Laminin α2 muscular dystrophy

Congenital muscular dystrophies (CMD) are a heterogeneous group of muscle disorders characterized by muscular dystrophy and different degrees of CNS involvement.^5,6 At least four different forms of CMD have been identified to date: Fukuyama type CMD,^7,8 linked to chromosome 9q; muscle-eye-brain disease(MEB)^10,11; Walker-Warburg syndrome¹²; and the classic (occidental) or "pure" form of CMD.¹³ Laminin α2 deficiency has been identified in about 40 to 50% of the cases of classic "occidental" type CMD.^1,2 Although muscle involvement is similar in these different forms, CNS involvement ranges from normality (lamininα2-positive CMD) to abnormal white matter signal by MRI (lamininα2-negative CMD) to severe structural brain abnormalities(Walker-Warburg syndrome, MEB disease, Fukuyama type CMD).

The clinical hallmarks of laminin α2-negative CMD are neonatal onset of muscle weakness and hypotonia, severe clinical course with inability to achieve independent ambulation, and abnormal white matter signal by MRI.^14-20 Laminin α2-positive CMD generally presents with a less severe phenotype and no MRI abnormalities.⁶ Recently, the clinical spectrum of laminin α2-negative CMD has become more heterogeneous than initially thought. Epilepsy has been reported in five lamininα2-negative patients^16,18 and abnormalities of brain neuronal migration in four patients.^21,22 Partial laminin α2 deficiency with a mild clinical phenotype linked to the laminin α2 gene on chromosome 6q^23-25 has been reported. Also, there is some overlap of the clinical features of laminin α2-positive CMD with laminin α2-negative CMD.^26,27

Laminin is a major component of the extracellular matrix and is composed of three different subunits: one heavy chain (α) and two light chains(β, γ).²⁸ The muscle-specific laminin isoform is α2β1γ1, also known as laminin 2.^29,30 Laminin 2 is thought to provide a link between the extracellular matrix and the intracellular cytoskeleton via its putative muscle receptors, dystroglycan, and integrin α7β1.^3,4,31

Relatively few laminin α2 mutations have been identified to date in laminin α2-negative CMD patients^32-36 or in partial laminin α2-deficient CMD.^21,23 The specificity of biochemical abnormalities for the corresponding gene mutations (primary versus secondary laminin α2 deficiency) has not been addressed in these articles. Secondary deficiencies are a concern as laminin α2 assembles in a heterotrimer complex, and it also takes part in an elaborate supramolecular assembly with various other extracellular matrix proteins. Thus, the primary defect of some interacting proteins might result in a destabilization of laminin α2 itself, and thus in a secondary deficiency. Many examples of such secondary deficiencies are now well described with regards to the sarcoglycans and dystroglycans.³⁷ The issue of a secondary laminin α2 deficiency is particularly important for genetic counseling and prenatal diagnosis. Most prenatal diagnosis done to date relies on laminin α2 immunostaining of chorionic villi sample (CVS) or linkage analysis in 6q and assumes that all patients with deficient lamininα2 protein on muscle biopsy have mutations in the laminin α2 gene.^38-40

Here we studied 22 CMD patients showing complete deficiency of lamininα2 at the molecular, biochemical, and clinical levels. We identified at least one laminin α2 gene mutation in 95% of the CMD patients studied, showing that biochemical laminin α2 deficiency detected in the patient's muscle biopsy via laminin α2 immunofluorescence or immunoblotting is most often due to primary laminin α2 gene mutations. Moreover, we describe the development of a protein truncation test (PTT) for this disorder, and show that this new test is the fastest and most reliable method for mutation analysis in completely laminin α2-negative CMD.

Materials and methods. Patient population. We screened our database of over 2,700 frozen muscle biopsies referred to this laboratory for biochemical studies for patients meeting clinical (early onset of muscle weakness and hypotonia, joint contractures, markedly delay motor milestones, association of muscle weakness, and CNS involvement) and histopathologic (muscle histopathology consistent with a dystrophic process) criteria for congenital muscular dystrophy.

Clinical data. Each patient was evaluated by the referring physician. Each physician was asked to report on standardized clinical information including sex, age at biopsy, clinical presentation, best motor achievement, presence of contractures, cognitive function, serum creatine kinase (CK), respiratory function, cardiac function, and MRI studies.

Laminin α2 immunofluorescence and immunoblotting. Lamininα2 immunofluorescence was done using a commercially available antibody directed against the carboxyl terminus of the laminin α2 (mAb 1922, Chemicon, Temecula, CA; used 1:5,000 dilution) on 4-µm-thick cryosections from each patient muscle biopsy as previously described.³⁴ Laminin α2 immunoblot was done as previously described.³⁴

DNA analysis. Blood was collected in ethylenediaminotetraacetic acid (EDTA) tubes from the parents of nine patients. DNA was isolated from peripheral blood or from cryosections of muscle biopsy specimen as previously described.⁴¹

Mutation detection. RNA extraction, cDNA synthesis, and reverse transcription (RT)-PCR was done as previously described.³⁴ RT-PCR was done using 20 overlapping primer sets covering the entire coding sequence of the laminin α2 gene. RT-PCR products were then denatured and loaded on native acrylamide gels for single strand conformational polymorphisms (SSCP) analysis. Three different conditions were used for SSCP: 5% acrylamide, 10% glycerol, room temperature; 5% acrylamide, 4 °C; and MDE (FMC Bio-Products, Rockland, ME) matrix at room temperature. Conformers were directly excised from the dried gel, eluted in water, and reamplified for direct automated sequencing using cycle sequencing and dyedeoxyfluor kits(ABI, Foster City, CA).

PTT analysis. Eight sets of overlapping primers covering most(∼97%) of the laminin α2 cDNA coding sequence were designed as follows:

• PTT 8F: 5′ GGATCCTAATACGACTCACTATAGGGAGACCACCATG TCC TCT GGC TCC CGA GAA GTG 3′-1350R 5′ GTG TCA AGG ATG AGA AAC ATG CTC G 3′;

• PTT 1001F: 5′ GGATCCTAATACGACTCACTATAGGGAGACCACCATG TGT GAT CAG TGC TGT CCA GGA T 3′-1750R 5′ ACT TGG ACT CAC CTC AGC AGA TC 3′;

• PTT 1451F: 5′ GGATCCTAATACGACTCACTATAGGGAGACCACCATG GCC TGT AAC TGC AGT GGG TTA GG 3′-3293R 5′ AAT ACA GGC CAA TGC AAC TGT C 3′;

• PTT 2981F: 5′ GGATCCTAATACGACTCACTATAGGGAGACCACCATG GAC TGT GAA GAG AGT GGA CAA TG 3′-4527R 5′ GCA GAA GGA CTT GAC GAC TAC CG 3′;

• PTT 4101F: 5′ GGATCCTAATACGACTCACTATAGGGAGACCACCATG CAA AGC AGG ATT TCT GAA ATC TC 3′-5942R 5′ CAG GTC CTC GGG GTT TAT TAA AGG 3′;

• PTT 5552F: 5′ GGATCCTAATACGACTCACTATAGGGAGACCACCATG GGC AAT GAC ATA CTC GAT GAA GC 3′-7300R 5′ TTC CGT TGT CAG CAA TCA AAA CCA T 3′;

• PTT 7001F: 5′ GGATCCTAATACGACTCACTATAGGGAGACCACCATG TTC CGA GAA AAA GAA GGT GAC TGC 3′-8537R 5′ CAT GGC TGC GAT CAA TCA TGC TG 3′;

• PTT 7741F: 5′ GGATCCTAATACGACTCACTATAGGGAGACCACCATG TTG GGA AGT GGA GGG ACA CCA G 3′-9245R 5′ TGA CAC AAA TGA CCC TGT GTT TG 3′

The forward primer of each set included signals for transcription by T7 polymerase (TAATACGACTCACTATAGGG, bacteriophage promoter T7 sequence) and an in vitro translation initiation sequence (AGCCACCATG) at their 5′ ends.

To obtain adequate PCR product for transcription-coupled translation, it was necessary for all regions (except PTT 7741F-9245R) to be amplified by nested PCR, with an initial amplification using external primers. For the pre-amplification, seven sets of overlapping primers were designed to cover laminin α2 cDNA sequence between nucleotide 1 to 9,217. PCR was conducted for 25 cycles under stringent conditions (annealing and extension at 68 °C).

The second round of nested PCR reactions contained about 1 µL of the first PCR amplification, 50 ng of each primer, 100 nM of each dNTP, and 2.5 U of AmpliTaq (Perkin-Elmer; Branchburg, NJ) in 25 µL total volume. Reaction conditions were 3 minutes at 94 °C to denature; 10 seconds at 94 °C, 30 seconds at 65 °C, and 4 minutes at 68 °C for 10 cycles; and 10 seconds at 94 °C, 30 seconds at 65 °C, and 4 minutes at 68 °C with 20-second extension for 16 cycles with an extension at 68 °C for 10 minutes for all primer sets except primer set PTT 1001F-1750R, where annealing was at 60 °C. For primer set PTT 7741F-9245R, 0.5 U of Elongase (Gibco-BRL; Gaithersburg, MD) were used, and PCR conditions were: 94 °C for 3 minutes to denature; and 94°C for 30 seconds, 65 °C for 30 seconds, and 68 °C for 2.5 minutes for 35 cycles with an extension at 68 °C for 10 minutes.

The RT-PCR products were then transcribed and translated in vitro using [³⁵S]methionine and the TNT Coupled Reticulocyte Lysate Systems (Promega; Madison, WI) following the manufacture's suggestions. Briefly, the PCR products included the transcription initiation, signal for T7 RNA polymerase, and the translation initiation in frame with the laminin α2 coding sequence. The PCR products were used as template for the in vitro synthesis of RNA and the simultaneous translation into protein using rabbit reticulocyte lysate. A 15% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) polyacrylamide mini gel(Biorad; Hercules, CA) was used to size fractionate the translation products. After electrophoresis, the SDS-PAGE gels were treated with EN³HANCE autoradiography enhancer (DuPont; Wilmington, DE) to increase the signal. The dried gels were then exposed to X-ray films.

Truncated protein sizes were calculated using the molecular weight marker as reference. Putative stop codons were then localized on the lamininα2 cDNA sequence based on the truncated protein size. The corresponding sequence was then directly PCR amplified and cycle sequenced.

Confirmation of laminin α2 mutations. All the identified mutations were confirmed on the genomic DNA of the patients and the parents(where available) using exon-specific primers. When the mutation resulted in an endonuclease restriction enzyme site change, the PCR product was digested with the appropriate enzyme. Insertions and duplications were directly PCRed and size fractionated on denaturing acrylamide gels.

Results. Clinical and biochemical features in lamininα2-negative CMD. We screened our database of over 2,700 frozen muscle biopsies referred to this laboratory for biochemical studies for patients meeting clinical and histopathologic criteria for congenital muscular dystrophy (see Methods). A total of 75 patients meeting these criteria were selected.

Laminin α2 immunofluorescence, using an antibody directed against the carboxyl terminus of the protein, was performed in all 75 biopsies. Thirty-two patients showed normal laminin α2 immunostaining(32/75; 43%), 30 showed complete laminin α2 deficiency (30/75; 40%), and 13 showed partial laminin α2 deficiency (13/75; 17%) in their muscle biopsies (figure 1). Laminin α2 mutation studies were conducted only on the completely laminin α2-deficient patients. Of the 30 laminin α2-deficient muscle biopsies, 22 were selected for lamininα2 gene mutation analysis based on the availability of adequate muscle specimen. In these biopsies, laminin α2 immunoblot was done to confirm laminin α2 deficiency. No residual laminin α2 was detected in any of the samples (complete deficiency).

Thirteen boys and nine girls were studied (all clinical information is detailed in table 1). The M:F ratio was 1.75:1. The age at biopsy ranged between 3 to 144 months (mean, 26 ± 40). Floppiness was the clinical presentation in all. One patient required mechanical ventilation, and in three, a nasogastric tube was placed for a few days soon after birth. Contractures were infrequently present at birth (4/22; 18% of patients), but subsequently developed in all and included elbows, wrists, hips, knees, and ankles. In three patients minor skeletal deformities were present at birth, and included funnel chest, arched palate, and clubfoot. One patient presented with bilateral wrist drop. Motor milestones were severely delayed in all. The average age for sitting unsupported was 20 months (20± 7.3; range, 7 to 36 months) in the 12 patients who achieved this task. Of the remaining 10, only two were older than 36 months. The best motor achievement by any patient was ability to control the head when sitting(4/22; 18%). Two patients were able to walk a few steps with bilateral support (Patients 6 and 10). CK was markedly increased in the early stage of the disease (4,260 ± 2,292 U/L), but declined with advancing age. Restrictive lung disease was present in 12 patients: three required bi-level positive airway pressure (BiPAP) or intermittent positive pressure ventilation (IPPB), and one had a tracheostomy. Four patients died at ages 5, 6, 9, and 10 years from respiratory failure; the remaining 18 patients are still living.

All patients were cognitively normal. In 14 patients, MRI showed abnormal white matter signal. In three patients, results of CT scan of the brain done at 6 days and 3 and 4 months were normal. In all three, subsequent MRI studies showed the typical white matter changes. In an additional four patients, imaging studies were reported normal in an early stage of the disease, and were not repeated at an older age.

Seizures were reported in three patients. Patient 1 had a single episode of generalized tonic-clonic seizures at 3 months, with no other episodes (the patient is now 4.5 years old). Patient 21 had her first seizure at age 9 years. It was described as her seeing red and green in the right field of vision, with subsequent staring, drooling, and being unresponsive for approximately 1 minute. She then showed eye jerking to the right upper gaze, fell asleep, and was incontinent of urine. Postictal state lasted 15 minutes. EEG showed occipital slowing of delta frequency on the left side. No evident paroxysmal features were recorded. One month later, she had another seizure with similar features. She began taking carbamazepine, and has been seizure free since then. Patient 22 had her first generalized tonic-clonic seizure at 10 years of age, and has continued to have seizures (3 to 4 per month) until her current age of 12. EEG showed a focus in the medial posterior right parietal lobe. None of these three patients show evidence of cortical dysplasia on their MRI studies.

Laminin α2 mutation studies. SSCP analysis was first used to screen for laminin α2 mutations in our cohort of lamininα2-negative CMD. The coding sequence of the laminin α2 gene was covered by 20 sets of overlapping primers. PCR reactions were run under three different electrophoresis conditions. We identified 22 laminin α2 gene mutations of 44 putatively mutated chromosomes (22/44; 50%). In nine patients (9/22; 41%), both mutations were identified: five patients were compound heterozygotes and four were homozygotes. In four patients(4/22; 18%), only one mutation, and in nine patients (9/22; 41%), no laminin α2 mutations were identified (figure 2). The deletion mutation at nucleotide position 2096-2097, resulting in a frame-shift, was present in 23% of the patients (16% of the chromosomes) either as a homozygous (two patients) or heterozygous (three patients) mutation. A large number of polymorphisms distributed over the entire sequence of laminin α2 gene were also detected (see figure 2).

PTT analysis was conducted to verify if the nine patients in whom no laminin α2 gene mutations were identified by SSCP were mutation-negative. Eight sets of nested, overlapping primers were designed to cover 97% of the laminin α2 coding sequence. PCR products were transcribed and translated in vitro, and proteins run on SDS-PAGE acrylamide gels (figure 3). PTT identified lamininα2 mutations in 21 of 22 patients studied (95%). Seventy-five percent of the all putative laminin α2 mutations were identified (33 lamininα2 mutations of 44 chromosomes screened). All the mutations previously detected by SSCP were confirmed by PTT: size-appropriate truncated proteins were identified for each SSCP nonsense mutation. In 12 patients (12/22; 55%), both mutations were identified. The three additional patients identified by PTT were homozygous. In nine patients (9/22; 41%), only one mutation, and in one patient (1/22; 4%), no laminin α2 mutations were identified (figure 4).

In two patients, a clear truncated protein was detected in at least two different PCR/in vitro transcription/translation assays (Patients 10 and 19). The sequence of the entire region cover by each PCR was then sequenced, but no mutations could be identified.

In total, nine deletions, two insertions, one deletion/insertion, three nonsense mutations, and one missense mutation resulting in the change of the initiator methionine to a threonine were identified (table 2). These mutations were present both in the globular and in the rod-like domains of the protein. Thirty-three percent were in the carboxyl terminus (G domain) and 24% in the globular IVb domain (figure 4). All the mutations were confirmed at the genomic DNA level. In one patient (Patient 5), a homozygous 9154-9157 insertion at cDNA level was heterozygous at genomic DNA level. We hypothesize that in this patient, a mutation in the lamininα2 promoter region might interfere with the normal expression of that allele.

The genomic DNA of parents was collected for nine CMD patients. In all six, an appropriate inheritance pattern was shown between the patient and his or her parents.

Discussion. We studied 22 complete laminin α2-negative CMD patients and showed, for the first time, that in the vast majority of the cases biochemical laminin α2 deficiency is due to primary lamininα2 gene mutations. We also showed that PTT analysis is the faster and more accurate method to study laminin α2 mutations in CMD. We found the clinical phenotype to be relatively homogeneous. Epilepsy was present in at least 10% of the cases. Moreover, we found that CT scan or MRI studies of the brain can be normal at very young ages, but all older patients showed dramatic white matter changes by MRI.

Since the first identification of congenital muscular dystrophy with laminin α2 deficiency,¹ a series of reports has narrowed the clinical spectrum and biochemical aspects of this subset of CMDs. However, there have been few molecular mutation studies correlating primary laminin α2 gene mutations with observed protein deficiency. Molecular studies have been hindered by the large size of the coding sequence of laminin α2 (∼10 Kb), the complex genomic organization of the laminin α2 gene (64 exons),⁴² the labor-intensive SSCP techniques used for mutation detection, and the relatively rare occurrence of laminin α2-deficient CMD.

We studied 22 patients with complete laminin α2-negative CMD, using both SSCP and PTT methods, and were able to identify at least one lamininα2 gene mutation in 21 of them (95%). Using these two mutation detection systems, we determined that PTT is faster, more convenient, and more sensitive than SSCP. PTT not only identified all the previously SSCP-detected changes, but also identified an additional 10 laminin α2 gene abnormalities (25% of the total putative mutant alleles). Two patients showed clearly truncated PTT products; however, we were unable to identify nucleotide changes at the DNA level. We attribute this to difficulties in sequencing the mutant allele, and further investigations on these two patients are underway. We identified both mutant alleles in 12 patients (55%), but only one mutation in 9 heterozygous patients (41%). Because laminin α2-negative CMD is an autosomal recessive disorder, nine other laminin α2 mutations remained undetected by our analyses. According to these findings, the sensitivity of our PTT analysis is about 80%(33 mutations identified of 42 putatively mutated chromosomes). This incomplete sensitivity of the PTT assay is probably due to a failure to amplify the mutated allele (preferential amplification of the normal allele, or underexpression of the mutated allele), or to a failure to detect very small deletions/insertions mutations that do not alter the reading frame of the RNA. Other explanations are mutations in the primer binding sites and the inability of PTT to detect missense mutations. However, all mutations identified to date in laminin α2-negative CMD, by us and others, are nonsense mutations.

Our data showing that primary laminin α2 gene mutations underlie the large majority of laminin α2 protein deficiency are clinically relevant. Prenatal diagnosis and genetic counseling in lamininα2-negative CMD are done through CVS laminin α2 immunostaining or linkage analysis using microsatellite markers in 6q. If secondary lamininα2 deficiency exists, then linkage analysis on 6q, where lamininα2 maps, might result in false negative or false positive linkage data. CVS immunostaining, on the other hand, could similarly be inaccurate when no information about the developmental expression of the defective primary protein is known. Our data suggest that prenatal diagnosis by CVS or linkage should be accurate in >95% of patients showing complete laminin α2 deficiency on biopsy. Studies similar to that presented here must be done on patients showing partial laminin α2 deficiency before statements on biochemical/molecular correlations can be made.

Interestingly, in our cohort of laminin α2-negative patients, epilepsy has been reported in three. In one of these, a seizure was reported. In a second patient, complex partial seizures, secondarily generalized, were reported. No clinical details of the seizures were available in the third patient. Thus, at least 10% of laminin α2-negative CMD patients might have epilepsy in their lifetime. It is also interesting to note that the onset of seizures in these two patients was after 6 years of age. Because the average age of our patient population is below this threshold, it is possible that the percentage of laminin α2-negative patients who might suffer epileptic seizures would be higher than 10%. We tried to correlate specific laminin α2 gene mutations to the occurrence of epilepsy in the two patients with reliable evidence of epilepsy. Mutations in the IVb domain were present in one, and mutation in I+II domain in the other. These two domains of the protein are substantially different. The I+II domain is involved in coiled-coil assembly, and thus is probably extremely important in the chain to chain assembly; the IVb domain is a globular domain in the short arm of the heterotrimer complex. Because the biochemical consequences of any of the laminin α2 gene mutations we described is complete lack of expression of the protein, we believe that genotype/phenotype correlation in complete laminin α2-negative CMD is not significant.

In 8 of the 22 patients we studied early (2.85 ± 1.53 months [n = 5], with a range between 6 days and 4 months), brain imaging studies have been reported as normal, both by CT or MRI. In four of these eight patients, subsequent MRI studies showed the classic abnormal white matter signal; the analysis was not repeated in the remaining four. These data, consistent with an earlier study,⁴³ suggest that early scans(before 4 months of age) in congenital muscular dystrophy patients appear normal, and this can be diagnostically misleading.

Why laminin α2 deficiency results in abnormal white matter signal by MRI in CMD remains unclear. Oligodendrocytes in the brain lack a basal lamina, and histopathologic studies have shown that laminin α2 in the adult brain is localized only in the basement membrane of blood vessels.⁴⁴ In fetal brain, laminin α2 may play an essential role in serving as a substrate promoting proper neurite outgrowth, as has been shown in chicken and mouse models.^45-47

To elucidate the role of laminin α2 in brain, it would be interesting to study laminin α2 function in partial laminin α2 deficiency CMD patients further. In this category of patients, it is likely that missense mutations are common. The identification of different missense mutations involving different domains of the mature protein may allow their use as substrate for in vitro neurite outgrowth assays, and provide invaluable insight into the neurodevelopmental functions of lamininα2.

Acknowledgments

The authors thank David Duggan and Hisashi Kobayashi for thoughtful suggestions and help. We are grateful to Bo Liu for his expert assistance in Protein Truncation Test.