A survey on mutation spectrum in Iranian patients with limb-girdle muscular dystrophies

Abstract

Limb–girdle muscular dystrophies (LGMD) designate diverse types of muscular dystrophies that predominantly affect proximal skeletal muscles. Although both autosomal recessive and dominant forms exist, the majority of cases are inherited in an autosomal recessive manner. Since the spectrum of genetic variants that cause this disorder is quite broad, next-generation sequencing techniques are the best diagnostic tools for LGMD. In this study, we provide an overview of mutation spectrum of LGMD-related genes in the Iranian patients using whole exome sequencing. Notably, CAPN3 and LAMA2 genes were the genes encompassing the highest frequencies of pathogenic or likely pathogenic variants in this cohort. Pathogenic and likely pathogenic variants were identified in CAPN3 gene in total of 10 cases out of 48 cases tested (20%). In addition, different variants in each of POMGNT1 and TTN genes were detected in five and four patients, respectively. Three patients had DYSF variants (6%). While the inheritance of the majority of cases was supposed to be in an autosomal recessive manner, in three cases, the disease inheritance was best explained by the dominant type (c.947 C > T variant in the DNAJB6, c.746G > A variant in the LMNA, and c.1417G > A variant in the TNPO3). The current study broadens the spectrum of LGMD-related mutations among Iranian patients and facilitates genetic counseling in the affected families.

Introduction

Limb–girdle muscular dystrophies (LGMD) is a term to designate diverse types of muscular dystrophies that predominantly affect proximal skeletal muscles [1]. They are inherited as autosomal disorders, either dominant or recessive. They are extremely heterogeneous in terms of underlying genetic causes [1]. From a clinical point of view, they are usually nonsyndromic, with clinical manifestations classically limited to the skeletal muscle. Their age of onset varied from childhood to adulthood displaying a spectrum of a very severe forms in early life to very mild types being appeared in late life [2]. LGMDs are caused by pathogenic genetic variants in genes that encode different proteins of the muscle fibers, located in various subcellular loci including nucleus, sarcomere, sarcolemma, sarcoplasm, and extracellular matrix [3, 4]. The advent of whole exome sequencing (WES) method has facilitated identification of genetic causes of LGMDs with more than 30 autosomal loci identified for this type of disorders [1]. For instance, targeted sequencing of 18 LGMD-related genes in a cohort of Korean patients has led to identification of LGMD2B, LGMD1B, LGMD2A, and LGMD2G as the most common subtypes of LGMD in this country [5]. In addition, application of targeted next generation sequencing covering 420 genes has facilitated identification of genetic cause of LGMD among Chinese patients. The most common subtypes in this cohort have been LGMD2B, LGMD2A, LGMD2D and LGMD1B, respectively [6]. Similar strategy has been used to identify disease-causing variants among LGMD patients from USA, showing contribution of CAPN3, DYSF, FKRP and ANO5 to the pathogenesis of LGMD in this country [7]. A more recent study has shown the efficacy of combined sequence and copy number analysis in diagnosis of LGMD and other myopathies, revealing the important roles of CAPN3, DYSF, GAA, ANO5, and FKRP as major contributors to LGMD [8].

Most of cases are inherited in an autosomal recessive manner and designated as different types of LGMD2. Studies in different populations have shown that 10–33% of LGMD2 cases are due to mutations in the CAPN3 gene that encodes an intracellular nonlysosomal cysteine protease modulated by calcium ions [2]. Besides, 5–30% of cases exhibit mutations in the DYSF gene that encodes a ubiquitous transmembrane protein contributing to the calcium-mediated sarcolemma resealing [2]. Mutations in different genes that encode Sarcoglycans are responsible for 10–20% of adult-onset cases and up to 68% of cases with childhood onset [2].

Appropriate genetic counseling and therapy of LGMD patients rely on efficient genetic diagnosis. Since the spectrum of genetic variants that cause this disorder is quite broad, next-generation sequencing techniques are the best diagnostic tools for LGMD. In this study, we provide an overview of mutation spectrum of LGMD-related genes in the Iranian patients using WES.

Case presentation

This study was performed on 48 Iranian cases of LGMD. Cases were referred to the Comprehensive Genomic Center, Tehran, Iran during 2018–2024 for molecular diagnosis and counseling. All methods were carried out in accordance with relevant guidelines and regulations. All experimental protocols were approved by ethical committee of Shahid Beheshti University of Medical Sciences. Informed consent forms were signed by patients or their legal representatives. Clinical signs and symptoms, biochemical tests and EMG were compatible with the diagnosis of LGMD in most cases. Thus, patients were investigated by WES technique.

Molecular diagnosis

Genomic DNA was obtained from the peripheral blood of LGMD cases using the standard salting-out procedure. The concentration and quality of DNA were evaluated using a NanoDrop 1000 (Thermo Fisher Scientific, USA). Genomic DNA of probands was subjected to WES using an Illumina HiSeq4000 system with paired-end reads of 101 bp and 100X coverage. Exonic and adjoining exon-intron border regions were enriched using SureSelectXT2 V6 kits. After exclusion of low-quality reads, the reads were mapped to the human genome reference (hg37 build) using the Burrows-Wheeler Aligner. Next, Sequence Alignment/Map (SAM) tools were used for detection and removal of duplicates. Then, recalibration and single nucleotide polymorphism/indel calling were conducted. Variant calling and filtering were done using the Genome Analysis Toolkit. The primary step involved manual filtration process. First, formerly reported variants in ClinVar were evaluated for their association with the clinical finding. At that point, we considered the allele state. Finally, the called variants were annotated and ranked using ANNOVAR software. The identified variants were verified by Sanger sequencing in the probands. Furthermore, segregation analyses were conducted in the families that requested prenatal diagnosis for subsequent pregnancies.

Results

Tables 1–7 show a summary of clinical and molecular data. Patients had variable degrees of muscle weakness and elevated CPK levels. A total of 42 patients out of 49 were born to consanguineous parents.

Molecular tests revealed different variants in LGMD-related genes. Ten patients had pathogenic/likely pathogenic variants in the CAPN3 gene and were categorized as Calpainopathy (LGMD 2 A/R1) (Table 1). Another 13 patients were categorized as Dystroglycanopathy (LGMD type C, 9) (Table 2). In addition, four patients were classified as Sarcoglycanopathy (LGMD 2 C-2 F/R3-R6) (Table 3).

Notably, CAPN3 and LAMA2 genes were the genes encompassing the highest frequencies of pathogenic or likely pathogenic variants in this cohort. Pathogenic and likely pathogenic variants were identified in CAPN3 gene in total of 10 cases out of 48 cases tested (20%). In addition, different variants in each of POMGNT1 and TTN genes were detected in five and four patients, respectively. Three patients had DYSF variants (Table 4). While the inheritance of the majority of cases was supposed to be in an autosomal recessive manner, in three cases, the disease inheritance was best explained by the dominant type (c.947 C > T variant in the DNAJB6, c.746G > A variant in the LMNA, and c.1417G > A variant in the TNPO3) (Table 7).

In total, eight novel variants were identified, including c.1532 A > G in the POMGNT1 gene; c.49 A > G in the TRAPPC11 gene; c.500G > T in the SGCB gene; c.6020G > A and c.5680G > T variants in the DYSF gene; and c.8548-1G > C, c.225del and c.2540G > T variants in the LAMA2 gene.

In some cases, the clinical or brain MRI features were not completely compatible with LGMDs. For instance, case 11 represented developmental delay, severe regression, seizure, and dystonia. Molecular tests revealed that she was homozygote for a variant of uncertain significance (VUS) in the DAG1 gene and heterozygote for another variant in the SLC20A2 gene which was associated with idiopathic basal ganglia calcification. Another female patients with short stature (Case 29) was found to be compound heterozygote for a likely benign (c.5102 C > T) and a likely pathogenic variant (c.6020G > A) in the DYSF gene. Another challenging issue was the molecular report of Case 33 who represented developmental delay, hypotonia, decreased DTR, and decreased level of Biotinidase. His EMG-NCV showed myopathy and his brain MRI showed atrophy. Molecular tests revealed that he was compound heterozygote for two VUS in the LAMA2 gene, heterozygote for a pathogenic variant in the LMNA gene and heterozygote for a VUS in the COL12A1.

White matter change in the brain MRI was reported in Case 35 who was further demonstrated to have a homozygote pathogenic variant in the LAMA2 gene (c.283 + 1G > A). Similarly, Case 39 had leukoencephalopathy in the brain MRI in addition to a homozygote pathogenic variant in the LAMA2 gene (c.7147 C > T). Two cases with white matter abnormalities in the brain MRI (Cases 18 and 20) had variants of uncertain significance in the POMGNT1 gene (c.355-3T > G and c.1489 C > G, respectively).

Table 1 Summary of clinical and molecular findings in patients with LGMD 2 A/R1 (Calpainopathy with autosomal recessive inheritance pattern)

Table 2 Summary of clinical and molecular findings in patients with dystroglycanopathy (LGMD type C, 9 and autosomal recessive inheritance pattern)

Table 3 Summary of clinical and molecular findings in patients with sarcoglycanopathy (LGMD 2 C-2 F/R3-R6) with autosomal recessive inheritance pattern

Table 4 Summary of clinical and molecular findings in patients with dysferlinopathy (LGMD 2B/R2) with autosomal recessive inheritance pattern

Table 5 Summary of clinical and molecular findings in Merosin-deficient patients (Autosomal recessive 23)

Table 6 Summary of clinical and molecular findings in patients with titinopathy (LGMD 2 J or autosomal recessive 10)

Table 7 Summary of clinical and molecular findings in patients with autosomal dominant inheritance pattern (Transportinopathy (LGMD 1 F/D2), laminopathy (LGMD 1B), and LGMD1D, respectively)

Discussion

In the current study, we summarized the results of molecular genetics investigations in patients with different variants in the LGMD-related genes. As expected, the majority of patients were born to consanguineous parents and the mode of inheritance was supposed to be autosomal recessive. Among 49 patients, five patients were compound heterozygote for variants in the LGMD-related genes. While the majority of variants were located in the exonic regions, two intronic variants, namely c.2861 + 1G > A and c.355-3T > G were identified in the DYSF and POMGNT1 genes, respectively. CAPN3 and LAMA2 genes were the genes encompassing the highest frequencies of pathogenic or likely pathogenic variants in this cohort. Pathogenic and likely pathogenic variants were identified in CAPN3 gene in total of 10 cases. The CAPN3 c.2107 C > T variant was identified in two cases in the current study as homozygous variant and classified as likely pathogenic according to ACMG guidelines. This variant is classified as VUS in ClinVar by all laboratories. While ClinVar labels this variant as a VUS, our evidence meets ≥ 2 Moderate (PM1, PM2) + 2 Supporting (PP2, PP3) ACMG criteria, warranting likely pathogenic. Additional supporting evidence is that no other pathogenic variants were found in these patients. In fact, the variant affects a conserved residue in the calpain-3 protease domain, but functional studies are lacking. Homozygosity in consanguineous families supports pathogenicity, but more cases are needed. In addition, while the clinical manifestations are typical for calpainopathy, milder phenotypes can be explained by the presence of a missense variant.

A number of other cases had VUS in other genes, such as DAG1 (Case 11), SGCB (Case 26), and POMGNT1 (Cases 16, 18, 20). Among those with VUS, segregation analysis was only performed in Case 26. Case 11 had c.185 C > T (p.P62L) variant in DAG1. Clinical features were severe developmental delay, seizures, dystonia, and family history of affected siblings. This rare variant has no functional data. Co-occurrence of this variant with SLC20A2 (AD) VUS complicates interpretation. Phenotype overlaps with DAG1-related disorders but is more severe (e.g., brain involvement). Thus, it may represent a severe spectrum or digenic inheritance.

The POMGNT1 c.1532 A > G (p.Y511C) variant in Case 16 was associated with motor delay and seizures, consistent with dystroglycanopathy, but its VUS status highlights gaps in our understanding of genotype-phenotype correlations. Similarly, Cases 18 and 20 had VUS in this gene. All of these variants affect conserved residues or splice sites but lack functional validation. While phenotypes align with POMGNT1-related dystroglycanopathy (e.g., brain-eye-muscle involvement), conflicting classifications in ClinVar complicates the interpretation.

Case 23 had TRAPPC11 c.49 A > G (p.M17V). This variant has low frequency in gnomAD and is predicted to be benign by in silico tools. Patients’ phenotype overlaps with TRAPPC11-related LGMD but includes severe neurological features. Co-occurrence with LMNB2 (AD) VUS may modify the phenotype. It is unclear if variant contributes to muscle pathology or is incidental.

Case 26 had SGCB c.500G > T (p.G167V) variant. The phenotype was consistent with sarcoglycanopathy, but pathogenicity was unconfirmed. This variant is absent from population databases; but in silico predictions conflict (e.g., SIFT: damaging, PolyPhen-2: benign). No functional studies have assessed its impact on sarcoglycan complex stability.

Finally, Case 45 had TNPO3 c.1417G > A (p.V473I) variant which was listed as VUS in ClinVar. In silico predictions are conflicting. TNPO3 variants typically cause adult-onset LGMD1F, but this patient presented earlier. Family segregation studies are needed to confirm pathogenicity. While phenotype matches transportinopathy, variant significance remains uncertain.

These cases emphasize the importance of integrating clinical data with genetic findings to refine variant interpretation and guide patient management.

The high frequency of CAPN3 variants in the current study is in line with the reported frequencies in other populations [2]. Furthermore, CAPN3, LAMA2 and DYSF were the genes comprising the highest frequency of mutations in another investigation in the Iranian LGMD patients [9]. Notably, the c.1469G > A variant in the CAPN3 gene was also reported in another Iranian patient [9]. In addition, molecular investigations in six unrelated Iranian families who presented with progressive muscle weakness, highly suggestive of calpainopathies led to identification of five causative variants in the CAPN3 gene revealed which had not been reported in the Iranian population before including a novel 6 bp deletion (c.795_800delCATTGA) and four previously reported mutations, including c.2257delGinsAA [10]. Notably, we identified a similar mutation in this gene (c.2256_2257insA, p.D753Rfs*12) in our cohort of patients.

Another study among Iranian patients revealed two novel mutations in the DYSF genes, namely c.2419 C > T (p.Gln807*) and c. (1,053 + 1_1,054-1)_(1,397 + 1_1,398‐1)del [11], which were distinct from the identified mutations in the current study. Meanwhile, genetic investigations in another cohort of Iranian patients showed that SGCB, SGCA, SGCG, CAPN3 and ISPD genes encompass the higher rates of mutations [12]. Finally, another molecular investigation in a large cohort of Iranian patients suspicious for LGMDs revealed that the most prevalent form of LGMD was calpainopathy followed by sarcoglycanopathy, particularly beta-sarcoglycanopathy [13]. Moreover, exon 2 deletion in the SGCB gene was reported to be the most frequent mutation in the mentioned study [13]. Notably, the same deletion was also reported by our group in another family with two affected siblings [14]. Evidence of possible founder effects were also reported in families with mutations in SGCB [13] and DYSF [11, 13] genes. Thus, the spectrum of LGMD-related mutations among Iranian patients is quite broad, necessitating further studies to find the putative repeated mutations. While this study identified recurrent mutations in CAPN3 and LAMA2, the relatively small cohort might limit the generalizability of variant frequencies to broader Iranian or global LGMD populations. Larger, multi-center studies are needed to validate these results. Moreover, the assignment of AD inheritance in three cases relied on variant pathogenicity predictions and published evidence, but functional validation such as segregation analysis or in vitro assays was not performed. Uncertainties remain about the mechanistic impact of these variants, particularly DNAJB6 c.947 C > T and TNPO3 c.1417G > A variants. In addition, WES may miss non-coding, deep intronic, or structural variants (e.g., copy-number variations) in LGMD-related genes. Complementary techniques like RNA sequencing or MLPA could improve diagnostic yield. Clinical severity and progression data were not consistently available for all cases, impeding genotype-phenotype correlation assessments. Future studies should integrate detailed phenotyping to elucidate variant-specific disease manifestations. The high frequency of certain variants (particularly in the CAPN3 and LAMA2 genes could reflect founder effects in the Iranian population or selection bias in recruitment. Haplotype analysis would help distinguish recurrent mutations from shared ancestry. While pathogenicity classifications were according to ACMG guidelines, the functional consequences were not experimentally validated. Therefore, functional studies (e.g., protein expression assays for CAPN3 truncations) are suggested to strengthen variant interpretations. Despite these limitations, our findings expand the mutational spectrum of LGMD in Iran and underscore the utility of WES in diagnostic approaches of this group of heterogeneous muscular dystrophies, particularly in the consanguineous families. Addressing these gaps through collaborative studies with longitudinal phenotyping and multi-omics approaches will refine genotype-phenotype associations and therapeutic strategies.