J Neuropathol Exp Neurol
Copyright Ó 2008 by the American Association of Neuropathologists, Inc.
Vol. 67, No. 7
July 2008
pp. 711Y719
ORIGINAL ARTICLE
Endothelial and Myogenic Differentiation of Hematopoietic
Progenitor Cells in Inflammatory Myopathies
David Hollemann, MD, Herbert Budka, MD, Wolfgang N. Löscher, MD, PhD,
Genya Yanagida, MD, Michael B. Fischer, MD, and Julia V. Wanschitz, MD
Abstract
INTRODUCTION
From the Department of Blood Group Serology and Transfusion Medicine
(DH, GY, MBF), the Institute of Neurology, Medical University Vienna
(HB); and the Clinical Department of Neurology, Innsbruck Medical
University (WNL, JVW), Innsbruck, Austria.
Send correspondence and reprint requests to: Julia V. Wanschitz, MD,
Clinical Department of Neurology, Innsbruck Medical University,
Anichstrasse 35, A-6020 Innsbruck, Austria; E-mail: julia.wanschitz@
i-med.ac.at
Drs. Fischer and Wanschitz contributed equally to this work.
This study was supported by Project Grant No. P13481 from the Austrian
Science Fund (FWF).
Autoimmune inflammatory myopathies (IMs) are the
largest group of acquired and potentially treatable muscle
diseases. The 3 major subtypes of IM (i.e. dermatomyositis
[DM], polymyositis [PM], and inclusion body myositis [IBM])
have distinct clinical, histopathologic, and immunologic features (1). In DM, a complement-mediated microangiopathy
leads to endothelial cell (EC) hyperplasia and obliteration of
intramuscular blood vessels that result in reduction of capillary density, microinfarcts, and perifascicular atrophy (2).
Myocytotoxicity mediated by CD8+ T cells surrounding and
invading major histocompatibility class I antigen-expressing
myofibers is suggested as one of the pathogenic mechanisms
of PM and IBM (3). In IBM, the most frequent IM in elderly
patients, chronic inflammation occurs in conjunction with a
degenerative process that is related to protein misfolding and
deposition of A-amyloid and is associated with a poor response to current immunotherapies and progressive disability
of most patients (4).
Maintenance of adult skeletal muscle and regeneration
upon damage are accomplished by satellite cells, a population
of tissue-resident committed myogenic progenitors located in
anatomic cell niches beneath the basal lamina of mature
myofibers (5, 6). In response to muscle injury, satellite cells
enter a well-defined sequence of activation, proliferation,
myogenic differentiation, and fusion events to repair or replace damaged muscle fibers (7). The entry of quiescent
satellite cells into the myogenic differentiation program is
critically regulated by the paired box transcription factors
Pax3/Pax7 (8Y10) and myogenic regulatory factors, including
Myf5, MyoD, myogenin, and MRF4 (11, 12). Satellite celldependent repair of muscle fibers is also crucial for recovery
of muscle strength in IM after suppression of inflammation.
In IBM, however, repair mechanisms are insufficient to compensate for the persistent degenerative process (13).
In addition to muscle-committed satellite cells, bone
marrow-derived multilineage stem cells have been shown to
exhibit myogenic potential in experimental conditions; this
suggests that these cells are a possible alternative source for
muscle regeneration (14). After transplantation, donor cells
derived from unfractionated bone marrow (15, 16) or purified
hematopoietic and myelomonocytic progenitor cells were
found to be capable, albeit at a low frequency, of integrating
into regenerating myofibers in transplant recipient mice
J Neuropathol Exp Neurol Volume 67, Number 7, July 2008
711
Incorporation of circulating hematopoietic progenitor cells (HPCs)
into damaged skeletal muscle has been proposed as a novel
mechanism of tissue repair complementary to satellite cellYdependent
regeneration. We studied the occurrence and myoendothelial differentiation of HPCs in muscle of patients with inflammatory myopathies. Muscle biopsies from untreated patients with dermatomyositis,
polymyositis, inclusion body myositis, and controls were investigated
for the expression of endothelial (CD31, von Willebrand factor,
vascular endothelial growth factor receptor 2), hematopoietic (CD34,
CD133, CD45), and myogenic (Pax7, MyoD) markers by immunohistochemistry and reverse-transcriptaseYpolymerase chain reaction.
Confocal laser scanning microscopy was used to visualize coexpression of CD34, CD133, von Willebrand factor, or Pax7 on
individual cells. Morphometric analysis revealed significantly
increased numbers of CD133+ cells per square millimeter in
polymyositis and inclusion body myositis compared with controls
(p G 0.001); this correlated with the density of CD45+ infiltrates (p G
0.001). By confocal laser scanning microscopy, we detected several
mononuclear cells that coexpressed either CD34/von Willebrand
factor or CD133/Pax7 with or without CD34 reactivity, indicating
endothelial or myogenic commitment of some HPCs in skeletal
muscle. Rarely, CD133+/CD34T/Pax7+ cells seemed to occupy
satellite cell niches or to incorporate into preexisting myofibers.
Our findings suggest that circulating HPCs colonize skeletal muscle
in inflammatory conditions and provide evidence for in situ
myoendothelial differentiation of some of these cells.
Key Words: Hematopoietic progenitors, Myositis, Neovascularization, Pax7, Satellite cells.
Copyright @ 2008 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
Hollemann et al
J Neuropathol Exp Neurol Volume 67, Number 7, July 2008
(17, 18). Survival and incorporation of engrafted cells into
host muscle requires a supportive microenvironment provided
by prosurvival signals from neighboring cells and sufficient
vascularization to ensure oxygen and nutrient supply. Conversely, differentiating myogenic cells are proangiogenic and
reciprocally interact with ECs to support synchronized
vasculomyogenesis (19), which is mandatory for a functionally relevant reconstitution of damaged muscle (20).
A subpopulation of primitive hematopoietic progenitor
cells (HPCs) that are defined by the surface molecules CD34
and CD133 (21Y23) and have a high potential to differentiate
along myoendothelial lineages in vitro has been found in
adults (24, 25). These HPCs contribute in vivo to regenerating muscle fibers (25, 26) and neovascularization (27).
They predominantly reside in bone marrow but circulate in
the blood stream at low numbers under physiologic conditions (21). During maturation induced by several growth
factors that include vascular endothelial growth factor
(VEGF)(28), these cells gradually lose CD133 as they start
to differentiate into small-vessel ECs (22), but they still
express CD34 (29). In human peripheral blood, approximately 92% of CD133+/CD34+ cells coexpress the panhematopoietic marker CD45 (25). Numerous cytokines and
mediators of inflammation increase the numbers of circulating progenitors, which constitutively migrate through the
blood and colonize peripheral organs for renewal of tissuespecific stem cells (30). Additionally, the upregulation of
chemo-attractant receptors (e.g. L-selectin, vascular cell
adhesion molecule 1) on small-vessel endothelium of
inflamed muscle stimulates migration of circulating progenitors through the blood vessel wall (31).
The objective of this study was to investigate whether
circulating HPCs enter skeletal muscle during immunemediated inflammation and to determine whether there is
evidence for endothelial and myogenic differentiation of these
cells in human skeletal muscle in vivo.
for Blood Group Serology and Transfusion Medicine, the
Clinical Institute of Neurology, Medical University Vienna,
and the Clinical Department of Neurology, Innsbruck Medical
University (no. 563/2007). Archival muscle tissue specimens
were retrospectively collected in Vienna and Innsbruck
after written informed consent of patients to participate in
the study.
MATERIALS AND METHODS
Patients
Skeletal muscle was obtained by open biopsy from
untreated patients with DM (n = 5; mean age, 56.8 T
21.8 years), PM (n = 5; mean age, 59.0 T 13.6 years), and
IBM (n = 6; mean age, 70 T 14.0 years) for diagnostic purposes
after informed consent. Classification of the myositis was
based on the distribution and immunophenotype of inflammatory infiltrates, the pattern of major histocompatibility
type I antigen expression, and the deposition of complement
C5b-9 complex (1). The diagnosis of IBM was confirmed by
the presence of 16- to 18-nm tubulofilamentous inclusions by
electron microscopic examination. Muscle biopsies from
another 5 individuals (mean age, 39.2 T 18.0 years) in whom
the diagnostic workup excluded the clinical suspicion of
malignant hyperthermia or metabolic myopathy were used
as controls. The tissue was immediately snap-frozen in
liquid nitrogen and stored at j80-C until analysis. The
study was approved by the Ethics Commission of the Medical
University Vienna as a cooperation between the Department
712
Immunohistochemistry
Frozen sections of deltoid, biceps brachii, or quadriceps
muscles were studied by standard histologic procedures, and
the pattern of inflammation was analyzed by immunohistochemistry for CD45 (1.2 Kg/ml; Dako, Glostrup, Denmark),
CD8 (0.25 Kg/ml; Dako), CD20 (1.4 Kg/ml; Dako), CD68
(0.4 Kg/ml; Dako), HLA-I (0.35 Kg/ml; Dako), and membrane attack complex (0.4 Kg/ml; Dako). Primary antibodies
were detected with a Chemmate detection kit (Dako) and
visualized with 3,3¶-diaminobenzidine as chromogen. Furthermore, immunohistochemistry for endothelial, hematopoietic, and myogenic markers was performed as previously
reported (32) with minor modifications. Acetone-fixed sections were blocked with avidin/biotin blocking reagent
(Vector Laboratories, Burlingame, CA) and incubated overnight at 4-C with rabbit anti-von Willebrand factor (vWF; 5.7
Kg/ml; Dako), mouse anti-CD31 (4.5 Kg/ml; Dako), mouse
anti-CD34 (0.5 Kg/ml; Beckman Coulter, Immunotech,
Marseille, France), mouse anti-CD133 (5 Kg/ml, detecting
AC133-1 and AC133-2; R&D Systems, Minneapolis, MN),
mouse anti-Pax7 (7.1 Kg/ml; R&D Systems), rabbit antiMyoD (0.2 Kg/ml; Santa Cruz Biotechnology, Santa Cruz,
CA), or biotinylated goat antiYVEGF receptor 2 (VEGFR-2;
2 Kg/ml; R&D Systems) diluted in Tris-buffered saline/1%
bovine serum albumin. The reactivity of the primary unlabeled antibodies was revealed using biotinylated goat F(ab¶)2
anti-mouse immunoglobulin (Ig)G + IgM (H + L; 4.2 Kg/ml;
Jackson ImmunoResearch, West Grove, PA) or biotinylated
donkey anti-rabbit IgG (H + L; 1.2 Kg/ml; Jackson
ImmunoResearch) diluted in 500 Kg/ml normal human
Ig (Biotest, Dreieich, Germany) followed by streptavidinhorseradish peroxidase complex (20 Kg/ml; Sigma, St. Louis,
MO). For the detection of CD133 and VEGFR-2, a supersensitive streptavidinYhorseradish peroxidase conjugate (Biocare Medical, Walnut Creek, CA) was used. Endogenous
peroxidase was blocked during the incubation period with
secondary antibody by adding D -glucose (50 mg/ml)
and glucose oxidase Type VII (dilution of normal human
IgG, D-glucose, and glucose oxidase Type VII was used at a
rate of 7:2:1). The primary antibody was omitted for control
purposes, and isotype controls were included in the protocol;
for Pax7, an IgG1 isotype control was used. To visualize
antibody staining, the sections were exposed to 3-amino-9ethyl-carbazole (Sigma) and subsequently counterstained
with Mayer hematoxylin.
Morphometric Analysis
To quantify CD31+ and vWF+ ECs, CD34+ and
CD133+ HPCs, CD45+ leukocytes, and Pax7+ cells, 16 representative areas per section within the endomysium, including inflammatory foci, were selected. Numbers of cells were
Ó 2008 American Association of Neuropathologists, Inc.
Copyright @ 2008 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 67, Number 7, July 2008
Hematopoietic Progenitors in Myositis
counted with a 40 objective using an ocular morphometric
grid covering a total area of 1 mm2 (33).
thermocycler (Hybaid; PCR express, Ashford, Middlesex,
UK). Complementary DNA template (3 Kl; 1/20 of reversetranscribed RNA) and 2 Kl of 10 Kmol/L primer mix (CD31,
CD34, MyoD1, VEGFR-2, hypoxanthine guanine phosphoribosyltransferase [HPRT]) were applied for the amplification
protocol. For reverse-transcriptase-PCR with primer pair
CD133, one fourth cDNA of reverse-transcribed RNA was
used. Intron-spanning oligonucleotide primer pairs specific for
CD31 (platelet EC adhesion molecule 1; sense, 5¶-TCC AGT
GTC CCC AGA AGC-3¶; antisense, 5¶-CAA GGG AGC CTT
CCG TTC-3¶), CD34 (sense, 5¶-CTT GCT GAG TTT GCT
GCC TTC-3¶; antisense, 5¶-ACA TTT CCA GGT GAC AGG
Confocal Laser Scanning Microscopy
Frozen sections 4-Km thick were fixed in acetone,
blocked with avidin/biotin blocking reagent and 12.5% donkey serum, incubated overnight at 4-C with mouse anti-Pax7
(2.5 Kg/ml; R&D Systems) or rabbit anti-vWF (5.7 mg/ml)
diluted in Tris-buffered saline/1% bovine serum albumin, and
subsequently incubated with 500 Kg/ml normal human Ig and
Alexa Fluor (AF) 555 donkey anti-mouse IgG (H + L; for
mouse anti-Pax7; all secondary AF antibodies 4 Kg/ml;
Invitrogen Life Technologies, Carlsbad, CA) or AF 555
donkey anti-rabbit IgG (H+L; for rabbit anti-vWF). After
blocking with 20% mouse serum (for anti-Pax7), tissue
samples were then incubated with fluorescein isothiocyanate
(FITC)-conjugated anti-CD34 (2.5 Kg/ml; Becton Dickinson
PharMingen, San Jose, CA) or FITC-conjugated anti-CD45
(1.25 Kg/ml; Becton Dickinson PharMingen) diluted in Trisbuffered saline/1% bovine serum albumin/5% mouse serum.
An incubation with polyclonal goat anti-CD133 (2 Kg/ml;
Santa Cruz Biotechnology) overnight was followed by a step
using secondary antibody AF 647 donkey anti-goat IgG (H +
L; 4 Kg/ml; Invitrogen Life Technologies), which finished
the staining procedure. Serial dilutions of each primary and
secondary antibody were performed to minimize nonspecific
binding, assure separation of the fluorescent signals, and optimize fluorophore concentration to preclude self-quenching. A
purified mouse IgG1, a mouse IgG1 FITC, and a purified
mouse IgG2-biotin (all in a concentration of 2.5 Kg/ml;
Becton Dickinson PharMingen) were included as isotypespecific controls instead of primary antibodies.
Stained samples were covered with 4¶,6-diamidino-2phenylindole (DAPI) diluted 1:4 with Vectashield mounting
medium (Vector Laboratories). Sections were analyzed with
a confocal laser scanning microscope (Zeiss LSM 510;
Oberkochen, Germany) with a multiphoton laser (argon laser,
488 nm for FITC [green]; Helium Neon 1, 543 nm for AF
555 [red]; Helium Neon 2, 633 nm for AF 647 [blue]) and a
63 Zeiss Plan-Apochromat differential interference contrast oil immersion objective with a numeric aperture of 1.40
using a multitracking scan (32).
Reverse-Transcriptase-Polymerase Chain
Reaction
Total RNA from snap-frozen muscle tissue samples of
patients with PM, DM and IBM, and controls was isolated using
TRIzol reagent (Invitrogen Life Technologies) according to the
manufacturer’s protocol. Contaminating genomic DNA was
removed by DNase I treatment using the DNA-free kit
(Ambion, Austin, TX) (32). RNA was quantified at 260 nm,
and equal amounts were reverse transcribed into complementary DNA (cDNA) using the SuperScript preamplification
system kit for first-strand cDNA synthesis (Invitrogen Life
Technologies) (32). For reverse-transcriptaseYpolymerase
chain reaction (PCR), the PCR reagent system kit and Taq
DNA polymerase (Invitrogen Life Technologies) were used.
Amplification of cDNA was performed using a gradient
Ó 2008 American Association of Neuropathologists, Inc.
FIGURE 1. (A) Area with reduced microvascular density and
remnants of destroyed endothelial cells (ECs; arrow) in
dermatomyositis (DM). (B) Inflammatory infiltrate in polymyositis (PM) with proliferating capillaries that are strongly
reactive for CD31. Round and spindle-shaped mononuclear
cells stain for CD31 less intensely than mature ECs. (C) Regular
capillary network in control muscle (AYC; anti-CD31). (D)
CD34-reactive capillaries in DM. (E) Irregular capillary architecture indicating vascular remodeling (arrow) and numerous
interstitial cells (arrowhead) show weak CD34 expression in
inclusion body myositis [IBM]. (F) Control (DYF; anti-CD34).
(G) Focal accumulation of CD133+ hematopoietic progenitor
cells (HPCs) colocalizes with inflammatory infiltrates in PM. (H)
CD133+ HPCs surround a viable myofiber (asterisk) in IBM. (I)
Single interstitial CD133+ HPCs (arrowheads) in a control
muscle (GYI; anti-CD133). (J, K) Increased numbers of Pax7+
cells in PM and IBM, respectively, compared with control (L).
Most Pax7+ cells are located in sublaminal niches (arrowheads), but, occasionally, they are also detected in the
interstitial compartment (J, K, arrows; JYL, anti-Pax7). Original
magnification: 400 (all images).
713
Copyright @ 2008 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 67, Number 7, July 2008
Hollemann et al
TABLE. Semiquantitative Analysis of Immunohistochemical Staining
vWF
DM
PM
IBM
C
145.8
214.4
242.3
148.8
T
T
T
T
84
59.7
76.2
70.2
CD31
308.4
382.6
363.3
277.2
T
T
T
T
62.6
197
88.6
54.5
CD34
221.4
254.4
326.3
229.8
T
T
T
T
CD133
59.5
76.8
99.4
55.8
37.2
75
98.1
0.4
T
T
T
T
21.9
24.6
32.7
0.5
CD45
130.8
208.6
373
5.2
T
T
T
T
53.3
112
72.9
7.2
Pax7
66.4
104.4
105.7
21
T
T
T
T
15.6
91.7
46.8
6.4
Table shows mean numbers T SD of vWF+, CD31+, CD34+, CD133+, CD45+, and Pax7+ cells per square millimeter in DM, PM, IBM, and control muscle.
C, control; DM, dermatomyositis; IBM, inclusion body myositis; PM, polymyositis; SD, standard deviation; vWF, von Willebrand factor.
CTA-3¶), CD133 (sense, 5¶-TAC CAA GGA CAA GGC GTT
CAC-3¶; antisense, 5¶-CAG TCG TGG TTT GGC GTT GTA3¶), MyoD1 (sense, 5¶-CGG CGG AAC TGC TAC GAA-3¶;
antisense, 5¶-GAT GCG CTC CAC GAT GCT-3¶), and
VEGFR-2 (sense, 5¶-CAT GTG GTC TCT CTG GTT GTG3¶; antisense, 5¶-TCC CTG GAA GTC CTC CAC ACT-3¶).
The low-expressing housekeeping gene primer HPRT (sense,
5¶-TGA AAA GGA CCC CAC GAA-3¶; antisense 5¶-ACA
ACA ATC CGC CCA AAG G-3¶) was used to check the
integrity of the RNA (all primers were synthesized by MWGBiotech, Ebersberg, Germany). The amplification program
included 35 cycles of denaturation for 20 seconds at 94-C,
primer annealing for 30 seconds at 57-C for CD31, 59-C for
CD34 and CD133, 58-C for MyoD1, 60-C for VEGFR-2, and
55-C for HPRT, primer extension for 45 seconds at 72-C,
with a final extension step for 10 minutes. Ten-microliter
aliquots of the PCR generated products were resolved by
electrophoresis on 1.5% agarose gels containing ethidium
bromide (10 mg/ml; Bio-Rad, Hercules, CA) in Tris acetate
EDTA buffer, exposed to ultraviolet light in an EpiChemi II
darkroom, and photographed using DigiDoc-It System software (UVP, Upland, CA) and the GelDoc 2000 Documentation System (Bio-Rad). The predicted sizes of the PCR
products were 462 bp for CD31, 364 bp for CD34, 449 bp
for CD133, 114 bp for MyoD1, 251 bp for VEGFR-2, and
390 bp for HPRT, confirmed using a 100-bp DNA ladder
(1 Kg/Kl; Invitrogen Life Technologies).
was disorganized, showing either focal loss that was most
pronounced in DM (Fig. 1A) or focal proliferation of
capillaries in areas of endomysial inflammation and tissue
damage or repair (Fig. 1B). The surface marker CD34 is
expressed on mature ECs and EC progenitors, and there was
intense staining for it in most microvascular ECs in IM and
controls (Figs. 1DYF). Additionally, numerous round or
Statistical Analysis
For statistical analysis, SPSS (release 14.0; SPSS, Inc.,
Chicago, IL) software was used. Distribution of variables was
analyzed by Kolmogorov-Smirnov test. Between-group comparisons were performed with analysis of variance and
Bonferroni post hoc test. Correlations between variables
were assessed by linear regression analysis. p G 0.05 was
considered as statistically significant. Values are expressed as
mean T standard deviation.
RESULTS
Distribution of Endothelial and Myogenic
Progenitor Cells in IM and Control Muscle
Immunohistochemistry for the EC markers CD31 and
vWF visualized the microvascular architecture within the
endomysial connective tissue and demonstrated a regular
distribution of capillaries surrounding each individual
myofiber in controls (Fig. 1C). In IM, the capillary network
714
FIGURE 2. (A) Box plots show numbers of CD34+ (open
boxes), CD45+ (shaded boxes), and CD133+ (solid boxes) cells
per square millimeter in dermatomyositis, polymyositis, inclusion body myositis, and controls. Horizontal bars, mean values;
boxes, standard errors of the means; whiskers, standard
deviations. (B) There was a highly significant correlation
between numbers of CD45+ and CD133+ cells per square
millimeter (continuous line, linear regression line; dashed lines,
95% confidence intervals).
Ó 2008 American Association of Neuropathologists, Inc.
Copyright @ 2008 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 67, Number 7, July 2008
Hematopoietic Progenitors in Myositis
spindle-shaped mononuclear cells without lumina expressed
CD34 (Fig. 1E) and CD31 (Fig. 1B) and, less frequently,
vWF. These stained cells were present in the interstitial spaces
of IM muscle and probably represent differentiating ECs. A
small subset of these cells expressed VEGFR-2 at low levels
(not shown). In IM, mononuclear cells reactive for the early
HPC marker CD133 were mainly observed as focal aggregates
(Fig. 1G) in the interstitial connective tissue and occasionally
surrounding viable muscle fibers (Fig. 1H) or sparsely
spreading into the tissue; only single interstitial CD133+ cells
FIGURE 3. (A) Arrangement of Pax7+ satellite cells (red; arrowhead) around CD34+ capillaries (green; arrow). (B, C) Pax7+ cells
(arrowheads) do not coexpress the panhematopoietic marker CD45 (arrows). (AYC) Inclusion body myositis (IBM);
magnifications: (A, B) 990; (C) 630. (D) Mononuclear CD34+ cell with elongated cell processes (arrow) and a satellite cell
(arrowhead) closely attached to a muscle fiber (asterisk; IBM 990). Inset, 4¶,6-diamidino-2-phenylindole (DAPI) stain (yellow) for
nuclei. (E) Group of mononuclear CD34+ cells, 1 of which coexpresses von Willebrand factor (vWF; arrow), indicating endothelial
cell differentiation, in a case of IBM. The arrowhead marks a mature vWF+ microvessel (1,440). Inset, nuclei of CD34+
mononuclear cells are visualized with DAPI (purple). (F) Cytoplasmic Pax7 reactivity of interstitial mononuclear cells coexpressing
either CD34 (G), indicated by the solid arrow, or CD133 antigens (H), indicated by open arrows. (I) Merged image. (FYI)
Polymyositis (1,440). Inset, 4¶,6-diamidino-2-phenylindole stain for nuclei (yellow); (AYE, I) and insets are merged images.
Isotype controls showed no specific binding (data not shown).
Ó 2008 American Association of Neuropathologists, Inc.
715
Copyright @ 2008 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
Hollemann et al
FIGURE 4. A CD133+/CD34+/Pax7+ cell seems to be incorporated into a preexisting muscle fiber, inclusion body
myositis (1,530; A, Pax7; B, CD34; C, CD133; D, merged
image). Inset, 4¶,6-diamidino-2-phenylindole stain demonstrates the nucleus (arrowhead) of the incorporated cell.
Isotype controls showed no specific binding (data not shown).
were detectable in controls (Fig. 1I). Immunoreactivity for
CD133 was restricted to cells with an immature appearance
characterized by round to oval nuclei and scant cytoplasm.
These cells were predominantly colocalized with CD45+
inflammatory infiltrates; CD133 was not observed on mature
capillary ECs. The markers of myogenic differentiation, Pax7
and MyoD1, were expressed by nuclei located in satellite cell
niches of mature muscle fibers and occasionally by interstitial
mononuclear cells (Figs. 1J, K).
The results of morphometric analysis are summarized in
the Table. Despite marked irregularities of capillary density in
IM cases, the total numbers of CD31+, CD34+, and vWF+
cells did not differ significantly between DM, PM, IBM, and
controls. Inflammatory myopathy samples contained numerous
CD31+, CD34+, and, less frequently, vWF+ mononuclear cells
within the interstitium, but expression of these markers in
controls was largely restricted to mature ECs of capillaries.
CD133+ mononuclear cells were much less numerous than
CD31+ and CD34+ cells, but they were significantly increased
in PM and IBM compared with controls (Fig. 2A; p G 0.001).
In DM, interstitial CD133+ cell numbers were lower than in
IBM but showed considerable variability among individual
cases and did not differ significantly from PM and controls.
Numbers of CD133+ cells strongly correlated with the density
of CD45+ inflammatory infiltrates (Fig. 2B; p G 0.001). The
ratio of CD133+ to CD45+ cells was significantly elevated in
inflamed (mean, 0.302 T 0.084) versus control (mean, 0.07 T
0.13) muscle (p G 0.001). Furthermore, mean numbers of
Pax7+ cells per square millimeter were increased in IM com-
716
J Neuropathol Exp Neurol Volume 67, Number 7, July 2008
pared with controls (Table; p G 0.01) but did not differ significantly between IM subtypes.
Confocal laser scanning microscopy was used to
characterize further the expression patterns of CD133, CD34,
CD45, vWF, and Pax7 in IM. Antibody against Pax7 visualized satellite cells in their niches (Figs. 3AYC); less
frequently, Pax7 was coexpressed by interstitial CD133+/
CD34j, CD133+/CD34+, or CD133j/CD34+ cells, suggesting their myogenic commitment (Figs. 3F, I). Unexpectedly,
Pax7 reactivity was localized in the cytoplasm of these cells.
Within the interstitial compartment, heterogeneous CD133+/
CD34+ or CD133j/CD34+ (Fig. 3G) or CD133+/CD34j
(Fig. 3H) HPCs with triangular shapes or elongated cytoplasmic processes were most often detected near blood
vessels but sometimes also in intimate contact with mature
myofibers (Fig. 3D). Triple staining for CD133, CD34, and
vWF delineated a subset of CD133j/CD34+ cells that delicately coexpressed vWF, indicating EC differentiation
(Fig. 3E). Occasionally, CD133+/CD34j/Pax7+ or CD34+/
CD133+/Pax7+ cells seemed to be incorporated into preexisting myofibers (Figs. 4AYD). Coexpression of CD45 by
CD133+/Pax7+ cells was not detected. No specific binding
was observed when isotype controls instead of primary antibodies were used. Autofluorescence of inflammatory cells
was not observed with isotype controls. When triple staining
for CD133/CD34/Pax7 was performed, the yellow channel
was used to visualize the original blue DAPI signal of nuclei.
Reverse-Transcriptase-Polymerase Chain
Reaction
Extraction of total RNA from muscle tissue of patients
with PM and IBM resulted in an approximately 2-fold higher
FIGURE 5. Reverse transcriptase-polymerase chain reaction
(PCR) analysis of CD31, CD34, CD133, vascular endothelial
growth factor receptor 2, and MyoD1 expression. Muscle
tissue of 2 representative patients with dermatomyositis,
polymyositis, and inclusion body myositis, and 2 controls
show bands with the predicted size of the PCR product. The
low-expressing housekeeping gene hypoxanthine guanine
phosphoribosyltransferase served as control; a 100-bp DNA
ladder size marker is shown in the left lane of each panel.
Ó 2008 American Association of Neuropathologists, Inc.
Copyright @ 2008 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 67, Number 7, July 2008
Hematopoietic Progenitors in Myositis
yield compared with samples obtained from DM and controls
(data not shown). The amount of extracted total RNA
corresponded to the higher numbers of cells per square
millimeter in muscle tissue of patients with PM and IBM
because of the leukocyte infiltration. Message for CD31,
CD34, CD133, and MyoD1 was found in all disease groups
and controls (Fig. 5), although the levels of expression varied.
With the exception of 1 PM case, message for VEGFR-2 was
detected in all cases. Of the 2 isoforms of CD133 (AC133-1
and AC133-2), only AC133-1 mRNA was found.
regenerative cycles in IBM. Corresponding to the higher
density of cellular infiltration per square millimeter in PM and
IBM, an approximately 2-fold higher yield of total RNA was
extracted from muscle tissue of those patients compared with
DM and controls. When equivalent amounts of total RNA
were reverse transcribed to cDNA that was then amplified by
PCR, disease subgroups and controls showed comparable
PCR products specific for CD133-1, an isoform of CD133.
The presence of CD133+ cells in inflammatory foci and the
close correlation between numbers of CD133+ HPCs and
CD45+ leukocytes suggest that CD133+ HPCs enter skeletal
muscle via the circulation along with inflammatory cells, as
shown by experimental studies (25, 42). The endomysial
CD133+/CD45+ cell ratio was significantly elevated in
inflamed versus control tissue. Given the normally low
numbers of bone marrowYderived HPCs less than 0.01% of
the total mononuclear fraction in human peripheral blood
(43), we speculate that CD133+ HPCs are mobilized from the
bone marrow into the circulation and recruited along with
myelomonocytic cell populations to injured muscle, presumably to augment repair mechanisms after inflammatory
damage. In accordance with enhanced regeneration during
inflammation, Pax7+ satellite cell numbers were significantly
increased in IM patients compared with controls.
To initiate tissue repair, sufficient vascularization is
necessary not only to ensure oxygen and nutrition supply but
also to facilitate migration of circulating progenitor cells to sites
of injury. In skeletal muscle, satellite cell niches organize near
capillaries, suggesting a functional relationship between myogenesis and vasculogenesis (19). Extensive experimental data
support the novel concept that the formation of new blood
vessels in postnatal life occurs simultaneously by angiogenesis
(which involves the proliferation and migration of resident EC
and remodeling of preexisting vessels) and vasculogenesis,
that is, the recruitment and incorporation of circulating bone
marrowYderived endothelial progenitor cells (EPCs) into the
vasculature (44). The origin and precise phenotypic and functional characteristics of progenitor cells with EC properties,
however, are incompletely defined. Peripheral blood cells that
coexpress the surface markers CD34, CD133, and VEGFR-2
have been described as a subset of circulating human EPCs(45);
these cells have a high potential to differentiate along the
endothelial lineage in vitro (23). Controversy regarding the role
of these cells in vasculogenesis, however, arose in recent work
that reported the failure of highly selected CD133+/CD34+/
VEGFR-2+ human umbilical cord blood cells to form perfused
vessels in vivo (46). In that study, only CD34+/CD45j cells
from umbilical cord blood formed endothelial colonies in a
colony-forming unit assay and demonstrated vessel-forming
activity in vivo. It seems likely that EPCs show distinct clonogenic potential and express certain surface markers depending
on the source and in different culture systems. Furthermore, the
interaction of CD133+/CD34+/VEGFR-2+ cells with cocultured cells may influence their differentiation.
In skeletal muscle of patients with IM, we observed
numerous interstitial mononuclear cells that were diffusely
interspersed in areas of inflammation and tissue injury and
expressed CD34 and/or CD31 antigens, although with less
intensity than mature ECs. CD133+ and VEGFR-2+ cells
DISCUSSION
Regeneration and repair of damaged muscle are
fundamental processes for rebuilding muscle integrity and
functional recovery of muscle strength after immunosuppressive treatment in patients with IMs (13). Such treatment is
poorly effective in IBM (4). Satellite cells are considered to
be the main source for muscle repair after injury in postnatal
individuals (34). These tissue-specific myogenic stem cells
reside under the basal lamina of mature myofibers (5) and
efficiently repair muscle damage in healthy individuals but
fail to compensate the progressive loss of functional muscle
tissue when confronted with degenerative muscle diseases
(35). Another population of lineage-uncommitted multipotent
stem cells that express CD34 and Sca-1 has been described in
the interstitial spaces of skeletal muscle. These cells are
distinct from satellite cells and are capable of forming various
cell constituents such as myogenic cells, ECs, adipocytes, and
fibroblasts (36Y38). It has been proposed that these so-called
Bside population stem cells[ constitute a reservoir of satellite
cells (37, 38) and originate from myelomonocytic progenitors
(18). In animal models, both satellite and interstitial stem cell
compartments can be repopulated by circulating bone
marrow-derived donor cells after transplantation (39, 40).
The physiologic relevance of this phenomenon and relevance
to human disease remain unclear (41).
We investigated the occurrence of HPCs with putative
myoendothelial potential in human skeletal muscle in
immune-mediated inflammation. In IM, we detected clusters
of endomysial CD133+ cells, which are considered to represent primitive HPCs (22), and determined that they colocalize with CD45+ mononuclear infiltrates. This contrasted
with the presence of only single CD133+ interstitial cells in
noninflammatory control muscle. Morphometric evaluation
demonstrated significantly increased CD133+ cell numbers in
PM and IBM compared with controls. Dermatomyositis
patient samples had the lowest numbers of endomysial
CD133+ HPCs among the IM patients studied, but there
was high interindividual variability in these samples such that
some HPC numbers overlapped with PM and controls. This
observation can be explained by the higher density of
endomysial inflammatory infiltrates in IM subtypes driven
by T-cell myocytotoxicity (i.e. PM and IBM) because
numbers of CD133+ cells strongly correlated with the density
of infiltrating CD45+ leukocytes. Considering the higher ages
of IBM patients, the high frequency of CD133+ HPCs in IBM
muscle was unexpected. This finding might be due to the
chronic inflammatory process and sustained degenerative and
Ó 2008 American Association of Neuropathologists, Inc.
717
Copyright @ 2008 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
Hollemann et al
J Neuropathol Exp Neurol Volume 67, Number 7, July 2008
invading the interstitial tissue were rarely observed. To characterize the coexpression pattern of CD34+ and CD133+ cells,
we performed confocal laser scanning microscopy and detected different subsets of CD133+/CD34j, CD133+/CD34+, and
CD133j/CD34+ cells that might reflect the maturation of
some invading primitive CD133+ cells into CD34+ endothelial progenitors. Interstitial CD34+ cells in muscle, however,
very likely represent a heterogeneous population originating
from different sources as they also may derive from resident
ECs (47) or invade skeletal muscle directly via the circulation
(48). Additionally, monocytes/macrophages themselves were
shown to express EC markers after stimulation with vascular
endothelial growth factor and basic fibroblast growth factor
(49). Independent from their progeny, in situ endothelial
differentiation of several interstitial CD34+ mononuclear cells
in skeletal muscle was indicated by coexpression of vWF.
In addition to their ability to improve neovascularization
in ischemic tissues (44), human circulating CD133+ and
CD34+ HPCs were shown to undergo myogenic differentiation when exposed to certain cytokines in culture (24, 25).
Moreover, they participate in muscle regeneration in vivo after
systemic or intramuscular delivery to transgenic scid/mdx-mice,
a murine model of Duchenne muscular dystrophy, causing partial restoration of dystrophin expression and functional improvement of muscle strength (25, 50). To assess the myogenic
potential of HPCs in human skeletal muscle, we used triple
staining for CD133, CD34, and the early myogenic marker
Pax7, which was shown to be necessary for the induction of
myogenic differentiation of murine satellite cells (8) and the
Sca-1+ side population stem cells (51). Immunostaining for
Pax7 identified satellite cells at typical anatomic sites at the
periphery of myofibers but also rounded Pax7+ mononuclear
cells scattered in the interstitial connective tissue. These observations are in agreement with a previous study that isolated
satellite cells and another distinct subpopulation of musclederived myogenic progenitors from human material (52). The
origin of interstitial myogenic progenitors, however, and their
functional relationship with sublaminal satellite cells remained
undetermined (53). Coexpression of CD133/CD34 and Pax7 by
cells in skeletal muscle of patients with IM suggests recruitment
of circulating HPCs into the myogenic pathway in humans as a
response to inflammatory damage. Interestingly, we found that
these cells express the transcription factor Pax7 in the cytoplasm
instead of the nucleus. Although we cannot explain this staining
pattern, the antibody to Pax7 might recognize different isoforms
of the molecule with different functions. Similarly, OCT-4,
another transcription factor that plays an important role in the
maintenance of the pluripotent state of embryonic stem cells,
was initially believed to be exclusively intranuclear. More recently, a second isoform, OCT-4B, was found that was mainly
localized to the cytoplasm (54). CD133+/Pax7+ or CD133+/
CD34+/Pax7+ cells were detectable in both the interstitial compartment and rarely in satellite cell niches. This is consistent
with demonstration of transplanted bone marrowYderived
myogenic progenitors in muscle connective tissue and satellite
cell niches and their fusion with preexisting myofibers (40).
Although rarely observed in human tissue, our study is
in accordance with experimental data supporting the concept
that circulating myoendothelial progenitors can colonize
skeletal muscle after injury, creating a Bsteady state[ of
equilibrium with resident satellite cells and EPCs and
participating in the permanent restoration of damaged muscle.
These complementary mechanisms of tissue regeneration
seemed to be particularly enhanced in patients with IBM,
which might reflect an attempt to compensate for the failure to
rebuild muscle after sustained degeneration and regeneration
cycles in the context of an Baged milieu[ of IBM muscle.
718
ACKNOWLEDGMENT
The authors thank Monika Zelle and Beate Rüger for
expert technical support.
REFERENCES
1. Dalakas MC, Hohlfeld R. Polymyositis and dermatomyositis. Lancet
2003;362:971Y82
2. Dalakas MC. Polymyositis, dermatomyositis and inclusion-body myositis. N Engl J Med 1991;325:1487Y98
3. Dalakas MC. Inflammatory disorders of muscle: Progress in polymyositis, dermatomyositis and inclusion body myositis. Curr Opin Neurol
2004;17:561Y67
4. Askanas V, Engel WK. Inclusion-body myositis: A myodegenerative
conformational disorder associated with Abeta, protein misfolding, and
proteasome inhibition. Neurology 2006;66:S39Y48
5. Mauro A. Satellite cell of skeletal muscle fibers. J Biophys Biochem
Cytol 1961;9:493Y95
6. Schultz E, McCormick KM. Skeletal muscle satellite cells. Rev Physiol
Biochem Pharmacol 1994;123:213Y57
7. Buckingham M, Bajard L, Chang T, et al. The formation of skeletal
muscle: From somite to limb. J Anat 2003;202:59Y68
8. Seale P, Sabourin LA, Girgis-Gabardo A, et al. Pax7 is required for the
specification of myogenic satellite cells. Cell 2000;102:777Y86
9. Relaix F, Rocancourt D, Mansouri A, et al. A Pax3/Pax7-dependent
population of skeletal muscle progenitor cells. Nature 2005;435:948Y53
10. Zammit PS, Relaix F, Nagata Y, et al. Pax7 and myogenic progression
in skeletal muscle satellite cells. J Cell Sci 2006;119:1824Y32
11. Tajbakhsh S, Buckingham M. The birth of muscle progenitor cells in the
mouse: Spatiotemporal considerations. Curr Top Dev Biol 2000;48:225Y68
12. Kassar-Duchossoy L, Gayraud-Morel B, Gomes D, et al. Mrf4 determines skeletal muscle identity in Myf5:Myod double-mutant mice.
Nature 2004;431:466Y71
13. Morosetti R, Mirabella M, Gliubizzi C, et al. MyoD expression restores
defective myogenic differentiation of human mesoangioblasts from
inclusion-body myositis muscle. Proc Natl Acad Sci U S A 2006;103:
16995Y7000
14. Long MA, Corbel SY, Rossi FM. Circulating myogenic progenitors and
muscle repair. Semin Cell Dev Biol 2005;16:632Y40
15. Ferrari G, Cusella-De Angelis G, Coletta M, et al. Muscle regeneration by
bone marrow-derived myogenic progenitors. Science 1998;279:1528Y30
16. Gussoni E, Soneoka Y, Strickland CD, et al. Dystrophin expression in the
mdx mouse restored by stem cell transplantation. Nature 1999;401:390Y94
17. Corbel SY, Lee A, Yi L, et al. Contribution of hematopoietic stem cells
to skeletal muscle. Nat Med 2003;9:1528Y32
18. Doyonnas R, LaBarge MA, Sacco A, et al. Hematopoietic contribution
to skeletal muscle regeneration by myelomonocytic precursors. Proc Natl
Acad Sci U S A 2004;101:13507Y12
19. Christov C, Chretien F, Abou-Khalil R, et al. Muscle satellite cells and
endothelial cells: Close neighbors and privileged partners. Mol Biol Cell
2007;18:1397Y409
20. Tamaki T, Uchiyama Y, Okada Y, et al. Functional recovery of damaged
skeletal muscle through synchronized vasculogenesis, myogenesis, and
neurogenesis by muscle-derived stem cells. Circulation 2005;112:2857Y66
21. Asahara T, Murohara T, Sullivan A, et al. Isolation of putative progenitor
endothelial cells for angiogenesis. Science 1997;275:964Y67
22. Yin AH, Miraglia S, Zanjani ED, et al. AC133, a novel marker for human
hematopoietic stem and progenitor cells. Blood 1997;90:5002Y12
23. Gehling UM, Ergun S, Schumacher U, et al. In vitro differentiation of
endothelial cells from AC133-positive progenitor cells. Blood 2000;95:
3106Y12
Ó 2008 American Association of Neuropathologists, Inc.
Copyright @ 2008 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 67, Number 7, July 2008
Hematopoietic Progenitors in Myositis
24. Pesce M, Orlandi A, Iachininoto MG, et al. Myoendothelial differentiation
of human umbilical cord blood-derived stem cells in ischemic limb
tissues. Circ Res 2003;93:e51Ye62
25. Torrente Y, Belicchi M, Sampaolesi M, et al. Human circulating AC133(+)
stem cells restore dystrophin expression and ameliorate function in dystrophic skeletal muscle. J Clin Invest 2004;114:182Y95
26. Jankowski RJ, Deasy BM, Cao B, et al. The role of CD34 expression
and cellular fusion in the regeneration capacity of myogenic progenitor
cells. J Cell Sci 2002;115:4361Y74
27. Kawamoto A, Iwasaki H, Kusano K, et al. CD34-positive cells exhibit
increased potency and safety for therapeutic neovascularization after
myocardial infarction compared with total mononuclear cells. Circulation 2006;114:2163Y69
28. Ferrara N, Gerber HP, LeCouter J. The biology of VEGF and its
receptors. Nat Med 2003;9:669Y76
29. Fina L, Molgaard HV, Robertson D, et al. Expression of the CD34 gene
in vascular endothelial cells. Blood 1990;75:2417Y26
30. Wright DE, Wagers AJ, Gulati AP, et al. Physiological migration of
hematopoietic stem and progenitor cells. Science 2001;294:1933Y36
31. Peault B, Rudnicki M, Torrente Y, et al. Stem and progenitor cells in
skeletal muscle development, maintenance, and therapy. Mol Ther 2007;
15:867Y77
32. Ruger B, Giurea A, Wanivenhaus AH, et al. Endothelial precursor cells
in the synovial tissue of patients with rheumatoid arthritis and osteoarthritis. Arthritis Rheum 2004;50:2157Y66
33. Wanschitz J, Maier H, Lassmann H, et al. Distinct time pattern of
complement activation and cytotoxic T cell response in Guillain-Barré
syndrome. Brain 2003;126:2034Y42
34. Dhawan J, Rando TA. Stem cells in postnatal myogenesis: Molecular
mechanisms of satellite cell quiescence, activation and replenishment.
Trends Cell Biol 2005;15:666Y73
35. Buckingham M. Skeletal muscle progenitor cells and the role of Pax
genes. C R Biol 2007;330:530Y33
36. Young HE, Steele TA, Bray RA, et al. Human reserve pluripotent
mesenchymal stem cells are present in the connective tissues of skeletal
muscle and dermis derived from fetal, adult, and geriatric donors. Anat
Rec 2001;264:51Y62
37. Asakura A, Seale P, Girgis-Gabardo A, et al. Myogenic specification of
side population cells in skeletal muscle. J Cell Biol 2002;159:123Y34
38. Tamaki T, Akatsuka A, Ando K, et al. Identification of myogenicendothelial progenitor cells in the interstitial spaces of skeletal muscle.
J Cell Biol 2002;157:571Y77
39. LaBarge MA, Blau HM. Biological progression from adult bone marrow
to mononucleate muscle stem cell to multinucleate muscle fiber in
response to injury. Cell 2002;111:589Y601
40. Dreyfus PA, Chretien F, Chazaud B, et al. Adult bone marrow-derived
stem cells in muscle connective tissue and satellite cell niches. Am J
Pathol 2004;164:773Y79
41. Sherwood RI, Christensen JL, Weissman IL, et al. Determinants of
skeletal muscle contributions from circulating cells, bone marrow cells,
and hematopoietic stem cells. Stem Cells 2004;22:1292Y304
42. Palermo AT, Labarge MA, Doyonnas R, et al. Bone marrow
contribution to skeletal muscle: A physiological response to stress.
Dev Biol 2005;279:336Y44
43. Nolta JA, Jordan CT. Spotlight on hematopoietic stem cells: Looking
beyond dogma. Introduction. Leukemia 2001;15:1677Y80
44. Ribatti D. The discovery of endothelial progenitor cells. An historical
review. Leuk Res 2007;31:439Y44
45. Peichev M, Naiyer AJ, Pereira D, et al. Expression of VEGFR-2 and
AC133 by circulating human CD34(+) cells identifies a population of
functional endothelial precursors. Blood 2000;95:952Y58
46. Yoder MC, Mead LE, Prater D, et al. Redefining endothelial progenitor
cells via clonal analysis and hematopoietic stem/progenitor cell
principals. Blood 2007;109:1801Y9
47. Vailhe B, Vittet D, Feige JJ. In vitro models of vasculogenesis and
angiogenesis. Lab Invest 2001;81:439Y52
48. Asahara T, Kawamoto A. Endothelial progenitor cells for postnatal
vasculogenesis. Am J Physiol Cell Physiol 2004;287:C572Y79
49. Scavelli C, Nico B, Cirulli T, et al. Vasculogenic mimicry by bone
marrow macrophages in patients with multiple myeloma. Oncogene
2008;27:663Y74
50. Torrente Y, Tremblay JP, Pisati F, et al. Intraarterial injection of
muscle-derived CD34(+)Sca-1(+) stem cells restores dystrophin in mdx
mice. J Cell Biol 2001;152:335Y48
51. Seale P, Ishibashi J, Scime A, et al. Pax7 is necessary and sufficient for
the myogenic specification of CD45+:Sca1+ stem cells from injured
muscle. PLoS Biol 2004;2:E130
52. Alessandri G, Pagano S, Bez A, et al. Isolation and culture of human
muscle-derived stem cells able to differentiate into myogenic and
neurogenic cell lineages. Lancet 2004;364:1872Y83
53. Pavlath GK, Gussoni E. Human myoblasts and muscle-derived SP cells.
Methods Mol Med 2005;107:97Y110
54. Lee J, Kim HK, Rho JY, et al. The human OCT-4 isoforms differ in their
ability to confer self-renewal. J Biol Chem 2006;281:33554Y65
Ó 2008 American Association of Neuropathologists, Inc.
719
Copyright @ 2008 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.