Forced aggregation and defined factors allow highly uniform-sized embryoid bodies and functional cardiomyocytes from human embryonic and induced pluripotent stem cells

ORIGINAL ARTICLE

Forced aggregation and defined factors allow highly uniform-sizedembryoid bodies and functional cardiomyocytes from humanembryonic and induced pluripotent stem cells

Martin Pesl • Ivana Acimovic • Jan Pribyl • Renata Hezova • Aleksandra Vilotic •

Jeremy Fauconnier • Jan Vrbsky • Peter Kruzliak • Petr Skladal •

Tomas Kara • Vladimir Rotrekl • Alain Lacampagne • Petr Dvorak •

Albano C. Meli

Received: 25 May 2013 / Accepted: 25 October 2013 / Published online: 21 November 2013

� Springer Japan 2013

Abstract In vitro human embryonic stem cells (hESCs)

and human induced pluripotent stem cells (hiPSCs) can

differentiate into functional cardiomyocytes (CMs). Pro-

tocols for cardiac differentiation of hESCs and hiPSCs

include formation of the three-dimensional cell aggregates

called embryoid bodies (EBs). The traditional suspension

method for EB formation from clumps of cells results in an

EB population heterogeneous in size and shape. In this

study we show that forced aggregation of a defined number

of single cells on AggreWell plates gives a high number of

homogeneous EBs that can be efficiently differentiated into

functional CMs by application of defined growth factors in

the media. For cardiac differentiation, we used three hESC

lines and one hiPSC line. Our contracting EBs and the

resulting CMs express cardiac markers, namely myosin

heavy chain a and b, cardiac ryanodine receptor/calcium

release channel, and cardiac troponin T, shown by real-

time polymerase chain reaction and immunocytochemistry.

Using Ca2? imaging and atomic force microscopy, we

demonstrate the functionality of RyR2 to release Ca2?

from the sarcoplasmic reticulum as well as reliability in

contractile and beating properties of hESC-EBs and hiPSC-

EBs upon the stimulation or inhibition of the b-adrenergic

pathway.

Keywords Human pluripotent stem cell � Embryoid

body � Differentiation � Cardiomyocyte � Calcium

Introduction

Human embryonic stem cells (hESCs) and human induced

pluripotent stem cells (hiPSCs) have the ability to grow

indefinitely and differentiate in all somatic cell types. Both

of these cell types can be differentiated into functional

cardiomyocytes (CMs) in vitro. Differentiation of hESCs

and hiPSCs across the cardiac lineage can be achieved

using different methods: coculture with mouse visceral

endoderm-like cells (END-2 cells), formation of three-

dimensional aggregates called embryoid bodies (blasto-

cyst-like structures [1, 2]), or culturing as a monolayer on

Matrigel (BD Biosciences, San Jose, CA, USA) in medium

supplemented with defined factors. The efficiency of these

methods depends on the developmental potential of each

cell line, which may vary between lines [3–5]. While the

differentiation of CMs in embryoid bodies (EBs) has been

reported as reliable, resulting in up to 60 % contracting

M. Pesl and I. Acimovic contributed equally to this work.

Electronic supplementary material The online version of thisarticle (doi:10.1007/s00380-013-0436-9) contains supplementarymaterial, which is available to authorized users.

M. Pesl � I. Acimovic � A. Vilotic � J. Vrbsky � V. Rotrekl �P. Dvorak � A. C. Meli

Department of Biology, Faculty of Medicine, Masaryk

University, Brno, Czech Republic

M. Pesl � J. Vrbsky � P. Kruzliak � T. Kara � P. Dvorak

ICRC, St Anne’s University Hospital, Brno, Czech Republic

J. Pribyl � R. Hezova � P. Skladal

CEITEC, Masaryk University, Brno, Czech Republic

J. Fauconnier � A. Lacampagne � A. C. Meli (&)

INSERM U1046, University of Montpellier I, University

of Montpellier II, 371 Avenue du Doyen G. Giraud,

CHU Arnaud de Villeneuve, Building INSERM Crastes de

Paulet, 34295 Montpellier, France

e-mail: [email protected]

P. Skladal

Department of Biochemistry, Faculty of Science, Brno,

Czech Republic

123

Heart Vessels (2014) 29:834–846

DOI 10.1007/s00380-013-0436-9

EBs [4, 6, 7], the heterogeneity in size and shape of the

EBs may eventually contribute to inconsistency in effi-

ciency of cardiac differentiation.

By using well-characterized CCTL12, CCTL13, and

CCTL14 hESC lines [8, 9] as well as clone 4 hiPSC line

[10], we have established a method to form highly homo-

geneous EBs containing equal numbers of stem cells that

can be differentiated to CMs. This was achieved through

forced aggregation in AggreWell plates followed by three

specific steps of induction [11]. Importantly, while the first

contracting EBs appear between days 14 and 20 of dif-

ferentiation, we observed interline variability in differen-

tiating efficiency between lines. Quantitative reverse

transcription–polymerase chain reaction (qRT-PCR) and

immunocytochemistry indicate that the contracting hESC-

EBs and hiPSC-EBs of ±200 lm in diameter express

cardiac markers including atrial and ventricular myosin

heavy chain a (MYH6) and b (MYH7), and cardiac ryan-

odine receptor/calcium (Ca2?)-release channel (RYR2),

yielding a population of up to 95 % of cardiac troponin T

(cTnT)-positive cells. Using a Ca2? imaging technique and

atomic force microscopy (AFM), we observed spontaneous

Ca2? transients and demonstrated the functionality of

RyR2 in releasing Ca2? from the sarcoplasmic reticulum

(SR). We also observed reliability in contractile and beat-

ing properties of EBs from both hESCs and hiPSCs in

response to pharmacologic modulators of the b-adrenergic

pathway (i.e., isoproterenol and metoprolol). We showed

that caffeine, a RyR2 agonist, increases contraction force

(CF) in correlation with the depletion of Ca2? from the SR

and through RyR2.

Taken together, these data indicate that using both hESC

and hiPSC lines, AggreWell plates are suitable for for-

mation of highly uniform EBs with similar starting cell

numbers, which can be differentiated to contracting EBs

containing functional CMs.

Materials and methods

Maintenance of pluripotent stem cells

The CCTL12, CCTL13, and CCTL14 hESC lines [8, 9] as

well as the clone 4 hiPSC line [10] used in this study were

maintained in an undifferentiated state as colonies on the

mitomycin C (10 lg/ml; Molekula, Gillingham, UK)

inactivated mouse embryonic fibroblast (MEF) feeder

(seeded at 5 9 105 MEFs per 60-mm diameter Petri dish)

in hESC medium consisting of Dulbecco’s modified

Eagle’s medium (DMEM)/F-12 (Gibco, Grand Island, NY,

USA), 15 % KnockOut Serum Replacement (KO-SR;

Gibco), 1 % L-glutamine (Gibco), 1 % nonessential amino

acids (PAA, Pasching, Austria), 0.5 % penicillin–

streptomycin (PAA), 0.1 mM b-mercaptoethanol (Sigma,

St Louis, MO, USA) and 4 ng/ml human fibroblast growth

factor 2 (FGF2; R&D Systems, Minneapolis, MN, USA).

All cells were passaged with 1 mg/ml collagenase type IV

(Gibco). Passages were: CCTL12 (33–65), CCTL13

(34–60), CCTL14 (36–50), hiPSC clone 4 (63–96).

Formation of embryoid bodies

Prior to formation of EBs, cells treated with collagenase type

IV were plated on Matrigel-coated plates and cultured in

MEF-conditioned medium (MEF-CM) supplemented with

10 ng/ml human FGF2 and 1 % L-glutamine. The next two

passages of cells were done using 19 TrypLE Express

(Gibco). Incubation of the cells with 19 TrypLE was for 2

min at 37 �C, followed by centrifugation of cells at 170g for 4

min (Eppendorf centrifuge 5804R), and cells were grown as

monolayers on Matrigel. When cells reached 100 % con-

fluent monolayer, they were treated with 19 Accutase

(Sigma) for 4 min at 37 �C, and centrifuged at 350g for 7 min.

EB formation was induced by seeding of ±2.5 9 106 cells in

2 ml of EB medium consisting of KO-DMEM (Invitrogen,

Carlsbad, CA, USA), 10 % heat-inactivated fetal bovine

serum (Invitrogen), 1 % L-glutamine, 1 % nonessential

amino acids, 1 % penicillin–streptomycin, and 0.1 mM b-

mercaptoethanol per well on an AggreWell 400 plate

(Stemcell Technologies, Grenoble, France). EB medium was

supplemented with 25 lg/ml Y-27632 (Sigma), and forced

aggregation of cells was achieved by centrifugation of the

plate at 100g for 3 min at room temperature (23 �C). The

plate was then held under hypoxic conditions (90 % N2, 5 %

O2, and 5 % CO2) at 37 �C.

Cardiac differentiation

Twenty-four hours after seeding of the cells, formed EBs

were transferred to low-adhesive 60-mm diameter Petri

dishes (ratio 1:2) in Induction 1 medium composed of EB

medium supplemented with 10 ng/ml bone morphogenetic

protein 4 (BMP4; R&D), 5 ng/ml FGF2, and 6 ng/ml activin

A (R&D). After 4 days, medium was replaced with Induction

2 medium composed of EB medium supplemented with 10

ng/ml vascular endothelial growth factor (VEGF; R&D) and

10 lM inhibitor of Wnt response 1 (IWR-1; Sigma) for 3

days, followed by Induction 3 medium (EB medium with 10

ng/ml VEGF and 5 ng/ml FGF2). Induction 3 medium was

changed every fourth day until day 28 of differentiation, after

which EBs were kept only in EB medium without growth

factors. Dishes were held in the hypoxic conditions (5 % O2

and 5 % CO2) at 37 �C for the first 12 days of the differen-

tiation and then transferred to normoxic conditions (20 % O2,

75 % N2, 5 % CO2). First beating EBs appeared between days

14 and 20 of differentiation.

Heart Vessels (2014) 29:834–846 835

123

Quantitative RT-PCR

One-month-old beating CCTL12 hESC-EBs and undiffer-

entiated CCTL12 hESCs on Matrigel were used for RNA

extraction. Total RNA was isolated from the cells using a

PureLink RNA Mini Kit (Ambion, Carlsbad, CA, USA)

followed by cDNA synthesis using Revert Aid H Minus

First Strand cDNA Synthesis Kit (Fermentas, Gummers-

bach, Germany) according to the manufacturer’s protocols.

Real-time PCR was performed using the Applied Biosys-

tems 7500 Sequence Detection System. The 20-ll PCR

reaction mixture consisted of 10 ll TaqMan Gene Expres-

sion Master Mix (Applied Biosystems, Foster City, CA,

USA), 8 ll nuclease-free water, 1 ll primer and probe mix

of the TaqMan Gene Expression assay (MYH6, MYH7 and

RYR2; Applied Biosystems), and 1 ll cDNA. Reactions

were running on 96-well optical plates (Applied Biosys-

tems) at 50 �C for 2 min, followed by 95 �C for 10 min, and

40 cycles at 95 �C for 15 s and 60 �C for 1 min. Expression

levels were normalized to the housekeeping gene glycer-

aldehyde-3-phosphate dehydrogenase (GAPDH).

Immunocytochemistry

To dissociate cells from EBs, 10 beating EBs were col-

lected in 1 ml culture medium and washed three times in 10

ml Ca2?-free solution (120 mM NaCl, 5.4 mM KCl, 5 mM

MgSO4, 5 mM sodium pyruvate, 20 mM glucose, 20 mM

taurine, and 10 mM HEPES; pH 6.9). EBs were then

incubated in Ca2?-free solution for 20 min before centri-

fugation at 15g for 3 min. After removing the supernatant,

EBs were incubated in 1 ml digestion solution (Ca2?-free

solution supplemented with 0.8 mg/ml of collagenase from

Clostridium histolyticum (Sigma) and 0.04 mg/ml protease

from Streptomyces griseus (Sigma)) prewarmed to 37 �C,

for 30 min at 37 �C, with periodic shaking. Dissociated

cells were centrifuged at 170g for 3 min and the cell pellet

was resuspended in 1 ml KB solution (85 mM KCl, 30 mM

K2HPO4, 1 mM EGTA, 2 mM ATP-Na2, 5 mM sodium

pyruvate, 5 mM creatine, 20 mM taurine, and 20 mM

glucose; pH 7.3) prewarmed to 37 �C and incubated for 30

min at 37 �C, with periodic gentle shaking. To stain dis-

sociated CMs, cells were plated onto gelatin-coated 12-mm

diameter glass coverslips in EB medium. The fixation of

the samples was done with 4 % paraformaldehyde for 1 h

on ice and washed with 19 phosphate-buffered saline

(PBS). A blocking solution 1 (1 % bovine serum albumin

(BSA), 0.1 % Triton in 19 PBS) was used for 15 min at

room temperature and then washed out with 19 PBS. A

blocking solution 2 (1 % BSA, 0.03 % Tween, 0.01 %

NaN3 in 19 PBS) was used for 1 h at room temperature

and then washed out with 19 PBS. Incubation with pri-

mary antibody anti-human cTnT (1:500, Cell Signaling,

Danvers, MA, USA) took place overnight at 4 �C. Incu-

bation with donkey anti-rabbit Alexa 594 secondary anti-

body (1:500, Invitrogen) was for 1 h at room temperature.

Samples were then mounted in Mowiol–diamidino-2

phenylindole (DAPI; Sigma) and were observed under a

living-cell confocal Carl Zeiss LSM 700 microscope

(Oberkochen, Germany). Images were analyzed using

ImageJ, version 4.6 (rsweb.nih.gov/ij/). For counting of

cTnT-positive cells, images with DAPI and cTnT signals

were overlapped. The total number of nuclei and cTnT-

negative cells were counted manually and were used to

calculate the percentage of cTnT-positive cells. cTnT-

positive cells were scored in each experiment on at least

three slides. The average percentage of positive cells and

standard deviation were calculated from a minimum of 500

cells per experiment.

For immunostaining of cut sections of EBs, beating EBs

were embedded in 1 % low melting agarose (Promega,

Madison, WI, USA) in 19 PBS and stored at -80 �C until

sectioning. Prior cryosectioning (Leica CM 1850; Leica,

Wetzlar, Germany), EBs in low melting agarose were

frozen in Jung Tissue Freezing Medium (Leica). Cryosec-

tions (5 lm) were fixed in 4 % paraformaldehyde for 1 h on

ice and washed with 19 PBS. A blocking solution 1 (1 %

BSA, 0.1 % Triton in 19 PBS) was used for 15 min at

room temperature and then washed out three times for 5

min with 19 PBS. A blocking solution 2 (1 % BSA, 0.03 %

Tween in 19 PBS) was used for 1 h at room temperature

and then washed out with 0.05 % Tween in 19 PBS.

Samples were incubated with primary antibody anti-human

cTnT (1:200, Santa Cruz Biotechnologies, Santa Cruz, CA,

USA) for 1.5 h at room temperature, washed three times for

5 min with 0.05 % Tween in 19 PBS, and incubated with

donkey anti-goat Alexa 488 secondary antibody (1:400,

Invitrogen) for 1 h at room temperature. After incubation

with secondary antibody, samples were incubated for 5 min

with DAPI (1:1000 in 19 PBS; Sigma), washed three times

for 5 min with 19 PBS, and mounted in Mowiol (Sigma).

All solutions were administered in drops. Pictures of the

samples were taken using a living-cell confocal Carl Zeiss

LSM 700 microscope.

Measurements of cytosolic Ca2? variation

Cells were incubated in a Tyrode’s solution (135 mM

NaCl, 10 mM HEPES, 5.4 mM KCl, 0.9 mM MgCl2; pH

7.4) containing 1 mM CaCl2 and loaded with 3–5 lM

Fluo4-AM (Invitrogen) for 20 min. Cells were then placed

in an experimental chamber on the stage of an inverted

microscope. Ca2? images were recorded with a laser

scanning confocal microscope (Zeiss LSM Exciter, 409

water immersion objective, in x–y mode, 1 image/0.495 s).

To enable comparisons between cells, changes in the

836 Heart Vessels (2014) 29:834–846

123

Fluo-4 fluorescence signal (DF) were divided by basal

fluorescence (F0) using ImageJ. Ca2? transient amplitudes

were analyzed by Prism (version 6.0; GraphPad, San

Diego, CA, USA).

Atomic force microscopy

Beating EBs were plated on a 60-mm diameter Petri dish in

the presence of Tyrode’s solution. Prior to the experiments,

1.8 mM CaCl2 and 10 mM glucose were added freshly to

the solution maintained at 37 �C. To avoid the cantilever

thermic deflection, the parts of AFM located in liquid were

heated up to 37 �C and maintained for 15 min. Landing

procedure was finalized with the deflection (DFL) param-

eter changed to value of 2 nA, which was proportional to the

force value of 25 nN (force acting on the tip surface). In all

AFM experiments, we used a silicon nitride triangular-lever

probe SNL-10 (Bruker AFM Probes, Camarillo, CA, USA)

with a silicon cantilever of very low spring constant (0.06

N/m). Data acquisition was performed using an AFM

microscope (NTgra Vita) equipped with scanning by probe-

measuring head (NT-MDT, Zelenograd, Russia). Temper-

ature inside the measuring chamber was driven by Petri dish

thermostatic module SU045NTF (NT-MDT). Nova soft-

ware version 1.0.26.1297 (NT-MDT) was used to drive the

AFM microscope, capture the data, and partially evaluate

the measured data.

According to Hook’s law, the cantilever deflection

(bending) is proportional to the force acting between the

EB and cantilever:

F ¼ �k � Dh ð1Þ

where F is the acting force, k is stiffness of the cantilever

material, and Dh is the change of cantilever position caused

by the force. The recorded DFL–height calibration curve

was fitted using the Boltzmann function allowing simple

calculation of Dh (height):

DFL ¼ A1 � A2

1þ eðh�h0=dhÞ þ A2 ð2Þ

where A1 (initial value), A2 (final value), h0 (center), and dh

are the parameters of regression function, height h is the

function argument, and DFL is the dependent variable. The

data (DFL vs time curves) recorded by AFM microscope

were analyzed by Microcal Origin software, version 8.07

(OriginLab, Northampton, CA, USA), using an internal

peak-fitting module, whereby the proper peaks in the

recorded curves can be marked by setting peak orientation,

width, height, and minimum height as the peak parameters.

Position of peak maxima in time was exported and used for

further analysis of beat frequency and force (using Eqs. 1

and 2, and cantilever spring constant k equal to 0.06 N/m).

Young’s modulus describing stiffness of the cells in cluster

was calculated from the measured force-distance (FD)

curves (DFL vs height). DFL was further transformed to

the force values and fitted with the Hertzian model,

described by the equation:

F ¼ 4Ka2

2ð1� vÞp tan Hð3Þ

where F is the force acting on cell surface, K is Young’s

modulus, a is the depth of indentation, m is Poisson’s ratio

(equal to 0.5 for incompressible materials), and H is the

angle between tip walls and EB surface.

Statistical analysis

Statistical comparisons were performed with the nonpara-

metric Wilcoxon signed-rank test or the Mann–Whitney test

when specified, using Prism (version 6.0; GraphPad). Data

are expressed as mean ± standard error of the mean (SEM),

and differences were considered significant at P\0.05. To

compare percentage of contracting EBs in all lines, one-way

analysis of variance (ANOVA) followed by post hoc Bon-

ferroni analysis with a two-sided method was used.

Results

Efficiency in generating contracting EBs from hESCs

and hiPSCs

EBs were derived from CCTL12, CCTL13, and CCTL14

hESC lines [8, 9] and clone 4 hiPSC line [10] using stem

cells grown on Matrigel-coated plates. EBs were formed

with the same initial amount of stem cells per EB (2000

cells) by centrifugation of AggreWell 400 plates. Forced

aggregation by a mild centrifugation step (100g, 3 min at

room temperature) using AggreWell 400 plates and 24 h of

hypoxic incubation allows highly uniformly sized and

shaped EBs of ±200 lm in diameter (Fig. 1a).

To quantify the efficiency in obtaining beating EBs from

the different stem cell lines, we adapted a protocol of

cardiac differentiation with three steps of induction under

hypoxia using defined cytokines, as previously published

(Fig. 1b) [11]. EBs from all tested cell lines started to beat

from days 14 to 20 of differentiation.

To evaluate the efficiency of our method using Aggre-

Well 400 plates, we counted manually the number of

beating EBs over the total number of EBs per plate and per

line, and at least three times. From days 20 to 30 of dif-

ferentiation, CCTL12 hESC, CCTL13 hESC, and clone 4

hiPSC lines exhibit between 33 and 46 % of beating EBs

on average (CCTL13 hESC mean 34 ± 6 %, n = 6 versus

clone 4 hiPSC mean 44 ± 5 %, n = 6; not significant;

Heart Vessels (2014) 29:834–846 837

123

Fig. 1c). CCTL14 hESC line delivers fewer beating EBs on

average, with high variability, and significantly different to

the CCTL12 hESC line (CCTL12 mean 46 ± 6 %, n = 3

versus CCTL14 mean 10 ± 9 %, n = 3; P\0.05; Fig. 1c),

whereas EBs that were not treated with growth factors

produced \1 % (data not shown). Thus, while interline

variability is observed, CCTL12 hESC and clone 4 hiPSC

lines display the highest rate of contracting EBs.

Fig. 1 a Formed embryoid bodies (EBs; from CCTL13 hESC line) after

24 h on AggreWell 400 plate and after transfer to low-adhesive 60-mm

diameter Petri dish. b Schematic outlines of cardiac differentiation

protocol. c Percentage of beating EBs over the total number of EBs formed

from days 20 to 30 of differentiation. For each line displayed, 3–6 dishes

were used for counting. Data presented as mean ± SEM. *P\0.05. BMP4,

bone morphogenetic protein 4; FGF2, fibroblast growth factor 2; VEGF,

vascular endothelial growth factor; IWR1, inhibitor of Wnt response 1

838 Heart Vessels (2014) 29:834–846

123

We then tried to form EBs using V96 microwell plates,

which are used to generate beating EBs from many lines

without interline variability, as previously published [4].

When using V96 microwell plates with similar conditions of

media and cell density, EBs do not form (data not shown).

Interestingly, we noticed that the 24-h postformation

density of plated EBs per Petri dish modulates the ratio of

‘‘attached’’ EBs over floating EBs. Thus, when ±500 EBs

are placed in a low-adhesive 60-mm diameter Petri dish,

more than 90 % of them result in floating EBs eventually.

A density of ±1000 EBs per low-adhesive 60-mm Petri

dish results in more than 50 % of ‘‘attached’’ EBs over the

total number of EBs (data not shown). These observations

indicate the importance of EB density, in agreement with

observations made for murine ESCs [12].

To assess relative expression of cardiac markers in

1-month-old EBs, we performed qRT-PCR to detect specific

atrial myosin heavy-chain a (MYH6) and ventricular myosin

heavy-chain b (MYH7). We also evaluated the expression of

cardiac ryanodine receptor (RYR2). The comparison of the

relative expression of these three genes between CCTL12

hESC-EBs and the corresponding undifferentiated stem cells

indicates that all genes are significantly upregulated (Table 1;

Fig. 2a). Correspondingly, CCTL12 hESC-EBs at day 30 of

differentiation exhibit significantly higher levels of MYH6,

MYH7, and RYR2 mRNA with a 98,661-fold, 507-fold, and

38-fold increase, respectively.

In addition, we immunostained enzymatically dissoci-

ated cells from the contracting EBs by using an antibody

against the sarcomeric cTnT and detected its expression

(Fig. 2b). Dissociated cells from beating CCTL13 hESC-

EBs are cTnT-positive, exhibiting a typical striated pattern

[13, 14]. However, we observed that these cells exhibit

heterogeneous shapes (e.g., rectangular, triangular, or

round), likely due to the EB spheroid spatial organization

as previously described [15] (Fig. 2c). We then counted the

number of cTnT-positive cells over the total number of

cells and based on the number of DAPI-stained nuclear

DNA, and found a cardiac efficiency of 95.2 ± 0.9 % (n =

3) for the CCTL13 hESC line. To confirm that most of cells

contained in beating EBs are CMs, we also immunostained

cryosections of beating EBs for cTnT. Cut sections of

beating EBs appear to contain mainly cTnT-positive cells,

confirming the percentage of cTnT-positive cells obtained

from dissociated EBs (Fig. 2d). Taken together, these

results indicate that our beating EBs are mostly composed

of CMs expressing typical cardiac markers.

Spontaneous intracellular Ca2? transients

in hESC-CMs and hiPSC-CMs are RyR2 dependent

To monitor the spontaneous intracellular Ca2? transients

and assess the functionality of the SR in hESC-CMs and

hiPSC-CMs, we performed laser scanning confocal

microscopy on enzymatically dissociated CMs. Cells were

first loaded with a fluorescent Ca2? indicator (Fluo-4AM)

to monitor these transients. Under control conditions, with

1 mM CaCl2 in the external compartment and without any

stimulation, clusters of enzymatically dissociated hESC-

CMs exhibit heterogeneous and spontaneous rhythmic

Ca2? transients (Fig. 3a, b). In several clusters, diffusion of

Ca2? transients can be seen between CMs composing a

cluster, suggesting the existence of cardiac electrical cou-

pling. When performed on enzymatically dissociated single

hESC-CMs, spontaneous Ca2? transients occur more fre-

quently on average as shown in Fig. 3d (mean frequency of

clusters of 0.15 ± 0.01/s, n = 10, versus mean frequency of

single CM of 0.38 ± 0.03/s, n = 3, P \ 0.01). A similar

phenomenon was observed using clusters of hiPSC-CMs

and single hiPSC-CMs (mean frequency of clusters of 0.13

± 0.02/s, n = 7, versus mean frequency of single CM of

0.30 ± 0.04/s, n = 6, P\0.05; Fig. 3c, d). Overall, there is

no difference in frequencies of spontaneous Ca2? transients

between hESC-CMs and hiPSC-CMs, indicating a similar

maturity for both cell types to release Ca2? spontaneously

from the SR stores.

To assess the SR Ca2? loading, we used caffeine as a RyR2

agonist. Exposure to 10 mM caffeine induces a quick eleva-

tion of Ca2? transient immediately followed by absence of

Ca2? events (Fig. 3a, b) in both types of CMs. These results

suggest that the recorded spontaneous Ca2? events are mainly

initiated by RyR2 dependent and not by inositol triphosphate

receptor (IP3R) activation that could, however, be second-

arily activated. These data also suggest that hESC-CMs and

hiPSC-CMs contain a high amount of Ca2? in the intracel-

lular SR stores that is totally depleted upon caffeine-induced

RyR2 activation, consistent with previous reports [16, 17].

hESC- and hiPSC-derived CMs exhibit consistent

and similar contractile and beating properties

We then focused on the functional maturity of our beating

EBs by measuring their mechanical properties rather than

their electrophysiologic properties, which do not imply that

the resulting CMs actually contract properly. We set up an

AFM-based method to monitor the contracting force (CF)

and beat rate (BR) properties among the homogeneous

beating EBs from hESC and hiPSC lines. Thus, by gently

touching the whole beating EB with the AFM cantilever, we

were able to measure CF and BR in physiologic solution

(Tyrode solution with 1.8 mM CaCl2 at 37 �C) (Fig. 4a).

Under basal conditions, hESC-EBs and hiPSC-EBs contract

similarly on average (31 ± 7 vs 39 ± 9 nN), both with sub-

stantial reliability (Fig. 4b). hESC-EBs exhibit a BR (i.e.,

beats per unit of time) with an average of 51 ± 5 beats/min,

which is significantly slower than the mean BR of hiPSC-EBs

Heart Vessels (2014) 29:834–846 839

123

Table 1 mRNA levels of MYH6, MYH7, and RYR2 in CCTL12 hESC-derived contracting EBs (EBc) and undifferentiated CCTL12 hESCs (SC)

Gene EBc (n = 4)b SC (n = 4)b Fold change P valuea

MYH6 4.128 ± 1.375 4.184e-005 ± 1.044e-005 98661 0.0286

MYH7 0.004092 ± 0.001419 8.057e-006 ± 6.900e-006 507 0.0286

RYR2 0.02429 ± 0.008714 0.0006365 ± 0.0002177 38 0.0286

a According to Mann–Whitney testb Expression levels (mean ± SEM) were normalized to glyceraldehyde-3-phosphate dehydrogenase expression

Fig. 2 a Gene expression in contracting CCTL12 hESC-EBs (EBc) and

undifferentiated CCTL12 hESCs (SC). MYH6 (n = 4), MYH7 (n = 4),

and RYR2 (n = 4) are overexpressed in EBc in comparison with SC.

Expression levels were normalized to glyceraldehyde-3-phosphate

dehydrogenase expression. *P \ 0.05, according to Mann–Whitney

test. b Representative immunocytochemistry of cardiac troponin T

(cTnT) in CCTL13 hESC-derived cardiomyocytes (CMs) from enzy-

matically dissociated beating EBs. The cTnT is stained in red while the

nuclear DNA is stained in blue (diamidino-2 phenylindole (DAPI)). The

scale bar represents 100 lm. c CCTL13 hESC-derived CMs from

enzymatically dissociated beating EBs visualized under higher mag-

nification. The cTnT is stained in red while the nuclear DNA is stained

in blue (DAPI). The scale bar represents 20 lm. d Cryosection of

CCTL13 hESC beating EBs stained for cTnT (green) and nuclear DNA

(blue). Both panels show the immunostaining performed on cryosec-

tions of CCTL13 hESC beating EBs. For immunostaining experiments,

negative controls were used with significantly lower signal. The scale

bar represents 50 lm

840 Heart Vessels (2014) 29:834–846

123

Fig. 3 a Ca2? imaging showing cytosolic Ca2? events in dissociated

clusters of hESC-CMs. Representative images recorded in a series of

acquisitions showing variation in fluorescence intensity in some cells

with the time (t) in seconds. The time courses of fluorescence, relative

to the colored dots (green, blue, and red) and normalized to the basal

level, show spontaneous cytosolic Ca2? elevations. The bottom right

image represents the cytosolic Ca2? elevations upon 10 mM caffeine.

The scale bar at the bottom right is 50 lm. Illustrations of

spontaneous Ca2? transients in dissociated clusters of hESC-CMs

followed by SR Ca2? depletion induced by 10 mM caffeine are shown

below the images. b Bar graph representing the normalized fluores-

cence intensity related to spontaneous Ca2? transients (white bar, n =

11) and caffeine-induced Ca2? transients (black bar, n = 11). c Ca2?

imaging showing spontaneous cytosolic Ca2? events in dissociated

clusters of hiPSC-CMs. Representative images recorded in a series of

acquisitions showing variation in fluorescence intensity in some cells

with the time in seconds. The time courses of fluorescence, relative to

the colored dots (green, light blue, dark blue, red) and normalized to

the basal level, show spontaneous cytosolic Ca2? elevations. d Bar

graph summarizes the frequency (per second) of spontaneous Ca2?

transients in dissociated clusters of hESC-CMs (white bar, n = 10),

single dissociated hESC-CMs (white and striped bar, n = 3),

dissociated clusters of hiPSC-CMs (gray bar, n = 7), and single

dissociated hESC-CMs (gray and striped bar, n = 6). Data are

presented as mean ± SEM. *P \ 0.05, **P \ 0.01

Heart Vessels (2014) 29:834–846 841

123

(74 ± 7 beats/min, Fig. 4c). Thus, these data indicate

consistent CF and BR and reliability in the recording

parameters, and difference in mean BR between hESC-EBs

and hiPSC-EBs (Fig. 4b, c and Supplementary Movie 1).

Furthermore, we observed a 10–20 % change in both CF

and BR values when the temperature was 37 ± 0.5 �C

while a drop in frequency of more than 50 % was observed

when the temperature was 32 �C (data not shown). This

variation is likely due to the high thermal sensitivity of the

resulting CMs forming the contracting EBs, which is

known to slow down contractions and the intracellular

Ca2? release [18–20].

Because the ‘‘fight-or-flight’’ response is an important

mechanism occurring in the heart under stress conditions,

we tested the b-adrenergic pathway in the beating EBs.

Thus, addition of metoprolol (specific b1-receptor inhibi-

tor) significantly reduces both CF and BR in comparison

with basal conditions (BC) in hiPSC-EBs (mean CF 13 ± 3

nN and mean BR 23 ± 6 beats/min, Fig. 4b, c). Similarly,

metoprolol induces a trend toward a decrease in both CF

and BR in hESC-EBs in comparison with basal conditions

(Fig. 4b, c), with no difference when compared with

hiPSC-EBs. Addition of isoproterenol (b-receptor agonist)

20 min after metoprolol induces an increase of both

Fig. 3 continued

842 Heart Vessels (2014) 29:834–846

123

parameters when compared with metoprolol-treated EBs,

and actually recovers both parameters to the values

obtained under BC in hESC-EBs and hiPSC-EBs (Fig. 4b,

c). No difference was observed in mean BR between

hESC-EBs and hiPSC-EBs upon addition of isoproterenol

(Fig. 4c).

When the millimolar range (1–5 mM) of isoproterenol

was applied directly on the contracting hESC-EBs, we

Fig. 4 a Illustrative scheme of the typical pattern seen on recording

contraction force (CF) and beat rate (BR) of a cardiac EB in contact

with the AFM cantilever tip. The contraction of the cardiac EB

corresponds to the peaks in the trajectory, initially measured as an

intensity of current (nA), while the x-axis corresponds to the time in

seconds. The height and reciprocal of beat-to-beat separation of peaks

correspond to the force and frequency of EB beat, respectively. The

dashed red signal represents the basic noise from the plastic dish.

b Bar graph summarizing the CF of hESC-EBs and hiPSC-EBs under

basal conditions (white bars), in the presence of 70 lM metoprolol

(white bars with oblique lines), and in the presence of 1 mM

isoproterenol (grey bars). The number in each bar indicates the

number of experiments. c Bar graph summarizing the BR (beats/min)

of hESC-EBs and hiPSC-EBs under basal conditions (white bars), in

the presence of 70 lM metoprolol (white bars with oblique lines), and

in the presence of 1 mM isoproterenol (gray bars). The number in

each bar indicates the number of experiments. d Graph summarizing

the normalized CF of hESC-EBs and hiPSC-EBs prior and after

treatment with 1–5 mM isoproterenol (Iso). For each condition, the

CF is normalized by the maximal value. From 4 to 5 EBs were tested

per condition. e Graph summarizing the normalized BR of hESC-EBs

and hiPSC-EBs prior to and after treatment with 1–5 mM isoprote-

renol (Iso). For each condition, the BR is normalized by the maximal

value. From 4 to 5 EBs were tested per condition. f Bar graph

summarizing the CF of hESC-EBs and hiPSC-EBs prior to caffeine

(white bar), during 1 mM caffeine-induced peak (white bar with

oblique lines), and after caffeine-induced peak (gray bar). The

number in each bar indicates the number of experiments. g Bar graph

summarizing the BR (beats/min) of hESC-EBs and hiPSC-EBs prior

to caffeine (white bar), during 1 mM caffeine-induced peak (white

bar with oblique lines), and after caffeine-induced peak (gray bar).

The number in each bar indicates the number of experiments. Data

are presented as mean ± SEM. *P \ 0.05, **P \ 0.01

Heart Vessels (2014) 29:834–846 843

123

observed a slight increase, although nonsignificant, in both

normalized mean CF and BR when compared with basal

conditions, while a similar effect of isoproterenol was

observed in CF of hiPSC-EBs but with no significant

change in BR (Fig. 4d, e).

Taken together, these results indicate a functional b-

adrenergic pathway in both contracting hESC-EBs and

hiPSC-EBs, with no difference between the two types of

stem cell clusters.

We then sought to evaluate the effect of caffeine on the

contractile properties of the EBs, in relation with the SR

Ca2? depletion effect observed in dissociated CMs (Fig. 3).

It is believed that caffeine potentiates CF in skeletal and

cardiac muscle by penetrating the cell membrane and

potentiating Ca2? release from the SR, which contributes

to its stimulating and ionotropic effect [21–23]. When

hESC-EBs and hiPSC-EBs are treated with 1 mM caffeine,

a quick and pronounced increase in the CF immediately

occurs, followed by a progressive decrease of the CF (e.g.,

mean CF 20 ± 8 nN prior to caffeine, 100 ± 42 nN under

caffeine, 19 ± 7 nN post caffeine for hiPSC-EBs, Fig. 4f).

Interestingly, in this particular case application of 1 mM

caffeine does not change the BR (e.g., mean BR 45 ± 20

nN prior to caffeine, 58 ± 5 nN under caffeine, 62 ± 5 nN

post caffeine for hiPSC-EBs, Fig. 4g), suggesting that the

molecular mechanisms underlying the repeatable beats are

not affected by the caffeine-induced RyR2 activation.

These data indicate that caffeine quickly causes an increase

of the contraction of the CMs, due to the release of SR

Ca2? through RyR2, followed by a progressive decrease in

contraction, due to the depletion of Ca2? in the SR stores as

well as the Ca2? efflux through the Na?/Ca2? exchanger at

the sarcolemma. Thus, these data are then in agreement

with the SR Ca2? depletion and the absence of Ca2? events

after caffeine application observed by fluorescent micros-

copy in dissociated CMs (Fig. 3).

Discussion

In the this study, we present a forced aggregation method

using AggreWell 400 plates as a relatively homogeneous

starting point for stem cells. Using the same number of

cells offers the advantage to form uniformly sized and

shaped EBs. These EBs can be differentiated to beating

EBs upon application of defined factors with up to 50 %

efficiency from both hESCs and hiPSCs. Similarly to pre-

vious studies, we observed significant variability between

stem cell lines (hESC and iPSC) on differentiation in CMs

[24]. Our results indicate that the CMs composing the

beating EBs express specific cardiac markers, and we

found that in fact approximately 95 % of total number of

cells is positive for cTnT. Using fluorescent confocal

microscopy, we observed that enzymatically dissociated

CMs release Ca2? spontaneously from the intracellular SR

and mainly via the cardiac RyR2. Our results also dem-

onstrate that the resulting homogenous EBs, from both

hESC and hiPSC, can be used to investigate the contracting

force and the beating rate upon pharmacologic treatment in

a reliable manner. Thus, our results indicate that these EBs

can serve as suitable models for drug development and

testing because of their homogeneity.

We tested for the first time CCTL12, CCTL13, and

CCTL14 hESC lines [8, 9] for their cardiac differentiation

capacity. Our results indicate a substantial variability in

these lines, with the best efficiency of CM differentiation

for CCTL12 hESC line over CCTL13 and CCTL14 hESC

lines, estimated by the number of beating EBs following

treatment with specific induction media as previously

published [11]. Furthermore, the clone 4 hiPSC line used in

this study exhibits an efficiency similar to that of CCTL12

hESC, confirming its suitability for cardiac differentiation

as previously published [10].

Some groups revealed that the differentiation efficiency

is dependent of the EB size and the extracellular environ-

ment [25, 26]. Another group reported that some hESC

lines have a higher rate of cardiac differentiation in small

EBs [27]. Thus, it is reasonable to assume that a lower

amount of cells (i.e., \2000 cells per EB) might give a

higher percentage of CMs. In agreement with the afore-

mentioned published results, our preliminary and unpub-

lished data indicate that the smallest EBs developed using

the suspension method have a higher probability to beat in

comparison with the largest ones. While the suspension

method is cheaper than the AggreWell method, the

resulting EBs exhibit a high heterogeneity in shape and

size, which limits standardized protocols [28].

We also observed that the density of freshly formed EBs

is an important factor that leads to a different ratio of

‘‘attached’’ EBs versus floating EBs in low-adhesive Petri

dishes. Although counting of beating ‘‘attached’’ EBs is

easier to perform for practical reasons, they quickly lose

their three-dimensional shape. In the same way, a study has

reported that high density of EBs facilitates the cardiac

differentiation as well as the stimulation of BMP signaling

in murine ESCs [12].

It must be noted that among the different published

methods to generate human CMs through EB formation,

the composition of the media strongly varies. While the

universal system from Zambidis’ group uses, among oth-

ers, poly(vinyl alcohol), insulin and RPMI, and fetal bovine

serum (FBS) in the late phase of induction [4], in this study

we used, among others, FBS in media at early stage, activin

A, and VEGF as previously published [11]. Of note, our

EB formation medium contained FBS instead of KO-SR,

which was recently shown to improve cardiac

844 Heart Vessels (2014) 29:834–846

123

differentiation and contractility of CMs [29]. Endogenous

Wnt/b-catenin signaling is required for cardiac differenti-

ation of hESCs at an early stage of differentiation protocol

directed by BMP4 and activin A, while late inhibition of

Wnt/b-catenin pathway enhanced cardiogenesis. Thus,

while Keller’s group uses the Dickkopf 1 (Dkk1) protein as

a Wnt-signaling inhibitor through its specific affinity for

LRP6 [11], we used IWR-1, which acts as an inhibitor of

Wnt response 1. Moreover, it has been shown that the late

inhibition of Wnt signaling improves the efficiency of

BMP4 signaling-dependent cardiac differentiation from

hESCs and hiPSCs [30].

In this study, we demonstrated that the intracellular

Ca2? release through RyR2 channels appears functional in

both hESC and hiPSC lines, although further studies are

needed to fully characterize the Ca2? handling at the SR

level in these cells. For instance, other groups have shown

that the intracellular Ca2? homeostasis is immature in

hiPSC-CMs in comparison with hESC-CMs [17]. Further

evaluation of the contribution of both RyR2 and IP3R in

the calcium signaling of hESCs and hiPSCs is important,

since it is known that IP3R are highly expressed in

development and heart failure [31, 32]. Using AFM, we

verified that the inhibition of the b-adrenergic receptors

decreases the CF and BR in both hESC-CMs and hiPSC-

CMs composing the contracting EBs. The fact that

metoprolol inhibits both parameters without any prior b-

adrenergic stimulation suggests a high basal adrenergic

tonus in our hESC-EBs and hiPSC-EBs. Such a tonus

could also explain the mild effect of isoproterenol in our

experiments, in addition of the penetration rate of the

drugs into the multilayer EBs. This aspect needs further

exploration. Furthermore, we observed that caffeine

induces a trend toward increased CF but not BR in both

hESC-EBs and hiPSC-EBs, followed by a decrease. This

effect is likely due to the quick increase of cytosolic Ca2?

and by the final depletion of Ca2? from the SR stores [33,

34]. Overall, our AFM data match with those of the Ca2?

imaging upon caffeine treatment. However, it should be

considered that the effect of caffeine on EBs overall must

be relative to the accessibility of the contracting CMs

within the EB structure, also composed of extracellular

matrix.

Recently, Liu et al. [15] used a similar method to

investigate the mechanobiological properties of hESC-CMs

and hiPSC-CMs. Our results demonstrate that the EBs

formed using AggreWell 400 plates generate consistent

contractile and electrical properties with or without phar-

macologic stimulation. Of note, our results showed one-

order higher force of the EBs when compared with the

results from Liu et al. [15]. The same group has published

typical noise levels during force measurements of around

20 pN. This value is unusually low, as similar values are

usually obtained for interaction of single biomolecules.

Such a force for movement (noise connected with move-

ment) of a cell cluster is unexpected. The difference in

measurements can be due to the size of the EBs as well as

their final amount of CMs. Our data also indicate that the

beating EBs are highly sensitive to temperature variation,

suggesting that a constant temperature is a major factor in

maintaining reliability in recordings. This observation may

be related to alternative measurements that calibrate body

temperature as 1 �C per 10 % rise in heart beat (pulse-wave

frequency).

Finally, from a clinical point of view, the present study

offers a method to generate a high quantity of functional

CMs that could be potentially used for clinical application

and drug screening. Our data thus indicate the high

reproducibility of experiments on the resulting homoge-

neous EBs.

Acknowledgments We would like to thank Dr Livia Eiselleova,

Stanislava Koskova, Professor Ales Hampl, Dana Stritecka, and Eva

Peslova for their assistance, as well as Professor Majlinda Lako for

kindly providing the clone 4 hiPSC. This work was supported by

grants from the Ministry of Education, Youth, and Sports of the

Czech Republic (CZ.1.07/2.3.00/20.0011 and MSM0021622430),

project FNUSA-ICRC (no. CZ.1.05/1.1.00/02.0123) from the Euro-

pean Regional Development Fund, SoMoPro—Marie Curie

Actions—South Moravian Region, and by the European Society of

Cardiology (ESC) to Albano C. Meli. The research leading to these

results obtained a financial contribution from the European Com-

munity within the Seventh Framework Program (FP/2007-2013)

under Grant Agreement No. 229603. This work was supported by

CEITEC—Central European Institute of Technology (CZ.1.05/1.1.00/

02.0068) from the European Regional Development Fund. Albano C.

Meli was supported by a French Muscular Dystrophy Association

Research Grant (AFM). Ivana Acimovic was supported by a PLU-

RICELL grant (CZ.1.07/2.3.00/20.0011).

Conflict of interest There is no conflict of interest.

References

1. Doetschman TC, Eistetter H, Katz M, Schmidt W, Kemler R

(1985) The in vitro development of blastocyst-derived embryonic

stem cell lines: formation of visceral yolk sac, blood islands and

myocardium. J Embryol Exp Morphol 87:27–45

2. Doi K, Itoh H, Nakagawa O, Igaki T, Yamashita J, Chun TH,

Inoue M, Masatsugu K, Nakao K (1997) Expression of natriuretic

peptide system during embryonic stem cell vasculogenesis. Heart

Vessels Suppl 12:18–22

3. Burridge PW, Anderson D, Priddle H, Barbadillo Munoz MD,

Chamberlain S, Allegrucci C, Young LE, Denning C (2007)

Improved human embryonic stem cell embryoid body homoge-

neity and cardiomyocyte differentiation from a novel V-96 plate

aggregation system highlights interline variability. Stem Cells

25:929–938

4. Burridge PW, Thompson S, Millrod MA, Weinberg S, Yuan X,

Peters A, Mahairaki V, Koliatsos VE, Tung L, Zambidis ET

Heart Vessels (2014) 29:834–846 845

123

(2011) A universal system for highly efficient cardiac differen-

tiation of human induced pluripotent stem cells that eliminates

interline variability. PLoS One 6:e18293

5. Kattman SJ, Witty AD, Gagliardi M, Dubois NC, Niapour M,

Hotta A, Ellis J, Keller G (2011) Stage-specific optimization of

activin/nodal and BMP signaling promotes cardiac differentiation

of mouse and human pluripotent stem cell lines. Cell Stem Cell

8:228–240

6. Xu XQ, Graichen R, Soo SY, Balakrishnan T, Rahmat SN, Sieh

S, Tham SC, Freund C, Moore J, Mummery C, Colman A,

Zweigerdt R, Davidson BP (2008) Chemically defined medium

supporting cardiomyocyte differentiation of human embryonic

stem cells. Differentiation 76:958–970

7. Yang L, Soonpaa MH, Adler ED, Roepke TK, Kattman SJ,

Kennedy M, Henckaerts E, Bonham K, Abbott GW, Linden RM,

Field LJ, Keller GM (2008) Human cardiovascular progenitor

cells develop from a KDR ? embryonic-stem-cell-derived pop-

ulation. Nature 453:524–528

8. Dvorak P, Dvorakova D, Koskova S, Vodinska M, Najvirtova M,

Krekac D, Hampl A (2005) Expression and potential role of

fibroblast growth factor 2 and its receptors in human embryonic

stem cells. Stem Cells 23:1200–1211

9. Eiselleova L, Peterkova I, Neradil J, Slaninova I, Hampl A,

Dvorak P (2008) Comparative study of mouse and human feeder

cells for human embryonic stem cells. Int J Dev Biol 52:353–363

10. Armstrong L, Tilgner K, Saretzki G, Atkinson SP, Stojkovic M,

Moreno R, Przyborski S, Lako M (2010) Human induced plu-

ripotent stem cell lines show stress defense mechanisms and

mitochondrial regulation similar to those of human embryonic

stem cells. Stem Cells 28:661–673

11. Dubois NC, Craft AM, Sharma P, Elliott DA, Stanley EG,

Elefanty AG, Gramolini A, Keller G (2011) SIRPA is a specific

cell-surface marker for isolating cardiomyocytes derived from

human pluripotent stem cells. Nat Biotechnol 29:1011–1018

12. Lee MY, Cagavi Bozkulak E, Schliffke S, Amos PJ, Ren Y, Ge

X, Ehrlich BE, Qyang Y (2011) High density cultures of

embryoid bodies enhanced cardiac differentiation of murine

embryonic stem cells. Biochem Biophys Res Commun 416:51–57

13. Fujita E, Nakanishi T, Nishizawa T, Hagiwara N, Matsuoka R

(2013) Mutations in the cardiac troponin T gene show various

prognoses in Japanese patients with hypertrophic cardiomyopa-

thy. Heart Vessels. doi:10.1007/s00380-013-0332-3

14. Zwi-Dantsis L, Huber I, Habib M, Winterstern A, Gepstein A,

Arbel G, Gepstein L (2013) Derivation and cardiomyocyte dif-

ferentiation of induced pluripotent stem cells from heart failure

patients. Eur Heart J 34:1575–1586

15. Liu J, Sun N, Bruce MA, Wu JC, Butte MJ (2012) Atomic force

mechanobiology of pluripotent stem cell-derived cardiomyo-

cytes. PLoS One 7:e37559

16. Itzhaki I, Rapoport S, Huber I, Mizrahi I, Zwi-Dantsis L, Arbel G,

Schiller J, Gepstein L (2011) Calcium handling in human induced

pluripotent stem cell derived cardiomyocytes. PLoS One 6:e18037

17. Lee YK, Ng KM, Lai WH, Chan YC, Lau YM, Lian Q, Tse HF,

Siu CW (2011) Calcium homeostasis in human induced pluripo-

tent stem cell-derived cardiomyocytes. Stem Cell Rev 7:976–986

18. Liu B, Wohlfart B, Johansson BW (1990) Effects of low tem-

perature on contraction in papillary muscles from rabbit, rat, and

hedgehog. Cryobiology 27:539–546

19. Fu Y, Zhang GQ, Hao XM, Wu CH, Chai Z, Wang SQ (2005)

Temperature dependence and thermodynamic properties of Ca2?

sparks in rat cardiomyocytes. Biophys J 89:2533–2541

20. Wang SQ, Huang YH, Liu KS, Zhou ZQ (1997) Dependence of

myocardial hypothermia tolerance on sources of activator cal-

cium. Cryobiology 35:193–200

21. Degubareff T, Sleator W Jr (1965) Effects of caffeine on mam-

malian atrial muscle, and its interaction with adenosine and cal-

cium. J Pharmacol Exp Ther 148:202–214

22. Axelsson J, Thesleff S (1958) Activation of the contractile

mechanism in striated muscle. Acta Physiol Scand 44:55–66

23. Weber A, Herz R (1968) The relationship between caffeine

contracture of intact muscle and the effect of caffeine on retic-

ulum. J Gen Physiol 52:750–759

24. Bellin M, Marchetto MC, Gage FH, Mummery CL (2012)

Induced pluripotent stem cells: the new patient? Nat Rev Mol

Cell Biol 13:713–726

25. Bratt-Leal AM, Carpenedo RL, McDevitt TC (2009) Engineering

the embryoid body microenvironment to direct embryonic stem

cell differentiation. Biotechnol Prog 25:43–51

26. Mohr JC, Zhang J, Azarin SM, Soerens AG, de Pablo JJ,

Thomson JA, Lyons GE, Palecek SP, Kamp TJ (2010) The mi-

crowell control of embryoid body size in order to regulate cardiac

differentiation of human embryonic stem cells. Biomaterials

31:1885–1893

27. Bauwens CL, Peerani R, Niebruegge S, Woodhouse KA,

Kumacheva E, Husain M, Zandstra PW (2008) Control of human

embryonic stem cell colony and aggregate size heterogeneity

influences differentiation trajectories. Stem Cells 26:2300–2310

28. Antonchuk J (2013) Formation of embryoid bodies from human

pluripotent stem cells using AggreWell plates. Methods Mol Biol

946:523–533

29. Preda MB, Burlacu A, Simionescu M (2013) Defined-size

embryoid bodies formed in the presence of serum replacement

increases the efficiency of the cardiac differentiation of mouse

embryonic stem cells. Tissue Cell 45:54–60

30. Ren Y, Lee MY, Schliffke S, Paavola J, Amos PJ, Ge X, Ye M,

Zhu S, Senyei G, Lum L, Ehrlich BE, Qyang Y (2011) Small

molecule Wnt inhibitors enhance the efficiency of BMP-4-

directed cardiac differentiation of human pluripotent stem cells.

J Mol Cell Cardiol 51:280–287

31. Gutstein DE, Marks AR (1997) Role of inositol 1,4,5-trisphos-

phate receptors in regulating apoptotic signaling and heart failure.

Heart Vessels Suppl 12:53–57

32. Satin J, Itzhaki I, Rapoport S, Schroder EA, Izu L, Arbel G,

Beyar R, Balke CW, Schiller J, Gepstein L (2008) Calcium

handling in human embryonic stem cell-derived cardiomyocytes.

Stem Cells 26:1961–1972

33. Shibata S, Hollander PB (1967) Effects of caffeine on the con-

tractility and membrane potentials of rat atrium. Experientia

23:559

34. Shinohara T, Kim D, Joung B, Maruyama M, Vembaiyan K,

Back TG, Wayne Chen SR, Chen PS, Lin SF (2013) Carvedilol

analog modulates both basal and stimulated sinoatrial node

automaticity. Heart Vessels doi. doi:10.1007/s00380-013-0378-2

846 Heart Vessels (2014) 29:834–846

123