Content uploaded by Hananeh Fonoudi
Author content
All content in this area was uploaded by Hananeh Fonoudi on Nov 05, 2015
Content may be subject to copyright.
Protocols and Manufacturing for Cell-Based Therapies
A Universal and Robust Integrated Platform for the
Scalable Production of Human Cardiomyocytes From
Pluripotent Stem Cells
HANANEH FONOUDI,
a,b,c,
*
HASSAN ANSARI,
a,
*
SAEED ABBASALIZADEH,
a
MEHRAN REZAEI LARIJANI,
a
SAHAR KIANI,
a
SHIVA HASHEMIZADEH,
a
ALI SHARIFI ZARCHI,
a
ALEXIS BOSMAN,
b,c
GILLIAN M. BLUE,
d,e,f
SARA PAHLAVAN,
a
MATTHEW PERRY,
c,g
YISHAY ORR,
d,e
YAROSLAV MAYORCHAK,
e
JAMIE VANDENBERG,
c,g
MAHMOOD TALKHABI,
a
DAVID S. WINLAW,
d,e,f
RICHARD P. HARVEY,
b,c,h
NASSER AGHDAMI,
a
HOSSEIN BAHARVAND
a,i
Key Words. Human pluripotent stem cells xEmbryonic stem xInduced pluripotent stem x
Cardiomyocytes xDirected differentiation xCell therapy xSmall molecules xBioreactor
ABSTRACT
Recent advances in the generation of cardiomyocytes (CMs) from human pluripotent stem cells (hPSCs), in
conjunction with the promising outcomes from preclinical and clinical studies, have raised new hopes for
cardiac cell therapy. We report the development of a scalable, robust, and integrated differentiation plat-
form for large-scale production of hPSC-CM aggregates in a stirred suspension bioreactor as a single-unit
operation. Precise modulation of the differentiation process by small molecule activation of WNT signal-
ing, followed by inactivation of transforming growth factor-band WNT signaling and activation of sonic
hedgehog signaling in hPSCs as size-controlled aggregates led to the generation of approximately 100%
beating CM spheroids containing virtually pure (∼90%) CMs in 10 days. Moreover, the developed differ-
entiation strategy was universal, as demonstrated by testing multiple hPSC lines (5 human embryonic stem
cell and 4 human inducible PSC lines) without cell sorting or selection. The produced hPSC-CMssuccessfully
expressed canonical lineage-specific markers and showed high functionality, as demonstrated by micro-
electrode array and electrophysiology tests. This robust and universal platform could become a valuable
tool for the mass production of functional hPSC-CMs as a prerequisite for realizing their promising poten-
tial for therapeutic and industrial applications, including drug discovery and toxicity assays. STEM CELLS
TRANSLATIONAL MEDICINE 2015;4:1–13
SIGNIFICANCE
Recent advances in the generation of cardiomyocytes (CMs) from human pluripotent stem cells
(hPSCs) and the development of novel cell therapy strategies using hPSC-CMs (e.g., cardiac patches)
in conjunction with promising preclinical and clinical studies, have raised new hopes for patients with
end-stage cardiovascular disease, which remains the leading cause of morbidity and mortality glob-
ally. In this study, a simplified, scalable, robust, and integrated differentiation platform was devel-
oped to generate clinical grade hPSC-CMs as cell aggregates under chemically defined culture
conditions. This approach resulted in approximately 100% beating CM spheroids with virtually pure
(∼90%) functional cardiomyocytes in 10 days from multiple hPSC lines. This universal and robust bio-
processing platform can provide sufficient numbers of hPSC-CMs for companies developing regen-
erative medicine technologies to rescue, replace, and help repair damaged heart tissues and for
pharmaceutical companies developing advanced biologics and drugs for regeneration of lost heart
tissue using high-throughput technologies. It is believed that this technology can expedite clinical
progress in these areas to achieve a meaningful impact on improving clinical outcomes, cost of care,
and quality of life for those patients disabled and experiencing heart disease.
INTRODUCTION
Human pluripotent stem cells (hPSCs), including
human embryonic stem cells (hESCs) and indu ced
pluripotent stem cells (hiPSCs), can be consid-
ered an unlimited source for the production of
cardiomyocytes (CMs) owing to their self-renewal
and directed differentiation capabilities [1, 2].
Several approaches, including embryoid body for-
mation [3], coculture [4], cytokine-directed dif-
ferentiation [5], and protein transduction [6],
have been introduced to generate hPSC-derived
a
Department of Stem Cells and
Developmental Biology, Cell
Science Research Center,
Royan Institute for Stem Cell
Biology and Technology, and
i
Department of
Developmental Biology,
University of Science and
Culture, Academic Center for
Education, Culture and
Research, Tehran, Iran;
b
Developmental and Stem Cell
Biology Division and
g
Molecular Cardiology and
Biophysics Division, Victor
Chang Cardiac Research
Institute, Darlinghurst, Sydney,
New South Wales, Australia;
c
St. Vincent’s Clinical School,
Faculty of Medicine, and
h
School of Biotechnology and
Biomolecular Sciences,
University of New South
Wales, Kensington, New South
Wales, Australia;
d
Kids Heart
Research and
e
The Heart
Centre for Children, The
Children’s Hospital at
Westmead, Sydney, New
South Wales, Australia;
f
Sydney Medical School,
University of Sydney, Sydney,
New South Wales, Australia
*
Contributed equally.
Correspondence: Hossein
Baharvand, Ph.D., Department of
Stem Cells and Developmental
Biology, Cell Science Research
Center, Royan Institute for Stem
Cell Biology and Technology,
Academic Center for Education
Culture and Research, P.O. Box
19395-4644, Tehran, Iran.
Telephone: 98-21-2230-6485;
E-Mail: Baharvand@RoyanInstitute.
org
Received December 4, 2014;
accepted for publication July 8,
2015.
©AlphaMed Press
1066-5099/2015/$20.00/0
http://dx.doi.org/
10.5966/sctm.2014-0275
STEM CELLS TRANSLATIONAL MEDICINE 2015;4:1–13 www.StemCellsTM.com ©AlphaMed Press 2015
P
ROTOCOLS AND
M
ANUFACTURING FOR
C
ELL
-B
ASED
T
HERAPIES
by AlphaMed Press on November 5, 2015http://stemcellstm.alphamedpress.org/Downloaded from
Published Ahead of Print on October 28, 2015 as 10.5966/sctm.2014-0275.
cardiomyocytes since the first report in 2001. Despite significant
methodological advances, the hPSC-based production platforms
still contain critical disadvantages, including high costs and low
efficiency and reproducibility. This has limited the application
of hPSC-derived cardiomyocytes for clinical and industrial appli-
cations such as drug discovery and toxicity testing [7].
Most of the promising applications of hPSCs-CMs, including
those that are closer to commercialization such as high-throughput
drug screening and toxicity testing, followed by clinical applica-
tions, necessitate the production of massive numbers of pure
and functional CMs [8]. Therefore, developing robust and afford-
able technologies for large-scale expansion of hPSCs and their
integrated differentiation into cardiomyocytes in scalable culture
systems would largely facilitate their commercial applications.
To date, a number of different protocols and technologies
have been introduced for the production of CMs from hPSCs with
the aim of optimizing different aspects of their bioprocessing, in-
cluding culture conditions that favor the production of mature
and fully functional CMs, and robust, integrated, and scalable dif-
ferentiation and purification methodologies [9, 10].
To develop fully defined culture conditions for generation of
hPSC-CMs, the exploration of novel and efficient small molecules
(SMs) for chemically induced cardiac differentiation has recently
emerged as a viable alternative to recombinant cytokines and un-
known factors in serum [11–13]. Manipulation of the signaling
pathways required for normal heart development has guided
the development of efficient hPSC-CM differentiation protocols
[9, 14–16].
In contrast, the current experimental methods for the differ-
entiation of hPSCs often rely on the production of heterogeneous
cellular aggregates termed “embryoid bodies”(EBs) or two-
dimensional (2D) small-scale static cultures [17–19]. These proto-
cols are typically not scalable and/or result in the generation of
EBs with a highly heterogeneous size and low CM yield. In order
to establish more robust bioprocesses, different approaches,
such as microwell-mediated control, microprinting technologies
[20–23], and microcarrier cultures [20], have been applied to
address these issues. However, these techniques have limited dif-
ferentiation potential, scalability (microwell-mediated control,
hanging drop, and microprinting technologies), universality,
and/or reproducibilitybecause of the differentiationprotocol itself
or the low throughput of the methods used such as forced aggre-
gation techniques before transferring cells to dynamic culture
conditions. In addition, most of the protocols depend on using
expensive and complex media or reagents (mTeSR1; StemCell
Technologies, Vancouver, BC, Canada, http://www.stemcell.com;
or StemPro-34; Thermo Fisher Scientific, Waltham, MA, http://
www.thermofisher.com) or microcarriers for expansion of hPSCs
and their directed differentiation to cardiomyocytes [24, 25].
Recently, efforts have been made to develop scalable culture
systems for the large-scale production of hPSCs-CMs under
three-dimensional (3D) culture conditions. Kempf et al. tried to
generate hPSC-CMs in a 100-ml stirred suspension bioreactor
with batch and perfused modes using mTeSR medium (StemCell
Technologies) for the hPSC expansion phase and Roswell Park
Memorial Institute (RPMI)/B27 medium (Thermo Fisher Scien-
tific) for the differentiation phase. However, batch cultures failed
to generate contracting EBs [24], and perfused cultures resulted
in contracting EBs with a heterogeneous size (350–600 mm in di-
ameter), which could limit their future application for clinical and
industrial use.
In a previous study, we developed a robust and cost-effective
culture system for mass production of size-controlled hPSC aggre-
gate cultures in stirred suspension bioreactors. Our protocols
have paved the way for mass production of these unique cells
under xeno-free conditions with superior scalability (review avail-
able in [26]). Subsequently, we have shown that this platform can
be easily integrated for large-scale generation of hepatocyte-like
cells that improved hepatic failure in an animal model after trans-
plantation [27, 28].
In the present report, we have used this platform for the
development of an integrated, simplified process for large-
scale production of highly homogenous hPSC-CM aggregates in
a cost-effective single-unit operation with high efficacy, repro-
ducibility, and universality. This scalable production system can
be easily integrated with an efficient scalable purification system,
including culture-based methods, such as those using lactate-
enriched medium for selective purification of CMs, to generate
a highly pure population of cardiac cells for biomedical applica-
tions [29]. The development of such integrated platforms can
be considered an important step toward the commercialization
of hPSCs-CM-based technologies for clinical, pharmaceutical, tis-
sue engineering, and in vitro organ development applications.
MATERIALS AND METHODS
Generation of Fibroblast Cultures
The Sydney Children’s Hospital Network Human Research Ethics
Committee provided ethics approval (reference no. HREC/10/
CHW/44). Suitable participants were selected from the Kids Heart
Research DNA Bank, and the participants’parents or guardians
provided consent. Skin biopsies were performed by cardiotho-
racic surgeons and sampled from the upper arm of the partici-
pants. Fibroblasts were cultured from the skin biopsy specimens
to establish cell lines for each participant.
Culture of hPSCs as Aggregates
hESC lines (RH5, RH6, R725.1, R661.5, and R662.2) [30, 31]
and hiPSC lines (VC645-9, VC913-5, VC618-3, and VC646-1), the
latter generated at the Victor Chang Cardiac Research Institute,
Australia, were used in the present study. hPSCs were expanded
using a previously described suspension culture method [27, 32].
In brief, to initiate suspension cultures, 2 310
5
viable cells per
milliliter were transferred to nonadhesive bacterial plates
(60 mm; catalog no. 628102; Greiner Bio-One, Sigma-Aldrich,
St. Louis, MO, http://www.sigmaaldrich.com) in 5 ml of hESC-
conditioned medium containing 100 ng/ml basic fibroblast
growth factor (bFGF; Royan Institute for Stem Cell Biology
and Technology, Academic Center for Education Culture and
Research, Tehran, Iran). The cells were incubated under standard
conditions (37°C, 5% CO
2
and saturated humidity). The medium
was renewed daily. The hESC medium included Dulbecco’s
modified Eagle’s medium/F12 medium (catalog no. 21331-020;
Gibco, Grand Island, NY, http://www.lifetechnologies.com) sup-
plemented with 20% Knockout Serum Replacement (catalog
no. 10828-028; Gibco), 2 mM L-glutamine (catalog no. 25030-
024; Gibco), 0.1 mM b-mercaptoethanol (catalog no. M7522;
Sigma-Aldrich), 1% nonessential amino acids (catalog no.
11140-035; Gibco), 1% penicillin and streptomycin (catalog
no. 15070-063; Gibco), 1% insulin-transferrin-selenite (catalog
no. 41400-045; Gibco).
2Scalable Production of hPSC-Derived Cardiomyocytes
©AlphaMed Press 2015 STEM CELLS TRANSLATIONAL MEDICINE
by AlphaMed Press on November 5, 2015http://stemcellstm.alphamedpress.org/Downloaded from
Published Ahead of Print on October 28, 2015 as 10.5966/sctm.2014-0275.
Conditioned medium was prepared by overnight incubation
of hESC medium (without bFGF) with confluent human foreskin
fibroblasts in 75-cm
2
T flasks, previously inactivated by treatment
with mitomycin C (catalog no. M0503; Sigma-Aldrich).
Optimizing Differentiation Process for Cardiac
Differentiation in Static Suspension Culture
The hPSC aggregates cultured in static suspension mode (i.e.,
nonadhesive bacterial plates) were directly differentiated into
CMs for optimization trials. In order to induce CM differentia-
tion from hPSCs in suspension culture systems, 3-, 5-, and 7-
day-old hPSC size-controlled spheroids (average size, 90 630,
175 625, and 250 632 mm, respectively) were treated for 24
hours in differentiation medium (RPMI 1640; catalog no. 31870-
022; Gibco) supplemented with 2% B27 without retinoic acid (cat-
alog no. 12587-010; Gibco) or without insulin (catalog no. A18956-01;
Gibco), 2 mM L-glutamine, 0.1 mM b-mercaptoethanol, 1% non-
essential amino acids, 1% penicillin and streptomycin, and differ-
ent concentrations (3, 6, 9, 12, and 15 mM) of the SM CHIR99021
(CHIR; catalog no. 041-0004; Stemgent, Cambridge, MA, http://
www.stemgent.com) as glycogen synthase kinase inhibitor and
canonical WNT/b-catenin pathway activator. The spheroids were
washed with Dulbecco’s phosphate-buffered saline (DPBS) and
then maintained in fresh differentiation medium without SMs
for 1 day. After 1 day, the medium was exchanged for new differ-
entiation medium that contained 5 mM IWP2 (catalog no. 3533;
Tocris Bioscience, Bristol, U.K., http://www.tocris.com) as a WNT
antagonist, 5 mM SB431542 (catalog no. S4317; Sigma-Aldrich) as
an inhibitor of transforming growth factor-bsuperfamily type I acti-
vin receptor-like kinase receptors, and 5 mM purmorphamine (Pur;
catalog no. 04-0009; Stemgent) as a sonic hedgehog (SHH) agonist.
Spheroids were cultured for 2 days in this medium. After washing
the spheroids with DPBS, fresh differentiation medium without
SMs was added and refreshed every 2–3 days until the end of
the differentiation process at day 30.
Integrated Differentiation of hPSCs Toward CMs in
a Stirred Suspension Bioreactor
In order to develop an integrated differentiation process to gen-
erate human CMs from hPSCs in a scalable manner, we expanded
hPSCs as a single cell inoculated suspension culture in 125-ml
spinner flasks (Cellspin; Integra Biosciences AG, Zizers, Switzer-
land, http://www.integra-biosciences.com) with a 100-ml work-
ing volume at a 40 rpm agitation rate, as previously described
[27]. In brief, to initiate the dynamic cultures, 2 310
5
cells per
milliliter were transferred to 100 ml of hPSC medium, conditioned
on human foreskin fibroblast, and containing freshly added
100 ng/ml bFGF [27] in the stirred bioreactor. Cell aggregates
were treated with 10 mMr-activated kinase inhibitor (ROCKi; cat-
alog nos. Y-27632 and Y0503; Sigma-Aldrich) 1 hour before enzy-
matic dissociation with Accumax cell aggregate dissociation
medium (catalog no. AM105; Innovative Cell Technologies, Inc.,
San Diego, CA, http://www.accutase.com). The bioreactor was
agitated at 35 rpm (increased to 40 rpm after 24 hours), and me-
dium refreshing began after 48 hours of culture with the same
medium, without ROCKi. The seeded spinner flasks were incu-
bated under standard conditions. After passages in dynamic sus-
pension culture, the cells were cryopreserved as previously
described [28] or induced for differentiation.
To induce CM differentiation from hPSCs in dynamic suspen-
sion culture, the time points, differentiation media, and SMs were
exactly similar to the static system, with the exception of the ad-
dition of 0.1% polyvinyl alcohol (PVA; catalog no. 363073; Sigma-
Aldrich) for the first 48 hours and 10 mM ROCKi for the first
24 hours. In addition, aggregates were washed twice with 25 ml
of DPBS after mesodermal and cardiac induction steps by stop-
ping agitation for 5–10 minutes and removing the spent media
containing SM cocktails. After washing the cardiac-induced aggre-
gates, fresh differentiation medium (100 ml) was added to the
culture vessel without SMs and totally refreshed every 2–3 days
until the end of the differentiation process at day 30.
Exploring the Optimum hPSC Aggregate Size for
Integrated Cardiac Differentiation
In order to achieve size-controlled hPSC aggregates and, subse-
quently, hPSC-CMs with high differentiation efficacy, the RH5 cell
line was cultured in dynamic suspension culture for 3, 5, and
7 days under optimal hydrodynamic culture conditions. The diam-
eters of size-controlled aggregates generated in each culture
were quantified using a phase-contrast inverted microscope
(model no. CKX41; Olympus, Center Valley, PA, http://www.
olympusamerica.com) and Olysia Bioreport software (Soft Imag-
ing System, Olympus). Next, hPSC aggregates of a defined size
were transferred to differentiation media under static conditions
to explore the optimum size of dynamic culture systems for
differentiation.
Quantification of Beating Spheroids
We determined the cardiogenic differentiation efficiency using an
inverted cell culture microscope (model no. CKX41; Olympus) to
count the number of beating hESC and hiPSC spheroids through-
out the experiment. The numbers of beating spheroids were nor-
malized to the total numbers of spheroids at each time point. All
quantification experiments and analyses were performed using
at least three independent biological replicates.
RNA Isolation and Quantitative Real-Time Polymerase
Chain Reaction
We collected the hPSC spheroids at various time points in the dif-
ferentiation process in both static and dynamic systems. Total
RNA was extracted using the TRIzol reagent (Invitrogen, Carlsbad,
CA, http://www.invitrogen.com). RNA quality and concentration
were analyzed using a Biochrom WPA spectrophotometer
(Biochrom, Holliston, MA, http://www.biochrom.com). Possible
genomic DNA contamination was removed by DNase I (Invitro-
gen) treatment for 15 minutes at room temperature after which
2mg of total RNA was used for reverse transcription with an oligo
(dT)
20
primer and Super Script III First-Strand Synthesis System
(Invitrogen), according to the manufacturer’s instructions. Quan-
titative reverse transcription polymerase chain reaction (RT-PCR)
was performed with the Power SYBR Green PCR Master Mix (Applied
Biosystems, Foster City, CA, http://www.appliedbiosystems.com) in
triplicate for each sample and each gene. The PCR conditions included
denaturation at 94°C for 30 seconds, annealing at 60°C for 30 seconds,
and extension at 72°C for 1 minute for 35 cycles, with a 72°C extension
for 7 minutes at the end. The expression of genes of interest was nor-
malized according to GAPDH expression. The relative gene expression
levels were quantified using the 2
(2DDCt)
method. The primer se-
quences are listed in supplemental online Table 1.
Fonoudi, Ansari, Abbasalizadeh et al. 3
www.StemCellsTM.com ©AlphaMed Press 2015
by AlphaMed Press on November 5, 2015http://stemcellstm.alphamedpress.org/Downloaded from
Published Ahead of Print on October 28, 2015 as 10.5966/sctm.2014-0275.
In order to analyze quantitative RT-PCR data, we used R sta-
tistical language (R Foundation for Statistical Computing, Vienna,
Austria, http://www.r-project.org) [33]. Principal component
analysis (PCA) was performed on the scaled data. For the time-
course analysis, the genes were clustered according to the ex-
pression values in different samples using a K-means algorithm.
Visualization of the data was performed using the R packages
ggplot2 [34] and heatmap.
Flow Cytometry
RH5 hESC spheroids were collected at different time points after
differentiation initiation in static and dynamic systems, washed
twice with PBS, incubated with 0.05% trypsin-EDTA (catalog no.
25300-054; Gibco) at 37°C for 4–5 minutes and then pipetted
5–12 times. After neutralizing trypsin activity by the addition of
medium, the cell suspension was passed through a 40-mm filter
mesh (catalog no. 352340; BD Falcon, BD Biosciences, San Diego,
CA, http://www.bdbiosciences.com) to remove clumps and un-
dissociated spheroids. After trypsinization and achievement of
single-cell suspensions, the cells were washed twice in ice-cold
staining buffer (PBS supplemented with 1% heat-inactivated fetal
bovine serum[FBS], 0.1% sodium azide, and 2 mM EDTA) and fixed
in high-grade 4% paraformaldehyde (PFA) for 15 minutes at 4°C.
The cells were washed again with staining buffer, permeabilized
with 0.2% (vol/vol) Triton X-100 in PBS for 20 minutes, and
blocked for 15 minutes at 4°C with a combination of 10%
heat-inactivated goat serum in staining buffer. The cells were in-
cubated overnight at 4°C (or 30 minutes at 37°C) with the suit-
able primary antibodies (1:100) or appropriate isotype matched
controls, and then washed three times with staining buffer, after
which secondary antibodies (1:500) were added to the cells. After
45 minutes of incubation at 4°C, the cells were washed three
times with staining buffer and analyzed using a flow cytometer
(FACSCalibur; BD Biosciences) and flowing software, version
2.5.1 (BD Biosciences). For each analysis, 0.5–1310
6
cells were
used per sample. All experiments were replicated at least three
times. The primary and secondary antibodies used for flow cy-
tometry are listed in supplemental online Table 2.
Immunostaining and Imaging
hPSC differentiated beating spheroids were collected at different
time points after differentiation initiation in static and dynamic
systems. The collected spheroids were washed twice with PBS
and dissociated into single cells with 0.05% trypsin-EDTA (catalog
no. 25300-054; Gibco) at 37°C for 4–5 minutes. Individualized
cells were cultured on gelatin-coated chamber slides (catalog
no. 177437; Nunc; Thermo Scientific) in RPMI/B27 medium. After
5 days, the attached cells were washed once with PBS, fixed with
4% (wt/vol) PFA at room temperature for 15 minutes, washed
once with washing buffer (PBS/0.1% Tween 20), permeabilized
with 0.2% Triton X-100 in PBS for 15 minutes, and blocked with
5% (vol/vol) goat serum for 1 hour. Primary antibodies diluted
in blocking buffer (1:100) were added to the cells, followed by
an overnight incubation at 4°C. After incubation, the cells were
washed three times with washing buffer, each for 5 minutes. Sec-
ondary antibodies diluted in blocking buffer (1:500) were added
to cells, after which they were incubated for 1 hour at room tem-
perature. The cells were subsequently washed three times with
washing buffer, then covered with a Vectashield mounting
medium that contained 496-diamidino-2-phenylindole (Vector
Laboratories, Inc., Burlingame, CA, http://www.vectorlabs.com).
Imaging was performed using an upright confocal microscope
(Zeiss LSM700, Zeiss, Jena, Germany).
For immunohistochemistry analysis, the beating RH5 hESC
spheroids were collected, washed with PBS, fixed with 4% (wt/
vol) PFA at room temperature for 15 minutes, and prepared for
paraffin-embedded tissue blocks. Paraffin-embedded spheroids
were cut into 6-mm sections using a microtome (Microm
HM325; Thermo Scientific) and kept at room temperature until
use. For staining, we dewaxed and hydrated the spheroid section
slides, followed by heat-mediated antigen retrieval using a Dako
target retrieval solution (catalog no. S2367; Dako, Glostrup, Den-
mark, http://www.dako.com). Permeabilization, blocking steps,
and incubation with primary and secondary antibodies were per-
formed as described for individualized cultured cells. The primary
and secondary antibodies used for cultured cells and staining the
spheroids are listed in supplemental online Table 2.
Microelectrode Array Recording
We characterized the functional properties of hESC spheroid-
derived CMs by performing an extracellular recording of field
potentials (FPs) using a microelectrode array (MEA) data acquisi-
tion system (Multi Channel Systems, Reutlingen, Germany). The
MEA plates contained a matrix of 60 titanium nitride electrodes
(30 mm) with an interelectrode distance of 200 mm. MEA plates
were sterilized and hydrophilized with FBS for 30 minutes,
washed with sterile water and coated with 0.1% gelatin for 1 hour.
For this analysis, areas of plated beating spheroids were
mechanically dissected and plated on the middle of a sterilized
MEA plate in medium that contained 20% FBS. On the day of the
experiment, coated MEAs were connected to a head stage am-
plifier. Extracellular potentials were sampled at 50 KHz, and all
recordings were performed at 37°C. Recordings were per-
formed for 100 seconds at baseline and at 5 minutes after drug
application. FP signals were analyzed forFP duration (defined as
the interval between the minimum FP and maximum FP), inter-
spike intervals, and beating frequency. Data were analyzed
using AxoScope software (Molecular Devices, Sunnyvale, CA,
http://www.moleculardevices.com).
Patch Clamp Electrophysiology
Action potentials (APs) were recorded from spontaneously beat-
ing hESC- or hiPSC-derived CMs using the current clamp in the
whole cell patch clamp configuration. Beating spheroids were
dissociated into single cells using 0.05% trypsin-EDTA (catalog
no. 25300-054; Gibco) at 37°C for 4–5 minutes. Single beating
cardiomyocytes were plated onto glass coverslips coated with
ECM Gel (catalog no. E1270; Sigma-Aldrich) and maintained at
37°C. Before experimentation, the coverslips were transferred
to a recording chamber mounted on the stage of an Olympus
inverted microscope (catalog no. CKX41; Olympus). The extracel-
lular bath solution within the chamber contained 150 mM NaCl,
5.4 mM KCl, 15 mM HEPES, 15 mM D-glucose, 1 mM MgCl
2
, and
1.8 mM CaCl
2
; adjusted to pH7.4 with NaOH. Recording pipettes
were pulled from borosilicate glass capillaries using a P-97 hori-
zontal puller (Sutter Instrument, Novato, CA, http://www.
sutter.com) and had a tip resistance of 3–6MV. The pipette so-
lution contained 150 mM KCl, 10 mM NaCl, 2 mM CaCl
2
,5mM
EGTA, 10 mM HEPES, and 5 mM MgATP; adjusted to pH 7.2 with
KOH. Data were acquired using a multiclamp 700B amplifier (Axon
4Scalable Production of hPSC-Derived Cardiomyocytes
©AlphaMed Press 2015 STEM CELLS TRANSLATIONAL MEDICINE
by AlphaMed Press on November 5, 2015http://stemcellstm.alphamedpress.org/Downloaded from
Published Ahead of Print on October 28, 2015 as 10.5966/sctm.2014-0275.
Instruments/Molecular Devices Corp., Union City, CA, http://
www.moleculardevices.com), a Digidata 1440 analog-to-digital
board and pClamp 10 software (Axon Instruments), at a sampling
frequency of 10 kHz and low-pass filtered at 2 kHz. Data analysis
was performed using Clampfit 10 (Axon Instruments) and Prism 6
(GraphPad Software, Inc., San Diego, CA, http://www.graphpad.
com) software.
Statistical Analysis
All quantifications were performed using at least three indepen-
dent replicates. Data are presented as mean 6SD, in which data
with symmetrical (normal) or nonsymmetrical distributions
were analyzed using one-way analysis of variance, followed
by the post hoc least significant difference or Dunnett’ssignif-
icant difference test. Values of p,.05 were considered statis-
tically significant.
RESULTS
Modulation of WNT, Transforming Growth Factor-b,
and SHH Signaling Pathways to Enhance
Cardiac Differentiation
In order to develop an efficient protocol for large-scale genera-
tion of CMs from hPSCs, we attempted to manipulate the
most important cardiogenic signaling pathways by using SMs in ad-
herent culture. First, we optimized chemically induced CM differen-
tiation methods using recently reported SM cocktails [14, 15]. After
several optimization trials using multiple different co mbinatio ns
of SMs and treatment durations, we determined that the most
effective protocol for CM differentiation was exposure of hESCs to
12 mM CHIR for 1 day, followed by a 1-day rest period in the same
medium without SMs. Subsequently, IWP2, SB431542, and Pur (5
mM each) were used for an additional 2 days. The SMs inhibited
WNT/b-catenin and transforming growth factor-b(TGF-b) and
activated the SHH signaling pathways. This method led to the ap-
pearance of beating clusters in the adherent cultures at days 7–10
after the onset of differentiation (data not shown).
In order to transform the optimized adherent culture protocol
to scalable suspension culture conditions, we induced 7-day spher-
oids of hPSCs that had been generated in low-attachment dishes
under static suspension culture conditions with the optimized
protocol (Fig. 1). For differentiation, the cells were treated the
same as for the adherent culture protocol.
Daily analysis of differentiation in spheroids showed that the
first beating appeared after 10 days of differentiation induction
(Fig. 1B). To determine the optimal size of spheroids for CM differ-
entiation, 3-, 5-, and 7-day hESC spheroids (average size, 90 630,
175 625, and 250 632 mm, respectively) generated in dynamic
suspension culture were used for differentiation induction in the
static suspension system. In the case of the 3-day spheroids, cell
viability was significantly decreased; the spheroids became disrupted
and finally dispersed after CHIR treatment (data not shown). We
excluded this group from further experiments. We observed that
approximately 50% of the 5-day spheroids started beating only
7 days after differentiation induction; approximately 100% of spher-
oids formed beating structures after 10 days of differentiation in-
duction and maintained spontaneous contractile activity for at
least 30 days in culture (Fig. 1B). The average size of the 5-day spher-
oids was 175 625 mm; therefore, this was chosen as the optimal
size for the subsequent experiments in our study.
Optimization of Chemically Induced Cardiomyocyte
Differentiation in Static Suspension System
Different concentrations of CHIR have been reported that favor
mesendoderm induction. The 3D structure of spheroids might affect
the diffusion rates of CHIR throughout the cell aggregates, which
could, in turn, result in inefficient and heterogeneous differentia-
tion. The sensitivity of cells in these compact structures might be
completely different from that of cells in monolayer adherent cul-
ture. Therefore, to explore the optimal concentration of CHIR for in-
duction of hPSC aggregates, we treated the 5-day aggregates with
different concentrations of CHIR (3, 6, 9, 12, and 15 mM), and the
spheroids were assessed after 10 days of differentiation.
Our data showed that by increasing the CHIR concentration, the
percentage of beating spheroids also increased. Approximately
100% of the spheroids began beating at both 12 mMand15mM
CHIR (Fig. 1C). Flow cytometry analysis showed that a2myosin
heavy chain (aMHC) expression increased and reached a plateau
(.90%) after treatment with 12 mMand15mMCHIR(Fig.1C);this
was also supported by the quantification of beating. These results
suggested CHIR at 12 mM would be the optimal concentration for
mesoderm induction. This concentration was used in the first step
of our differentiation protocol in suspension cultures.
We sought to determine whether manipulating one or two
signaling pathways was sufficient to induce efficient CM differen-
tiation or whether the complete cocktail of three chemicals
(IWP2, SB431542, and Pur) was necessary. We examined different
combinations of these three chemicals and calculated the num-
ber of beating spheroids after 10 days of differentiation (Fig.
1D). The results indicated that treatment of cell aggregates with
SB431542 or Pur resulted in the formation of only a few beating
spheroids, and IWP2 treatment led to higher numbers of beating
spheroids (up to 20%), confirming a role for WNT signaling in car-
diogenesis in this static suspension system. The combination of
SB431542 and Pur resulted in a lower percentage of beating
spheroids compared with IWP2 and the other two chemical com-
binations. The combination of all three chemicals dramatically
increased the number of beating spheroids to approximately
100% and was identified as the most effective combination at
this step for CM differentiation in the suspension culture condi-
tion. Flow cytometry analysis of aMHC
+
cardiac cells at day 10
after differentiation initiation supported the beating results
and revealed that a large population of CMs (.90% positive
for aMHC) formed when using all three chemicals (Fig. 1D). These
results suggest that inhibition of WNT signaling is necessary at this
step for CM differentiation in suspension cell aggregates. To achieve
an efficient differentiation process, inhibition of TGF-bsignaling
and activation of SHH signaling was also required.
Gene Expression Pattern of Cardiogenic Genes in a Static
Suspension System
We determined the expression profiles of cardiac lineage-specific
genes throughout differentiation by performing quantitative
RT-PCR on spheroids collected at different time points after differ-
entiation initiation (days 0, 1, 2, 4, 6, 8, 10, 20, and 30). Rapid
induction of T (Brachyury), a mesodermal marker, and Mesp1,
a marker for the earliest steps of cardiovascular progenitor cell
specification, was observed at the first 2 days of differentiation,
confirming the role of WNT signaling in mesoderm lineage specifi-
cation (supplementalonline Fig. 1). Downregulationof these genes
occurred simultaneously with significant upregulation of later
Fonoudi, Ansari, Abbasalizadeh et al. 5
www.StemCellsTM.com ©AlphaMed Press 2015
by AlphaMed Press on November 5, 2015http://stemcellstm.alphamedpress.org/Downloaded from
Published Ahead of Print on October 28, 2015 as 10.5966/sctm.2014-0275.
cardiac progenitor transcription factors (HAND1, TBX5, ISL1,
MEF2C, and NKX2-5) at day 4 of differentiation. The expression
of GATA4 increased earlier than that of the other cardiogenic
transcription factors and displayed a steady level of expression
over time. The expression of cTNT,a-MHC,andb-MHC (struc-
tural protein-coding genes) were upregulated at day 6 of differ-
entiation and maintained until the end of the experiment.
Flow cytometry analysis of differentiated cells at day 15 sho wed
that most individualized cells (up to 90%) were cTNT
+
, and endothe-
lial cells (von Willebrand factor-positive) and vascular smooth mus-
cle cells (a2smooth muscle actin-positive) represented a small
proportion of the differentiated cells (∼3% and ∼8%, respectively;
supplemental online Fig. 2).
Taken together, these results show that CM differentiation in
the static suspension system passed through the main steps of cardio-
genesis seen in normal heart development, as well as in embryoid
body- and monolayer-based differentiation methods.
Validation of Optimized Cardiomyocyte Differentiation
for Different hPSC Lines in the Static Suspension
Culture System
To test whether our developed protocol supported robust CM dif-
ferentiation in different hPSC lines, we differentiated five hESC
and four hiPSC lines in the static suspension system using an
optimal induction procedure. Analysis of the percentage of
beating spheroids showed that almost all lines started sponta-
neous beating at day 7 of differentiation, which increased rap-
idly to a plateau, with nearly 100% beating spheroids at days
10–15 (Fig. 2A). These data demonstrate the reproducibility
and universality of our protocol for all tested hPSC lines.
Flow cytometry analysis of dissociated beating spheroids dem-
onstrated that approximately 90% of cells were cTNT
+
in
four evaluated cell lines at day 15 of differentiation (Fig.
2B ). Subsequent experiments were performed with the RH5 hESC
line.
Integrated Generation of Human CMs in a Stirred
Suspension Bioreactor
In order to develop an integrated platform for large-scale produc-
tion of human CMs, we used our optimized static suspension dif-
ferentiation strategy in a stirred bioreactor with a 100-ml working
volume. Five-day hPSC spheroids that were 175 625 mm in diam-
eter were induced in a spinner flask by replacing hPSC expan-
sion medium with the same volume of differentiation medium
(100 ml) for differentiation induction (Fig. 3A). After several rounds
of experiments, we found that the addition of 0.1% PVA and 10 mM
ROCK inhibitor Y-27632 were essential for the first 2 days of
differentiation culture to increase cell viability and spheroid
Figure 1. Experimental design and optimization of chemically induced hPSC differentiation to cardiomyocytes (CMs) in a static suspension
system. We used the human embryonic stem cell line RH5 in this step. (A): Five-day spheroids formed in suspension culture (spinner flasks)
were transferred to low-attachment dishes that contained differentiation medium. For mesoderm induction, hPSC spheroids were treated
for 1 day with 12 mM CHIR. Then, CHIR was removed, and the cells were cultured for 1 more day in differentiation media without small molecules.
At the end of this stage, precardiac mesoderm had formed. To obtain cardiac progenitors, we treated the spheroids with IWP2, SB431542, and
Pur (5 mM each) for 2 days, after which the media were renewed every 2–3 days until the end of the study (day 30). *, First beating at day 7 of
differentiation. (B): The effect of spheroid size on CM differentiation. The first beating in the 5- and 7-day spheroids was observed at days 7 and
10. All the 5-day spheroids were beating by day 10. The percentage of beating spheroids increased slowly and had reached 100% by day 10 in the
7-day spheroids. (C): Evaluation of cardiac differentiation by counting the number of beating spheroids (%) and flow cytometry analysis of
a-MHC-expressing cells (%) indicated that increasing the CHIR concentration to 12 mM resulted in increased CM differentiation. (D): Flow cytom-
etry analysis of a-MHC-positive cells (%) and calculating the number of beating spheroids (%) showed that the combination of IWP2, SB431542,
and Pur led to the most efficient CM differentiation in a static suspension system. All data are presented as mean 6SD (n= 3). Abbreviations:
bFGF, basic fibroblast growth factor; CHIR, CHIR99021; DMEM, Dulbecco’s modified Eagle’s medium; hPSCs, human pluripotent stem cells; MHC,
myosin heavy chain; Pur, purmorphamine; RPMI, Roswell Park Memorial Institute; SB, SB431542.
6Scalable Production of hPSC-Derived Cardiomyocytes
©AlphaMed Press 2015 STEM CELLS TRANSLATIONAL MEDICINE
by AlphaMed Press on November 5, 2015http://stemcellstm.alphamedpress.org/Downloaded from
Published Ahead of Print on October 28, 2015 as 10.5966/sctm.2014-0275.
integrity (data not shown). The efficiency of CM differentiation was
determined by calculating the number of beating spheroids in the
first 2 weeks of differentiation (Fig. 3B). Compared with the static
differentiation system, the first spontaneous beating appeared at
day 7. The percentage of beating spheroids increased progressively
and reached a plateau with approximately 100% beating spheroids
by day 10 (supplemental online Video 1). Immunostaining for cTNT,
MLC2a, and NKX2-5 in sections of beating spheroids collected at
day 10 showed cytoplasmic- and nuclear-localized immunola-
bel ing for cTNT and MLC 2a and NK X2-5, respectively, indicating
the efficiency of CM differentiation (Fig. 3C). Furthermore, immu-
nostaining for MLC2a and costaining for NKX2-5 and cTNT in cells
from both hESC and hiPSC lines demonstrated the nuclear accu-
mulation of NKX2-5 and defined the sarcomeric structures in dif-
ferentiated CMs (Fig. 3D).
Cardiac-Specific Gene Expression Pattern and Cardiac
Lineage Induction in a Dynamic Suspension System
We examined the pluripotency and cardiac-specific gene ex-
pression patterns and transcript enrichment in hPSCs during CM
differentiation in a dynamic culture system by performing
quantitative RT-PCR at eight time points (supplemental online
Fig. 3). The results showed that the pluripotency genes OCT4
and NANOG were significantly downregulated and the mesoder-
mal marker Twas upregulated after differentiation induction. The
highest level of MESP1 expression was observed at day 2. The ex-
pression of cardiac progenitor markers c-KIT, ISL1, and PDGFR-a
increased during differentiation, with the highest level of ex-
pression at day 8. Late cardiac progenitor markers NKX2-5, TBX5,
MEF2C, and GATA4 were upregulated exactly after differentiation
induction, had peaked at days 6–8, and remained at a steady high
level of expression until the end of the study (day 30). Cardiac
structural and maturation markers such as cTNT,aMHC, and
MLC2v began expression at day 6 and reached a plateau with sus-
tained expression until the end of study at day 30.
We grouped the pluripotency and differentiation genes into
three clusters according to their expression pattern (Fig. 4A). Cluster
1 included the pluripotency genes (OCT4 and NANOG)andmeso-
dermal marker T, with all showing downregulation during CM dif-
ferentiation. The genes in cluster 2 showed a later peak after
differentiation induction, followed by downregulation over the dif-
ferentiation process. Cluster 3 included genes that encode CM-
specific transcription factors and structural proteins. These genes
showed upregulation and sustained expression throughout
CM differentiation. Replicated samples were clustered together
in PCA that showed a clear roadmap of differentiation (Fig. 4B,
left, arrow). This map started from day 0 (hPSCs) and terminated
at day 30. Although the pluripotency genes NANOG and OCT4
were located in the same direction of the day 0 replicates,
MLC2v was found at the end of the differentiation roadmap.
These analyses demonstrated the quality of the obtained re-
sults and reproducibility of the experiment.
To quantify the percentage of cardiac lineage cells, we per-
formed flow cytometry analysis at different time points in which
the cardiac lineage genes showed the highest level of expression
(Fig.4C).T
+
mesodermal cells constituted a high fraction of spheroids
cells (∼90%) at day 2 of differentiation, and this percentage de-
creased during differentiation. MEF2C
+
cells represented ∼50%
of total cells at day 6 and had increased to approximately 80%
at day 8 of differentiation. A rapid increase in the percentage
of aMHC
+
was observed after day 6, which had increased further
to 85% by day 10. Taken together, these flow cytometry analyses
corroborated the gene expression results discussed and showed
the stepwise induction of different cardiac lineage states in a
dynamic differentiation system.
Electrophysiological Properties of Human CMs
Differentiated in a Dynamic Suspension System
We investigated the electrophysiological properties of hPSC-
derived CMs using MEA and patch clamp techniques. We sought
to determine whether integrated differentiation in a dynamic
suspension culture system will produce functional CMs. hPSC
spheroid-derived CMs in a dynamic suspension system devel-
oped spontaneous electrical activity, indicated by the typical
Figure 2. Reproducibility and cardiomyocyte (CM) differentiation pattern of human pluripotent stem cells in a static suspension system. (A):
Five-day spheroids of different human ESC (hESC) and human iPSC (hiPSC) lines followed a similar differentiation pattern in which all cell lines
showed a beating initiation point at day 7 and a plateau with nearly 100% beating spheroids at day 10. Error bars represent SD (n= 3). (B): Flow
cytometry analysis of cTnT
+
cells indicated that the efficiency of CM generation from hESCs (RH5) and hiPSCs (618-3, 913-5, and 616-1) had
reached 90% by day 15 of differentiation. Error bars represent SD (n= 3). Abbreviations: D, day; ESC, embryonic stem cell; iPSC inducible
pluripotent stem cell.
Fonoudi, Ansari, Abbasalizadeh et al. 7
www.StemCellsTM.com ©AlphaMed Press 2015
by AlphaMed Press on November 5, 2015http://stemcellstm.alphamedpress.org/Downloaded from
Published Ahead of Print on October 28, 2015 as 10.5966/sctm.2014-0275.
extracellular FPs formed at different areas of the beating spher-
oids. The presence of a sharp and a slow component of the FP
was noted (Fig. 5A). The FPs could be divided into a rapid compo-
nent, reflecting depolarization, a plateau phase, and a slow com-
ponent, representing repolarization [35]. We observed the
chronotropic response of the spheroid-derived CMs after admin-
istration of the b-agonist isoproterenol (Iso). Application of 100 nM
Iso resulted in a typical and comparable increase in the FP frequency
compared with the basal condition (Fig. 5B). Moreover, the FP
duration was significantly shortened in the Iso-treated beatingspher-
oids (525 65 at baseline vs. 485 66 at Iso treatment). Thus, the
dynamic suspension culture system supported the development
of spontaneous contractile activity and electrophysiological
functionality.
We also studied action potentials in single beating cells using
the whole cell mode of the patch clamp technique. A primary
study of single beating cells at day 30 of differentiation showed
nodal-like APs in most patched cells (supplemental online Fig. 4).
Figure 3. Experimental design of chemically induced hESC differ entiation to CMs in a dynamic system. (A): Five-day spheroids formed in suspension
culture (spinner flasks) were transferred to a new spinner flask that contained differentiation medium (Fig. 1A). ROCK inhibitor wasused for 1 day and
PVA forthe first 2 days. *, First beating at day 7 of differentiation. (B):Time course of development of spontaneously beating spheroids. The 5-day RH5
hESC spheroids began beating at 7 days after induction of different iation. The maximum number of beating spheroids (.90%) was observed at day 10.
Error bars represent SD (n=3).(C): Immunostaining of hESC-derived beating spheroids sectioned at day 30 for CM-specific markers. (D): Dissociated
beating spheroids were stained for CM-specific transcription factor (NKX2-5) and structural proteins (MLC2a and cTNT). Abbreviations: bFGF, basic
fibroblast growth factor; CHIR, CHIR99021; DAPI, 49,6-diamidino-2-phenylindole; DMEM, Dulbecco’s modified Eagle’s medium; hPSCs, human
pluripotent stem cells; Pur, purmorphamine; PVA, polyvinyl alcohol; RPMI, Roswell Park Memorial Institute.
8Scalable Production of hPSC-Derived Cardiomyocytes
©AlphaMed Press 2015 STEM CELLS TRANSLATIONAL MEDICINE
by AlphaMed Press on November 5, 2015http://stemcellstm.alphamedpress.org/Downloaded from
Published Ahead of Print on October 28, 2015 as 10.5966/sctm.2014-0275.
This classification was based on the morphology of APs
(supplemental online Fig. 4A), AP parameters (Vm ,10 V/s and
AP duration at 90% [APD
90
]of #150) [36, 37] and mean beating
rate .100 (supplemental online Fig. 4B, 4C). The application of
quinidine at 100 mM resulted in deceleration (supplemental
Fig. 4Db) of spontaneous APs followed by complete abolition
of beating, demonstrating the contribution of I
Na
(sodium cur-
rent) in the depolarization phase of these APs. The cells showed
partial recovery after washout (supplemental Fig. 4Dd). Quinidine
also caused prolongation of the AP duration, showing the contri-
bution of hERG channels in the repolarization phase of these cells.
The effect of quinidine was reversible, although it had not fully
reversed after 5 minutes, consistent with the slow dissociation
kinetics for quinidine.
AP recording from single beating hPSC-derived CMs was also
repeated at days 60 and 90 of differentiation. We observed APs of
the three main cardiac cell types in cells differentiated from hPSCs
in the dynamic suspension system (Fig. 6A–6C). Vm .10 (20.5 6
1.9 V/s) and APD
90
/APD
50
#1.6 (1.6 60.09) was measured in
morphologically ventricle-like APs, which further confirmed their
subtype(Fig. 6A). Nodal-likeAPs showed a Vm ,10 (8.7 60.96 V/s)
and APD
90
/APD
50
.1.6 but ,2 (1.8 60.03; Fig. 6C). Atrial-like APs
showed similar properties to ventricle-like APs but with shorter
AP durations (Fig. 6B). Thus, integrated differentiation of hiPSCs
Figure 4. Expression patterns of pluripotency and cardiac lineage-related markers in a dynamic suspension system. (A): Differentially
expressed genes could be grouped into three clusters. Cluster 1 included pluripotency genes (OCT4 and NANOG) and a mesodermal marker
(T) and showed a reduction in gene expression over the time course of cardiomyocyte (CM) differentiation. The genes collected in cluster 2
showed upregulation after differentiation induction followed by downregulation over the differentiation process. Cluster 3 included genes that
encode CM-specific transcription factors and structural proteins. These genes showed sustained upregulation throughout CM differentiation.
(B): Principal components analysis (PCA) of expression values in the dynamic suspension system. Left: Each point represents a replicate sample,
and the colors show the days after differentiation. Right: The location of each gene in the same PCA plot. This analysis shows the expression of
pluripotency and CM-related genes matched the days after differentiation induction. (C): Quantitative data on dissociated cells from RH5 human
embryonic stem cell spheroid showing cardiac lineage markers at different time points during differentiation in the dynamic system. Error bars
represent SD (n= 3). Abbreviation: MHC, myosin heavy chain.
Fonoudi, Ansari, Abbasalizadeh et al. 9
www.StemCellsTM.com ©AlphaMed Press 2015
by AlphaMed Press on November 5, 2015http://stemcellstm.alphamedpress.org/Downloaded from
Published Ahead of Print on October 28, 2015 as 10.5966/sctm.2014-0275.
in a dynamic suspension culture system supported the generation
of different cardiac cell types.
DISCUSSION
Effective cardiac cell therapies and the development of pharmacol-
ogy and toxicology screening platforms require large numbers of
pure and functional cardiac cells that cannot be easily produced
by most current protocols for the generation of CMs from hPSCs.
For example, to achieve an effective cardiac cell therapy 1–23
10
9
of hPSC-CMs are required for myocardial infarction (review pro-
vided in [38]). Therefore, developing flexible and robust protocols
and platforms for cost-effective, reproducible, and large-scale gen-
eration of the desired hPSC-derivative cells is necessary to realize the
great potential of hPSC-CMs for clinical and commercial applications.
To date, different CM differentiation protocols have been reported
that manipulated the WNT, TGF-b, and SHH signaling pathways
using different protein factors, matrix components, or SMs [14,
15]. However, most of the protocols resulted in large variations in
CM differentiation efficacy among different cell types and lines, such
as were experienced in our preliminary experiments. After numerous
trials, we have developed a new procedure using CHIR (12 mM, WNT/
b-catenin signaling activator) for 1 day to induce mesendodermal
differentiation, followed by a 1-day rest period, and then 2 days of
treatment with a SM cocktail composed of IWP1 (WNT/b-catenin
inhibitor), SB431542 (TGF-breceptor inhibitor), and Pur (SHH ago-
nist) to generate CMs. It is known that WNT signaling has a biphasic
role in cardiogenesis [39]. The addition of the WNT protein or IWP2
can reduce endogenous WNT heterogeneity and provide a condi-
tion for stable expansionand efficient differentiation ofWNT
high
or
WNT
low
hESC populations, respectively [40]. However, our differ-
entiation method has eliminated the variability in CM differenti-
ation between hPSCs lines reported in other related studies. This
universality of the production process will largely facilitate the
future widespread application of the protocol and development
of hPSC-CM-based therapeutic products.
More recently, another universal CM differentiation protocol
for hiPSCs was developed using a chemically defined medium that
consisted of three components—basal medium RPMI 1640, L-
ascorbic acid 2-phosphate, and rice-derived recombinant human
albumin. This protocol resulted in generation of CMs with high ef-
ficacy and productivity. However, the proposed culturing platform
was based on a 2Dadherent culture on peptide-modified surfaces,
which are very expensive and offer poor scalability [9].
To overcome the complexity and limited scalability issues of the
previously developed CM generation protocols, we have successfully
developed a cost-effective and robust process that allows large-scale
expansion of hPSCs as aggregates and their integrated differentia-
tion for production of homogenous hPSC-CM aggregates. hPSCs
were initially expanded in a stirred bioreactor as size-controlled ho-
mogenous aggregates. Next, the aggregates with different sizes
were directly induced to differentiate into CMs by a simple stepwise
protocol optimized for suspension culture under carrier-free condi-
tions without the use of additional extracellular matrix or high-cost
recombinant proteins or peptides. The time-course analysis of beat-
ing in variously sized spheroids (90 630, 175 625, and 250 632
mm) showed that cardiac induction occurred in 5-day spheroids
(175 625 mm in diameter) at an earlier time and with greater effi-
ciency than in 7-day spheroids (250 632 mm); the 3-day spheroids
disrupted after induction. These results support previous reports
that recognized the size of hPSC aggregates or EBs as a key parameter
influencing the induction efficiency of cardiac and other lineages [41,
42]. The low diffusion rate of growth factors, chemicals, and
nutrients and of oxygen gradients inside the cell aggregates might
underpin the delayed differentiation observed in the 7-day spher-
oids. However, treating cell aggregates with CHIR decreased hPSC
aggregate integrity in dynamic culture conditions and increased cell
loss after differentiation induction, explaining the disruption of 3-day
aggregates after induction. Therefore, the generation of size-
controlled aggregates in hPSC expansion cultures by controlling
the hydrodynamic culture conditions and defining the optimum
aggregate size for a specific differentiation protocol is crucial to
achieving homogenous and efficient CM differentiation.
In optimized culture conditions, a stirred bioreactor contain-
ing hPSC aggregates (80–90 310
6
of cells in a 100-ml working
volume after 5 days of culture) directly differentiated into cardi-
omyocytes with optimized differentiation cocktails and strategy.
With this approach, approximately 100% of the undifferentiated
aggregates generated cardiospheres that showed spontaneous
contractility and contained highly enriched CMs (up to 90% cTNT
+
and MHC
+
cells) with high functionality in vitro. This high purity
and functionality could facilitate the development of an inte-
grated, cost-effective purification system to generate the large-
scale hPSC-CMs required for further applications.
Although our proposed strategy provides a robust, cost-effective,
and universal platform for large-scale generation of hPSC-CMs, we be-
lieve the efficacy of the current proposed strategy could be improved
further by optimization of the bioprocess parameters in fully con-
trolled conditions. These parameters include oxygen tension, hydrody-
namic culture conditions, feeding and media refreshment strategy,
and the development of innovative SM delivery technologies to in-
crease the differentiation efficacy and homogeneity and minimize
hPSC cell loss after differentiation induction. More recently, it has been
demonstrated that optimizing the bioprocess parameters, including
oxygen tension (hypoxia), and bioreactor hydrodynamics can boost
mouse iPSC differentiation toward CMs [43]. However, these finding
should also be validated for differentiation of hPSC cell lines.
Figure 5. Extracellular field potentials recorded from beating spher-
oids using microelectrode array. (A): Microelectrode array recording
showed typical electrical properties of human embryonic stem cell
spheroid-derived cardiomyocytes (CMs). (B): Drug response of line
RH5-derived CMs showed increased chronotropy when challenged
with Iso, a b-adrenergic agonist. Abbreviations: FPD, field potential
duration; Iso, isoproterenol.
10 Scalable Production of hPSC-Derived Cardiomyocytes
©AlphaMed Press 2015 STEM CELLS TRANSLATIONAL MEDICINE
by AlphaMed Press on November 5, 2015http://stemcellstm.alphamedpress.org/Downloaded from
Published Ahead of Print on October 28, 2015 as 10.5966/sctm.2014-0275.
While our study was in preparation, Kempf et al. reported the
development of a similar small molecule CM differentiation and cul-
turing protocol applied to three hESC and hiPSC lines as aggregates
in 100-ml stirred bioreactors [43]. Their study applied sequential
CHIR99021 (7.5 mM compared with 12 mM in our study) and
IWP2 (5 mM) treatments separated by 2 days (rather than the 1
day used in our study) and did not include the TGF-bsignaling inhib-
itor SB431542 or SHH pathway inducer Pur. After scaling up the op-
timized differentiation protocol from 12-well plates to Erlenmeyer
flasks and, finally, the stirred bioreactor, the hPSCs expanded in
the batch stirred bioreactor culture as aggregates with a 283 6
9.6-mm average diameter failed to produce hPSC-CMs in the differ-
entiation phase. However, hPSC aggregates with a larger diameter
(470–530 mm for different cell lines) that were produced in a contin-
uous culture using a cyclic perfusion feeding strategy generated con-
tracting hPSC-CMs after 6–10 days of differentiation induction. In the
present study, we have demonstrated that functional and beating
hPSC-CMs can be generated in simplified batch and single-unit op-
eration from 5-day aggregates with a 175 625-mm average diame-
ter after 7 days of induction. This appears to be universally applicable,
because it was reproducible for 9 hESC and hPSC cell lines. The sim-
plified and batch operation mode of our protocol could offer an
advantage over the strategy of continuous culture using perfusion
feeding, which has a higher cost, greater complexity of process con-
trol in large-scale culture, and considerable cell loss during perfusion
(∼25%). In addition, the hPSC-CMs produced in our culture system
had a smaller average diameter (150–200 mm) compared with the
hPSC-CMs generated from the perfusion feeding strategy (about 1
mm). The smaller and more homogenous cardiosphere sizes will fa-
cilitate their downstream processing (e.g., enzymatic diss ociation
and purification) and future use for cell therapy and drug discov-
ery applications.
Regarding yield and universality, the continuous culture
mode resulted in hPSC-CM yields that were similar to but lower
than ours for the hESC line (67%–81% MHC
+
cells in three trials
after 10 days of induced differentiation), with yields apparently
more variable with the hiPSC lines (range, 27%–83% MHC
+
cells
in three trials). Further work is required to determine the practical
basis for these differences.
CONCLUSION
hPSCs likely have the potential to differentiate into all cell types of
the body and constitute an extremely attractive tool for the gen-
eration of cells for cell therapy and other biomedical applications
once the safety and scale-up issues have been overcome. An in-
tegrated, robust bioprocess for mass production of hPSC-CMs will
pave the way for commercial and clinical applications.
To date, three reports [44–46] have been published, and 11 ap-
proved clinical trials involving hPSC-based therapies have been reg-
istered at the U.S. National Institutes of Health clinical trials website
(http://www.clinicaltrials.gov), with 9 for ocular indications (8 from
hESCs, 1 from hiPSCs [47]), 1 for diabetes, and 1 for severe heart fail-
ure. Although hPSC-CMs still seem to be some distance from clinical
application, their large-scale production will also provide an oppor-
tunity for the development and refinement of screening platforms
for drug toxicity and modeling of diseases, including congenital
heart disease, long-QT syndromes, and Timothy syndrome (reviews
provided in [48, 49]). The availability of massive numbers of human
CMs will also provide broad scope for cardiac tissue engineering, in
vitro organ development, molecular cardiovascular research, and
the development of safer, more effective drugs for cardiovascular
therapies using high-throughput technologies.
The protocol we have described allows the production of hPSC-
CMs in a simplified and robust process from different hPSCs lines. It
offers advantages over currently demonstrated suspension proto-
cols, which are variously limited in scalability, complexity, affordabil-
ity, efficacy, or CM functionality. Billions of hPSC-CMs can be produced
using the proposed scale-out and scale-up strategies using hiPSCs or
hESCs as the starting cells for production and purification (Fig. 7). Op-
timizing the key bioprocessing parameters is now an imperative that
should be implemented.
ACKNOWLEDGMENTS
This study was funded by grants provided from the Royan Insti-
tute, Iranian Council of Stem Cell Research and Technology, Iran
National Science Foundation, National Health and Medical
Research Council of Australia (NHMRC; Grant 354400), National
Heart Foundation of Australia/Heart Kid Australia (Grant
G11S5629), and New South Wales Cardiovascular Research Net-
work. H.F. was supported by a University International Postgrad-
uate Scholarship from the University of New South Wales,
Australia. J.V. was supported by a Senior Research Fellowship
from the NHMRC (Grant 1019693). R.P.H. was supported by an
NHMRC Australia Fellowship (Grant 573705). We express our
gratitude to the human subjects who participated in this research.
The Victor Chang Cardiac Research Institute does not engage in,
nor does it condone, the destruction of human embryos for re-
search. Its contribution to this study was limited to work on hu-
man induced pluripotent stem cells.
AUTHOR CONTRIBUTIONS
H.F. and H.A.: experiment design, manuscript writing, cell culture,
real-time polymerase chain reaction analysis, immunocytofluorescence,
Figure 6. Action potentials(APs) recorded from single beating human
pluripotent stem cell-derivedcardiomyocytes (CMs).APs of singlebeat-
ing human embryonic stem cell-derived CMs: ventricle-like APs (A),
atrial-like APs (B), and nodal-like APs (C). The classification of different
cardiac cell types was based on the morphology of APs (left) and AP
parameters. All AP recordings were performed at room temperature.
Fonoudi, Ansari, Abbasalizadeh et al. 11
www.StemCellsTM.com ©AlphaMed Press 2015
by AlphaMed Press on November 5, 2015http://stemcellstm.alphamedpress.org/Downloaded from
Published Ahead of Print on October 28, 2015 as 10.5966/sctm.2014-0275.
flow cytometry; S.A.: experiment design, manuscript writing; M.R.L.:
human pluripotent stem cell suspension culture in bioreactor; S.K. and
S.H.: electrophysiology; A.S.Z.: real-time polymerase chain reaction
data analysis; A.B.: human induced pluripotent stem cell line genera-
tion, human induced pluripotent stem cell design; G.M.B.: clinical
translation, provision of study material or patients, human ethics
and sample procurement, fibroblast culture initiation; S.P. and
M.P.: single cell electrophysiology experiment performance, manuscript
writing; Y.O. and Y.M.: skin biopsy of subjects; J.V.: electrophysiology
data analysis, manuscript writing; M.T.: manuscript writing; D.S.W.: clin-
ical translation, provision of study material or patients, human ethics
and sample procurement; R.P.H.: human induced pluripotent stem cell
design, manuscript writing; N.A.: experiment design; H.B.: experiment
design, manuscript writing.
DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
The authors indicated no potential conflicts of interest.
REFERENCES
1Thomson JA, Itskovitz-Eldor J, Shapiro SS
et al. Embryonic stem cell lines derived from
human blastocysts. Science 1998;282:1145–
1147.
2Yu J, Vodyanik MA, Smuga-Otto K et al. In-
duced pluripotent stem cell lines derived from
human somatic cells. Science 2007;318:1917–
1920.
3Kehat I, Kenyagin-Karsenti D, Snir M et al.
Human embryonic stem cells can differentiate
into myocytes with structural and functional
properties of cardiomyocytes. J Clin Invest
2001;108:407–414.
4Mummery C, Ward-van Oostwaard D,
Doevendans P et al. Differentiation of human
embryonic stem cells to cardiomyocytes: Role
of coculture with visceral endoderm-like cells.
Circulation 2003;107:2733–2740.
5Laflamme MA, Chen KY, Naumova AV et al.
Cardiomyocytes derived from human embryonic
stem cells in pro-survival factors enhance func-
tion of infarcted rat hearts. Nat Biotechnol 2007;
25:1015–1024.
6Fonoudi H, Yeganeh M, Fattahi F et al.
ISL1 protein transduction promotes cardi-
omyocyte differentiation from human em-
bryonic stem cells. PLoS One 2013;8:
e55577.
Figure 7. An integrated, robust bioprocessing platform for large-scale production of hPSC-CMs. Production of large-scale hPSC-CMs require
hESCs (derived from human blastocysts) and hiPSCs (generated from patient fibroblasts) as starting material. Expansion of these cell lines
as single-cell inoculated suspension cultures in stirred suspension bioreactors using scale-up and scale-out strategies resulted in a few billion
to billions of undifferentiated hPSCs. The hPSCs could be easily differentiated in an integrated single-unit operation to hPSC-CMs during
10–15 days. Purification of hPSC-CMs could be further achieved by culturing the produced cells within lactate-enriched medium. The resultant
cells offer tremendous advantages for developing different applications that require large numbers of cells such as high-throughput screening
and drug discovery, in vitro organ development, and cardiac tissue engineering. Abbreviations: CMs, cardiomyocytes; hESC, human embryonic
stem cell; hPSC, human pluripotent stem cell.
12 Scalable Production of hPSC-Derived Cardiomyocytes
©AlphaMed Press 2015 STEM CELLS TRANSLATIONAL MEDICINE
by AlphaMed Press on November 5, 2015http://stemcellstm.alphamedpress.org/Downloaded from
Published Ahead of Print on October 28, 2015 as 10.5966/sctm.2014-0275.
7Zhu WZ, Hauch KD, Xu C et al. Human em-
bryonic stem cells and cardiac repair. Trans-
plant Rev (Orlando) 2009;23:53–68.
8Desbordes SC, Studer L. Adapting human
pluripotent stem cells to high-throughput and
high-content screening. Nat Protoc 2013;8:
111–130.
9BurridgePW, Matsa E, ShuklaP et al. Chem-
ically defined generation of human cardiomyo-
cytes. Nat Methods 2014;11:855–860.
10 Karakikes I, Senyei GD, Hansen J et al.
Small molecule-mediated directed differentia-
tion of human embryonic stem cells toward
ventricular cardiomyocytes. STEM CELLS TRANSLA-
TIONAL MEDICINE 2014;3:18–31.
11 Zhu S, Wei W, Ding S. Chemical strategies
for stem cell biology and regenerative medi-
cine. Annu Rev Biomed Eng 2011;13:73–90.
12 Zhang Y, Li W, Laurent T et al. Small mol-
ecules, big roles—The chemical manipulation of
stem cell fate and somatic cell reprogramming. J
Cell Sci 2012;125:5609–5620.
13 Yuan X, Li W, Ding S. Small molecules in
cellular reprogramming and differentiation.
Prog Drug Res 2011;67:253–266.
14 Gonzalez R, Lee JW, Schultz PG. Stepwise
chemically induced cardiomyocyte specifica-
tion of human embryonic stem cells. Angew
Chem Int Ed Engl 2011;50:11181–11185.
15 Lian X, Hsiao C, Wilson G et al. Robust car-
diomyocyte differentiation from human pluripotent
stem cells via temporal modulation of canonical
Wnt signaling. Proc Natl Acad Sci USA 2012;109:
E1848–E1857.
16 Minami I, Yamada K, Otsuji TG et al. A
small molecule that promotes cardiac differen-
tiation of human pluripotent stem cells under
defined, cytokine- and xeno-free conditions.
Cell Reports 2012;2:1448–1460.
17 Chen VC, Couture SM, Ye J et al. Scalable
GMP compliant suspension culture system for
human ES cells. Stem Cell Res (Amst) 2012;8:
388–402.
18 Singh H, Mok P, Balakrishnan T et al. Up-
scaling single cell-inoculated suspension cul-
ture of human embryonic stem cells. Stem Cell
Res (Amst) 2010;4:165–179.
19 Lecina M, Ting S, Choo A et al. Scalable
platform for human embryonic stem cell
differentiation to cardiomyocytes in suspended
microcarrier cultures. Tissue Eng Part C Meth-
ods 2010;16:1609–1619.
20 Niebruegge S, Bauwens CL, Peerani R
et al. Generation of human embryonic stem
cell-derived mesoderm and cardiac cells using
size-specified aggregates in an oxygen-controlled
bioreactor. Biotechnol Bioeng 2009;102:493–
507.
21 Bauwens CL, Song H, Thavandiran N et al.
Geometric control of cardiomyogenic induction
in human pluripotent stem cells. Tissue Eng Part
A 2011;17:1901–1909.
22 Hwang YS, Chung BG, Ortmann D et al.
Microwell-mediated control of embryoid body
size regulates embryonicstem cell fate via differ-
ential expression of WNT5a and WNT11. Proc
Natl Acad Sci USA 2009;106:16978–16983.
23 Nguyen DC, Hookway TA, Wu Q et al. Mi-
croscale generation of cardiospheres promotes
robust enrichment of cardiomyocytes derived
from human pluripotent stem cells. Stem Cell
Rep 2014;3:260–268.
24 Kempf H, Olmer R, Kropp C et al. Controlling
expansion and cardiomyogenic differentiation of
human pluripotent stem cells in scalable suspen-
sion culture. Stem Cell Rep 2014;3:1132–1146.
25 Hemmi N, Tohyama S, Nakajima K et al. A
massive suspension culture system with meta-
bolic purification for human pluripotent stem
cell-derived cardiomyocytes. STEM CELLS TRANS-
LATIONAL MEDICINE 2014;3:1473–1483.
26 O’BrienC, Laslett AL. Suspended in culture—
Human pluripotent cells for scalable techno lo-
gies. Stem Cell Res (Amst) 2012;9:167–170.
27 Abbasalizadeh S, Larijani MR, Samadian A
et al. Bioprocess development for mass produc-
tion of size-controlled human pluripotent stem
cell aggregates in stirred suspension bioreactor.
Tissue Eng Part C Methods 2012;18:831–851.
28 Vosough M, Omidinia E, Kadivar M et al.
Generation of functional hepatocyte -like cells from
human pluripotent stem cells in a scalable suspen-
sion culture. Stem Cells Dev 2013;22:2693–2705.
29 Tohyama S, Hattori F, Sano M et al. Dis-
tinct metabolic flow enables large-scale purifi-
cation of mouse and human pluripotent stem
cell-derived cardiomyocytes. Cell Stem Cell
2013;12:127–137.
30 Baharvand H, Ashtiani SK, Taee A et al.
Generation of new human embryonic stem cell
lines with diploid and triploid karyotypes. Dev
Growth Differ 2006;48:117–128.
31 Taei A, Hassani SN, Eftekhari-Yazdi P et al.
Enhanced generation of human embryonic
stem cells from single blastomeres of fair and
poor-quality cleavage embryos via inhibition
of glycogen synthase kinase band Rho-associated
kinase signaling. Hum Reprod 2013;28:2661–2671.
32 Larijani MR, Seifinejad A, Pournasr B et al.
Long-term maintenance of undifferentiated hu-
man embryonic and induced pluripotent st em cells
in suspension. Stem Cells Dev 2011;20:1911–1923.
33 Ihaka R, Gentleman RR. A language and
environment for statistical computing. J Com-
put Graph Stat 1996;5:299–314.
34 Wickham H. ggplot2: Elegant Graphics
for Data Analysis. New York: Springer, 2009.
35 Asakura K, Hayashi S, Ojima A et al. Im-
provement of acquisition and analysis methods
in multi-electrode array experiments with iPS
cell-derived cardiomyocytes. J Pharmacol Toxicol
Methods 2015 [Epub ahead of print].
36 Fatima A, Xu G, Shao K et al. In vitro mod-
eling of ryanodine receptor 2 dysfunction using
human induced pluripotent stem cells. Cell
Physiol Biochem 2011;28:579–592.
37 Fatima A, Kaifeng S, Dittmann S et al. The
disease-specific phenotype in cardiomyocytes
derived from induced pluripotent stem cells
of two long QT syndrome type 3 patients. PLoS
One 2013;8:e83005.
38 Serra M, Brito C, Correia C et al. Process
engineering of human pluripotent stem cells for
clinical application. Trends Biotechnol 2012;30:
350–359.
39 Ueno S, Weidinger G, Osugi T et al. Bi-
phasic role for Wnt/beta-catenin signaling in
cardiac specification in zebrafish and embryonic
stem cells. Proc Natl Acad Sci USA 2007;104:
9685–9690.
40 Blauwkamp TA, Nigam S, Ardehali R et al.
Endogenous Wnt signalling in hu man embryonic
stem cells generates an equilibrium of distinct
lineage-specified progenitors. Nat Commun
2012;3:1070.
41 Burridge PW, Anderson D, Priddle H et al.
Improved human embryonic stem cell embryoid
body homogeneity and cardiomyocyte differenti-
ation from a novel V-96 plate aggregation system
highlights interline variability. STEM CELLS 2007;25:
929–938.
42 Mohr JC, Zhang J, Azarin SM et al. The
microwell control of embryoid body size in or-
der to regulate cardiac differentiation of human
embryonic stem cells. Biomaterials 2010;31:
1885–1893.
43 Correia C, Serra M, Espinha N et al. Com-
bining hypoxia and bioreactor hydrodynamics
boosts induced pluripotent stem cell differenti-
ation towards cardiomyocytes. Stem Cell Rev
2014;10:786–801.
44 Schwartz SD, Hubschman JP, Heilwell G
et al. Embryonic stem cell trials for macular de-
generation: A preliminary report. Lancet 2012;
379:713–720.
45 Schwartz SD, Regillo CD, Lam BL et al.
Human embryonic stem cell-derived retinal
pigment epithelium in patients with age-
related macular degeneration and Stargardt’s
macular dystrophy: Follow-up of two open-
label phase 1/2 studies. Lancet 2015;385:
509–516.
46 Song WK, Park KM, Kim HJ et al. Treat-
ment of macular degeneration using embryonic
stem cell-derived retinal pigment epithelium:
Preliminary results in Asian patients. Stem Cell
Rep 2015;4:860–872.
47 Reardon S, Cyranoski D. Japan stem-cell
trial stirs envy. Nature 2014;513:287–288.
48 Sallam K, Kodo K, Wu JC. Modeling
inherited cardiac disorders. Circ J 2014;78:
784–794.
49 Dell’Era P, Benzoni P, Crescini E et al. Car-
diac disease modeling using induced pluripotent
stem cell-derivedhuman cardiomyocytes. World
J Stem Cells 2015;7:329–342.
See www.StemCellsTM.com for supporting information available online.
Fonoudi, Ansari, Abbasalizadeh et al. 13
www.StemCellsTM.com ©AlphaMed Press 2015
by AlphaMed Press on November 5, 2015http://stemcellstm.alphamedpress.org/Downloaded from
Published Ahead of Print on October 28, 2015 as 10.5966/sctm.2014-0275.
