Search

The examination was performed whether aquaporin (AQP) 9 is expressed in normal skeletal muscle at mRNA and protein levels. Gel electrophoresis of the reverse transcription-polymerase chain reaction (RT-PCR) product of total RNA samples of human normal muscles by oligonucleotide primers for human AQP9 showed a band of 221 basepairs, which corresponded to the basepair length between two primers of AQP9. The nucleotide sequence of RT-PCR product coincided with that of human AQP9. Immunoblot analysis revealed that the rabbit and sheep antibodies against the synthetic peptide of the C-terminal cytoplasmic domain of human AQP9 molecule reacted with a protein of approximately 30 kDa molecular weight in extracts of human normal skeletal muscles. Immunohistochemistry with our anti-AQP9 antibodies showed an immunoreaction at the myofiber surface of both type 1 and type 2 fibers with almost equal staining intensity in human skeletal muscles. The implication of AQP9 expression in skeletal myofibers was discussed.

We generated the muscle aquaporin 4 (AQP4) overexpressing transgenic mice in order to investigate the skeletal muscle pathology at RNA and protein levels. At RNA level, the AQP4 mRNA expression of soleus, EDL and cardiac muscles in Tg mice was statistically significantly higher than that in wild mice by the real-time reverse transcription polymerase chain reaction method. At protein level examinations, we used the immunoblot, immunohistochemistry and freeze-fracture electron microscopy. The immunoblot showed the single band of 31kDa with anti-AQP4 antibody in the extracts of soleus and EDL muscles of wild mice but not in extract of wild cardiac muscle; while the reaction band was noted in cardiac muscle of Tg mice and the reaction band was stronger in the extracts of soleus and EDL muscles of Tg mice. The immunohistochemistry showed that the expression of AQP4 at the myofiber surface of soleus and EDL muscles of Tg mice was more marked than that of wild mice and, interestingly, the AQP4 expression of these muscles of Tg mice appeared to be more remarkable in type 1 slow twitch myofibers as judged by the positive slow myosin immunostaining of adjacent serial sections. The immunofluorescence staining with anti-AQP4 antibody of cardiac muscles of wild mice revealed the scarcely immunopositive myofibers; whereas the immunostaining cardiac muscles of Tg mice contained the numerous AQP4 immunopositive myofibers. The freeze-fracture electron microscopy demonstrated that the orthogonal array densities in soleus and EDL muscle plasma membranes of Tg mice were significantly higher than those of wild mice and that the orthogonal array like particle density of cardiac muscle plasma membranes of Tg mice appeared to be more numerous than that of cardiac myofibers of wild mice. Finally the clinical phenotype of Tg mice appeared to be similar to that of wild mice. Further physiological examination with devices may suggest some about the physiological difference.

To examine aquaporin 1 (AQP1) expression in skeletal muscle tissue precisely, we performed reverse transcription-polymerase chain reaction (RT-PCR) at RNA level and immunoblot analysis, immunohistochemistry and immunoelectron microscopy at protein level. The RT-PCR study of total RNA from normal human skeletal muscle showed a strong single band of AQP1. At the protein level we used two commercially available antibodies, both of which recognize the cytoplasmic domain of the AQP1 molecule. One antibody gave positive results. Immunoblot of muscle extract showed a 30-kDa band protein, the molecular weight of which corresponded to that of AQP1. Immunohistochemically, AQP1 was immunostained at the myofiber surface both in type 1 and type 2 myofibers with almost the same intensity, and its staining pattern was rather diffuse and irregular compared with that of the anti-dystrophin antibody. The endomysial endothelial cells were also immunolabeled. Immunoelectron microscopy revealed that the immunogold particles indicating the presence of the AQP1 molecule were present along the inside surface of the muscle plasma membrane. However, another antibody showed negative results except for the endomysial endothelial cells which were positively stained. We drew the conclusion that AQP1 is expressed at the endomysial capillary endothelial cell and further AQP1 may be expressed at the human skeletal myofiber plasma membrane.

Myoferlin is a novel protein of unknown function with high homology to dysferlin, the gene mutations of which cause limb girdle muscular dystrophy type 2B and Miyoshi myopathy. The myoferlin gene seems to be a candidate for the modifier, and because of the high homology to dysferlin myoferlin may work as a compensator for the absence of dysferlin in dysferlinopathy. This hypothesis is based on the observation that utrophin, which has 80% homology with dystrophin, is overexpressing in the dystrophin deficient myofibers. To test this hypothesis, we investigated the myoferlin expression by immunoblot and immunohistochemical analysis in muscles of five patients with dysferlinopathy. For this aim, we generated a myoferlin specific antibody that does not cross react with dysferlin, and performed the immunoblot, immunohistochemical and immunoelectron microscopic studies. Immunohistochemical analysis showed that the antibodies against myoferlin and dysferlin clearly stained the normal human myofiber surface membranes. The electron microscopy of single immunogold labeled samples for myoferlin showed the presence of the molecular signal along the normal muscle cell membrane. Immunoblot analysis showed that the intensity of 230-kDa myoferlin band of dysferlinopathy muscle extracts was similar to that of normal muscle extracts. The immunostaining of dysferlinopathy muscles with anti-myoferlin antibody revealed a weak immunoreactivity along the muscle cell surface. Thus, the compensatory overexpression of myoferlin was not detected in muscles with dysferlinopathy.

It is likely that orthogonal arrays (OAs) and caveolae seen on the replicas of freeze-fractured muscle plasma membranes are involved in maintaining osmotic homeostasis. Therefore, using the freeze-fracture technique, we examined the ultrastructural changes in OAs and caveolae of the skeletal muscle plasma membrane of dehydrated and water-loaded mice. In the muscle plasma membranes of 6 dehydrated and 6 water-loaded mice, caveolar distribution was not altered, and the densities of caveolae and OAs did not show statistically significant differences when compared with those in 6 control mice, although the skeletal muscles of water-loaded mice sometimes had muscle plasma membranes with extremely numerous OAs. In contrast, the muscle plasma membranes of dehydrated mice often revealed changes in the distribution of OAs, which existed in a group at the confined area of the muscle plasma membranes and were frequently accompanied by the aggregations of intramembranous particles (IMPs) around OAs. Thus, on the basis of the present study, we suggest that OAs in skeletal muscles as well as those in brain may play an important role in maintaining osmotic homeostasis of these organs under abnormal water balance.

This study was undertaken to investigate the expression of aquaporin 4 (AQP4) in the muscle plasma membrane of children with Fukuyama-type congenital muscular dystrophy (FCMD) at protein and mRNA levels. The biopsied six muscles with FCMD, six histochemically normal muscles and eight disease control muscles were analyzed by means of immunoblots, immunohistochemistry and reverse-transcription polymerase chain reaction (RT-PCR). Immunoblots showed that the band of FCMD muscle extracts stained with anti-AQP4 antibody was faint in comparison with that of normal muscle extracts. The immunohistochemistry revealed that most of the FCMD myofibers showed negative immunostaining with anti-AQP4 antibody, although the partially positive immunostaining of sporadic FCMD myofibers was noted. The immunoreactivity was positive with anti-dystrophin and anti-beta-spectrin antibodies in almost all FCMD myofibers. The quantitative RT-PCR demonstrated that the AQP4 mRNA level of the FCMD muscles was markedly reduced. On the basis of these findings, we conclude that the expression of AQP4 in FCMD myofibers is reduced and the reduced content of AQP4 mRNA in FCMD muscles may be related to the decreased expression of AQP4 at the muscle plasma membrane of FCMD myofibers.

Our previous freeze-fracture study demonstrated the decreased density of intramembranous particles (IMPs) on the protoplasmic (P) face of muscle plasma membranes in mdx mice. However, the molecular mechanism is unknown. In the present freeze-fracture study, we examined whether the reduced P-face IMP density in mdx mice would be caused by depletion of the rosette-like IMP clusters, which are IMP aggregations differing from crystal-like orthogonal arrays (OAs). By comparison with control mice, the P-face plasma membranes of extensor digitorum longus (EDL) and soleus (SOL) muscles of mdx mice demonstrated the following findings: (1) decreased IMP density with subunit particles of OAs and IMP clusters, (2) decreased IMP density without subunit particles of OAs, (3) normal IMP density without subunit particles of OAs and IMP clusters, (4) decreased OA density in EDL muscles and normal OA density in SOL muscles, and (5) decreased IMP cluster densities in both muscles. Thus, the reduced IMP density of P-face muscle plasma membranes in mdx mice may result from the decreased IMP clusters, suggesting the relationship between IMP clusters and the integral membrane proteins is influenced by dystrophin deficiency such as that of dystrophin-associated glycoproteins or other membrane proteins.

To analyze the molecular mechanism of the increased caveolin 3 activities in dystrophin-deficient muscles, we investigated three-dimensionally the changes in caveolin 3 molecular distribution and density at the sarcolemma of mdx mice by the fracture-label electron microscopic technique. At the sarcolemma of skeletal muscles from mdx mice, the densities of gold particles associated with caveolae, non-associated with caveolae and arranged circularly without caveolae were higher than those in control mice (P

The question whether aquaporin 3 (AQP3) is expressed in normal human skeletal muscle at mRNA and protein levels has been examined, since AQP3 has been reported to be coexpressed with AQP4 in various kinds of tissues other than skeletal muscle. The gel electrophoresis of the reverse transcription polymerase chain reaction (RT-PCR) product of total RNA samples extracted from normal human muscle specimens by using the oligonucleotide primers for AQP3 contained a band of 629 base pairs which corresponded to the base pair length between two primers of AQP3. The nucleotide sequence of this RT-PCR product coincided with that of AQP3. At the protein level, immunoblot, immunohistochemical and immunoelectron microscopical studies were done by using rabbit antibody against the synthetic peptide of the cytoplasmic domain of the human AQP3 molecule. Immunoblot analysis showed that rabbit antibody against the human AQP3 reacted with a protein of approximately 30 kDa molecular weight in extracts of normal human skeletal muscles. The immunoreaction for the anti-AQP3 antibody with normal human muscle was noted at the myofibre surface. Immunogold labelling electron microscopy revealed that the gold particles indicating the presence of AQP3 molecules were located mainly at the inside surface of muscle plasma membrane.

Aciculin is a novel adherens junction antigen extracted from human uterine smooth muscle that is reported to associate biochemically with dystrophin. We attempted to determine (i) the immunostainability of anti-aciculin antibody for the 6 histochemically normal human muscles and seven muscles from boys with Duchenne muscular dystrophy (DMD) and 11 disease control muscles, (ii) the ultrastructural localization of aciculin in normal skeletal myofibers, (iii) aciculin's spacial relationship with dystrophin and beta-spectrin, and (iv) if the aciculin is ultrastructurally colocalized with dystrophin, the distance from the aciculin epitope to the epitope of the dystrophin N- or C-terminal domain. For this, rabbit anti-aciculin antibody was generated against the synthetic peptide of aciculin fragment D [4]. Immunohistochemical staining showed that the immunostainability of DMD muscles for anti-aciculin antibody was markedly decreased as compared with normal and disease control muscles. Single and double immunogold labeling electron microscopy of 6 histochemically normal human quadriceps femoris muscles revealed that aciculin was present along the inner surface of muscle plasma membrane and that aciculin formed doublets more frequently with dystrophin (23.5 +/- 1.8%; group mean +/- SE) than with beta-spectrin (12.8 +/- 1.1%; P0.01 two tailed t test). Rabbit anti-aciculin antibody frequently formed doublets with monoclonal antibodies against the N- or C-terminal domain of dystrophin at the muscle cell surface. These results suggest that aciculin is associated with dystrophin and may interact with both the N- and C-terminal domains of dystrophin.

α1-Syntrophin, a member of dystrophin-associated proteins, is expressed at the sarcolemma and at perivascular astrocytes, and participates in protein-protein interactions through its PDZ domain. Aquaporin-4 (AQP4) is the predominant water channel protein in the brain, and also expressed at the sarcolemma of fast-twitch muscle fibers. AQP4 is concentrated in orthogonal array particles (OAPs), and its expression has been reported to be decreased at the sarcolemma of dystrophin-deficient mdx mice. We examined whether α1-syntrophin targets AQP4 at the sarcolemma. Immunohistochemistry showed that AQP4 is absent at the sarcolemma in α1-syntrophin knockout mice and that its expression is also lost from the perivascular astrocyte endfeet. On the other hand, expression of AQP4 is not decreased at the sarcolemma of the knockout mice in the neonatal stage. Moreover, AQP4 is expressed in lung, stomach, and kidney of wild-type and α1-syntrophin null mice. Our results show that α1-syntrophin is a key molecule to localize AQP4 to the sarcolemma of mature fast myofibers and astrocyte endfeet, but AQP4 is targeted to the plasma membrane by different molecules in lung, stomach, and kidney.

To examine the influence of thyroid hormone on the skeletal muscle plasma membrane, we analyzed the changes in ultrastructural architecture and membrane area complexed with digitonin of muscle plasma membrane in myopathic patients with hypo-and hyperthyroidism by the conventional freeze-fracture (F-F) technique and F-F cytochemistry using the sterol-specific ligand digitonin. The densities of flask-shaped invaginations, which are mainly thought to correspond to caveolae, intramembranous particles, and orthogonal arrays, and the changes of digitoninreacted membrane areas in the muscle plasma membranes in three patients with hypothyroid myopathy and one patient with both myasthenia gravis and hyperthyroidism were compared with those in age-matched controls. In the conventional F-F study, the muscle plasma membrane of hypothyroid patients showed increased invagination density, whereas that of the hyperthyroid patient was normal ultrastructurally. In the F-F cytochemistry study, however, the ratio of digitonin-reacted membrane areas versus fractured membrane areas was not different between hypothyroid patients and controls, whereas that of the hyperthyroid patient was lowered in comparison with that of control. These results suggest that thyroid hormone may alter the biochemical properties and ultrastructural architecture of muscle plasma membrane.

Dystrophin is the Duchenne muscular dystrophy gene product and is a membrane cytoskeletal protein present in the network of the plasma membrane undercoat. Adhalin (50 kDa dystrophin-associated glycoprotein) and beta-dystroglycan (43 kDa dystrophin-associated glycoprotein) are the transmembrane components of the normal muscle plasma membrane, and beta-dystroglycan has been demonstrated to bind dystrophin at the inside surface of normal muscle plasma membrane. This investigation was undertaken to test whether the epitopes of dystrophin, beta-dystroglycan and adhalin are closely associated with each other by using triple immunogold labelling electron microscopy on normal human skeletal myofibers. Although closely associated signals of triplet immunogold particles were observed, there were less numerous than expected. However, closely associated signals of two epitopes of dystrophin and beta-dystroglycan, dystrophin and adhalin, or adhalin and beta-dystroglycan were frequently observed. These ultrastructural findings are consistent with biochemical evidence implying that dystrophin, beta-dystroglycan and adhalin are closely associated with each other at the normal muscle plasma membrane.

The ultrastructural localization of adhalin and its relations to dystrophin, beta-dystroglycan, and beta-spectrin were studied in normal murine skeletal myofibers. The C-terminal peptides of adhalin and beta-dystroglycan were synthesized based on their cDNAs, and the affinity-purified antibodies against these peptides were produced. Single-immunolabeling electron microscopy showed that the adhalin was located just inside the muscle plasma membrane or inside the myofiber a short distance from the plasma membrane. The adhalin signal was also noted at the sarcoplasmic side of plasmalemmal invaginations or at vesicular structures in subsarcolemmal areas. Double-immunogold-labeling electron microscopy disclosed a similar localization of dystrophin, beta-dystroglycan, and beta-spectrin. The close association of adhalin with dystrophin or beta-dystroglycan was demonstrated by formation of doublets by signals of antibodies of adhalin with those of dystrophin or beta-dystroglycan and was confirmed by statistical analyses. This study demonstrated that the location of adhalin is close to that of dystrophin and beta-dystroglycan at the muscle plasma membrane.

Muscle plasma membrane undercoats were investigated by conventional electron microscopy in both Duchenne muscular dystrophy (DMD) and Fukuyama congenital muscular dystrophy (FCMD). The densities of the plasma membrane undercoats were rarefied in the parts of the plasma membranes overlying the degenerating focus in both DMD and FCMD myofibers. The degree of rarefaction tended to be parallel to the degree of degeneration in the myofibers. It was hard to distinguish the undercoat densities of normal-looking myofibers of DMD and FCMD muscles from those of control myofibers from histochemically-normal muscles. On the other hand, the undercoats of regenerating myofibers in DMD and FCMD muscles were denser than normal.

The intramembranous particle (IMP), orthogonal array (OA) and orthogonal array subunit particle (OASP) densities in skeletal muscle plasma membranes from eight patients with Becker's muscular dystrophy (BMD) were analysed by the freeze-fracture technique. The results showed almost normal IMP density with the significant decrease of OA and OASP densities in BMD. The group mean densities +/- SE of IMPs on the protoplasmic faces with and without OASPs, and on extracellular faces/microns 2 were 2137 +/- 207, 1839 +/- 68 and 895 +/- 108, respectively in controls; whereas those of BMD were 1989 +/- 259, 1837 +/- 203 and 900 +/- 239, respectively (P > 0.1 by two-tailed t-test). The group median density of OAs and their pits/microns 2 was 4.89 with mid-ranges (25-75% values of the counts) of 2.66-10.18 in controls; whereas that in BMD was 2.15 with mid-ranges of 1.14-4.31 (P < 0.01 by Wilcoxon rank-sum test). The group mean density +/- SE of OASPs in controls was 15.99 +/- 1.83; whereas that in BMD was 13.47 +/- 1.07 (P < 0.01 by two-tailed t-test). However, the diminution of OA and OASP densities in BMD muscle plasma membranes was not as severe as in Duchenne's muscular dystrophy. There was a relationship between OA density and clinical severity in BMD patients; the decrease of OA density in a severe BMD patient was more marked than that in mildly affected BMD patients. Therefore, it seems that marked depletion of OA density may lead to the severe disability in muscular dystrophies.

Skeletal muscle samples from four patients with myopathies showing autophagic vacuole formation were examined by immunohistochemical, biochemical and immunoblot analyses. Immunochemical studies demonstrated strongly positive reactions of cathepsins B and L in and around the rimmed vacuoles and weak reactions in intramyofibral portions of degenerating muscle fibres. Faint staining of cathepsin H was also seen at the sites of rimmed vacuoles. Biochemical analyses showed increased activity of cathepsin B and L in muscle specimens from the patients with rimmed vacuole formation. However, there was no remarkable cathepsin H activity in the muscle specimens from these patients. Cathepsin L immunoblot analysis of muscle extracts showed three bands with molecular masses of 30 kDa, 36 kDa and 39 kDa. The immunostaining of cathepsin L from patients with rimmed vacuole formation was stronger than that of normal controls. Cathepsin B immunoblot analysis showed only faint bands in samples from patients with rimmed vacuole formation and normal controls. This study demonstrated, for the first time, strong immunoreactivity of cathepsin L at the sites of rimmed vacuoles. Possible mechanisms of rimmed vacuole formation are discussed.

Syntrophins are the cytoplasmic peripheral proteins of dystrophin glycoprotein complex, of which five (alpha l, beta 1, beta 2, gamma 1 and gamma 2) isoforms have been identified so far. Respective syntrophin isoforms are encoded by different genes but have similar domain structures. At the sarcolemma of skeletal muscle, the most abundant alpha l-syntrophin was shown to interact at its PDZ domain with many membrane proteins. Among them, the AQP4 interaction with alpha 1-syntrophin PDZ domain was demonstrated by a Tg mouse study, prompting us to investigate the interaction between mouse alpha l-syntrophin (BC018546: nt.267-492, PDZ domain) pEXP-AD502 as prey vector and mouse AQP4 (NM009700: nt.805-969) pDBLeu as bait vector by the yeast two-hybrid assay, resulting in a negative study. We further studied the binding partner of another sarcolemma located beta 1-syntrophin, and performed a yeast two-hybrid experiment. With human beta 1-syntrophin as bait and human skeletal muscle cDNA library as prey, we obtained one positive clone which turned out to be alpha-dystrobrevin. Although the interaction of human beta 1-syntrophin with alpha-dystrobrevin has already been shown by immunoprecipitation assay, we have here confirmed this interaction by a yeast two-hybrid experiment.

Aquaporin (AQP) 4 is a water-specific channel protein and is abundant in central nervous tissues and skeletal muscles. Recently, the AQP4 molecule has been increasingly highlighted in its pathophysiological role of several neurological diseases, such as stroke, muscular dystrophy and neuromyelitis optica. We therefore measured the levels of AQP4 mRNA and glyceraldehyde-3 phosphate dehydrogenase mRNA (an internal control) in muscle and brain tissues of wild-type mice (C57BL10/ScSn) and age-matched dystrophin-deficient mdx mice (C57BL10/ScSn mdx) by real-time quantitative RT-PCR. The relative AQP4 mRNA level was highest in the spinal cord among the neuromuscular tissues examined in wild-type mice. Among the muscle tissues of wild-type mice, the relative AQP4 mRNA level was higher in extensor digitorum longus (EDL) muscles, and its descending order was EDL, quadriceps femoris, soleus and heart muscles. It is noteworthy that there was no difference in the relative AQP4 mRNA levels in the brain tissues between wild-type mice and age-matched mdx mice. In contrast, the AQP4 mRNA level in the quadriceps femoris muscle was significantly lower in mdx mice than in wild-type mice. The fact that the spinal cord contains the highest AQP4 mRNA may be related to the pathogenesis of neuromyelitis optica, in which AQP4 protein is the target antigen. In addition, the low expression level of AQP4 mRNA in the mdx mouse muscle suggests a functional link between AQP4 and dystrophin in the muscle tissue. We suggest that a similar pathomechanism may underlie the phenotypic consequences of the mdx mouse and Duchenne muscular dystrophy.

The plasma membrane of skeletal muscles contains water channels such as aquaporin 4 (AQP4), aquaporin 3 (AQP3) and aquaporin 7 (AQP7). In dehydrated mice, we have recently reported the altered distribution of the aggregations of intramembranous particles (IMPs), such as orthogonal array (a crystal-like structure) and IMP cluster (a rosette-like structure) on the freeze-fractured skeletal muscle plasma membranes. In this fracture-label study, we first tested whether the orthogonal arrays (OAs) were composed of AQP4 in skeletal muscles and further analyzed the relationship between IMPs including IMP clusters and AQP3 molecules. As a result, many of the gold particles indicating AQP4 was associated with OAs (79%) by our fracture-label technique. On the other hand, approximately 50% of gold particles indicating AQP3 were associated with IMP clusters. Thus we confirmed that the OAs are composed of AQP4 in skeletal muscles, and further demonstrated that some of the IMP clusters are composed of AQP3 and may participate in maintaining osmotic homeostasis in skeletal myofibers. The fracture-label method is useful in investigating the molecular identification of membrane proteins such as AQP3 and AQP4.

Sarcospan is a 25 kDa transmembrane component of dystrophin-associated glycoprotein. We generated a rabbit polyclonal antibody against synthetic peptide of the N-terminal domain of human sarcospan. Using this antibody we investigated the localization of sarcospan and its spacial relation to the components of sarcoglycan subcomplex in normal human skeletal myofibers by immunofluorescent microscopy and immunogold electron microscopy. In immunofluorescence the reaction was observed continuously at the myofiber surface. Ultrastructurally the gold signals of rabbit anti sarcospan antibody were present along the muscle plasma membrane, mainly at its inside surface. The triple immunogold labeled muscle samples showed that the signals of rabbit or sheep polyclonal anti alpha-, beta-, gamma- and delta-sarcoglycan antibodies and/or mouse monoclonal anti beta-, gamma- and delta-sarcoglycan antibodies were located along the muscle plasma membrane, and the cluster formation of different two or three sarcoglycan molecules was observed. The triple immunogold labeling also revealed that the signal of sarcospan molecules are present frequently in doublets and/or triplets with the components of sarcoglycan subcomplex, resulting in the cluster formation of signals of sarcoglycan and sarcospan molecules. The result of this study showed that sarcospan was expressed at the myofiber surface and that sarcospan was present in close association with alpha-, beta-, gamma- and delta-sarcoglycans and formed a functional unit with sarcoglycan subcomplex.

Aquaporin 4 (AQP4) is a water channel protein that is widely distributed in human tissues. However, the precise functional role of AQP4 in skeletal muscle tissue has not yet been determined. Expression of AQP4 was reported to be reduced in muscle tissue from Duchenne muscular dystrophy patients. In the regenerating phase of skeletal muscle, AQP4 expression was reduced when nerve supply was not present. However, in diseased human muscles with neurogenic atrophy including amyotrophic lateral sclerosis, there has been no data on the changes in AQP4 expression. In the present study, we investigated the expression of AQP4 at mRNA and protein levels in human muscles with neurogenic atrophy. The mean level of AQP4 mRNA was significantly lower in muscles with neurogenic atrophy than that in muscles from normal controls. The myofiber surface immunostaining with anti-AQP4 antibody in muscles with neurogenic atrophy was reduced on the surface of scattered myofibers, small angulated myofibers, and myofibers in small- and large-group atrophy despite the presence of dystrophin. Based on the present findings, we conclude that the expression of AQP4 is affected by nerve supply and is down-regulated in human muscles with neurogenic atrophy.

Primary deficiency of merosin causes a severe congenital muscular dystrophy (CMD) and a mouse dystrophy (dy/dy mouse). Also, its secondary deficiency is seen in some CMD with abnormal glycosylation of Alpha-dystroglycan, an extracellular membrane protein, which is the receptor of merosin and binds to dystrophin underlying the sarcolenma via Beta-dystroglycan, a transmembrane protein. In immunogold and freeze-etch electron microscopic studies, merosin in basal lamina of normal skeletal muscles has a zonation in the distribution and is localized at the lamina lucida of muscle basal lamina, and dystrophin molecules are often closed to merosin molecules at the inside and outside surface of muscle plasma membrane. Moreover, merosin molecules exist as the short fine cross-bridge fibrils connecting the basal lamina to the neighboring outer leaflet of the muscle plasma membrane. In freeze-fracture studies, the changes in muscle plasma membranes of dy/dy mice reveal a markedly decreased density of orthogonal arrays (OAs) but normal density of intramembranous particles (IMPs), whereas depletions of IMPs with decreased OAs have been found in Fukyama-type congenital muscular dystrophy, Duchenne muscular dystrophy, and mdx mice. Thus, further studies including the functional role of OAs would be required to understand the pathomechanism of merosin-deficient CMD.

The question whether aquaporin 3 (AQP3) is expressed in normal human skeletal muscle at mRNA and protein levels has been examined, since AQP3 has been reported to be coexpressed with AQP4 in various kinds of tissues other than skeletal muscle. The gel electrophoresis of the reverse transcription polymerase chain reaction (RT-PCR) product of total RNA samples extracted from normal human muscle specimens by using the oligonucleotide primers for AQP3 contained a band of 629 base pairs which corresponded to the base pair length between two primers of AQP3. The nucleotide sequence of this RT-PCR product coincided with that of AQP3. At the protein level, immunoblot, immunohistochemical and immunoelectron microscopical studies were done by using rabbit antibody against the synthetic peptide of the cytoplasmic domain of the human AQP3 molecule. Immunoblot analysis showed that rabbit antibody against the human AQP3 reacted with a protein of approximately 30 kDa molecular weight in extracts of normal human skeletal muscles. The immunoreaction for the anti-AQP3 antibody with normal human muscle was noted at the myofibre surface. Immunogold labelling electron microscopy revealed that the gold particles indicating the presence of AQP3 molecules were located mainly at the inside surface of muscle plasma membrane.

We examined whether AQP7 molecules are expressed in the normal skeletal muscle at mRNA and protein levels. Gel electrophoresis of the reverse transcription-polymerase chain reaction (RT-PCR) product of total RNA samples of normal human or mouse muscles by using oligonucleotide primers for human or mouse AQP7 showed a band of 328 or 369 basepairs, which corresponded to the basepair length between two primers of AQP7. The nucleotide sequence of these RT-PCR products coincided with those of human and mouse AQP7. Immunoblot, immunohistochemical and immunoelectron-microscopic studies of the protein were done by using the rabbit antibody against the synthetic peptide of the N-terminal cytoplasmic domain of the human AQP7 molecule. Immunoblot analysis showed that the rabbit antibody against human AQP7 reacted with a protein of approximately 30 kDa molecular weight in extracts of normal human and mouse skeletal muscles, and normal mouse liver. Immunohistochemistry with our anti-AQP7 antibody showed an immunoreaction at the myofiber surface of type 1 and type 2 fibers in human muscles and of type 2 fibers in mouse muscles.

The double contrast pharyngogram by use of computed radiography (DCP-CR) has been found to be useful in detection of asymptomatic swallowing dysfunction. Following the DCP-CR examination, we investigated the incidence of aspiration pneumonia in 143 patients with chronic cerebrovascular disease (CVD) for 3 years and the effects of ACE inhibitors on the prevention of pneumonia. Aspiration pneumonia occurred in 29 out of 143 patients, and more frequently in the elderly chronic CVD patients with multiple brain lesions. Aspiration pneumonia was confirmed in 26 out of 85 patients (30.6%) with abnormal barium adhesion to the pharyngeal wall on the double contrast pharyngogram image by DCP-CR; whereas pneumonia occurred in 3 out of 58 patients (5.2%) with normal findings of DCP-CR pharyngogram. Among chronic CVD patients with abnormal findings of DCP-CR pharyngogram, the incidence of aspiration pneumonia was significantly lower in the patients treated with ACE inhibitors than in those treated with other antihypertensive agents or without antihypertensive agents (chi 2 value = 7.163, p0.05). Accordingly, ACE inhibitors may prevent the aspiration pneumonia and reduce the incidence of aspiration pneumonia in the chronic CVD patients with abnormal DCP-CR pharyngogram images.

Aquaporin 4 (AQP4) is a recently discovered membrane bound water-selective channel and has been described at the light microscopic level to be predominantly expressed in the astrocytes of the brain, especially at the perivascular astrocyte endfoot processes. Alpha1-syntrophin, a member of dystrophin-associated protein, has also been reported at the light microscopic to be expressed level in the same site of astrocytes as AQP4 and interacts with other molecules through its PDZ domain. AQP4 expression has been reported to be absent at the sarcolemma and the perivascular astrocyte endfoot processes of alpha1-syntrophin knockout mice. Based on these observations, the molecular association between AQP4 and alpha1-syntrophin could be speculated. To test this hypothesis, we investigated the ultrasturctural localization of AQP4 and alpha1-syntrophin in the brain astrocytes by using double immunogold labeled electron microscopy. The results showed that AQP4 and alpha1-syntrophin colocalized frequently at the astrocyte membrane, especially at the perivascular astrocyte endfoot processes and suggested the presence of linkage between AQP4 and alpha1-syntrophin at the astrocyte plasma membrane.

In order to investigate the mode of existence of the sarcoglycan complex, neuronal nitric oxide synthase (nNOS), β-dystroglycan, and dystrophin in the normal skeletal myofiber, we examined the ultrastructural localization and mutual spatial relationship of nNOS, β-dystroglycan, dystrophin, and the individual components of the sarcoglycan complex by using triple immunogold labeling electron microscopy. Each molecule of α-, β-, γ- and δ-sarcoglycans is located intracellularly or extracellularly near the muscle plasma membrane mostly in accordance with the sarcoglycan antigenic sites against which the antibodies were generated. The association of different two and/or three sarcoglycan molecules out of α-, β-, γ- and δ-sarcoglycan molecules was frequently observed. Each molecule of nNOS, β-dystroglycan, and dystrophin was ultrastructurally noted along the cell surface of normal skeletal myofibers. Moreover, the close relation of a sarcoglycan molecule with β-dystroglycan and dystrophin, and the association of nNOS with dystrophin were also confirmed ultrastructurally. Thus, this study demonstrated that the constituting molecules of the sarcoglycan complex, nNOS, β-dystroglycan, and dystrophin existed in the form of a cluster at the normal muscle plasma membrane. The association of nNOS with dystrophin and its associated glycoproteins may form a macromolecular signaling complex at the muscle plasma membrane. Microsc. Res. Tech. 55:154–163, 2001. © 2001 Wiley-Liss, Inc.

Changes in muscle plasma membranes in mice lacking exon 52 of the dystrophin gene (mdx52 mouse) were studied using the freeze-fracture technique. The extensor digitorum longus (EDL) muscle plasma membrane of the mdx52 mouse at 8 weeks of age showed significantly increased caveola density (p < 0.05 by two-tailed t-test) and significantly decreased densities of intramembranous particles (IMPs), orthogonal arrays (OAs) and orthogonal array subunit particles (OASPs) (p < 0.05 by two-tailed t-test, p < 0.01 by Wilcoxon rank-sum test, p < 0.05 by two-tailed t-test, respectively) on the protoplasmic face when compared with those of control EDL muscles. These changes are more similar to those seen in DMD than those in the mdx mouse at the same age as reported previously. Thus, the gene abnormality in the different exon of the mouse dystrophin gene seems to induce somewhat different changes in the muscle plasma membrane.

The recently identified water channel aquaporin 4 is a major component of the orthogonal arrays observed with freeze-fracture electron microscopy. We examined the expression of aquaporin 4 mRNA and protein in rat regenerating muscle under innervated and denervated conditions. We found decreased sarcolemmal immunostaining of aquaporin 4 in denervated regenerating muscle as opposed to innervated muscle. Quantitative reverse transcription-polymerase chain reaction revealed that aquaporin 4 mRNA was expressed in the innervated regenerating muscle; whereas it was not expressed in denervated muscle. Thus, lack of aquaporin 4 protein may be due to lack of aquaporin 4 mRNA in the denervated regenerating muscle. We conclude that the nerve supply influences expression of aquaporin 4 at the mRNA level in regenerating muscle.

Object Adhesion molecules are suggested to play important roles in the pathogenesis of inflammatory diseases. We examined the expression of adhesion molecules in the muscles of human inflammatory myopathies. Methods We immunohistochemically studied the expression and distribution of two molecules in the selectin family (E- and P-selectin) and their common ligand sialyl Lewis X in 18 inflammatory myopathies, 13 disease controls, and 16 normal controls. Results In inflammatory myopathies, E- and P-selectin were upregulated on the surface of blood vessels, especially on the endothelial cells of the venules. Sialyl Lewis X was upregulated in the blood vessels, infiltrating leukocytes, and the surface of some atrophic myofibers. Some control muscles also showed weakly positive staining with these molecules, however, expression of these molecules was most striking in the muscles of inflammatory myopathies. Conclusion The results suggested that these molecules are upregulated in inflammatory myopathies and might play a role in the pathogenesis of inflammatory myopathies.(Internal Medicine 38: 632-635, 1999)

Immunostainability of anti aquaporin 4 antiserum was investigated in the muscles of dystrophin deficient mdx mice. Western blot analysis showed that the rabbit antiserum against aquaporin 4 reacted with a 28 kDa protein in extracts of normal mouse quadriceps femoris muscles but did not react with the protein in extracts of quadriceps femoris muscles of mdx mice. Immunoperoxidase staining of the muscles from normal and mdx mice revealed the positive immunoreaction at the myofiber surface of normal mice and the negative, or the faint and discontinuous immunostaining at the surface of mdx myofibers. Immunogold electron microscopy disclosed the localization of aquaporin 4 molecules at the myofiber plasma membranes of normal mice and the localization was consistent with that of orthogonal array particles in the protoplasmic face of normal muscle plasma membrane seen in freeze fracture replicas. This study demonstrated that the density of aquaporin 4 molecules was decreased in the muscle plasma membranes of mdx mice, resulting in the faulty function of mdx myofibers.

Ultrastructural localization of alpha-, beta- and gamma-sarcoglycan and their mutual relation, and their relation to dystrophin, beta-dystroglycan and beta-spectrin were investigated in normal skeletal myofibers. Single-immunogold labeling electron microscopy showed that the signals of rabbit and sheep polyclonal antibodies against the synthetic peptide of the cytoplasmic domain of alpha-, beta or gamma-sarcoglycan were present along the inside surface of muscle plasma membrane and at the sarcoplasmic side of plasma membrane invaginations and vesicular structures in subsarcolemmal areas. These localizations were similar to that of dystrophin, beta-dystroglycan and beta-spectrin. Double-immunogold labeling disclosed the close association of alpha-, beta- and gamma-sarcoglycan each other and alpha-, beta-, gamma-sarcoglycan with dystrophin or beta-dystroglycan, and this was confirmed by statistical analysis. Monoclonal antibody against the extracellular domain of alpha-sarcoglycan was used with above-mentioned polyclonal anti-beta- and -gamma-sarcoglycan antibodies for triple-immunogold labeling, in which signals of alpha-sarcoglycan localized at the outer surface of muscle plasmalemma and those of beta- and gamma-sarcoglycans were present at the inside surface of plasma membrane. The triple immunolabeling showed an occasional closely associated presence of the three signals for alpha-, beta-and gamma-sarcoglycans, and a more frequent association for two signals out of alpha-, beta- and gamma-sarcoglycans. This study demonstrated that alpha-, beta- and gamma-sarcoglycan are closely located to one another and to dystrophin and beta-dystroglycan at the muscle plasma membrane.

We investigated the ultrastructural localization of alpha 1-syntrophin and neuronal nitric oxide synthase (nNOS) in normal human skeletal myofibers and analyzed their relation to each other and to dystrophin using single and double immunogold-labeling electron microscopy. Single immunolabeling showed antibodies to alpha 1-syntrophin and nNOS on the inner surface of the muscle plasma membrane, the sarcoplasmic side of plasma membrane invaginations, and the sarcoplasm near mitochondria of subsarcolemmal areas. The epitopes of alpha 1-syntrophin and nNOS tended to be present in clusters. Double immunolabeling revealed that epitope combinations of alpha 1-syntrophin-dystrophin, alpha 1-syntrophin-nNOS, and nNOS-dystrophin occurred more frequently in doublet form than did other epitope combinations, such as alpha 1-syntrophin-beta- spectrin and nNOS-beta-spectrin. These increased frequencies were noted both at the muscle plasma membrane undercoat and near mitochondria of subsarcolemmal areas. A significantly higher percentage of doublets comprised antibodies against alpha 1-syntrophin and dystrophin (28.5 +/- 1.5%, group mean +/- SE) than those against alpha 1-syntrophin and beta-spectrin (9.2 +/- 0.8%, P0.01). Furthermore, nNOS formed doublets significantly more frequently with dystrophin (25.2 +/- 3.3%) and alpha 1-syntrophin (26.0 +/- 4.1%) than with beta-spectrin (13.9 +/- 2.3%; P0.05). These data support the association of dystrophin, alpha 1-syntrophin, and nNOS at the inner surface of the muscle plasma membrane and near mitochondria of subsarcolemmal areas of normal human skeletal myofibers.

Merosin is a basement-membrane-associated protein found in striated muscle, peripheral nerve and placenta, the deficiency of which causes the muscle wasting condition in C57BL/6J-dy/dy, so-called dy/dy mouse. Moreover, merosin is the binding protein of 156 kDa alpha-dystroglycan which binds dystrophin by way of 43 kDa beta-dystroglycan. Therefore, merosin is an important component of the basal lamina of normal skeletal myofibers. We investigated the ultrastructural localization of merosin antibody in normal human skeletal myofibers by using immunogold electron microscopy and freeze etch electron microscopy. The ultrastructure of the basal lamina showed the presence of the lamina lucida, lamina densa and lamina reticularis. The lamina lucida appeared electron translucent with the exception of fuzzy fibrils. The immunogold electron microscopy disclosed that the merosin was present at the innermost layer (lamina lucida) of the basal lamina of normal human skeletal myofibers. With freeze etch replica electron microscopy, short cross-bridge fine fibrils were noted in the lamina lucida, connecting the basal lamina to the outer leaflet of the muscle plasma membrane. They measured 3-13 nm in diameter, 20-90 nm in length and were distributed with a spacing of 30-40 nm. The immunogold particles showing the presence of the merosin epitope were associated with these connecting structures.

Ultrastructural localization of the C-terminus of a 43-kDa dystrophin-associated glycoprotein (43-DAG) and its relation to dystrophin were investigated in normal human skeletal myofibers by using single and double immunogold labelling methods. The C-terminal end of 43-DAG antibody was localized ultrastructurally at the inside surface of muscle plasma membrane and the sarcoplasmic side of the plasma membrane invaginations. Double immunolabelling electron microscopy disclosed that signals of the C-terminal end of 43-DAG and dystrophin were often associated with each other and formed a doublet at the inside surface of plasma membranes in normal human skeletal myofibers.

The ultrastructure and mode of existence of the dystrophin molecule and its relations to actin filaments were examined in murine skeletal myofibers. Electron microscopy of freeze-etched replicas of gold-labelled dystrophin molecules in quick-freeze, deep-etch, rotary-shadow preparations revealed rod-like structures 108.2 +/- 16.3 nm long and 3.1 +/- 1.5 nm thick. Some dystrophin molecules appeared to link their ends to form anastomosing networks; others were separate from each other. The dystrophin molecules were parallel or nearly parallel to the inner surface of the muscle plasma membrane. Double immuno-labelling transmission electron microscopy using N- and C-terminal dystrophin antibodies showed that the group mean distances of the N- and C-terminal signals from the muscle plasma membrane were 52.7 +/- 8.1 nm and 45.9 +/- 11.3 nm, respectively, which were not significantly different. Histograms of the distribution of the N- and C-terminal distances from the muscle plasma membrane had similar patterns with peaks 10-20 nm from the membrane. This was consistent with the findings of the mode of existence of dystrophin molecules seen in freeze-etched replicas. Finally, the dystrophin molecules were linked with the most peripheral sarcoplasmic actin like filaments, end to side as well as end to end.

Concentrations of several proteins that are characteristic of the nervous system were time-sequentially analyzed by radio- and enzyme-immunoassay in the cerebrospinal fluid (CSF) of patients with Creutzfeldt-Jakob disease (CJD). We found abnormally high levels of several proteins, such as neuron-specific enolase (NSE), S-100b protein, brain-type isozyme of creatine kinase (CK-BB) and alpha subunit of GTP binding protein G0 (G0 alpha) in the early stage of the disease. Generally, these protein levels were far higher in CJD patients than in normal controls and other neurological patients in the early stage before the typical clinical manifestations were evident. These levels increased to maxima when the disease activity was most prominent and returned to normal or mildly elevated levels in the terminal stage. The results imply that these protein levels can serve as biochemical markers for the presence of an active destructive process in CJD brain and provide us with a useful indicator for early diagnosis of CJD.

Dystrophin is the protein product of Duchenne muscular dystrophy gene which is defective in this genetic disorder. Here we identified ultrastructurally the dystrophin molecule from the various cytoskeletons at the cytoplasmic surface of murine myofiber plasma membrane by using quick-freeze, deep-etch, rotary-shadow replica of anti-dystrophin antibody-decorated muscle samples. The molecule was really cytoskeleton and incorporated in the meshwork of the plasma membrane-associated cytoskeletons. The molecule appeared to connect directly and/or through another cytoskeletal molecule with actin filament.

The Duchenne muscular dystrophy product 'dystrophin' has been shown to be located at the inner surface of normal muscle plasma membrane. This study was undertaken to visualize the shape of dystrophin molecules and their topographical distribution at the inner surface of murine skeletal muscle plasma membrane. The immunogold electron microscopy of plastic-embedded quadriceps femoris muscles of six mdx mice and six control mice showed the presence of gold particles along the muscle plasma membrane undercoat of all muscle samples from the control mice without any antibody reaction in the mdx mice muscles. The gold-labelled muscles of six mdx and six control mice were quickly frozen by liquid helium in a rapid-freeze apparatus. High magnification electron microscopy of the quick-freeze, deep-etch, rotary-shadow replicas of the gold-labelled muscles demonstrated the presence of dystrophin molecules associated with gold particles at the cytoplasmic surface of mdx control mice. The dystrophin molecules displayed a variety of shapes, such as rods with a reduction in diameter from one end to the other end and/or with the enlargement of their end(s). These dystrophin molecules were incorporated in the meshwork of muscle plasma membrane-associated cytoskeletons.

We synthesized 3 peptide fragments predicted by residues 2354–2368 (peptide I), 2310–2324 (peptide II) and 2255–2269 (peptide III) on the mid-portion of the human dystrophin cDNA map where the most frequent intragenic deletions occurred in Duchenne muscular dystrophy. Rabbit antibodies against these peptides were raised and cryosections of 47 biopsied muscles were studied immunohistochemically. The 47 biopsied muscles included the quadriceps femoris muscles of 8 Duchenne muscular dystrophy patients, 8 child and 5 adult normal controls, 1 facioscapulohumeral dystrophy, 2 limb girdle dystrophy, 3 myotonic dystrophy, 3 polymyositis, 1 mitochondrial myopathy, 1 nemaline myopathy, 3 amyotrophic lateral sclerosis and the extensor digitorum longus muscles of 6 mdx mice (C57BL/10ScSn-mdx) and 6 normal control mice (C57BL/10ScSn). The peptide I antiserum continuously stained the myofiber surface membranes in 8 child and 5 adult normal control muscles, and in 14 other muscles from various neuromuscular diseases, but failed to stain the surface membranes in normal control mice. The surface membranes of 8 Duchenne muscles were not stained by the peptide I antiserum except for a few myofibers. Although the ELISA titers of peptide I, II and III antibodies were high, immunostaining by peptide II antiserum showed no reaction in the myofibers of any of the biopsied muscles, and immunostaining by peptide III antiserum revealed faint reactions on the myofiber surface membranes of all biopsied muscles, including the mdx control mouse muscles except for the Duchenne and mdx myofibers.

The Duchenne muscular dystrophy gene product "dystrophin" is a sarcolemma-associated cytoskeleton which is present just inside the sarcolemma. We investigated the ultrastructure and the relationship of this cytoskeleton to the muscle plasma membrane by immunoelectron microscopy and freeze-etch electron microscopy of liquid helium-frozen fresh muscles. The immunoelectron microscopy of the extensor digitorum longus muscles of six mdx mice and six control mice showed the location of anti-dystrophin antibody along the muscle plasma membrane undercoat of all the muscle samples from the control mice without any antibody reaction in the mdx mice muscles. Fresh extensor digitorum longus muscles of seven mdx and six control mice were quick-frozen in liquid helium in the rapid-freeze device. High-magnification electron microscopy of the deep-etch, rotary-shadow replicas of the frozen muscles showed the network formation and attachment of individual rod-shaped cytoskeletons of variable size to the cytoplasmic surface of the muscle plasma membrane in both mdx mice and control mice. The length of cytoskeletons attached to the muscle plasma membranes was measured and mean length +/- SE in mdx mice and control mice were 98 +/- 4 nm and 101 +/- 3 nm, respectively. Although these values were not statistically different (P greater than 0.1), the distribution frequency of 130-150 nm muscle plasma membrane-associated cytoskeletons was 7.9% in mdx mice versus 14.5% in control mice. Since the predicted length of dystrophin is 125-150 nm, the 130- to 150-nm plasma membrane-associated cytoskeletons of mdx control mice may include dystrophin.

Background In Duchenne muscular dystrophy (DMD), previous freeze-fracture electron microscopic studies demonstrated that muscle plasma membrane contained markedly decreased numbers of orthogonal arrays. Recent investigations showed that orthogonal arrays were composed of aquaporin 4 (AQP4) molecules, a member of the water channel protein family. Objectives To study whether the immunostainability of anti-AQP4 antibody is reduced in muscles of patients with DMD and whether, if it is reduced, the problem is at the genomic DNA, messenger RNA (mRNA), or posttranscriptional level. Patients and methods We analyzed the muscle and blood samples from 6 boys with DMD, 6 normal control subjects, and 12 patients with neuromuscular diseases at the protein, genomic DNA, and mRNA levels. At the protein level, immunohistochemical staining and immunoblot analysis were performed. At the genomic DNA and mRNA levels, the polymerase chain reaction and reverse transcription polymerase chain reaction, respectively, were used to screen for mutations in the AQP4 gene. Results At the protein level, immunohistochemical staining of our originally generated rabbit anti-AQP4 antibody in DMD muscles was markedly reduced. Most of the DMD myofibers showed negative staining with sporadic partially positive fibers at their myofiber surface, whereas the control muscles displayed continuous myofiber surface staining. Immunoblot analysis showed that the content of AQP4 in DMD muscles was remarkably decreased. Amplification of leukocyte genomic DNA by polymerase chain reaction showed that the patients with DMD had genomic DNA of the AQP4 molecule. Quantitative reverse transcription polymerase chain reaction demonstrated that DMD skeletal muscles contained markedly decreased AQP4 mRNA compared with controls. Conclusion The reduction in AQP4 in DMD muscles results from decreased levels of AQP4 mRNA in DMD myofibers.