Use of SNP array analysis to identify a novel TRIM32 mutation in limb-girdle muscular dystrophy type 2H

Abstract

Molecular diagnosis of monogenic diseases with high genetic heterogeneity is usually challenging. In the case of limb-girdle muscular dystrophy, multiplex Western blot analysis is a very useful initial step, but that often fails to identify the primarily affected protein. We report how homozygosity analysis using a genome-wide SNP array allowed us to solve the diagnostic enigma in a patient with a moderate form of LGMD, born from consanguineous parents. The genome-wide scan performed on the patient’s DNA revealed several regions of homozygosity, that were compared to the location of known LGMD genes. One such region indeed contained the TRIM32 gene. This gene was previously found mutated in families with limb-girdle muscular dystrophy type 2H (LGMD2H), a mild autosomal recessive myopathy described in Hutterite populations and in 4 patients with a diagnosis of sarcotubular myopathy. A single missense mutation was found in all these patients, located in a conserved domain of the C-terminal part of the protein. Another missense mutation affecting the N-terminal part of TRIM32, observed in a single consanguineous Bedouin family, was reported to cause the phenotypically unrelated and genetically heterogeneous Bardet–Biedl syndrome, defining the BBS11 locus. Sequencing of TRIM32 in our patient revealed a distal frameshift mutation, c.1753_1766dup14 (p.Ile590Leu fsX38). Together with two recently reported mutations, this novel mutation confirms that integrity of the C-terminal domain of TRIM32 is necessary for muscle maintenance.

Introduction

Molecular diagnosis of autosomal recessive diseases with extensive non allelic genetic heterogeneity is a challenging task because clinical presentation seldom suggests a particular defective gene. Autosomal recessive limb-girdle muscular dystrophies (LGMD) represent such a case, with 13 genes identified up to now (Table 1) and extensive analysis of these genes fail to identify causative mutations in 25–50% of the cases (depending whether restrictive or wider initial diagnostic criteria are used) [1], [2]. Western blot analysis of muscle biopsies, often performed in a multiplex format [3], or immunohistology allows the direct testing of 6 proteins encoded by LGMD genes (see Table 1) and can thus orient the search for mutations when the protein band is missing or altered in migration or quantity. In addition, testing of anomalies of dystroglycan glycosylation can orient towards LGMD2H (FKRP), LGMD2K (POMT1), or the recently proposed LGMDL (fukutin) and LGM2N (POMT2) [4]. However, this approach can fail to detect alterations caused by missense mutations (especially in the case of calpain-3), or can detect secondary quantitative alterations not linked to mutations in the cognate gene [1]. This approach also requires a muscle biopsy conserved in proper conditions and is thus often not applicable to retrospective diagnosis of deceased patients. Thus exhaustive sequencing of all genes would be warranted in many cases for accurate diagnosis and genetic counseling, but this is very costly and usually not performed in clinical practice.

The frequency of rare autosomal recessive disease is increased in inbred populations and the proportion of patients born from consanguineous parents increases with the decreasing frequency of a given recessive disease. We have taken advantage of this fact and of the availability of genome-wide SNP arrays to identify, using homozygosity mapping, the molecular cause in a patient with slowly progressive LGMD and onset at 25 y of age. We found a homozygous frameshift mutation in TRIM32, a gene associated to an extremely rare form of LGMD, LGMD2H (OMIM 254110). At the time of this observation, only a single missense mutation causing LGMD and arising on a single haplotype had been reported in this gene, in Hutterite families and in two German patients with a diagnosis of sarcotubular myopathy [5], [6], while a missense mutation in the same gene had been reported as causing Bardet–Biedl syndrome (BBS11) in a single consanguineous family [7]. Recently, Saccone et al. [2] reported two additional homozygous TRIM32 mutations in two LGMD probands out of 310 tested.