Expert Panel Curation of 31 Genes in Relation to Limb Girdle Muscular Dystrophy

Introduction

Limb girdle muscular dystrophies (LGMDs) comprise a group of disorders characterized by weakness and progressive wasting of proximal limb muscles¹. Although phenotypically similar with regard to pattern of weakness, LGMDs can show variable expressivity, ranging from severe, early onset to mild, late onset forms². Prior to advances in massively parallel sequencing (MPS) methods, LGMD diagnoses were based on linkage studies in large pedigrees, clinical phenotyping, muscle pathology findings and candidate gene screening via Sanger sequencing³. More than 30 genes are currently identified as being implicated in LGMD, with both autosomal dominant (AD) and autosomal recessive (AR) inheritance patterns⁴. LGMDs have been classified into subtypes based on the causative gene and inheritance pattern, and this classification system and nomenclature have undergone revisions over the years⁵. Revised, current nomenclature defines AD (LGMD D1–5) and AR (LGMD R1–28) disease entities. Notably, the LGMD nomenclature includes some genes that are related to a broader phenotypic spectrum. For example, a group of myopathies associated with a reduction in α-dystroglycan glycosylation can range from congenital onset with early lethality to mild , adult-onset weakness. In these cases, the single disease entity for curation may be more appropriately defined as a phenotypic spectrum that encompasses LGMD rather than as pure LGMD. The rationale for ascribing a disease gene to an LGMD subtype has not been systematic and most recently was achieved by expert consensus. This recent effort to reclassify LGMDs removed 10 genes previously deemed LGMD and added an additional 5 genes⁵.

With the advances in MPS technologies, the ability to identify genomic variants has increased tremendously. However, accurate interpretation of the detected genomic variants is crucial for molecular diagnostics. Moreover, accurate variant interpretation is not possible without understanding the clinical validity of the implicated GDR. To this end, the Clinical Genome Resource (ClinGen)⁶, an NIH-funded initiative to build an authoritative central resource to define the clinical relevance of genes and variants for use in precision medicine and research, has developed a semi-quantitative framework to assign clinical validity to gene-disease relationships⁷. The ClinGen Muscular Dystrophies and Myopathies gene curation expert panel (MDM GCEP, formerly Limb girdle muscular dystrophy GCEP) applied this framework to assess the strength of evidence associated with genes asserted as LGMD in the revised LGMD nomenclature. Gene-disease validation is an essential step towards generating targeted testing panels⁸ and establishing variant pathogenicity.

Methods

Results

The ClinGen MDM GCEP evaluated 31 genes implicated in LGMD, leading to 35 GDRs. Of these GDRs, 30 (86%) were classified Definitive, 4 (11%) were classified Moderate and 1 (3%) was classified Limited (Figure 1A). No gene curations reached a Strong, No Known Disease Relationship, Disputed or Refuted classification. Twenty-six genes were curated for a single disease entity, including 14 genes for which the curated disease was a lumped entity, while four genes were split into separate curations for different disease entities, based on the inheritance pattern (see Table 1).

Eighty-six percent of LGMD GDRs reached a Definitive classification, with a majority (25/30) receiving the maximum 12 points for the genetic evidence category (Figure 1A and Table 1). Scores for experimental evidence ranged from 1 – 6 points, and 14/30 Definitive curations obtained all 6 points for the experimental category. Twelve Definitive curations received maximum scores in both genetic and experimental categories. Among the four curations with Moderate classifications, though genetic evidence scores ranged between 4.5 – 7.1 points, experimental evidence was scarce, amounting to 1 – 2 points. In the case of the GDR for AD-CAPN3 (LGMDD4) that had a Limited classification, both genetic and experimental evidence were limited, accumulating less than 2 points for genetic and experimental evidence (see Table 1).

The MDM GCEP re-curated four GDRs that reached a Limited or Moderate classification when they were originally curated at least two years prior. Curation of additional published evidence in the time elapsed, and rescoring of evidence based on the latest version (SOP v9) of the ClinGen gene-disease validity framework, led to the upgrading of the DAG1 curation from Moderate to Definitive and of the TNPO3 and POPDC3 curations from Limited to Moderate classifications. There was no change to the classification of PLEC (Moderate) despite an increase in the genetic evidence scores.

Five genes overlapping Congenital Myopathies and MDM GCEPs (COL6A1, COL6A2, COL6A3, LAMA2, and TTN), were linked to a disease spectrum that included both myopathy and LGMD phenotypes and thus were curated by the two panels in collaboration. The MDM GCEP reviewed preliminary curation data provided by the Congenital Myopathy GCEP and offered suggested modification to the lumping and splitting assessments associated with COL6 and TTN genes. The MDM GCEP provided evidence for LAMA2 and TTN gene curation related to published cases meeting the MDM GCEP phenotype definition of LGMD. Thus, the MDM GCEP is recognized as a secondary contributor to all eight GDR curations.

Discussion

Using the ClinGen gene-disease validity framework, the MDM GCEP (formerly LGMD GCEP) classified 35 GDRs for 31 genes asserted in the ENMC revised LGMD nomenclature, providing the community with quality assessments of the strength of evidence supporting a gene-disease relationship and LGMD as a specific presentation within the disease spectrum. For genes that were asserted to be involved in only a specific LGMD subtype (e.g. LGMDR1), the MDM GCEP chose to curate for the disease entity of “Limb-Girdle Muscular Dystrophy”, and depending on the inheritance pattern, included “autosomal recessive” or “autosomal dominant” as a prefix. For genes implicated in more than a single disease entity, including phenotypes on a spectrum beyond LGMD, more inclusive nomenclature was used to indicate the lumped disease entity, also incorporating the gene name to be more specific and consistent with the recommendations from the ClinGen disease naming advisory committee, e.g., collagen 6-related myopathy, myopathy caused by variation in POMT1.

Four genes (CAPN3, COL6A1, COL6A2, COL6A3) were split based on differing inheritance patterns and curated as two separate GDRs for each gene. For each of these genes, the pathogenic mechanism of AR versus AD was deemed distinct. The mechanism for CAPN3, COL6A AR disease is a loss of function relating to a loss of protein expression or function. However in the case of AD inheritance, the mechanism is dominant negative for the COL6A, resulting in impaired collagen assembly¹¹. Although less clear, the dominant CAPN3 mutations may also behave in dominant-negative manner resulting in decreased proteolytic activity or stability of the wild-type calpain-3 protein¹².

A primary goal of the MDM GCEP was to specify phenotypic criteria for assertion of an LGMD phenotype. For curation purposes, we defined LGMD phenotypic criteria as: 1) progressive weakness in limb musculature (pelvic and shoulder girdle); 2) independent ambulation by the age 2 years; 3) elevated CK of >2 times the upper limit of normal; 4) muscle biopsy with dystrophic or myopathic features. It is important to emphasize that these phenotypic criteria are not an attempt to define the phenotypic spectrum of LGMD disease genes, which often are part of a broader phenotypic spectrum that overlaps to varying degrees with congenital and distal myopathies and other phenotypic patterns of weakness. Moreover, the criteria were not intended to supplant diagnositic approaches that may include other modalities utilized in clinical practice such as electrodiagnostic studies or muscle magnetic resonance imaging. Rather, these criteria were refined by the MDM GCEP to provide a pragmatic, minimal list of core phenotypic criteria necessary to assert a gene-LGMD relationship.

Of the GDRs evaluated for an LGMD disease entity, extensive curation evidence was available for 12 Definitive GDRs (Figure 1B), which allowed scoring the maximum scores for genetic and experimental evidence. CAPN3, which has been reported in the autosomal dominant (AD) inheritance pattern in a few multi-generational pedigrees, was assessed as split curations for LGMD inherited in AD and AR patterns. The AR curation reached a Definitive classification, while the AD curation ended up being Limited due to the paucity of families identified and an unclear pathogenic mechanism.

Variation in multiple genes encoding proteins involved in the addition of O-linked oligosaccharides to α-dystroglycan (CRPPA, FKRP, FKTN, GMPPB, POMGNT1, POMGNT2, POMT1, and POMT2) have been implicated in muscular dystrophy, with the spectrum of phenotypes ranging from a severe muscle disorder also linked to brain and eye anomalies and, in some cases, intellectual disability, to a milder isolated LGMD presentation. The MDM GCEP recommended changes to MONDO classifications in order to reflect this spectrum (Figure 2). Each of these genes was curated by the MDM GCEP for their relationship with myopathies and reached a definitive classification. Importantly, although the LGMD phenotype has been asserted in the literature for variants in the POMGNT1 gene, the reported individual did not meet the LGMD phenotype definition set forth by the GCEP¹³ or the reported variants were considered unlikely to be causative^{14, 15}. DAG1 was curated for the entity “qualitative or quantitative defects of α-dystroglycan”. This curation reached a Definitive classification upon recuration in March 2023. However, there were <4 probands in the curated literature meeting the LGMD phenotype criteria^16–18, and similar to POMGNT1, DAG1 has limited evidence to support a causative relationship with an LGMD phenotype. It is important to note that the availability of more convincing evidence in the future could change this relationship. The MDM GCEP urges the community to publish more cases meeting the LGMD phenotype carrying causative variants in these genes.

AR LGMD due to PLEC variation is typically characterized by early childhood onset of proximal muscle weakness and atrophy, notably without skin involvement¹⁹. In addition to an isolated LGMD phenotype, PLEC has been reported in relation to epidermolysis bullosa simplex with muscular dystrophy (EBS5B). The MDM GCEP distinguished the molecular mechanism of EBS5B as primarily nonsense, out-of-frame insertions or deletions within exon 31 and 32, leading to premature protein termination of all plectin isoforms. In contrast, the underlying mechanism for isolated LGMD without skin involvement appears to be recessive truncating variants specifically in exon 1f that encodes the plectin 1f isoform that is only expressed in skeletal muscle. A homozygous founder variant, a 9-base pair deletion, has been reported in all the probands scored for this curation^{19, 20}. Additional variants in PLEC have been noted in the literature in patients with asserted LGMD^21–24 however, the curation did not include these probands since they harbored missense variants or variants in other transcripts that did not solely impact the plectin 1f isoform and were considered to have an EBS syndrome in addition to muscle weakness.

HNRNPDL reached a Moderate classification in relation to AD LGMD, and all probands and families scored for genetic evidence were heterozygous for variants at the same residue, Asp378²⁵. Additional caution is warranted when interpreting variants at other positions in this gene. TNPO3 was recurated in March 2023, and while this curation was initially classified as Limited, it was upgraded to Moderate upon recuration. Evidence that aided in this upgrade came from a Drosophila model published in 2021²⁶, which showed partial recapitulation of the impaired muscle function typical of patients with AD LGMD due to TNPO3 variation.

The collaboration between the MDM GCEP and the Congenital Myopathies GCEP on five genes (COL6A1, COL6A2, COL6A3, LAMA2, and TTN) led to the MDM GCEP being a secondary contributor on the eight Definitive curations. Genetic analysis in patients with collagen 6-related myopathies revealed both dominant and recessive defects in one of the COL6A1, COL6A2, or COL6A3 genes that affect the final Collagen VI protein. On the basis of clinical as well as genetic findings, Ullrich congenital muscular dystrophy, Bethlem myopathy and LGMD can justifiably be regarded as a continuous spectrum of disease ranging from congenital to late adolescent/early adulthood onset rather than distinct clinical entities²⁷. For each of these curations, the MDM GCEP identified at least 4 independent probands that supported a phenotypic presentation consistent with LGMD.

In conclusion, the majority of GDRs reached a Definitive classification upon assessment by the MDM GCEP. These expert-reviewed assertions on the clinical validity of genes implicated in LGMDs form an invaluable resource for clinicians and laboratory diagnosticians. The Moderate and Limited relationships will be reevaluated by the GCEP in 2–3 years’ time, along with assessing any additional and/or novel genes reported in relation to LGMD. Indeed, since the inception of the MDM GCEP, three new LGMD genes have been published and assigned LGMDR27–29 (JAG2, HMGCR and SNUPN respectively) by OMIM^28–31. Systematic evaluation of GDRs related to LGMD is necessary to standardize the genes included on genetic tests for LGMD and related phenotypes.

Supplementary Material