive below the critical limit of 300 µm for the entire run (Figure
impact of controlled pH values and glucose feeding with 4C). These measures combined with the gentle stirring
up to double the cell density compared to an uncontrolled properties of the pitched 8-blade impeller led to cell
culture system. densities comparable to the 9.5 × 106 cells/mL achieved by
Furthermore, the observation that approach IV with 60 rpm at 7 days of incubation in this setting (Figure 4B). It
additional DO control (see Table 1) yielded final cell is also worth noting, that despite the higher agitation speed
densities comparable to the ones obtained from setting cell viability remained comparable between all conditions
III with glucose feeding alone (see Table 1) suggested a (Figure 4B). Thus, in order to strike a balance between
possible compensatory effect of DO supplementation at aggregate size control and minimal shear forces on the
later timepoints able to make up for the initial cell loss. Still, cells, 80 rpm were used from here on out for the following
further DO strategy adaptation was necessary to reduce early experiments (V, see Table 1).
cell loss and control aggregate size.
Optimized feeding and perfusion rate
Controlling cell aggregate size and viability by pre-culture As stated earlier, addition of glucose is beneficial to the cell
optimization, DO adaptation, and agitation adjustment culture growth but also increases the levels of the growth-
In order to overcome the initial cell loss and increased inhibiting metabolite lactate. Thus, in order to enable
aggregate size in setting IV (see Table 1), a four-stage increasing glucose levels and to prevent the rise of lactate
parameter adaption approach was conducted (Figure 4A): to cell growth-inhibiting levels, the medium perfusion rate
(i) Pre-culture optimization was carried out by shortening was stepwise increased from 1 to 2 culture-volumes/day
the 2D culture period from 96 hours to 72 hours, resulting in between culturing day 2 and 5. At the same time, glucose
a cell confluence of 65-75% instead of 90-95%. concentration was increased from 3.15 to 6.15 g/L between

A B C

60, 80, 100,

or 120 rpm

Figure 4: Preventing initial stem cell loss and controlling cell aggregate size by (A) pre-culture optimization, as well as DO and
agitation control. Effect of the optimization steps described in (A) on (B) viable cell density and (C) aggregate size.
APPLICATION NOTE I No. 472 I Page 7

A B C

Figure 5: Effects of optimized medium perfusion (VI, see Table 1) and glucose feeding on (A) glucose and (B) lactate levels, as well as
(C) viable stem cell density.

day 1 and 3 to 6.15 to 7.65 g/L from day 4 onwards (VI, see Thus, compared to uncontrolled conditions (I, see Table 1)
Table 1) (Figure 5A). With this optimized glucose feeding with 3.04 × 106 cells/mL, the combination of pH, glucose
strategy, also stronger lactate production was to be expected feeding, DO, agitation, and perfusion rate control along with
that should be counteracted by the elevated medium an optimized pre-culture resulted in a sixfold increase in cell
exchange facilitated by the higher medium flow rate. density at day 7 post inoculation.
Indeed, lactate levels in this setting remained comparable
to the non-feeding-optimized approaches (Figure 5B). Further culture optimization by in silico modelling
Yet, cell density increased once more to 18 × 106 cells/mL At this point, the precise parameter control provided by the
(Figure 5C), almost double the amount achieved during the DASbox Mini Bioreactor System and the gentle cell mixing
last optimization step. with the 8-blade impeller of the DASbox Mini Bioreactor

A B
viable cells [× 106/mL]

viable cells [× 106/mL]

viable cells [× 106/mL]

Glucose, Lactate [mM]

Glucose, Lactate [mM]

Glucose, Lactate [mM]

process day process day

process day
Glucose, Lactate [mM]

Glucose, Lactate [mM]

Glucose, Lactate [mM]

viable cells [× 106/mL]

viable cells [× 106/mL]

viable cells [× 106/mL]

process day process day

process day
Glucose, Lactate [mM]

Glucose, Lactate [mM]

viable cells [× 106/mL]

viable cells [× 106/mL]

process day
process day

Figure 6: (A) In silico modelling-supported cell density increase shown (B) for 3 different stem cell lines.
APPLICATION NOTE I No. 472 I Page 8

resulted already in a remarkable stem cell number increase. Stem cell properties of hiPSCs cultured in a stirred-tank
To further push the achievable cell density and to reduce the bioreactor
workload of testing each parameter adaptation in a wet-lab The previous results demonstrate that process optimization
setting, culture optimization was next supported by in silico through precise parameter control in a stirred-tank
modelling. bioreactor enables both higher cell numbers and robust cell
Using the Berkely-Madonna software viability.
(https://berkeley-madonna.myshopify.com/), the latest wet- Now, the next step was to test if the resulting stem cells
lab results achieved by setting VI (see Table 1) were fed were still retaining their expected stem cell properties
into an algorithm to predict further parameter optimization after such a strong growth period. For that, cells were
approaches (Figure 6A, first row). analyzed for the typical pluripotency markers TRA-1-60,
This resulted in model VII (see Table 1) with a predicted SSEA-4, OCT-3/4, NANOG, SOX2 and KI-67 after 7 days
cell density of 70 × 106 cell/mL on day 7 of culture. However, of culture. As shown in Figure 7A and B, cells postive for
wet-lab results of the suggested model reached cell densities each marker could be identified by flow cytometry and
of 23 × 106 cells/mL (Figure 6A, second row). Even though immunofluorescence microscopy.
these numbers surpassed all the previous approaches, they Another important property of iPSC is the ability to
were lower than expected, probably due to an unexpected differentiate into various cell types of the endo-, meso-
glucose peak by day 6 of culture, indicating the need for and ectoderm germ layer. Therefore, the presence of cell
further modelling adaptation. differentiation capabilities was tested by conventional
To further focus the precision of the modelling process, undirected differentiation, which revealed the expression
model VII (see Table 1) wet-lab results were fed into the of germ layer-specific markers Desmin (DES), SOX17 and
software once more, resulting in the model VIII approach TUBB3 (Figure 7C). Furthermore, application of specific
(see Table 1). Now, the wet-lab results resembled the differentiation protocols to hiPSCs cultured for 4 and 7 days
predicted 40 × 106 cells/mL much closer, reaching to generate definitive endoderm, intestinal progenitors and
33 × 106 cells/mL, almost 10× the density of uncontrolled cardiomyocytes resulted in the reproducible induction of
conditions (Figure 6A, third row). 85 – 95 % of these progenies (Figure 7D). Additionally, it
These unmatched numbers of almost 5 × 109 cells in is worth noting that no chromosomal abnormalities were
150 mL culture volume were confirmed by using three observed for the process-derived cells after 7 days of culture,
different stem cell line culture runs under the model VIII as exemplarily shown for hHSC_1285 in Figure 7E.
(see Table 1) conditions with cell viabilities comparable to Together, these data suggest that the pluripotent stem
previous runs (Figure 6B). cell population maintained all expected key properties
after cultivation to high cell densities achieved by process
optimization in a stirred-tank bioreactor.
APPLICATION NOTE I No. 472 I Page 9

A A

B
A

C
A

D
A E
A

Figure 7: Stem cell properties of hiPSCs cultured in a stirred-tank bioreactor. Analysis of pluripotency-associated marker expression
in undifferentiated hiPSCs by (A) flow cytometry and (B) immunofluorescence microscopy, as well as (C) in hiPSCs differentiated
towards the three germ layers. (D) Differentiation of process-derived (d4 and d7) GMPDU_8 aggregates into definitive endoderm
(DE), intestinal progenitors and cardiomyocytes. (E) Karyotype of cells cultured for 7 days under condition VIII (see Table 1).
APPLICATION NOTE I No. 472 I Page 10

Conclusion

The data described in this study demonstrate the power of a enabled efficient mixing, aggregate size control, and high
controllable and adjustable growth environment. Using the cell viability, while keeping shear force stress low at the
precise parameter control capabilities of the DASbox Mini same time.
Bioreactor System in combination with systematic adaptation We believe that this step-by-step parameter adaptation
and in silico process modelling enabled the increase of stem represents one of the most efficient ways to approach
cell densities from about 3 × 106 cells/mL in an uncontrolled process optimization and development for stem cell culture
setting to almost 35 × 106 cells/mL (Figure 8). This translated in bioreactors. The procedure described here can act as a
to a total cell number of almost 5 × 109 cells within the whole roadmap to identify and overcome cultivation bottlenecks,
culture volume. Furthermore, the specifically designed increase stem cell numbers, and advance the field of stem
60° pitched 8-blade impeller of the DASbox Mini Bioreactor cell applications.

I Uncontrolled conditions I Cell numbers per mL after 7 days of incubation

Process optimization stages

II + pH control II

III + Glucose feeding III

IV + DO control IV

V + (Pre-)culture/DO optimization V

VI + Perfusion/Feeding optimization VI

VII/VIII + in silico modelling VIII

0 5 10 15 20 25 30 35
viable cells [× 106]/mL

Figure 8: Summary of the process optimization steps along with the achieved cell densities.
APPLICATION NOTE I No. 472 I Page 11

Literature

[1] A. Y. L. Wang, ‘Human induced pluripotent stem cell‐derived exosomes as a new therapeutic strategy for various
diseases’, International Journal of Molecular Sciences, vol. 22, no. 4. MDPI AG, pp. 1–17, Feb. 02, 2021. doi: 10.3390/
ijms22041769.
[2] K. Takahashi and S. Yamanaka, ‘Induction of Pluripotent Stem Cells from Mouse Embryonic and Adult Fibroblast
Cultures by Defined Factors’, Cell, vol. 126, no. 4, pp. 663–676, Aug. 2006, doi: 10.1016/j.cell.2006.07.024.
[3] A. Haase et al., ‘Generation of Induced Pluripotent Stem Cells from Human Cord Blood’, Cell Stem Cell, vol. 5, no. 4, pp.
434–441, Oct. 2009, doi: 10.1016/j.stem.2009.08.021.
[4] R. Zweigerdt, I. Gruh, and U. Martin, ‘Your Heart on a Chip: iPSC-Based Modeling of Barth-Syndrome-Associated
Cardiomyopathy’, Cell Stem Cell, vol. 15, no. 1, pp. 9–11, Jul. 2014, doi: 10.1016/j.stem.2014.06.015.
[5] A. E. H. El-Kadiry, M. Rafei, and R. Shammaa, ‘Cell Therapy: Types, Regulation, and Clinical Benefits’, Frontiers in
Medicine, vol. 8. Frontiers Media S.A., Nov. 22, 2021. doi: 10.3389/fmed.2021.756029.
[6] G. M. Pigeau, E. Csaszar, and A. Dulgar-Tulloch, ‘Commercial scale manufacturing of allogeneic cell therapy’, Frontiers
in Medicine, vol. 5, no. AUG. Frontiers Media S.A., 2018. doi: 10.3389/fmed.2018.00233.
[7] R. Zweigerdt, R. Olmer, H. Singh, A. Haverich, and U. Martin, ‘Scalable expansion of human pluripotent stem cells in
suspension culture’, Nat Protoc, vol. 6, no. 5, pp. 689–700, May 2011, doi: 10.1038/nprot.2011.318.
[8] C. Jensen and Y. Teng, ‘Is It Time to Start Transitioning From 2D to 3D Cell Culture?’, Frontiers in Molecular
Biosciences, vol. 7. Frontiers Media S.A., Mar. 06, 2020. doi: 10.3389/fmolb.2020.00033.
[9] F. Manstein et al., ‘High density bioprocessing of human pluripotent stem cells by metabolic control and in silico
modeling’, Stem Cells Transl Med, vol. 10, no. 7, pp. 1063–1080, Jul. 2021, doi: 10.1002/sctm.20-0453.
[10] F. Manstein, K. Ullmann, W. Triebert, and R. Zweigerdt, ‘Process control and in silico modeling strategies for enabling
high density culture of human pluripotent stem cells in stirred tank bioreactors’, STAR Protoc, vol. 2, no. 4, p. 100988,
Dec. 2021, doi: 10.1016/j.xpro.2021.100988.
[11] R. Olmer et al., ‘Suspension culture of human pluripotent stem cells in controlled, stirred bioreactors’, Tissue Eng Part C
Methods, vol. 18, no. 10, pp. 772–784, Oct. 2012, doi: 10.1089/ten.tec.2011.0717.
[12] C. Kropp et al., ‘Impact of Feeding Strategies on the Scalable Expansion of Human Pluripotent Stem Cells in Single-Use
Stirred Tank Bioreactors’, Stem Cells Transl Med, vol. 5, no. 10, pp. 1289–1301, Oct. 2016, doi: 10.5966/sctm.2015-
0253.
[13] H. Kempf, C. Kropp, R. Olmer, U. Martin, and R. Zweigerdt, ‘Cardiac differentiation of human pluripotent stem cells in
scalable suspension culture’, Nat Protoc, vol. 10, no. 9, pp. 1345–1361, Oct. 2015, doi: 10.1038/nprot.2015.089.
[14] M. Ackermann et al., ‘Continuous human iPSC-macrophage mass production by suspension culture in stirred tank
bioreactors’, Nat Protoc, vol. 17, no. 2, pp. 513–539, Feb. 2022, doi: 10.1038/s41596-021-00654-7.
[15] S. Hartung et al., ‘Directing cardiomyogenic differentiation of human pluripotent stem cells by plasmid-based transient
overexpression of cardiac transcription factors’, Stem Cells Dev, vol. 22, no. 7, pp. 1112–1125, Apr. 2013, doi: 10.1089/
scd.2012.0351.
[16] A. Haase, G. Göhring, and U. Martin, ‘Generation of non-transgenic iPS cells from human cord blood CD34+ cells under
animal component-free conditions’, Stem Cell Res, vol. 21, pp. 71–73, May 2017, doi: 10.1016/j.scr.2017.03.022.
[17] A. Haase et al., ‘GMP-compatible manufacturing of three iPS cell lines from human peripheral blood’, Stem Cell Res, vol.
35, Mar. 2019, doi: 10.1016/j.scr.2019.101394.
[18] A. Sahabian et al., ‘Chemically-Defined, Xeno-Free, Scalable Production of hPSC-Derived Definitive Endoderm
Aggregates with Multi-Lineage Differentiation Potential’, Cells, vol. 8, no. 12, p. 1571, Dec. 2019, doi: 10.3390/
cells8121571.
[19] W. A. Renner, M. Jordan, H. M. Eppenberger, and C. Leist, ‘Cell-cell adhesion and aggregation: Influence on the growth
behavior of CHO cells’, Biotechnol Bioeng, vol. 41, no. 2, pp. 188–193, Jan. 1993, doi: 10.1002/bit.260410204.
APPLICATION NOTE I No. 472 I Page 12

Supplementary Information

Supplementary Table 1: : Culture condition abbreviations used in this

application note and the equivalent from the original publication [9].
Abbreviations used Nomenclature in the
in text original publication
I pUC
II p7
III p7G
IV p7GO
V p7GOS80
VI p7GOS80oF
VII Stg1M
VIII Stg2M

Ordering information
Description Order no.
DASbox® Mini Bioreactor System, for cell culture, 4-fold system 76DX 04C C
DASbox® Mini Bioreactor, cell culture, glass, stainless steel head plate, temperature, pH, DO, and level sensors 76DS 025 0OD SS
Pitched-Blade Impeller, 8-blade, 60° pitch, stainless steel, O.D. 34 mm, I.D. 5 mm 7810 060 4
DASbox® overhead drive, 76DX OHD
DASbox® exhaust system, 76DX OFF
DASbox® exhaust condenser, Peltier 76DX CON D
Compression Fitting, complete, with Pg 13.5 male thread, I.D. 12 mm 7853 228 4
Triple Port, Pg 13.5 thread, 3 tubes with O.D. 4 mm × L 85 mm, all parts included 7870 641 4
Pipe, stainless steel, with barb, O.D. 4 mm/I.D. 2 mm, L 225 mm 7810 702 3
Compression Fitting, complete, with Pg 13.5 male thread, I.D. 6 mm 7853 228 3
L-Sparger, stainless steel, complete, O.D. 6 mm, L 300 mm, W 63 mm 7710 202 2
Pump Head Tubing, for DASGIP MP8 pump, Bioprene, I.D. 0.5/W 1.05 mm, female/female 7851 011 8
Pump Head Tubing, for DASGIP MP8 pump, Bioprene, I.D. 1.0/W 1.05 mm, male/female 7851 010 9
Feed Line, with 2× Luer lock fittings, male/male, C-Flex, I.D. 0.8 mm, L 1 m 7851 030 9
Feed Line, with 2× Luer lock fittings, male/male, C-Flex, I.D. 0.8 mm, L 2 m 7851 031 0
Sampling Accessory, with swabable valve 7851 014 5
DASware® control software, including PC, OS, and licenses, for 4-fold DASbox® Mini Bioreactor System 7860 016 7
Eppendorf Safe-Lock Tubes, 1.5 and 2.0 mL 0030 120 086

Your local distributor: www.eppendorf.com/contact

Eppendorf SE · Barkhausenweg 1 · 22339 Hamburg · Germany
[email protected] · www.eppendorf.com

www.eppendorf.com
Sigma-Aldrich® is a registered trademarks of Merck KGAA, Germany. Thermo Fisher Scientific® is a registered trademark of Thermo Fisher Scientific Inc., USA. Eppendorf® and the Eppendorf Brand Design are registered trademarks of
Eppendorf SE, Germany. DASbox® and DASware® are registered trademarks of DASGIP Information and Process Technology GmbH, Germany. All rights reserved, including graphics and images. Copyright © 2023 by Eppendorf SE.

Eppendorf SE reserves the right to modify its products and services at any time. This application note is subject to change without notice. Although prepared to ensure accuracy, Eppendorf SE assumes no liability for errors, or for any
damages resulting from the application or use of this information. Viewing the application note alone cannot as such provide for or replace reading and respecting the current version of the operating manual.

AA472-020-01-092023-STS