Article Text
PDF
Abstract
Objective: To identify pathogenic mutant alleles of the PYGM gene in “genetic manifesting heterozygous” patients with McArdle disease—that is, those in whom we could only find a sole mutant allele by genomic DNA analysis.
Methods: We studied four unrelated patients. PCR-RFLP, gene sequencing, and muscle cDNA analysis were performed to search for mutations in the PYGM gene. The effects of the mutations were evaluated by in silico analysis, and gene expression was assessed by real-time polymerase chain reaction (PCR).
Results: Patient 1 was a compound heterozygous for the p.G205S missense mutation and for a novel “in frame” mutation, p.Q176_M177insVQ, resulting from a retention of six nucleotides from the 3′-end sequence of intron 4. Patient 2 was heterozygous for the common nonsense mutation p.R50X, and for a 1094 bp, c.1969+214_2177+369del mutation, spanning from intron 16 to intron 17 sequences. Furthermore, mRNA expression level was dramatically reduced consistent with nonsense mediated decay. Patient 3 was heterozygous for the p.R50X substitution, and patient 4 was heterozygous for the relatively common private Spanish mutation p.W798R. These two patients harboured a heterozygous exonic synonymous variant, p.K215K. Quantification of gene transcripts in patient 3 revealed a drastic decrease in the relative expression of the gene, which strongly supports the possibility of nonsense mediated decay.
Conclusions: Our results indicate that skeletal muscle cDNA studies in “genetic manifesting heterozygous” patients with McArdle disease are prone to identify their second mutant allele.
Statistics from Altmetric.com
Mutations in the skeletal muscle isoform of the glycogen phosphorylase gene (PYGM, myophosphorylase; MIM 608455) result in glycogen storage disease type V (McArdle disease, GSDV; MIM 232600), the most common glycogenolytic disorder in skeletal muscle. The diagnosis is often in the second or third decade of life manifesting as exercise intolerance, premature fatigue and muscle weakness during exercise, frequently accompanied by myalgia and cramps, and sometimes myoglobinuria.1 Molecular heterogeneity has been demonstrated by the identification of more than 95 different mutations described to date.2–6 The most common mutation in Caucasian populations is p.R50X (with an allelic frequency ranging from 31–72%).6 7 Although McArdle disease is thought to be inherited as a recessive disorder, in some patients, the so-called “genetic manifesting heterozygotes”, only one allele with a pathogenic mutation has been identified.1 8
Disturbances at mRNA processing were described as the underlying molecular mechanism for some PYGM gene mutations. In particular the p.R50X and most of frameshift mutations resulting in premature termination codons (PTCs) undergo mRNA nonsense mediated decay,9 and an exonic synonymous variant (p.K609K) leads to abnormal mRNA splicing species.10
We report four unrelated patients with McArdle disease in whom we found a sole mutant pathogenic allele in the PYGM gene by genomic DNA analysis.11 We studied cDNA from these patients in an attempt to identify the second mutant allele. Different mutations and mutant molecular mechanisms are described, suggesting that transcript analysis is a valuable approach in the definitive molecular characterisation of those patients with McArdle disease previously considered to be “genetic manifesting heterozygotes”.
PATIENTS AND METHODS
Patients and controls
We studied five patients belonging to four unrelated families (one patient was the brother of patient 4) with McArdle disease proven by histochemical and biochemical deficiency of myophosphorylase in skeletal muscle (table 1). Age at diagnosis ranged from 25–58 years. Common features of the disease were observed in all patients—that is, long term history of exercise intolerance (including the “second-wind phenomenon”) since childhood presenting with myalgia, cramps, and persistent elevated resting serum creatine kinase followed by sharp increases after crisis (range 2300–15000 U/l, normal <170 U/l). All patients had at least one episode of myoglobinuria. Muscle weakness was present in three patients (patients 2, 4 and his brother), all of them >45 years of age. In three patients (patients 3, 4 and his brother) an ischaemic forearm exercise test was performed showing a flat-response curve of lactate and an exaggerated increase in serum ammonia profile. Electromyogram was performed in two patients (patients 1 and 2) revealing signs of mild myopathy.
Genomic DNA samples from muscle or blood from 200 healthy individuals and cDNA from muscle of 30 controls were analysed to discard the presence of the novel mutations in the general population.
Written consent was obtained from all individuals. The study was approved by the institutional ethics committee (Hospital Universitario 12 de Octubre, Madrid, Spain) and was in accordance with the Declaration of Helsinki for Human Research.
Molecular analysis
DNA studies
Genomic DNA was isolated from muscle by proteinase K and phenol/chloroform extraction. Screening for the three most common mutations in the Spanish population—p.R50X, p.G205S, and p.W798R—were carried out by polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) methods described elsewhere.12 13 The entire coding sequence and intron/exon boundaries of the PYGM gene were amplified by using primers described by Kubisch et al,11 purified by GFX Gel Band Purification Kit, GE Healthcare Europe and forward/reverse sequenced by ABIDyeDeoxy Terminator Cycle sequencing Applied Biosystems kit in the ABI 3100XL genetic analyzer (Applied Biosystems, Foster City, California, USA). Deletion mapping was performed by long range PCR amplification using TaKaRa LA Taq polymerase (Takara Biotechnology, Otsu, Japan) and then by sequencing amplicons using a primer walking analysis (table 2)
mRNA and cDNA analysis
Muscle total RNA was isolated by the ToTALLY RNA Kit (Ambion, Applied Biosystems). Thereafter, PYGM cDNA was amplified in two overlapping fragments by the high capacity cDNA archive kit (Applied Biosystems) as described elsewhere,9 using a fragment in the porphobilinogen deaminase gene (PBGD) as a control. The PYGM cDNA and PBGD fragments were run in 1.5% agarose gel. PCRs amplicons were subcloned into pGEM-T Easy Vector Systems (Promega, Fitchburg, Wisconsin, USA) following the manufacturer’s instructions. After the transformation in the ultracompetent cells JM109, the insertions were directly amplified from each colony, and purified using Wizard Plus SV Minipreps (Promega). For sequencing the PYGM fragments, eight internal primers in an Applied Biosystem 3100XL automated DNA sequencer were used.
PYGM mRNA levels were quantitated by real-time PCR, using TaqMan fluorogenic probes and a 7500 Real Time PCR System (Applied Biosystems) as described elsewhere.9 All reactions were run in triplicate and analysed using the Applied Biosystems SDS 7500 system software (Applied Biosystems). Expression values were recalculated considering 100% the average of the results obtained in the control samples.
In silico analysis
The Splice Site Finder algorithm (www.UMD.be/SSF) was used to predict if intronic mutant sequences were putative novel splicing sites.14 The Mfold web server was used to analyse the folding and stability of mutant mRNAs.15
RESULTS
In patient 1 (table 1) the common heterozygous p.G205S amino acid change was found by PCR-RFLP, while sequencing of PYGM gene at gDNA level displayed a novel intronic heterozygous nucleotide transition, c.529-8g>a (fig 1). This intronic mutation was located in a non-conserved nucleotide residue of the acceptor splicing sequence. By analysing a 1000 nt sequence surrounding the mutation, the Splice Site Finder algorithm14 identified the mutant region both as a novel donor sequence (ag/gtgcag, score 66 vs 92 for a wild-type donor sequences) and as a novel acceptor sequence (ctgacctgccttag/g, score 76 vs 79 for the wild-type acceptor sequence). To assess whether this mutation could be the underlying cause of any disturbance in the PYGM gene splicing machinery we amplified and cloned the cDNA from the patient’s muscle. cDNA sequencing of some clones showed a retention of six nucleotides in the coding region from the 3′-end of wild-type intron 4 (fig 1). In addition, we found a relative over-expression (154% above controls) of myophosphorylase transcripts quantified by real-time PCR.
In patient 2, the common p.R50X mutation was identified in one allele by gDNA analysis. Because the patient had complete absence of myophosphorylase activity in muscle, we searched for another mutation using muscle cDNA amplification and cloning, identifying a skipping of exon 17. A long range PCR, spanning from intron 14 to intron 18, revealed two fragments of 4.5 Kb and 3.5 Kb. We subcloned the long range PCR amplicons in a pGEM-T vector and then isolated and sequenced the 3,5 kb clones to map the deletion by a primer-walking method. These studies demonstrated a large deletion of 1094 bp, c.1969+214_2177+369del, that spanned from intron 16 to intron 17 and was flanked by an intronic direct repeat “agacca” (fig 2). Muscle quantification of PYGM gene transcripts by real-time PCR showed a relative expression of 3% with regard to control values (table 1).
The p.R50X was found in one allele in patient 3 whereas the p.W798R change was observed in a sole allele in patient 4 and his brother. Sequencing of the PYGM gene revealed a novel heterozygous synonymous variant, c.645G>A (p.K215K) in all these patients. This nucleotide change was localised 16 nucleotides upstream from the 3′-end of exon 5. Amplification and sequencing of the muscle cDNA in patient 3 (the sole patient in whom remaining muscle was available for transcript analysis) revealed a normal size cDNA harbouring an apparent homozygous c.645G>A change. Relative expression of the gene quantified by real-time PCR showed a 5% transcripts level with respect to mean control values. By using the mfold web server15 to predict the folding of PYGM-mRNA we found that nucleotide 645G was located in a secondary structure interacting with a uracile at position 725 and that the wild-type folding structure had a free energy lower than that of the mutant mRNA (−1380.30 KJ vs −1283.18 KJ).
Each of the three novel mutations described here were absent in 200 gDNA and 30 cDNAs from healthy controls of the same ethnic origin.
DISCUSSION
Mutations affecting the mRNA splicing or maturation processes (including intronic mutations) often escape from genetic diagnosis as most patients are screened by sequencing gDNA. A few studies9 10 have demonstrated that a number of mutations in this gene may lay on non-coding regions and affect transcripts processivity. In this regard, in 62 Italian index patients 10.5% of mutant alleles were not identified by sequencing,5 whereas 2.6% of mutant alleles failed to be found by sequencing analysis in 95 Spanish index patients.6 By contrast, some other studies were successful in identifying both putative mutant alleles in all patients with McArdle disease.3 7 Now we have found seven “genetic manifesting heterozygotes” in a series of 125 unrelated patients who were characterised at the molecular level, representing 5.6% of the patients and 2.8% of mutant alleles. Here, we have only been able to study four out of seven of these patients with histochemical and biochemical deficiency of myophosphorylase in skeletal muscle. All these patients presented with the cardinal symptoms of the disease including myoglobinuria.
Screening of gDNA by PCR-RFLP analysis revealed that patient 1 was heterozygous for the p.G205S missense mutation, that patients 2 and 3 were heterozygous for the most common nonsense mutation p.R50X, and that patient 4 was heterozygous for the relatively common (in Spanish populations) p.W798R mutation. In addition, sequencing of the exons and intron/exon boundaries of the PYGM gene showed an intronic heterozygous transition (c.529-8g>a) in a non-conserved position of the splice canonical sequence in patient 1, and an apparent heterozygous exonic neutral synonymous variant16 (p.K215K) in patients 3 and 4.
Although myophosphorylase activity was negligible in the four patients, we considered these patients to be “genetic manifesting heterozygotes” relying on their gDNA molecular genetic data. The occurrence of manifesting heterozygotes with McArdle disease has been a long lasting controversy that could have important implications in diagnosis and genetic counselling.17–19 Andersen et al19 showed that non-symptomatic heterozygotes for PYGM gene mutations had maximal oxidative capacities and peak lactate responses identical to control subjects, suggesting that they are not prone to develop symptoms of McArdle disease. Therefore, given that our “genetic manifesting heterozygotes” patients showed a clear-cut pattern consistent with McArdle disease, it was reasonable to suspect: (1) a pathogenic role of both the intronic heterozygous transition (c.529-8g>a) in patient 1, and the exonic synonymous variant (p.K215K) in patients 3 and 4; and (2) the existence of another mutation(s) undetectable by routine gDNA analysis of the PYGM gene.
To address these possibilities, we carried out muscle cDNA analysis. Patient 2 harboured a 1094 bp deletion, c.1969+214_2177+369del, spanning from intron 16 to intron 17 sequences, thus including exon 17. This nearly 1.1 Kb large deletion probably originated from the presence of a flanking direct repeat “agacca” at the breakpoints of the deletion which led to exon 17 skipping in the mature mRNA. The removal of approximately 10% of the catalytic site of the enzyme results in a frameshift that predicts the synthesis of a truncated protein 69 amino acids smaller than the wild-type one. Moreover, mRNA expression level was dramatically reduced (3% of mean control) which is consistent with the effect of nonsense mediated decay in the expression of the two alleles harbouring PTCs (for example, the p.R50X and the frameshift mutation reported here), as previously documented.9 Consistently, Bruno et al5 reported skipping of exon 17 in two patients, but the size and precise mechanisms of the rearrangement were not established.
Patient 1 showed a heterozygous intronic change in the c.529-8 g>a located in a non-conserved site of the intron 4 consensus splicing canonical sequence. Furthermore, cDNA analysis showed a retention of six nucleotides in the coding region from the 3′-end sequence of wild-type intron 4 (fig 1). Consistent with in silico analysis, this aberrant molecule is supposed to be built by the activation of a newly generated acceptor splicing site sequence “agGT” (fig 1). This nucleotide retention predicts an insertion of two amino acid residues in the mutant protein resulting in an “in frame” mutation, p.Q176_M177insVQ, which could alter the polarity of the myophosphorylase dimerisation region.20 Other in-frame mutations have been previously reported, such as the p.V239del in Italian patients,5 and the p.F710del21 in Japanese patients. However, the one identified here is the first in-frame mutation resulting in the insertion of two amino acids in the PYGM protein.
Patients 3 and 4 harboured a novel heterozygous synonymous variant, c.645G>A (p.K215K), localised 16 nucleotides upstream from the 3′-end of exon 5. Muscle for further studies was available only in patient 3. cDNA analysis in this patient showed a normal size molecule carrying an apparently homozygous c.645G>A synonymous variant, suggesting that only the mRNA carrying this change was transcribed. These results seem not to be consistent with the possibility that exonic silent synonymous variants may affect splicing processes in patients with McArdle disease.10 16 However, quantification of gene transcripts by real-time PCR revealed a drastic decrease (that is, 5% with respect to control values) in the relative expression of the gene, which strongly supports the possibility of nonsense mediated decay in the expression of his two alleles (p.R50X and p.K215K).9 Unfortunately, the lack of muscle sample in patient 4 did not allow us to confirm these data in this patient. Evidence for alteration in folding of PYGM-mRNA came from the use of the mfold program15 which revealed that the nucleotide 645G is located in mRNA secondary structure, and that the free energy of the folding structure was higher in the mutant than in the wild-type, thus predicting instability for the mutant. Taken together, our data suggest that the low level of expression of the p.K125K allele might be related to the synthesis of aberrant mRNAs with exon (or exons) skipping leading to a frameshift, activation of PTCs, and subsequent nonsense mRNA decay.
Our results indicate that an approach using skeletal muscle cDNA studies in genetic manifesting heterozygous patients with McArdle disease is more reliable for identifying a putative second mutant allele. The satisfactory examination of the molecular basis of this monogenic disorder might have relevant clinical implications in molecular diagnosis,5 11 in genetic counselling, and in searching for new effective strategies of molecular therapies.22
Acknowledgments
We appreciate Dr Hernández (Ciudad Real Hospital) and Dr Pérez-Calvo (Lozano Blesa Hospital) for referral of patients to our institution.
REFERENCES
Footnotes
Competing interests: None declared.
Funding: G-CI was supported by a fellowship from Fondo de Investigacion Sanitaria (FIS) (PI04/0487) and by a contract from Fundación de Investigación Biomédica Hospital 12 de Octubre, Madrid, Spain. R-JC was supported by a contract from FIS (CA05-0039). GN-G was supported by a fellowship from FIS (PI04/0362), Instituto de Salud Carlos III. Grant sponsor: Fondo de Investigación Sanitaria (FIS); Grant numbers: PI040487; PI041157; PI040362. Grant sponsor: Spanish Network for Rare Diseases (CIBERER), Madrid, Barcelona.
Patient consent: Obtained