Medical Research Council (MRC) Muscle and Cell Motility
Unit and MRC Centre for Developmental Neurobiology, Guy's Campus,
King's College London, London SE1 1UL, United Kingdom
To investigate the cause of skeletal
muscle weakening during aging we examined the sequence of cellular
changes in murine muscles. Satellite cells isolated from single muscle
fibers terminally differentiate progressively less well with increasing
age of donor. This change is detected before decline in satellite cell
numbers and all histological changes examined here. In
MSVski transgenic mice, which show type IIb fiber
hypertrophy, initial muscle weakness is followed by muscle degeneration
in the first year of life. This degeneration is accompanied by a
spectrum of changes typical of normal muscle aging and a more marked
decline in satellite cell differentiation efficiency. On a
myoD-null genetic background, in which satellite cell
differentiation is defective, the MSVski muscle
phenotype is aggravated. This suggests that, on a wild-type genetic
background, satellite cells are capable of repairing
MSVski fibers and preserving muscle integrity in early
life. We propose that decline in myogenic cell differentiation
efficiency is an early event in aging-related loss of muscle function,
both in normal aging and in some late-onset muscle degenerative conditions.
 |
INTRODUCTION |
AGING HUMAN
SKELETAL MUSCLE becomes weaker and undergoes atrophy, changes
that frequently lead to physical dependence in the elderly. Muscle
aging atrophy results from the combined effects of loss of muscle
fibers and atrophy of the remaining fibers (2, 4, 28, 31, 40, 44,
51). Decline in fiber number is paralleled by the loss of motor
neurons occurring throughout life (39). Consequently,
because each motor neuron is thought to have a limited innervation
capacity, useful reversal of fiber number decline may be difficult. In
contrast, various interventions can reverse muscle atrophy in aging
humans or rodents by enhancing fiber size. Exercise is effective at
preserving human muscle mass and function in the elderly (24,
25). In rodents, overexpression of IGF-I in muscle has also been
shown to enhance fiber size, mitigating muscle atrophy and weakness
(4, 16). Nevertheless, muscle mass declines with age even
in individuals who exercise, consistent with the view that although
factors such as neural or hormonal changes contribute to muscle aging
(11, 37, 39), changes in muscle tissue itself, such as
reduced force, enhanced susceptibility to injury, and poor repair
capacity, are important (3, 8, 10, 21, 27, 35, 41, 55, 66,
67). In particular, in pathological situations, such as Duchenne
muscular dystrophy, in which muscle function is gradually lost,
enhanced cellular turnover has been implicated in exacerbation of the
deficit and control of disease progression (20, 73).
However, the causal links between the various cellular changes leading
to muscle weakness and atrophy in aging or disease are unclear.
The cellular mechanisms underlying fiber atrophy are poorly understood.
Fiber size consists of two elements, the number of nuclei inside each
multinucleated muscle fiber and the volume of cytoplasm supported by
each nucleus, hereafter referred to as the nuclear domain size
(35). Both elements vary among fibers of different type,
during fiber development and in response to changes in innervation
(23, 65). During development, ablation of the transforming
growth factor (TGF)-
superfamily member growth and differentiation
factor (GDF)-8 leads to increased nuclear number and fiber size
(48, 76). Conversely, ablation of IGF-I leads to
decreased muscle mass (56). Whether these pathways only
control nuclear number or also affect domain size is unclear. To
understand the long-term effects of increase in domain size in the
absence of change in nuclear number on muscle integrity and function
and their consequences for aging, we examined transgenic mice
overexpressing a truncated form of the Ski protooncogene, a histone
deacetylase complex component that has been implicated in various
signaling pathways, including that activated by TGF- superfamily
members (1, 46, 52, 71). MSVski transgenic mouse muscle hypertrophies without increase in fiber or myonuclear number (69). Expression levels of the transgene are
highest in fast muscles, and hypertrophy specifically occurs in fast
type IIb fibers (43). A similar, but poorly characterized,
hypertrophic muscle phenotype occurs in MSVski
transgenic cattle, leading to death within 10-15 wks of birth
accompanied by muscle weakness and elevated serum creatine
phosphokinase, a marker of muscle damage (9). Other
studies have demonstrated muscle damage after hypertrophy
(13), and these findings, together with the observations made in the transgenic cattle, suggest that muscle nuclear domain enlargement may be detrimental to muscle integrity.
We report that MSVski-induced nuclear domain enlargement
is accompanied by loss of force but no other detectable changes in young mice. However, over time, muscle degeneration becomes apparent with defective regeneration and other signs typical of aggravated muscle aging. The premature aginglike changes in MSVski
muscle are preceded by decreased satellite cell differentiation
potential, a change similar to that observed in aging wild-type mice.
The data raise the possibility that long-term muscle nuclear domain enlargement may have deleterious effects in later life and that decreased satellite cell differentiation efficiency may contribute to
the changes occurring during muscle aging.
| |
MATERIALS AND METHODS |
Generation of transgenic mice.
Fertilized mouse eggs were injected with the FB29 MSVski
transgene encoding a truncated version of chicken c-ski
protooncogene driven by the murine sarcoma virus LTR as previously
described (69). Three of four founder mice carrying the
MSVski transgene identified by Southern blot analysis
transmitted the transgene to F1 progeny, termed lines 6, 16, and 33. Transgenic mice were backcrossed onto a CBA/J background
for at least four generations before further analysis or crossing.
Northern blot analysis of skeletal muscle total RNA from each line at
postnatal day 60 (P60) showed that line 33, the highest expresser,
accumulated ~3.3-fold more MSVski mRNA than line 6, the lowest expresser, and that MSVski mRNA was
undetectable in soleus and high in extensor digitorum longus (EDL), as
described previously (43, 69). MSVski line 16 heterozygous mice were bred to homozygous
myoDm1 (hereafter referred to as
myoD
/) mice (59).
F1 progeny heterozygous for MSVski and
myoD/ were backcrossed with homozygous
myoD/ mice, and progeny were genotyped for
MSVski by Southern blot analysis and for myoD
by PCR on tail DNA. Mice were kept in plastic cages with wire mesh lids
in a 12:12-h light-dark cycle and fed ad libitum.
Histology, immunocytochemistry, and morphometry.
For morphometric analysis, full-length EDL muscle was isolated, clamped
at rest length, and frozen in isopentane cooled in liquid nitrogen. For
total fiber number and fiber type proportion, whole lower hind legs
were frozen as above and fiber counts were made on the entire
transverse sections of the midbelly EDL. Ten-micrometer serial
cryosections were stained with myosin heavy chain (MyHC) antibodies
specific for MyHC IIb (BF-F3), MyHC IIa (A4.74), embryonic fast MyHC
(F1.652), and slow -cardiac MyHC (A4.951) (47, 64) as
described previously (33). Total nuclei counts were made on BF-F3-stained EDL sections after DNA staining with
4',6- diamidino-2-phenylindole (DAPI) on areas containing ~50
fibers. Cross-sectional areas (CSAs) of at least 200 IIb and non-IIb
fibers contained within the same area were determined by outlining
fibers manually in NIH Image captured with a Pulnix TM-765E camera
attached to a Zeiss Axiophot. To ensure transverse sectioning, images
with fiber profiles showing no significant major axis orientation were analyzed.
Northern blot analysis.
Total RNA was extracted from skeletal muscles of a single upper leg
with a single-step method (15). Approximately 30 µg of
total RNA were separated on 1% agarose gels containing 0.41 M
formaldehyde in 1× MOPS buffer, transferred to a nylon membrane, and
probed with a 32P random-primed MSVski cDNA
and human -actin cDNA (which readily cross-reacts to mouse
-actin) as a loading control.
Nuclear turnover and creatine kinase assay.
Mice were injected intraperitoneally with 30 mg/kg body wt of
5-bromo-2'-deoxyuridine (BrdU) and 100 mg/kg Evans blue dye in saline.
After 17 h, lower hindlimbs were collected and cryosectioned as
above. BrdU staining of acid-treated sections followed
(72). Terminal deoxynucleotidyl transferase-mediated nick
end-labeling (TUNEL) staining on fixed cryosections used the Apoptag
kit (Invitrogen). Serum was prepared by coagulation and centrifugation
of blood samples, stored frozen, and assayed with a Boehringer Mannheim creatine kinase (CK) NAC-activated kit using UV detection.
Satellite cell cultures derived from single muscle fibers.
Single fibers dissociated by collagenase treatment of EDL muscle of
MSVski or wild-type mice of various ages were plated, and outgrowing mononucleate cells were analyzed as described previously (57). Fibers derived from at least two mice of each
genotype and age group were compared. Single EDL fibers were
individually plated on Matrigel (Becton Dickinson)-coated wells of
Permanox chamber slides incubated in a humid environment at 37°C and
5% CO2. After plating, fibers were incubated for 3 days in
plating medium consisting of 10% (vol/vol) horse serum (GIBCO) and
0.5% chick embryo extract (CEE, Imperial) in DMEM containing 2%
L-glutamine and 1% penicillin and streptomycin (GIBCO). At
day 3, fibers were removed and medium was replaced with a
rich medium containing 20% FCS (GIBCO), 10% horse serum, and 2% CEE
in DMEM to promote cell growth. After a further 2 days, cultures were
differentiated for 2 days in a medium consisting of 2% FBS in DMEM.
Cells were rinsed in PBS, fixed in cold methanol, blocked in 5% horse
serum in PBS, and incubated with MAb against desmin (1/500, clone no. DE-U-10, IgG1; Sigma) and MyHC (1/10, A4.1025, IgG2a; Ref.
19). Primary antibodies were successively detected with
rat anti-mouse IgG1 (1:1,000; Serotec), FITC-conjugated goat anti-mouse
IgG2a (1:100; Serotec), and Cy3-conjugated donkey anti-rat IgG (1:100; Jackson). In the last antibody incubation, the DNA dye DAPI was added.
Cells were fixed in cold methanol before being mounted with antifading
agent. At day 3 after plating and after 2 days in growth
medium, the total number of cells was analyzed for each single-fiber
culture. After two additional days in differentiation medium, a
representative number of cells were analyzed by randomly analyzing 10 fields of view equivalent to a total area of 5.45 mm2. In
this study, differentiation efficiency is the ratio of nuclei associated with MyHC-desmin over the total number of nuclei associated with desmin.
Single permeabilized fiber isometric force measurement.
The protocol has been described by Sabido-David et al.
(61). Briefly, single-fiber 2- to 3-mm segments were
dissected from permeabilized EDL of P60 MSVski
heterozygote or wild-type littermates and aluminum T clips were crimped
to their ends and attached to a fixed hook and to a force transducer
(AE801; Aksjelskapet). Sarcomere lengths were adjusted to 2.4 µm, and
fiber CSAs were calculated by assuming a circular circumference. Peak
tension was recorded in maximally Ca-activated and Mg-rigor conditions, as was the time to reach 90% of maximum calcium activation
(t90) and relaxation to 50%
(t50).
Muscle DNA and protein contents.
Each weighed EDL muscle was homogenized with 1 ml of lysis buffer (0.1 M KPO4 pH 7.8, 0.2% Triton X-100, 1 mM DTT, 0.2 mM phenylmethylsulfonyl fluoride, and 5 µg/ml leupeptin). Protein concentrations were determined with the bicinchoninic acid (BCA) protein assay (Pierce). DNA concentrations were determined with a
modification of the Hoechst 33258 assay (26). Briefly, a
50-µl sample was diluted in 450 µl of DNA assay buffer (0.15 M
NaCl, 15 mM Na citrate) before incubation with 250 µl of Hoechst
33258 (0.8 mg/l in distilled H2O) for 10 min at room
temperature in the dark. Fluorescence was read with a
fluorometer (Spex; FluoroMax Instruments) at an excitation
wavelength of 360 nm and an emission wavelength of 450 nm, and DNA
concentration was determined from a standard curve generated with
salmon sperm DNA.
Statistical analysis.
Fiber CSA data were analyzed with a multiple ANOVA test. Data on fiber
type proportion, total fiber number, and nuclear number and data from
single permeabilized fiber force experiments were analyzed with an
unpaired t-test. Data from single-fiber cultures were
analyzed with the Kolmogorov-Smirnov two-sample (1-tailed) test and the
Spearman rank-order correlation coefficient rs.
| |
RESULTS |
MSVski IIb fiber size increase is not paralleled by
increased force generation.
To examine the effect of nuclear domain enlargement, we made
MSVski transgenic mice. Each of three independent lines
of such mice showed increases in CSA of postnatal fast muscle fibers
compared with control littermates. At P17, wild-type and
MSVski transgenic EDL IIb muscle fibers were similar
(Fig. 1, A-C,
J). By P60, IIb fibers in
transgenic animals had increased in size over twofold, from 1,325 ± 52 µm2 (n = 6) in controls to
3,042 ± 360 µm2 (n = 5) in
MSVski (Fig. 1, D-F, J). The
distribution of IIb fiber sizes in MSVski mice is skewed
toward larger size, with some very large fibers present (>6,000
µm2; Fig. 1F). In MSVski lines
16 and 33 the type IIb fiber size increase was paralleled by a doubling
of EDL wet mass and extractable protein but without a significant
increase in muscle DNA content or nuclear number (Fig.
2). Thus MSVski causes
approximately twofold nuclear domain enlargement, in all respects
confirming published results (43, 69). In our lines, in
contrast to previous reports, MSVski-induced hypertrophy
occurred without bone malformation or significant increase in body
mass, so hypertrophy is unlikely to be a response to increased load
(38). We also analyzed the ability of
MSVski fibers to generate force. Isometric force
measurements on single permeabilized fibers demonstrated that
hypertrophied MSVski fibers from P60 mice generate less
force per unit CSA than control fibers, in both maximally activated and
Mg-rigor conditions (Table 1 and data not
shown). However, the kinetics of activation (t90) and relaxation
(t50) between the MSVski and
control fibers remained similar (Table 1 and data not shown). The
reduction in specific force without change in kinetics of
MSVski EDL fibers suggests either that each active
myosin molecule produces less force or that fewer myosins are
functioning. Immunohistochemical fiber typing using MAbs to MyHC
isoforms revealed no differences between young adult control and
MSVski limbs (see below). Evans blue dye perfusion
failed to detect a change in fiber integrity: all fibers excluded the
dye in EDL muscles of P90 control and MSVski mice (data
not shown). Similarly, histological comparisons of P60
MSVski and control muscles showed no signs of
pathology (Figs. 1 and 3). However,
occasional isolated MSVski fibers had misalignment of
sarcomeres in adjacent myofibrils. Thus, although MSVski
fibers are larger than controls, they do not generate greater force.
View larger version (94K):
[in this window]
[in a new window]
|
Fig. 1.
MSVski-induced IIb fiber hypertrophy declines with
age. Transverse cryosections of extensor digitorum longus (EDL) from
wild-type (A, D, G) and
MSVski (B, E, H) mice
at postnatal day (P)17 (A, B), P60 (D,
E), and aging (P >550; G, H) stained
with MAb specific to type IIb myosin heavy chain (MyHC). In
MSVski mice IIb fibers are hypertrophied (arrows) from
P60, but with aging there is additional appearance of small IIb fiber
profiles (arrowhead). Bar, 50 µm. Distribution of mean ± SE IIb
fiber cross-sectional (CSA) at P17 (C), P60 (F),
and aging (I) in wild-type and MSVski mouse
EDL is shown (n = 3, 6, and 3 for wild type and 4, 5, and 5 for MSVski at P17, P60, and P >550,
respectively). With aging, there is a decrease in very large and an
increase in very small IIb fibers in MSVski mice.
J: average ± SE IIb fiber CSA with age in wild-type
and MSVski mouse EDL. Data are from C,
F, and I (**P < 0.01). On aging,
the mean IIb fiber size is comparable between wild-type and
MSVski mice. K: Northern analysis of total
RNA from skeletal muscle shows a 2.5-kb chicken ski
transcript from the MSVski transgene detectable from
P17, reaching high levels at P60 that are maintained in aging (P550)
mice. Actin loading control is shown at bottom.
|
|
View larger version (43K):
[in this window]
[in a new window]
|
Fig. 2.
Increase in nuclear domain size of adult MSVski
mice. A: muscle mass is doubled (left),
extractable DNA concentration is halved (center), and
extractable protein concentration is unaltered (right) in
P60 MSVski EDL compared with wild type. Values are
means ± SE (n = 4, 9 for P17 and
n = 5, 11 for P60 wild type and MSVski,
respectively). B: total nuclear numbers per fiber
[calculated by dividing number of 4',6-diamidino-2-phenylindole
(DAPI)-stained nuclei by number of fibers in an area of >50 fibers
without visible interstitial tissue] are little changed in P60
MSVski EDL. Values are means ± SE
(n = 1, 3 for line 33 and n = 4, 5 for
line 16 wild type and MSVski, respectively).
C: both IIb fiber CSA per nucleus (left) and
perimeter per nucleus (right) are significantly increased in
P60 MSVski EDL (P = 0.01 and 0.0004, respectively). Fiber area and perimeter of IIb fibers were measured in
each of several cryosections from EDL midbelly of 3 MSVski (265, 22, and 24 total fibers) and 2 wild-type
littermate (89 and 27 total fibers) mice with NIH Image. Values are
means ± SE (n = 116 and 311, wild type and
MSVski, respectively).
|
|
View larger version (183K):
[in this window]
[in a new window]
|
Fig. 3.
Nonperipheral myonuclei are abundant in aging
MSVski mice. A: hematoxylin and eosin (H & E)
staining on EDL transverse sections of wild-type (A,
C) or MSVski (B, D)
mice at P60 (A, B) or aging (C,
D) shows an increase in nonperipheral myonuclei (arrowheads)
in aging MSVski mice. Bar, 50 µm.
|
|
Aging MSVski mice undergo muscle damage and regeneration.
Histological examination of P60 MSVski mice showed no
signs of damaged fibers or nonperipheral myonuclei (Fig. 3). Control aging mice (defined as animals over 550 days of age) showed relatively few nonperipheral nuclei, although slightly more than at P60 (Fig. 3C). However, aging MSVski muscle revealed
that all kinds of fibers (large and small, IIb and non-IIb) have
abundant nonperipheral nuclei (>70% of fibers analyzed in some
sections; Fig. 3D), suggesting that regeneration had
occurred within the muscle. Infiltration by cells of the immune system
was not observed, nor were serum CK levels raised in mice of either
genotype or age group [CK levels were 632 ± 208, 352 ± 93, 477 ± 94, and 409 ± 101 U/l in P42-P106 control
(n = 6), MSVski expressing
(n = 6) and aging control (n = 2), and
MSVski expressing (n = 4),
respectively]. Therefore, MSVski mouse muscle has a
contractile deficit but apparently healthy histology in early life but
shows marked histological damage in aging animals.
Exacerbated aging-related changes in MSVski muscle.
Accompanying the damage in aging MSVski mice were a
series of changes typical of normal muscle aging but more severe.
Depending on genetic strain, mice begin to show aging changes late in
their second year of life and are often defined as "aged" over 2 yr of age. Average IIb fiber CSA begins to decline in aging CBA/J mice,
and this change was enhanced in aging MSVski: although
highly hypertrophied fibers (>6,000 µm2) were still
present, many smaller fiber profiles were also observed (0-1,199
µm2) (Fig. 1, G-I). The mean
MSVski IIb fiber size was greatly reduced to a value
similar to the wild-type level, despite continued expression of the
transgene (Fig. 1, J and K). These data suggest
that a long-term effect of MSVski is to promote the
appearance of small IIb fibers typical of muscle aging (2,
44).
Exacerbation of other changes typical of normal muscle aging was also
apparent in MSVski muscle (Fig.
4). Type IIb fibers become less frequent
compared with other fiber types as rodent muscle ages (2, 4,
40). In our control cohort this trend was not apparent, perhaps
because of strain differences in the onset of aging changes. In control
mice, ~65% of EDL fibers were of type IIb at all ages [61%
(n = 3), 66% (n = 4) and 69%
(n = 3) at P17, P60, and >P550, respectively]. In P17
and P60 MSVski mice, the EDL had a proportion of IIb
fibers comparable with that of control littermates [67%
(n = 3) and 60% (n = 5),
respectively]. However, in aging MSVski, IIb fibers
were less frequent than in controls (Fig. 4, E-G),
dropping to 48% (n = 3) from 69% in controls (P < 0.05). In place of IIb fibers, significantly
increased numbers of IIx and slow fibers were observed (20 to 38% for
IIx and 0.4 to 2.2% for slow in aging control and
MSVski, respectively), with little change in IIa fiber
frequency (10.9% in controls and 11.3% in MSVski)
(Fig. 4, A-D and G). These non-IIb fibers
were increased in size but were significantly smaller than normal IIb
fibers in wild-type animals (Figs. 4H and 1J). At
P17 and P60, when the non-IIb fiber type proportion was unchanged from
control, the non-IIb fiber size was also unaffected by the
MSVski transgene (Fig. 4H). One
interpretation of this spectrum of changes is that reduced IIb fiber
number combined with increased non-IIb fiber size and frequency
reflects a shift of some hypertrophied IIb fibers to a non-IIb
phenotype. Thus aging MSVski mice show an enhancement of
changes in fiber type proportion that occur during normal aging.
View larger version (97K):
[in this window]
[in a new window]
|
Fig. 4.
MSVski expression leads to increase in
non-IIb fiber type proportion and size in aging mice. Serial EDL
transverse sections from aging wild-type (A, C,
E) and MSVski (B, D, F)
mice stained with MAb specific to slow MyHC (A,
B), type IIa MyHC (C, D), and type IIb
MyHC (E, F). Type I, IIa, and IIb are
specifically stained by these antibodies, whereas type IIx fibers are
not stained by any of these antibodies. Arrowheads show examples of the
3 most abundant fiber types present in the EDL. Arrows highlight a
branched fiber. Bar, 100 µm. G: quantification of fiber
type frequency was performed by scoring all slow (dark gray bars), IIa
(light gray bars), IIx (open bars), and IIb (filled bars)
MyHC-containing fibers present at the muscle midbelly. Data are
means ± SE (n = 3). H: analysis of
non-IIb fiber size in wild-type and MSVski mice EDL with
age. Wild-type non-IIb fiber CSA increases during growth and plateaus
between P60 and aging. MSVski non-IIb fibers are
indistinguishable from those in wild-type littermates until P60 but
increase in size in aging mice. Size increase in aging
MSVski IIa fibers compared with control IIa fibers is
similar to that in all non-IIb fibers, suggesting that IIa and IIx
fibers increase in size to a similar extent in aging
MSVski mice. Data are means ± SE
(n = 3, 4 and 3 for wild type and 3, 5, and 5 for
MSVski at P17, P60, and P550, respectively). wt, Wild
type. **P < 0.01.
|
|
Counts of the total number of fiber profiles present at the EDL
midbelly at P60 revealed no significant difference between MSVski and control littermates [1,391 ± 62 (n = 8) and 1,268 ± 47 (n = 6),
respectively]. Like humans, rodents lose a significant number of
fibers during aging (31, 39, 44). This appeared to occur
in our control cohort, with total fiber profile number dropping to
995 ± 56 (n = 3) in aging animals
(P = 0.01). In aging MSVski mice, in
contrast, total fiber profile numbers were unchanged compared with P60
[1,268 ± 32 (n = 3) in aging
MSVski, P = 0.27], showing that IIb
fiber loss alone does not account for the change in fiber proportions.
On the contrary, the increased number of fiber profiles in aging
MSVski compared with aging controls suggests that the
transgene protects fibers from death, triggers generation of additional
fibers, or causes fiber branching. Analysis of serial sections
demonstrated the presence of some branched IIb fibers in aging
MSVski muscle (Fig. 4B, D,
and F). To test the hypothesis that enhanced apparent fiber
number was due to fiber branching, aging EDL muscle was dissociated
into single fibers and large IIb fibers were analyzed optically for
branching after plating in culture (Fig.
5A). Increased fiber branching
was a characteristic of aging mice (Fig. 5B). Single fibers
from young (P75-P150) MSVski mice were
indistinguishable from wild-type littermates with respect to branching
(Fig. 5B). However, as MSVski mice aged,
fibers showed earlier onset and quantitative increase in branching
(Fig. 5, A and B). By P720-P785 many
individual fibers had more than one branch (data not shown). Thus
formation of these apparently branched fibers is likely to account for
the apparent maintenance of fiber number and to contribute to the
apparent decline to wild-type mean IIb fiber size in aging
MSVski mice.
View larger version (57K):
[in this window]
[in a new window]
|
Fig. 5.
Exacerbation of aging-related fiber branching in
aging MSVski mice. A: isolated single EDL
fibers show that branching (arrows) is observed more commonly in fibers
from P240 MSVski mice than from age-matched wild-type
mice. Mononucleate cells migrate away from single fibers (arrowheads).
B: analysis of isolated EDL fiber branching in wild-type and
MSVski mice at P75-P150, P220-P240, and
P720-P785. Fiber branching increases slightly with normal aging
but markedly in MSVski mice from P220 onwards. Data are
means ± SE [n = 2 animals for all controls (110, 75, and 102 total fibers) and 3, 3, and 2 animals for
MSVski (87, 136 and 164 fibers) at P75-P150,
P220-P240, and P720-P785, respectively].
|
|
Rapid muscle damage in myoD/ mice induced by
MSVski hypertrophy.
The spectrum of changes present in aging MSVski mice,
involving nonperipheral myonucleation, increase in slower fiber type proportion, loss of IIb fibers, decreased size of the remaining IIb
fibers, and enhanced fiber branching, is reminiscent of the changes
observed in aging muscle. Moreover, the poor contractile properties of
fibers from young MSVski mice suggested that an early
defect may precipitate the later degenerative pathology. Human
satellite cells isolated from aging people show decreased replicative
capacity, suggesting that defective repair may contribute to the
changes observed in aging muscle (55). A similar
phenomenon has been suggested to contribute to impaired muscle
regeneration in aging rodents (66) and in the later stages
of Duchenne muscular dystrophy (73). These considerations
raised the possibility that the damage and regeneration in aging
MSVski mice might result from prolonged covert muscle
damage due to the overexpression of MSVski or the
resultant hypertrophy. To determine whether the MSVski
transgene leads to covert muscle damage in young animals, MSVski mice were bred onto a myoD/
background, which has impaired regeneration because of a reduced satellite cell differentiation capacity (49,
62).
We first demonstrated that the absence of myoD did not
prevent the MSVski-induced hypertrophy. The
MSVski transgene is expressed at similar levels in the
absence or presence of myoD (data not shown). Hypertrophy of
IIb fibers is seen in all MSVski genotypes, with or
without myoD (Fig.
6A). Moreover, quantitative
fiber size analysis showed that the removal of myoD does not
prevent IIb hypertrophy or significantly alter non-IIb fiber size.
However, MSVski-induced hypertrophy may be slightly
reduced in the absence of myoD (Figs. 6B and
1F). In particular, there appear to be fewer IIb fibers of
the largest sizes (>5,600 µm2) and somewhat more IIb
fibers of very small size (<1,600 µm2). Thus a
functional myoD gene is not required for
MSVski-induced hypertrophy, although
MSVski.myoD/ muscles are
slightly different from those of MSVski mice.
View larger version (92K):
[in this window]
[in a new window]
|
Fig. 6.
Muscle hypertrophy and damage in young
MSVski.myoD/ mice. A: EDL
transverse sections from P60 wild-type and MSVski mice
in a myoD+/ or myoD/
background stained for IIb MyHC. B: analysis of
P60 EDL IIb fiber size distribution as in Fig. 1 (n = 6 for each genotype). The absence of MyoD does not prevent the
Ski-induced hypertrophy (hatched bars). C: spaced transverse
sections of a P60 MSVski.myoD/ EDL
stained with MAb to embryonic (Emb), IIb MyHC, and H & E. Distances
between sections are: Emb-IIb and IIb-H & E = 48 µm, H & E-IIb = 80 µm. For orientation, a IIb fiber ( )
and a IIa fiber (*) are marked. Nonperipheral myonuclei are clearly
visible after H & E stain (arrowheads), and characteristic fibers that
appear split are apparent (highlighted on the drawings,
right). D: quantitation of TUNEL and BrdU
labeling in MSVski and control mice. At least 5 sections/leg were analyzed. To avoid ambiguity, no attempt was made to
distinguish myogenic from nonmyogenic nuclei. BrdU-labeled nuclei were
scored in whole EDL cross sections of MSVski mice or
wild-type littermates. TUNEL-labeled nuclei were less numerous and were
consequently scored in entire lower hindlimb sections from sibling
myoD+/ or myoD/
mice with or without the MSVski transgene. Values are
means ± SE; from left to right,
n = 6, 3, 4, 4, 2, and 2 mice. No differences were
significant, chiefly owing to high interanimal variation that appeared
to be independent of genotype. Scale bars, 100 µm.
|
|
MSVski.myoD/ EDL muscle
displayed two signs of degeneration. First, analysis of serial sections
showed apparent fiber branching in young
MSVski.myoD/ mice not
observed in myoD/ or
MSVski.myoD+/ mice (Fig. 6C). The
small branched fibers expressed IIb MHC and did not show signs of
acute regeneration, such as the expression of embryonic MHC (Fig.
6C). Second, nonperipheral nuclei were more numerous in
the MSVski.myoD/ mice (Fig.
6C) than in wild-type mice or mice carrying either myoD/ or MSVski alone. Up to
50 nonperipheral nuclei were observed in a single EDL section of one
MSVski.myoD/ mouse and at least 24 ± 6 (n = 6) nonperipheral nuclei/EDL section on
average. This compared with an average of 2.2 ± 0.5 (n = 6), 1.5 ± 0.3 (n = 6) and
4.5 ± 1.4 (n = 8) nonperipheral nuclei/EDL section in myoD+/,
myoD/, and
MSVski.myoD+/,
respectively. Although myoD/ mice have fewer
total fibers, the EDL IIb fiber proportion is unchanged (34,
60). The MSVski transgene does not alter either property (data not shown). However, the number of nuclei/fiber profile
is increased from 2.7 ± 0.1 (n = 4) in
myoD/ (similar to wild type, see Fig.
2B) to 4.2 ± 0.4 (n = 5) in
MSVski.myoD/ (P < 0.01), consistent with the view that hyperplasia is triggered by the
MSVski transgene in young
myoD/ animals. Presumably, many of these
additional nuclei are mononucleate cells (see below). So, by several
criteria, young MSVski.myoD/ mice appear
to undergo significant muscle degeneration and show other changes
reminiscent of older MSVski mice containing functional MyoD.
The slight increase in nonperipheral nuclei in
MSVski.myoD+/ compared with
myoD+/ animals suggested that increased
nuclear turnover might be occurring in young MSVski
mice. We assessed this possibility in a cohort of MSVski
and wild-type littermates by analyzing BrdU incorporation into nuclei
in S phase and TUNEL staining for apoptotic nuclei within lower
hindlimb muscle. No gross turnover of nuclei was detected, although a
slight increase in nuclear turnover in MSVski animals
could not be excluded (Fig. 6D). Both BrdU-labeled and TUNEL-labeled nuclei were rare in muscle tissue of either genotype (averaging around 1 BrdU-labeled nucleus per EDL cryosection and many
fewer TUNEL-labeled nuclei). Thus, although functioning satellite cells
appear to be required to maintain MSVski muscle
integrity, enhanced nuclear turnover is not detected.
Decreased satellite cells with aging and increased
desmin cells in MSVski EDL.
The phenotypes of young MSVski.myoD/
mice, older MSVski mice, and aged wild-type mice share
many similarities. It is possible, therefore, that a defect in
satellite cells contributes to the failure of muscle regeneration in
aging MSVski and/or wild-type mice, as it appears to do
in MSVski.myoD/ mice. Skeletal muscle
satellite cell number decreases with increasing age (8, 27,
67), and this may contribute to the poor regeneration capacity
observed in aging muscle. Individual culture of single intact fibers
and their associated mononucleate cells induces activation,
proliferation, and, subsequently, differentiation of satellite cells
and permits assessment of fiber-associated cell properties without the
cell selection inherent in bulk culture methods (7, 8,
57). In single-fiber cultures from young and adult wild-type
mice, most fibers yielded numerous cells (Table 2, Fig.
7),
generally over 10 cells/fiber, most of which contained desmin (Table 2,
Fig. 7), an intermediate filament protein expressed in myoblasts and
differentiated muscle fibers (36). In cultures from aging
wild-type mice, most mononucleate cells still contain desmin (Fig. 7),
but desmin+ cell numbers are significantly reduced
(P < 0.001; Table 2, Fig. 7). Consistent with this,
almost 50% of fibers from aging wild-type mice yielded no mononucleate
cells, whereas almost all fibers from younger mice yield myogenic cells
(Table 2). Thus normal aging leads to a decline in myogenic cells
recoverable in single-fiber cultures, confirming the finding of Bockold
et al. (8) on C57Bl/10 mice.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 7.
Changes in mononucleate cell yield with aging from
wild-type and MSVski single EDL fiber. Data from
individual EDL fiber cultures, ranked according to the number of
desmin+ cells present after 3 days in culture are
displayed, following Bockold et al. (8), by representing
the group of cells derived from a single fiber as a point on the
y-axis. The number of desmin+ cells and the
total number of cells derived from each fiber are read from the
x-axis. To permit visual comparison between conditions from
which distinct numbers of fibers were analyzed, the data are normalized
for sample size by expressing individual rank as a percentage of the
total number of fibers analyzed for each condition. In wild-type
cultures (left), desmin+ cell numbers were a
high proportion of total cell numbers from each fiber at all ages, but
total yield declined with age (P < 0.01). In
MSVski cultures (right), the number of
desmin+ cells shows no significant decline with age,
although cultures are more heterogeneous than wild type. Also, total
cell yield is more variable, with many individual MSVski
fibers at P220-P240 and P720-P785 yielding more
desmin cells than wild-type fibers (P < 0.05 and P < 0.01, respectively). Number of
nonhypercontracted fibers analyzed (from n animals) was 27 (2) and 25 (3) at P75-P150, 16 (2) and 13 (3) at P220-P240, and 32 (2) and 32 (2) at P720-P785 for control
and MSVski, respectively. All trends discussed were also
observed from hypercontracted fibers (data not shown; such fibers
constituted approximately one-half those cultured) and from single
fibers analyzed for desmin expression from these and 2 further mice
after between 2 and 5 days in culture.
|
|
Young MSVski mice appear to have normal mononucleate
cells. The numbers of both desmin+ and desmin
cells are comparable to those in young wild-type mice. On average, each
cultured wild-type fiber yielded 30 ± 6 (n = 27)
desmin+ cells (84% ± 4 of total cells; n = 27), whereas MSVski fibers yielded 32 ± 5 (n = 25) desmin+ cells (81% ± 7 of total
cells; n = 23). However, as MSVski mice mature, changes in mononucleate cells become apparent. By 8 mos of age,
two changes are observed. First, the number of desmin
cells obtained from MSVski fibers increases to 45 ± 24 (n = 13) from 2 ± 1 in wild type
(P < 0.01). Second, although desmin+ cell
numbers are unchanged overall, their distribution between fibers is
less uniform: some fibers yielded more desmin+ cells,
whereas more fibers yielded no desmin+ cells. These changes
were also observed in aged MSVski mice. Aged
MSVski fibers yielded more desmin+ and also
more desmin cells than age-matched controls
(P < 0.001; Fig. 7, Table 2). The normal aging-related
decline in desmin+ cells is not so marked in
MSVski mice, which are undergoing continual regeneration.
Decreased myoblast differentiation efficiency with normal aging and
with MSVski.
The increased number of desmin+ cells in aged
MSVski mice compared with age-matched controls is
clearly not sufficient to repair the Ski-induced damage efficiently and
prevent the exacerbated aging-related changes in these mice. This is
surprising in view of the efficient repair of damage that occurs in
young MSVski mice and raises the possibility that
decrease in muscle regenerative capacity may be the result of a
decrease in myoblast differentiative capacity. We therefore examined
whether the properties of myogenic cells were altered in aging
wild-type and MSVski mice.
Analysis of proliferative capacity of desmin+ or
desmin cells from all three ages of both wild-type and
MSVski mice revealed no significant differences in the
rate of cell doubling, which occurred in ~13 h in all cases (data not
shown). No evidence of apoptosis was observed in any cultures.
Thus changes in intrinsic myoblast proliferation capacity are unlikely
to account for either the decline in desmin+ cell yield
with age or the enhanced numbers of desmin+ cells in aged
MSVski mice.
In contrast to proliferative capacity, myogenic differentiative
capacity showed marked changes with age, and these were exacerbated in
MSVski mice. After challenge with low-growth factor
conditions for 2 days, aging wild-type fiber cultures yielded fewer
differentiated myocytes than cultures from younger animals (Fig.
8). In cultures from P75-P150 mice,
nearly all desmin+ cells differentiated, as detected by
MyHC staining, forming long multinucleated myotubes (Fig.
8A). Quantitation of the proportion of nuclei in
desmin+ and MyHC+ cytoplasm revealed that
differentiation efficiency gradually decreased in wild-type cultures
with increasing age of the mice (P < 0.01; Fig.
8B). Desmin+ cell number itself was
significantly decreased at P720-P785 (P < 0.01),
with 99 ± 9 (n = 18) desmin+ cells
compared with 430 ± 56 (n = 20) at P75-P150
and 554 ± 69 (n = 19) at P220-P240, probably
because of the lower numbers of desmin+ cells yielded by
aged fibers. This observation raised the possibility that the decline
in differentiation efficiency could be due to the reduced density of
desmin+ cells in the well. However, several observations
argue against this possibility. First, the number of
desmin+ cells per culture was similar at P75-P150 and
P220-P240, yet the differentiation efficiency declined
(P < 0.01, Fig. 9A,
right). Second, all but one
aged fiber culture contained between 51 and 200 desmin+
cells, for which the mean differentiation efficiency was 46.5% ± 6.7 (n = 17) compared with 88.6% ± 7.5 (n = 4) for young fiber cultures that yielded similarly low numbers of
desmin+ cells (Fig. 9A, right).
Third, analysis of individual fiber cultures showed no significant
correlation between desmin+ cell number and differentiation
efficiency at any age (Fig. 9B). The number of
desmin cells was low in wild-type cultures at all ages
and did not affect differentiation efficiency of desmin+
cells (Fig. 9A, left). Thus, in wild-type
cultures, the differentiation efficiency of myocytes declines in
parallel with increasing age, and this decline appears to be intrinsic
to the myocytes.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 8.
Decreased satellite cell
differentiation with aging and MSVski. A:
fluorescence micrographs of single-fiber cultures stained for the
expression of the myogenic cell marker desmin (Cy3, red), all MHC
(FITC, green) and DNA (DAPI, blue). Cultures were established from EDL
fibers of wild-type and MSVski mice aged P75-P90
and P720. After 3 days in plating medium and 2 days in growth medium,
cells were induced to differentiate in a medium low in growth factors
for a further 2 days. Differentiated desmin+ cells
(arrowheads) were numerous in the cultures from young muscles, whereas
in the older cultures, there were more nondifferentiated
desmin+ cells (arrows). B: proportion of
desmin+ cells expressing MHC in cultures from wild type and
MSVski EDL fibers of P75-P150, P220-P240, and
P720-P785 mice. Values are means ± SE (n = 20, 19, 18 for wild type and 22, 22, 29 for MSVski at
P75-P150, P220-P240, and P720-P785, respectively).
Significant differences: *P < 0.05, **P < 0.01.
|
|
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 9.
Satellite cell differentiation defect is independent of
both myogenic and nonmyogenic cell density. A: independence
of differentiation efficiency from desmin+
(right) and desmin (left) cell
numbers at each age. Values are means ± SE with n as
indicated in bars. B: comparison of differentiation
efficiency of wild-type and aged MSVski fiber cultures
yielding similar numbers of desmin+ cells.
|
|
In MSVski cultures, a more marked trend of decreased
differentiation efficiency was observed with aging (Fig. 8). Young
MSVski cultures did not differ significantly from
wild-type cultures in any respect, with 85% ± 2 (n = 22) of desmin+ cells containing MHC (Figs. 8, 9). However,
cultures from P220-P240 and P720-P785 MSVski
fibers showed progressively worse differentiation compared with
age-matched wild-type fiber cultures (P < 0.05 and P < 0.01, respectively). Eight-month-old
MSVski myoblasts differentiated as poorly [57% ± 5 (n = 22) desmin+ cells containing MyHC] as
those of aged wild-type cultures, and aged MSVski
myogenic cells differentiated half as well as [23% ± 4 (n = 29) desmin+ cells containing MyHC].
Thus the decline in differentiative efficiency occurs earlier and is
more dramatic as MSVski mice age. As in fiber cultures
assessed 3 days after explant, older MSVski fibers yielded significantly more desmin cells after the
differentiation period (Fig. 9A, left;
P < 0.05 at P220-P240 and P < 0.01 at P720-P785). However, correlation analysis showed that
differentiation efficiency declined with age, independent of
desmin cell number. No significant correlation was
apparent in any age group (Fig. 9A, left).
Similarly, the decline in differentiation efficiency of
MSVski fiber cultures with age was not accounted for by
decreasing desmin+ cell numbers (Fig. 9A,
right). When young and aged MSVski fibers yielding similarly low numbers of desmin cells
(1-400) were compared, aged cells differentiated
poorly (Fig. 9A, right). Moreover, the aged
MSVski cultures had more desmin+ cells than
age-matched wild-type cultures yet differentiated less well independent
of desmin+ cell number (Fig. 9). Together, the data show
that MSVski desmin+ cells undergo a more
marked and earlier loss of differentiation efficiency than wild-type
myogenic cells, paralleling the earlier onset of a spectrum of
aging-related changes in MSVski muscles.
| |
DISCUSSION |
Four major findings arise from this study of aging and the
long-term effects of MSVski-induced muscle hypertrophy.
First, fiber-associated myogenic cells lose differentiation efficiency with age. Second, although MSVski mice have larger
fibers, these fibers are relatively weak. Third, the
MSVski transgene leads to muscle changes suggestive of
premature muscle aging, possibly triggered by "covert" muscle
damage. Fourth, the premature aging and onset of obvious degeneration
in MSVski hypertrophied muscle is preceded by decline in
differentiation efficiency of fiber-associated mononucleate myogenic cells.
Defective differentiation of myoblasts from aging mice.
Aging mammalian muscle undergoes a characteristic set of changes
including loss of force, slower contraction, and poor regenerative ability (44, 63). The primary causes of and the sequence
of events leading to such changes are unclear (29). Our
finding that myogenic cells from aging muscle have a decreased
intrinsic efficiency of differentiation focuses attention on whether
this change contributes to the deficits of aging muscle.
Despite extensive analysis in the past of myoblasts derived from aging
muscle, poor differentiation has not been described hitherto. Yet the
decrease in satellite cell numbers, which we also observed, is well
known (8, 66, 67). We believe the key to our analysis is
the single-fiber culture method, which allows a high proportion of
satellite cells associated with individual fibers to be assessed in a
unique culture environment. Satellite cell abundance in mouse EDL
muscle is ~1-2% of myonuclei, i.e., five per fiber, and this is
consistent with the number of distinct proliferative foci observed
during the early days of single-fiber cultures (5, 68).
Whole muscle dissociation methods usually have estimated yields of
~0.01% of satellite cells and, therefore, may select for subsets of
myogenic cells (57). Such cells frequently show reduced
replication capacity but satisfactory differentiation when derived from
older tissue (73). We observed neither altered morphology
nor altered proliferation in the generality of myogenic cells derived
from aged mouse muscle fibers, yet we found poor differentiation.
Together with recent demonstrations of the heterogeneity of the
myogenic cell compartment in postnatal muscle tissue (5, 6, 30,
42), these findings raise the possibility that distinct myogenic
precursor cell pools may show different changes during aging. In
situations in which aged satellite cells are forced to undergo
myogenesis in a young environment, they perform well (11, 12, 37,
39), so perhaps environmental factors can revive the
differentiation capacity of aged satellite cells in a way similar to
the reported effect of short-term IGF-I exposure on replicative
capacity of myoblasts isolated from aging rats (14).
Nevertheless, during the slow aging of muscle the lack of such factors
may account for the gradual loss of satellite cell performance and thus
contribute to other aging-related changes.
Previously, the large numbers of nonmyogenic fibroblastic cells in
whole muscle dissociates hampered analysis of differentiation because
fibroblastic cells are more abundant in aging muscle and can inhibit
differentiation (53). Our analysis provides two arguments
against an inhibitory effect of nonmyogenic cells accounting for
reduced differentiation. First, very few desmin
nonmyogenic cells are present in single-fiber cultures from wild-type mice at any age that we examined, and the few there are do not significantly increase with age, either in absolute numbers or as a
proportion of desmin+ cells. Second, there is no
correlation between differentiation efficiency of desmin+
cells and number or proportion of desmin cells in
cultures from individual fibers of the same mouse at any age examined.
Neither does the known cooperativity of myoblast differentiation,
together with the reduced numbers of desmin+ cells from
aging muscle, account for the poor differentiation: single-fiber
cultures yielding similar total desmin+ cell numbers still
showed reduced differentiation if derived from aging mice compared with
younger animals. Moreover, there was no correlation between
desmin+ cell number and differentiation efficiency in
single-fiber cultures from young mice. Thus, in contrast to data from
selected populations of highly proliferative myogenic cells isolated
from aging muscle, which show defects in proliferation
(66), our analysis of the behavior of the general
population of fiber-associated satellite cells reveals little
alteration in satellite cell proliferation. However, deficits in
myoblast differentiation efficiency become more marked with age.
Does decreased myoblast differentiation efficiency result from or
contribute toward other changes in aging muscle? Our data provide
evidence that it is an early marker of muscle aging. CBA/J mice show
very limited signs of aging by P550. Yet already by P240, there was a
significant reduction in myoblast differentiation efficiency, which
preceded any detectable drop in myogenic cell numbers. When other
markers of aging such as fiber branching and decreased numbers of
satellite cells become apparent, myogenic cell differentiation
efficiency is even poorer. As discussed in more detail below,
MSVski mice show premature changes similar to those
occurring during aging, and these are preceded by premature decline in
myoblast differentiation capacity. Thus it appears that decline in
myoblast differentiation efficiency is a very early marker of
aging-related changes and progressively worsens with age. This suggests
that decrease in differentiation efficiency is not a consequence of
other known changes. On the contrary, it has the potential to cause or
contribute to the later changes.
MSVski, MyoD, and muscle hypertrophy.
Our data confirm that the MSVski transgene leads to
specific IIb fiber hypertrophy (43, 69). This hypertrophy
results from a postnatal increase in type IIb fiber cytoplasmic volume, accompanied by proportionate increase in muscle protein, without significant changes in total fiber number, fiber type, number of nuclei
per fiber, or extractable DNA content, suggesting strongly that the
over twofold increase in cytoplasmic volume of IIb fibers occurred
without a change in the number of myofiber nuclei. Consistent with the
view that extra myoblast differentiation is not required for the
hypertrophy, we find similar hypertrophy in
MSVski.myoD/ mice that have defective
myogenic cell differentiation (49, 62). Moreover, our
findings show that MyoD is not required either for the larger size of
type IIb fibers (which normally express more MyoD; Ref.
34) or for the differential expression of
MSVski between fast and slow muscle (18,
70) or for MSVski-driven muscle hypertrophy. In
contrast, manipulation of myogenin, myocyte enhancer factor-2 (MEF-2),
and Ras proteins has implicated each in fiber size regulation in
rodents (32, 50, 75).
MSVski transgene causes late-onset muscle degeneration.
Muscle from young MSVski transgenic mice appears
healthy, despite marked hypertrophy. However, we report that subtle
early defects in muscle function are followed by a profound late-onset degeneration. During both maximal activation and Mg-rigor conditions, young MSVski fibers generated less force per CSA than
control fibers. The reason for this apparent force deficit is unclear, although the lack of change in activation and relaxation kinetics in
permeabilized fibers argues against changes in regulatory machinery and
suggests a decrease in force generation by myosin heads. Decrease in
rigor force, assuming that all myosin heads are attached to actin in
normal rigor, also suggests either a decrease in strongly attached
myosin heads or a decrease in the tension generated per myosin head
(17, 45). Disorganization of intracellular structure, as
suggested by the misaligned myofibrils, may prevent force generation or
effective transmission of force along the fiber.
In MSVski mice, altered muscle function is followed by
onset of muscle degeneration in later life. We have been unable to detect any increase in classic features of fiber damage in young MSVski mice. In particular, BrdU labeling experiments in
young MSVski mice failed to demonstrate a significant
activation of satellite cell proliferation. Only around one nucleus per
whole muscle cross section was labeled in these experiments, and it is
unclear whether in control animals these nuclei were satellite cells or
nonmuscle cells, so it is possible that a significant increase on
satellite cell activation has been overlooked. Whatever the case, the
data suggest that muscle degeneration/regeneration in young
MSVski muscle does not involve extensive satellite cell activation comparable to that caused by muscle crush or toxin injection. Nevertheless, our observation that genetic ablation of the
myoD gene causes premature onset of muscle damage in
MSVski mice suggests that satellite cells can mitigate
the effects of the transgene in early life. The major described
function of myoD in postnatal muscle is to promote satellite
cell differentiation during muscle regeneration, so inefficient
satellite cell-mediated repair of covert damage could explain the
MSVski.myoD/ phenotype, just as
described for mdx.myoD/ mice
(49). Although we cannot rule out some more complex
genetic interaction of MSVski and myoD, the
simplest view of these findings suggests that covert muscle
degeneration occurs in young MSVski mice, which is
normally efficiently repaired. As MSVski mice age, abundant central nucleation, fiber branching, and very small fibers appear. These probable degenerative signs are accompanied by a spectrum
of changes characteristic of normal murine muscle aging, including
fiber atrophy, a shift to slower fiber types, and altered myogenic cell
behavior (see below). Because aging-related changes, such as fiber type
shifts, may in part be attributed to changes of innervation, the
MSVski mice raise the possibilities either that
MSVski affects the motor neurons directly or that
changes in innervation with age may be "myogenic" in origin.
Whether the observed decline in specific force is pathological
and might trigger later changes in MSVski muscle is
unclear. Muscle hypertrophy resulting from either overload or
compensatory hypertrophy or dystrophy is also accompanied by a decline
in specific force (22, 58), so it is uncertain whether the
high expression of c-ski is directly harmful to fibers or
whether the hypertrophy induced by c-ski expression is the
cause. However, the literature is replete with examples of degenerative
changes associated with muscle fiber hypertrophy. Muscle overload
hypertrophy in rodents is associated with muscle fiber damage
(74). In addition, muscle hypertrophy after weight lifting
in humans, which involves increase in myonuclear number as well as
fiber volume, is associated with elevated serum and urinary markers for
muscle damage (54). Hypertrophy-associated damage is often
attributed to the training regimes or forces applied to the muscle that
elicit hypertrophy. This interpretation may be partially correct.
However, our data together with the elevated CK and muscle weakness in
MSVski transgenic cattle (9) suggest that
muscle with nuclear domain enlargement caused by genetic manipulation
may be more fragile and, without a change in stress, may become
damaged. No comparative studies of muscle aging in bodybuilders or
weight lifters, proper analysis of the relative contribution of nuclear
domain enlargement and increase in myonuclear number, or long-term
follow-ups of the effects of strength training to promote muscle
function in the elderly have been reported. So it is important to
determine whether, in susceptible individuals, long-term muscle
hypertrophy elicited through training regimes may also contribute,
through long-term covert damage and resultant alteration in
regenerative capacity of satellite cells, to an ultimately worsened
aging prognosis.
Myoblast changes precede most late-onset degenerative changes.
Regardless of the exact relationship between early force deficit and
the late-onset muscle degeneration in MSVski mice, what triggers muscle degenerative changes when they appear? Several lines of
argument point to a contribution of changes in the mononucleate cell
population of MSVski mice. First, single-fiber cultures
of MSVski muscle show premature aging-related decline in
differentiative capacity of myogenic cells. In the context of an
enhanced need for muscle repair caused by the transgene, such a decline
could account for the appearance of degeneration. Second, the fibers in
MSVski mice are highly branched even at P240, before
most degenerative changes are apparent, suggesting that the
differentiation deficit observed in cell culture is also occurring in
vivo. We never observed embryonic myosin in MSVski
muscle, so de novo muscle formation is unlikely. Instead, the branched
fibers are consistent with inefficient repair of damaged fiber regions.
Third, no differences in mononucleate cells between young
MSVski and wild-type mice were observed, suggesting that
MSVski does not act directly in mononucleate cells but
rather triggers the changes indirectly. However, significant changes
were present by P240, and these worsened in older MSVski
mice. Fourth, nonmyogenic desmin cell numbers are
enhanced in P240 MSVski compared with wild-type mice.
Thus mononucleate cell changes arise early, whereas most fiber changes
occur later.
We examined the myogenic desmin+ and nonmyogenic
desmin cells for their contribution to the
MSVski phenotype. Myogenic cells from P240
MSVski are defective in differentiation compared with both age-matched controls and MSVski myoblasts from
younger mice. This defect is not accounted for by decline in myogenic
cell number because such decline does not occur in
MSVski mice, nor it is caused by increase in nonmyogenic
cells because some single-fiber cultures yield no nonmyogenic cells yet
still show poor differentiation of myogenic cells. It is possible that
the marked increase in nonmyogenic cells is important. However, the
clear changes in myogenic cells make it unnecessary to evoke effects of
nonmyogenic cells. Instead, premature changes in myogenic cells akin to
those occurring in normal aging are sufficient to account for both the exacerbated aging-related changes and the late onset of muscle degeneration in MSVski mice. In future, it will be
essential to test this hypothesis and to examine whether human aging
and late-onset muscle degenerations (such as muscular dystrophies or
spinal muscular atrophies) share elements in common with these murine models.
We thank Charlotte Peterson, Steve Hughes, Bernd Krippl, Michael
Rudnicki, Kim Wells, Terry Partridge, and Louise Heslop for help and reagents.
This work was supported by the MRC and European Commission
BMH4-CT96-0174 and QLK6-2000-530 to S. M. Hughes.
S. B. P. Chargé and A. S. Brack held MRC PhD studentships.
Address for reprint requests and other correspondence:
S. M. Hughes, MRC Centre for Developmental Neurobiology, 4th
floor south, New Hunt's House, Guy's Campus, King's College London, London SE1 1UL, UK (E-mail:
simon.hughes{at}kcl.ac.uk).
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
TOP
ABSTRACT
INTRODUCTION
