ARTICLES
Mesoangioblast stem cells ameliorate
muscle function in dystrophic dogs
Maurilio Sampaolesi
1,2
*, Stephane Blot
3
*, Giuseppe D’Antona
2
, Nicolas Granger
3
, Rossana Tonlorenzi
1
,
Anna Innocenzi
1
, Paolo Mognol
4
, Jean-Laurent Thibaud
3
, Beatriz G. Galvez
1
, Ines Barthe
´
le
´
my
3
, Laura Perani
1
,
Sara Mantero
4
, Maria Guttinger
5
, Orietta Pansarasa
2
, Chiara Rinaldi
2
, M. Gabriella Cusella De Angelis
2
,
Yvan Torrente
6
, Claudio Bordignon
1
, Roberto Bottinelli
2
& Giulio Cossu
1,5,7
Duchenne muscular dystrophy remains an untreatable genetic disease that severely limits motility and life expectancy in
affected children. The only animal model specifically reproducing the alterations in the dystrophin gene and the full spectrum
of human pathology is the golden retriever dog model. Affected animals present a single mutation in intron 6, resulting in
complete absence of the dystrophin protein, and early and severe m uscle degeneration with nearly complete loss of motility
and walking ability. Death usually occurs at about 1 year of age as a result of failure of respiratory muscles. Here we report
that intra-arterial delivery of wild-type canine mesoangioblasts (vessel-associated stem cells) results in an extensive
recovery of dystrophin expression, normal muscle morphology and function (confirmed by measurement of contrac tion
force on single fibres). The outcome is a remarkable clinical amelioration and preservation of active motility. These data
qualify mesoangioblasts as candidates for future stem cell therapy for Duchenne patients.
Duchenne muscular dystrophy primarily affects skeletal muscle,
causing fibre degeneration, progressive paralysis and death
1
.No
effective treatment exists although novel therapeutic strategies, rang-
ing from new drugs to gene and cell therapy, hold promise for sig-
nificant advance in the future
2
. In particular, different types of stem
cell have been shown to induce dystrophin synthesis and partial
rescue of the pathology in dystrophic mice
3–8
. However, dystrophic
mice do not display clinical signs of the disease, and to proceed to a
clinical trial it is imperative to show efficacy in a large, non-syngeneic
animal model of muscular dystrophy. Golden retriever muscular
dystrophy (GRMD)
9,10
is a very severe form of dystrophy, which
affects not only limb, respiratory and heart muscles but also pharyn-
geal muscles, resulting in a severe involvement of the digestive tract;
although variability exists between individuals, by 8 months of age
most dogs walk with great difficulty (Supplementary Movie 1). To
test the efficacy of cell or gene therapy, we transplanted GRMD dogs
with either autologous genetically corrected or donor wild-type
mesoangioblasts, under different regimes of immune suppression.
Ten dystrophic dogs were treated in three experiments and a gen-
eral scheme of treatments and outcome is reported in Table 1. Four
dogs received autologous mesoangioblasts, transduced in vitro with a
lentiviral vector expressing human microdystrophin (Supplementary
*These authors contributed equally to this work.
1
San Raffaele Scientific Institute, Universita
`
Vita e Salute, Stem Cell Research Institute, Via Olgettina 58, 20132 Milan, Italy.
2
Department of Experimental Medicine and Interuniversity
Institute of Myology, University of Pavia, Via Forlanini 6-8, 27100 Pavia, Italy.
3
Neurobiology Laboratory, E
´
cole Ve
´
te
´
rinaire d’Alfort, 7 Avenue Ge
´
ne
´
ral de Gaulle, 94704 Maisons-Alfort
cedex, France.
4
Department of Bioengineering, Politecnico di Milano, Piazza Leonardo Da Vinci, 20130 Milan, Italy.
5
Institute of Cell Biology and Tissue Engineering, San Raffaele
Biomedical Science Park of Rome, Via Castel Romano 100, 00128 Rome, Italy.
6
IRCCS Fondazione Policlinico di Milano, Department of Neurological Sciences, University of Milan, Via
Sforza 35, 20122 Milan, Italy.
7
Department of Biology and Centre for Stem Cell Research, University of Mil an, Via Celoria 28, 20130 Milan, Italy.
Table 1
|
Summary of treatment
Dog Dog Cell treatment Lentiviral vector Onset of
treatment
Immune suppression
(time)
Dystrophin
expression
Motility Outcome of experiment
(at time P400)no. name
01A Ucal Autologous, gene therapy CK-mdys-ires
–
GFP P118
–
1/2 Loss Euthanasia (P272)
02H Vrillie Heterologous, WT donor
–
P80 CYC A (P78) 1 Loss Euthanasia (P235)
03H Valgus Heterologous, WT donor
–
P75 CYC A (P73) 111 No decline Alive and well
04H Varus Heterologous, WT donor
–
P75 RAP (P73) 111 Modest decline Alive and well
05H Viko Heterologous, WT donor
–
P77 RAP 1 IL-10 (P74) ND ND (sudden death) Myocarditis (P186)
06A Vaccin Autologous, gene therapy MLC1F-mdys P113
–
11 Major decline Euthanasia (P326)
07A Valium Autologous, gene therapy MLC1F-mdys P113
–
ND Loss Pneumonia (P245)
08A Vampire Autologous, gene therapy MLC1F-mdys P113
–
11 Major decline Pneumonia (P154)
09H Azur Heterologous, WT donor
–
P159 CYC A (P157) 11 Restored Alive and well
10H Azor Heterologous, WT donor
–
P159 CYC A (P157) 111 Restored Alive and well
11U Akan None
–––
2 Loss Euthanasia (P380)
12U Vulcano None
–––
2 Loss Euthanasia (P376)
13U Viking None
–––
2 Loss Euthanasia (P340)
Each dog was given a specific name and a sequential number, followed by A (transplanted with autologous cells), H (transplanted with heterologous cells) or U (untreated). The nature of the lentiviral
vector is indicated (CK-mdys-ires
–
GFP, creatine kinase promoter driving microdystrophin-ires
–
GFP; MLCF1-mdys, myosin light chain 1 fast promoter driving microdystrophin). Dystrophin expression
was quantified as follows: 2, average dystrophin (or micro-dystrophin) expression in less than 1% of positive fibres; 1/2, less than 10% of positive fibres; 1, less than 20% of positive fibres; 11,
less than 50% of positive fibres; 111, more than 50% of positive fibres. CYC A, cyclosporine; ND, not determined; RAP, rapamycin; WT, wild-type. Euthanasia was administered when clinical
conditions worsened.
Vol 444
|
30 November 2006
|
doi:10.1038/nature05282
574
Nature
Publishing
Group
©2006
Fig. 1), a truncated but functional version of dystrophin
11
; six dogs
received wild-type mesoangioblasts from a single DLA (dog leuko-
cyte antigen)-unrelated donor under treatment with either cyclos-
porine or rapamycin. Three other dystrophic dogs were not treated
and served as controls. All dystrophic dogs received steroids daily as
standard treatment. Overall, the results showed that donor wild-type
mesoangioblasts significantly ameliorate many symptoms of canine
muscular dystrophy, whereas autologous genetically corrected cells
are much less effective.
Isolation and characterization of canine mesoangioblasts
Both wild-type and dystrophic mesoangioblasts were isolated from
the outgrowth of small, vessel-containing, tissue fragments from
muscle biopsies performed after diagnosis, at about 15 days postnatal
(P15). The cells show a morphology very similar to that of mouse
mesoangioblasts
12
(Fig. 1a), proliferate efficiently in a medium
devised for stem cells (Fig. 1b) and show a euploid kariotype of 78
chromosomes both at early and late passage (Fig. 1c); cells undergo
senescence after about 25 population doublings. Canine mesoangio-
blasts express CD44 and CD13 but not CD34, CD45, CD117 or CD31
(not shown). Dog mesoangioblasts differentiate into multinucleated
myotubes when co-cultured with C2C12 mouse myoblasts or when
transfected with MyoD. For the gene transfer experiments, prolif-
erating mesoangioblasts isolated from dystrophic dogs were trans-
duced with lentiviral vectors expressing human microdystrophin and
(only for Ucal, the first dog treated) enhanced green fluorescent
protein (EGFP). Both proteins became readily detectable after myo-
genic differentiation (Fig. 1d–f). Finally, to test the ability of these
cells to reconstitute muscle fibres in vivo, both wild-type and GRMD
genetically corrected mesoangioblasts were injected into SCID
(severe combined immunodeficiency)-mdx mice, which do not
reject xenogenic cells; the cells migrated from the femoral artery to
the downstream muscles with an efficiency similar to that of their
wild-type mouse counterparts
13
(Fig. 1g). Three weeks after injection,
canine mesoangioblasts gave rise to dystrophin-positive fibres con-
taining dog nuclei, identified by anti-human lamin A-C antibody,
which recognizes human and dog but not mouse nuclei (Fig. 1h, i).
Thus, dog mesoangioblasts seem similar to their mouse postnatal
counterparts by all the parameters tested, with the notable exception
of a finite lifespan, a predictable difference between cells from rodents
and other mammals.
Feasibility experiment
Dogs are identified by name and also by a sequential number fol-
lowed by a letter (A for autologous cell transplantation, H for het-
erologous cell transplantation, and U for untreated) (Table 1).
Two dogs, Ucal and Vrillie, were treated with three consecutive (at
1-month intervals) injections of 5 3 10
7
cells into the femoral artery.
Ucal (01A) received autologous cells, transduced with the lentiviral
vector expressing human microdystrophin (Supplementary Fig. 1a).
Vrillie (02H) received wild-type donor cells under a regimen of
cyclosporine treatment.
During and after the treatment, Ucal (01A) and Vrillie (02H) did
not show appreciable sign of clinical amelioration and underwent a
progressive decline in their walking ability. Biopsies, taken 1 month
after the third injection, revealed variable morphology in different
muscles of the injected legs, varying from severe and advanced degen-
eration in the tibialis cranialis of Ucal (01A), shown in Fig. 2a, to
an intermediate severity in the same muscle of Vrillie (02H), shown
in Fig. 2b. In general, the morphology of Vrillie (02H) was better
but still variable, with several areas being quite well preserved.
Dystrophin expression also showed variability and was in general
correlated with morphology. At 8 months of age, the proportion of
revertant fibres in dystrophic dogs varies from 0.02% to 0.3% (ref.
14). In the biopsies collected, dystrophin-positive fibres ranged from
2% to 7% in Ucal (01A) and from 4% to 10% in Vrillie (02H) (not
shown). Figure 3a, a9 shows a biopsy of Ucal in which clusters of
dystrophin-positive fibres with several centrally located nuclei can be
observed. Biopsies from contralateral, non-injected legs showed
poorer morphology and a smaller proportion (2% or less) of dystro-
phin fibres. Unexpectedly, we found areas of dystrophin expression
in the triceps brachialis of Vrillie (02H) (not shown), indicating that
cde
f
Migrated cells (%
)
iQd
0
5
10
30
iGs iTA uQd uGs uTA Lv Sp
g
i
Time (days)
10 15 2050
No. of cells
h
d′ e′
f′ f′′
10
9
10
8
10
7
10
6
10
5
10
4
10
3
ab
Figure 1
|
Characterization of dog mesoangioblasts. a, Morphology of
canine mesoangioblasts isolated from muscle biopsies of a golden retriever
dog at P15.
b, Proliferation curves of wild-type (filled circles) and dystrophic
(open circles) canine mesoangioblasts.
c, Karyotype of canine
mesoangioblasts, which are consistently euploid until senescence.
d–f, Transduction of dystrophic canine mesoangioblasts with a lentiviral
vector expressing human microdystrophin and the EGFP gene reporter
under the control of muscle-specific creatine kinase promoter. GFP
expression is undetectable in proliferating mesoangioblasts (
d, d9) but is
readily detected in myotubes derived from the fusion of transduced
mesoangioblasts with C2C12 myoblasts (
e, e9). Similar results were obtained
after MyoD-induced differentiation: GFP-positive cells (
f0) also express
MyHC (
f9) as confirmed in the merged image (f), which also shows
multinucleation.
g, Migration of canine mesoangioblasts into skeletal
muscle: 5 3 10
5
mouse (black bars) or dog (grey bars) mesoangioblasts,
previously transduced with a GFP-expressing lentiviral vector
4
, were
injected into the right femoral artery of SCID-mdx mice. Six hours after
injection, several muscles were isolated and the presence of donor cells was
measured by real-time PCR analysis for GFP as detailed elsewhere
13
. Qd,
quadriceps; Gs, gastrocnemius; Tc, tibialis cranialis, Lv, liver; Sp, spleen; the
letter i (injected) before the muscle name indicates muscle isolated from the
injected leg; the letter u (uninjected) indicates muscles isolated from the
contralateral leg.
h, Top: immunofluorescence with antibodies against
human lamin A-C (green) and dystrophin (red), revealing dog
mesoangioblasts inside the muscle fibres of SCID-mdx mice, 21 days after
intra-arterial injection. Bottom: nuclei were stained with 4,6-diamidino-2-
phenylindole (DAPI).
i, Immunofluorescence of fibres with antibodies
against dystrophin (red, top) and laminin (green, bottom) in the muscle of
SCID-mdx mice, 21 days after intra-arterial injection. Scale bar, 20 mm. Error
bars, 1 s.d.
NATURE
|
Vol 444
|
30 November 2006 ARTICLES
575
Nature
Publishing
Group
©2006
injected cells must to a certain extent transit through the capillaries of
the injected leg and then through the filter organs, finally entering the
arterial circulation to reach the forelimb muscles. Overall, the intens-
ity of staining was weak; indeed, western blot analysis confirmed very
low, barely detectable, levels of microdystrophin and dystrophin,
respectively. Immunohistochemistry detected several inflammatory
infiltrates containing mainly macrophages and lymphocytes (Supple-
mentary Fig. 2a, b). Although these data demonstrate donor-cell-
dependent dystrophin expression in dystrophic dogs and thus the
feasibility of this therapeutic approach, the results were modest
overall. To improve the efficacy of the treatment, the number of
injections was increased to five (always at 1-month intervals), a
new lentivector was produced in which muscle-specific creatine
kinase was replaced by the stronger myosin light chain 1F pro-
moter and GFP was deleted because we could not detect GFP in
unfixed cryostat sections (Supplementary Fig. 1b).
Efficacy of heterologous wild-type mesoangioblasts
Three dogs, namely Valgus, Varus and Viko (03H, 04H and 05H,
respectively), were treated with donor cells. One of these dogs
(Valgus, 03H) received five arterial systemic injections (5 3 10
7
cells
each) through a catheter that was introduced in the left femoralis
artery and reached the aortic arch at the level of the left subclavia: cells
were released mainly in the two large arteries. From a clinical point of
view, Valgus (03H) had optimal performance and was still walking
well 5 months after the last injection and the termination of immune
suppression, at the age of 13 months (Supplementary Movie 2).
Valgus (03H) was treated with cyclosporine, whereas the other two
dogs, Varus (04H) and Viko (05H), were treated with rapamycin and
with rapamycin and interleukin (IL)-10, respectively. Different pro-
tocols of immune suppression were tested to evaluate efficacy versus
toxicity in this model, but the results did not show significant differ-
ences between cyclosporine and rapamycin. In fact, Varus (04H;
Supplementary Movie 3) and Viko (05H) also had good clinical
performance but after 2 months of treatment Viko (05H) died sud-
denly of a fulminans myocarditis whose cause remained unexplained.
Varus (04H), in contrast, progressively lost walking ability after the
end of immune suppression. Unexpectedly, however, this animal had
no detectable anti-dystrophin antibodies and his circulating lympho-
cytes did not react to donor mesoangioblasts or to protein extracts
from the transplanted muscle (Supplementary Fig. 1c, d).
At the end of the treatment, biopsies were taken from several
muscles of the injected and contralateral leg of these dogs, treated
with heterologous wild-type cells. Histological analysis of biopsies
from Valgus (03H) revealed generally well-preserved morphology
(Fig. 2c), although areas of degeneration and regeneration were
detected infrequently. Supplementary Fig. 3 shows a large area of
the tibialis cranialis of Valgus (03H) (Supplementary Fig. 3a, a9),
better preserved than a corresponding area from Vampire (08A), a
dog transplanted with autologous cells (Supplementary Fig. 3b, b9),
cc′
bb′
a′a
d′d
e′e
Figure 3
|
Immunofluorescence analysis of tissue from treated dogs.
Double immunofluorescence analysis of muscle biopsies from the tibialis
cranialis of Ucal (01A;
a, a9), Valgus (03H; b, b9), Varus (04H; c, c9), Vampire
(08A;
d, d9) and Azor (10H; e, e9), stained with anti-laminin antibody (green
in
a9–e9 and inset in d9) and anti-dystrophin (red in a–d) or anti-b-
sarcoglycan (red in inset in
d). Nuclear staining with DAPI appears in blue.
Note clusters of dystrophin-expressing fibres in Ucal, 01A (
a) and extensive
reconstitution of dystrophin-expressing fibres in Valgus (03H;
b), Varus
(04H;
c) and Azor (10H; e). Several fibres expressing microdystrophin in
Vampire, 08H (
d) also express b-sarcoglycan in the same fibres stained on
serial, non-adjacent sections (inset in
d). Scale bar, 100 mm.
d
b
c
a
Figure 2
|
Morphology of muscle in treated dogs. Azan Mallory staining of
several muscle biopsies from the tibialis cranialis of dogs transplanted with
heterologous mesoangioblasts or autologous, genetically corrected,
mesoangioblasts.
a–c, Examples show variability from severely affected
tissue of Ucal (01A;
a) and Vrillie (02H; b), with many infiltrates, collagen
and fat deposition, to the almost normal appearance of tissue from Valgus
(03H;
c), showing only thickened interstitial tissue. d, A biopsy from Vaccin
(06A) shows an intermediate situation with well-preserved morphology.
Bar 5 100 mm.
ARTICLES NATURE
|
Vol 444
|
30 November 2006
576
Nature
Publishing
Group
©2006
described in detail below, which is itself better preserved than the
biopsy from Akan (11U), an untreated dog (Supplementary Fig. 3c,
c9). Both forelimb muscles (not shown) and the diaphragm
(Supplementary Fig. 5a) of Valgus (03H) also showed a relatively
preserved morphology. Similarly, dystrophin was expressed in most
of the examined biopsies from several different muscles of Valgus
(03H9), such as tibialis cranialis (Fig. 3b, b9), gastrocnemius and
biceps femoralis (not shown). Large areas, containing several hun-
dred dystrophin-expressing fibres (up to 70% of total fibres), were
frequently detected: an example of the sartorius of Valgus (03H) is
shown in Supplementary Fig. 5. Several clusters of dystrophin pos-
itive fibres could be detected even in the diaphragm (Supplementary
Fig. 4e).
The biopsies of Varus (04H) also showed well-preserved morpho-
logy although signs of muscle necrosis and regeneration were still
present (Fig. 2d). Overall, most of the biopsies in all muscles were
morphologically less affected than those of untreated dogs.
Dystrophin expression in these biopsies was variable but large areas
of dystrophin-positive fibres could be detected, as shown in Fig. 3c,
c9. In addition, the diaphragm of Varus was partly preserved
(Supplementary Fig. 4b) and contained clusters of dystrophin-pos-
itive fibres (Supplementary Fig. 4f). Figure 4 shows the percentage of
dystrophin-expressing fibres, which ranged from 10% to at least 70%
in two distant sections of three different biopsies each of selected
muscles from Valgus (03H) and Varus (04H). Western blot analyses
of extracts from different biopsies of the same muscles confirmed the
presence of significant amounts of dystrophin, varying from an
undetectable signal to about 60% of a control wild-type canine mus-
cle (Fig. 4c). When dystrophin was clearly detected, b-sarcoglycan
was also detected, indicating reconstitution of the dystrophin-asso-
ciated complex.
Efficacy of autologous mesoangioblasts
Three dogs, Vaccin, Valium and Vampire (06A, 07A and 08A,
respectively), received their own mesoangioblasts, transduced with
the new lentiviral vector expressing human microdystrophin. Like
Ucal (01A) in the first experiment, they received the first injection at
about 4 months of age (P113). Valium (07A) and Vampire (08A)
died of pneumonia after the third and fifth injection, respectively.
Because of early death, no force measurement was performed but a
movie showing reduced walking ability in Vampire (08A) was taken
few days before his death (Supplementary Movie 4) and biopsies were
taken at autopsy. Valium (07A) received systemic arterial delivery,
like Valgus (03H), and maintained a certain walking ability, albeit
very stiff. The biopsies of the three dogs revealed partly preserved
morphology; for example, tibialis cranialis from Vaccin (06A) is
shown in Fig. 2d. In addition, microdystrophin expression in the
tibialis cranialis of Vampire (08A) was significantly widespread
(Fig. 3d, d9), despite the poor clinical performance. The micro-
dystrophin was able to recruit at least some member of the dystro-
phin-associated complex such as b-sarcoglycan (shown in a serial,
non-adjacent section in insets to Fig. 3d, d9 and by western blot
analysis in Fig. 4b). A quantitative analysis of microdystrophin for
Vaccin (06A) and Vampire (08A) is reported in Fig. 4. Thus, all three
dogs treated with autologous, genetically corrected cells performed
poorly, even though two of them showed amelioration of morpho-
logy and microdystrophin-expressing fibres, ranging from 5% to
50%.
Efficacy of late transplantation of donor mesoangioblasts
To verify whether the less effective results obtained with autologous
cells were due to the later onset of the treatment, two dogs, Azur
(09H) and Azor (10H), received the first of five injections at P159 (5
months), 1 month later than those animals treated with their own
genetically modified cells. Although both dogs were already severely
affected at the onset of treatment, Azor (10H) showed a striking
improvement of motility (while still limping) and was even able to
run at the end of treatment (Supplementary Movie 5). Azur (09H)
showed a less evident but clearly detectable amelioration (Supplem-
entary Movie 6). In particular, although there was variation between
dogs in the progression of the disease, spontaneous improvement was
never observed. Biopsies taken at the end of treatment also showed
relatively well-preserved morphology (although with sclerosis and
infiltrates) and widespread dystrophin expression in t he tibialis cra-
nialis (Fig. 4e, e9), biceps femoralis, gastrocnemius, sartorius and even
diaphragm (Supplementary Fig. 6 and data not shown). Thus, we
conclude that even with a later onset of treatment, donor cells seem
to produce a greater amelioration of muscular dystrophy than is pro-
duced by autologous microdystrophin-expressing cells. After the end
of treatment with cyclosporine, Azor (10H) continued to walk actively
until the end of the experiment, whereas Azur (09H) rapidly lost
walking ability, much as had previously been observed with Valgus
(03H) and Varus (04H).
Vampire
Vaccin
Valgus
ui
Sar
ui
TC
ui
Gas
ui
Sar
ui
TC
ui
Gas
ui
Sar
ui
TC
ui
Gas
ui
BF
ui
Sar
ui
TC
ui
Gas
ui
BF
100
80
60
40
20
0
100
80
60
40
20
0
Dys
MyHC
Dys
Dys
β-SG
MyHC
MyHC
µDys
µDys
β-SG
MyHC
MyHC
Vaccin Vampire
Varus Valgus
Valgus
Varus
WT
Akan
100
a
b
c
80
60
40
20
0
100
80
60
40
20
0
Varus
Dystrophin-positive fibres (%)
Vampire
Vaccin
Valgus
Varus
Figure 4
|
Quantitative analysis of dystrophin content in tissue from
treated dogs. a
, Three separate sections for two different biopsies of the
injected (i, blue symbols) and non-injected (u, red symbols) sartorius (Sar),
tibialis cranialis (TC) and gastrocnemius (Gas) for two dogs injected with
donor cells (Valgus (03H) and Varus (04H)) and two dogs injected with
autologous, genetically corrected, cells (Vaccin (06A) and Vampire (08A))
were analysed for dystrophin expression. A total of 200 laminin-positive
fibres (of any size) were counted in randomly selected fields, and the
percentage of these fibres also expressing dystrophin was calculated.
b, Western blot analysis of different biopsies from the muscles shown in
a plus the biceps femoralis (BF) showing the expression of dystrophin (Dys,
Valgus (03H) and Varus (04H)) or microdystrophin (mDys, Vaccin (06A)
and Vampire (08A)). Each lane was loaded with 80 mg of proteins. Myosin
heavy chains (MyHC) are also shown as an internal standard in the bottom
panel. For Varus (03H) and Vampire (08A) the expression of b-sarcoglycan
(b-SG) is also shown in the middle panel. There is a good correlation
between the expression of dystrophin or microdystrophin and
b-sarcoglycan.
c, Expression of dystrophin and myosin heavy chains in the
injected biceps femoralis of GMRD untreated (Akan (U11)), wild type,
Varus (04H) and Valgus (03H); only 30 mg of total proteins was loaded in
each lane for direct comparison.
NATURE
|
Vol 444
|
30 November 2006 ARTICLES
577
Nature
Publishing
Group
©2006
Analysis of muscle-released serum enzymes
Throughout treatment, serum collected from treated dogs was used
to detect the levels of creatine kinase and other enzymes released by
damaged muscle fibres (Supplementary Fig. 7); in general, the inject-
ion of mesoangioblasts corresponded to rapid and often profound
decrease in enzyme activity. This may be due to improved survival of
fibres reconstituted with transplanted cells but also to factors such as
insulin-like growth factor 1, basic fibroblast growth factor and others
released by donor cells
15
and contributing to fibre survival. At the end
of treatment, the level of creatine kinase in all treated dogs was well
below the those reported for this form of muscular dystrophy
9
. Other
enzymes released by damaged muscle fibres, such as aspartate amino-
transferase, showed an overall similar pattern (data not shown).
Enhancement of contraction force
To test whether morphological and biochemical changes would
correspond to increased force of contraction we performed two
different types of analysis: tetanic force of the tibialis cranialis and
extensor digitorum longus muscle for Valgus (03H), Varus (04H),
Vaccin (06A), Azur (09H) and Azor (10H), and force of contraction
on isolated single fibres from Varus (03H), Valgus (04H), Vaccin
(06A) and Valium (07A). Figure 5a shows the tetanic force of a
normal dog (black squares), which increases progressively with age.
Dystrophic dogs have variable tetanic force (ranging from 90% to
40% of wild type) but, in the absence of treatment (red squares),
progressively lose it. In the experiment reported, Vaccin (06A),
who received autologous cells (green triangles), showed a strong
tetanic force at the onset of treatment and maintained it up to the
fourth injection but later showed a significant decrease. In contrast,
the four dogs receiving donor cells (blue squares and triangles) started
from a decreased tetanic force but maintained it though the treatment
and eventually showed a modest increase in the last measure-
ment, at the end of treatment. To avoid the problems deriving
from this variability, force was also measured and reported as the
percentage increase in treated over untreated legs. Figure 5b
shows that all dogs receiving donor cells showed an increase in
tetanic force ranging from 50% to 100% in the treated leg,
whereas the dog receiving his own cells (Vaccin, 06A) did not
show any increase. To measure force of contraction at the single-
fibre level, a large population of single skinned fibres was dis-
sected from the tibialis cranialis, gastrocnemius and sartorius
muscles of Valgus (03H) and Varus (04H) and of two control
wild-type dogs. After determination of specific force, each fibre
was also analysed for the expression of dystrophin and of fast or
slow myosin heavy chains (MyHC). The results reported in Fig. 5c
show that fast fibres from dystrophic dogs had a complete recov-
ery in force when expressing dystrophin, up to the level of fibres
from wild-type dogs. Fibres showing partial expression of dystro-
phin also showed partial force recovery. A picture showing fully
or partly reconstituted as well as dystrophin-negative fibres is
shown in Fig. 5d. In contrast, dystrophic slow fibres did not
show reduced force in comparison with wild-type fibres, as
shown previously
16
(data not shown). A similar measurement
on the single skinned fibres of the tibialis cranialis and sartorius
of Vaccin and Valium (06A and (07A) showed a similar trend,
but the increase in force of contraction was modest and not
statistically significant (data not shown).
Immune reaction against dystrophin and donor cells
To test the possible occurrence of an immune reaction in the trans-
planted dogs against donor cells and/or the transgene (dog dystro-
phin and human microdystrophin, respectively), we performed an
immunocytochemical analysis of cellular infiltrates, a western blot
analysis to test the reactivity of dog sera and a lymphocyte prolifera-
tion assay to test the appearance of cellular immunity against donor
cells and/or the transplanted muscle tissue
17–19
. Results indicate a low
frequency of infiltrates, an absence of serum antibodies and modest
activation only for local lymphocytes exposed to transplanted muscle
tissue (Supplementary Information).
Conclusions and future perspectives
Gene or cell therapy approaches for GRMD have until now produced
negative
20
or modestly positive results
21–23
. We show here that it is
possible to transplant mesoangioblasts into dystrophic dogs and
obtain an extensive reconstitution of fibres expressing dystrophin,
an improvement in the contraction force and, in many cases, a pre-
servation of walking ability. Previous work in the mouse
4,13
showed
that mesoangioblasts express some of the protein that leukocytes use
to cross the vessel wall and so they invade the interstitial tissue,
eventually to fuse with and contribute sarcoglycan to regenerating
fibres. Donor wild-type mesoangioblasts seemed to be more efficient
than autologous, genetically corrected cells. Possibly microdystro-
phin produces a modest functional rescue when delivered late
through donor cells in contrast with the excellent functional rescue
induced when delivered as a transgene
11
. A different onset of treat-
ment should not be crucial because two late-transplanted dogs (Azor
0.16
ac
bd
0.12
0.08
0.04
0
250
200
150
100
50
0
511
576810911
+ +/–
GRMD
–WT
Force kg
–1
(%)
0
30
60
90
Specific force (kN m
–2
)
Force (%)
+ +/– –
79
Months
Months 42.12 31.9154.61
–
+
+/–
Figure 5
|
Physiology. Functional properties of skeletal muscles in vivo and
of individual muscle fibres in vitro of treated dystrophic dogs after five
consecutive injections of donor wild-type mesoangioblasts.
a, Tetanic force
of the flexor muscles of the cranial tibial compartment of different dogs
measured after maximal stimulation of the common peroneal nerve and
recorded with an ergometer developed in house. Black squares, normal wild-
type dog; red squares, dystrophic untreated dog (Akan (11U)); green
triangles: dystrophic dog transplanted with autologous genetically corrected
mesoangioblasts (Vaccin (06A)); blue symbols, dystrophic dogs
transplanted with donor wild-type mesoangioblasts (squares, Valgus (03H)
and Varus (04H); triangles, Azur (09H) and Azor (10H)). Tetanic force was
normalized to each dog’s body weight to obtain weight-corrected specific
force.
b, Relative increase in force of the flexor muscles of the cranial tibial
compartment of the injected leg versus the untreated leg. Blue squares,
Varus (04H); green squares, Vaccin (06A); filled blue diamonds, Azur (09H);
open blue diamonds, Azor (10H).
c, Specific force (maximum isometric
force/cross-sectional area) of a population of 199 single muscle fibres
dissected from tibialis cranialis, gastrocnemius, and sartorius muscles of
Valgus (03H) and Varus (04H), indicated together as GRMD, and of 148
fibres dissected from the same muscles of two control golden retriever dogs
(WT). All fibres shown are fast fibres (type 2A, 2AX or 2X) (n 5 71 for
GRMD; n 5 106 for WT). Type 2A, 2AX and 2X fibres were pooled together
because no difference in specific force was observed between these fast fibre
types in either GRMD or WT dogs. Fibres from Valgus (03H) and Varus
(04H) were grouped into dystrophin-positive (1), partly positive (1/2) and
negative (2) on the basis of the presence of dystrophin detected, after force
determination, by immunostaining by anti-dystrophin antibody.
d, Examples of positive (1), partly positive (1/2) and negative (2) single
skinned fibres immunostained for dystrophin after force determination. The
specific force (maximum isometric force/cross sectional area) developed by
each fibre is shown under each panel. Scale bar, 100 mm. Error bars, 1 s.d.
ARTICLES NATURE
|
Vol 444
|
30 November 2006
578
Nature
Publishing
Group
©2006
