вход по аккаунту


Elevated basic fibroblast growth factor in the serum of patients with Duchenne muscular dystrophy.

код для вставкиСкачать
Elevated Basic Fibroblast
Growth Factor in the
Serum of Patients with
Duchenne Muscular
P. A. DAmore, PhD,"t'f R. H. Brown, Jr, MD,'I
P-T. Ku, BA,+ E. P. Hoffman, PhD,'X"
H. Watanabe, MS,W K. Arahata, MD#
T. Ishihara, MD,@ and J. Folkman, MD*B
The mechanism whereby dystrophin deficiency leads to
excessive fibrosis and muscle degeneration is not known.
The absence of dystrophin in skeletal muscle is associated with reduced plasma membrane stability as evidenced by elevated serum levels of the cytoplasmic enzyme creatine kinase. Basic fibroblast growth factor, a
cytoplasmic polypeptide growth regulator that stimulates connective tissue synthesis, induces satellite cell
proliferation, and suppresses myogenic differentiation,
is made by skeletal muscle. We hypothesize that dystrophin deficiency leads to the constant release of basic
fibroblast growth factor, which in turn contributes to
fibrosis and muscle weakness by stimulating connective
tissue and suppressing skeletal muscle differentiation.
As an initial step in testing this hypothesis, we measured
basic fibroblast growth factor in the serum of Duchenne
muscular dystrophy patients. We found that whereas basic fibroblast growth factor was undetectable in the serum of normal individuals (n = 200), levels were elevated in 11 of 18 patients with Duchenne muscular
DAmore PA, Brown RH Jr, Ku P-T, Hoffman EP,
Watanabe H, Arahata K, Ishihara T, Folkman J.
Elevated basic fibroblast growth factor in the
serum of patients with Duchenne muscular
dystrophy. Ann Neurol 1994;35:362-365
Duchenne muscular dystrophy (DMD) is known to be
caused by mutations of the dystrophin gene. Dys-
From rhe 'Laboratory for Surgical Research, Children's Hospital,
Boston; tProgram in Cell and Developmental Biology and Departments of $Pathology and $Anatomy, Harvard Medical School, Boston; (Cecil B. Day Laboratory for Neuromuscular Research, Massachusetts General Hospital, Boston, MA; **Departmentof Molecular
Genetics and Biochemistry, University of Pittsburgh Medical School,
Pittsburgh, PA; ' i f a k e d a Chemical Corporation, Osaka; $$National
Institute of Neuroscience, Tokyo; and 83Narional Higashi Saitama
Hospital, Saitama, Japan.
Received Mar 15, 1993, and in revised form Jul 30 and Sep 1.
Accepted for publication Sep 2, 1993.
Address correspondence co Dr D' Amore, Laboratory for Surgical
Research, Children's Hospital, 300 Longwood Avenue, Boston, MA
trophin is part of the plasma membrane cytoskeleton
of muscle fibers; its deficiency in Duchenne dystrophy
is associated with skeletal membrane instability that
characterizes D M D {l}.It is hypothesized that muscle
membrane fragility in humans may trigger aggressive
proliferation of fibrotic tissue in the muscle, which in
turn leads to failure of muscle regeneration and weakness 121.
The fibroblast growth factors (FGFs) are a family of
growth regulators that have a variety of effects on cells
of mesenchymal and neural lineage, including stimulation of connective tissue proliferation and satellite cell
proliferation, and suppression of myogenic differentiation in vitro {3]. Basic FGF (bFGF) does not possess
the leader sequence required for secretion by classic
secretory pathways {4]. It has been suggested that
bFGF acts as a "wound hormone"; following local release as a result of injury, cell necrosis, or mobilization
from the extracellular matrix, where it has been shown
to be sequestered, bFGF initiates or participates in regeneration and the healing process [Sf. Thus, as part
of the normal repair process, bFGF may be released
from skeletal or cardiac muscle "injured" by exercise
[ b ] or ischemia [7}, resulting in increased vascular supply and enhanced muscle regeneration.
Assuming that one mode of bFGF release is through
compromised plasma membranes, and knowing that an
absence of dystrophin leads to altered membrane integrity, we hypothesize that dystrophin deficiency leads
to the release of muscle bFGF. In addition, we speculate that the inappropriate, continuous release of bFGF
contributes to excessive fibrosis and scarring and suppresses muscle regeneration. As an initial test of this
hypothesis we assayed the serum of DMD patients for
the presence of bFGF.
Materials and Methods
Blood samples were obtained by standard venipuncture technique. Cellular and clotted elements were removed by centrifugation (800 rpm, 10 minutes, 4°C). Aliquots of serum
were removed for immediate determination of creatine kinase (CK) and the remainder was frozen at -20°C until it
was assayed for bFGF using an enzyme-linked immunoassay
that has been described It;}. This assay employs three munoclonal antibodies; the Fab fragment of one monoclonal antibody is conjugated to horseradish peroxidase, while a mkture of the other two monoclonal antibodies is in the solid
phase. The bound peroxidase activity is developed with 0phenylenediamine, which is detected at 492 nm. The assay
is sensitive to 30 pgiml and does not derect acidic FGF. Tho
recovery of human recombinant bFGF, when added to serum
at find concentrations of 82 and 740 pg/ml, was 103 2
15.0% and 102.9 9.5%, respectively [S]. The intraassay
variation is 4 to 10% and the interassay variation is 6 to
The inclusion criteria for DMD were as follows. For Patients l , 3 to 6 , 9 to 13, and l ? , immunofluorescence, immu-
Copyright 0 1994 by the American Neurological Association
noblot, and DNA analyses were conducted. For patients 2,
7, 8, and 18, the criteria included male sex, infancy onset
of severe progressively proximal weakness, presence of calf
hypertrophy, markedly elevated CK level, and dystrophic
histological abnormalities as seen in muscle biopsy specimen
and electromyography (EMG), findings consistent with m u cular dystrophy. For Patients 14 and 15, muscle biopsy was
not possible and there was no DMD gene deletion.
Tuble I . Busic Fibroblust Growth Factor (bFGF) and Creatine
Kinase (CK) in Serum of Patients with Duchenne Afuscdap
Sample No.
CK (unirs/L)a
Age (yr)
< 30
2,02 1
4,3 12
< 30
< 30
< 30
W e examined the sera of 200 normal individuals and
found that none had levels of bFGF above 30 pg/ml.
bFGF levels in the serum of 18 individuals with D M D
(Table 1) and 20 individuals with other neuromuscular
disorders (Table 2) were also assayed. Eleven of these
18 D M D serum samples contained elevated levels of
bFGF; the mean bFGF level was 80 pg/ml and the
range was 40 to 165 pgiml. Five of 20 serum samples
from patients with other neuromuscular disorders had
elevated levels of bFGF; the mean was 54 pgiml with
a range of 36 to 70 pgiml.
One consequence of the reduced plasma membrane
stability associated with dystrophin deficiency is elevated serum levels of the skeletal muscle cytoplasmic
enzyme CK [ 11. Since both CK and bFGF are localized
in the cytoplasm, we predicted that they would be released together. The individuals with DMD whose sera
we measured in this study had a mean CK level of
7,358 k 6,330 unitslliter with a range of 1,046 to
24,100 unitdliter. (The variances noted here and
throughout this report are standard deviations.) T o determine if the observed variability in the serum bFGF
levels of the D M D patients correlated with the variation in CK, we compared bFGF and CK levels in the
D M D patients. In every D M D serum sample in which
bFGF was not detectable (7/18), the CK levels were
below 6,000 unitsiliter (see Table 1; Samples 1-7).
The two additional D M D samples with CK levels below 6,000 unitdliter (see Table 1; Samples 8 and l l )
had relatively low bFGF levels (40 and 50 pg/ml, respectively). Although there is not a direct correlation
between the CK and bFGF levels in all of the D M D
sera, there is a clear distinction between CK levels of
the samples with and those without detectable bFGF.
The mean CK level in the D M D sera samples without
detectable bFGF was 2,985
1,867, whereas the
mean CK level in the samples with a bFGF value above
30 pglml was 10,141 2 6,642. The difference between these groups is significant at p = 0.0033 (unpaired nonparametric test; Mann-Whitney two-sample
There was also a relationship between the age of the
D M D patient and serum CK and bFGF levels; older
patients (10-20 years old) had significantly lower CK
and bFGF levels than did the younger patients (4-9
years old). Patients 10 to 20 years old had a mean CK
Serum bFGF (pg/ml)
< 30
< 30
< 30
"The upper limit of normal for the laboratory that assayed CK in
serums of Patients 1, 3-6, 9-13, and 17 was 75 units/L. The upper
limit of normal for the other samples was 400 units/L.
Table 2. Basic Fibroblast Grouth Factor IbFGF) In the S e r m
of PutrentJ wzth Nenrommcular Di~ordersOther Than
Durhenne Musczllur Dystrophy
Sample No
M yopathic
Serum bFGF
Charcot-Marie-Tooth (neuronai)
Motor neuroparhy
Demyeiinating neuropathy
Charcor-Marie-Tooth (dernyelinative)
Post-polio syndrome
Becker dystrophy
Scapuloperoneai myopathy
Limb-girciie myopathy
Myoparhy-type unknown
lMyoshi distal myoparhy
Mitochondria1 myopathy
Diffuse rhabdomyolysis
Myoparhy-type unknown
Myoshi distal myopathy
Mysthenia gravis
Neuroacanthoc ytosis
< 30
< 30
< 30
< 30
< 30
< 30
< 30
< 30
< 30
amyotrophic lateral sclerosis.
Brief Communication: DAmore et al: bFGF and Duchenne Muscular Dystrophy 363
value of 3,134 k 1,932 whereas patients 4 to 9 years
old had a mean CK value of 12,638 2 5,732 (p =
0.0005; Mann-Whitney two-sample rest; unpaired nonparametric). All of the serum samples with CK values
of 6,000 or less were from patients 10 years or older
(see Table 1;Samples 1-43, 10, and 11). Of the patients
10 years or older, 3 had detectable bFGF (see Table
1; Samples 8, 10, and ll), though these bFGF levels
were among the lowest measured. All of the patients
younger than 10 years had CK levels over 7,000 units/
liter and all of the sera from these patients had detectable bFGF. The highest bFGF values measured, 95 to
165 pg ml, about 16 to 28 times that reported for
normal serum bFGF (see Table 1; samples 15-18),
were from patients 7 years or younger.
The observation of abnormally elevated bFGF levels
in the sera of some DMD patients is consistent with
our hypothesis that instability of the muscle plasma
membrane due to dystrophin deficiency leads to leakage of bFGF. This finding is remarkable in light of our
observation that bFGF was not measurable in the serum of 200 normal individuals. The only other circumstances in which serum bFGF has been detected are in
certain cancers [9}, following surgery, during pregnancy, and following myocardial infarction (J. Folkman, unpublished data, 1792). All of these conditions
are characterized by tissue injury or necrosis, and/or
rapid angiogenesis. We have been unable to detect
bFGF in the serum of dystrophin-deficient mice (mdx)
of ages 2 weeks through adulthood (data not shown).
This observation is of interest since these animals lack
dystrophin, but do not develop endomysial fibrosis.
Not all of the DMD sera samples had elevated bFGF
levels; 7 had a bFGF level below 30 pgiml, the lower
limits of detection of our assay. The CK levels in these
7 serum samples were all relatively low (< 6,000 units!
liter). We know from data obtained by others that normal serum contains approximately 5 pg/ml of bFGF
[9]. Given the correlation between CK and bFGF levels, we suspect that these sera might have elevated,
although relatively low, bFGF levels. Our inability to
detect bFGF levels between 6 and 30 pg/ml prevents
us from examining this possibility.
All of the younger patients (< 10 years old) had
both high serum CK levels (> 9,000 unitdliter) and
detectable serum bFGF, whereas the older patients had
lower CK levels and generally undetectable bFGF. We
suspect that the loss of muscle mass associated with the
progression of the disease process accounts for these
findings. Interestingly, CK and FGF levels were not
strictly correlated. The highest CK values were not
always associated with the highest bFGF levels. Sera
with CK levels of 24,100 and 19,390 unitsAiter had
bFGF levels of 120 and 138 pglml, respectively,
364 Annals of Neurology
Vol 35
No 3 March 1994
whereas a serum sample with a CK level of 9,622
yielded the highest measured bFGF (165 pgiml). We
are aware that there is great variability among the serum CK levels for D MD patients. It is known that the
time of day at which DMD sera are collected influences
CK levels (CK levels are higher at the end of the day).
Since the DMD serum samples included in our study
were drawn by two individuals at different times of the
day (Samples 1, 3-6, 7-13, and 17 were collected in
the morning, whereas the remainder were collected in
the late afternoon), we suspected that the different
times might contribute to the lack of strict correlation
between the CK and bFGF levels. However, separation of serum samples on the basis of their collection
time did not reveal any greater correlation between
bFGF and CK. An alternate explanation for the nonlinearity of the relationship between bFGF and CK levels
may be unequal degradation by proteases that are released or activated as part of the disease process.
bFGF was detected in 5 patients with neuromuscular
disorders other than DMD. The explanation for these
elevated levels is not known at this time but may be
related to alterations in membrane permeability or integrity associated with these diseases.
The ability to measure circulating bFGF under any
circumstance is surprising in light of our previous observation that the clearance time for intravenously administered bFGF is extremely short (in the order of
minutes) [lo}. One possible explanation for the elevated bFGF levels in D MD is that although bFGF is
rapidly cleared, circulating levels are maintained because bFGF is chronically released from muscle tissue
throughout the entire body. This release could be due
to myofiber necrosis. In addition, the very large syncytial muscle fibers appear to be able to withstand membrane damage without becoming necrotic. As evidence
of this fact, serum albumin is found in viable dystrophin-deficient myofibers [ 1 I}. bFGF released from
muscle would be expected to be bound by heparan
sulfate in the basal lamina of muscle fibers, a speculation supported by the finding of elevated bFGF levels
in the basement membranes of muscle from mdx mice
{12]. Thus, the bFGF-binding sites may be overwhelmed by excessive bFGF release, leading to spillover into the serum. Alternatively, bFGF, once
released, may complex with a heparin-like glycosaminoglycan or another binding protein that stabilizes it
against clearance. Since mast cells are known to contain
and release heparin-rich granules, it is possible that the
glycosaminoglycan released by mast cells binds to and
stabilizes the bFGF released by the dystrophic muscle
tissue. Findings in humans previously suggested an important role for mast cells in the early phases of DMD
[I 3}. More recently, histological analyses of mast cells
in normal and dystrophin-deficient muscle in humans,
dogs, and mice revealed a strong correlation between
mast cell localization and disease progression
Gorospe, M. D. Tharp, J. Hinckley, et al, unpublished
data). Thus, whereas the local release of bFGF from
injured or exercised muscle may normally lead to local
muscle healing and regeneration, the chronic release
of bFGF by d ystrophin-deficient muscle may result in
the excess fibrosis and scarring associated with DMD.
This work w1\s supported by NIH grant EY 05985 (to P. A. D.)
and grants from the Muscular Dystrophy Association and C.B. Day
Foundation (to R. H. B.).
The authors are grateful to D r T. L. Munsat for assisrance in obtaining D M D blood samples, to D r Lou Kunkel and Sandra R. Smith
for their critical reading of the manuscript, and to Carlene Pavlos
for help in preparation and editing of the manuscript
1. Rowland LP. Biochemistry of muscle membranes in Duchcnne
muscular dystrophy. Muscle Nerve 1980;3:3-20
2. Hoffman BP, Gorospe JR. The animal models of Duchenne
dystrophy: windows on the pathophysiological consequences of
dystrophin deficiency. In: Mooseker MS. Morrow JS, eds. Topics in membranes. New York: Academic, 1991:113-154
'3. Saunders KB, DAmore PA. FGF and TGF-P: actions and intrractions in bicilogical systems. Crir Rev Eukary Gene Express
4 . Abraham JA, Mergia A, Whang JL, et al. Nucleotide sequence
of a bovine clone encoding the angiogenic protein, basic fibroblast growth factor. Science 1986;233:545-348
5. hlcNeil PL. Cell wounding and healing: the opening and closing
of wounds in the cell membrane pruvides a route for the release
of moleculrs that maintain and repair tissue. Am Sci 1991;79:
6. Morrow NG, Kraus WE. Moore JW, e t al. Increased expression
of fibroblast growth facrors in a rabbit skeleral muscle model of
exercise conditioning. J Clin Invest 1990;85:1816-1820
7. Schaper W. Development and role of coronary collaterals.
Trends Cardiovasc Med 1991;1:256-261
8. Watanabe H, Hori A, Seno M, e t al. A sensitive enzyme immunoassay for human basic fibroblast growth fixtor. Biochem Biophys Res Commun 1991;175:229-235
9. Ii M, Yoshida H , A r m & Y, e t al. Improved enzyme immunoassay for human basic fibroblast growth factor using a new enhanced chemiluminescence system. Biochem Biophys Res Commun 1993;193:540-545
10. Thompson RW, Whalen GF, Saunders KB, et al. Heparinmediated release of fibroblast growth factor-like activity into rhe
circulation of rabbits. Growth Factors 1990;3:22 1-229
11. Morandi L, Mora M, Gussoni E, e t al. Dystrophin analysis in
Duchenne and Becker muscular dystrophy carriers: correlation
with intercellular calcium and albumin. Ann Neurol 1990;28:
12. DiMario J, Buffinger N, Yamada S, et al. Fibroblast growth
factor in the extracellular matrix of dystrophic (mdx) mouse
muscle. Science 1989;224:688-690
13. Hellia~ellTR, Gunhan 0, Edwards RHT. Mast cells in neuromuscular diseases. Neurol Sci 1990;98:267-276
in Twins: Study of
324 3-Myopathy,
Encephalopathy, Lactic
Acidosis, and Strokelike
Episodes Mitochondrid
DNA Mutation
F. Degoul, PhD," M. Diry," A. Pou-Serradell, MD;)
J. Lloreta,? a n d C. Marsac, MD'
Two dizygotic twins with myopathy and leukoencephaIopathy are described. The female twin had an incomplete form of MELAS syndrome (myopathy, encephalopathy, lactic acidosis, and strokelike episodes)with severe
myopathy, epileptic seizures without strokelike episodes. The male twin presented clinical features exciusively of myopathy and subclinical leukoencephalopathy. The MELAS mitochondrial DNA point mutation
(MELAS-3243)was found by southern blot and polymerase chain reaction in muscle, skin fibroblasts, and blood
of the female twin and was not detected in the skin
fibroblasts nor in the blood of the mother, nor in any of
the tissues tested in the male twin. The absence of mutation in male twin tissues raises questions about the
pathogenetic significance of the mutation in this family.
Degoul F, D i r y M, Pou-Serradell A , Lloreta J,
Marsac C. Myo-leukoencephalopathy in twins:
study of 3243-myopathy, encephalopathy, lactic
acidosis, a n d strokelike episodes mitochondrial
DNA mutation. A n n Neurol 1994;35:365-370
Within the past 3 years, at least three specific point
mutations (3243,3271, 3250) have been described in
the mitochondrial tRNALeu(UUR'
gene associated with
MELAS (myopathy, encephalopathy, lactic acidosis,
and strokelike episodes) syndrome, a maternally inherited syndrome [l, 21 or with mitochondria1 myopathy
131. However, it has been demonstrated recently that
in spite of being a high level of concordance between
the clinical diagnosis of MELAS syndrome and the
presence of the 3243 mutation, this mutation is also
exhibited in patients with incomplete MELAS syndrome 141, in asymptomatic relatives of MELAS paFrom "INSERM U75, Paris, France, and tSemice of Neurology,
Hospital del Mar, Universitat Autonoma de Barcelona, Barcelona,
Received Mar 5 , 1993, and in revised form Jul 20 and Aug 27
AcLepted for publication Aug 31, 1993
Address correspondence to Ur Degoul, INSERM U 75, 156 rue de
vaugirard, F-75015 Pans, France
Copyright 0 1994 by t h e American Neurological Association
Без категории
Размер файла
440 Кб
factors, patients, growth, basic, duchenne, serum, elevated, muscular, dystrophy, fibroblasts
Пожаловаться на содержимое документа