healthmedicinet

Derivation and ex vivo expansion of satellite cells using Pax3

To determine whether SCs maintained engraftment potential when expanded ex vivo using
conditional expression of Pax3, we followed the strategy summarized in Fig. 1a, in which SCs from the transgenic Pax7-ZsGreen reporter mouse 27] were (I) purified by flow cytometry, (II) genetically modified with a lentiviral
vector encoding a doxycycline-inducible Pax3 transgene, (III) expanded ex vivo in
the presence of doxycycline, and then (IV) transplanted into immune-deficient, dystrophin-deficient
NSG-mdx ^4Cv31] mice. After enzymatic digestion, the muscle mononuclear fraction of Pax7-ZsGreen
was FACS-purified based on ZsGreen expression, which reflects Pax7
⁺
cells (Fig. 1b), and accordingly gave rise to a homogeneous SC population (Fig. 1c). These cells were immediately transduced with a doxycycline-regulated conditional
Pax3-IRES-mCherry-expressing lentivector (Pax3 induced) 32]. As a control, SCs were transduced with empty vector (mCherry only). Pax3
⁺
(mCherry
⁺
) cells were detected only when doxycycline (dox) was added to the culture medium
(Additional file 1). To determine the effect of Pax3 on the expansion of transduced SCs, we evaluated
the proliferation rate of Pax3-induced cells side-by-side with control cell preparations
(empty vector) grown under identical culture conditions: proliferation medium with
basic fibroblast growth factor (bFGF) and dox. Notable expansion advantage was observed
in Pax3-induced cultures when compared to control counterparts (Fig. 1d). Although under these proliferation conditions, both control and Pax3-induced cells
displayed similar morphology (Fig. 1e, f, panel I), only Pax3-induced cells showed abundant Pax3 expression, as evidenced by immunofluorescence
staining (Fig. 1e, f, panel II) and gene expression analyses (Additional file 2). As expected, Pax3 overexpression in SCs was accompanied by upregulation of its
target gene Myf5 35] (Additional file 2). Under proliferation conditions, Pax3-induced cells showed no signs of myotube formation,
as indicated by the absence of signal for myosin heavy chain (MHC) (Fig. 1f, panel III, and Additional file 2), whereas the control uninduced population spontaneously differentiated into MHC-positive
myotubes (Fig. 1e, panel III, and Additional file 2). Nevertheless, when Pax3-induced and control cells were subjected to differentiation
conditions (5 % horse serum and withdrawal of dox and bFGF), both cultures gave rise
to multinucleated myotubes displaying abundant expression of MHC (Fig. 1e, f, panels IV and V, and Additional file 2). Control cultures expressed significant levels of MHC under proliferation conditions,
suggesting the propensity of these cells to begin differentiation as soon as they
have reached confluence. We next quantified the fusion index of control and Pax3-induced
cultures. Upon in vitro differentiation, Pax3-induced SCs exhibited elevated fusion
index (67 %) relative to control cultures (47 %). Thus, under the conditions tested
here, Pax3 induction allows the in vitro expansion of less differentiated SCs, without
affecting their ability to terminally differentiate into fusing myotubes.

Fig. 1. Derivation and characterization of Pax3-induced satellite cells. a (I) FACS purification of satellite cells based on ZsGreen expression (Pax7), (II) transduction of Pax7
⁺
cells with an inducible expression system encoding Pax3, (III) in vitro expansion of Pax3-induced cells and control empty vector counterparts,
and (IV) transplantation of iPax3 and control cells into NSG-mdx ^4Cv
. b Representative FACS profile for ZsGreen (Pax7) expression in digested muscles isolated
from Pax7-ZsGreen reporter mice. Sorting gate for ZsGreen
⁺
(Pax7
⁺
) satellite cells is shown. c Phase-contrast image of sorted ZsGreen
⁺
(Pax7
⁺
) satellite cells. d Cell growth curve of Pax3-induced cells and control counterparts at several passages
(P1–P4) (?=?2, mean?±?SD). e, f In vitro characterization of ex vivo expanded satellite cells grown under proliferation
and differentiation culture conditions. Phase-contrast images of control empty vector
(e) and Pax3-induced (f) monolayers. Representative immunofluorescence staining for Pax3 (red, upper panels) and MHC (red, lower panels) in control empty vector SCs (e) and Pax3-induced SCs (f). Cells are co-stained with DAPI (blue). Scale bar 200 ?m. g Fusion index calculation. Error bars represent s.e.m. (?=?3). **P??0.01

In vivo regenerative potential of ex vivo expanded satellite cells

To evaluate in vivo repopulation potential after 1 week of ex vivo expansion, Pax3-induced
and respective control cell preparations were transplanted into the TA muscles of
NSG-mdx ^4Cv
mice. Prior to cell transplantation, both hind limbs were subjected to irradiation
(12 Gy/leg) to deplete endogenous SCs 31] and injury with cardiotoxin (CTX). While the contra-lateral TA was injected with
PBS, 350,000 Pax3-induced or control cells were injected into the right TA. Five weeks
after transplantation, TA muscles were harvested and evaluated for engraftment by
immunofluorescence staining for dystrophin. Whereas DYS
⁺
myofibers were virtually undetectable in PBS-injected muscles (Fig. 2a, c), dystrophin expression was observed in TA muscles that had been transplanted with
control (Fig. 2b) or Pax3-induced (Fig. 2d) cell preparations, with the latter showing higher engraftment levels (Fig. 2e, 14?±?7.4 vs. 37?±?5.7 %, respectively).

Fig. 2. Regenerative potential of Pax3-induced satellite cells following their transplantation
into NSG-mdx ^4Cv
mice. Engraftment analysis of control empty vector (a, b) and Pax3-induced cells (iPax3) (c, d). Cross sections of TA muscles harvested from NSG-mdx ^4Cv
mice that had been injected with PBS (a, c) or satellite cells (b, d) were stained with antibody to dystrophin (red). Engrafted tissues from control and Pax3-induced cells are represented by mice #03 and #05 and #07 and #09, respectively. DAPI is shown in blue. Scale bar, 50 ?m. e Quantification of DYS
⁺
myofibers in treated muscles. Error bars represent s.e.m. (?=?6). *P??0.03

Next, we determined whether myofiber engraftment was accompanied by improvement in
muscle strength. As expected, the maximum isometric force for PBS-injected TA muscles
(contra-lateral legs) was low (Fig. 3a, gray lines). In contrast, engrafted TA muscles showed enhanced isometric force (Fig. 3a, red lines). Cell transplantation of both control and Pax3-induced preparations resulted
in improved absolute (Fig. 3b) and specific (Fig. 3c) force of engrafted muscles when compared with their respective PBS-injected contra-lateral
muscles. However, muscles that had been transplanted with Pax3-induced cells displayed
significantly superior functional improvement (Fig. 3b, c) when compared to control cells (1.52-fold). No statistical difference was observed
in forces between the contra-lateral legs (PBS) of the two groups of mice. These results
demonstrate that 7-day cultured SCs expanded with Pax3 have a superior ability to
improve muscle function, compared to control empty vector transduced counterparts.

Fig. 3. Contractile properties of transplanted muscles and satellite cell homing. a Representative examples of maximum isometric tetanic force in TA muscles that had
been injected with PBS (contra-lateral leg, gray line) and control or Pax3-induced cells (red lines). Wild-type Bl6 mice were used for reference control (dashed line). b, c Cell transplantation produces an improvement in absolute (F₀
, b) and specific (sF₀
= F₀
normalized to CSA, c) force. Error bars represent s.e.m. from a total of six mice. *P??0.05, **P??0.01, ***P??0.001. d In situ analysis reveals the presence of donor-derived satellite cells (ZsGreen/Pax3-induced
cells) in the host stem cell pool, as shown by the presence of cells co-stained for
both Pax7 (red) and ZsGreen (green) (white arrow) beneath the basal lamina (gray). e Upon reinjury, engrafted donor-derived satellite cells give rise to newly formed
myofibers, as indicated by the co-expression of DYS (red) and embryonic MHC (green) (white arrow). Arrowheads denote DYS
^?
/eMHC
⁺
host-derived new formed myofibers. DAPI is shown in blue. Scale bar, 50 ?m

To assess whether Pax3-induced cells have the capacity to engraft the host SC compartment,
and therefore contribute to ongoing regeneration, engrafted TA muscles were stained
for ZsGreen and Pax7 to identify donor-derived SC contribution. Histological analysis
of transverse sections of TA muscles 1 month after transplantation clearly identified
the presence of Pax7
⁺
ZsGreen
⁺
cells beneath the basal lamina, suggesting that Pax3-induced cells can engraft the
SC pool (Fig. 3d). To investigate whether donor-derived iPax3 SCs would be able to contribute to ongoing
muscle regeneration, a cohort of mice transplanted with unlabelled Pax3-induced cells
were reinjured with CTX 1 month after cell transplantation. Ten days after reinjury,
we detected donor-derived newly regenerated myofibers, as indicated by the presence
of DYS
⁺
/embryonic MHC
⁺
myofibers (Fig. 3e, white arrows). Since we have used half of the usual dose of CTX (5ul/5uM, instead
of 10ul/10uM) for these reinjury studies, CTX injection did not result in degeneration
of the whole tissue, and accordingly the presence of DYS
⁺
/eMHC
^?
fibers was detected. These results suggest that at least some of engrafted Pax3-induced
cells remain less differentiated and are able to respond to a second round of muscle
injury.

Genetic repair of dystrophic Pax3-induced cells

We next applied genetic correction to ex vivo expanded dystrophic SCs following the
protocol outlined in Fig. 1a, but using SCs harvested from mdx mice bred to carry the Pax7-ZsGreen reporter (Fig. 4a). For genetic repair, we used the human micro-dystrophin
^{?R4–23/?CT}
(?DYS) transgene lacking the spectrin-like repeats 4–23 and the C-terminus 36] and the non-viral Sleeping Beauty system for transduction. First, we generated a Tn vector (pKT2-Neo selection marker driven
by the-Neo/hH2 ?DYS; Fig. 4b) containing two divergent genes: a GFP/Neo selection marker driven by the hEF1a-eIF4g
promoter and the human ?Dystrophin (?DYS) gene under the control of a pHSA 37].

Fig. 4. Correction of dystrophin-deficient Pax3-induced satellite cells using a human ?DYS transgene. a FACS plot shows gate for the purification of ZsGreen
⁺
(Pax7
⁺
) satellite cells from Pax7-ZsGreen/mdx mice. b The Sleeping Beauty transposon system consists of transposon (Tn) and transposase (SB100X) vectors. The Tn is a
bicistronic promoter vector of 11.3 Kb containing the ubiquitin hEF1a-eIF4g (Pr, in
gray) and the skeletal muscle-specific skeletal ?-actin promoter (pHSA, in black). The ubiquitin promoter drives a GFP-2A-Neo. This selection marker cassette is flanked
by lox P sequences (red). The human ?DYS gene is under control of the pHSA. SB100X transposase proteins (red spheres) bind the DR sequences (yellow arrows) within the two inverted repeats (IR/DR, arrowheads) and catalyze integration of the whole transposon transgene into the genome with
high efficiency. c Representative FACS profiles for enrichment steps used to isolate a pure and stable
population of corrected GFP
⁺
cells (?DYS-Pax3-induced cells) following transfection with pKT2/?DYS and SB100X. Control consisted of dystrophin-deficient Pax3-induced cells (CTL) nucleofected with pKT2 transposon vector only (no transposase). d RT-PCR analysis for uncorrected (UNC, dystrophin-deficient Pax3-induced cells) and corrected (Corr, ?DYS-Pax3-induced cells) cells grown under proliferation (P) and differentiation (D) culture conditions shows the expression of human ?DYS solely in corrected cells. GAPDH was used as housekeeping gene

SCs were isolated by flow cytometry from Pax7-ZsGreen/mdx mice (Fig. 4a), immediately transduced with the doxycycline-inducible Pax3 vector, and grown in
doxycycline to induce Pax3 expression. It should be noted that almost immediately
upon placing the Pax7-ZsGreen SCs into culture, the ZsGreen fluorescence is lost.
We now then transduced these non-fluorescent cells with the ?-dystrophin correction
vector, which contained a GFP reporter, and sorted on this signal; therefore, the
culture was now constitutively green. Dystrophin-deficient Pax3-induced cells were
subsequently nucleofected with Tn vector and transposase (engineered hyperactive variant
SB100X 38]; Fig. 4b, upper panel), using a plasmid ratio of 4:1, respectively, which we have previously
found to provide optimal in vitro gene transfer for a large transgene 32]. Five days after nucleofection, flow cytometry analysis revealed a cell sub-population
positive for GFP/?DYS (~1.2 %) (Fig. 4b, lower panel). Following two rounds of sorting, a highly enriched ?DYS⁺
(GFP
⁺
) population was obtained (96 %) (Fig. 4b, lower panel). Expression of the transgene in corrected cells was confirmed by RT-PCR
analysis using specific primers for the human ?DYS transgene (Fig. 4c). These results demonstrate the capacity for the Sleeping Beauty system to deliver a large transgene (11.3 Kb) into dystrophic activated SCs.

Regenerative potential of ?DYS-Pax3-induced cells

To assess the regenerative potential of corrected ?Dys-Pax3-induced cells in vivo,
these cells were transplanted into CTX-injured TA muscles of NSG-mdx ^4Cv
mice. We did not irradiate these mice as irradiation would be discouraged in the clinical
setting. One month following transplantation, TA muscles were harvested and sections
were evaluated for engraftment by immunostaining using a human DYSTROPHIN antibody
that recognizes the N-terminal epitope, which is preserved in the human ?DYS transgene. While no DYS expression was detected in PBS-injected muscles (Fig. 5a), muscles that had been transplanted with ?DYS-Pax3-induced cells generated large engrafted areas with DYS
⁺
myofibers (Fig. 5b). Quantification of engraftment revealed that approximately 20 % of fibers in transplanted
muscles were ?DYSTROPHIN
⁺
, confirming the regeneration potential of ex vivo corrected activated SCs.

Fig. 5. Engraftment of ?DYS-Pax3-induced cells into NSG-mdx ^4Cv
mice. TA muscles harvested from NSG-mdx ^4Cv
mice that had been injected with PBS (a) or corrected ex vivo expanded satellite cells (?DYS-Pax3-induced cells) (b) were stained using an antibody specific for human DYSTROPHIN (red). The DYS protein was detected only in the transplanted muscles. Two representative
transplanted mice (b) are shown. DAPI is shown in blue. Scale bar, 50 ?m. c Quantification of human ?DYSTROPHIN
⁺
myofibers in these transplanted muscles. Error bars represent s.e.m (?=?6)

We next investigated whether engraftment of corrected ?DYS-Pax3-induced cells was accompanied by functional improvement. Engrafted muscles showed
superior isometric (Fig. 6a), absolute (Fig. 6b), and specific (Fig. 6c) force when compared to PBS-injected TA muscles.

Fig. 6. Contractile function and response to reinjury by muscles engrafted with ?DYS-Pax3-induced cells. a Representative examples of maximum isometric tetanic force in TA muscles that had
been injected with PBS (contra-lateral leg, gray line) or Pax3 induced (red line). b, c?DYS-Pax3-induced cell transplantation produced a significant improvement in absolute
(F₀
, b) and specific (sF₀
= F₀
normalized to CSA, c) forces. Error bars represent s.e.m. from a total of six mice. **P??0.01. d Immunofluorescence staining for embryonic MHC and ?DYS in engrafted TA muscles analyzed
10 days after CTX reinjury indicates the presence of newly formed donor myofibers
as denoted by co-expression of human ?DYS (red) and eMHC (green) (arrows). Arrowheads show ?DYS
^?
/eMHC
⁺
host-derived newly formed myofibers. Alexa-647 was used to detect eMHC. DAPI is shown
in blue. Scale bar, 50 ?m

To determine whether engrafted ?DYS-corrected Pax3-induced cells would have the same ability to respond to injury as
shown above for WT cells and would therefore be capable of providing ?DYSTROPHIN continuously,
we reinjured muscles that had been previously transplanted with ?DYS-Pax3-induced cells. Ten days following CTX injection, we stained muscle sections
with embryonic MHC and human DYS antibodies. This clearly showed the presence of donor-derived
newly regenerated muscle fibers that were double-positive for ?DYS and embryonic MHC
(Fig. 6d, white arrows and Additional file 3). Altogether, these results show that transplantation of corrected ?DYS-Pax3-induced cells provides functional improvement of dystrophic muscles, both in
terms of muscle force generation and in terms of their ability to respond to ongoing
muscle injury and stably express ?DYS protein.

SCs isolated by flow cytometry have been demonstrated to possess a tremendous capacity
to improve muscle function in mdx mice 31]; however, the impracticality of isolating large numbers of SCs from living donors
as well as the requirement for gene correction, if considering an autologous transplantation
setting, necessitates ex vivo expansion. To date, only one study has reported a combined
cell/gene therapy approach using SCs in the context of muscular dystrophy 18]. In this study, the authors isolated SCs from a dystrophic mouse, transduced them
with a lentiviral vector encoding the mouse ?DYS transgene, and immediately transplanted them into the dystrophic muscle and found
that they were able to differentiate into DYS+ fibers.

Several studies have investigated the transplantation of cultures derived from prospectively
isolated SCs. Blau and colleagues demonstrated that culturing mouse SCs on a substrate
that mimics muscle tissue elasticity, and in the presence of an inhibitor for p38MAPK,
helped maintain “stemness” features 10], 39]. Following a different approach, Tapscott and colleagues expanded freshly isolated
canine SCs by activating the Notch signaling pathway, which bestowed superior in vivo
regenerative ability upon SC-initiated cultures compared to controls 20]. In a recent study, Rudnicki and colleagues reported that short treatment of SCs
with Wnt7a resulted in enhanced engraftment that was accompanied by improved muscle
function 40].

Herein, we demonstrate that upon conditional expression of Pax3, freshly isolated
SCs can be successfully expanded when compared to their cultured empty vector control
counterparts (Fig. 1d). Following their intramuscular transplantation into dystrophic mice, Pax3-induced
cells display greater regenerative potential than control SCs, and engraftment levels
correlated with a significant improvement in muscle strength (Fig. 3a–c). Importantly, we also show that engrafted Pax3-induced cells are capable of seeding
the SC pool and responding to a second round of CTX-induced damage by generating newly
formed DYS
⁺
fibers (Fig. 4d, e). In addition, we show that Pax3-induced dystrophic SCs are amenable to genetic correction.
Using a non-viral Sleeping Beauty system carrying a human ?DYS transgene, we corrected SCs from dystrophin-deficient mice and found that these were
capable of differentiating into functional muscle fibers in vivo (Fig. 5), increasing force generation capacity of dystrophic muscles (Fig. 6a–c), and producing new myofibers upon CTX reinjury that remain positive for the ?DYS transgene.