| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Department of Clinical Pathology,* Institute of Clinical Research, Odense University Hospital, Odense; Immunology and Microbiology, Institute of Medical Biology, University of Southern Denmark, Odense; and the Institute of Molecular Pathology, University of Copenhagen, Copenhagen, Denmark
 |
Abstract |
|---|
 Top Abstract Experimental ProceduresResultsDiscussionReferences |
|---|
7B integrin, and dystroglycans. Ectopic ADAM12 expression in normal mouse skeletal muscle also improved regeneration after freeze injury, presumably by the same mechanism. Hence, it was suggested that ADAM12 could be a candidate for nonreplacement gene therapy of Duchenne muscular dystrophy. We therefore evaluated the long-term effect of ADAM12 overexpression in muscle. Surprisingly, we observed loss of skeletal muscle and accelerated fibrosis and adipogenesis in 1-year-old mdx mice transgenically overexpressing ADAM12 (ADAM12+/mdx mice), even though their utrophin levels were mildly elevated compared with age-matched controls. Thus, membrane stabilization was not sufficient to provide protection during prolonged disease. Consequently, we reinvestigated skeletal muscle regeneration in ADAM12 transgenic mice (ADAM12+) after a knife cut lesion and observed that the regeneration process was significantly impaired. ADAM12 seemed to inhibit the satellite cell response and delay myoblast differentiation. These results discourage long-term therapeutic use of ADAM12. They also point to impaired regeneration as a possible factor in development of muscular dystrophy.
ADAM12 is expressed in skeletal muscle at the embryonic stage and in the early postnatal period, where it is present in the developing myofibers.11-14 It is absent from normal adult skeletal muscle11,12 but re-expressed in the fibers during skeletal muscle regeneration.15 ADAM12 protein was proposed to participate directly in the cell to cell fusion process to generate multinucleated cells, both during embryonic growth and regeneration.14-16
Studies on ADAM12-null mice suggested a regulatory role for ADAM12 in development of both skeletal muscle and adipose tissue because viable mutants showed reductions in interscapular muscles and brown adipose tissue; however, the ADAM12 gene was proven dispensable for myoblast fusion.9 Ectopic expression of ADAM12 in an ADAM12 transgenic (TG) mouse model10 resulted in a faster and more efficient regeneration after freeze injuries.10 Overexpression of ADAM12 also induced adipogenesis in these TG mice,17 in accordance with the ADAM12-null mouse study.9
ADAM12 was shown to alleviate the pathology of the mdx mouse when overexpressed in this model for Duchenne muscular dystrophy.10,12
Analyses indicated that ADAM12 stabilized the sarcolemma through an up-regulation of utrophin thereby protecting the mdx fibers and thus modulating the effect of lack of dystrophin.12
In accordance with this, the effect of ADAM12 on muscle damage seemed to be specific for the mdx phenotype, because overexpression of ADAM12 in the laminin 2-deficient (dy/dy) muscular dystrophic mice had no effect on the damaged muscle.18
Based on the ability of ADAM12 to boost utrophin gene levels, ADAM12 was suggested as a candidate for therapy of Duchenne muscular dystrophy.7
With the aim of further characterizing the therapeutic potential of ADAM12 in substitution or nonreplacement gene therapy in dystophinopathies, we decided to evaluate the long-term effect of ectopic ADAM12 expression. Analysis of skeletal muscle samples obtained from 1-year-old ADAM12+/mdx mice, however, revealed an aggravating effect of the transgene on the mdx muscle pathology, and investigations indicated a reduction of muscle stem cell mobilization response in these old mice. This prompted us to investigate skeletal muscle regeneration after major acute injury in ADAM12+ TG mice, with focus on satellite cell mobilization. The results point to ADAM12 as a negative regulator of the satellite cell response during regeneration promoting adipogenic and fibrogenic substitution of muscle tissue.
|
Experimental Procedures |
|---|
|
TopAbstractExperimental ProceduresResultsDiscussionReferences |
|---|
ADAM12+ TG mice and age-matched normal littermate controls (LC), 10 to 12 weeks of age, were used for the knife injury experiment. The generation of TG mice expressing human ADAM12 under the control of the muscle creatine kinase promoter has been described previously.12 ADAM12+/dystrophin– animals were obtained by mating homozygous female mdx mice [C57BL/10ScSn-Dmd(mdx); Jackson Laboratories, Bar Harbor, ME] with male ADAM12+ mice.12 One-year-old individuals were used for analysis along with 1-year-old homozygous mdx mice and normal controls.
Animal Experiments
All animal experiments were performed in accordance with Danish legislation on animal welfare and approved by the Danish Council for Supervision with Experimental Animals. ADAM12+ TG (n = 23) and LC mice (n = 24) (10 to 12 weeks of age) were anesthetized using Hypnorm/Dormicum (fentanyl:fluanisone/midazolam), and experimental regeneration of skeletal muscle was induced by a stab incision in both calf muscles (musculus triceps surae). The mice were sacrificed by cervical dislocation 0.25, 1, 2, 3, 4, 5, 7, 10, and 14 days after injury. The injured muscles were removed and processed for either histological analysis or RNA purification. Uninjured muscle tissue was obtained from three ADAM12+ mice and three control mice. For histological analysis, muscles were fixed in formalin and embedded in paraffin and for RNA analysis kept in RNAlater (Invitrogen, Tåstrup, Denmark) until further processing. From the 1-year-old mice (ADAM12+/mdx, mdx, and control) musculus quadriceps and musculus tibialis muscles were formalin-fixed and embedded in paraffin whereas musculus gastrocnemius and musculus tibialis muscles were snap frozen for cryosectioning.
RT-PCR and Quantitation
Total RNA from muscle samples kept in RNAlater was isolated using TRIzol reagent (Invitrogen) according to the manufacturer’s instructions. One µg of pooled total RNA was used to synthesize cDNA using the SuperscriptIII first-strand synthesis system (Invitrogen). The RNA samples from each genotype were pooled (10 µg of each sample) for each time point after surgery.
Two µl of cDNA were used as template for each PCR experiment, and primers were added at a concentration of 5 pmol/20 µl each. Primer sequences used: Pax7 sense: 5'-GCTACCAGTACAGCCAGTATG-3', Pax7 antisense: 5'-GTCACTAAGCATGGGTAGATG-3', MyoD sense: 5'-GCCCGCGCTCCAACTGCTCTGAT-3', MyoD antisense: 5'-CCTACGGTGGTGCGCCCTCTGC-3', Myogenin sense: 5'-GGGCCCCTGGAAGAAAAG-3'; Myogenin antisense: 5'-AGGAGGCGCTGTGGGAGT-3'; NCAM sense: 5'-ATTGTCTGC- TCCTCGGTCCATT-3'; NCAM antisense: 5'-GAGGGTAGAGGAGTCGTCATCA-3'; GAPDH sense: 5'-AACGACCCCTTCATTGAC-3', GAPDH antisense: 5'-TCCACGACATACTCAGCAC-3', 18s rRNA sense: 5'-GTAACCCGTTGAACCCC- ATT-3' and 18s rRNA antisense: 5'-CCATCCAATCGGTAGTGACG-3'. All PCR reactions were run as multiplex reactions with GAPDH. 18s rRNA was run as an additional control gene (not shown). Reaction conditions: initial denaturation 94°C for 5 minutes; denaturation: 94°C for 30 seconds; annealing: specific temperature for 30 seconds; extension: 72°C for 30 seconds; and final extension: 72°C for 7 minutes; Pax7: 56°C, 35 cycles; MyoD: touchdown 67 to 57°C, 35 cycles; Myogenin: 60°C, 35 cycles; NCAM: 57°C, 31 cycles; and 18s rRNA: 55°C, 30 cycles. Both a nontemplate (water) control and a reverse transcriptase control (–RNA) were run as negative controls for all primer pairs. Uninjured muscle was included in the analysis as normal background control. PCR products were visualized on a 2% agarose gel stained with ethidium bromide. DNA sequence analysis confirmed the specificity of the PCR reaction products. Each data point was quantified using the GelDoc 1000 and the software program Quantity One (Bio-Rad, Herlev, Denmark) and normalized to the abundance of GAPDH (glyceraldehyde-3-phophate-dehydrogenase) expressed as log2 ratio followed by calculation of fold expression compared with normal uninjured muscle.19
Histological Analysis and Immunohistochemical Stainings
Formalin-fixed and paraffin-embedded muscle tissue was cut in 4-µm sections in a longitudinal direction, to ensure inclusion of the zones of interest in the injured area, and stained with hematoxylin and eosin (H&E) for morphological analysis. A Sirius red staining was performed to detect fibrosis. For immunohistochemistry (IHC), antigen retrieval was done by microwave oven treatment, and sections were subsequently blocked to eliminate endogenous peroxidase and endogenous biotin. The following primary antibodies were used: monoclonal mouse--myogenin at 1:200 (F5D; DAKO, Glostrup, Denmark); monoclonal mouse--pax7 at 1:10 (Developmental Studies Hybridoma Bank, Iowa City, IA); polyclonal rabbit--chicken NCAM (neural cell adhesion molecule) at 1:2000 (AB5032; Chemicon International, Inc., Temecula, CA); rat--mouse ki67 at 1:50 (Tec3; DAKO), monoclonal mouse--dystrophin at 1:10 (MAB 1692; Chemicon, AH Diagnostics, Arhus, Denmark), polyclonal rat--mouse CD45 at 1:50 [30-F11 (Ly 5); BD Pharmingen, Brondby, Denmark], polyclonal rabbit--human myostatin 1:200 (Chemicon, AH Diagnostics), monoclonal mouse--human transforming growth factor (TGF)-ß1 at 1:10 (NCL-TGFB; NovoCastra, Newcastle, UK), monoclonal mouse--human follistatin at 1:10 [85918 (MAB669); R&D Systems, TriChem ApS, Skanderborg, Denmark).
For detection of mouse monoclonal antibodies, the ARK kit (DAKO) was used. The labeled streptavidin-biotin-system with goat--rat (E468 at 1:200; DAKO) and the Envision+ horseradish peroxidase system (K4003, DAKO) were used as detection systems for polyclonal antibodies. DAB+ was used as chromogen. Nuclei were counterstained using Mayer’s hemalum with 4.5% chloral hydrate. Controls with exclusion of the primary antibodies were included in all experiments.
Immunofluorescence Stainings
Crysections of musculus tibialis (n = 6) and musculus gastrocnemius (n = 6) from ADAM12+/mdx and mdx mice were stained with monoclonal utrophin at 1:50 (NCL-DRP2; NovoCastra) and monoclonal ß1D integrin at 1:25 [2B1 (MAB1900); Chemicon, AH Diagnostics] using an ARK-immunofluorescence protocol using a fluorescein isothiocyanate-labeled Tyramid (TSA kit, NEL701; Perkin-Elmer, Waltham, MA) procedure. Before staining, sections were blocked to eliminate endogenous biotin and peroxidase. Nuclei were detected with 4,6-diamidino-2-phenylindole mounting medium (Vectashield; Vector Laboratories, Burlingame, CA).
Morphometrics
The number of cells expressing a particular protein during the course of regeneration was estimated on sections by counting the number of positive nuclei (pax7, ki67, and myogenin) or nuclei surrounded by positive cytoplasm in mononuclear cells (NCAM). Morphometrics was performed using a Leica microscope equipped with a camera connected to a computer and a motorized cross board. By means of the CAST software (2000; Olympus Denmark A/S) systematic random fields for counting were selected. Fields with positive events (positive nuclei or stained cytoplasm) were accepted as lesion area and included. The number of nuclei in cells with positive events were compared with the total number of nuclei in the counted fields to give an estimate of the number of cells expressing the various proteins at the time points investigated.
Statistics
P values were calculated for the comparison of TG animals with LC animals using two-way factorial analysis of variance analysis with an -value of 0.05. P values <0.05 were considered statistically significant.
|
Results |
|---|
|
TopAbstractExperimental ProceduresResultsDiscussionReferences |
|---|
Having seen the beneficial effect of short-term ADAM12 expression on the mdx phenotype,12
the next step in evaluating the therapeutic potential of ADAM12 was to study its long-term effects. For this purpose musculus quadriceps and musculus tibialis from 1-year-old ADAM12+/mdx mice, mdx mice, and normal mice was analyzed by IHC (Table 1
). A Sirius staining performed on muscle in the 1-year-old ADAM12+/mdx mice to detect fibrosis (Figure 1, A–C)
demonstrated that the ADAM12/mdx mice had more extensive fibrosis and more adipocytes, thus having a more dystrophic phenotype compared with the mdx mice. A recent study indicated a relationship between myostatin, fibrosis, and TGF-ß1 in skeletal muscle.20
Therefore we decided to evaluate the expression of myostatin, a known inhibitor of myogenesis,21
TGF-ß1, and follistatin, a myostatin inhibitor,20
because these factors could participate as co-regulators in the fibrosis reaction observed. TGF-ß1 (data not shown) and follistatin (data not shown) were both expressed in the connective tissue of ADAM12+/mdx and mdx mice. However, we did not observe any difference in expression of either TGF-ß1 or follistatin in ADAM12+/mdx mice compared with mdx mice. Thus, expression of these two factors seems to have no apparent correlation to the ADAM12+/mdx phenotype observed in this study. However, in contrast to mdx mice, most of the myofibers in the ADAM12+/mdx mice expressed myostatin.
|
|
These results implied that long-term expression of ADAM12 in mdx dystrophic mice compromised muscle restoration seen as reduced stem cell function, increased inflammation, fibrosis, and adipogenesis. The mechanisms underlying this could be exhaustion of the stem cell population; however, the expression patterns seen, in particular the myostatin expression, could suggest that the ADAM12+/mdx mice had an overall suppressed regeneration.
Membrane stabilization, rather than an increase in regeneration capacity, was proposed to be the mechanism behind early alleviation of the dystrophic phenotype in the ADAM12+/mdx mice.10
We therefore analyzed the expression of utrophin (Figure 2)
and ß1D integrin (data not shown) in the membranes and found that expression of utrophin seemed mildly elevated in the aged ADAM12+/mdx mice compared with age-matched controls. Expression of ß1D integrin reflected the fibrosis reaction observed in the Sirius stain (data not shown). Thus a down-regulation of these compounds could not account for the late dystrophic phenotype.
|
. In brief, we found few pax7+ve but many myogenin+ve nuclei in 6-week-old ADAM12+/mdx compared with age-matched mdx mice. Identical staining patterns were observed for NCAM and myostatin in fibers from both genotypes. Although the presence of many CD45+ve nuclei indicated pronounced inflammation in mdx mice, a Sirius staining revealed more fibrosis in the young ADAM12+/mdx mice than in the mdx mice. These results altogether suggested an altered regeneration in the young ADAM12+/mdx mice, too. The Regeneration Process Is Impaired in ADAM12 TG Mice after Severe Injury
To address the question about stem cell function in an ADAM12-enriched milieu, we studied the regenerative response after a stab lesion in 10- to 12-week-old ADAM12+ mice. Regeneration after freeze injury has previously been reported for ADAM12+ mice10
but not with specific focus on myogenic cell parameters. The knife cut injury resulted in a total loss of architecture of the injured muscle, but on day 5, the injured area in the control mice had regained most of its original structure (Figure 3A)
, and on day 10 (Figure 3B)
the regeneration was almost complete in these animals. The only indication of a previous injury was the centrally located nuclei in the repaired fiber segments (Figure 3B)
. By contrast, no obvious regeneration had occurred in the ADAM12+ mice on day 5 (Figure 3C)
, and on day 10 their muscle structure was still incomplete, although some regenerative activity could be observed (Figure 3D)
. These findings strongly indicated an impaired regenerative capacity in the ADAM12+ mice compared with their LCs. To analyze this effect further, immunohistochemical stainings and RT-PCR were performed.
|
; however, the difference between the two groups was not statistically distinguishable (P value, 0.104). Compared with uninjured control muscle, the fold-expression of pax7, myoD, and myogenin mRNA differed significantly between ADAM12+ and LC mice (Figure 5, A–C
; P < 0.01 for all three markers). In ADAM12+ mice, the mRNA expression of pax7 (Figure 5A)
varied little during the regeneration period. By contrast, the LCs displayed an up-regulation of pax7 initiating a few hours after surgery, peaking at days 3 to 5 and declining slowly during the later phases of regeneration (Figure 5A)
. The same expression pattern was observed for myoD (Figure 5B)
. The difference in pax7 expression was also visualized immunohistochemically (Figure 6A
, P < 0.01). These data indicated that the recruitment of activated satellite cells/proliferating myogenic cells was depressed in ADAM12+ mice compared with the controls.
|
|
|
and by IHC (Figure 6B)
revealed that fold expression of myogenin mRNA when compared with uninjured control muscle was significantly different between ADAM12+ and LC mice (Figure 5C
, P < 0.01). The analysis of cells expressing myogenin protein in their nuclei showed a similar expression pattern (Figure 6B
, P < 0.05). Thus, fewer differentiating myogenic cells were present during regeneration in the ADAM12+ mice, which supported the result from the pax7 and myoD analyses.
The expression pattern of NCAM followed that of pax7, myoD, and myogenin on both mRNA and protein level during the regeneration process in LC mice (Figures 5D and 6C)
. In the ADAM12+ mice (Figure 5D)
we observed a constitutive expression of NCAM mRNA, and the difference in NCAM mRNA expression patterns between LC and TG mice were supported statistically (P < 0.0001) (Figure 5D)
. The number of NCAM+ cells during regeneration recorded by IHC appeared reduced in the ADAM12+ mice compared with the controls even though no significant difference could be observed when the entire observation period was included (P = 0.0265).
Myotubes were visualized by NCAM IHC (Figure 7, A and B)
. On day 5, the LC mice showed regenerating areas filled with NCAM+ myotubes (Figure 7A)
. Such fibers were scarcely present in ADAM12+ mice; instead, we observed mononuclear, nonfusing NCAM+ cells (Figure 7B)
. The NCAM patterns thus indicated a difference in the temporal development of regeneration in ADAM12+ mice compared with controls, with a delayed fusion and myotube formation in ADAM12+ mice.
|
The analyses of ki67, pax7, myoD, myogenin, and NCAM expression indicated that satellite cell activation and subsequent differentiation was repressed during skeletal muscle regeneration in the ADAM12+ mice. The morphological investigations (Figure 3)
also supported these findings, suggesting that the normal regeneration was substituted by connective tissue formation. To confirm this, a Sirius red staining was performed (Figure 8)
.
|
but could not be observed in the control mice (Figure 8, D–F)
. The ADAM12+ mice displayed massive connective tissue reaction within the regenerating area on day 5 (Figure 8B)
, whereas the control mice formed new myotubes and only normal amounts of connective tissue at the same time point (Figure 8E)
. The connective tissue component was still very prominent day 10 after injury in the ADAM12+ mice (Figure 8C)
, whereas the LCs completely regenerated the destroyed muscle tissue leaving no excess connective tissue at day 10 (Figure 8F)
. Differences in the late stages of regeneration could also be seen from the dystrophin IHC staining. The control mice had at day 10 regained their muscle architecture completely (Figure 8F
, inset), whereas the ADAM12+ mice displayed a compromised regeneration and less structured skeletal muscle compartment at this time point (Figure 8C
, inset). |
Discussion |
|---|
|
TopAbstractExperimental ProceduresResultsDiscussionReferences |
|---|
The mechanism of ADAM12 alleviation in the young mdx mice has been linked to the up-regulation of utrophin,10 a functional homologue of dystrophin, known to reduce the myopathic effect resulting from lack of dystrophin in mdx muscle.24 Although there continues to be mildly elevated utrophin expression in aged ADAM12+/mdx mice compared with their age-matched mdx controls, it seems from our studies that the ADAM12 induced inhibition of the regeneration throughout time overrules the positive effect of utrophin up-regulation. This may be attributable to the mild degree of up-regulation achieving only subtherapeutic levels. When a negative effect is not seen initially, this could be attributable to a higher regeneration capacity in young animals in combination with the retardation of the degeneration provided by utrophin up-regulation.
Previously a positive effect of ADAM12 on regeneration has been reported.10 Interestingly, it has been found that different types of muscle injury may result in different regenerative responses.25 The study in which ADAM12 was reported to support regeneration10 used smaller freeze injuries, which leaves the basement lamina as a scaffold, whereas the lesions in our present study completely disrupted the basal lamina. This difference could to some extent explain the observed difference because the conserved basal lamina scaffolds in the smaller freeze injuries could create a more myogenic milieu and thus assist the myogenic cells26 in muscle repair. Moreover, ADAM12 is a membrane-spanning protein, which in itself could provide a structural membrane support when the basement membrane is intact.
The skeletal muscle of mice transgenically overexpressing ADAM12 demonstrates a slightly myopathic pattern observed as centrally placed nuclei, scattered necrosis, and accumulation of adipose tissue in the muscles.10 In our study, we found a constitutive expression of NCAM mRNA, further supporting the myopathic phenotype of the ADAM12+ transgenic muscles. If ADAM12 has an inhibitory effect on regeneration, it could explain this observed myopathic phenotype. The negative effect would accumulate throughout time and become accelerated when combined with the mdx phenotype, resulting in an overt dystrophic phenotype. Thus, the ADAM12+/mdx mouse seems to develop its more severe myopathy because of inhibition of the recruitment of progenitor cells and fusion/maturation of myogenic cells into fibers.
Lack of, or mutations in proteins necessary for maintenance of the mature muscle fiber integrity are common defects observed in different kinds of myopathies.27 The present observations suggest inhibition of stem cell activation as a novel mechanism that could lead to myopathy/dystrophy. It has previously been observed that transgenic expression of ADAM12 in skeletal muscle results in the occurrence of adipocytes within the muscles.17 In our study we observed fibrosis and adipose accumulation in an environment where the satellite cells are not lost but rather inhibited, thus indicating that adipogenesis is not linked to loss of myogenic stem cells/precursors.
ADAM12 has been presented as a possible candidate in nonreplacement gene therapy of muscular dystrophies providing structural support alleviating the mdx phenotype.7 However, as observed in the present study, a structural support is not enough if the satellite cell response is constantly depressed not allowing an effective regeneration to occur. In long-term disease, this will eventually result in adipogenesis and fibrogenesis. The use of ADAM12 as booster gene thus needs to be revised. Perhaps, ADAM12 could be useful in short-term boosting to yield a short-term inhibition of satellite cell recruitment combined with a structural support both directly and through utrophin up-regulation to protect the stem cell population against exhaustion. However, based on the present results, long-term treatment is not recommended.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by The Lundbeck Foundation, the Carlsberg Foundation, and the Danish Muscle Dystrophy Organization.
Accepted for publication August 2, 2007.
|
References |
|---|
|
TopAbstractExperimental ProceduresResultsDiscussionReferences |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
