d
s
im
5
and engraftment is present over 11 months post-
associated with human ES cells are eliminated, and the possi-
bility of generating patient-specific iPS cells for autologous ther-
tions (Emery, 2002). Current treatment options are only palliative,
and thus far there is no cure for any type of MD. Therapeutic
strategies that focus on the replacement of the diseased muscle
between transplantation studies involving mouse and human
pluripotent stem cells. Proof-of-principle studies using human
iPS cells are required in order to begin seriously considering
tissue with stem cells that can give rise to healthy myofibers, as
well as self-renew, are particularly attractive. This strategy has
been used in the hematopoietic system for the past 40 years
potential therapeutic applications of these cells.
Here we describe the efficient derivation of a proliferating pop-
ulation of human skeletal myogenic progenitors from both ES
transplant. This study provides the proof of principle
for the derivation of functional skeletal myogenic
progenitors from human ES/iPS cells and highlights
their potential for future therapeutic application in
muscular dystrophies.
INTRODUCTION
Muscle wasting affects millions of individuals worldwide and is
caused by a variety of conditions, including cachexia, sarcope-
nia, and muscular dystrophies (MDs). The latter comprises
more than 30 genetically distinct disorders that culminate in
paralysis and, in many instances, cardiopulmonary complica-
apies is enabled. Whereas safety issues still need to be carefully
addressed before these cells can be used in the clinical setting,
a critical prerequisite for a potential therapeutic application is the
generation of abundant engraftable tissue-specific cell prepara-
tions. Although the use of mouse iPS-derived cells to correct
a disease phenotype has been documented for several models
of disease through derivation of hematopoietic (Hanna et al.,
2007), endothelial (Xu et al., 2009), neural (Wernig et al., 2008),
pancreatic (Alipio et al., 2010), liver (Espejel et al., 2010), and
myogenic (Darabi et al., 2011a; Mizuno et al., 2010) precursor
cells, the human iPS field lags far behind in this regard. To
date, there is only one study documenting functional improve-
ment from human iPS cells, using a rat model of Parkinson’s
disease (Hargus et al., 2010). Thus, there is clearly a huge gap
exhibit superior strength. Importantly, transplanted
cells also seed themuscle satellite cell compartment, 2008; Takahashi et al., 2007; Yu et al., 2007), ethical concerns
Human ES- and iPS-Derive
Restore DYSTROPHIN and
upon Transplantation in Dy
Radbod Darabi,1 Robert W. Arpke,2 Stefan Irion,3 John T. D
and Rita C.R. Perlingeiro1,*
1Department of Medicine
2Department of Pediatrics
Lillehei Heart Institute, University of Minnesota, Minneapolis, MN 554
3iPierian, 951 Gateway Blvd, South San Francisco, CA 94080, USA
*Correspondence: perli032@umn.edu
DOI 10.1016/j.stem.2012.02.015
SUMMARY
A major obstacle in the application of cell-based
therapies for the treatment of neuromuscular disor-
ders is obtaining the appropriate number of stem/
progenitor cells to produce effective engraftment.
The use of embryonic stem (ES) or induced pluripo-
tent stem (iPS) cells could overcome this hurdle.
However, to date, derivation of engraftable skeletal
muscle precursors that can restore muscle function
from human pluripotent cells has not been achieved.
Here we applied conditional expression of PAX7 in
human ES/iPS cells to successfully derive large
quantities of myogenic precursors, which, upon
transplantation into dystrophic muscle, are able to
engraft efficiently, producing abundant human-
derived DYSTROPHIN-positive myofibers that
610 Cell Stem Cell 10, 610–619, May 4, 2012 ª2012 Elsevier Inc.
Cell Stem Cell
Short Article
Myogenic Progenitors
Improve Contractility
trophic Mice
os,3 Marica Grskovic,3 Michael Kyba,2
5, USA
with great success. A major caveat with muscle tissue is the
impossibility of obtaining enough skeletal muscle stem cells
(satellite cells) without causing severe and permanent damage
to the muscle of the donor, in contrast to hematopoietic stem
cells (HSCs), which can be harvested from mobilized peripheral
blood or marrow with minimal harm to the donor. Small muscle
biopsies allow for the ex vivo expansion of satellite cell progeny;
however, as observed for HSCs (Guenechea et al., 1999), ex vivo
expansion of myoblasts from satellite cells results in loss of
engraftment ability (Montarras et al., 2005). Consistently, early
clinical trials involving the transplantation of ex-vivo-expanded
myoblasts failed to improve strength in patients with Duchenne’s
MD (Mendell et al., 1995; Vilquin, 2005). Therefore, alternate
sources of early skeletal muscle progenitors are required for
the feasibility of a stem cell therapy approach for MD.
One of the major advantages of pluripotent stem cells is the
prospect of generating large quantities of specific cell popula-
tions for regenerative purposes. In particular, with the recent
breakthrough of reprogramming somatic cells (Park et al.,
no
A
VVI B
sBEsllecSE9H-7XAPi EB-derived mo
IIIIII
+
Purification of PAX7+
(GFP+) cells
Cell Stem Cell
Muscle Engraftment from Human ES and iPS Cells
and iPS cells, which, upon transplantation into dystrophin-defi-
cient mice, promote extensive and long-term regeneration that
is accompanied by functional improvement.
RESULTS
PAX7 Induces the Myogenic Program in Differentiating
Human ES and iPS Cells
To assess whether PAX7, a paired-box transcription factor well
known for its role in the maintenance of the adult satellite cell
compartment (Oustanina et al., 2004; Seale et al., 2000), can
efficiently induce the myogenic program in human ES- and
iPS-derived embryoid bodies, as observed in mouse cultures
(Darabi et al., 2011a; Darabi et al., 2011b), we modified the
C
e
ll
c
o
u
n
t
H9
IPRN 14.57
IPRN 13.13
Days
10
10
10
10
0 4 8 12 16
8
7
6
5
C
PAX7 MYOGENIN MHC
D
if
f
e
r
e
n
t
ia
t
io
n
P
r
o
li
f
e
r
a
t
io
n
iPAX7-H9
92 ± 1.7
3.6 ± 0.2
12.6 ± 2.7
89 ± 1.2
5 ± 1.2
94 ± 1.02
iPAX7-IPRN
13.13
95 ± 0.7
3.1 ± 0.5
10 ± 1.4
91 ± 1.3
2.8 ± 0.5
96 ± 1.5
D
if
f
e
r
e
n
t
ia
t
io
n
P
r
o
li
f
e
r
a
t
io
n
D
if
f
e
r
e
n
t
ia
t
io
n
P
r
o
li
f
e
r
a
t
io
n
iPAX7-IPRN
14.57
94 ± 1.01
2.9 ± 0.3
7.9 ± 1.01
87 ± 2.3
3.09 ± 0.8
93 ± 1.9
Expansion of PAX7
progenitors
14%
GFP
D
E
vector encoding PAX
was detected by inc
stream of the PAX7 g
PAX7 induction in the
cence analyses, whic
GFP upon doxycyclin
modification did not al
or their ability to di
(Figure 1A).
In embryogenesis, P
myogenic fate within p
tiated iPAX7 human (h
lowed by 3 days in m
(Figure 1A). This time
Cell Stem Cell 10, 610
differentiation by adding dox to the myogenic
medium. GFP+ (PAX7+) cells emerge in these
cultures and begin to proliferate. GFP+ cells are
purified by FACS (IV). Representative FACS profile
shows PAX7 (GFP) expression after 4 days of dox
induction in H9 differentiating ES cells. The
percentage indicated represents the fraction of
GFP+ cells (IV). PAX7+ myogenic progenitors are
layer
PAX7 induction
+ dox
Figure 1. Myogenic Induction of Human ES/
iPS Cells by PAX7
(A) Schematic of differentiation protocol with
representative morphological aspects of iPAX7
H9: in the undifferentiated state as ES cell colonies
in mTeSR medium (I), and in the EB stage (II). At
day 7 of differentiation, EBs are collected and
plated on a gelatinized flask to grow as a mono-
layer (III). PAX7 induction is initiated at day 10 of
expanded in myogenic induction medium sup-
plemented with dox and human bFGF (V). Scale
bars represent 100 mm.
(B) Growth curve of PAX7-induced ES- and iPS-
derived myogenic progenitors during in vitro
expansion. Data represent mean ± SE of four
independent experiments.
(C–E) Immunostaining of PAX7-induced human
ES-derived (C) and iPS-derived (D and E)
myogenic cells for PAX7, MYOGENIN, andMHC in
proliferation (top) and differentiation (bottom)
conditions. With PAX7 induction under prolifera-
tion conditions, most cells express PAX7 and only
a few express markers of terminal differentiation
(top panels), whereas under differentiation condi-
tions (and dox withdrawal), almost all of the cells
become positive for MYOGENIN and MHC, form-
ing multinucleated myotubes (bottom panels).
Cells were costained with DAPI (blue). Numbers on
each panel represent the percentage of cells ex-
pressing PAX7, MYOGENIN, or MHC. Data are
mean ± SE. For each condition, four slides were
used for quantification. Scale bars represent
100 mm. See also Figure S1.
human H9 ES cell line and two well-char-
acterized human iPS cell lines,
IPRN13.13 and IPRN14.57 (see Figures
S1A–S1F available online), generated
from fibroblasts from normal donors,
with a doxycycline-inducible lentiviral
7 (iPAX7). Expression of the transgene
orporating an ires-GFP reporter down-
ene (Figure S1G). Further confirmation of
se cells was provided by immunofluores-
h showed coexpression of PAX7 and
e (dox) induction (Figure S1H). Genetic
ter the morphology of the pluripotent cells
fferentiate into embryoid bodies (EBs)
AX7 and its homolog PAX3 act to confer
araxial mesoderm. We therefore differen-
) ES and iPS cells for 7 days as EBs fol-
onolayer before inducing PAX7 with dox
point is well into the peak of mesoderm
–619, May 4, 2012 ª2012 Elsevier Inc. 611
CD
IPRN
13.13
98% 88% 98% 100%
100%
CD56 α 7 INTEGRIN M- CADHERIN CD29
70% 98% 100%
H9
IPRN
14.57
98% 93% 98% 100%
A
generation, as indicated by Brachyury expression (Figure S1I).
Following 4 days of induction, PAX7+GFP+ cells were purified
by fluorescence-activated cell sorting (FACS) and expanded in
secondary monolayer culture in proliferation medium containing
dox and bFGF (Figure 1A and Figure S1J). Both ES- and iPS-
derived myogenic progenitors demonstrated notable expansion
potential, averaging 86-fold by week 2 (Figure 1B), with a total of
six to seven doublings during this period. Under these prolifera-
tion conditions, iPAX7 hES and hiPS cells expressed PAX7
abundantly (Figures 1C–1E and Figures S1K and S1L).
MYOGENIN and MYOSIN HEAVY CHAIN (MHC), markers of
terminal muscle differentiation, were barely detectable (Figures
1C–1E). This profile changed when iPAX7 hES and hiPS cells
were subjected to differentiation (5% horse serum and with-
drawal of dox and bFGF; differentiation medium). In these
culture conditions for 2 weeks, human myogenic progenitors
differentiated into multinucleated myotubes, with abundant
expression of MYOGENIN and MHC, while rare cells expressed
PAX7 (Figures 1C–1E). These results were confirmed by gene
Human DYSTROPHIN Human / pan-dystrophin Human DYSTROPHIN Human / pan-dystrophin
I
P
R
N
1
3
.1
3
C
o
n
t
r
o
l
I
P
R
N
1
4
.4
7
H
9
B
0
80
160
H9 iPS1 iPS2
Total Number of Human DYSTROPHIN
Positive Fibers/TA Section
A
v
e
ra
g
e
N
u
m
b
e
r
iPS1: IPRN 13.13
iPS2: IPRN 14.57
C
ED
F
myogenic progenito
(Figure S1M).
Human ES- and iPS-
Display Similar Surf
We characterized s
myogenic progenitors
Our results show a r
ure 2A and Figure S2A
(Figure 2A and Figures
showed homogenous
M-CADHERIN, and a
markers are associat
myogenic progenitors
et al., 2008; Sherwoo
has not yet been defi
been considered to be
(Pe´ault et al., 2007).
high levels of CD63,
612 Cell Stem Cell 10, 610–619, May 4, 2012 ª2012 Elsevier Inc.
dystrophin (left, in green), as evidenced by the use
of a pandystrophin antibody. (C–E) Engraftment of
proliferating myogenic progenitors obtained from
PAX7-induced human ES-derived (C) and iPS-
derived (D and E) cells in TA muscles of NSG (n = 4
for each cell line) 2 months after intramuscular
transplantation. Immunofluorescence stainingwith
anti-human (in red) and anti-pandystrophin (in
green) antibodies reveals presence of donor-
98%
44
98%
98%
Figure 2. Phenotypic Profile and Regenera-
tive Potential of Human ES/iPS-Derived
Myogenic Cells
(A) Representative FACS profile of PAX7-induced
humanES- and iPS-derived proliferatingmyogenic
progenitors. Histogram plots show isotype control
staining profile (gray line) versus specific antibody
staining profile (red line). Percentages represent
the fraction of cells that express a given surface
antigen. See also Figure S2.
(B–F) Transplantation ofmyogenic progenitors into
cardiotoxin-injured NSG mice. (B) PBS-injected
control muscles show no staining for human-
specific DYSTROPHIN (right, in red), but, as ex-
pected, do show uniform expression of mouse
Cell Stem Cell
Muscle Engraftment from Human ES and iPS Cells
derived myofibers expressing human DYSTRO-
PHIN in recipient muscles (in red). Scale bars
represent 100 mm. (F) Quantification of human
DYSTROPHIN+ fibers in engrafted muscles
shows similar engraftment of human ES- versus
iPS-derived myogenic progenitors. For this, the
total number of human DYSTROPHIN+ fibers in
cross-sections of TA muscles (sections spanned
entire muscles) was counted. Data are shown as
mean ± SE.
expression analyses, which showed
high levels of PAX7 expression solely
under proliferation conditions (in the
presence of dox) (Figure S1M) and upre-
gulation of MYOD and late skeletal
muscle-specific markers, MYOGENIN,
DYSTROPHIN, and MHC, when these
rs had undergone final maturation
Derived Myogenic Progenitors
ace Marker Profile
urface marker expression of these
by FACS using a panel of antibodies.
emarkable similarity between hES- (Fig-
) and hiPS-derived myogenic progenitors
S2B and S2C). Cells in each preparation
expression of CD56, CD29, CD44,
7-INTEGRIN. Although most of these
ed with murine satellite cells and early
(Cornelison and Wold, 1997; Sacco
d et al., 2004), the human satellite cell
ned by flow cytometry. Only CD56 has
a reliable marker of human satellite cells
These cells were also found to express
CD146, CD105, CD90, and CD13; the
Human LAMIN AC Human DYSTROPHIN Merge
Control
H9
IPRN 13.13
IPRN 14.57
0
50
100
H9 iPS1 iPS2
Total Number of Human DYSTROPHIN
Positive Fibers/TA Section
A
ve
ra
ge
N
um
be
r
s
0
20
40
60
gr
am
2 6 10 14 16
H9 iPS1 iPS2NSG NSG
mdx iPS1: IPRN 13.13
iPS2: IPRN 14.57
Cell
PBS
PBS
Cell
NSG
NSG- mdx
Keys:
0
25
50
75
1
F0
gr
am
* ***+
H9 iPS1 iPS2NSG NSG
-mdx
0
2.5
5
7.5
1
CSA
m
m
2
+++
H9 iPS1 iPS2NSG NSG
-mdx
Specific force (sF0)
K
N
/m
2
0
45
90
135
1
* *
**+++
H9 iPS1 iPS2NSG NSG
-mdx
0
10
20
30
1
Fatigue Index
T
im
e
(s
) +++
H9 iPS1 iPS2NSG NSG
-mdx
Weight
0
30
60
90
1
m
g
+++
H9 iPS1 iPS2NSG NSG
-mdx
A
CB
FED
HG
Figure 3. Efficient Engraftment and Functional Recovery after Transplantation of Human ES/iPS-Derived Myogenic Progenitors into Dystro-
phic Mice
(A) While no staining for human LAMIN AC or DYSTROPHIN is detected in PBS-injected control TA muscles of NSG-mdx4Cv mice (top), abundant expression for
human LAMIN AC (in green) and DYSTROPHIN (in red) is observed (bottom) in dystrophic muscles treated with human ES/iPS-derived myogenic progenitors
1 month after the transplantation (n = 5 for H9, n = 6 for IPRN13.13, and n = 7 for IPRN 14.47). Note that nuclear LAMIN AC staining occurs predominantly within
human DYSTROPHIN+ myofibers. Scale bars represent 100 mm. See also Figure S3.
Cell Stem Cell
Muscle Engraftment from Human ES and iPS Cells
Cell Stem Cell 10, 610–619, May 4, 2012 ª2012 Elsevier Inc. 613
last three are antigens known to be present in mesenchymal
stem cells (Pittenger and Martin, 2004). CD34 labeled a discrete
subfraction of these cells. Other screened antigens, including
CD45, CD33, KDR, and CD31, were undetectable in these
myogenic progenitor populations (Figure S2), indicating the
absence of hematopoietic and endothelial cells. The adhesion
Functional Improvement in Dystrophic Mice
To determine the regenerative potential of these myogenic
progenitors in the context of muscular dystrophy, we trans-
planted them into mdx mice engineered to lack B, T, and NK
cells. These mice were generated by crossing mice carrying
the mdx4Cv mutation, an ENU-induced stop codon in exon 53
ice
se
jur
Cell Stem Cell
Muscle Engraftment from Human ES and iPS Cells
molecules CXCR4 and CD106 were also not detected. We
examined MAJOR HISTOCOMPATIBILITY COMPLEX (MHC)
expression because the lack of MHC class I expression on
other embryonic and ES-derived cells has limited engraftability,
even in immunodeficient mice, due to NK cell-mediated
responses (Rideout et al., 2002; Tabayoyong et al., 2009).
This analysis revealed that, regardless of ES or iPS origin,
proliferating myogenic progenitors express MHC class I mole-
cules (Figure S2). This pattern is beneficial from the perspective
of avoiding an NK-mediated lack of self-MHC response but
indicates the importance of HLA matching.
In Vivo Regenerative Potential of HumanES/iPS-Derived
Myogenic Progenitors
Next we examined the in vivo skeletal muscle regenerative
potential of iPAX7 hES- and hiPS-derived myogenic progenitors
by transplanting these cells directly into the tibialis anterior (TA)
muscles of NOD/SCID gamma-c (NSG) mice, an immune-defi-
cient strain commonly used as a recipient of human hematopoi-
etic cells. The gamma-c mutation (IL2Rg) ablates NK cells,
rendering NSG mice unable to reject human cells due to lack
of self-MHC presentation, resulting in better hematopoietic
engraftment than in mice bearing the NOD/SCID mutation alone
(Shultz et al., 2005). NSG mice were injured with cardiotoxin
(CTX) 24 hr prior to cell transplantation. The contralateral TA
muscle, which served as a control, was also preinjured with
CTX but injected only with PBS. Two months after transplanta-
tion, muscle sections were harvested and evaluated for engraft-
ment by immunostaining with both pandystrophin and human-
specific DYSTROPHIN antibodies. No expression of human
DYSTROPHIN could be detected in PBS-injected control
muscles (Figure 2B); staining was only observed with a pandy-
strophin antibody (Figure 2B). On the other hand, muscles that
had been treated with iPAX7 hES- (Figure 2C) and hiPS-derived
(Figures 2D and 2E) myogenic progenitors demonstrated
engraftment of human-derived myofibers, as evidenced by the
clear expression of human-specific DYSTROPHIN in recipient
muscles (Figures 2C–2E). We did not observe major differences
in terms of engraftment between ES- and iPS-derived myogenic
progenitors (Figure 2F). No tumor formation was observed in
transplanted mice, even in a long-term (46 weeks) cohort.
(B) Quantification of human DYSTROPHIN+ fibers in NSG-mdx4Cv engrafted m
progenitors. For this, the total number of human DYSTROPHIN+ fibers in cross-
shown as mean ± SE.
(C) Representative example of force tracings in TAmuscles of nontreated, nonin
4Cv
NSG-mdx mice that had been injected with PBS (control, red line) or human E
(D and E) Effect of iPAX7 human ES/iPS-derivedmyogenic cell transplantation on
nontreated, noninjured NSG (purple) and NSG-mdx4Cv (brown) mice are shown f
(F and G) Weight and CSA of control and transplanted muscles, respectively. Valu
shown for reference. See also Figure S4. Data are shown as mean ± SE.
(H) Fatigue index: time for force to decline to 30%of itsmaximal value shows no sig
as mean ± SE. *p < 0.05, **p < 0.01 compared to its PBS control. +p < 0.05, +++p
614 Cell Stem Cell 10, 610–619, May 4, 2012 ª2012 Elsevier Inc.
(Im et al., 1996) with very low reversion frequency (Danko
et al., 1992), t