- Email: [email protected]
Profound skeletal muscle depletion of α-dystroglycan in Walker-Warburg syndrome
doi:10.1016/S1090-3798(03)00042-4 European Journal of Paediatric Neurology 2003; 7: 129–137
ORIGINAL ARTICLE
Profound skeletal muscle depletion of a-dystroglycan in Walker-Warburg syndrome CECILIA JIME´NEZ-MALLEBRERA,1 SILVIA TORELLI,1 SUSAN C BROWN,1 LUCY FENG,1 MARTIN BROCKINGTON,1 CAROLINE A SEWRY,1 DANIEL BELTRA´N-VALERO DE BERNABE´,2 FRANCESCO MUNTONI1 1
Neuromuscular Unit, Imperial College School of Medicine, Hammersmith Hospital, London, UK; 2Department of Human Genetics, University Medical Centre Nijmegen, The Netherlands
Walker-Warburg syndrome (WWS) is an autosomal recessive disorder characterized by the combined involvement of the central nervous and skeletal muscle systems. Although the molecular basis of WWS remains unknown, defects in the muscle fibre basal lamina are characteristic of other forms of congenital muscular dystrophy (CMD). In agreement with this, some forms of CMD, due to glycosyltransferase defects, display a reduction in the immunolabelling of a-dystroglycan, whilst b-dystroglycan labelling appears normal. Here we describe an almost complete absence of a-dystroglycan using both immunohistochemistry and immunoblotting in two patients with WWS. In addition, there was a mild reduction of laminin-a2. In contrast, immunohistochemical labelling of perlecan and collagen VI was normal. Linkage analysis excluded the recently identified POMT1 locus, responsible for a proportion of WWS cases. These results confirm that WWS is a genetically heterogeneous condition and suggest that disruption of the a-dystroglycan/laminin-a2 axis in the basal lamina may play a role in the degeneration of muscle fibres in WWS—also in cases not due to POMT1 defects. Keywords: Walker-Warburg syndrome. a-Dystroglycan. Protein glycosylation. Extracellular matrix. Immunohistochemistry. Immunoblotting.
Introduction Walker-Warburg syndrome [MIM 236670] is a lethal autosomal recessive disorder characterized by the combined involvement of the central nervous and skeletal muscle systems. Brain malformations with congenital muscular dystrophy are also found in other forms of congenital muscular dystrophy such as muscle-eye-brain disease (MEB, [MIM 253280]) and Fukuyama congenital muscular dystrophy (FCMD, [MIM 253800]). In addition, severe ocular abnormalities involving the retina and the anterior chamber are present in WWS and MEB but not in FCMD. The diagnosis of WWS is based on clinical (congenital hypotonia and weakness, retinal malformation), radiological (type II lissencephaly,
cerebellar malformation) and pathological observations (congenital muscular dystrophy).1 Since the clinical features of WWS and MEB overlap significantly (to a higher extent than to FCMD) it was suggested that MEB and WWS may be allelic disorders. However, after the MEB locus was mapped to 1p32-p34,2 linkage analysis showed that both disorders are genetically distinct.3 Similarly, FCMD can be genetically excluded by linkage analysis to 9q31-q33.4 The genes defective in FCMD ( fukutin gene) and in MEB (POMGnT1 gene) have been identified,4,5 and encode proteins with a potential role in protein modification/ glycosylation.5,6 The genetic and molecular basis of WWS are just starting to be elucidated. Very recently, mutations in the POMT1 gene have been reported in a proportion of WWS patients.7 This gene encodes a glycosyltransferase which
Received 18.12.02. Accepted 28.2.03. Correspondence: Cecilia Jime´nez-Mallebrera, Neuromuscular Unit, Hammersmith Hospital, Du Cane Road, London W12 0NN, UK. Tel: þ 44-20-8383-2126; Fax: þ 44-20-8746-2187; e-mail: [email protected]
1090-3798/03/07/0129+9 $35.00
Q 2003 European Paediatric Neurology Society
130 catalyses the first step in O-mannosyl glycan synthesis. However, genome-wide linkage analysis suggests that WWS is genetically heterogeneous and indeed mutations in the POMT1 were only found in 20% of families studied.7 The expression of extracellular matrix proteins is abnormal (reduced or absent) in several forms of congenital muscular dystrophy. This includes laminin-a2 chain of laminin-2 in MDC1A8 [MIM 156225]; integrin a7 in Integrin a7 deficiency9 [MIM 600536]; collagen VI in Ullrich CMD10 (UCMD, [MIM 254090]). In addition to these primary defects, secondary changes can be seen. For example, a secondary reduction in the expression level of laminin-a2 chain has been described in both FCMD11 and MEB.12 Furthermore, in an increasing number of CMD forms with and without brain involvement, which include FCMD, MEB, MDC1B and MDC1C, the expression of a-dystroglycan, but not b-dystroglycan, is reduced to a variable extent.13 – 16 Regarding laminin-a2 chain expression in WWS, some authors have shown preserved expression17 while others found a decreased expression.18 There is also one report of reduced laminin-b2 expression in WWS muscle.19 Using electron microscopy, previous authors have noted basal lamina abnormalities (described as thinning and disruption in non-necrotic fibres) in the skeletal muscle fibres of WWS patients.20 Interestingly, disruption of the BM has also been described in FCMD21 and in primary laminin-a2 deficient congenital muscular dystrophy.22,23 To gain insight into the pathogenesis of WWS we have examined the expression of a number of sarcolemmal, trans-sarcolemmal and extracellular matrix proteins in the skeletal muscle of two patients with classical WWS features. Genetic analysis in one of the two cases allowed us to rule out an involvement of the recently identified POMT1 gene, mutated in a proportion of cases of WWS. We found an almost complete absence of a-dystroglycan and a mild reduction of laminin-a2 using both immunohistochemistry and immunoblotting. In contrast, expression of perlecan and collagen VI was essentially normal. These results suggest that impaired processing of a-dystroglycan may be at the centre of the dystrophic process in WWS.
Materials and methods Immunocytochemistry Frozen 8 mm sections were incubated with monoclonal antibodies to spectrin and dystrophin
Original article: C Jime´nez-Mallebrera et al. (NCL-SPEC1, NCL-DYS1, NCL-DYS2 and NCLDYS3, Novocastra laboratories Ltd), the neonatal form of myosin heavy chain (NCL-MHCn, Novocastra laboratories Ltd), laminin-a2 (MAB1922 Chemicon and 4H8, Alexis corporation), collagen VI (MAB1944 and MAB3303, Chemicon), perlecan (MAB1948, Chemicon), b-dystroglycan (NCL-bDG, Novocastra Laboratories Ltd), a-dystroglycan (VIA4-1, Upstate Biotechnology) and sheep polyclonal raised against a 20 aa peptide from the chick a-dystroglycan sequence that recognizes the core protein of a-dystroglycan24 (kind gift of Dr Stephan Kro¨ger). VIA4-1 is a monoclonal antibody to an as yet unidentified epitope (possibly a carbohydrate). Prior to staining with VIA4-1, sections were fixed in 50% acetic acid/ethanol for 6 minutes at room temperature and non-specific binding of this antibody was blocked treating sections with 3% BSA for 30 minutes at room temperature. All primary antibodies were applied for 1 hour and revealed with an appropriate biotinylated secondary antibody (Amersham 1:200) for 30 minutes, followed by streptavidin conjugated to Alexa 594 (Molecular Probes) for 15 minutes. All dilutions and washings were made in phosphate buffered saline. Sections were mounted in aqueous mountant and viewed with epifluorescence using a Leica DMR microscope linked to Metamorph (Universal Imaging). Control sections were labelled without primary antibodies, and all sections were compared with age-matched control samples from other neuromuscular disorders and with normal muscle.
Western blotting Muscle proteins were extracted in sample buffer consisting of 1 M Tris HCl, 1% SDS, glycerol, 2-mercaptoethanol plus a cocktail of protease inhibitors (antipain, aprotinin and leupeptin) and quantified using the BCA assay (Pierce). Soluble protein (50 mg) was resolved using a NuPage Precast gel (4 –12% Bis-Tris) (Invitrogen) and then electrophoretically transferred to nitrocellulose membrane (Amersham). Membranes were blocked in 3% BSA to detect a-dystroglycan and 5% milk powder to detect laminin-a2, both in TBST Buffer. Membranes were probed with antibodies to adystroglycan (VIA4-1, 1:1000), b-dystroglycan (Novocastra 1:25), laminin-a2 (MAB1922 Chemicon 1:1000), washed, and incubated with HRP-anti mouse (Jackson Laboratories). Polypeptides were visualized using chemiluminescence (ECL þ Plus, Amersham).
Original article: a-Dystroglycan abnormalities in Walker-Warburg muscle
Haplotype analysis DNA for genetic analysis was available from patient 1 and his family but not from family 2. DNA was extracted from blood lymphocytes collected from patient 1, parents and the healthy sibling. The MEB, FCMD and FKRP loci, that have been associated with CMD and brain involvement, or with abnormal processing of a-dystroglycan, were studied using linkage analysis in this consanguineous family. The markers used were: for the MDC1C locus the dinucleotide repeat markers D19S219 and D19S606;15 for the FCMD locus we used microsatellite markers D9S306, D9S2105, D9S2107, D9S2170, D9S2171 and D9S172, which map to chromosome 9q31. It is thought that the Fukutin gene is present in very close proximity to centromeric D9S2170. The order of these markers is as follows: cen-D9S306-D9S2105-FCMD-D9S2170D9S2171-D9S2107-D9S172-tel; the distance between D9S2105 and D9S2107 is approximately 230 kb. Regarding the MEB locus, this was studied with the markers D1S211, D1S2677, D1S427, D1S2652, and D1S200.2 In addition, patient 1 was also studied for the flanking markers to the POMT1 locus, D9S260 and D9S1793, which lie 3.1 cm centromeric and 3.9 telomeric to the locus, respectively.
Results Patients Patient 1 Patient 1 was a male infant who was referred to us at the age of 2 months and 1 week with severe neonatal hypotonia, developmental delay and poor visual behaviour. He was the third child of a Bengali consanguineous family; his older siblings were healthy. Congenital hydrocephalus had been noticed in the past that required shunt at the age of 2 weeks. His serum creatine kinase (CK) levels were markedly elevated in the first few weeks of life and an opthalmological examination revealed bilateral lamellar cataracts. His feeding had been very poor and he had been nasogastrically fed since birth. On examination at the age of 5 weeks, he was extremely floppy with a ‘frog-like’ posture, showing no antigravity proximal movements in the lower and upper limbs, but only distal hands and feet movements. He had dysmorphic features with macrocephaly despite the shunting, prominent eyes (bupthalmos) secondary to anterior
131
chamber abnormalities, and poorly formed muscle bulk. He had no head control on traction or ventral suspension and no apparent response to visuoauditory stimuli. His reflexes were absent. A brain MRI at the age of 5 weeks showed a severe hydrocephalus and the thin residual cortical mantle was smooth in appearance. The cerebellar hemispheres were also severely hypoplastic. A muscle biopsy was obtained from his quadriceps at the age of 2 weeks. The patient died at the age of 8 months during an episode of upper respiratory tract infection; he had shown no psychomotor development until then. A few years later the family was referred back because mother was pregnant. The pregnancy was carefully followed with serial ultrasound scans from the age of 12 weeks. At the age of 18 weeks gestation a severe hydrocephalus was noticed in the fetus, which was therefore considered to be similarly affected. Pregnancy was terminated a few weeks afterwards. The muscle of this fetus was not stored for immunocytochemical analysis and is not included in this study.
Patient 2 Patient 2 is the first child of a consanguineous family and presented with similar features to patient 1 which can be summarized as prenatal onset with ventriculomegaly detected with an ultrasound at 23 weeks, which was confirmed by the later findings (at 30 weeks) of ventriculomegaly, absent corpus callosum and enlarged third and fourth ventricle detected. At birth the baby had poor spontaneous respiratory efforts requiring elective intubation. Examination revealed extended legs because of breech presentation, contractures at the knees and poor spontaneous movements. Hydrocephalus with a posterior fossa cyst was confirmed postnatally and the cerebellum appeared hypoplastic. Ophthalmological examination revealed a bilateral micocornea, posterior lens opacities, vitreal dysplasia and bilateral retinal dysplasia. She was ventilated for a week but then extubated in air. She had a ventriculoperitoneal shunt inserted aged 3 months. She was discharged home at that stage and is stable, although her feeding difficulties persist. A muscle biopsy of the quadriceps was performed at the age of 6.5 weeks and analysed by standard histological and histochemical techniques. This biopsy was referred to us for specialized analysis when the patient was 3 months old.
132
Haplotype analysis The lack of homozygosity at the corresponding loci allowed us to exclude the involvement of the MEB, FCMD and FKRP loci. In addition, patient 1 was also shown to be heterozygous for the flanking markers to the POMT1 locus, i.e. D9S260 and D9S1793.7 These results are in accordance with recently published data that suggest that WWS is a genetically heterogeneous condition. We hypothesize that patient 1 suffers from a subtype of WWS not caused by mutations in the POMT1 gene.
Histopathology and immunocytochemistry The muscle biopsy of patient 1 showed an abnormal variation in fibre size with several small fibres and some larger ones and features of immature muscle. Fibre type differentiation as indicated by oxidative enzyme staining was indistinct. Most fibres co-expressed the neonatal and fast isoforms
Original article: C Jime´nez-Mallebrera et al. of myosin heavy chain (Fig. 2). There was no apparent increase in connective tissue, no excess glycogen and no apparent accumulation of fat. Patient 2 showed a wider variation in fibre size than patient 1 and in contrast to patient 1, an excess of connective tissue. A large proportion of fibres expressed neonatal myosin (Fig. 2). Both patients showed a similar pattern of immunolabelling with all the antibodies used although some differences were noted. Labelling of the sarcolemmal proteins spectrin and dystrophin was within normal limits although some variation in the labelling intensity of individual fibres was seen with the antibody directed against the COOH-terminus of dystrophin (Fig. 1). This variation was more marked in the muscle from patient 2 which showed some very weak/ almost negative fibres (Fig. 1). The significance of this observation is unclear but it could relate to immaturity. In both cases, a-dystroglycan staining using two different antibodies showed a profound reduction in the level of immunolabelling on all fibres
Fig. 1. Muscle sections from patients 1 and 2 and from a neonate patient with a neurometabolic condition used as a control. Sections were immunolabelled with antibodies to spectrin and dystrophin (NCL-DY2 carboxy-terminus and NCL-DYS3 amino-terminal portion of rod domain). Scale bar (bottom left) ¼ 200 mm.
Original article: a-Dystroglycan abnormalities in Walker-Warburg muscle (Fig. 2). Labelling of the transmembrane protein b-dystroglycan, however, was well preserved although with some variability (Fig. 2, some bright large fibres and some small weak fibres). Expression of the laminin-a2 chain of laminin-2 isoform was assessed using an antibody against the 300 kDa fragment and an antibody against its COOH-terminal 80 kDa fragment. In both cases the labelling was mildly reduced with some isolated large fibres (Fig. 3) that were not labelled with either of the two antibodies or were only labelled very weakly. These are the same fibres that labelled very intensely with the b-dystroglycan antibody. b-Spectrin appeared normal in these large fibres indicating that this was not due to non-specific damage of the plasma membrane. This is also unlikely to be due to regeneration since these fibres were not labelled with the
133
neonatal myosin antibody. The reduction of laminin-a2 immunolabelling was more significant in patient 1 and in both cases it was more apparent with the antibody against the NH2-terminal 300 kDa fragment (Fig. 3). In the muscle of patient 1 labelling of the laminin-b1 and -g1 chains was comparable with the control muscle while in patient 2 some fibres showed weak laminin-b1 labelling compared with the control muscle and to laminin-g1 chain labelling (data not shown). In addition to laminin-a2, a-dystroglycan also binds to perlecan (proteoglycan heparan sulphate) agrin and collagen VI. However, the immunolabelling of perlecan and collagen VI was not significantly different to controls in either patient despite the almost complete absence of a-dystroglycan (Fig. 3).
Fig. 2. a-Dystroglycan is absent in the muscle of patients 1 and 2 using both antibodies (VIA4-1 and core) whereas bdystroglycan appears normal. The proportion of neonatal myosin positive fibres in patients 1 and 2 seems larger than in the control muscle. Scale bar (bottom left) ¼ 200 mm.
134
Original article: C Jime´nez-Mallebrera et al.
Fig. 3. In WWS muscle some fibres are weakly labelled with two anti-laminin-a2 antibodies. In contrast, collagen VI and perlecan appear essentially normal (note that collagen VI is slightly reduced in patient 2 but not in patient 1). Scale bar (bottom left) ¼ 200 mm.
Western blotting In order to investigate further the reduction of a-dystroglycan an SDS extract from the patient’s muscle was analysed by Western blotting using the VIA4-1 antibody (Fig. 4A). This antibody recognized a broad band of 150 – 200 kDa, which corresponds to the highly glycosylated form of a-dystroglycan, in normal adult and neonatal muscle (and in an extract prepared from C2C12 cell culture and used as a positive control). No reactive band was detected in the either patient 1 or patient 2 extracts. For further comparisons, the WWS patients’ extract was run alongside an extract from a Duchenne muscular dystrophy (DMD) patient and an aged-matched patient with an undiagnosed form of laminin-a2 positive CMD. The intensity and molecular weight of the a-dystroglycan polypep-
tide in DMD and CMD muscle were comparable with the control sample which confirms that the reduction of a-dystroglycan in WWS muscle is not a non-specific consequence of the dystrophic process. A reduction of the 80 kDa fragment of laminin-a2 in WWS muscle was also revealed by Western blotting (Fig. 4B). The antibody that recognizes the 300 kDa laminin-a2 band is not suitable for Western blotting. The laminin-a2 band was almost undetectable in patient 2 and was significantly reduced in patient 1 when compared with the intensities of the bands seen in control fetal muscle (16 weeks gestation), control neonate muscle and in a muscle extract from an age-matched patient with an unclassified form of laminin-a2 positive CMD (Fig. 4B). To ensure that these findings were not due to protein degradation or low protein content,
Original article: a-Dystroglycan abnormalities in Walker-Warburg muscle
135
Fig. 4. Muscle proteins analysed by Western blotting using antibodies to A, a-dystroglycan, B, laminin-a2 and C, bdystroglycan. a-Dystroglycan (.150 kDa) is undetectable in patient 1 (P1) and patient 2 (P2). The 80 kDa fragment of laminin-a2 is almost absent in patient 2 and reduced in patient 1. b-Dystroglycan levels in WWS extracts are comparable with those in the control extracts. A.ctrol ¼ adult control muscle; N.ctrol ¼ neonate control muscle; CMD ¼ laminin-a2 positive congenital muscular dystrophy muscle; DMD ¼ Duchenne muscular dystrophy muscle. 16w ¼ 16 weeks control fetal muscle; C2C12 ¼ human skeletal muscle cell line C2C12.
the patient’s muscle extract was also probed with an antibody against b-dystroglycan. The intensities of the corresponding 43 kDa b-dystroglycan polypeptide bands were within normal limits (Fig. 4C).
Discussion The dystrophin associated complex has been attributed with a role in stabilizing the sarcolemma25 although there is now increasing evidence that this protein assembly of which dystroglycan is a central part, also plays a role in cell signalling.26 In this paper we have identified a severe depletion of a-dystroglycan, but not b-dystroglycan, in two patients with the characteristic clinical features of Walker-Warburg syndrome. In one of these patients all the molecularly defined CMD variants with similar clinicopathological features were excluded at the molecular level, including WWS caused by mutations in the POMT1 gene.7 The almost complete absence of immunolabelling using either the antibody to the core protein or to
the carbohydrate epitope recognized by VIA4-1 differs from what has been described in other forms of CMD. In these patients there is either a secondary reduction in a-dystroglycan in which the expression of the core protein is well preserved (as seen in MEB14 and in patients with mutations in the POMT1 gene7) or there is some residual immunolabelling with all the antibodies used (as seen in MDC1C15). Furthermore, on immunoblotting, VIA4-1 did not detect a hypoglycosylated species of a-dystroglycan of reduced molecular weight as reported for example in MDC1C muscle.15 Taken together, these observations indicate a severe depletion of a-dystroglycan and not necessarily only of the glycosylated form. The deficient core staining we observed using an antibody that was raised against core a-dystroglycan might indicate that the stability of the entire molecule is affected in this patient. However this should be confirmed with antibodies that recognize a different epitope within the core protein or that work on Western blot. Unfortunately these additional antibodies are currently not freely available and therefore our interpretation remains speculative.
136 a-Dystroglycan is the main receptor for laminina2 in skeletal muscle.25 However, in the present study, we found almost normal or partly reduced immunolabelling of laminin-a2 in the basement membrane. Given the complete absence of a-dystroglycan it is possible to speculate that laminina2 may be bound to additional receptors such as integrin a7b1d in the muscle fibre membrane. Interestingly, the immunohistochemistry results were not matched by the marked reduction in the intensity of the laminin-a2 band on immunoblots seen in both WWS patients. Although the significance of this observation is unclear it is worth noting that a similar discrepancy has been described in the muscle of LGMD2I patients.27 Except for laminin-a2, we found no major alterations in the expression pattern of the other extracellular matrix proteins examined including laminin chains b1 and g1, and perlecan and collagen VI. In future studies it would be of interest to assess the expression in WWS muscle of agrin and biglycan, which also bind to a-dystroglycan. There is increasing evidence in the literature that abnormal glycosylation of a-dystroglycan, in vitro and in vivo, disrupts its binding to laminin and other ligands14,28 and thus the link between the extracellular matrix and the intracellular cytoskeleton. This has been recently proposed in MEB and FCMD muscle and in the muscle and brain tissue from myd mouse.12 It remains to be seen if any laminin binding activity remains in the muscle of WWS patients with such a severe depletion of adystroglycan as the case described here. a-Dystroglycan is also expressed as a heavily glycosylated protein of 120 kDa in brain cortex and cerebellum.29 The brain abnormalities seen in WWS affect these two regions (cortical dysgenesis and cerebellar malformations) and are a consequence of a cell migration defect. Interestingly, a conditional gene knock-out of a-dystroglycan in brain develops a neuronal migration disorder resembling the pachygyria seen in FCMD.30Thus, the abnormal post-translational modification of a-dystroglycan may be central to both the muscle degeneration and to the cell migration defect characteristic of WWS. In addition to the abnormalities in a-dystroglycan and laminin-a2 we found that both WWS cases had also in common a mild reduction of dystrophin labelling and an abnormally large proportion of fibres positive for fetal/neonatal myosin. There are indeed a few reports in the literature of reduced/variable dystrophin labelling in WWS muscle17,20 although both observations could reflect muscle immaturity.
Original article: C Jime´nez-Mallebrera et al. The present study shows that the combined assessment of the expression of a-dystroglycan, laminin-a2 and other extracellular matrix and membrane associated proteins by immunological techniques may be useful to the differential diagnosis of WWS and provides further data towards the current endeavour to elucidate the role of adystroglycan in muscle disease.
Acknowledgements We are very grateful to Dr J Geddes, Dr M Hird and Dr P Mannix from the Royal London Hospital for referring these patients/biopsies to our service. We wish to thank the Muscular Dystrophy Campaign of Great Britain and Northern Ireland, the European Community (Myo-Cluster: GENRE grant QLG1 CT 1999 00870) and the NSCAG grants for the financial support to this work.
References 1 Dobyns WB, Pagon RA, Armstrong D, et al. Diagnostic criteria for Walker-Warburg syndrome. Am J Med Genet 1989;32:195 – 210. 2 Cormand B, Avela K, Pihko H, et al. Assignment of the muscle-eye-brain disease gene to 1p32-p34 by linkage analysis and homozygosity mapping. Am J Hum Genet 1999;64:126– 35. 3 Cormand B, Pihko H, Bayes M, et al. Clinical and genetic distinction between Walker-Warburg syndrome and muscle-eye-brain disease. Neurology 2001;56:1059 – 69. 4 Toda T, Segawa M, Nomura Y, et al. Localization of a gene for Fukuyama type congenital muscular dystrophy to chromosome 9q31-33. Nature Genet 1993;5:283 –6. 5 Yoshida A, Kobayashi K, Manya H, et al. Muscular dystrophy and neuronal migration disorder caused by mutations in a glycosyltransferase, POMGnT1. Dev Cell 2001;1:717– 24. 6 Aravind L, Koonin EV. The fukutin protein family— predicted enzymes modifying cell-surface molecules. Curr Biol 1999;9:R836 – 7. 7 Beltra´n-Valero De Bernabe´ D, Currier S, Steinbrecher A, et al. Mutations in the O-mannosyltransferase gene POMT1 give rise to the severe neuronal migration disorder Walker-Warburg syndrome. Am J Hum Genet 2002;71. 8 Tome´ FMS, Evangelista T, Leclerc A, et al. Congenital muscular dystrophy with merosin deficiency. C R Acad Sci III 1994;317:351 – 7. 9 Hayashi YK, Chou FL, Engvall E, et al. Mutations in the integrin alpha-7 gene cause congenital myopathy. Nature Genet 1998;19:94 – 7. 10 Camacho Vanegas O, Bertini E, Zhang RZ, et al. Ullrich scleroatonic muscular dystrophy is caused by
Original article: a-Dystroglycan abnormalities in Walker-Warburg muscle
11
12
13
14
15
16
17
18
19
recessive mutations in collagen type VI. Proc Natl Acad Sci USA 2001;98:7516 – 21. Hayashi YK, Engvall E, Arikawa-Hirasawa E, et al. Abnormal localization of laminin subunits in muscular dystrophies. J Neurol Sci 1993;119:53 –64. Haltia M, Leivo I, Somer H, et al. Muscle-Eye-Brain disease, a neuropathological study. Ann Neurol 1997;41:173– 80. Hayashi YK, Ogawa M, Tagawa K, et al. Selective deficiency of alpha-dystroglycan in Fukuyama-type congenital muscular dystrophy. Neurology 2001;57: 115 –21. Michele DE, Barresi R, Kanagawa M, et al. Posttranslational disruption of dystroglycan ligand interactions in congenital muscular dystrophies. Nature 2002;418:417– 21. Brockington M, Blake DJ, Prandini P, et al. Mutations in the fukutin-related protein gene (FKRP) cause a form of congenital muscular dystrophy with secondary laminin alpha2 deficiency and abnormal glycosylation of alpha-dystroglycan. Am J Hum Genet 2001;69:1198– 209. Brockington M, Sewry CA, Herrmann R, et al. Assignment of a form of congenital muscular dystrophy with secondary merosin deficiency to chromosome 1q42. Am J Hum Genet 2000;66:428 – 35. Voit T, Sewry CA, Meyer K, et al. Preserved merosin M-chain (or laminin-alpha 2) expression in skeletal muscle distinguishes Walker-Warburg syndrome from Fukuyama muscular dystrophy and merosindeficient congenital muscular dystrophy. Neuropediatrics 1995;26:148 – 55. Kanoff RJ, Curless RG, Petito C, et al. Walker-Warburg syndrome: neurologic features and muscle membrane structure. Pediatr Neurol 1998;18:76 –80. Wewer UM, Durkin ME, Zhang X, et al. Laminin beta 2 chain and adhalin deficiency in the skeletal muscle of Walker-Warburg syndrome (cerebro-ocular dysplasia-muscular dystrophy). Neurology 1995;45:2099– 101.
20
21
22
23
24
25
26
27
28 29
30
137
Vajsar J, Ackerley C, Chitayat D, Becker LE. Basal lamina abnormality in the skeletal muscle of Walker-Warburg syndrome. Pediatr Neurol 2000;22: 139– 43. Ishii H, Hayashi YK, Nonaka I, Arahata K. Electron microscopic examination of basal lamina in Fukuyama congenital muscular dystrophy. Neuromuscul Disord 1997;7:191 –7. Minetti C, Bado M, Morreale G, et al. Disruption of muscle basal lamina in congenital muscular dystrophy with merosin deficiency. Neurology 1996;46: 1354– 8. Osari S, Kobayashi O, Yamashita Y, et al. Basement membrane abnormality in merosin-negative congenital muscular dystrophy. Acta Neuropathol 1996;91: 332– 6. Herrmann R, Straub V, Blank M, et al. Dissociation of the dystroglycan complex in caveolin-3-deficient limb girdle muscular dystrophy. Hum Mol Genet 2000;9:2335– 40. Ervasti JM, Campbell KP. A role for the dystrophinglycoprotein complex as a transmembrane linker between laminin and actin. J Cell Biol 1993;122: 809– 23. Yang B, Jung D, Motto D, et al. SH3 domainmediated interaction of a-dystroglycan and Grb2. JBC 1995;270:11711 – 4. Bushby K, Anderson LV, Pollitt C, et al. Abnormal merosin in adults. A new form of late onset muscular dystrophy not linked to 6q2. Brain 1998;121: 581– 8. Sugita S. A stochiometric complex of neurexins and dystroglycan in brain. J Cell Biol 2001;154:435 –45. Zaccaria ML, Di Tommaso F, Brancaccio A, et al. Dystroglycan distribution in adult mouse brain: a light and electron microscopy study. Neuroscience 2001;104:311 – 24. Moore SA, Saito F, Chen J, et al. Deletion of brain dystroglycan recapitulates aspects of congenital muscular dystrophy. Nature 2002;418:422 –5.
ORIGINAL ARTICLE
Profound skeletal muscle depletion of a-dystroglycan in Walker-Warburg syndrome CECILIA JIME´NEZ-MALLEBRERA,1 SILVIA TORELLI,1 SUSAN C BROWN,1 LUCY FENG,1 MARTIN BROCKINGTON,1 CAROLINE A SEWRY,1 DANIEL BELTRA´N-VALERO DE BERNABE´,2 FRANCESCO MUNTONI1 1
Neuromuscular Unit, Imperial College School of Medicine, Hammersmith Hospital, London, UK; 2Department of Human Genetics, University Medical Centre Nijmegen, The Netherlands
Walker-Warburg syndrome (WWS) is an autosomal recessive disorder characterized by the combined involvement of the central nervous and skeletal muscle systems. Although the molecular basis of WWS remains unknown, defects in the muscle fibre basal lamina are characteristic of other forms of congenital muscular dystrophy (CMD). In agreement with this, some forms of CMD, due to glycosyltransferase defects, display a reduction in the immunolabelling of a-dystroglycan, whilst b-dystroglycan labelling appears normal. Here we describe an almost complete absence of a-dystroglycan using both immunohistochemistry and immunoblotting in two patients with WWS. In addition, there was a mild reduction of laminin-a2. In contrast, immunohistochemical labelling of perlecan and collagen VI was normal. Linkage analysis excluded the recently identified POMT1 locus, responsible for a proportion of WWS cases. These results confirm that WWS is a genetically heterogeneous condition and suggest that disruption of the a-dystroglycan/laminin-a2 axis in the basal lamina may play a role in the degeneration of muscle fibres in WWS—also in cases not due to POMT1 defects. Keywords: Walker-Warburg syndrome. a-Dystroglycan. Protein glycosylation. Extracellular matrix. Immunohistochemistry. Immunoblotting.
Introduction Walker-Warburg syndrome [MIM 236670] is a lethal autosomal recessive disorder characterized by the combined involvement of the central nervous and skeletal muscle systems. Brain malformations with congenital muscular dystrophy are also found in other forms of congenital muscular dystrophy such as muscle-eye-brain disease (MEB, [MIM 253280]) and Fukuyama congenital muscular dystrophy (FCMD, [MIM 253800]). In addition, severe ocular abnormalities involving the retina and the anterior chamber are present in WWS and MEB but not in FCMD. The diagnosis of WWS is based on clinical (congenital hypotonia and weakness, retinal malformation), radiological (type II lissencephaly,
cerebellar malformation) and pathological observations (congenital muscular dystrophy).1 Since the clinical features of WWS and MEB overlap significantly (to a higher extent than to FCMD) it was suggested that MEB and WWS may be allelic disorders. However, after the MEB locus was mapped to 1p32-p34,2 linkage analysis showed that both disorders are genetically distinct.3 Similarly, FCMD can be genetically excluded by linkage analysis to 9q31-q33.4 The genes defective in FCMD ( fukutin gene) and in MEB (POMGnT1 gene) have been identified,4,5 and encode proteins with a potential role in protein modification/ glycosylation.5,6 The genetic and molecular basis of WWS are just starting to be elucidated. Very recently, mutations in the POMT1 gene have been reported in a proportion of WWS patients.7 This gene encodes a glycosyltransferase which
Received 18.12.02. Accepted 28.2.03. Correspondence: Cecilia Jime´nez-Mallebrera, Neuromuscular Unit, Hammersmith Hospital, Du Cane Road, London W12 0NN, UK. Tel: þ 44-20-8383-2126; Fax: þ 44-20-8746-2187; e-mail: [email protected]
1090-3798/03/07/0129+9 $35.00
Q 2003 European Paediatric Neurology Society
130 catalyses the first step in O-mannosyl glycan synthesis. However, genome-wide linkage analysis suggests that WWS is genetically heterogeneous and indeed mutations in the POMT1 were only found in 20% of families studied.7 The expression of extracellular matrix proteins is abnormal (reduced or absent) in several forms of congenital muscular dystrophy. This includes laminin-a2 chain of laminin-2 in MDC1A8 [MIM 156225]; integrin a7 in Integrin a7 deficiency9 [MIM 600536]; collagen VI in Ullrich CMD10 (UCMD, [MIM 254090]). In addition to these primary defects, secondary changes can be seen. For example, a secondary reduction in the expression level of laminin-a2 chain has been described in both FCMD11 and MEB.12 Furthermore, in an increasing number of CMD forms with and without brain involvement, which include FCMD, MEB, MDC1B and MDC1C, the expression of a-dystroglycan, but not b-dystroglycan, is reduced to a variable extent.13 – 16 Regarding laminin-a2 chain expression in WWS, some authors have shown preserved expression17 while others found a decreased expression.18 There is also one report of reduced laminin-b2 expression in WWS muscle.19 Using electron microscopy, previous authors have noted basal lamina abnormalities (described as thinning and disruption in non-necrotic fibres) in the skeletal muscle fibres of WWS patients.20 Interestingly, disruption of the BM has also been described in FCMD21 and in primary laminin-a2 deficient congenital muscular dystrophy.22,23 To gain insight into the pathogenesis of WWS we have examined the expression of a number of sarcolemmal, trans-sarcolemmal and extracellular matrix proteins in the skeletal muscle of two patients with classical WWS features. Genetic analysis in one of the two cases allowed us to rule out an involvement of the recently identified POMT1 gene, mutated in a proportion of cases of WWS. We found an almost complete absence of a-dystroglycan and a mild reduction of laminin-a2 using both immunohistochemistry and immunoblotting. In contrast, expression of perlecan and collagen VI was essentially normal. These results suggest that impaired processing of a-dystroglycan may be at the centre of the dystrophic process in WWS.
Materials and methods Immunocytochemistry Frozen 8 mm sections were incubated with monoclonal antibodies to spectrin and dystrophin
Original article: C Jime´nez-Mallebrera et al. (NCL-SPEC1, NCL-DYS1, NCL-DYS2 and NCLDYS3, Novocastra laboratories Ltd), the neonatal form of myosin heavy chain (NCL-MHCn, Novocastra laboratories Ltd), laminin-a2 (MAB1922 Chemicon and 4H8, Alexis corporation), collagen VI (MAB1944 and MAB3303, Chemicon), perlecan (MAB1948, Chemicon), b-dystroglycan (NCL-bDG, Novocastra Laboratories Ltd), a-dystroglycan (VIA4-1, Upstate Biotechnology) and sheep polyclonal raised against a 20 aa peptide from the chick a-dystroglycan sequence that recognizes the core protein of a-dystroglycan24 (kind gift of Dr Stephan Kro¨ger). VIA4-1 is a monoclonal antibody to an as yet unidentified epitope (possibly a carbohydrate). Prior to staining with VIA4-1, sections were fixed in 50% acetic acid/ethanol for 6 minutes at room temperature and non-specific binding of this antibody was blocked treating sections with 3% BSA for 30 minutes at room temperature. All primary antibodies were applied for 1 hour and revealed with an appropriate biotinylated secondary antibody (Amersham 1:200) for 30 minutes, followed by streptavidin conjugated to Alexa 594 (Molecular Probes) for 15 minutes. All dilutions and washings were made in phosphate buffered saline. Sections were mounted in aqueous mountant and viewed with epifluorescence using a Leica DMR microscope linked to Metamorph (Universal Imaging). Control sections were labelled without primary antibodies, and all sections were compared with age-matched control samples from other neuromuscular disorders and with normal muscle.
Western blotting Muscle proteins were extracted in sample buffer consisting of 1 M Tris HCl, 1% SDS, glycerol, 2-mercaptoethanol plus a cocktail of protease inhibitors (antipain, aprotinin and leupeptin) and quantified using the BCA assay (Pierce). Soluble protein (50 mg) was resolved using a NuPage Precast gel (4 –12% Bis-Tris) (Invitrogen) and then electrophoretically transferred to nitrocellulose membrane (Amersham). Membranes were blocked in 3% BSA to detect a-dystroglycan and 5% milk powder to detect laminin-a2, both in TBST Buffer. Membranes were probed with antibodies to adystroglycan (VIA4-1, 1:1000), b-dystroglycan (Novocastra 1:25), laminin-a2 (MAB1922 Chemicon 1:1000), washed, and incubated with HRP-anti mouse (Jackson Laboratories). Polypeptides were visualized using chemiluminescence (ECL þ Plus, Amersham).
Original article: a-Dystroglycan abnormalities in Walker-Warburg muscle
Haplotype analysis DNA for genetic analysis was available from patient 1 and his family but not from family 2. DNA was extracted from blood lymphocytes collected from patient 1, parents and the healthy sibling. The MEB, FCMD and FKRP loci, that have been associated with CMD and brain involvement, or with abnormal processing of a-dystroglycan, were studied using linkage analysis in this consanguineous family. The markers used were: for the MDC1C locus the dinucleotide repeat markers D19S219 and D19S606;15 for the FCMD locus we used microsatellite markers D9S306, D9S2105, D9S2107, D9S2170, D9S2171 and D9S172, which map to chromosome 9q31. It is thought that the Fukutin gene is present in very close proximity to centromeric D9S2170. The order of these markers is as follows: cen-D9S306-D9S2105-FCMD-D9S2170D9S2171-D9S2107-D9S172-tel; the distance between D9S2105 and D9S2107 is approximately 230 kb. Regarding the MEB locus, this was studied with the markers D1S211, D1S2677, D1S427, D1S2652, and D1S200.2 In addition, patient 1 was also studied for the flanking markers to the POMT1 locus, D9S260 and D9S1793, which lie 3.1 cm centromeric and 3.9 telomeric to the locus, respectively.
Results Patients Patient 1 Patient 1 was a male infant who was referred to us at the age of 2 months and 1 week with severe neonatal hypotonia, developmental delay and poor visual behaviour. He was the third child of a Bengali consanguineous family; his older siblings were healthy. Congenital hydrocephalus had been noticed in the past that required shunt at the age of 2 weeks. His serum creatine kinase (CK) levels were markedly elevated in the first few weeks of life and an opthalmological examination revealed bilateral lamellar cataracts. His feeding had been very poor and he had been nasogastrically fed since birth. On examination at the age of 5 weeks, he was extremely floppy with a ‘frog-like’ posture, showing no antigravity proximal movements in the lower and upper limbs, but only distal hands and feet movements. He had dysmorphic features with macrocephaly despite the shunting, prominent eyes (bupthalmos) secondary to anterior
131
chamber abnormalities, and poorly formed muscle bulk. He had no head control on traction or ventral suspension and no apparent response to visuoauditory stimuli. His reflexes were absent. A brain MRI at the age of 5 weeks showed a severe hydrocephalus and the thin residual cortical mantle was smooth in appearance. The cerebellar hemispheres were also severely hypoplastic. A muscle biopsy was obtained from his quadriceps at the age of 2 weeks. The patient died at the age of 8 months during an episode of upper respiratory tract infection; he had shown no psychomotor development until then. A few years later the family was referred back because mother was pregnant. The pregnancy was carefully followed with serial ultrasound scans from the age of 12 weeks. At the age of 18 weeks gestation a severe hydrocephalus was noticed in the fetus, which was therefore considered to be similarly affected. Pregnancy was terminated a few weeks afterwards. The muscle of this fetus was not stored for immunocytochemical analysis and is not included in this study.
Patient 2 Patient 2 is the first child of a consanguineous family and presented with similar features to patient 1 which can be summarized as prenatal onset with ventriculomegaly detected with an ultrasound at 23 weeks, which was confirmed by the later findings (at 30 weeks) of ventriculomegaly, absent corpus callosum and enlarged third and fourth ventricle detected. At birth the baby had poor spontaneous respiratory efforts requiring elective intubation. Examination revealed extended legs because of breech presentation, contractures at the knees and poor spontaneous movements. Hydrocephalus with a posterior fossa cyst was confirmed postnatally and the cerebellum appeared hypoplastic. Ophthalmological examination revealed a bilateral micocornea, posterior lens opacities, vitreal dysplasia and bilateral retinal dysplasia. She was ventilated for a week but then extubated in air. She had a ventriculoperitoneal shunt inserted aged 3 months. She was discharged home at that stage and is stable, although her feeding difficulties persist. A muscle biopsy of the quadriceps was performed at the age of 6.5 weeks and analysed by standard histological and histochemical techniques. This biopsy was referred to us for specialized analysis when the patient was 3 months old.
132
Haplotype analysis The lack of homozygosity at the corresponding loci allowed us to exclude the involvement of the MEB, FCMD and FKRP loci. In addition, patient 1 was also shown to be heterozygous for the flanking markers to the POMT1 locus, i.e. D9S260 and D9S1793.7 These results are in accordance with recently published data that suggest that WWS is a genetically heterogeneous condition. We hypothesize that patient 1 suffers from a subtype of WWS not caused by mutations in the POMT1 gene.
Histopathology and immunocytochemistry The muscle biopsy of patient 1 showed an abnormal variation in fibre size with several small fibres and some larger ones and features of immature muscle. Fibre type differentiation as indicated by oxidative enzyme staining was indistinct. Most fibres co-expressed the neonatal and fast isoforms
Original article: C Jime´nez-Mallebrera et al. of myosin heavy chain (Fig. 2). There was no apparent increase in connective tissue, no excess glycogen and no apparent accumulation of fat. Patient 2 showed a wider variation in fibre size than patient 1 and in contrast to patient 1, an excess of connective tissue. A large proportion of fibres expressed neonatal myosin (Fig. 2). Both patients showed a similar pattern of immunolabelling with all the antibodies used although some differences were noted. Labelling of the sarcolemmal proteins spectrin and dystrophin was within normal limits although some variation in the labelling intensity of individual fibres was seen with the antibody directed against the COOH-terminus of dystrophin (Fig. 1). This variation was more marked in the muscle from patient 2 which showed some very weak/ almost negative fibres (Fig. 1). The significance of this observation is unclear but it could relate to immaturity. In both cases, a-dystroglycan staining using two different antibodies showed a profound reduction in the level of immunolabelling on all fibres
Fig. 1. Muscle sections from patients 1 and 2 and from a neonate patient with a neurometabolic condition used as a control. Sections were immunolabelled with antibodies to spectrin and dystrophin (NCL-DY2 carboxy-terminus and NCL-DYS3 amino-terminal portion of rod domain). Scale bar (bottom left) ¼ 200 mm.
Original article: a-Dystroglycan abnormalities in Walker-Warburg muscle (Fig. 2). Labelling of the transmembrane protein b-dystroglycan, however, was well preserved although with some variability (Fig. 2, some bright large fibres and some small weak fibres). Expression of the laminin-a2 chain of laminin-2 isoform was assessed using an antibody against the 300 kDa fragment and an antibody against its COOH-terminal 80 kDa fragment. In both cases the labelling was mildly reduced with some isolated large fibres (Fig. 3) that were not labelled with either of the two antibodies or were only labelled very weakly. These are the same fibres that labelled very intensely with the b-dystroglycan antibody. b-Spectrin appeared normal in these large fibres indicating that this was not due to non-specific damage of the plasma membrane. This is also unlikely to be due to regeneration since these fibres were not labelled with the
133
neonatal myosin antibody. The reduction of laminin-a2 immunolabelling was more significant in patient 1 and in both cases it was more apparent with the antibody against the NH2-terminal 300 kDa fragment (Fig. 3). In the muscle of patient 1 labelling of the laminin-b1 and -g1 chains was comparable with the control muscle while in patient 2 some fibres showed weak laminin-b1 labelling compared with the control muscle and to laminin-g1 chain labelling (data not shown). In addition to laminin-a2, a-dystroglycan also binds to perlecan (proteoglycan heparan sulphate) agrin and collagen VI. However, the immunolabelling of perlecan and collagen VI was not significantly different to controls in either patient despite the almost complete absence of a-dystroglycan (Fig. 3).
Fig. 2. a-Dystroglycan is absent in the muscle of patients 1 and 2 using both antibodies (VIA4-1 and core) whereas bdystroglycan appears normal. The proportion of neonatal myosin positive fibres in patients 1 and 2 seems larger than in the control muscle. Scale bar (bottom left) ¼ 200 mm.
134
Original article: C Jime´nez-Mallebrera et al.
Fig. 3. In WWS muscle some fibres are weakly labelled with two anti-laminin-a2 antibodies. In contrast, collagen VI and perlecan appear essentially normal (note that collagen VI is slightly reduced in patient 2 but not in patient 1). Scale bar (bottom left) ¼ 200 mm.
Western blotting In order to investigate further the reduction of a-dystroglycan an SDS extract from the patient’s muscle was analysed by Western blotting using the VIA4-1 antibody (Fig. 4A). This antibody recognized a broad band of 150 – 200 kDa, which corresponds to the highly glycosylated form of a-dystroglycan, in normal adult and neonatal muscle (and in an extract prepared from C2C12 cell culture and used as a positive control). No reactive band was detected in the either patient 1 or patient 2 extracts. For further comparisons, the WWS patients’ extract was run alongside an extract from a Duchenne muscular dystrophy (DMD) patient and an aged-matched patient with an undiagnosed form of laminin-a2 positive CMD. The intensity and molecular weight of the a-dystroglycan polypep-
tide in DMD and CMD muscle were comparable with the control sample which confirms that the reduction of a-dystroglycan in WWS muscle is not a non-specific consequence of the dystrophic process. A reduction of the 80 kDa fragment of laminin-a2 in WWS muscle was also revealed by Western blotting (Fig. 4B). The antibody that recognizes the 300 kDa laminin-a2 band is not suitable for Western blotting. The laminin-a2 band was almost undetectable in patient 2 and was significantly reduced in patient 1 when compared with the intensities of the bands seen in control fetal muscle (16 weeks gestation), control neonate muscle and in a muscle extract from an age-matched patient with an unclassified form of laminin-a2 positive CMD (Fig. 4B). To ensure that these findings were not due to protein degradation or low protein content,
Original article: a-Dystroglycan abnormalities in Walker-Warburg muscle
135
Fig. 4. Muscle proteins analysed by Western blotting using antibodies to A, a-dystroglycan, B, laminin-a2 and C, bdystroglycan. a-Dystroglycan (.150 kDa) is undetectable in patient 1 (P1) and patient 2 (P2). The 80 kDa fragment of laminin-a2 is almost absent in patient 2 and reduced in patient 1. b-Dystroglycan levels in WWS extracts are comparable with those in the control extracts. A.ctrol ¼ adult control muscle; N.ctrol ¼ neonate control muscle; CMD ¼ laminin-a2 positive congenital muscular dystrophy muscle; DMD ¼ Duchenne muscular dystrophy muscle. 16w ¼ 16 weeks control fetal muscle; C2C12 ¼ human skeletal muscle cell line C2C12.
the patient’s muscle extract was also probed with an antibody against b-dystroglycan. The intensities of the corresponding 43 kDa b-dystroglycan polypeptide bands were within normal limits (Fig. 4C).
Discussion The dystrophin associated complex has been attributed with a role in stabilizing the sarcolemma25 although there is now increasing evidence that this protein assembly of which dystroglycan is a central part, also plays a role in cell signalling.26 In this paper we have identified a severe depletion of a-dystroglycan, but not b-dystroglycan, in two patients with the characteristic clinical features of Walker-Warburg syndrome. In one of these patients all the molecularly defined CMD variants with similar clinicopathological features were excluded at the molecular level, including WWS caused by mutations in the POMT1 gene.7 The almost complete absence of immunolabelling using either the antibody to the core protein or to
the carbohydrate epitope recognized by VIA4-1 differs from what has been described in other forms of CMD. In these patients there is either a secondary reduction in a-dystroglycan in which the expression of the core protein is well preserved (as seen in MEB14 and in patients with mutations in the POMT1 gene7) or there is some residual immunolabelling with all the antibodies used (as seen in MDC1C15). Furthermore, on immunoblotting, VIA4-1 did not detect a hypoglycosylated species of a-dystroglycan of reduced molecular weight as reported for example in MDC1C muscle.15 Taken together, these observations indicate a severe depletion of a-dystroglycan and not necessarily only of the glycosylated form. The deficient core staining we observed using an antibody that was raised against core a-dystroglycan might indicate that the stability of the entire molecule is affected in this patient. However this should be confirmed with antibodies that recognize a different epitope within the core protein or that work on Western blot. Unfortunately these additional antibodies are currently not freely available and therefore our interpretation remains speculative.
136 a-Dystroglycan is the main receptor for laminina2 in skeletal muscle.25 However, in the present study, we found almost normal or partly reduced immunolabelling of laminin-a2 in the basement membrane. Given the complete absence of a-dystroglycan it is possible to speculate that laminina2 may be bound to additional receptors such as integrin a7b1d in the muscle fibre membrane. Interestingly, the immunohistochemistry results were not matched by the marked reduction in the intensity of the laminin-a2 band on immunoblots seen in both WWS patients. Although the significance of this observation is unclear it is worth noting that a similar discrepancy has been described in the muscle of LGMD2I patients.27 Except for laminin-a2, we found no major alterations in the expression pattern of the other extracellular matrix proteins examined including laminin chains b1 and g1, and perlecan and collagen VI. In future studies it would be of interest to assess the expression in WWS muscle of agrin and biglycan, which also bind to a-dystroglycan. There is increasing evidence in the literature that abnormal glycosylation of a-dystroglycan, in vitro and in vivo, disrupts its binding to laminin and other ligands14,28 and thus the link between the extracellular matrix and the intracellular cytoskeleton. This has been recently proposed in MEB and FCMD muscle and in the muscle and brain tissue from myd mouse.12 It remains to be seen if any laminin binding activity remains in the muscle of WWS patients with such a severe depletion of adystroglycan as the case described here. a-Dystroglycan is also expressed as a heavily glycosylated protein of 120 kDa in brain cortex and cerebellum.29 The brain abnormalities seen in WWS affect these two regions (cortical dysgenesis and cerebellar malformations) and are a consequence of a cell migration defect. Interestingly, a conditional gene knock-out of a-dystroglycan in brain develops a neuronal migration disorder resembling the pachygyria seen in FCMD.30Thus, the abnormal post-translational modification of a-dystroglycan may be central to both the muscle degeneration and to the cell migration defect characteristic of WWS. In addition to the abnormalities in a-dystroglycan and laminin-a2 we found that both WWS cases had also in common a mild reduction of dystrophin labelling and an abnormally large proportion of fibres positive for fetal/neonatal myosin. There are indeed a few reports in the literature of reduced/variable dystrophin labelling in WWS muscle17,20 although both observations could reflect muscle immaturity.
Original article: C Jime´nez-Mallebrera et al. The present study shows that the combined assessment of the expression of a-dystroglycan, laminin-a2 and other extracellular matrix and membrane associated proteins by immunological techniques may be useful to the differential diagnosis of WWS and provides further data towards the current endeavour to elucidate the role of adystroglycan in muscle disease.
Acknowledgements We are very grateful to Dr J Geddes, Dr M Hird and Dr P Mannix from the Royal London Hospital for referring these patients/biopsies to our service. We wish to thank the Muscular Dystrophy Campaign of Great Britain and Northern Ireland, the European Community (Myo-Cluster: GENRE grant QLG1 CT 1999 00870) and the NSCAG grants for the financial support to this work.
References 1 Dobyns WB, Pagon RA, Armstrong D, et al. Diagnostic criteria for Walker-Warburg syndrome. Am J Med Genet 1989;32:195 – 210. 2 Cormand B, Avela K, Pihko H, et al. Assignment of the muscle-eye-brain disease gene to 1p32-p34 by linkage analysis and homozygosity mapping. Am J Hum Genet 1999;64:126– 35. 3 Cormand B, Pihko H, Bayes M, et al. Clinical and genetic distinction between Walker-Warburg syndrome and muscle-eye-brain disease. Neurology 2001;56:1059 – 69. 4 Toda T, Segawa M, Nomura Y, et al. Localization of a gene for Fukuyama type congenital muscular dystrophy to chromosome 9q31-33. Nature Genet 1993;5:283 –6. 5 Yoshida A, Kobayashi K, Manya H, et al. Muscular dystrophy and neuronal migration disorder caused by mutations in a glycosyltransferase, POMGnT1. Dev Cell 2001;1:717– 24. 6 Aravind L, Koonin EV. The fukutin protein family— predicted enzymes modifying cell-surface molecules. Curr Biol 1999;9:R836 – 7. 7 Beltra´n-Valero De Bernabe´ D, Currier S, Steinbrecher A, et al. Mutations in the O-mannosyltransferase gene POMT1 give rise to the severe neuronal migration disorder Walker-Warburg syndrome. Am J Hum Genet 2002;71. 8 Tome´ FMS, Evangelista T, Leclerc A, et al. Congenital muscular dystrophy with merosin deficiency. C R Acad Sci III 1994;317:351 – 7. 9 Hayashi YK, Chou FL, Engvall E, et al. Mutations in the integrin alpha-7 gene cause congenital myopathy. Nature Genet 1998;19:94 – 7. 10 Camacho Vanegas O, Bertini E, Zhang RZ, et al. Ullrich scleroatonic muscular dystrophy is caused by
Original article: a-Dystroglycan abnormalities in Walker-Warburg muscle
11
12
13
14
15
16
17
18
19
recessive mutations in collagen type VI. Proc Natl Acad Sci USA 2001;98:7516 – 21. Hayashi YK, Engvall E, Arikawa-Hirasawa E, et al. Abnormal localization of laminin subunits in muscular dystrophies. J Neurol Sci 1993;119:53 –64. Haltia M, Leivo I, Somer H, et al. Muscle-Eye-Brain disease, a neuropathological study. Ann Neurol 1997;41:173– 80. Hayashi YK, Ogawa M, Tagawa K, et al. Selective deficiency of alpha-dystroglycan in Fukuyama-type congenital muscular dystrophy. Neurology 2001;57: 115 –21. Michele DE, Barresi R, Kanagawa M, et al. Posttranslational disruption of dystroglycan ligand interactions in congenital muscular dystrophies. Nature 2002;418:417– 21. Brockington M, Blake DJ, Prandini P, et al. Mutations in the fukutin-related protein gene (FKRP) cause a form of congenital muscular dystrophy with secondary laminin alpha2 deficiency and abnormal glycosylation of alpha-dystroglycan. Am J Hum Genet 2001;69:1198– 209. Brockington M, Sewry CA, Herrmann R, et al. Assignment of a form of congenital muscular dystrophy with secondary merosin deficiency to chromosome 1q42. Am J Hum Genet 2000;66:428 – 35. Voit T, Sewry CA, Meyer K, et al. Preserved merosin M-chain (or laminin-alpha 2) expression in skeletal muscle distinguishes Walker-Warburg syndrome from Fukuyama muscular dystrophy and merosindeficient congenital muscular dystrophy. Neuropediatrics 1995;26:148 – 55. Kanoff RJ, Curless RG, Petito C, et al. Walker-Warburg syndrome: neurologic features and muscle membrane structure. Pediatr Neurol 1998;18:76 –80. Wewer UM, Durkin ME, Zhang X, et al. Laminin beta 2 chain and adhalin deficiency in the skeletal muscle of Walker-Warburg syndrome (cerebro-ocular dysplasia-muscular dystrophy). Neurology 1995;45:2099– 101.
20
21
22
23
24
25
26
27
28 29
30
137
Vajsar J, Ackerley C, Chitayat D, Becker LE. Basal lamina abnormality in the skeletal muscle of Walker-Warburg syndrome. Pediatr Neurol 2000;22: 139– 43. Ishii H, Hayashi YK, Nonaka I, Arahata K. Electron microscopic examination of basal lamina in Fukuyama congenital muscular dystrophy. Neuromuscul Disord 1997;7:191 –7. Minetti C, Bado M, Morreale G, et al. Disruption of muscle basal lamina in congenital muscular dystrophy with merosin deficiency. Neurology 1996;46: 1354– 8. Osari S, Kobayashi O, Yamashita Y, et al. Basement membrane abnormality in merosin-negative congenital muscular dystrophy. Acta Neuropathol 1996;91: 332– 6. Herrmann R, Straub V, Blank M, et al. Dissociation of the dystroglycan complex in caveolin-3-deficient limb girdle muscular dystrophy. Hum Mol Genet 2000;9:2335– 40. Ervasti JM, Campbell KP. A role for the dystrophinglycoprotein complex as a transmembrane linker between laminin and actin. J Cell Biol 1993;122: 809– 23. Yang B, Jung D, Motto D, et al. SH3 domainmediated interaction of a-dystroglycan and Grb2. JBC 1995;270:11711 – 4. Bushby K, Anderson LV, Pollitt C, et al. Abnormal merosin in adults. A new form of late onset muscular dystrophy not linked to 6q2. Brain 1998;121: 581– 8. Sugita S. A stochiometric complex of neurexins and dystroglycan in brain. J Cell Biol 2001;154:435 –45. Zaccaria ML, Di Tommaso F, Brancaccio A, et al. Dystroglycan distribution in adult mouse brain: a light and electron microscopy study. Neuroscience 2001;104:311 – 24. Moore SA, Saito F, Chen J, et al. Deletion of brain dystroglycan recapitulates aspects of congenital muscular dystrophy. Nature 2002;418:422 –5.