Measurement of myostatin concentrations in human serum:
Circulating concentrations in young and older men and effects
of testosterone administration
Kishore M. Lakshmana, Shalender Bhasina,***8727;, Christopher Corcoranb, Lisa A. Collins-Racie b,
Lioudmila Tchistiakova b, S. Bradley Forlowc, Katie St. Ledger c, Michael E. Burczynski c,
Andrew J. Dorner b, Edward R. LaVallie b
a Section of Endocrinology, Diabetes, and Nutrition, Boston University School of Medicine, Boston Medical Center, 670 Albany Street, Boston, MA 02118, United States b Department of Biological Technologies, Wyeth Research, 35 Cambridge Park Drive, Cambridge, MA 02140, United States
c Department of Clinical Translational Medicine, Wyeth Research, 500 Arcola Road, Collegeville, PA 19426, United States
article info
Article history:
Received 31 August 2008
Received in revised form 19 December 2008
Accepted 20 December 2008
Keywords:
Myostatin assay
Sarcopenia
Myostatin and lean body mass
Myostatin and testosterone
Age effects on myostatin
abstract
Methodological problems, including binding of myostatin to plasma proteins and cross-reactivity of assay
reagents with other proteins, have confounded myostatin measurements. Here we describe development
of an accurate assay for measuring myostatin concentrations in humans. Monoclonal antibodies that bind
to distinct regions of myostatin served as capture and detector antibodies in a sandwich ELISA that used
acid treatment to dissociate myostatin from binding proteins. Serum from myostatin-deficient Belgian
Blue cattle was used as matrix and recombinant human myostatin as standard. The quantitative range
was 0.15***8211;37.50 ng/mL. Intra- and inter-assay CVs in low, mid, and high range were 4.1%, 4.7%, and 7.2%,
and 3.9%, 1.6%, and 5.2%, respectively. Myostatin protein was undetectable in sera of Belgian Blue cattle
and myostatin knockout mice. Recovery in spiked sera approximated 100%. ActRIIB-Fc or anti-myostatin
antibody MYO-029 had no effect on myostatin measurements when assayed at pH 2.5. Myostatin levels
were higher in young than older men (mean ± S.E.M. 8.0 ± 0.3 ng/mL vs. 7.0 ± 0.4 ng/mL, P = 0.03). In men
treated with graded doses of testosterone, myostatin levels were significantly higher on day 56 than
baseline in both young and older men; changes in myostatin levels were significantly correlated with
changes in total and free testosterone in young men. Myostatin levels were not significantly associated
with lean body mass in either young or older men.
Conclusion: Myostatin ELISA has the characteristics of a valid assay: nearly 100% recovery, excellent precision,
accuracy, and sufficient sensitivity to enable measurement of myostatin concentrations in men and
women.
© 2009 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
A substantial body of evidence from experiments of nature and
man supports the view that myostatin is an important inhibitor
of skeletal muscle growth (Grobet et al., 1997; Kambadur et al.,
1997; McPherron et al., 1997; McPherron and Lee, 1997; GonzalezCadavid
et al., 1998; Reisz-Porszasz et al., 2003; Schuelke et al.,
2004). Naturally occurring mutations of myostatin in cattle, sheep,
mice, dogs, and humans are associated with hypermuscularity
Grant support: The collection of clinical samples and data analyses were
supported by NIH grants 1UO1AG114369, 1RO1DK59627, 1RO1DK70534, and
1P30AG31679-01. Measurements of serum myostatin levels were performed by
Wyeth Research Laboratories.
***8727; Corresponding author. Tel.: +1 617 414 2951; fax: +1 617 638 8217.
E-mail address:
bhasin@bu.edu (S. Bhasin).
(Grobet et al., 1997; Kambadur et al., 1997; McPherron and Lee,
1997; Reisz-Porszasz et al., 2003; Schuelke et al., 2004; Mosher et
al., 2007). Similarly, knockout mice carrying a targeted deletion of
the myostatin gene (Lee, 2008) and transgenic mice that hyperexpressmyostatin
prodomain (Yang et al., 2001; Bhasin et al., 2005), or
are unable to process myostatin into its active product have higher
skeletal muscle mass than their wild type controls (Zhu et al., 2000).
Conversely, transgenic mice which hyperexpress myostatin in the
skeletal muscle have lower muscle mass compared to wild type
controls (Reisz-Porszasz et al., 2003). Inhibition of myostatin action
in adult animals by myriad strategies ***8211; administration of monoclonal
antibodies, propeptide, follistatin, or soluble ActRIIB receptor
***8211; increases muscle mass, suggesting that myostatin also restrains
skeletal muscle mass during adult life (Whittemore et al., 2003; Lee,
2004; Lee et al., 2005; Welle et al., 2007; Nakatani et al., 2008).
Myostatin is a secreted protein derived by proteolytic cleavage
of its precursor protein at a dibasic site by a furin type protease;
0303-7207/$ ***8211; see front matter © 2009 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.mce.2008.12.019
K.M. Lakshman et al. / Molecular and Cellular Endocrinology 302 (2009) 26***8211;32 27
this proteolytic cleavage generates the mature myostatin protein
and a propeptide, which binds myostatin and inhibits its action
(Hill et al., 2002, 2003; Lee, 2008). Myostatin circulates in plasma
bound to several proteins, including its propeptide, follistatinrelated
protein, and growth and differentiation factor-associated
serum protein-1 (GASP-1) (Hill et al., 2002, 2003; Lee, 2008).
Myostatin bound to these plasma proteins in a latent complex is
unable to bind its receptor and activate signaling (Hill et al., 2002,
2003; Lee, 2008). The precise mechanisms leading to activation
of myostatin in its target tissues are not fully understood. Proteolytic
cleavage of the bound myostatin propeptide by members
of bone morphogenetic protein (BMP)-1-tollloid family of metalloproteinases
(Wolfman et al., 2003) or extracellular processing and
maturation of pro-myostatin by furin (Anderson et al., 2008) are
potential mechanisms for activation of myostatin.
Although several studies have reported the presence of myostatin
protein in blood (Zachwieja et al., 1999; Gonzalez-Cadavid
and Bhasin, 2004; Walker et al., 2004; Hosoyama et al., 2006),
methodological problems have confounded the measurements of
circulating myostatin concentrations. The earlier studies using
direct radioimmunoassays preceded the recognition that myostatin
circulates as a latent complex bound to plasma proteins
(Gonzalez-Cadavid et al., 1998; Zachwieja et al., 1999; Walker et
al., 2004; Hosoyama et al., 2006). Also, the cross-reactivity of other
plasma proteins in the myostatin assays was not fully appreciated
(Zachwieja et al., 1999; Walker et al., 2004; Hosoyama et al., 2006).
Therefore, there has been a paucity of accurate data on circulating
myostatin levels in healthy young and older individuals, and the
sparse data available have been contradictory. For instance, some
studies reported higher myostatin levels in older men and women
while others found lower or unchanged myostatin expression in
skeletal muscle of older men and women in comparison to young
men and women (Zachwieja et al., 1999; Welle et al., 2002; Kim et
al., 2005; Raue et al., 2006).
Here, we report the development and validation of an accurate
assay for the measurement of serum myostatin levels in humans.
The assay uses acid treatment to strip myostatin from its binding
proteins. Serum from Belgian Blue cattle (naturally devoid of
myostatin protein due to an inactivating mutation) was used as
the matrix, and recombinant human dimeric myostatin protein as
reference standard. Using this validated assay, we measured the circulating
myostatin concentrations in healthy young and older men.
We also determined the effects of testosterone treatment on circulating
myostatin levels using stored samples from a testosterone
dose response study. In that study, the details of which have been
published (Bhasin et al., 2001, 2005; Storer et al., 2003), administration
of graded doses of testosterone to healthy young and older men
was associated with dose-dependent increases in skeletal muscle
mass. We hypothesized that testosterone-associated increase
in muscle mass would be associated with suppression of serum
myostatin levels. However, we recognized that if myostatin serves
as a chalone to regulate muscle growth, as has been proposed by
Lee (2004) and others (Gaussin and Depre, 2005), myostatin levels
would be expected to increase after testosterone administration as
a counter-regulatory mechanism to restrain testosterone-induced
increase in muscle mass.
2. Methods
Human subjects. Serum samples were derived from healthy young men, aged
18***8211;35 years, and older men, aged 60***8211;75 years, with normal testosterone levels, who
were participants in a testosterone dose response study (Bhasin et al., 2001, 2005).
The design and main findings of this study have been reported previously (Bhasin
et al., 2001, 2005). The study protocols were approved by the institutional review
boards of Charles Drew University and Harbor-UCLA Research and Education Institute.
All participants provided informed consent. Exclusion criteria included history
of prostate cancer, PSA >4 ng/mL, AUA lower urinary tract symptom score >7, hematocrit
>48%, severe sleep apnea, diabetesmellitus, congestive heart failure,myocardial
infarction in the preceding 6 months, androgen use in the preceding year, or participation
in moderate to intense exercise training. After a 4-week control period,
participants were randomized to one of five treatment groups to receive monthly
injections of a GnRH agonist (leuprolide depot, 7.5 mg; TAP, North Chicago, IL) to suppress
endogenous testosterone production. The participants also received weekly
intramuscular injections of testosterone enanthate (TE, Delatestryl, 200 mg/mL;
Savient Pharmaceuticals, Iselin, NJ) in one of five doses: 25, 50, 125, 300, or 600 mg
(Bhasin et al., 2001, 2005). Treatment duration was 20 weeks. The Data and Safety
Monitoring Board stopped the 600 mg TE dose group in December 2002 due to a
number of adverse events in older men in this dose group. After this point, randomization
was limited to one of four TE dose groups: 25, 50, 125, or 300 mg weekly.
Healthy and menopausal women. Serum samples of healthy, menstruating
women (n = 33), 19***8211;21 years of age, and postmenopausal women (n = 37), 67***8211;87
years of age, were purchased from BioServe, Beltsville, MD. These participants had
consented to participate in an IRB-approved Bioserve study. Surgically menopausal
women were 18***8211;55 years of age (n = 24), who had ovarian surgery at least 6 months
before enrollment and serum FSH >30 U/L, BMI <35 kg/m2, a normal PAP smear and
mammogram in the preceding 12 months, and who had provided informed consent
approved by the Boston University IRB.
2.1. Myostatin assay
Reagents. Monoclonal antibodies to myostatin were raised in myostatin-null
mice (Wyeth Research). Antibodies that bind to distinct regions of myostatin (data
not shown) were used as the capture (RK35) and detector (RK22) antibodies. RK22
was biotinylated using EZ®-Link Sulfo-NHS-LCBiotinylation Kit (Pierce, Rockford,
IL). MYO-029 is a humanized murine monoclonal antibody that binds to ActRIIB
receptor-interaction site on myostatin (Girgenrath et al., 2005; Wagner et al., 2008).
Recombinant mature human myostatin, and ActRIIB extracellular domain fused
to Fc portion of IgG (del Re et al., 2004), was expressed and purified from CHO
cells (Wyeth BioPharma, Andover, MA). Myostatin-deficient serum from Belgian
Blue cattle (Grobet et al., 1997; Kambadur et al., 1997), normal bovine serum, and
cynomolgous monkey serum was obtained from Bioreclamation (Hicksville, NY).
Serum from myostatin-null mice (C57BL/6 background) (McPherron et al., 1997)
and normal littermates was obtained from Wyeth BioResources.
Myostatin assay method. Serum samples (or calibrator samples in Belgian Blue
serum) were mixed with acid dissociation buffer (0.2 M glycine***8211;HCl pH 2.5) at a ratio
of 1:13.3. For non-dissociative assays, samples were mixed with THST buffer (50 mM
Tris***8211;HCl pH 8.0, 500 mM NaCl, 1 mM glycine, 0.05% Tween-20; pH 8.0). Assay plates
were incubated with 2.0 g/mL RK35 in coating buffer (100 mM sodium borate pH
9.1) overnight at 4 ***9702;C, washed, and blocked with 200 L/well of SuperBlock-TBS
(Pierce). Diluted serum samples (100 L) were transferred to assay plate, incubated
at room temperature for 90 min, washed 4 times with THST and 100 L biotinlylated
RK-22 secondary antibody (0.1 g/mL) was added to each well for 90 min at room
temperature. Plates were washed 4 times with THST, and 100 L Streptavidin***8211;HRP
(SouthernBiotech) diluted 1:40,000 in THST buffer was added for 1 h at room temperature.
Plates were washed again 4 times with THST, and developed by addition of
100 L TMB substrate for 12 min. 100 L 0.5 M H2SO4 was added per well, and ELISA
plates were read at 450 nm with wavelength correction set at 540 nm. A calibration
curve was generated by plotting the OD against calibrator concentration and using
a 5-parameter logistic fit. The myostatin concentrations in human samples reported
in the manuscript were generated in glycine buffer at pH 2.5.
Calibration curve range. A calibration curve consisting of 2-fold dilutions of
recombinant human mature myostatin in Belgian Blue serum extending from 0.07
to 75.00 ng/mL was prepared. The intra- and inter-assay imprecision of read-back
values for the 11 calibrators were determined from 6 analytical runs. The mean, S.D.,
CV, and bias of the extrapolated concentrations were calculated for each analytical
run (to assess intra-assay imprecision) and for all analytical runs (to assess interassay
imprecision). The lower and upper limits of quantitation were defined as the
lowest (LLQ) and highest (ULQ) calibrator concentrations that could be measured
with an intra-assay CV and bias ***8804;30%.
Validation samples. Three sets of validation samples corresponding to low, midrange,
and high serum concentrations of myostatin were prepared using serum
samples from healthy subjects with endogenous myostatin concentrations in the
lower end and mid-range of the calibration curve, respectively. The high validation
sample was a serum sample from a healthy subject spiked with recombinant
myostatin protein.
Intra- and inter-assay imprecision. Intra- and inter-assay CVs were measured
in six separate aliquots of low, mid, and high validation samples in five independent
analytical runs. The QC analytical run acceptance range (total assay variation
mean ± 2 S.D.) was determined from myostatin concentrations measured in ***8805;3 separate
aliquots of each of the three QC samples in ***8805;5 independent analytical runs.
One aliquot of each of the three QC samples was analyzed in each analytical run of
samples. An analytical run was accepted if the measured myostatin concentration
in two out of three QC samples was within acceptance range.
2.2. Other assays
Serum total testosterone levels were measured by a specific radioimmunoassay
that has been validated previously against liquid chromatography***8211;tandem
28 K.M. Lakshman et al. / Molecular and Cellular Endocrinology 302 (2009) 26***8211;32
mass spectrometry (LC***8211;MS/MS) (Bhasin et al., 2005). The intra- and inter-assay
CVs for total testosterone assay were 8.2% and 13.2%, respectively. Free testosterone,
separated from serum by an equilibrium dialysis procedure, was measured
by a sensitive radioimmunoassay that had a sensitivity of 8 pmol/L, and intraand
inter-assay CVs 4.2% and 12.3%, respectively (Sinha-Hikim et al., 1998). The
radioimmunoassay and LC***8211;MS/MS methods were compared by analyzing samples
prepared in charcoal stripped serum to which known amounts of testosterone
had been added. These measurements demonstrated a correlation of 0.99 between
the radioimmunoassay and LC***8211;MS/MS measurement. Serum sex hormone binding
gobulin (SHBG) levels were measured by an immunofluorometric assay that has
a sensitivity of 6.25 nmol/L. Body composition was assessed at baseline and during
week 20 by dual-energy X-ray absorptiometry (DXA, Hologic 4500, Waltham,
MA). A body composition phantom was used to calibrate the machine before each
measurement.
2.3. Statistical analyses
All outcome variables were evaluated for distribution and homogeneity of variance;
variables that did not meet the assumptions of homogeneity of variance or
normal distribution were log-transformed. ANOVA was used to evaluate differences
across dose groups stratified by age, younger vs. older, at a single time point; Tukey***8217;s
multiple comparison test was used to determine which groups differed significantly
if a difference was identified by ANOVA. Changes within groups from baseline to
treatment were evaluated by using paired t-tests. Alpha was set at 0.05 for determining
statistical significance. Data are presented as mean ± S.E.M. or mean percent
change from baseline ± S.E.M., unless otherwise indicated.
3. Results
3.1. Myostatin assay characteristics
Linear range. The mean intra- and inter-assay imprecision was
determined from 6 analytical runs for each of the 11 calibrators
(0.07***8211;75.00 ng/mL) in the standard curve. The inter-assay CV for
the 0.07 ng/mL calibrator was 36.4%, which exceeded the acceptable
limit (<30%). Therefore, the LLQ of the assay was determined
by the next calibrator point (0.15 ng/mL) at which the inter-assay
CV and bias were 19.7% and +3.4%, respectively. The inter-assay
CV of 32.4% for the 75.00 ng/mL calibrator was also not within
the acceptable limit (<30%), thereby defining the ULQ to the next
calibrator point (37.50 ng/mL) at which the CV and bias were
3.6% and +0.8%, respectively. Thus, the quantitative range of the
assay extended from 0.15 to 37.50 ng/mL in a biologically relevant
matrix.
Intra- and inter-assay imprecision. The mean myostatin concentrations
(ng/mL, ±S.D.) in the low, mid, and high validation samples
in five analytical runs were 3.70 ± 0.15, 7.60 ± 0.13, and 18.27 ± 0.95,
respectively. The measured intra-assay CV for the low, mid, and high
validation samples (n = 5 for each validation sample) was 4.1%, 4.7%,
and 7.2%, respectively, and the inter-assay CV was 3.9%, 1.6%, and
5.2%, respectively.
Assay specificity. The mature myostatin protein has a high degree
of sequence conservation among mammalian species (McPherron
and Lee, 1997), allowing use of the myostatin assay on many nonhuman
samples, includingmouse, rat, dog, cow, andmonkey. Serum
samples from myostatin-deficient cattle (Belgian Blue) and from
mice with an inactivating mutation in the myostatin gene (mstn KO)
were assayed under dissociative conditions, and compared to normal
animals of the same species. Serum myostatin concentration
in wild type mice averaged 113.00 ± 20.70 ng/mL (mean ± S.E.M.),
and in normal cows 41.30 ± 1.10 ng/mL (mean ± S.E.M.). In contrast,
myostatin protein was undetectable in sera of Belgian Blue cattle
and the myostatin KO mice, confirming the specificity of the assay;
mutations in these myostatin-null animals abolish the synthesis
of myostatin protein (McPherron et al., 1997; McPherron and Lee,
1997).
Evaluation of binding protein dissociation after sample acidification.
Acidification of serum samples dissociates specific myostatin
binding proteins such as the myostatin propeptide and follistatin
resulting in activation of receptor binding and signaling activity
Table 1
Baseline characteristics of the subjects (mean ± S.D.).
Characteristic Young men (n = 50) Older men (n = 48)
Age (years) 26.5 ± 4.6 66.4 ± 4.7
Height (cm) 176.3 ± 6.4 175.9 ± 5.7
Weight (kg) 75.1 ± 10.9 83.2 ± 11.7
BMI (kg/m2) 24.1 ± 3.0 26.9 ± 3.5
Lean body mass (kg) 57.6 ± 7.2 57.9 ± 6.3
Percent fat mass (%) 18.0 ± 6.4 26.6 ± 5.4
Total testosterone (ng/dL) 578.4 ± 165.2 330.6 ± 96.1
(Zimmers et al., 2002). The anti-myostatin mAb MYO-029 and the
soluble myostatin receptor protein ActRIIB-Fc are both capable of
neutralizing myostatin activity, and both bind to regions of myostatin
that overlap with epitopes for RK22 and RK35 antibodies.
Therefore, addition of these myostatin-binding proteins to serum
would be expected to compete with the ELISA antibodies and block
the myostatin-specific signal in the assay. The RK35 capture antibody
is capable of binding to myostatin at both pH 2.5 and pH 8,
while MYO-029 and ActRIIB-Fc binding to myostatin is abolished
at pH 2.5 (data not shown). Increasing concentrations (0, 3, 30,
and 300 g/mL) of either MYO-029 or ActRIIB-Fc were added to
non-human primate serum and assayed in myostatin ELISA at pH 8
(non-dissociative conditions, Fig. 1B) and at pH 2.5 (dissociative
conditions, Fig. 1C). The standard curves under dissociative and
non-dissociative conditions are shown in Fig. 1A. Without addition
of binding proteins, the apparent myostatin concentration in
the serum at pH 8 was 0.8 ng/mL; under dissociative conditions
at pH 2.5 the measured myostatin concentration increased more
than 7-fold to 6 ng/mL. Addition of up to 300 g/mL of either MYO-
029 or ActRIIB-Fc had no effect on myostatin measurements when
assayed at pH 2.5; however, when assayed at pH 8 the presence of
these binding proteins diminished the myostatin signal in a dosedependent
manner to values approaching the LLQ of the assay,
providing further evidence that most or all of the signal in the ELISA
reflects myostatin protein concentration. Studies in human serum
provided similar results (data not shown).
To further characterize the effect of pH on assay performance,
we diluted human serum with glycine buffer ranging from pH 6.0 to
0.5. Myostatin levels were estimated based on a calibrator curve of
purified rGDF-8 dimer spiked into Belgian Blue serum. We obtained
stable myostatin concentrations in the pH range 3.0***8211;2.0 (data not
shown). The measured myostatin immunoreactivity was lower at
pH below 2.0 and above pH 3.5. Thus, myostatin binds the capture
antibody RK35 robustly at pH 2.5.
3.2. Baseline characteristics of human subjects
The baseline characteristics of the young and healthy men in the
parent study have been described (Bhasin et al., 2001, 2005). Fiftytwo
of the 61 randomized young men and 51 of 60 randomized
older men completed the treatment phase. The causes of treatment
discontinuation and loss to follow up have been described
(Bhasin et al., 2001, 2005). Sufficient serum for myostatin assays
and body composition data were available through week 20 for 50
young men and 48 older men; these subjects were included in this
secondary analysis and their baseline characteristics are shown in
Table 1. Drug compliance rate was >99%.
Baseline total and free testosterone, percent free testosterone,
and SHBG concentrations, did not differ among the five groups
at baseline in either the young or older groups. However, older
men had lower total and free testosterone, and higher SHBG than
younger men. Body weight, body mass index, and percent fat mass
were greater in the older men than younger men, while height was
similar in both.
K.M. Lakshman et al. / Molecular and Cellular Endocrinology 302 (2009) 26***8211;32 29
Fig. 1. (A) Serum myostatin calibrator curves under dissociative and non-dissociative conditions. Calibration curves were prepared by spiking recombinant myostatin into
serum, and then serially diluting with THST +1% BSA buffer (green line) or Belgian Blue serum (blue line). Each calibrator curve was generated in either THST buffer pH 8.0 or
glycine buffer pH 2.5 for non-dissociative or dissociating conditions, respectively. Standard curves of recombinant myostatin standard in buffer (not serum) and Belgian Blue
serum at pH 8 are different, presumably due to the interference in the assay by myostatin-binding proteins in the Belgian Blue serum. Standard curves at pH 2.5 shows very
little difference between buffer and Belgian Blue serum, and the lower limit of quantitation is lower at this pH. Each data point represents mean ± S.E.M. of three replicates.
(B) and (C) Myostatin levels in cynomolgous monkey serum measured in the myostatin ELISA under non-dissociative (pH 8, panel (B)) or dissociative (pH 2.5, panel (C))
conditions following addition of increasing concentrations of anti-myostatin antibody MYO-029 or soluble myostatin receptor ActRIIB-Fc. Bars represent mean ± S.D. of three
replicate samples. Dashed line indicates the LLQ. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)
3.3. Myostatin levels in young and older men
Serum myostatin levels were normally distributed in both young
and older men. Young men had significantly higher myostatin levels
than older men (8.0 ± 0.3 ng/mL vs. 7.0 ± 0.4 ng/mL, P = 0.03)
(Table 2A). Serum myostatin levels were not significantly correlated
with lean body mass in either young or older men (Fig. 2A and B).
Table 2A
Serum baseline myostatin levels in young and older men.
Group Myostatin (ng/mL)
Mean ± S.D. Median 25th percentile 75th percentile
Young men (n = 50) 8.0 ± 2.3a 7.8 6.5 9.8
Older men (n = 48) 7.0 ± 2.5a 6.8 5.7 8.5
a P = 0.03 for t-test
Similarly, there was no significant correlation between myostatin
levels and body weight, body mass index, or testosterone levels at
baseline (not shown).
3.4. Myostatin levels in women
Serum myostatin levels in young women were not significantly
different from those in young men. Myostatin levels in menstruating
women, surgically menopausal and naturally menopausal
women did not differ significantly (Table 2B).
3.5. Effects of testosterone administration on myostatin levels in
men
Serum myostatin levels at baseline did not differ significantly
across the five dose groups within either young or older men. Serum
30 K.M. Lakshman et al. / Molecular and Cellular Endocrinology 302 (2009) 26***8211;32
Fig. 2. (A) Regression plot showing correlation of baseline myostatin levels with
lean body mass in young men. (B) Regression plot showing correlation of baseline
myostatin levels with lean body mass in older men.
Table 2B
Serum myostatin levels in young menstruating women, surgically menopausal
women, and in older women.
Group Myostatin (ng/mL)
Mean ± S.D. Median 25th
percentile
75th
percentile
Young menstruating
women (n = 33)
7.0 ± 2.7a 6.1 5.0 9.2
Surgically menopausal
women (n = 37)
6.7 ± 2.7a 6.2 5.5 7.4
Older women (n = 24) 6.7 ± 2.8a 6.2 5.1 8.2
a P = 0.86 for ANOVA.
myostatin levels were significantly higher on day 56 compared to
baseline in both young and older men (Fig. 3A). Older men experienced
a significantly greater percent increase in myostatin levels
than young men (Fig. 3B). The increases in myostatin levels during
testosterone therapy were not sustained; thus, serum myostatin
levels on day 140 were not significantly different from those at
baseline.
Changes in myostatin levels from baseline to day 56 were
significantly positively correlated with changes in total (Fig. 4A)
and free (Fig. 4C) testosterone concentrations in young men,
but not in older men (Fig. 4B and D). As previously reported,
testosterone treatment was associated with significant gains in
lean body mass; the changes in lean body mass were significantly
correlated with testosterone dose and testosterone concentration
(Bhasin et al., 2001, 2005). However, changes in lean body mass
Fig. 3. Changes in myostatin levels in young men in response to administration of graded doses of testosterone (bar diagram showing mean ± S.E.M. levels at baseline, and
days 56 and 140). Panel (A) shows the myostatin levels at baseline, treatment day 56, and 140 in young (left panel) and older men (right panel). The data are mean ± S.E.M. *P
value as in comparison to baseline levels. Myostatin levels on day 140 were not significantly different from baseline levels. Panel (B) shows the percent change from baseline
in serum myostatin levels from baseline to day 56 in young and older men. *P = 0.03.
K.M. Lakshman et al. / Molecular and Cellular Endocrinology 302 (2009) 26***8211;32 31
Fig. 4. Regression plots showing correlation of the change in myostatin levels from baseline to day 56 and changes in total and free testosterone concentrations and lean
body mass in young and older men. Panel (A) shows the linear regression plot of percent change in myostatin levels from baseline to day 56 and percent change in serum
total testosterone concentrations in young men. Panel (B) shows the linear regression plot of percent change in myostatin levels from baseline to day 56 and percent change
in serum total testosterone concentrations in older men. Panel (C) shows the linear regression of percent change in myostatin levels from baseline to day 56 and percent
change in serum free testosterone concentrations in young men. Panel (D) shows the linear regression of percent change in myostatin levels from baseline to day 56 and
percent change in serum free testosterone concentrations in older men. Panel (E) shows the linear regression of percent change in myostatin levels from baseline to day 56
and percent change in lean body mass from baseline to day 140 in young men. Panel (F) shows the linear regression of percent change in myostatin levels from baseline to
day 56 and percent change in serum lean body mass from baseline to day 140 in older men.
were not significantly correlated with either absolute or percent
change (Fig. 4E and F) in myostatin concentrations.
4. Discussion
Previous data on circulating myostatin levels in healthy individuals
and in individuals with disease are confounded by issues of
specificity and the inability of these earlier assays to reliably take
into account the binding of circulating myostatin to its binding proteins
in plasma. In our assay, myostatin was stripped off its binding
proteins by acid treatment that effectively dissociates myostatin
from the binding proteins. Other serum proteins do not have significant
cross-reactivity in our assays; even the highly homologous
GDF-11 protein is not expected to cross-react in the assay because
the RK-22 detector antibody in the sandwich ELISA does not bind
to GDF-11. The assay has all the characteristics of a valid measurement
system: nearly 100% recovery of spiked myostatin, excellent
precision in the physiologic range, and accuracy. The assay also has
sufficient sensitivity to be able to measure circulating myostatin
concentrations in almost all men and women.
Using this validated myostatin assay, we report here the distribution
of myostatin levels in healthy men and women. Although
direct comparisons with previous assays are difficult because of
the differences in calibrating standard, we find that the circulating
concentrations in men and women are substantially lower
than those reported previously. We show that myostatin levels are
lower in older men than young men. Also, contrary to expectation,
myostatin levels increased transiently in response to testosterone
32 K.M. Lakshman et al. / Molecular and Cellular Endocrinology 302 (2009) 26***8211;32
administration, but returned to baseline by 20 weeks of treatment.
The increments in myostatin levels were correlated with
circulating testosterone concentrations. These observations support
the hypothesis initially proposed by Lee (2004) and later by
Gaussin and Depre (2005) that myostatin acts as a chalone ***8211; a
counter-regulatory hormone ***8211; to restrain skeletal muscle growth
in response to an anabolic stimulus.
The concept of chalones ***8211; inhibitors of cell growth that serve as
counter-regulatory mechanisms to control the size of specific tissues
***8211; was introduced initially by Bullough. However, Lee (2004)
was the first to articulate the hypothesis that myostatin functions
as a chalone for skeletal muscle: it is produced and secreted by
skeletal muscle and it circulates in plasma to restrain skeletal muscle
mass. Our data support Lee***8217;s prescient prediction; myostatin
levels are increased in response to the increase in skeletal muscle
mass induced by testosterone administration. The increments in
myostatin levels were correlated with testosterone concentrations.
Thus, testosterone administration increases muscle mass resulting
in increased myostatin production and secretion; it is possible that
the increased circulating myostatin levels in turn restrain unlimited
growth of skeletal muscle in response to continued testosterone
administration. Similarly, older men with lower skeletal muscle
mass have lower myostatin levels than young men; one could speculate
that aging is associated with loss of skeletal muscle mass,
leading to decreased myostatin secretion, which in turn brakes further
muscle loss. In separate studies, Shyu et al. (2005) reported that
cyclic mechanical stretch upregulates IGF-1 as well as myostatin
expression. The stretch-induced myostatin increase in cardiomyocytes
is mediated by IGF-1 in part through MAP kinase and MEF2
pathway. In an accompanying editorial, Gaussin and Depre (2005)
suggested that myostatin represents a chalone of IGF-1 pathway in
the heart. IGF-1 induces cardiac muscle hypertrophy resulting in
increased myostatin production that then checks further hypertrophy
of the cardiac muscle.
The availability of a reliable assay for accurate measurement
of myostatin levels provides an excellent opportunity to examine
the physiologic regulation of circulating myostatin levels in healthy
humans and in patients with clinical disorders associated with loss
of skeletal muscle mass. In this study, under basal steady-state conditions,
serum myostatin levels were not correlated with lean body
mass; thus,myostatin levelsmay not be a good biomarker of skeletal
muscle mass. However, observations that myostatin levels rise early
during the course of testosterone-induced muscle mass accretion
raise the possibility that myostatin levels might serve as a useful
early biomarker for the anabolic effects of promyogenic therapies
such as testosterone. This speculation needs further investigation.
Acknowledgements
We thank Neil Wolfman, Riyez Karim, John Nowak, and Paul
Yaworsky (Wyeth Research) for their contributions to this work,
and the staff of the Harbor-UCLA GCRC.
References
Anderson, S.B., Goldberg, A.L., et al., 2008. Identification of a novel pool of extracellular
pro-myostatin in skeletal muscle. J. Biol. Chem. 283 (11), 7027***8211;7035.
Bhasin, S., Woodhouse, L., et al., 2001. Testosterone dose***8211;response relationships in
healthy young men. Am. J. Physiol. Endocrinol. Metab. 281 (6), E1172***8211;E1181.
Bhasin, S., Woodhouse, L., et al., 2005. Older men are as responsive as young men
to the anabolic effects of graded doses of testosterone on the skeletal muscle. J.
Clin. Endocrinol. Metab. 90 (2), 678***8211;688.
del Re, E., Sidis, Y., et al., 2004. Reconstitution and analysis of soluble inhibin
and activin receptor complexes in a cell-free system. J. Biol. Chem. 279 (51),
53126***8211;53135.
Gaussin, V., Depre, C., 2005. Myostatin, the cardiac chalone of insulin-like growth
factor-1. Cardiovasc. Res. 68 (3), 347***8211;349.
Girgenrath, S., Song, K., et al., 2005. Loss of myostatin expression alters fiber-type
distribution and expression of myosin heavy chain isoforms in slow- and fasttype
skeletal muscle. Muscle Nerve 31 (1), 34***8211;40.
Gonzalez-Cadavid, N.F., Bhasin, S., 2004. Role of myostatin in metabolism. Curr. Opin.
Clin. Nutr. Metab. Care 7 (4), 451***8211;457.
Gonzalez-Cadavid, N.F., Taylor, W.E., et al., 1998. Organization of the human myostatin
gene and expression in healthy men and HIV-infected men with muscle
wasting. Proc. Natl. Acad. Sci. U.S.A. 95 (25), 14938***8211;14943.
Grobet, L., Martin, L.J., et al., 1997. A deletion in the bovine myostatin gene causes
the double-muscled phenotype in cattle. Nat. Genet. 17, 71***8211;74.
Hill, J.J., Davies,M.V., et al., 2002. Themyostatin propeptide and the follistatin-related
gene are inhibitory binding proteins of myostatin in normal serum. J. Biol. Chem.
277 (43), 40735***8211;40741.
Hill, J.J., Qiu, Y., et al., 2003. Regulation of myostatin in vivo by growth and differntiation
factor-associated serum protein-1: a novel protein with protease inhibitor
and follistatin domains. Mol. Endocrinol. 17, 1144***8211;1154.
Hosoyama, T., Yamanouchi, K., et al., 2006. Role of serum myostatin during the
lactation period. J. Reprod. Dev. 52 (4), 469***8211;478.
Kambadur, R., Sharma, M., et al., 1997. Mutations in myostatin in double muscled
Belgian Blue and Piedmontese cattle. Genome Res. 7, 910***8211;916.
Kim, J.S., Cross, J.M., et al., 2005. Impact of resistance loading on myostatin expression
and cell cycle regulation in young and older men and women. Am. J. Physiol.
Endocrinol. Metab. 288 (6), E1110***8211;E1119.
Lee, S.J., 2004. Regulation of muscle mass by myostatin. Annu. Rev. Cell Dev. Biol. 20,
61***8211;86.
Lee, S.J., 2008. Genetic analysis of the role of proteolysis in the activation of latent
myostatin. PLoS ONE 3 (2), e1628.
Lee, S.J., Reed, L.A., et al., 2005. Regulation of muscle growth by multiple ligands
signaling through activin type II receptors. Proc. Natl. Acad. Sci. U.S.A. 102 (50),
18117***8211;18122.
McPherron, A.C., Lawler, A.M., et al., 1997. Regulation of skeletal muscle mass in mice
by a new TGF-beta superfamily member. Nature 387 (6628), 83***8211;90.
McPherron, A.C., Lee, S.J., 1997. Double muscling in cattle due to mutations in the
mypostatin gene. Proc. Natl. Acad. Sci. U.S.A. 94, 12457***8211;12461.
Mosher, D.S., Quignon, P., et al., 2007. A mutation in the myostatin gene increases
muscle mass and enhances racing performance in heterozygote dogs. PLoS
Genet. 3 (5), e79.
Nakatani, M., Takehara, Y., et al., 2008. Transgenic expression of a myostatin inhibitor
derived from follistatin increases skeletal muscle mass and ameliorates dystrophic
pathology in mdx mice. FASEB J. 22 (2), 477***8211;487.
Raue, U., Slivka, D., et al., 2006. Myogenic gene expression at rest and after a bout
of resistance exercise in young (18***8211;30 yr) and old (80***8211;89 yr) women. J. Appl.
Physiol. 101 (1), 53***8211;59.
Reisz-Porszasz, S., Bhasin, S., et al., 2003. Lower skeletal muscle mass in male transgenic
mice with muscle-specific overexpression of myostatin. Am. J. Physiol.
Endocrinol. Metab. 285 (4), E876***8211;E888.
Schuelke, M., Wagner, K.R., et al., 2004. Myostatin mutation associated with gross
muscle hypertrophy in a child. N. Engl. J. Med. 350, 2682***8211;2688.
Shyu, K.G., Ko, W.H., et al., 2005. Insulin-like growth factor-1 mediates stretchinduced
upregulation of myostatin expression in neonatal rat cardiomyocytes.
Cardiovasc. Res. 68 (3), 405***8211;414.
Sinha-Hikim, I., Arver, S., et al., 1998. The use of a sensitive equilibrium dialysis
method for the measurement of free testosterone levels in healthy,
cycling women and in human immunodeficiency virus-infected women. J. Clin.
Endocrinol. Metab. 83 (4), 1312***8211;1318.
Storer, T.W., Magliano, L., et al., 2003. Testosterone dose-dependently increases maximal
voluntary strength and leg power, but does not affect fatigability or specific
tension. J. Clin. Endocrinol. Metab. 88 (4), 1478***8211;1485.
Wagner, K.R., Fleckenstein, J.L., et al., 2008. A phase I/IItrial of MYO-029 in adult
subjects with muscular dystrophy. Ann. Neurol. 63 (5), 561***8211;571.
Walker, K.S., Kambadur, R., et al., 2004. Resistance training alters plasma
myostatin but not IGF-1 in healthy men. Med. Sci. Sports Exerc. 36 (5),
787***8211;793.
Welle, S., Bhatt, K., et al., 2007. Muscle growth after postdevelopmental myostatin
gene knockout. Am. J. Physiol. Endocrinol. Metab. 292 (4), E985***8211;E991.
Welle, S., Bhatt, K., et al., 2002. Insulin-like growth factor-1 and myostatin mRNA
expression in muscle: comparison between 62***8211;77 and 21***8211;31 yr old men. Exp.
Gerontol. 37 (6), 833***8211;839.
Whittemore, L.A., Song, K., et al., 2003. Inhibition of myostatin in adult mice increases
skeletal muscle mass and strength. Biochem. Biophys. Res. Commun. 300 (4),
965***8211;971.
Wolfman, N.M., McPherron, A.C., et al., 2003. Activation of latent myostatin by the
BMP-1/tolloid family of metalloproteinases. Proc. Natl. Acad. Sci. U.S.A. 100 (26),
15842***8211;15846.
Yang, J., Ratovitski, T., et al., 2001. Expression of myostatin pro domain results in
muscular transgenic mice. Mol. Reprod. Dev. 60 (3), 351***8211;361.
Zachwieja, J.J., Smith, S.R., et al., 1999. Plasma myostatin-immunoreactive protein is
increased after prolonged bed rest with low-dose T3 administration. J. Gravit.
Physiol. 6 (2), 11***8211;15.
Zhu, X., Hadhazy, M., et al., 2000. Dominant negative myostatin produces hypertrophy
without hyperplasia in muscle. FEBS Lett. 474 (1), 71***8211;75.
Zimmers, T.A., Davies, M.V., et al., 2002. Induction of cachexia in mice by systemically
administered myostatin. Science 296 (5572), 1486***8211;1488.
