JOURNAL OF TISSUE ENGINEERING AND REGENERATIVE MEDICINE RESEARCH ARTICLE

J Tissue Eng Regen Med (2015)

Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/term.1995

Three-dimensional functional human myocardial

tissues fabricated from induced pluripotent
stem cells
Hyoe Komae1,2, Hidekazu Sekine2, Izumi Dobashi2, Katsuhisa Matsuura2, Minoru Ono1,
Teruo Okano2 and Tatsuya Shimizu2*
1
Department of Cardiac Surgery, University of Tokyo, Japan
2
Institute of Advanced Biomedical Engineering and Science, Tokyo Womens Medical University, Japan

Abstract
The most radical treatment currently available for severe heart failure is heart transplantation;
however, the number of donor hearts is limited. A better approach is to make human cardiac tissues.
We developed an original cell sheet-based tissue-engineering technology to fabricate human cardiac
tissue by layering myocardial cell sheets. Human induced pluripotent stem (iPS) cells were differen-
tiated into cardiomyocytes to fabricate cardiomyocyte sheets. Initially, three-layer human iPS cardio-
myocyte (hiPSCM) sheets were transplanted on subcutaneous tissues of nude rats. Next, to fabricate
thicker tissue, three-layer sheets were transplanted on one day, then additional three-layer sheets
were transplanted onto them the following day, after the rst sheets were vascularized. On day 3,
the nal three-layer sheets were again transplanted, creating a nine-layer graft (multi-step transplan-
tation procedure). In the last step, six-layer sheets were transplanted on fat tissues of the inguinal
portion, which were subsequently resected together with the femoral arteries and veins to make
transplantable grafts with connectable vessels. They were then transplanted ectopically to the neck
portion of other rats by anastomosing vessels with the hosts jugular arteries and veins. Transplanted
three-layer hiPSCMs were beating and, histologically, showed a cardiac muscle-like structure with
vascular systems. Moreover, transplanted hiPSCMs proliferated and matured in vivo. Signicantly
thicker tissues were fabricated by a multi-step transplantation procedure. The ectopically
transplanted graft survived and continued to beat. We succeeded in fabricating functional human
cardiac tissue with cell sheet technology. Transplanting this cardiac tissue may become a new
treatment option for severe heart failure. Copyright 2015 John Wiley & Sons, Ltd.

Received 31 July 2014; Revised 4 November 2014; Accepted 9 December 2014

Keywords cell sheet; iPS cells; cardiomyocyte; cardiac tissue; transplantaion; heart failure

1. Introduction In cardiac regenerative medicine, cell injection

Recent progress in the eld of regenerative medicine
has been remarkable, and many new therapeutic
methods have been developed. Within this eld, cell-
based regenerative therapy is thought to be the most
promising.
*Correspondence to: Tatsuya Shimizu, Institute of Advanced Bio-
medical Engineering and Science, TWIns, Tokyo Womens Medi-
cal University, 8-1 Kawada-cho, Shinjuku-ku, Tokyo 162-8666,
Japan. E-mail: shimizu.tatsuya@twmu.ac.jp
methods were the rst to be commonly applied (Chugh
et al., 2012; Hare et al., 2012; Makkar et al., 2012).
Some of these injection methods did show improve-
ments in the ejection fraction and reduction of scar size
but problems remain, such as low cell survival rates
and the difculty of keeping the cells in the designated
areas.
These problems have been overcome by creating
three-dimensional (3D) tissues from single-source cells,
and this method will become one of the most effective
treatments. Since (Langer and Vacanti, 1993) succeeded
in fabricating ear cartilage using a biodegradable

Copyright 2015 John Wiley & Sons, Ltd.


polymer scaffold, and suggested the concept of tissue
engineering, scaffold-based tissue engineering has be-
come predominant and has already been used in clinical
applications (Crowley, 2013; Nemeno-Guanzon et al.,
2012; Tuan et al., 2013; Yildirimer et al., 2012). How-
ever, cell-dense tissues, such as heart, were difcult to
reconstruct because the seeded cells were isolated,
creating a lower cell density, which became even lower
after the scaffold biodegraded.
On the other hand, we have originally developed a
cell sheet-based tissue-engineering technology that re-
quires no scaffolds. By culturing cells on temperature-
responsive culture dishes and simply lowering the tem-
perature, the conuently cultured cells are released in
a sheet form (Okano et al., 1993). By stacking the cell
sheets, we can successfully make scaffold-free, 3D
tissues with cellcell adhesive molecules (Shimizu
et al., 2002).
We have reported success in forming thick, beating
heart tissue with vascular networks by layering neonatal
rat cardiomyocyte sheets in vivo (Shimizu et al., 2006a;
Shimizu et al., 2006b). We believe that human car-
diomyocytes can also be used to create cell sheets,
which can reconstruct thick human myocardial tissues
when they were layered, opening new treatment options
in cardiac regenerative medicine. Recently, techniques
for differentiating human stem cells, such as embryonic
stem (ES) cells or induced pluripotent stem (iPS) cells,
into human cardiomyocytes have now been established.
Applying these techniques, a human heart graft can
now be fabricated.
Many groups, including ourselves, have tried to use this
new tissue-engineering technology in clinical practice. An-
other group has reported that cardiac tissue sheets made
up of human cardiomyocytes differentiated from iPS cells,
endothelial cells and vascular mural cells were transplanted
onto rat infarcted hearts, which signicantly improved car-
diac function (Masumoto et al., 2014). Moreover, we are
now trying to fabricate a thick human heart tissue graft to
strengthen this effect, or support cardiac contraction, which
will be an innovative treatment for heart failure and is our
ultimate goal.
In this research, cardiomyocytes differentiated from
human induced pluripotent stem (hiPS) cells were
made into cell sheets. Stacking these hiPS cardiomyo-
cyte (hiPSCM) sheets in vivo, we tried to make thick
functional human cardiac tissues and to transplant this
tissue ectopically, in order to improve the methods for
fabricating thick functional human cardiac tissues and
their ectopic transplantation.
2. Methods
All animal experiments were performed according to the
Guidelines of Tokyo Womens Medical University on Ani-
mal Use, and consistent with the Guide for the Care and
Copyright 2015 John Wiley & Sons, Ltd.
H. Komae et al.
Use of Laboratory Animals prepared by the Institute of
Laboratory Animal Resources (ILAR).
2.1. Cardiac differentiation from human iPS cells
and fabrication of human myocardial tissue
grafts
HiPS cells were purchased from RIKEN (253G1, Tsukuba,
Japan) and cultured using a bioreactor system to induce
differentiation into hiPSCMs, as previously reported
(Matsuura et al., 2012), with some modications
(Figure 1A). To make cell sheets, 2 106 differentiated
hiPSCMs in this method were seeded on 35 mm
temperature-responsive dishes (UpCell, CellSeed, Tokyo,
Japan) in Dulbeccos modied Eagles medium (DMEM;
Sigma-Aldrich, St. Louis, MO, USA) supplemented with
10% fetal bovine serum (FBS) and cultured until they
grew to conuency. The culture dishes were placed in an
incubator set at 20C; the hiPSCM sheets detached spon-
taneously within 1 h. A detached sheet was spread at,
the medium was aspirated and the dish was held at 37C
for 1 h to allow the cell sheet to adhere. Another detached
hiPSCM sheet was then transferred and stacked onto the
rst sheet, using a pipette, to make two layers. The re-
maining medium was aspirated and the dish was held at
37C for another 1 h. An identical procedure was repeated
to layer a third sheet, thus fabricating the three-layer
hiPSCM sheets (Figure 1B).
2.2. Transplantation of engineered human
myocardial tissue grafts
Male Fischer 344 athymic nude rats (Charles River Japan,
Tokyo, Japan), aged 45 weeks, were anaesthetized by
2% isourane inhalation. An L-shaped incision was made
in the dorsal skin, exposing the dorsal subcutaneous tissue.
Three-layer hiPSCM sheets on temperature-responsive
dishes were detached by lowering the temperature to 20
C, and washed with lactated Ringers solution. The hiPSCM
sheets were lifted onto a sterile polypropylene support
sheet (2 1.5 cm), transplanted onto the subcutaneous tis-
sue and covered with a 0.5 mm-thick silicone membrane
(Unique Medical, Tokyo, Japan) to prevent adhesion to
the skin, then the incisions were closed (Shimizu et al.,
2006b). The transplanted portions were observed 2 weeks
later (n = 5) or every month up to 6 months (n = 8)
(Figure 1C). When grafts of hiPSCM sheets with more than
three layers were needed, an initial three-layer hiPSCM
sheet was transplanted and, after 1 day, a second three-
layer sheet was placed on the rst transplanted sheet, then
a third three-layer hiPSCM sheet on the following day.
Allowing 1 day between each transplantation encouraged
vascularization, which supports the survival of the nal
hiPSCM tissue (Figure 1D). In this multi-step transplanta-
tion procedure, hiPSCM sheets were repeatedly stacked
three times to make nine layers and were observed 2 weeks
later (n = 4) (Figure 1C).
J Tissue Eng Regen Med (2015)
DOI: 10.1002/term
3D functional human myocardial tissues from iPS cells

Figure 1. Methods of hiPSCM sheet transplantation. hiPS cells (blue) were cultured using a bioreactor system to induce differentia-
tion into hiPSCMs in procedures, as shown. (A) About 8090% of differentiated cells were cardiomyocytes (red) and the rest were sup-
posed to be mural cells (yellow). (B) Differentiated cells were seeded on temperature-responsive dishes and cultured until they grew
to conuency; then, by lowering the temperature, the hiPSCM sheets spontaneously detached and were stacked to three layers
in vitro, then (C) the sheets were transplanted in vivo. (D) By stacking over three layers, with repeated transplantation of layered cell
sheets at 1 day intervals (which we call a multi-step transplantation procedure), we were able to overcome the thickness limitation

2.3. Fabrication of grafts with connectable
vessels and their ectopic transplantation
Male F344 nude rats, aged > 8 weeks, weight 300350 g,
were anaesthetized and incisions were made in the ingui-
nal portion. A six-layer graft was made, using the multi-
step procedure, on the fat tissue perfused with a branch
of the femoral artery and vein (n = 2). Two weeks later,
the fat tissue with the six-layer construct was resected,
with the femoral artery and vein, after injecting heparin
(1000 U). Subsequently, the graft was transplanted into
the neck of another F344 rat (300350 g) by anastomos-
ing vessels with the hosts jugular artery and vein, using
a cuff technique. The transplanted graft was observed 1
week later (Figure 1C).
2.4. Histological analysis
Rats were anaesthetized as described and transplanted
sites together with underlying tissues were resected. In
some cases, 1 mg lectin (DyLight 594 Labelled Tomato
Lectin, DyLight 488 Labelled Tomato Lectin, Abcam, Cam-
bridge, UK) was injected into the penile venous plexus.
The rats were then sacriced under adequate anaesthesia.
Frozen cross-sections of the resected grafts were sec-
tioned sagittally into 610 m slices and haematoxylin
and eosin (H&E) staining was performed by conventional
methods. For immunohistochemistry, mouse anti-troponin
T antibodies (Thermo Fisher Scientic, Waltham, MA,
USA), rabbit polyclonal anti-CD31 antibodies (Thermo
Fisher Scientic), mouse anti--actinin antibodies
(Sigma-Aldrich), rabbit monoclonal anti-Ki67 antibodies
(Abcam) and rabbit polyclonal anti-Histone H3 antibodies
(Abcam) were used for primary antibodies. Fluorescently
labelled secondary antibodies (AlexaFluor 568-labelled
goat anti-mouse IgG antibodies, AlexaFluor 488-labelled
goat anti-mouse IgG antibodies, AlexaFluor 488-labelled
goat anti-rabbit IgG antibodies (Molecular Probes, Eugene,
OR, USA) were used for secondary antibodies.
The sections were nally counterstained with anti-fade
solution (ProLong Gold Antifade Reagent with DAPI, Molecu-
lar Probes) and were observed using confocal laser scanning
microscopy (FV1200 IX83, Olympus, Tokyo, Japan).
2.5. Electron microscopy analysis
Transplanted grafts were harvested and specimens were
xed and processed by Tokai Electron Microscopy Inc.

Copyright 2015 John Wiley & Sons, Ltd. J Tissue Eng Regen Med (2015)
DOI: 10.1002/term

(Nagoya, Japan) for electron microscopy analysis. The
specimens were embedded into 100% resin and polymer-
ized to make blocks. The blocks were then ultra-thinly
sectioned at 70 nm and stained with 2% uranyl acetate.
They were observed using a transmission electron micro-
scope (JEM-1200EX; Jeol Ltd, Tokyo, Japan) at an acceler-
ation voltage of 80 kV. Digital images were taken with a
CCD camera (Veleta; Olympus Soft Imaging Solutions
GmbH, Mnster, Germany).
2.6. Electrophysiological analysis
Transplantation sites were opened, and two microelec-
trodes (29 G in diameter, ADInstruments, New South
Wales, Australia) were positioned over the grafts. Electrical
potentials were amplied by bioelectric ampliers (UA102,
Unique Medical) and recorded using a data acquisition sys-
tem (ML870, PowerLab 8/30, ADInstruments). At the same
time, the hosts electrocardiogram and body pressure were
recorded.
2.7. Membrane potential mappings
Grafts were made as ectopic transplantation models. Three
weeks later, the resected graft was placed in a tissue cham-
ber (ABLE). First, the graft was perfused with 2.5 ml culture
medium containing 2 mM voltage-sensitive dye RH237
(Molecular Probes) dissolved in dimethyl sulphoxide and
5 M blebbistatin (Tocris, Ellisville, MO, USA) to inhibit con-
traction, at 50 l/min. Second, it was perfused with me-
dium at the same ow rate in the bioreactor system,
where O2 levels were maintained at 20%, CO2 levels at
5%, pH level at 7.4 and temperature at 37C.
An hour after RH 237 perfusion, the graft was excited with
an LED light source (LEX2-G, BrainVision, Tokyo, Japan) at
532 nm, and uorescent images were imaged with a dichroic
mirror (650 nm cut-off wavelength). Emission uorescence
of 715 nm was collected using a CMOS camera (MiCAM
Ultima, BrainVision) and a Tandem-lens scope (BrainVision)
at 1 ms/frame and 0.16 mm2/pixel. Magnication of the ob-
jective lens was 0.63 power (Choi and Salama, 2000). The
data were analyzed using BV_Ana software (Brainvision) to
process it, using a spatial lter for activation time analysis
and activation mapping.
2.8. Video capture
Macroscopic images were recorded using a digital video
camera (DCR-TRV900, Sony, Tokyo, Japan) and edited
with Adobe Premiere Pro software.
2.9. Data analysis
Data are expressed as mean SD. Statistical analysis was
performed using Students unpaired t-test; p < 0.05 was
considered signicant.
Copyright 2015 John Wiley & Sons, Ltd.
H. Komae et al.
3. Results
3.1. Thick human cardiac tissue fabrication by
the multi-step transplantation procedure
The transplanted three-layer hiPSCM sheets were still beat-
ing macroscopically after 2 weeks. Synchronized electrical
potentials were detected (Figure 2A). The results of uores-
cent immunostaining suggested that the transplanted
hiPSCM sheets had survived as cardiac tissues, with sarco-
mere structures and microvascular networks (Figure 2E, G)
connecting with the hosts circulatory system (Figure 2I).
The mapping indicated that there was no arrhythmia
(see supporting information, Video S1). In three cases of
nine-layer hiPSCM transplantation, macroscopic beating
and synchronized electrical potentials were detected 2 weeks
later (Figure 2B). The uorescent immunostaining also
showed that the cardiac tissues with sarcomere structures
and microvascular networks had survived (Figure 2F, H).
Some vessels were observed penetrating the transplanted
graft (Figure 2J).
Measuring the thickness of the transplanted grafts,
the nine-layer transplanted sheets were signicantly
thicker (359 161 m, n = 4) than the three-layer
sheets (165 40 m, n = 5); p < 0.01 (Figure 2C, D).
3.2. Long-term observation of three-layer
transplantation
Macroscopic beating and synchronized electrical poten-
tials were conrmed at 6 months after transplantation
(Figure 3A). The beating appeared strong (see supporting
information, Video S2) and the grafts were well-
vascularized (Figure 3B). The results of uorescent immu-
nostaining indicated that the structure of the cardiac tis-
sues contained sarcomere structures and a microvascular
network (Figure 3C); in addition, gap junctions between
cardiac cells were observed at 6 months (Figure 3D).
The fabricated tissue was more cell-dense at 6 months
than the grafts at 2 weeks after transplantation, which im-
plied that hiPSCMs were proliferating after transplanta-
tion. To see the transplanted hiPSCMs proliferation,
uorescent immunostaining for Ki67 and phosphor-
histone H3 (PPH3) was performed. From the results,
many hiPSCMs were proliferating at 2 weeks after trans-
plantation (Figure 3G, H); however, hiPSCMs had
stopped proliferation by 6 months (Figure 3E, F).
3.3. Maturation of transplanted hiPSCMs
From the electron microscopy images, the hiPSCM
sheets had poor myobrils (Figure 4A, B), but myobrils
had fused and elongated as time passed and desmo-
somes could be seen at 6 months (Figure 4E). Moreover,
the number of mitochondria in the cytoplasm increased
in the transplanted grafts. Either a sarcoplasmic
J Tissue Eng Regen Med (2015)
DOI: 10.1002/term
3D functional human myocardial tissues from iPS cells

Figure 2. Human cardiac tissues fabricated from hiPSCM sheets transplanted in vivo; results of three-layer hiPSCM sheet transplantation
after 2 weeks (A, C, E, G) and of nine-layer sheets (B, D, F, H). Electrical potential changes (A, B; blue) are regular and completely independent
from host ECG (red) or breathing (green). The thickness of the three-layer graft was 165 40 m (C; n = 5; *) and 359 161 m in the nine-
layer cases (D; n = 4; **). Fluorescent immunostaining shows cardiac tissues with microvascular systems (E, F; red, cTnT; green, CD31) and
the striated structure of sarcomeres (G, H; red, -actinin). Grafts were perfused from the host circulatory systems. Fabricated graft was
stained by red lectin injected into the host (I, arrows). Some vessels are penetrating the transplanted hiPSCM sheets (J; arrows)

Copyright 2015 John Wiley & Sons, Ltd. J Tissue Eng Regen Med (2015)
DOI: 10.1002/term
 H. Komae et al.

Figure 3. Results at 6 months after transplantation and hiPSCM proliferation. Electrical potential change (A; blue) was regular and
completely independent from host ECG (red) and breathing (green). Transplanted hiPSCM sheets (B; circled with arrows) appear
to be vascular-rich tissue. HiPSCM sheets became cell-dense cardiac tissue with vascular networks (C; red, cTnT; green, CD31) and
gap junctions (D; red, -actinin; green, connexin43). While hiPSCMs were proliferating (G; red, cTnT; green, Ki67: H; red, cTnT; green,
PHH3) at 2 weeks after transplantation, proliferation stopped at 6 months (E; red, cTnT; green, Ki67: F; red, cTnT; green, PHH3)

reticulum or a T-tube was observed around a myobril
at 6 months (Figure 4F).
3.4. Successful ectopic transplantation of
vascularized tissue
The grafts resected from the inguinal portion were beating,
and the synchronized electric potentials were measured
(Figure 5A). One week after ectopic transplantation, the
grafts were also beating (Figure 5B). Histological study
showed human cardiac tissue with microvascular networks
that were perfused from the hosts circulatory system
(Figure 5C, D).
4. Discussion
This is the rst report of thick functional human myocar-
dial tissue to be successfully fabricated from hiPSCM
sheets in vivo. We succeeded in reconstructing functional
human myocardial tissue thicker than 350 m from
hiPSCM sheets by using a multi-step transplantation pro-
cedure, and transplanted this fabricated cardiac tissue
ectopically.
4.1. Fabrication of thick human cardiac tissue
from hiPSCM sheets
To fabricate thick tissue, vascularization within the con-
struct is critical. From our previous study we learned
that when more than three layers of cell sheets are
transplanted in one step, central necrosis will develop,
and the thickness of the construct could not exceed 80
m. Looking at the lower portions of the constructs that
did survive, a well-organized vascular network was
formed, which indicates perfusion from the host
(Shimizu et al., 2006b). Therefore, we hypothesized that
multiple three-layer sheets could be transplanted if it
was onto a vascularized construct, which was veried.
It became clear that three-layer sheets can be layered
at 1 day intervals; moreover, the resulting six-layer con-
struct also developed a functional vascular network,
which enabled more sheets to be transplanted. In this
way, repeated transplantation of layered cell sheets af-
ter an interval of 1 day produced neovascularization
within the previously implanted graft and overcame
the limit of thickness (Figure 1D, multi-step transplanta-
tion procedure).
If thick human cardiac tissues are to be fabricated, a
large number of hiPSCM sheets must be layered. There-
fore, we needed to clarify whether our multi-step trans-
plantation procedure could be used with hiPSCM sheets.
Initially, we conrmed that transplanted three-layer
iPSCM sheets did survive, and that functional vascular
networks connected to the hosts circulatory system
were formed in the sheets. Moreover, the electric mem-
brane potential of the fabricated grafts proved to ow in
one direction, like native human cardiac tissues. Then,
additional hiPSCM sheets could be transplanted onto
the rst vascularized sheets, and they survived. By re-
peating this, hiPSCM sheets were stacked up to nine
layers. This was largely because the myocardial tissues
had formed functional microvascular networks within
the transplanted layers. Therefore, we can now say that

Copyright 2015 John Wiley & Sons, Ltd. J Tissue Eng Regen Med (2015)
DOI: 10.1002/term
3D functional human myocardial tissues from iPS cells

Figure 4. Results of electron microscopy. Low-powered (A, C, E) and high-powered (B, D, F) images of hiPSCM sheets (A, B) at 2
weeks (C, D) and 6 months (E, F). The cells are myobril-poor and sparse in the hiPSCM sheets (B), but the myobrils became dense,
wide and elongated over time, and desmosomes were formed at 6 months (E; blank arrows). Mitochondria increased in number (A, C,
E; arrows); arrowhead indicates sarcoplasmic reticulum or T-tubes around myobrils (F)

much thicker human myocardial tissue can be fabricated
in vivo by layering hiPSCM sheets using a multi-step
procedure.
4.2. HiPSCMs after transplantation in vivo
Two weeks after transplantation, the fabricated tissue
from three-layer hiPSCM sheets had increased in thick-
ness to 165 m, which seemed inconsistent with our
previous report that layered sheets > 100 m could
not be transplanted at one time. However, our hypothe-
sis that hiPSCMs were proliferating after transplantation
in vivo now makes sense. HiPSCM sheets survived by
developing functional microvascular networks, and
hiPSCMs could keep proliferating (Figure 3G, H). There-
fore, transplanted three-layer sheets exceeded the previ-
ous thickness limit.
However, the tissue was cell-sparse and the orientation
was not orderly compared to that of the infant rat cardio-
myocyte sheet (Shimizu et al., 2002) (Figure 2C, D). This
is probably because the infant rat cardiomyocytes were al-
ready mature to some degree, and infant rat
cardiomyocyte sheets would contain broblasts or other
factors, which would let cells bind as they were. HiPSCMs
are rather young cells, approximately corresponding to
human embryonic cardiomyocytes, which freely prolifer-
ate. It is possible that cell adhesion or assembly will be
weaker because hiPSCM sheets contain a higher propor-
tion of cardiomyocytes (8090%) than native tissue. We
also observed transplanted sheets up to 6 months and
found that the beating became more obvious and the fab-
ricated cardiac tissues became more cell-dense over time.
At 6 months, cell proliferation had almost stopped, like
native human cardiomyocytes after birth.
Previous infant autopsy reports say that human
cardiomyocytes proliferate during gestation and that pro-
liferation stops at approximately 23 weeks post partum
(Austin et al., 1995; Huttenbach et al., 2001; Mayhew
et al., 1997; Mayhew et al., 1998). Moreover, the rate of
proliferating nuclei is highest during weeks 1228 of ges-
tation, and after 28 weeks it decreases (Huttenbach
et al., 2001). The human infant heart is formed at about
5 weeks of gestation and, if the hiPSCMs correspond to na-
tive cardiomyocytes, they should proliferate for 2 weeks
after transplantation. Six months after transplantation

Copyright 2015 John Wiley & Sons, Ltd. J Tissue Eng Regen Med (2015)
DOI: 10.1002/term

Figure 5. Ectopic transplantation: 2 weeks later, the transplanted six-layer hiPSCM sheets are beating (A; blue, graft potential) and
after 1 week the ectopically transplanted graft was also beating (B; blue). Histologically, the graft is perfused from the host circula-
tory system (C; arrow, penetrating vessel: D; arrows; red, CD31; green, lectin)
would then make them equivalent to around 30 weeks
gestation, when the native cardiomyocytes are proliferat-
ing less.
The electron microscopy results suggested that
hiPSCMs matured in vivo as cardiomyocytes; myobrils
became dense, wide and elongated and the mitochondria
increased in number over time (Figure 4A, C, E). More-
over, cardiac-specic intracellular structures, such as the
sarcoplasmic reticulum or T-tubes, could be seen at 6
months (Figure 4F), whereas they could not be detected
at 2 weeks. HiPSCMs cultured in vitro also become more
mature to some degree, but they matured even more if
cultured in vivo (see supporting information, Figure S1).
Myobrils were being fused and elongated in the case of
either in vitro culture or in vivo, but the density of sarco-
mere units in hiPSCMs cultured in vivo were obviously
higher. Intracellular structures of hiPSCMs cultured
in vivo were also more developed. Therefore, we can say
that hiPSCMs were well suited for growing in vivo.
In the rat experiments, reasonably mature structures
could be seen only 3 weeks after infant rat cardiomyocyte
sheets were transplanted (Shimizu et al., 2002), which
might support the previous hypothesis that infant rat
cardiomyocytes were initially more mature than the dif-
ferentiated hiPSCMs.
From the previous report, intracellular structures such
as myobrils, desmosomes and secretary granules were
formed by fetal month 2, and mitochondria increased
in number from at about fetal month 3. At fetal month 7,
Copyright 2015 John Wiley & Sons, Ltd.
H. Komae et al.
the cardiomyocytes had matured in structure and
became similar to those of an adult (Sako, 1971a, 1971b,
1971c, 1971d).
In this study, we succeeded in fabricating rather mature
cardiac tissues with mature myobrils and elongated sar-
comeres in vivo, although they were not as mature com-
pared with human heart tissues. One of our goals is to
fabricate mature myocardial tissue with strong contractile
forces. To achieve this goal, newer methods or factors will
be required.
4.3. Ectopic transplantation
If thick human cardiac tissue were fabricated on the
intended portion, for example on the heart surface, it
would require very invasive polysurgery, which is imprac-
tical. Therefore, the tissue should be fabricated in an-
other, less invasive, location and then transplanted
ectopically.
In the eld of clinical plastic surgery, various tissues are
reconstructed using free ap grafts by anastomosing
nearby vessels with microsurgery to substitute defective
tissue due to congenital malformation, surgical resection
or traumatic injury (Brown and Shaw, 2010; Cannon
et al., 2012; Mohan et al., 2013; Wong and Wei, 2010).
We thought that if our fabricated tissue could be
transplanted like a free ap, it would survive and main-
tain its function. Therefore, successful results of ectopic
J Tissue Eng Regen Med (2015)
DOI: 10.1002/term
3D functional human myocardial tissues from iPS cells

transplantation in this study demonstrated that our hy-
pothesis might be practical. When intending to transplant
fabricated tissues onto the heart surface, graft arteries
used in coronary artery bypass surgery can be anasto-
mosed with the graft. If the main artery is long enough,
it can be anastomosed from the aorta, like a great saphe-
nous vein graft. The main vein can also be anastomosed
with the accompanying vein, or directly with the right
atrium, so as not to create a shunt.
We successfully fabricated human cardiac tissue by
transplanting hiPSCM sheets in vivo. If this method con-
tinues to develop and fabricated tissues can be scaled up,
we may be able to fabricate human cardiac tissue or
even a human heart itself in the future. Although further
studies will be required, transplantation of this new fab-
ricated cardiac tissue may become a new treatment op-
tion for severe heart failure that will rescue more
patients.

4.4. Future clinical use

Although there are problems such as the need for a large
number of cells, and the impractical process of layering
that requires polysurgery in vivo, many new methods are
under development that may solve these challenges. For
example, many researchers are developing new, large-
scale, cell-culture systems (Chen et al., 2014; Matsuura
et al., 2012) and several procedures are being substituted
with ex vivo ones using bioreactor systems (Sakaguchi
et al., 2013; Sekine et al., 2013). When we overcome these
challenges, this method will provide groundbreaking
treatment procedures.
This new technique may resolve many of the problems
with treatments currently available for severe heart fail-
ure. When myocardial tissues or even hearts fabricated
from iPS cells become available, they will be made from
human cells and perhaps even the patients own cells.
We plan to combine current therapeutic methods with
fabricated human cardiac tissue, which will be able to re-
duce many of the adverse effects to provide more effective
and safer treatments.
Conict of interest
T.O. is a founder and director of the board of CellSheed
Inc., licensing technologies and patents from Tokyo
Womens Medical University; and T.O. and T.S. are share-
holders of CellSeed Inc.
Acknowledgements
We thank Allan Nisbet for his useful comments and editing as-
sistance. This study was supported by grants from the Japan
Society for the Promotion of Science (JSPS) through the
Funding Programme for World-leading Innovative R&D on Sci-
ence and Technology (FIRST Programme), initiated by the
Council for Science and Technology Policy (CSTP). We report
grants from: CSTEC, sponsored by MEXT during the conduct
of the study; Hitachi Ltd, Dai Nippon Printing Co. Ltd,
CellSheed Inc., Olympus Corp., Terumo Corp., Nihon Kohden
Corp., Asahi Kasei Corp., Panasonic Corp., the National Insti-
tute of Material Science, Nikon Corp. and Kowa Co. Ltd out-
side the submitted work.

References
Austin A, Fagan DG, Mayhew TM. 1995; A
stereological method for estimating the to-
tal number of ventricular myocyte nuclei
in fetal and postnatal hearts. J Anat 187
(3): 641647.
Brown JS, Shaw RJ. 2010; Reconstruction of
the maxilla and midface: introducing a
new classication. Lancet Oncol 11(10):
10011008.
Cannon TY, Strub GM, Yawn RJ, et al. 2012;
Oromandibular reconstruction. Clin Anat
25(1): 108119.
Chen KG, Mallon BS, McKay RD, et al. 2014;
Human pluripotent stem cell culture: con-
siderations for maintenance, expansion,
and therapeutics. Cell Stem Cell 14(1):
1326.
Choi BR, Salama G. 2000; Simultaneous
maps of optical action potentials and cal-
cium transients in guinea-pig hearts:
mechanisms underlying concordant
alternans. J Physiol 529(1): 171188.
Chugh AR, Beache GM, Loughran JH, et al.
2012; Administration of cardiac stem cells
in patients with ischemic cardiomyopathy:
the SCIPIO trial: surgical aspects and in-
terim analysis of myocardial function and
viability by magnetic resonance. Circula-
tion 126(11, suppl 1): S5464.
Crowley C, Wong JM, Fisher DM, et al. 2013; A
systematic review on preclinical and clinical
studies on the use of scaffolds for bone repair
in skeletal defects. Curr Stem Cell Res Ther 8
(3): 243252.
Hare JM, Fishman JE, Gerstenblith G, et al.
2012; Comparison of allogeneic vs autolo-
gous bone marrow-derived mesenchymal
stem cells delivered by transendocardial
injection in patients with ischemic
cardiomyopathy: the POSEIDON random-
ized trial. J Am Med Assoc 308(22):
23692379.
Huttenbach Y, Ostrowski ML, Thaller D, et al.
2001; Cell proliferation in the growing
human heart: MIB-1 immunostaining in
preterm and term infants at autopsy.
Cardiovasc Pathol 10(3): 119123.
Langer R, Vacanti JP. 1993; Tissue engineer-
ing. Science 260(5110): 920926.
Makkar RR, Smith RR, Cheng K, et al. 2012;
Intracoronary cardiosphere-derived cells for
heart regeneration after myocardial
infarction (CADUCEUS): a prospective,
randomised phase 1 trial. Lancet 379
(9819): 895904.
Masumoto H, Ikuno T, Takeda M, et al.
2014; Human iPS cell-engineered cardiac
tissue sheets with cardiomyocytes and
vascular cells for cardiac regeneration.
Sci Rep 4: 6716.
Matsuura K, Wada M, Shimizu T, et al. 2012;
Creation of human cardiac cell sheets using
pluripotent stem cells. Biochem Biophys Res
Commun 425(2): 321327.
Mayhew TM, Gregson C, Pharaoh A, et al.
1998; Numbers of nuclei in different tissue
compartments of fetal ventricular myocar-
dium from 16 to 35 weeks of gestation.
Virchows Arch 433(2): 167172.
Mayhew TM, Pharaoh A, Austin A, et al.
1997; Stereological estimates of nuclear
number in human ventricular
cardiomyocytes before and after birth ob-
tained using physical disectors. J Anat
191(1): 107115.
Mohan AT, Al-Ajam Y, Mosahebi A. 2013;
Trends in tertiary breast reconstruction:
literature review and single centre experi-
ence. Breast 22(2): 173178.
Nemeno-Guanzon JG, Lee S, Berg JR, et al.
2012; Trends in tissue engineering for
blood vessels. J Biomed Biotechnol 2012
(2012); 956345.
Okano T, Yamada N, Sakai H, et al. 1993;
A novel recovery system for cultured
cells using plasma-treated polystyrene
dishes grafted with poly(N-

Copyright 2015 John Wiley & Sons, Ltd. J Tissue Eng Regen Med (2015)
DOI: 10.1002/term
 H. Komae et al.

isopropylacrylamide). J Biomed Mater
Res 27(10): 12431251.
Sakaguchi K, Shimizu T, Horaguchi S,
et al. 2013; In vitro engineering of
vascularized tissue surrogates. Sci Rep
3: 1316.
Sako Y. 1971a; [Electron microscopic studies
on the development of the human embry-
onic myocardium. 1. Myobrils]. Jpn Circ
J 35(11): 14911496.
Sako Y. 1971b; [Electron microscopic studies
on the development of the human embry-
onic myocardium. 2. Mitochondria and
glycogen]. Jpn Circ J 35(11): 14971500.
Sako Y. 1971c; [Electron microscopic studies
on the development of the human embry-
onic myocardium. 3. Granules]. Jpn Circ J
35(11): 15011503.
Sako Y. 1971d; [Electron microscopic studies
on the development of the human embry-
onic myocardium. 4. Cell junction]. Jpn
Circ J 35(11): 15041508.
Sekine H, Shimizu T, Sakaguchi K, et al.
2013; In vitro fabrication of functional
three-dimensional tissues with
perfusable blood vessels. Nat Commun
4: 1399.
Shimizu T, Sekine H, Isoi Y, et al. 2006a;
Long-term survival and growth of
pulsatile myocardial tissue grafts
engineered by the layering of cardio-
myocyte sheets. Tissue Eng 12(3):
499507.
Shimizu T, Sekine H, Yang J, et al. 2006b;
Polysurgery of cell sheet grafts overcomes
diffusion limits to produce thick,
vascularized myocardial tissues. FASEB J
20(6): 708710.
Shimizu T, Yamato M, Isoi Y, et al. 2002;
Fabrication of pulsatile cardiac tissue
grafts using a novel three-dimensional cell
sheet manipulation technique and
temperature-responsive cell culture sur-
faces. Circ Res 90(3): e40.
Tuan RS, Chen AF, Klatt BA. 2013; Cartilage
regeneration. J Am Acad Orthop Surg 21
(5): 303311.
Wong CH, Wei FC. 2010; Microsurgical free
ap in head and neck reconstruction. Head
Neck 32(9): 12361245.
Yildirimer L, Thanh NT, Seifalian AM. 2012;
Skin regeneration scaffolds: a multimodal
bottom-up approach. Trends Biotechnol 30
(12): 638648.

Supporting information on the internet

The following supporting information may be found in the online version of this article:

Figure S1. HiPSCMs cultured in vitro and in vivo: electron microscopy images of hiPSCMs cultured for 1 month in vitro
(A) and in vivo (B). Myobrils (arrows) were being fused and elongated in both images, but the density of sarcomere
units in the hiPSCMs cultured in vivo was obviously higher; intracellular structures of hiPSCMs cultured in vivo were
also more developed
Video S1. Optical mapping of electric membrane potential; the electric conduction ows in a non-random pattern (400
consecutive frames, 0.001 s/frame)
Video S2. Three-layer hiPSCM sheets transplanted after 6 months; wholly synchronized regular beating can be seen in
the transplanted area

Copyright 2015 John Wiley & Sons, Ltd. J Tissue Eng Regen Med (2015)
DOI: 10.1002/term