Differential diagnosis of vacuolar myopathies in the NGS era.

Altered autophagy accompanied by abnormal autophagic (rimmed) vacuoles detectable by light and electron microscopy is a common denominator of many familial and sporadic non-inflammatory muscle diseases. Even in the era of next generation sequencing (NGS), late-onset vacuolar myopathies remain a diagnostic challenge. We identified 32 adult vacuolar myopathy patients from 30 unrelated families, studied their clinical, histopathological and ultrastructural characteristics and performed genetic testing in index patients and relatives using Sanger sequencing and NGS including whole exome sequencing (WES). We established a molecular genetic diagnosis in 17 patients. Pathogenic mutations were found in genes typically linked to vacuolar myopathy (GNE, LDB3/ZASP, MYOT, DES and GAA), but also in genes not regularly associated with severely altered autophagy (FKRP, DYSF, CAV3, COL6A2, GYG1 and TRIM32) and in the digenic facioscapulohumeral muscular dystrophy 2. Characteristic histopathological features including distinct patterns of myofibrillar disarray and evidence of exocytosis proved to be helpful to distinguish causes of vacuolar myopathies. Biopsy validated the pathogenicity of the novel mutations p.(Phe55*) and p.(Arg216*) in GYG1 and of the p.(Leu156Pro) TRIM32 mutation combined with compound heterozygous deletion of exon 2 of TRIM32 and expanded the phenotype of Ala93Thr-caveolinopathy and of limb-girdle muscular dystrophy 2i caused by FKRP mutation. In 15 patients no causal variants were detected by Sanger sequencing and NGS panel analysis. In 12 of these cases, WES was performed, but did not yield any definite mutation or likely candidate gene. In one of these patients with a family history of muscle weakness, the vacuolar myopathy was eventually linked to chloroquine therapy. Our study illustrates the wide phenotypic and genotypic heterogeneity of vacuolar myopathies and validates the role of histopathology in assessing the pathogenicity of novel mutations detected by NGS. In a sizable portion of vacuolar myopathy cases, it remains to be shown whether the cause is hereditary or degenerative.

Linkage analysis and whole exome sequencing reveals AHNAK2 as a novel genetic cause for autosomal recessive CMT in a Malaysian family.

Charcot-Marie-Tooth (CMT) disease is a form of inherited peripheral neuropathy that affects motor and sensory neurons. To identify the causative gene in a consanguineous family with autosomal recessive CMT (AR-CMT), we employed a combination of linkage analysis and whole exome sequencing. After excluding known AR-CMT genes, genome-wide linkage analysis mapped the disease locus to a 7.48-Mb interval on chromosome 14q32.11-q32.33, flanked by the markers rs2124843 and rs4983409. Whole exome sequencing identified two non-synonymous variants (p.T40P and p.H915Y) in the AHNAK2 gene that segregated with the disease in the family. Pathogenic predictions indicated that p.T40P is the likely causative allele. Analysis of AHNAK2 expression in the AR-CMT patient fibroblasts showed significantly reduced mRNA and protein levels. AHNAK2 binds directly to periaxin which is encoded by the PRX gene, and PRX mutations are associated with another form of AR-CMT (CMT4F). The altered expression of mutant AHNAK2 may disrupt the AHNAK2-PRX interaction in which one of its known functions is to regulate myelination.

Improved inherited peripheral neuropathy genetic diagnosis by whole-exome sequencing.

Inherited peripheral neuropathies (IPNs) are a group of related diseases primarily affecting the peripheral motor and sensory neurons. They include the hereditary sensory neuropathies (HSN), hereditary motor neuropathies (HMN), and Charcot-Marie-Tooth disease (CMT). Using whole-exome sequencing (WES) to achieve a genetic diagnosis is particularly suited to IPNs, where over 80 genes are involved with weak genotype-phenotype correlations beyond the most common genes. We performed WES for 110 index patients with IPN where the genetic cause was undetermined after previous screening for mutations in common genes selected by phenotype and mode of inheritance. We identified 41 missense sequence variants in the known IPN genes in our cohort of 110 index patients. Nine variants (8%), identified in the genes MFN2, GJB1, BSCL2, and SETX, are previously reported mutations and considered to be pathogenic in these families. Twelve novel variants (11%) in the genes NEFL, TRPV4, KIF1B, BICD2, and SETX are implicated in the disease but require further evidence of pathogenicity. The remaining 20 variants were confirmed as polymorphisms (not causing the disease) and are detailed here to help interpret sequence variants identified in other family studies. Validation using segregation, normal controls, and bioinformatics tools was valuable as supporting evidence for sequence variants implicated in disease. In addition, we identified one SETX sequence variant (c.7640T>C), previously reported as a putative mutation, which we have confirmed as a nonpathogenic rare polymorphism. This study highlights the advantage of using WES for genetic diagnosis in highly heterogeneous diseases such as IPNs and has been particularly powerful in this cohort where genetic diagnosis could not be achieved due to phenotype and mode of inheritance not being previously obvious. However, first tier testing for common genes in clinically well-defined cases remains important and will account for most positive results.

LRRK2 G2385R and R1628P mutations are associated with an increased risk of Parkinson's disease in the Malaysian population.

The LRRK2 gene has been associated with both familial and sporadic forms of Parkinson's disease (PD). The G2019S variant is commonly found in North African Arab and Caucasian PD patients, but this locus is monomorphic in Asians. The G2385R and R1628P variants are associated with a higher risk of developing PD in certain Asian populations but have not been studied in the Malaysian population. Therefore, we screened the G2385R and R1628P variants in 1,202 Malaysian subjects consisting of 695 cases and 507 controls. The G2385R and R1628P variants were associated with a 2.2-fold (P = 0.019) and 1.2-fold (P = 0.054) increased risk of PD, respectively. Our data concur with other reported findings in Chinese, Taiwanese, Singaporean, and Korean studies.

Analysis of dynein intermediate chains, light intermediate chains and light chains in a cohort of hereditary peripheral neuropathies.

The cytoplasmic dynein heavy chain (DYNC1H1) gene has been increasingly associated with neurodegenerative disorders including axonal Charcot-Marie-Tooth disease (CMT2), intellectual disability and malformations of cortical development. In addition, evidence from mouse models (Loa, catabolite repressor-activator (Cra) and Sprawling (Swl)) has shown that mutations in Dync1h1 cause a range of neurodegenerative phenotypes with motor and sensory neuron involvement. In this current study, we examined the possible contribution of other cytoplasmic dynein subunits that bind to DYNC1H1 as a cause of inherited peripheral neuropathy. We focused on screening the cytoplasmic dynein intermediate, light intermediate and light chain genes in a cohort of families with inherited peripheral neuropathies. Nine genes were screened and ten variants were detected, but none was identified as pathogenic, indicating that cytoplasmic dynein intermediate, light intermediate and light chains are not a cause of neuropathy in our cohort.

X-linked Charcot-Marie-Tooth disease predominates in a cohort of multiethnic Malaysian patients.

INTRODUCTION: Data regarding Charcot-Marie-Tooth disease is lacking in Southeast Asian populations. We investigated the frequency of the common genetic mutations in a multiethnic Malaysian cohort. METHODS: Patients with features of Charcot-Marie-Tooth disease or hereditary liability to pressure palsies were investigated for PMP22 duplication, deletion, and point mutations and GJB1, MPZ, and MFN2 point mutations. RESULTS: Over a period of 3 years, we identified 25 index patients. A genetic diagnosis was reached in 60%. The most common were point mutations in GJB1, accounting for X-linked Charcot-Marie-Tooth disease (24% of the total patient population), followed by PMP22 duplication causing Charcot-Marie-Tooth disease type 1A (20%). We also discovered 2 novel GJB1 mutations, c.521C>T (Proline174Leucine) and c.220G>A (Valine74Methionine). CONCLUSIONS: X-linked Charcot-Marie-Tooth disease was found to predominate in our patient cohort. We also found a better phenotype/genotype correlation when applying a more recently recommended genetic approach to Charcot-Marie-Tooth disease.