Abstract
Background: Mutations in the fukutin-related protein gene FKRP cause limb-girdle muscular dystrophy (LGMD2I) as well as a form of congenital muscular dystrophy (MDC1C).
Objective: To define the phenotype in LGMD2I.
Methods: The authors assessed 16 patients from 14 families with FKRP gene mutations and LGMD and collected the results of mutation analysis, protein studies, and respiratory and cardiac investigations.
Results: Thirteen patients, most with adult presentation, were homozygous for the common C826A mutation in FKRP. The three other cases were compound heterozygotes for C826A and two of them presented in childhood, with more progressive disease. The pattern of muscle involvement, frequently including calf hypertrophy, was similar to dystrophinopathy. Complications in patients with LGMD2I were common and sometimes out of proportion to the skeletal muscle involvement. Six patients had cardiac involvement, and 10 had respiratory impairment: five required nocturnal respiratory support. All patients had serum creatine kinase at least 5 to 70 times normal. The most consistent protein abnormality found on muscle biopsy was a reduction of laminin α2 immunolabeling, either on muscle sections or immunoblotting alone.
Conclusions: LGMD2I due to FKRP mutations appears to be a relatively common cause of LGMD, with respiratory and cardiac failure as prominent complications.
Additional material related to this article can be found on the Neurology Web site. Go to www.neurology.org and scroll down the Table of Contents for the April 22 issue to find the title link for this article.
At least 16 different genetically defined subtypes of limb-girdle muscular dystrophy (LGMD) are recognized and distinguished by specialized diagnostic techniques.1-5⇓⇓⇓⇓ A recent study demonstrated that mutations in the fukutin-related protein gene FKRP (MIM *606596) cause a novel form of congenital muscular dystrophy (MDC1C).6,7⇓ The FKRP gene was mapped to chromosome 19q13.3 at the LGMD2I locus, and confirmation of mutations in families with disease linked to this locus demonstrated that FKRP mutations are also responsible for this form of LGMD, previously described only in patients from Tunisia (MIM #606612).6-8⇓⇓
The FKRP gene consists of four exons containing a 1,488–base pair (bp) open reading frame that encodes a 495–amino acid protein. FKRP is ubiquitously expressed in human tissues with highest levels in skeletal muscle and heart.6 Like fukutin (MIM*253800), FKRP contains conserved sequence elements suggesting that it is a glycosyltransferase. Providing evidence for this glycosyltransferase function, patients with MDC1C typically show abnormalities of processing of α-dystroglycan, in addition to a secondary reduction in laminin α2 (merosin).6 To date, muscle biopsy findings appear to be less uniform in the patients with LGMD. In the original description of the mapping of the LGMD2I locus, no protein abnormalities were noted. Another subset of patients with LGMD now known to have FKRP mutations have reduced laminin α2 chain immunolabeling on Western blotting only, with a normal appearance on immunolabeling of muscle biopsy samples.7,9⇓
The different subtypes of LGMD can now be distinguished at the gene and protein level from each other and from disorders (such as dystrophinopathy) that may clinically overlap with some of the different types of LGMD. The different types of LGMD may carry important implications for genetic counseling and management. This study reports the clinical details of 10 newly diagnosed patients with LGMD2I and mutations in FKRP as well as further follow-up and investigative data on the six patients we have previously reported.7,9⇓ These data provide evidence for a recognizable phenotype in LGMD2I that, although variable in severity, is distinct from other forms of LGMD. In many of these patients, striking clinical similarities between the LGMD2I phenotype and dystrophinopathy, including the presence of cardiomyopathy and respiratory failure, had led to problems in the differential diagnosis.
Patients and methods.
Clinical assessment.
Sixteen patients aged 11 to 58 years from 14 white families have been followed for between 1 and 11 years. Eleven patients are women and five are men. Patients 13 and 14 and Patients 11 and 12 are siblings. All individuals had had a muscle biopsy that confirmed the diagnosis of muscular dystrophy. Most patients had serum creatine kinase (CK) measurements available and some had had an EMG. Case histories were reviewed, and muscle power assessed according to the Medical Research Center scale for manual muscle testing. Respiratory function was assessed clinically and spirometrically. Forced vital capacity (FVC) and median forced expiratory volume in the first second (FEV1) was performed in sitting in 16 patients and in sitting and supine in 10 individuals. Seven patients had at least one overnight oxygen saturation monitoring or full polysomnography. A formal cardiology assessment including electrocardiogram (EKG) and echocardiogram was performed on 14 patients.
Muscle MR imaging.
Axial T1-weighted images (repetition time 949 msec, echo time 10 msec, slice thickness 5 mm) were obtained from the shoulder girdle to the feet of Patient 2 at the age of 15 years using a 1.5 Tesla MR scanner (Philips Interna, Best, Netherlands).
Muscle biopsy.
Muscle biopsies were performed and processed according to standard histologic and histochemical techniques. Immunocytochemical analysis of unfixed frozen tissue sections was performed by the indirect peroxidase method. Electrophoresis and multiplex Western blotting techniques were performed as previously described.10-12⇓⇓ Commercial antibodies to laminin α1 chain (monoclonal antibody [mAb] 1924), α2 chain or merosin (mAb 1922), β1 chain (mAb 1921), and γ1 chain (mAb 1920) were obtained from Chemicon International (Temecula, CA). An additional antibody to laminin α2 chain was used on sections: Mer3/22B2 (commercial source NCL-Merosin from Novocastra Laboratories, Newcastle upon Tyne, UK). The α-dystroglycan antibody used on four muscle biopsies was a mouse monoclonal immunoglobulin M, clone IIH6-C4 (from Upstate Biotechnology, Buckingham, UK). The monoclonal antibodies to dystrophin and the associated proteins used have been described previously.10-13⇓⇓⇓
Genetic analysis.
Genomic DNA was extracted from peripheral blood according to standard procedures. A PCR fragment encompassing the common C826A mutation was amplified by PCR using the primers FKRP-F (5′GCGACCTCTTCAACCTCTC3′) and FKRP-R (5′CCTTCTCCCATACGAAGC3′). The PCR was performed using 50 to 200 ng genomic DNA, 0.5 mM forward and reverse primer, 0.2 mM dNTP, 2M Betaine (Sigma, St. Louis, MO), and 0.5 units Taq polymerase (Promega, Madison, WI) per reaction. PCR conditions were 32 cycles of denaturation at 94 °C for 60 seconds, annealing at 58 °C for 60 seconds, and extension at 72 °C for 60 seconds. The 576-bp fragment obtained was then cleaved by the restriction endonuclease BfaI (New England Biolabs, Beverly, MA) using the manufacturer’s protocol, and electrophoresed through 2% agarose gels. In the presence of the C826A mutation the fragment remains uncut whereas wild-type DNA was cut into two fragments of 212 bp and 364 bp. Subsequent direct sequencing of the PCR-amplified DNA confirmed the digest results. The methods used for mutation analyses on Patients 5, 7, 8, 9, 11, and 12 were described previously.7
Results.
Clinical features.
Presentation and disease progression.
The range of age at onset was 2 to 40 years (mean 19.2 years). The main clinical features in the group are summarized in table 1 and supplementary table E-1 (available online at www.neurology.org). Five patients reported symptoms in childhood but only Patient 1 exhibited slightly delayed motor development. She was never able to jump or to run. Patients 2, 4, 5, and 13 reached normal motor milestones but toe walked or had a peculiar gait and could never run fast. Patients 4 and 5 had also experienced recurrent muscle cramps and myalgia in childhood. Patient 1 has been wheelchair dependent from the age of 12 years and has severe scoliosis. Patient 2 is also showing progressive weakness in his teens. Patients 4 and 5 exhibited progressive deterioration from their early 20s and Patient 5 was wheelchair dependent from the age of 26 years after an accident followed by a phase of immobilization. Patient 13 showed very little deterioration in his muscle function until middle adult life.
Table 1 Summary of clinical data
The remaining 11 patients had onset of symptoms in the second to fourth decades. The predominant mode of presentation was with a waddling gait, walking difficulties, and difficulties in climbing stairs. These patients described an indolent period of 3 to 35 years followed by progressive deterioration over 5 to 10 years. Patients with onset in their fourth decade exhibited the mildest phenotype with very slow progression. Overall, the mean age of the patients at their last review was 40.1 years and all but Patients 1, 5, and 15 were still ambulant for at least short distances. Patient 6 reported progressive dysphagia from the age of 35 although this mainly resolved after nocturnal ventilation was instituted for respiratory failure. At least three of the women have developed urinary urgency and occasional incontinence. Investigations in one individual showed no evidence of detrusor instability or hyper-reflexia and bladder capacity was normal. All individuals have normal intelligence and no cognitive decline and three had normal results on brain MRI.
Pattern of muscle involvement and distribution of weakness.
Distribution of weakness in the limbs was predominantly proximal and affected the hips more than the shoulder girdle. In most patients there was some minor asymmetry of muscle involvement—for example, grades varying from side to side by half to one MRC grade—but this was usually restricted to one or two muscle groups. Scapular winging was not prominent. There was no evidence for clinical involvement of the distal muscles except in the cases previously noted.9 Three patients had mild weakness of eye closure. Detailed assessment of Patient 13 (aged 55 years) is illustrative of the typical pattern of muscle involvement seen in the condition at the stage where the patient is still ambulant, with no facial weakness or scapular winging. Neck flexion was reduced to power grade 4. In the upper limbs, there was no distal weakness. Biceps were stronger than triceps (5 vs 4) and shoulder adduction was weaker than abduction (3 vs 4). In the lower limbs plantar flexion was reduced4 whereas dorsiflexion was grade 5 on the right and grade 4 on the left. Inversion and eversion were of normal power. Quadriceps were weaker than hamstrings (3 vs 5). Hip flexion, adduction, and abduction were weaker than hip extension (2 vs 4).
Contractures were not usually a marked feature. All individuals had lumbar lordosis without signs of spinal rigidity. Only Patient 1 exhibited severe scoliosis. In all patients hypertrophy of the calves was noted and occasionally other muscles as well (figure 1). Tongue hypertrophy was noted only in Patient 1.
Muscle MRI studies.
Axial T1-weighted images of Patient 2 at the age of 15 years showed prominent symmetric fatty infiltration in the lower limbs affecting the gluteal, vastus medius, intermedialis and lateralis, sartorius, rectus femoris, biceps femoris, semimembranosus, and semitendinosus muscles with relative sparing of the adductors. Distally, there was selective fatty replacement of the peroneus longus muscles, with milder involvement of other muscle groups except for tibialis anterior (figure 2, A and B⇓).
Figure 1. Asymmetric calf hypertrophy in Patient 7.
Figure 2. MRI scans from Patient 2. (A) T1-weighted axial image through the upper leg demonstrates prominent symmetric fatty infiltration of the vastus and biceps femoris muscles, with less marked changes in the sartorius, rectus femoris, semi-membranosus, and semitendinosus muscles. The adductor muscles appear to be spared. (B) T1-weighted axial images through the lower leg demonstrate prominent symmetric selective fatty infiltration of the peroneus longus muscle. The remaining muscles in the peroneal and posterior compartments show mild fatty infiltration, but with sparing of the tibialis anterior muscle.
Respiratory and cardiac involvement.
FVC corrected for height ranged from 19% of the predicted value to 103%. Serial measurements in seven patients over 5 years showed a gradual deterioration of FVC. In 10 patients a drop of FVC between 8 and 22% of the predicted value from sitting to supine suggested the presence of diaphragmatic involvement. As FVC declined, symptoms of nocturnal hypoventilation became apparent, together with breathlessness while lying flat. Respiratory failure necessitated the institution of noninvasive nighttime ventilation in four patients (5, 6, 8, and 15) at the ages of 32, 36, 39, and 58 years. They showed typical symptoms of nocturnal hypoventilation, which were relieved by treatment. Patient 13 is also awaiting nocturnal ventilation (see supplementary table E-1 [available online at www.neurology.org]). Two cases (Patients 5 and 15) had become wheelchair dependent before requiring ventilation (Patient 5 6 years before, following an accident, and Patient 15 less than 6 months before), whereas in the other patients respiratory failure supervened at a time when they were still just able to walk.
Evidence of cardiac involvement was documented in six patients, although not in either of the children in this series. Three patients (4, 8, and 15) at the ages of 28, 39, and 51 years developed symptomatic left heart failure while they were still independently ambulant. They have responded to a combination of diuretics and angiotensin converting enzyme inhibitor therapy, although one patient (15) has also required amiodarone for ventricular tachycardia. Echocardiography confirmed left ventricular impairment. The EKG results were also abnormal, showing sinus rhythm with frequent atrial ectopic beats, Q-waves, and repolarization abnormalities. Signal averaged EKG in one patient (8) showed late potentials, compatible with myocardial fibrosis.
Subclinical left ventricular impairment was observed on routine echocardiogram in three further patients (6, 14, and 16) at 38, 57, and 58 years with mild global hypokinesis of the left ventricle without chamber enlargement. Left ventricular ejection fraction was also reduced to 50% (see supplementary table E-1 [available online at www.neurology.org]). EKG abnormalities were detectable on standard 12-lead EKG and transthoracic EKG and included bifid, notched P-waves in sinus rhythm, compatible with left atrial hypertrophy and widespread repolarization changes. The PR interval, QRS duration, and QT intervals were normal.
Laboratory investigations.
All patients had elevated CK levels ranging between 750 and 10,000 IU/L (normal <150 IU/L). EMG was performed in nine patients (quadriceps [n = 4], deltoid [n = 2], tibialis anterior [n = 1], not defined [n = 4]). In all patients the findings were myopathic.
Muscle biopsy findings.
Light microscope studies.
Muscle biopsies (16 patients) showed histologic changes characteristic of a muscular dystrophy.
Immunolabeling.
Sufficient material for immunoanalysis remained from 14 biopsies and 13 showed abnormal protein expression. Eleven had a marked reduction of laminin α2 labeling, but this was detected on immunocytochemistry in only two cases, one of whom also had reduced laminin α2 chain on immunoblotting. In the other nine cases the reduction in laminin α2 chain was detectable on blots alone (figure 3A, Patient 2 [P2]). In two patients the abnormal laminin α2 chain expression was accompanied by significantly reduced labeling for calpain 3. Two brothers had a patchy reduction in laminin β1 chain labeling on sections with normal labeling for all other proteins examined (figure 3B, Patient 13 [P7]).
Figure 3. Immunolabeling of muscle biopsies in patients with limb-girdle muscular dystrophy (LGMD2I). (A) Three lanes from a Western blot labeled with a multiplex of monoclonal antibodies to dystrophin (400 kDa with metabolites down to about 150 kDa), calpain 3 (94 and 50 to 60 kDa), β-dystroglycan (43 kDa), and γ-sarcoglycan (35 kDa). Additional bands at ∼120 kDa are labeled with the 35DAG/21B5 antibody to γ-sarcoglycan, as described previously. The arrow indicates severely reduced labeling for the laminin α2 chain band in a patient with LGMD2I (Patient 2), compared to controls (C). Indirect peroxidase labeling. (B) Strong uniform labeling for laminin β1 chain at the periphery of all muscle fibers in a control biopsy section, compared to the overall reduction and patchy distribution in a patient with LGMD2I (Patient 13), is shown. Indirect peroxidase labeling with size bars for 100 μm.
Muscle was available for analysis of α-dystroglycan in four patients who all had a band of reduced abundance compared to control. Bands labeling with the α-dystroglycan antibody run from approximately 140 to 200 kDa (table 2). We were not able to comment on a size reduction in the patients with LGMD2I due to interference from myosin heavy chain at 200 kDa.
Table 2 Summary of immunohistochemical appearance and mutation analysis
Genetic analysis.
FKRP mutational analysis.
Sequence analysis of the FKRP coding region identified mutations in 16 individuals from 14 families (see table 2). Patients from 14 apparently unrelated and nonconsanguineous families were homozygous for the same common mutation C826A. Two novel frameshifting mutations were identified as the second mutations in Patients 1 and 2. The second mutation in Patient 8 has not yet been identified (see table 2) but haplotype analysis in his (albeit small) family is consistent with linkage to the LGMD2I locus (data not shown). A mutation such as a deletion on the second allele would not have been detected by the techniques used here.
Discussion.
This study confirms that LGMD due to FKRP mutations is an important cause of LGMD and this is our largest subgroup of patients with LGMD. The phenotype is consistent, and important because of the frequency of associated complications.6,7⇓
FKRP-related muscular dystrophy is highly variable. MDC1C patients are hypotonic from birth and never achieve independent ambulation.6 These patients typically have two missense mutations or one missense and one nonsense FKRP mutation, whereas the C826A mutation is not seen in this group.6,7 In the current study, 13 patients were homozygous for this mutation and of the three compound heterozygotes, two had progressive childhood disease and the third had developed symptomatic respiratory and cardiac failure in his late 30s. Among the patients sharing the same homozygous mutations, there was considerable variability of age at onset, progression, and cardiorespiratory involvement even within a family. The 11 patients with disease onset in the second to fourth decade had an initially indolent course with progressive deterioration after 5 to 35 years. The pattern of predominant pelvifemoral involvement was independent of severity. Distal muscles were usually clinically spared, although involvement of peroneus longus was seen on muscle MRI of one patient with preservation of tibialis anterior.
Our data emphasize the potential importance of cardiac complications. Thirty-eight percent of our patients had cardiac involvement. Symptomatic cardiomyopathy has not previously been described in LGMD2I, although Brockington et al. reported left ventricular dysfunction in three families with LGMD2I and three children with MDC1C.6,7⇓ Patients with muscular dystrophy due to FKRP mutations are therefore at risk of cardiac failure.
This disorder also primarily involves the respiratory musculature including the diaphragm. In every patient over the age of 20 years in whom this was measured, a drop in FVC from sitting to supine of 6 to 20% was demonstrable. Two of the children previously reported with MDC1C had also developed respiratory insufficiency. Respiratory failure in our patients supervened in all but two cases when the patients were still ambulant, in contrast to many other muscular dystrophies.
Clearer definition of the phenotype in LGMD2I allows comparison with other forms of autosomal recessive LGMD.1 The range in age at onset in LGMD2I is very broad and individuals such as Patient 1 overlap with the milder end of the MDC1C phenotype. Calpainopathy and sarcoglycanopathy can both present from early childhood to adult life, but usually have a tighter band of age at onset than LGMD2I, and in these conditions Achilles tendon contractures and scapular winging are relatively frequent. In dysferlinopathy, there tends to be an onset around the end of the second decade, and the frequent distal involvement is another distinction.1-3,14-17⇓⇓⇓⇓⇓⇓ Like in sarcoglycanopathy, muscle hypertrophy is common in LGMD2I. This is not typically seen in calpainopathy or dysferlinopathy.16,17⇓
Mutation analysis is the gold standard for diagnosis in LGMD, and the uniformity of mutation here is striking and different from the other forms of LGMD where recurrent mutations are rare. Analysis of a muscle biopsy is important in excluding alternative causes of muscular dystrophy10,18⇓ but the protein abnormalities reported are secondary and variable. The finding of absent or reduced laminin α2 chain on immunoblotting suggests LGMD2I, but its absence does not exclude it. Unlike MDC1C, detection of abnormal laminin α2 chain on immunocytochemistry is rare. Two patients showed a reduction in laminin β1 chain labeling, which has also been reported secondarily in Bethlem myopathy, facioscapulohumeral muscular dystrophy, and AD Emery Dreifuss muscular dystrophy (EDMD),19-21⇓⇓ and therefore needs to be interpreted in its clinical context. We also saw a reduction in calpain 3 immunoblotting in two cases. In the patients we were able to study for α-dystroglycan, a reduction in abundance of this protein was detected, but not to the extent reported in MDC1C.6,7⇓ Failure of glycosylation of this and other muscle proteins is an attractive hypothesis for the mechanism by which FKRP mutations cause muscular dystrophy.6,7,22,23⇓⇓⇓
Acknowledgments
Supported by the Muscular Dystrophy campaign of Great Britain and Northern Ireland (F.M. and K.B.), the University Newcastle upon Tyne (Luccock studentship, L.C.), Deutsche Forschungsgemeinschaft (M.P.), and the European Community (Myo-Cluster: GENRE grant OLGI CT 1999 00870) (F.M.).
Footnotes
-
See also pages 1230 & 1341
- Received October 4, 2002.
- Accepted January 17, 2003.
References
Disputes & Debates: Rapid online correspondence
NOTE: All authors' disclosures must be entered and current in our database before comments can be posted. Enter and update disclosures at http://submit.neurology.org. Exception: replies to comments concerning an article you originally authored do not require updated disclosures.
- Stay timely. Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- 200 words maximum.
- 5 references maximum. Reference 1 must be the article on which you are commenting.
- 5 authors maximum. Exception: replies can include all original authors of the article.
- Submitted comments are subject to editing and editor review prior to posting.