- Email: [email protected]
Immunocytochemical studies of aquaporin 4 in the skeletal muscle of mdx mouse
Journal of the Neurological Sciences 164 (1999) 24–28
Immunocytochemical studies of aquaporin 4 in the skeletal muscle of mdx mouse a
a,
a
a
a
Jian Wu Liu , Yoshihiro Wakayama *, Masahiko Inoue , Seiji Shibuya , Hiroko Kojima , Takahiro Jimi a , Hiroaki Oniki b a
Division of Neurology, Department of Medicine, Showa University Fujigaoka Hospital, 1 -30, Fujigaoka, Aoba-ku, Yokohama 227 -8501, Japan b Electron Microscope Laboratory, Showa University Fujigaoka Hospital, 1 -30 Fujigaoka, Aoba-ku, Yokohama, Japan Received 25 November 1998; accepted 8 February 1999
Abstract 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. 1999 Elsevier Science B.V. All rights reserved. Keywords: Anti aquaporin 4 antibody immunoreactivity; Skeletal muscle; Mdx mouse; Freeze fracture; Muscle plasma membrane; Orthogonal array
1. Introduction Duchenne muscular dystrophy (DMD) is a X-linked, dystrophin deficient disorder and the affected boys show proximal muscle weakness of 4 limbs and calf pseudohypertrophy from childhood, and inevitably die at around the age of 20 [7,10,14]. Although lack of dystrophin leads to muscle cell degeneration and necrosis in DMD, the precise mechanism of muscle cell death has not yet been fully clarified. The DMD gene product, dystrophin, is a membrane cytoskeletal protein located at the inside surface of muscle plasma membrane and does not have transmembranus domain [14,18]. Therefore the freeze fracture preparation of normal muscle does not reveal the dystrophin molecule, since the fracture plane of freeze *Corresponding author. Tel.: 181-45-971-1151; fax: 181-45-9742204.
fracture preferentially goes through the hydrophobic interior of muscle plasma membrane. Schotland et al. [21] and we [27] previously investigated the ultrastructure of muscle plasma membranes of both histochemically normal muscles and DMD muscles by using freeze fracture method, and found marked depletion of orthogonal arrays in DMD muscle plasma membranes. Thereafter Shibuya and Wakayama [23] studied by freeze fracture electron microscopy the muscle plasma membrane of dystrophin deficient mdx (C57BL / 10ScSn-mdx) mice and showed the similar depletion of orthogonal arrays but not as conspicuous as DMD muscles [23]. Orthogonal array is reported to exist more numerously in the plasma membrane of white muscle fibers in rats than in that of red muscle fibers [19]. Recent investigation demonstrated that aquaporin 4 molecule, which is a water channel protein present in a variety of cell types [5], is a 28 kDa integral membrane protein and is seen as orthogon-
0022-510X / 99 / $ – see front matter 1999 Elsevier Science B.V. All rights reserved. PII: S0022-510X( 99 )00051-9
J.W. Liu et al. / Journal of the Neurological Sciences 164 (1999) 24 – 28
al array in freeze fracture electron microscopy [9,26,29]. A cDNA for aquaporin 4 water channel protein was isolated recently from rat brain [8,13]. We generated a polyclonal antibody against the synthetic peptide of cytoplasmic domain of aquaporin 4, and investigated immunohistochemically the muscle plasma membranes of normal (C57BL / 10ScSn) and mdx mice.
2. Materials and methods
2.1. Peptide synthesis of aquaporin 4 General procedures for peptide synthesis and antibody generation were similar to those described previously [28]. Briefly, the peptide (CEKKGKDSSGEVLSSV) of the Cterminal end of the cytoplasmic domain in the rat aquaporin 4 molecule [4,8] was synthesized and extra cysteine was added to the N-terminus of this peptide. Bovine thyroglobulin was added at an extra cysteine residue. The antibody against this peptide was generated in rabbit. Solid phase enzyme-linked immunosorbent assay was used to determine the rabbit polyclonal antibody titer, which was 3204 800.
2.2. Western blot analysis of anti aquaporin 4 antiserum Western blot analysis of antiserum against aquaporin 4 in the normal and mdx mouse quadriceps femoris muscles was carried out by a previously described method with minor modifications [28]. Sodium dodecyl sulfate polyacrylamide gel electrophoresis was performed with a 3– 10% gradient gel. The protein was transferred from the gel to a clear blot P membrane (ATTO, Tokyo Japan) sheet by horizontal electrophoresis at 108 mA for 90 min at room temperature. Immunostaining was done with rabbit anti aquaporin 4 antiserum diluted 1:500.
2.3. Muscle samples and immunohistochemistry Extensor digitorum longus (EDL) and soleus muscles from 6 normal and 6 mdx mice were excised and immediately frozen in isopentane cooled by liquid nitrogen. Frozen 6 mm cross sections of the muscles were placed on coverslips and incubated with primary rabbit antiserum against aquaporin 4 diluted 1:400 in PBS for 24 h at cold room. The sections were then incubated for 30 min with horseradish peroxidase conjugated swine anti rabbit antibody (DAKO A / S, Denmark) diluted 1:40 in PBS. After washing in PBS, the sections were put in 0.2% 3,39diaminobenzidine solution in PBS containing a final concentration of 0.005% H 2 O 2 , and were reacted further for 3 min at room temperature. Control section were stained with normal rabbit serum diluted 1:400 in PBS for 24 h at cold room instead of the primary antiserum against
25
aquaporin 4 and the procedures after this were performed similarly to the antibody positive staining.
2.4. Immunoelectron microscopy and freeze fracture electron microscopy The quadriceps femoris muscles of 6 normal and 6 mdx mice were excised and immediately put in a U-shaped muscle clamp, and immersed for fixation in chilled 4% paraformaldehyde solution in 0.1 M phosphate buffer (pH 7.4) for 1 h. The fixed muscles were washed three times in PBS and frozen in liquid nitrogen-cooled isopentane. The muscles were then cut into thin sections in a cryostat and washed three times in PBS. To eliminate non-specific reactions, the slices were incubated for 30 min at room temperature in PBS containing 5% normal goat serum. For immunolabeling, 1:100 diluted rabbit antiserum against aquaporin 4 was applied as primary antibody to the sections for 24 h at 48C. After thorough rinsing, 5 nm goldlabeled goat anti-rabbit secondary antibody (Amersham UK) was diluted 1:20 in PBS and applied to the sections for 24 h at 48C. Then the sections were washed thoroughly. Control sections were incubated with nonimmune rabbit serum instead of the primary antiserum. The antibody labeled and control muscle samples were further fixed in chilled 2.5% glutaraldehyde solution in 0.1 M phosphate buffer (pH 7.4) for 30 min. These samples were post-fixed in chilled 2% O s O 4 solution for 1 h, dehydrated in an ascending series of ethanol and propylene oxide, and embedded in Epon. The unstained ultrathin sections were observed under the electron microscope. For freeze fracture electron microscopy, the quadriceps femoris muscles of normal mice were excised and immediately put in a U-shaped muscle clamp, and immersed for fixation in chilled 2.5% glutaraldehyde solution in 0.1 M phosphate buffer (pH 7.4) for 1 h. The fascicles of muscles were carefully removed under the dissecting microscope, cut into small blocks and gradually infiltrated in glycerol up to a concentration of 30%. Freezing was carried out in liquid nitrogen. Fracture was performed at 21208C in Eiko FD II A freeze fracture apparatus at a vacuum of 2.5–43 10 27 Torr and immediately replicated with electron beamgunned platinum and carbon. The thickness of the replicas was regulated with a quartz crystal thin film monitor. The muscle tissue was digested in a commercial bleaching solution. The detached replicas were washed twice in distilled water, picked up on uncoated 200-mesh grids and were examined under the electron microscope.
3. Results
3.1. Western blot analysis of anti aquaporin 4 antiserum Western blot analysis showed that the rabbit antiserum against aquaporin 4 reacted with a 28 kDa protein in
26
J.W. Liu et al. / Journal of the Neurological Sciences 164 (1999) 24 – 28
extracts of normal mouse quadriceps femoris muscles (Fig. 1(1)) but did not react with the protein in extracts of quadriceps femoris muscles of mdx mouse (Fig. 1(2)).
3.2. Immunohistochemistry In cross sections of EDL and soleus muscles of normal mice, aquaporin 4 was localized at the surface of individual myofiber (Fig. 2A,B). The cell surface of each myofiber showed a thin layer of immunoreaction product in both EDL and soleus muscles. The staining intensity in both EDL and soleus muscles was almost same in this study (Fig. 2A,B). No immunostain of intracellular structures was noted except for occasional nuclear stain (Fig. 2A,B) which was also seen without staining of myofiber surface membrane in the immunocontrol stain (Fig. 2E,F). Muscle samples from mdx mice revealed a moderate to marked reduction of immunostain for aquaporin 4 in most
Fig. 2. Immunohistochemical staining with anti aquaporin 4 antiserum in EDL (A) and soleus (B) muscles and their negative control staining of EDL (E) and soleus (F) muscles of normal mouse, and EDL (C) and soleus (D) muscles of mdx mouse. In both muscles of normal mouse, the myofiber surface membranes were clearly stained (A, B); while in mdx mouse muscles, the staining pattern was faint and discontinuous (C, D) and the surface membranes of most of the small calibereil myofibers scarcely showed the immunoreaction product of aquaporin 4 in both EDL (C) and soleus (D) muscles. Although occasional nuclear staining was noted in the EDL (A, C) and soleus (B, D) muscles of both normal (A, B) and mdx (C, D) mice, the similar nuclear staining was seen in negative control staining of EDL (E) and soleus (F) muscles of both normal (E, F) and mdx mice. (A–F 3400)
myofibers especially in small calibered muscle fibers (Fig. 2C,D). The mdx myofibers with positive immunostain showed faint and patchy patterns of immunoreaction (Fig. 2C,D). The stainability of aquaporin 4 in EDL muscles (Fig. 2C) of mdx mice was slightly stronger than that in soleus muscles (Fig. 2D) of mdx mice in this study.
3.3. Immunoelectron microscopy and freeze fracture electron microscopy
Fig. 1. Immunoblot analysis of aquaporin 4 expression in quadriceps femoris muscle extracts of normal (lane 1) and mdx (lane 2) mouse. Electrophoresis and blotting were performed as described in ‘Materials and Methods’. Immunostaining was with rabbit anti aquaporin 4 antiserum diluted 1:500 in PBS. In normal mouse muscle extract, the aquaporin 4 band is observed at 28kDa molecular weight; while in mdx muscle extract, the band is not observed. Numbers to the left indicate molecular masses of standards.
Immunoelectron microscopic study showed that the gold particles indicating the signals of aquaporin 4 molecules were present at the inside surface of muscle plasma membrane of normal mice (Fig. 3A) and that they also existed occasionally near caveolae of muscle plasma membrane (Fig. 3B). The signals of aquaporin 4 molecules were not noted as the muscle plasma membrane of mdx mice (Fig. 3C) and immunocontrol muscle samples.
J.W. Liu et al. / Journal of the Neurological Sciences 164 (1999) 24 – 28
Fig. 3. Immunogold electronmicrographs (A, B) show that the signals of aquaporin 4 molecules are noted along the inside surface of muscle plasma membranes of normal mice and they also occasionally existed near caveolae of muscle plasma membrane (arrow in B). The signals of aquaporin 4 molecules are not seen at the muscle plasma membrane of mdx mice (C). Figures D and E depict the higher power magnification views of muscle plasma membrane P and E faces, respectively. The P face contains the individual intramembranous particles of variable size and orthogonal arrays, and E face contains less numerous intramembranous particles and pits of orthogonal arrays. Some of orthogonal arrays and pits of them are present near caveolae (arrow in D, E). C in figures D and E: caveolae. (A–E 3160 000)
The splitting plane of freeze fractured muscle plasma membrane preferentially goes along the hydrophobic interior and yields two leaflets, protoplasmic(P) and extracellular(E) faces. The higher power magnification images of fractured muscle plasma membranes displayed the presence of individual intramembranous particles of variable diameters and orthogonal array particles in P face (Fig. 3D) and less numerous intramembranous particles and orthogonal array pits in E face (Fig. 3E). The presence of orthogonal array particles or pits near the rim of caveolae was occasionally noted (Fig. 3D,E) just as seen in immunoelectron micrograph (Fig. 3B).
4. Discussion Previous investigations using freeze fracture electron microscopy revealed the remarkable depletion of orthogonal arrays in the muscle plasma membrane of DMD and
27
their less extensive decrease in dystrophin deficient mdx mice [21,23,27]. Recent studies indicated that the orthogonal array particles of muscle plasma membrane are water channel protein named aquaporin 4 [9,26,29]. Although Frigeri et al. [6] have very recently reported the expression of aquaporin 4 in fast twitch fibers of normal mouse skeletal muscles, we independently initiated this work and obtained the similar result that the reduced number of fibers was stained by anti aquaporin 4 antiserum in the mdx mice. With regard to the immunostainability of muscle fiber type in normal mice, the immunoreaction was almost the same in both red and white muscle fibers in this study. However, in mdx mice the immunostainability of anti aquaporin 4 antiserum in EDL muscles was slightly stronger than that in soleus muscles. It is interesting to correlate this intensity of immunostainability of mdx muscles with the density of orthogonal arrays in mdx muscles seen by freeze fracture electron microscopy [23]. In mdx muscles, the immunostainability was less intense than that of control muscles in the present study and the orthogonal arrays are described to be decreased in mdx myofibers [23]. Moreover the immunoreactivity of antibody in the small calibered, seemingly regenerating myofibers of mdx mice was very faint in this study. In fact the orthogonal array appeared to be absent in muscle plasma membranes of rat regenerating myofibers [12]. Therefore, these findings correlate well each other. However the array density is described to be more numerous in EDL muscles than in soleus muscles in rats [19] and the immunoreactivity of the anti aquaporin 4 antiserum in normal mice was almost the same between EDL and soleus muscles in this study. The reason of this phenomenon is unclear. The possible explanation is that, although the array density is lower in soleus muscles than in EDL muscles in normal mice, the substantial amount of orthogonal arrays is present in the soleus muscle plasma membrane in normal mice and these orthogonal arrays of soleus muscles may display the normal intensity of immunoreaction for the antiserum against aquaporin 4. The occasional nuclear staining seen in the muscles of normal and mdx mice was considered to be non specific phenomenon, since both positive and control stainings were similarly performed. The immunoelectron microscopy demonstrated that the immunoreaction seen in light microscopic level of this study occurred at the inside surface of the muscle plasma memebrane of normal skeletal myofiber. This finding is reasonable, since the anti aquaporin 4 antiserum was raised against the cytoplasmic domain of aquaporin 4 molecule. The orthogonal arrays seen in freeze fractured skeletal myofibers are present at the protoplasmic half of the muscle plasma membrane. So the immunoreaction of anti aquaporin 4 antiserum at the ultrastructural level in the muscle plasma membranes in this study is considered to represent the location of orthogonal arrays in normal skeletal myofibers.
28
J.W. Liu et al. / Journal of the Neurological Sciences 164 (1999) 24 – 28
Why orthogonal arrays are decreased and, as a result, the immunoreactivity of anti aquaporin 4 antiserum is weak in mdx mice is unknown. Although the dystrophin is deficient in mdx mice [1,3,28,30], the binding between aquaporin 4 and dystrophin, or aquaporim 4 and dystrophin associated (glyco)proteins has not been investigated so far. Beside dystrophin associated (glyco)proteins, talin [22], aciculin [2] and calmodulin [11] have been shown to bind dystrophin. Although aquaporin 4 is not known to associate with dystrophin or with dystrophin associated (glyco)proteins, it is interesting to investigate this question. In immunohistochemical stain, dystrophin is enriched at the costamere [15,17,25], neuromuscular junction [16,24] and the myotendinous junction [20,24]. So the immunohistochemical studies with antibody against aquaporin 4 at these myofiber sites will throw light into the functional significance of aquaporin 4 molecule and its relationship to dystrophin or dystrophin associated (glyco)proteins in normal skeletal myofibers.
Acknowledgements This work was partly supported by grants from Kanagawa intractable disease fund and the National Center for Nervous, Mental and Muscular Disorders of the Ministry of Health and Welfare (8A-2-6), Japan.
References [1] Arahata K, Ishiura S, Ishiguro T, Tsukahara T, Suhara Y, Eguchi C, Ishihara T, Nonaka I, Ozawa E, Sugita H. Immunostaining of skeletal and cardiac muscle surface membrane with antibody against Duchenne muscular dystrophy peptide. Nature 1988;333:861–3. [2] Belkin AM, Burridge K. Association of aciculin with dystrophin and utrophin. J Biol Chem 1995;270:6328–37. [3] Bonilla E, Samitt CE, Miranda AF, Hays AP, Saliiati G, DiMauro S, Kunkel LM, Hoffman EP, Rowland LP. Duchenne muscular dystrophy: deficiency of dystrophin at the muscle cell surface. Cell 1988;54:447–52. [4] Frigeri A, Gropper MA, Turck CW, Verkman AS. Immunolocalization of the mercurial-insensitive water channel and glycerol intrinsic protein in epithelial cell plasma membranes. Proc Natl Acad Sci USA 1995;92:4328–31. [5] Frigeri A, Gropper MA, Umenishi F, Kawashima M, Brown D, Verkman AS. Localization of MIWC and GLIP water channel homologs in neuromuscular, epithelial and glandular tissues. J Cell Sci 1995;108:2993–3002. [6] Frigeri A, Nicchia GP, Verbavatz JM, Velenti G, Svelto M. Expression of aquaporin-4 in fast-twitch fibers of mammalian skeletal muscle. J Clin Invest 1998;102:695–703. [7] Gardner-Medwin D, Walton J. The muscular dystrophies. In: Walton JN, editor. Disorders of Voluntary Muscle, 6th ed, Churchill Livingstone, London, 1994, pp. 543–94. [8] Hasegawa H, Ma T, Skach W, Matthay MA, Verkman AS. Molecular cloning of a mercurial-insensitive water channel expressed in selected water-transporting tissues. J Biol Chem 1994;269:5497– 500. [9] Hatton JD, Cox GF, Miller AL, Nichol JA, Elilisman MH. Identification of polypeptides associated with sarcolemmal vesicles enriched in orthogonal arrays. Biochem Biophys Acta 1987;904:373– 80.
[10] Hoffman EP, Brown RH, Kunkel LM. Dystrophin: the protein product of the Duchenne muscular dystrophy locus. Cell 1987;51:919–28. [11] Jarrett HW, Foster JL. Alternate binding of actin and calmodulin to multiple sites on dystrophin. J Biol Chem 1995;270:5578–86. [12] Jimi T, Wakayama Y. Effect of denervation on regenerating muscle plasma membrane integrity: freeze-fracture and dystrophin immunostaining analyses. Acta Neuropathol (Berl) 1990;80:401–5. [13] Jung JS, Bhat RV, Preston GM, Guggino WB, Baraban JIM, Agre P. Molecular characterization of an aquaporin cDNA from brain: candidate osmoreceptor and regulator of water balance. Proc Natl Acad Sci USA 1994;91:13052–6. [14] Koenig M, Monaco AP, Kunkel LM. The complete sequence of dystrophin predicts a rod-shaped cytoskeletal protein. Cell 1988;53:219–26. [15] Masuda T, Fujimaki N, Ozawa E, Ishikawa H. Confocal laser microscopy of dystrophin localization in guinea pig skeletal muscle fibers. J Cell Biol 1992;119:543–8. [16] Miike T, Miyatake M, Zhao J, Yoshioka K, Uchino M. Immunohistochemical dystrophin reaction in synaptic regions. Brain Dev 1989;11:344–6. [17] Minetti C, Beltrame F, Marcenaro G, Bonilla E. Dystrophin at the plasma membrane of human muscle fibers shows a costameric localization. Neuromuscul Disord 1992;2:99–109. [18] Ozawa E, Noguchi S, Mizuno Y, Hagiwara Y, Toshida M. From dystrophinopathy to sarcoglycanopathy: evolution of a concept of muscular dystrophy. Muscle Nerve 1998;21:421–33. [19] Rash JE, Elilisman MH. Studies of excitable membranes I. Macromolecular specializations of the neuromuscular junction and the nonjunctional sarcolemma. J Cell Biol 1974;63:567–86. [20] Samitt CE, Bonilla E. Immunocytochemical study of dystrophin at the myotendinous junction. Muscle Nerve 1990;13:493–500. [21] Schotland DL, Bonilla E, Wakayama Y. Freeze-fracture studies of muscle plasma membrane in human muscular dystrophy. Acta Neuropathol (Berl) 1981;54:189–97. [22] Senter L, Luise M, Presotto C, Betto R, Teresi A, Ceoldo S, Salviati G. Interaction of dystrophin with cytoskeletal proteins: binding to talin and actin. Biochem Biophys Res Commun 1993;192:899–904. [23] Shibuya S, Wakayama Y. Freeze-fracture studies of myofiber plasma membrane in x chromosome-linked muscular dystrophy (mdx) mice. Acta Neuropathol (Berl) 1988;76:179–84. [24] Shimizu T, Matsumura K, Sunada Y, Mannen T. Dense immunostainings on both neuromuscular and myotendon junctions with an anti-dystrophin monoclonal antibody. Biomed Res 1989;10:405–9. [25] Straub V, Bittner RE, Leger JJ, Voit T. Direct visualization of the dystrophin network on skeletal muscle fiber membrane. J Cell Biol 1992;119:1183–91. [26] Verbavatz JIM, Van Hoek AN, Ma T, Sabolic I, Valenti G, Ellisman MIT, Ausiello DA, Verkman AS, Brown D. A 28kDa sarcolemmal antigen in kidney principal cell basolateral membranes: relationship to orthogonal arrays and MIP26. J Cell Sci 1994;107:1083–94. [27] Wakayama Y, Jimi T, Misugi N, Kumagai T, Miyake S, Shibuya S, Miike T. Dystrophin immunostaining and freeze-fracture studies of muscles of patients with early stage amyotrophic lateral sclerosis and Duchenne muscular dystrophy. J Neurol Sci 1989;91:191–205. [28] Wakayama Y, Jimi T, Takeda A, Misugi N, Kuinagai T, Miyake S, Shibuya S. Immunoreactivity of antibodies raised against synthetic peptide fragments predicted from mid portions of dystrophin cDNA. J Neurol Sci 1990;97:241–50. [29] Yang B, Brown D, Verkman AS. The mercurial insensitive water channel (AQP-4) forms orthogonal arrays in stably transfected chinese hamster ovary cells. J Biol Chem 1996;271:4577–80. [30] Zubrzycka-Gaarn EE, Bulman DE, Karpati G, Bucghes AIT, Belfall B, Klamut HJ, Talbot J, Hodges RS, Ray PN, Wortor RG. The Duchenne muscular dystrophy gene product is localized in sarcolemma of human skeletal muscle. Nature 1988;333:466–9.
Immunocytochemical studies of aquaporin 4 in the skeletal muscle of mdx mouse a
a,
a
a
a
Jian Wu Liu , Yoshihiro Wakayama *, Masahiko Inoue , Seiji Shibuya , Hiroko Kojima , Takahiro Jimi a , Hiroaki Oniki b a
Division of Neurology, Department of Medicine, Showa University Fujigaoka Hospital, 1 -30, Fujigaoka, Aoba-ku, Yokohama 227 -8501, Japan b Electron Microscope Laboratory, Showa University Fujigaoka Hospital, 1 -30 Fujigaoka, Aoba-ku, Yokohama, Japan Received 25 November 1998; accepted 8 February 1999
Abstract 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. 1999 Elsevier Science B.V. All rights reserved. Keywords: Anti aquaporin 4 antibody immunoreactivity; Skeletal muscle; Mdx mouse; Freeze fracture; Muscle plasma membrane; Orthogonal array
1. Introduction Duchenne muscular dystrophy (DMD) is a X-linked, dystrophin deficient disorder and the affected boys show proximal muscle weakness of 4 limbs and calf pseudohypertrophy from childhood, and inevitably die at around the age of 20 [7,10,14]. Although lack of dystrophin leads to muscle cell degeneration and necrosis in DMD, the precise mechanism of muscle cell death has not yet been fully clarified. The DMD gene product, dystrophin, is a membrane cytoskeletal protein located at the inside surface of muscle plasma membrane and does not have transmembranus domain [14,18]. Therefore the freeze fracture preparation of normal muscle does not reveal the dystrophin molecule, since the fracture plane of freeze *Corresponding author. Tel.: 181-45-971-1151; fax: 181-45-9742204.
fracture preferentially goes through the hydrophobic interior of muscle plasma membrane. Schotland et al. [21] and we [27] previously investigated the ultrastructure of muscle plasma membranes of both histochemically normal muscles and DMD muscles by using freeze fracture method, and found marked depletion of orthogonal arrays in DMD muscle plasma membranes. Thereafter Shibuya and Wakayama [23] studied by freeze fracture electron microscopy the muscle plasma membrane of dystrophin deficient mdx (C57BL / 10ScSn-mdx) mice and showed the similar depletion of orthogonal arrays but not as conspicuous as DMD muscles [23]. Orthogonal array is reported to exist more numerously in the plasma membrane of white muscle fibers in rats than in that of red muscle fibers [19]. Recent investigation demonstrated that aquaporin 4 molecule, which is a water channel protein present in a variety of cell types [5], is a 28 kDa integral membrane protein and is seen as orthogon-
0022-510X / 99 / $ – see front matter 1999 Elsevier Science B.V. All rights reserved. PII: S0022-510X( 99 )00051-9
J.W. Liu et al. / Journal of the Neurological Sciences 164 (1999) 24 – 28
al array in freeze fracture electron microscopy [9,26,29]. A cDNA for aquaporin 4 water channel protein was isolated recently from rat brain [8,13]. We generated a polyclonal antibody against the synthetic peptide of cytoplasmic domain of aquaporin 4, and investigated immunohistochemically the muscle plasma membranes of normal (C57BL / 10ScSn) and mdx mice.
2. Materials and methods
2.1. Peptide synthesis of aquaporin 4 General procedures for peptide synthesis and antibody generation were similar to those described previously [28]. Briefly, the peptide (CEKKGKDSSGEVLSSV) of the Cterminal end of the cytoplasmic domain in the rat aquaporin 4 molecule [4,8] was synthesized and extra cysteine was added to the N-terminus of this peptide. Bovine thyroglobulin was added at an extra cysteine residue. The antibody against this peptide was generated in rabbit. Solid phase enzyme-linked immunosorbent assay was used to determine the rabbit polyclonal antibody titer, which was 3204 800.
2.2. Western blot analysis of anti aquaporin 4 antiserum Western blot analysis of antiserum against aquaporin 4 in the normal and mdx mouse quadriceps femoris muscles was carried out by a previously described method with minor modifications [28]. Sodium dodecyl sulfate polyacrylamide gel electrophoresis was performed with a 3– 10% gradient gel. The protein was transferred from the gel to a clear blot P membrane (ATTO, Tokyo Japan) sheet by horizontal electrophoresis at 108 mA for 90 min at room temperature. Immunostaining was done with rabbit anti aquaporin 4 antiserum diluted 1:500.
2.3. Muscle samples and immunohistochemistry Extensor digitorum longus (EDL) and soleus muscles from 6 normal and 6 mdx mice were excised and immediately frozen in isopentane cooled by liquid nitrogen. Frozen 6 mm cross sections of the muscles were placed on coverslips and incubated with primary rabbit antiserum against aquaporin 4 diluted 1:400 in PBS for 24 h at cold room. The sections were then incubated for 30 min with horseradish peroxidase conjugated swine anti rabbit antibody (DAKO A / S, Denmark) diluted 1:40 in PBS. After washing in PBS, the sections were put in 0.2% 3,39diaminobenzidine solution in PBS containing a final concentration of 0.005% H 2 O 2 , and were reacted further for 3 min at room temperature. Control section were stained with normal rabbit serum diluted 1:400 in PBS for 24 h at cold room instead of the primary antiserum against
25
aquaporin 4 and the procedures after this were performed similarly to the antibody positive staining.
2.4. Immunoelectron microscopy and freeze fracture electron microscopy The quadriceps femoris muscles of 6 normal and 6 mdx mice were excised and immediately put in a U-shaped muscle clamp, and immersed for fixation in chilled 4% paraformaldehyde solution in 0.1 M phosphate buffer (pH 7.4) for 1 h. The fixed muscles were washed three times in PBS and frozen in liquid nitrogen-cooled isopentane. The muscles were then cut into thin sections in a cryostat and washed three times in PBS. To eliminate non-specific reactions, the slices were incubated for 30 min at room temperature in PBS containing 5% normal goat serum. For immunolabeling, 1:100 diluted rabbit antiserum against aquaporin 4 was applied as primary antibody to the sections for 24 h at 48C. After thorough rinsing, 5 nm goldlabeled goat anti-rabbit secondary antibody (Amersham UK) was diluted 1:20 in PBS and applied to the sections for 24 h at 48C. Then the sections were washed thoroughly. Control sections were incubated with nonimmune rabbit serum instead of the primary antiserum. The antibody labeled and control muscle samples were further fixed in chilled 2.5% glutaraldehyde solution in 0.1 M phosphate buffer (pH 7.4) for 30 min. These samples were post-fixed in chilled 2% O s O 4 solution for 1 h, dehydrated in an ascending series of ethanol and propylene oxide, and embedded in Epon. The unstained ultrathin sections were observed under the electron microscope. For freeze fracture electron microscopy, the quadriceps femoris muscles of normal mice were excised and immediately put in a U-shaped muscle clamp, and immersed for fixation in chilled 2.5% glutaraldehyde solution in 0.1 M phosphate buffer (pH 7.4) for 1 h. The fascicles of muscles were carefully removed under the dissecting microscope, cut into small blocks and gradually infiltrated in glycerol up to a concentration of 30%. Freezing was carried out in liquid nitrogen. Fracture was performed at 21208C in Eiko FD II A freeze fracture apparatus at a vacuum of 2.5–43 10 27 Torr and immediately replicated with electron beamgunned platinum and carbon. The thickness of the replicas was regulated with a quartz crystal thin film monitor. The muscle tissue was digested in a commercial bleaching solution. The detached replicas were washed twice in distilled water, picked up on uncoated 200-mesh grids and were examined under the electron microscope.
3. Results
3.1. Western blot analysis of anti aquaporin 4 antiserum Western blot analysis showed that the rabbit antiserum against aquaporin 4 reacted with a 28 kDa protein in
26
J.W. Liu et al. / Journal of the Neurological Sciences 164 (1999) 24 – 28
extracts of normal mouse quadriceps femoris muscles (Fig. 1(1)) but did not react with the protein in extracts of quadriceps femoris muscles of mdx mouse (Fig. 1(2)).
3.2. Immunohistochemistry In cross sections of EDL and soleus muscles of normal mice, aquaporin 4 was localized at the surface of individual myofiber (Fig. 2A,B). The cell surface of each myofiber showed a thin layer of immunoreaction product in both EDL and soleus muscles. The staining intensity in both EDL and soleus muscles was almost same in this study (Fig. 2A,B). No immunostain of intracellular structures was noted except for occasional nuclear stain (Fig. 2A,B) which was also seen without staining of myofiber surface membrane in the immunocontrol stain (Fig. 2E,F). Muscle samples from mdx mice revealed a moderate to marked reduction of immunostain for aquaporin 4 in most
Fig. 2. Immunohistochemical staining with anti aquaporin 4 antiserum in EDL (A) and soleus (B) muscles and their negative control staining of EDL (E) and soleus (F) muscles of normal mouse, and EDL (C) and soleus (D) muscles of mdx mouse. In both muscles of normal mouse, the myofiber surface membranes were clearly stained (A, B); while in mdx mouse muscles, the staining pattern was faint and discontinuous (C, D) and the surface membranes of most of the small calibereil myofibers scarcely showed the immunoreaction product of aquaporin 4 in both EDL (C) and soleus (D) muscles. Although occasional nuclear staining was noted in the EDL (A, C) and soleus (B, D) muscles of both normal (A, B) and mdx (C, D) mice, the similar nuclear staining was seen in negative control staining of EDL (E) and soleus (F) muscles of both normal (E, F) and mdx mice. (A–F 3400)
myofibers especially in small calibered muscle fibers (Fig. 2C,D). The mdx myofibers with positive immunostain showed faint and patchy patterns of immunoreaction (Fig. 2C,D). The stainability of aquaporin 4 in EDL muscles (Fig. 2C) of mdx mice was slightly stronger than that in soleus muscles (Fig. 2D) of mdx mice in this study.
3.3. Immunoelectron microscopy and freeze fracture electron microscopy
Fig. 1. Immunoblot analysis of aquaporin 4 expression in quadriceps femoris muscle extracts of normal (lane 1) and mdx (lane 2) mouse. Electrophoresis and blotting were performed as described in ‘Materials and Methods’. Immunostaining was with rabbit anti aquaporin 4 antiserum diluted 1:500 in PBS. In normal mouse muscle extract, the aquaporin 4 band is observed at 28kDa molecular weight; while in mdx muscle extract, the band is not observed. Numbers to the left indicate molecular masses of standards.
Immunoelectron microscopic study showed that the gold particles indicating the signals of aquaporin 4 molecules were present at the inside surface of muscle plasma membrane of normal mice (Fig. 3A) and that they also existed occasionally near caveolae of muscle plasma membrane (Fig. 3B). The signals of aquaporin 4 molecules were not noted as the muscle plasma membrane of mdx mice (Fig. 3C) and immunocontrol muscle samples.
J.W. Liu et al. / Journal of the Neurological Sciences 164 (1999) 24 – 28
Fig. 3. Immunogold electronmicrographs (A, B) show that the signals of aquaporin 4 molecules are noted along the inside surface of muscle plasma membranes of normal mice and they also occasionally existed near caveolae of muscle plasma membrane (arrow in B). The signals of aquaporin 4 molecules are not seen at the muscle plasma membrane of mdx mice (C). Figures D and E depict the higher power magnification views of muscle plasma membrane P and E faces, respectively. The P face contains the individual intramembranous particles of variable size and orthogonal arrays, and E face contains less numerous intramembranous particles and pits of orthogonal arrays. Some of orthogonal arrays and pits of them are present near caveolae (arrow in D, E). C in figures D and E: caveolae. (A–E 3160 000)
The splitting plane of freeze fractured muscle plasma membrane preferentially goes along the hydrophobic interior and yields two leaflets, protoplasmic(P) and extracellular(E) faces. The higher power magnification images of fractured muscle plasma membranes displayed the presence of individual intramembranous particles of variable diameters and orthogonal array particles in P face (Fig. 3D) and less numerous intramembranous particles and orthogonal array pits in E face (Fig. 3E). The presence of orthogonal array particles or pits near the rim of caveolae was occasionally noted (Fig. 3D,E) just as seen in immunoelectron micrograph (Fig. 3B).
4. Discussion Previous investigations using freeze fracture electron microscopy revealed the remarkable depletion of orthogonal arrays in the muscle plasma membrane of DMD and
27
their less extensive decrease in dystrophin deficient mdx mice [21,23,27]. Recent studies indicated that the orthogonal array particles of muscle plasma membrane are water channel protein named aquaporin 4 [9,26,29]. Although Frigeri et al. [6] have very recently reported the expression of aquaporin 4 in fast twitch fibers of normal mouse skeletal muscles, we independently initiated this work and obtained the similar result that the reduced number of fibers was stained by anti aquaporin 4 antiserum in the mdx mice. With regard to the immunostainability of muscle fiber type in normal mice, the immunoreaction was almost the same in both red and white muscle fibers in this study. However, in mdx mice the immunostainability of anti aquaporin 4 antiserum in EDL muscles was slightly stronger than that in soleus muscles. It is interesting to correlate this intensity of immunostainability of mdx muscles with the density of orthogonal arrays in mdx muscles seen by freeze fracture electron microscopy [23]. In mdx muscles, the immunostainability was less intense than that of control muscles in the present study and the orthogonal arrays are described to be decreased in mdx myofibers [23]. Moreover the immunoreactivity of antibody in the small calibered, seemingly regenerating myofibers of mdx mice was very faint in this study. In fact the orthogonal array appeared to be absent in muscle plasma membranes of rat regenerating myofibers [12]. Therefore, these findings correlate well each other. However the array density is described to be more numerous in EDL muscles than in soleus muscles in rats [19] and the immunoreactivity of the anti aquaporin 4 antiserum in normal mice was almost the same between EDL and soleus muscles in this study. The reason of this phenomenon is unclear. The possible explanation is that, although the array density is lower in soleus muscles than in EDL muscles in normal mice, the substantial amount of orthogonal arrays is present in the soleus muscle plasma membrane in normal mice and these orthogonal arrays of soleus muscles may display the normal intensity of immunoreaction for the antiserum against aquaporin 4. The occasional nuclear staining seen in the muscles of normal and mdx mice was considered to be non specific phenomenon, since both positive and control stainings were similarly performed. The immunoelectron microscopy demonstrated that the immunoreaction seen in light microscopic level of this study occurred at the inside surface of the muscle plasma memebrane of normal skeletal myofiber. This finding is reasonable, since the anti aquaporin 4 antiserum was raised against the cytoplasmic domain of aquaporin 4 molecule. The orthogonal arrays seen in freeze fractured skeletal myofibers are present at the protoplasmic half of the muscle plasma membrane. So the immunoreaction of anti aquaporin 4 antiserum at the ultrastructural level in the muscle plasma membranes in this study is considered to represent the location of orthogonal arrays in normal skeletal myofibers.
28
J.W. Liu et al. / Journal of the Neurological Sciences 164 (1999) 24 – 28
Why orthogonal arrays are decreased and, as a result, the immunoreactivity of anti aquaporin 4 antiserum is weak in mdx mice is unknown. Although the dystrophin is deficient in mdx mice [1,3,28,30], the binding between aquaporin 4 and dystrophin, or aquaporim 4 and dystrophin associated (glyco)proteins has not been investigated so far. Beside dystrophin associated (glyco)proteins, talin [22], aciculin [2] and calmodulin [11] have been shown to bind dystrophin. Although aquaporin 4 is not known to associate with dystrophin or with dystrophin associated (glyco)proteins, it is interesting to investigate this question. In immunohistochemical stain, dystrophin is enriched at the costamere [15,17,25], neuromuscular junction [16,24] and the myotendinous junction [20,24]. So the immunohistochemical studies with antibody against aquaporin 4 at these myofiber sites will throw light into the functional significance of aquaporin 4 molecule and its relationship to dystrophin or dystrophin associated (glyco)proteins in normal skeletal myofibers.
Acknowledgements This work was partly supported by grants from Kanagawa intractable disease fund and the National Center for Nervous, Mental and Muscular Disorders of the Ministry of Health and Welfare (8A-2-6), Japan.
References [1] Arahata K, Ishiura S, Ishiguro T, Tsukahara T, Suhara Y, Eguchi C, Ishihara T, Nonaka I, Ozawa E, Sugita H. Immunostaining of skeletal and cardiac muscle surface membrane with antibody against Duchenne muscular dystrophy peptide. Nature 1988;333:861–3. [2] Belkin AM, Burridge K. Association of aciculin with dystrophin and utrophin. J Biol Chem 1995;270:6328–37. [3] Bonilla E, Samitt CE, Miranda AF, Hays AP, Saliiati G, DiMauro S, Kunkel LM, Hoffman EP, Rowland LP. Duchenne muscular dystrophy: deficiency of dystrophin at the muscle cell surface. Cell 1988;54:447–52. [4] Frigeri A, Gropper MA, Turck CW, Verkman AS. Immunolocalization of the mercurial-insensitive water channel and glycerol intrinsic protein in epithelial cell plasma membranes. Proc Natl Acad Sci USA 1995;92:4328–31. [5] Frigeri A, Gropper MA, Umenishi F, Kawashima M, Brown D, Verkman AS. Localization of MIWC and GLIP water channel homologs in neuromuscular, epithelial and glandular tissues. J Cell Sci 1995;108:2993–3002. [6] Frigeri A, Nicchia GP, Verbavatz JM, Velenti G, Svelto M. Expression of aquaporin-4 in fast-twitch fibers of mammalian skeletal muscle. J Clin Invest 1998;102:695–703. [7] Gardner-Medwin D, Walton J. The muscular dystrophies. In: Walton JN, editor. Disorders of Voluntary Muscle, 6th ed, Churchill Livingstone, London, 1994, pp. 543–94. [8] Hasegawa H, Ma T, Skach W, Matthay MA, Verkman AS. Molecular cloning of a mercurial-insensitive water channel expressed in selected water-transporting tissues. J Biol Chem 1994;269:5497– 500. [9] Hatton JD, Cox GF, Miller AL, Nichol JA, Elilisman MH. Identification of polypeptides associated with sarcolemmal vesicles enriched in orthogonal arrays. Biochem Biophys Acta 1987;904:373– 80.
[10] Hoffman EP, Brown RH, Kunkel LM. Dystrophin: the protein product of the Duchenne muscular dystrophy locus. Cell 1987;51:919–28. [11] Jarrett HW, Foster JL. Alternate binding of actin and calmodulin to multiple sites on dystrophin. J Biol Chem 1995;270:5578–86. [12] Jimi T, Wakayama Y. Effect of denervation on regenerating muscle plasma membrane integrity: freeze-fracture and dystrophin immunostaining analyses. Acta Neuropathol (Berl) 1990;80:401–5. [13] Jung JS, Bhat RV, Preston GM, Guggino WB, Baraban JIM, Agre P. Molecular characterization of an aquaporin cDNA from brain: candidate osmoreceptor and regulator of water balance. Proc Natl Acad Sci USA 1994;91:13052–6. [14] Koenig M, Monaco AP, Kunkel LM. The complete sequence of dystrophin predicts a rod-shaped cytoskeletal protein. Cell 1988;53:219–26. [15] Masuda T, Fujimaki N, Ozawa E, Ishikawa H. Confocal laser microscopy of dystrophin localization in guinea pig skeletal muscle fibers. J Cell Biol 1992;119:543–8. [16] Miike T, Miyatake M, Zhao J, Yoshioka K, Uchino M. Immunohistochemical dystrophin reaction in synaptic regions. Brain Dev 1989;11:344–6. [17] Minetti C, Beltrame F, Marcenaro G, Bonilla E. Dystrophin at the plasma membrane of human muscle fibers shows a costameric localization. Neuromuscul Disord 1992;2:99–109. [18] Ozawa E, Noguchi S, Mizuno Y, Hagiwara Y, Toshida M. From dystrophinopathy to sarcoglycanopathy: evolution of a concept of muscular dystrophy. Muscle Nerve 1998;21:421–33. [19] Rash JE, Elilisman MH. Studies of excitable membranes I. Macromolecular specializations of the neuromuscular junction and the nonjunctional sarcolemma. J Cell Biol 1974;63:567–86. [20] Samitt CE, Bonilla E. Immunocytochemical study of dystrophin at the myotendinous junction. Muscle Nerve 1990;13:493–500. [21] Schotland DL, Bonilla E, Wakayama Y. Freeze-fracture studies of muscle plasma membrane in human muscular dystrophy. Acta Neuropathol (Berl) 1981;54:189–97. [22] Senter L, Luise M, Presotto C, Betto R, Teresi A, Ceoldo S, Salviati G. Interaction of dystrophin with cytoskeletal proteins: binding to talin and actin. Biochem Biophys Res Commun 1993;192:899–904. [23] Shibuya S, Wakayama Y. Freeze-fracture studies of myofiber plasma membrane in x chromosome-linked muscular dystrophy (mdx) mice. Acta Neuropathol (Berl) 1988;76:179–84. [24] Shimizu T, Matsumura K, Sunada Y, Mannen T. Dense immunostainings on both neuromuscular and myotendon junctions with an anti-dystrophin monoclonal antibody. Biomed Res 1989;10:405–9. [25] Straub V, Bittner RE, Leger JJ, Voit T. Direct visualization of the dystrophin network on skeletal muscle fiber membrane. J Cell Biol 1992;119:1183–91. [26] Verbavatz JIM, Van Hoek AN, Ma T, Sabolic I, Valenti G, Ellisman MIT, Ausiello DA, Verkman AS, Brown D. A 28kDa sarcolemmal antigen in kidney principal cell basolateral membranes: relationship to orthogonal arrays and MIP26. J Cell Sci 1994;107:1083–94. [27] Wakayama Y, Jimi T, Misugi N, Kumagai T, Miyake S, Shibuya S, Miike T. Dystrophin immunostaining and freeze-fracture studies of muscles of patients with early stage amyotrophic lateral sclerosis and Duchenne muscular dystrophy. J Neurol Sci 1989;91:191–205. [28] Wakayama Y, Jimi T, Takeda A, Misugi N, Kuinagai T, Miyake S, Shibuya S. Immunoreactivity of antibodies raised against synthetic peptide fragments predicted from mid portions of dystrophin cDNA. J Neurol Sci 1990;97:241–50. [29] Yang B, Brown D, Verkman AS. The mercurial insensitive water channel (AQP-4) forms orthogonal arrays in stably transfected chinese hamster ovary cells. J Biol Chem 1996;271:4577–80. [30] Zubrzycka-Gaarn EE, Bulman DE, Karpati G, Bucghes AIT, Belfall B, Klamut HJ, Talbot J, Hodges RS, Ray PN, Wortor RG. The Duchenne muscular dystrophy gene product is localized in sarcolemma of human skeletal muscle. Nature 1988;333:466–9.