| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
2-deficient mice
Biozentrum, University of Basel, Basel, Switzerland
1Correspondence: Biozentrum, University of Basel, Klingelbergstrasse 70, CH-4056 Basel, Switzerland. E-mail: markus-a.ruegg{at}unibas.ch
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2 subunit of laminins cause the severe "merosin-deficient congenital muscular dystrophy" (MDC1A). We have recently shown that overexpression of a miniaturized form of the molecule agrin (mini-agrin) counteracts the disease in dyW/dyW mice, a model for MDC1A. However, these mice express some residual truncated laminin-2, suggesting that the observed amelioration might be due to mini-agrin’s presenting the residual laminin-2 to its receptors. Here we show that the mini-agrin counteracts the disease in dy3K/dy3K mice, which are null for laminin-2. As in dyW/dyW mice, mini-agrin improves both the function and structure of muscle. We show that muscle regeneration after injury is severely impaired in dy3K/dy3K mice but is restored in the mini-agrin-expressing littermates. In summary, our results 1) show that the direct linkage of muscle basal lamina with the sarcolemma is the basis of mini-agrin-mediated amelioration and 2) provide unprecedented evidence that this linkage is important for proper regeneration of muscle fibers after injury. Our findings thus suggest that treatment with mini-agrin might be beneficial over the entire spectrum of the MDC1A disease, whose severity inversely correlates with expression levels and the size of the truncation in laminin-2.—Bentzinger, C. F., Barzaghi, P., Lin, S., Ruegg, M. A., Overexpression of mini-agrin in skeletal muscle increases muscle integrity and regenerative capacity in laminin-2-deficient mice.
Key Words: muscular dystrophy • laminin-2 deficiency • muscle regeneration
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2 chain (previously called merosin) and is characterized by early onset, severe hypotonia, demyelination of the peripheral nerve, and high levels of creatine kinase in the blood (for reviews, see refs 1
, 2
). Skeletal muscle of MDC1A patients is characterized by variations in fiber size, extensive fibrosis, and proliferation of adipose tissue. Despite alterations in the white matter of the central nervous system, the intellectual capacity of patients is often not affected. In contrast, demyelination of the peripheral nerve contributes to the severity of the disease. The availability of mice carrying mutations in the lama2 gene has greatly helped us to understand the mechanisms underlying the disease. These mice can be the result of spontaneous mutation, such as dystrophia muscularis (dy/dy) and the dy2J/dy2J mouse (3
, 4)
, or of targeted deletions in the lama2 gene, such as the dyW/dyW and the dy3K/dy3K mouse (5
, 6)
. The different mouse strains can be distinguished based on the severity of the disease. This difference is due to residual expression of truncated forms of laminin-2 chain in some of the mice (7)
. Similar variations have been reported in MDC1A patients (8)
.
Laminins belong to a rather large family of heterotrimeric, cruciform-like molecules composed of one , one ß, and one 
chain. So far, five , three ß, and three chains have been identified, giving rise to 15 different laminin isoforms (for a review, see ref 9
). The two key receptors to which the laminins bind are the integrins and dystroglycan (reviewed in ref 10
). Laminin-2 and laminin-4 both contain the 2 chain and are the main laminin isoforms expressed in skeletal and cardiac muscle, in peripheral nerve, and in the brain. In laminin-2-deficient muscle from mice or human patients, the laminin-4 chain is strongly up-regulated (11
12
13)
.
Laminin-4 is a shorthand chain, which makes it unlikely that 4 chain-containing laminins would polymerize. Moreover, it binds to -dystroglycan with low affinity (14)
. Loss of only one of these two functions is enough to cause muscle dystrophy, as seen in dy2J/dy2J mice that synthesize an N-terminally truncated form of the laminin-2 chain that binds to -dystroglycan (15
, 16)
. These mice still suffer from muscular dystrophy, but their phenotype is clearly less severe than that of dy/dy mice, which express only a little laminin-2. Thus, the binding of laminin-2 and -4 to cell surface receptors and their self-polymerization activity seem to be equally important to maintain muscle function.
On a cellular level, muscular dystrophies are often characterized by extensive degeneration and regeneration of muscle fibers. The successive replacement of muscle by nonmuscle tissue could be a consequence of depletion of the pool of satellite cells over time or a problem with muscle fiber regeneration per se. For example, muscle in mdx mice, a model for dystrophin-deficient Duchenne muscular dystrophy, is characterized by a high percentage of muscle fibers with centralized nuclei indicative of successful regeneration. In contrast, the increase in centralized myonuclei is not a distinctive feature in laminin-2-deficient muscle. Rather, muscle regeneration seems to be delayed, and it has been suggested that the muscle undergoes apoptosis during regeneration (17)
.
There is no therapy to treat MDC1A. However, interesting approaches have been taken in mouse models. For example, transgenic expression of the human laminin-2 chain in skeletal muscle improves the function of the muscle in dyW/dyW mice (5)
. Similarly, overexpression of a cDNA encoding mouse laminin-1 using the chicken ß-actin promoter profoundly improves the disease in dy3K/dy3K mice (18)
. Given the high structural similarity between the laminin-2 and the 1 chain, these results are not surprising, but may open new avenues for treating MDC1A patients. In contrast to the approaches with highly homologous proteins, we used a miniaturized form of the extracellular matrix molecule agrin specifically designed to contain high-affinity binding sites for the laminins (19)
and for -dystroglycan (20)
. By transgenic overexpression of such mini-agrin in muscle, we showed that it improved muscle function and overall health of dyW/dyW mice (13)
. We hypothesized that this improvement of muscle function is due to mini-agrin’s providing a link between laminin-8 (4, ß1, 1), which is up-regulated in dyW/dyW mice, and -dystroglycan. In agreement with this hypothesis, we found a stabilization of -dystroglycan at the muscle membrane. Moreover, a slight but significant increase of laminin-5 was detected in dyW/dyW mice that expressed the mini-agrin transgene. As laminin-5 can form a laminin network, this increase may have accounted for the re-establishment of the structure of the basement membrane.
The recent discovery that dyW/dyW express a truncated version of laminin-2 at low levels (7)
opened the possibility that the observed amelioration by the mini-agrin might be indirect. For example, mini-agrin could sequester the truncated laminin-2 in the muscle basement membrane and present it to its natural receptors such as -dystroglycan and the integrins. To test this, we investigated whether the mini-agrin transgene would have a beneficial effect in dy3K/dy3K mice devoid of laminin-2 expression (6
, 7)
. We show that the mini-agrin indeed improves the function, survival, and structure of muscle, thus providing strong support that mini-agrin acts directly. We show that the overexpression of mini-agrin greatly improves the regenerative capacity of laminin-2-deficient muscle upon injury. Thus, the mini-agrin not only serves as a structural anchor to prevent muscle degeneration but also activates intracellular signals that allow successful regeneration and prevent cell death. Moreover, this activity of the mini-agrin does not require the presence of residual levels of laminin-2. Thus, the use of mini-agrin could be a universal tool for treatment of MDC1A irrespective of the residual levels of laminin-2 expressed in the patient.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2-deficient mice
. The dy3K/dy3K mice were a kind gift from Drs. Shin’ichi Takeda and Yuko Miyagoe-Suzuki (6)
. Genotyping of heterozygous and homozygous dy3K mice was done by PCR with one primer within the neocassette, 5'-CCCGTGATATTGCTGAAG-3', and two primers specific for the exons in the lama2 gene: 5'-CAGGTGTTCCAGATTGCC-3' and 5'-CCTCTCCATTTTCTAAAG-3'. Unless specifically indicated, all experiments were done on 4-wk-old mice.
Locomotion
Locomotory behavior in animals of different genotypes was determined as described elsewhere (13)
. In brief, mice were placed into a new cage and motor activity was measured for 10 min. All movements such as walking, digging, or righting up were included. At least four animals of each genotype were tested.
Histology, immunohistochemistry, and antibodies
Muscles were immersed in 7% gum Tragacanth (Sigma, St. Louis, MO, USA) and rapidly frozen in liquid nitrogen-cooled isopentane (–150°C). Cross sections of 12 µm thickness were cut. General histology was performed using hematoxylin and eosin (H&E) (Merck, Rayway, NJ, USA). Membrane-bound and extracellular epitopes were visualized with Alexa-488- or Alexa-594-conjugated wheat germ agglutinin (WGA; Molecular Probes, Eugene, OR, USA). TUNEL staining was performed with TUNEL Enzyme and TUNEL Label solution (Roche Diagnostics, Nutley, NJ, USA) according to the manufacturer’s protocol for cryopreserved tissue. Polyclonal rabbit anti-mouse laminin-4 (Ab 377), polyclonal rabbit anti-mouse laminin-5 (Ab 405) and monoclonal rat anti-mouse laminin-1 (Ab 198) were all kind gifts from Dr. L. Sorokin, Lund University. Polyclonal sheep anti-mouse -dystroglycan was a kind gift from Dr. S. Kröger (University of Mainz) and the rat monoclonal antibody against mouse tenascin-C (21)
was given to us by R. Chiquet-Ehrismann (Friedrich Miescher Institute for Biomedical Research). The remaining antibodies were produced in-house or purchased from commercial sources as follows: monoclonal mouse anti-mouse ß-dystroglycan (Novocastra, Newcastle upon Tyne, UK), monoclonal rat anti-mouse NCAM (Chemicon, El Segundo, CA, USA), monoclonal mouse anti-rat developmental myosin heavy chain (dMyHC; Novocastra), monoclonal rat anti-mouse laminin-1 chain (Chemicon), polyclonal rabbit anti-chick agrin (Ab 3228; ref 22
). Depending on the source of the primary antibody, appropriate Cy3-conjugated (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) or Alexa488-conjugated (Molecular probes) secondary antibodies were used. 4',6'-Diamidino-2-phenylindole hydrochloride (DAPI) was used to stain for nuclei.
Reverse transcription and real-time PCR
Total RNA was isolated (FastRNA Pro Green Kit, Q-biogene) separately from both quadriceps muscles of three mice from each genotype. Reverse transcription was performed according to the manufacturer’s recommendation using 5 µg total RNA, with SuperScript II reverse transcriptase (Invitrogen, San Diego, CA, USA) and random hexamers as primers. For the subsequent Taqman real-time PCR, the following primer pairs and probes were used: laminin-4: 5'-AGCCGTGCCCCTGTCC-3', 5'-CCATTTTTCCTGTAGCAGGATTCT-3' and as probe 5'FAM-TTGCCCCACCTGGCCAACTTTG-TAMRA3'; laminin-5: 5'-CATTGGTCGTGTGCGCAA-3', 5'-GGATGTGGAACTTGAGGGCA-3' and as a probe 5'FAM-AGCAAGGTCAAGGTGTCCATGAAGTTCAATG-TAMRA3'; dystroglycan: 5'-CTACACCTGTTACTGC-3', 5'-ATGTAGGCGTAGGCACTATCCGCG-3' and as a probe 5'FAM-CCCAACCACGGCCATTCAGGAG-TAMRA3'. Real-time PCR using primers and probes for skeletal ß-actin and muscle creatine kinase was performed for normalization as described elsewhere (23
, 24)
. For the analysis, an ABI Prism 7700 Sequence Detector in conjunction with the TaqMan PCR Reagent Kit (Applied Biosystems, Foster City, CA, USA) was used. Expression levels for each gene of interest were normalized to the mean cycle number from real-time PCR of the two housekeeping genes. In all experiments, the ratio between the two housekeeping genes was constant (see Supplementary Fig. 1).
Quantifications
For all quantifications, pictures were collected using a Leica DM5000B fluorescence microscope, a digital camera (F-View; Soft Imaging System, Lakewood, CO, USA), and analySIS® software (Soft Imaging System). Muscle fiber size distribution was determined on WGA-stained muscle cross sections using the minimum distance of parallel tangents at opposing particle borders (minimal "Feret’s diameter") as described elsewhere (25)
. Measurement of minimal Feret’s diameter of notexin-treated muscle was done on cross sections stained for laminin-1. Normalization of the number of fibers in each fiber feret class and the number of TUNEL-positive nuclei, NCAM-positive muscle fibers, or dMyHC-positive was based on the total number of muscle fibers on each picture. In all quantification experiments, a minimum of 4000 muscle fibers per mouse and at least three mice of each genotype were analyzed.
Muscle regeneration
Skeletal muscle was injured by injection of the myotoxin notexin (Sigma) into the tibialis anterior of 23-day-old mice. Mice were anesthetized and an 
3 mm-long incision into the skin was made at the proximal end of the muscle. A Hamilton micro syringe was inserted and advanced to the distal end of the muscle. Seven microliters of notexin (50 µg/mL) were injected slowly into the muscle while pulling the needle back to assure delivery of the myotoxin along the entire muscle. Mice were killed by CO2 asphyxiation 5 days later; muscles were isolated and processed as described above.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2 chain (7)
. In our hands, the disease was less severe in dyW/dyW mice (5)
, which still express low levels of an N-terminally truncated form of laminin-2, than in dy3K/dy3K mice, which are completely null for laminin-2 (6)
. For example, dy3K/dy3K died earlier and showed a more pronounced degeneration of muscle (data not shown; see also Fig. 1
). To evaluate whether expression of the mini-agrin transgene ameliorates dystrophy in dy3K/dy3K mice, we mated the mini-agrin transgenic mice with mice heterozygous for the lama2 mutation. Offspring that were heterozygous for the lama2 mutation and carried the mini-agrin transgene were then further mated to eventually give rise to dy3K/dy3K mice that express the mini-agrin transgene (dy3K/mag-tg). Histological staining of several muscles including triceps brachii revealed fewer signs of degeneration in dy3K/mag-tg mice than in dy3K/dy3K mice (Fig. 1A
). As in the dyW/mag-tg mouse (13)
, the transgenic mini-agrin protein was deposited surrounding muscle fibers (Fig. 1A
, right panel). Next, we quantified the extent of amelioration by determining fiber size distribution. We measured the minimal Feret’s diameter, which represents the minimal distance of parallel tangents of a particle (25)
. Fibers were grouped into different size classes and the number of fibers in each class was determined. As shown in Fig. 1
, in triceps brachii (Fig. 1B
) and tibialis anterior (Fig. 1C
), the size distribution of the muscle fibers was shifted toward smaller classes in dy3K/dy3K (squares, dotted line) compared with wild-type mice (triangle, solid line). Minimal Feret’s diameter values from dy3K/mag-tg mice (circles, interrupted line) were distinct from dy3K/dy3K mice and were shifted toward wild-type mice. In tibialis anterior muscle (Fig. 1C
), we observed a significant increase in the number of large, potentially hypertrophic fibers in dy3K/mag-tg mice. The significant amelioration in the overall function of skeletal muscle was obvious in experiments measuring locomotory activity (Fig. 1D
): dy3K/mag-tg mice performed almost as well as wild-type mice whereas dy3K/dy3K mice performed poorly.
|
Transcriptional regulation of protein expression in dystrophic mice
The amount of laminin-4, laminin-5, and dystroglycan detected in muscle differs between wild-type, dyW/dyW, and dyW/mag-tg mice (13)
. To see whether similar changes were observed in the dy3K/dy3K mouse model and test whether any of the changes were due to transcriptional regulation, we performed immunofluorescence experiments in combination with quantitative reverse transcription real-time PCR (QRT-PCR) on quadriceps muscle. Compared with wild-type mice, staining intensity for laminin-4 was increased in dy3K/dy3K mice whereas no change for laminin-5 was observed (Fig. 2
A). Expression of the peripheral membrane protein -dystroglycan was strongly diminished in dy3K/dy3K mice although the expression level of the transmembranous ß-dystroglycan remained the same. Overexpression of mini-agrin in dy3K/dy3K did not affect laminin-4 and ß-dystroglycan levels (Fig. 2A
). However, laminin-5 and -dystroglycan were increased (Fig. 2A
). Thus, the changes in protein levels were similar to those observed on the dyW/dyW background (13)
. To evaluate whether these changes were based on alterations in transcription, we performed QRT-PCR using mRNA encoding ß-actin and muscle creatine kinase (MCK) for normalization. The two transcripts did not significantly differ between mice with different genotypes (see Supplementary Fig. 1). Thus, the relative contribution of muscle fibers (expressing MCK and ß-actin) and nonmuscle cells (expressing only ß-actin) to the total mRNA expressed in quadriceps muscle does not change significantly between mice with different genotypes. As shown in Fig. 2B
, mRNA encoding dystroglycan (black bars) was not changed. In contrast, laminin-4 mRNA expression was > 2-fold increased in dy3K/dy3K and dy3K/mag-tg mice compared with wild-type controls (Fig. 2B
, hatched bars). Similar to dystroglycan, no changes in mRNA expression of laminin-5 were observed in either dy3K/dy3K or dy3K/mag-tg mice (Fig. 2B
, open bars). These results indicate that the changes of laminin-5 and -dystroglycan in dy3K/mag-tg mice are likely due to posttranslational stabilization of the two proteins.
|
Mini-agrin prevents degeneration and promotes muscle regeneration upon injury
Among the predominant signs of muscle dystrophy in MDC1A and animal models is the high percentage of fibrotic tissue indicative of impaired regeneration. We therefore asked whether the beneficial effect of mini-agrin was based on an increased tolerance of muscle fibers to mechanical load and/or on an improved regenerative capacity. We stained muscle cross sections for neural cell adhesion molecule (NCAM) and developmental myosin heavy chain (dMyHC), both of which are expressed in regenerating but not mature muscle fibers. As shown in Fig. 3
A, muscle of dy3K/dy3K mice contained many NCAM-positive cells whereas dy3K/mag-tg and wild-type mice had only a few or no NCAM-positive cells. The same genotype-phenotype correlation was observed with dMyHC (Fig. 3B
). Quantification of NCAM- and dMyHC-positive fibers is shown in Fig. 3C
. We found that the number of immuno-positive fibers was significantly higher in triceps than in tibialis anterior. Moreover, the number of dMyHC-positive fibers was always lower than those positive for NCAM. Because NCAM, but not dMyHC, expression is regulated by innervation (26)
, the higher number of NCAM-positive fibers is probably because of the demyelination of peripheral nerves in dy3K/dy3K and dy3K/mag-tg mice. Most important, however, the number of NCAM- and dMyHC-positive muscle fibers in triceps and tibialis anterior was > 2-fold lower in dy3K/mag-tg than in dy3K/dy3Kmice (Fig. 3C
). Apoptosis of muscle fibers is a prominent feature of laminin-2 deficiency and is thought to reflect muscle degeneration (6
, 17)
. To evaluate whether mini-agrin affected these processes, we used TUNEL staining and found that the number of TUNEL-positive fibers in dy3K/mag-tg mice was < 40% of that detected in dy3K/dy3K mice (Fig. 3D
). Apoptotic myonuclei were not particularly enriched in NCAM- or dMyHC-positive fibers (data not shown). Finally, we measured the extent of fibrosis in the different mouse models using antibodies against tenascin-C, a specific marker for fibrosis and inflammation in adult muscle (12)
. As expected, muscle of wild-type mice was devoid of any staining whereas some tenascin-C-positive regions were detected in dy3K/mag-tg and prominent staining was observed in dy3K/dy3K mice (Supplementary Fig. 2
). Together, these results support the notion that mini-agrin is protective for the muscle fibers during contraction and thus prevents degeneration.
|
To determine the regenerative capacity of muscle in the different mouse models directly, we challenged muscles by forced injury. We destroyed the tibialis anterior muscle of 3-wk-old animals by injection of notexin, a snake venom compound with phospholipase A2 activity that provokes transient necrosis of mature muscle followed by vigorous myofiber regeneration (27
, 28)
. Five days after injury, muscles were isolated and analyzed. Although only a few small dMyHC-positive myofibers (red) were seen in dy3K/dy3K mice (Fig. 4
A), many more dMyHC-positive fibers that were considerably larger could be detected in dy3K/mag-tg and wild-type (wt) mice. In these two latter mice, most of the nuclei were localized within regenerating myofibers. In contrast, many nuclei were localized outside of regenerating muscle fibers in dy3K/dy3K muscle, suggesting that at the time of examination most of the cells were nonmuscle cells or muscle precursors. Indeed, WGA, a lectin that recognizes N-acetylglucosamine-containing epitopes in membrane proteins and in the extracellular matrix, stained these cells strongly (Fig. 4B
). In contrast, the large cytoplasm of regenerating muscle fibers was not stained in dy3K/mag-tg and wild-type mice (asterisks in Fig. 4B
). H&E staining of regenerating muscle confirmed the preponderance of mononucleated cells and the presence of only small muscle fibers in dy3K/dy3K mice (Supplementary Fig. 3). In contrast, regeneration was nearly complete in dy3K/mag-tg and wt mice (Supplementary Fig. 3). Thus, overexpression of mini-agrin can restore the capacity of laminin-2-deficient muscle to regenerate almost to wild-type levels.
|
The increased regenerative capacity of muscle in dy3K/mag-tg mice became apparent upon quantitative assessment of the minimal Feret’s diameter. Whereas the majority of muscle fibers in dy3K/dy3K mice had a minimal Feret’s diameter between 5 and 10 µm (Fig. 5
A, squares, dotted line), the majority of the regenerating fibers in dy3K/mag-tg (Fig. 5A
, circles, interrupted line) and wild-type (Fig. 5A
, triangles, solid line) had minimal Feret’s diameters between 15 and 20 µm. Regenerating muscle fibers, including small ones (arrowhead), were strongly positive for mini-agrin-like immunoreactivity (Fig. 5B
) and for laminin subunits (Fig. 5C
). No difference in the staining pattern was observed for the laminin-1 chain (Fig. 5C
, top), whereas expression of laminin-4 was strongly increased in dy3K/dy3K and dy3K/mag-tg mice compared with wild-type controls. Regenerating muscle showed an increased expression of laminin-5 in wild-type and dy3K/mag-tg mice but only a slight increase in dy3K/dy3K mice (Fig. 5C
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. A recent overview of published clinical descriptions of MDC1A patients indicates a strong variability in the severity of the disease (8)
. As in the different mouse models for MDC1A, which show differences in disease severity, the difference in the amount of laminin-2 or truncated versions expressed is probably the reason for the difference in the phenotype of human patients (8)
. In our previous work, we showed that mini-agrin expressed in dyW/dyW mice ameliorates the disease considerably (13)
. These mice still express low amounts of truncated laminin-2 and show a less severe phenotype than dy3K/dy3K mice (ref 7
and data not shown). We now provide evidence that mini-agrin improves the health of the most severely affected dy3K/dy3K mice. Thus, improvement of muscle function by mini-agrin is independent of any residual expression of laminin-2.
Although the amelioration in dy3K/dy3K mice is considerable, the extent of improvement appeared to be less than that in dyW/dyW mice (13)
. This could be seen clearly from the overall appearance of the dy3K/mag-tg mice, which were smaller and died earlier than dyW/mag-tg mice (data not shown). This suggests that mini-agrin does not reverse all of the disease-causing pathology. A major factor contributing to this is the lack of any mini-agrin expression outside of skeletal muscle. Tissues that contribute to the pathology in MDC1A, but where mini-agrin is not expressed, are peripheral nerve, the brain, and the heart (see ref 13
for a discussion). Thus, in these tissues the complete loss of laminin-2 is likely to worsen the disease phenotype when compared with dyW/dyW mice. Another possibility is that interactions of the laminin-2 chain with receptors that are not shared with the mini-agrin are important for the function of skeletal muscle. Interesting candidates for this are the 7 integrins, known to be involved in muscle dystrophy (29
, 30)
. Thus, the presence of an N-terminally truncated laminin-2 chain in dyW/dyW mice may bind to the integrins. Mini-agrin has not been shown to bind to 7 integrins and therefore may not compensate for the loss of this interaction in dy3K/dy3K mice. Indeed, the variation in fiber size is not completely restored to wild-type levels in 4-wk-old dy3K/mag-tg mice (Fig. 1)
but normal in dyW/mag-tg mice of the same age (13)
.
It is generally thought that muscle dystrophies are caused by the failure of muscle fibers to sustain mechanical load. This leads to muscle damage, subsequent increase in intracellular calcium concentration, and muscle death. The regenerative capacity of muscle will then allow the formation of new muscle fibers. Because of the reoccurring damage of muscle fibers, dystrophic muscle will undergo frequent cycles of degeneration/regeneration that deplete the pool of available satellite cells over time. The lack of satellite cells will result in muscle wasting and replacement of muscle by nonmuscle tissue. An alternative explanation for muscle wasting would be that the process of regeneration per se is not working properly. In agreement with what has been shown in dyW/dyW mice (17)
, regeneration is severely hampered in dy3K/dy3K mice (Fig. 4)
. Thus, a loss of the capability of muscle to regenerate appears to be a major cause of muscle wasting in laminin-2-deficient mice. Nevertheless, regenerating muscle fibers could still be detected in noninjured muscle of dy3K/dy3K mice (Fig. 3)
. Mechanistically, this could be based on a partial compensation by laminin-4, which is overexpressed in regenerating fibers (Fig. 5C
). Besides its importance for allowing the formation of heterotrimeric laminin in the absence of laminin-2, laminin-4 can bind to integrins, an interaction that may support regeneration. An involvement of integrins in myoblast fusion and sarcomere assembly has been shown in experiments in which all ß1 integrins were inactivated in developing muscle (31)
. Thus, the binding of laminin-4 to integrins may explain why dy3K/dy3K mice are still capable of some regeneration. It is less likely that the low number of regenerating muscle fibers after injury in dy3K/dy3K mice is due simply to a delay in regeneration. The considerable number of apoptotic cells found in nonchallenged muscle of dy3K/dy3K mice (Fig. 3D
) and the fact that dyW/dyW mice whose muscles were injured by freeze-thawing show signs of incomplete regeneration even 2 wk after injury (17)
strongly suggest that regeneration might be abortive and not just delayed in dy3K/dy3K muscle.
Our notexin injection experiments show that overexpression of mini-agrin restores the regenerative capacity of dy3K/dy3K muscle almost to wild-type levels (Fig. 4
and Fig. 5B
). Several lines of evidence strongly suggest that this activity can be directly attributed to the high-affinity binding of mini-agrin to -dystroglycan. Muscle in which satellite cells express normal levels of dystroglycan but lose expression after fusion undergo many cycles of degeneration and regeneration without pronounced signs of muscle dystrophy (32)
. In these mice, muscle regeneration after injury is indistinguishable from wild-type muscle and is much improved compared with mdx mice. Second, blockade of the -dystroglycan binding to laminin-2 with specific antibodies results in a high incidence of apoptosis in cultured myotubes (33)
. Although it is controversial whether -dystroglycan is a signaling receptor, recent work suggests that dystroglycan serves as a scaffold for ERK-MAP kinase signaling (34)
and that antibody-induced perturbation of the laminin binding to -dystroglycan, which prevents myoblast fusion, involves the PI3K/Akt(PKB) pathway (33)
. In summary, our results support the notion that engagement of -dystroglycan with its ligands is necessary for proper muscle regeneration. For example, binding to -dystroglycan might be necessary for the attachment of satellite cells to the basal lamina. This attachment would then assure the renewal of muscle fibers upon injury and eventually prevent cell death. The concomitant increase in laminin-5 expression may further support the function of mini-agrin by assembling a laminin network.
Overexpression of mini-agrin in dy3K/dy3K mice increased the amount of laminin-5. This increase of protein is regulated posttranscriptionally (Fig. 2)
. As laminin-5 is not truncated at the N terminus, it is highly probable that it can form a laminin network. Thus, the increased amount of laminin-5 in dy3K/mag-tg mice may be important to warrant proper formation and stability of basement membrane in regenerating and mature muscle. However, laminin-5 does not strongly bind to -dystroglycan (35)
and the amount of laminin-5 in the basement membrane of the dy3K/mag-tg mice is low. We therefore hypothesize that some of the improvements observed in dy3K/dy3K mice are a direct consequence of mini-agrin acting on the sarcolemma via its binding to -dystroglycan rather than an indirect consequence of the increased amount of laminin-5.
Taken together, our data provide strong evidence that the mechanisms that protect the muscle fibers against mechanical load and allow regeneration of satellite cells are greatly impaired in muscular dystrophies due to laminin-2 deficiency. Mini-agrin, via its binding to -dystroglycan and basement membrane, prevents muscle degeneration and greatly improves regeneration. The fact that mini-agrin influences both processes indicates that they are based on re-establishment of the linkage of -dystroglycan to muscle basement membrane and/or the recovery of ligand-mediated signaling. The high-affinity binding of agrin to the laminins via the NtA domain is likely to be important in maintaining sustained binding and/or signaling to -dystroglycan and for stabilizing laminin-10 (5,ß1,1) in the basement membrane. Our findings provide additional support for the fundamental role of -dystroglycan in the ontogeny and treatment of muscular dystrophies. Further work in our laboratory is devoted to establishing whether the expression of the mini-agrin would ameliorate the dystrophic phenotype when applied at later stages of the disease and to testing whether attachment of mini-agrin to the basement membrane is required for this effect.
|
ACKNOWLEDGMENTS |
|---|
Received for publication November 12, 2004. Accepted for publication January 27, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. Xu, K. Chandrasekharan, J. H. Yoon, M. Camboni, and P. T. Martin Overexpression of the Cytotoxic T Cell (CT) Carbohydrate Inhibits Muscular Dystrophy in the dyW Mouse Model of Congenital Muscular Dystrophy 1A Am. J. Pathol., July 1, 2007; 171(1): 181 - 199. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Meinen, P. Barzaghi, S. Lin, H. Lochmuller, and M. A. Ruegg Linker molecules between laminins and dystroglycan ameliorate laminin-{alpha}2-deficient muscular dystrophy at all disease stages J. Cell Biol., March 26, 2007; 176(7): 979 - 993. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Scotton, D. Bleckmann, M. Stebler, F. Sciandra, A. Brancaccio, T. Meier, J. Stetefeld, and M. A. Ruegg Activation of Muscle-specific Receptor Tyrosine Kinase and Binding to Dystroglycan Are Regulated by Alternative mRNA Splicing of Agrin J. Biol. Chem., December 1, 2006; 281(48): 36835 - 36845. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE |
