Charcot-Marie-Tooth (CMT) disease type 2A is known to be associated with mutations of MFN2, which is one of the leading causative genes for axonal CMT [1]. MFN2 encodes mitofusin 2 protein in the outer mitochondrial membrane, which has a crucial role in mitochondrial fusion by initiating the formation of antiparallel dimers between their second heptad repeat (HR2) domains [2]. Therefore, the loss-of-function of mitofusin 2 results in the failure of mitochondrial fusion and transport [2]. So far, more than 150 mutations in MFN2 have been identified (https://databases.lovd.nl/shared/genes/MFN2). Here, we report a CMT2A case with a novel mutation in MFN2. We believe this report may expand the mutation spectrum of CMT2A.

The proband, a 44-year-old man, was born from non-consanguineous parents. He visited the department of neurology presenting with slowly progressive weakness of the limbs. At the age of 10, he noticed the dropping of feet while walking on a flat surface. At the age of 20, Achilles tendon lengthening was undergone to minimize ankle contracture. At the age of 37, he noticed atrophy on the distal part of bilateral limbs. At the age of 43, he experienced difficulties writing, using chopsticks, and zipping up the jacket. He denied family history.

A neurologic examination at 45 years of age revealed a weakness in the distal part of bilateral arms and legs. Individually, the patient was not able to move the ankle joint, including dorsi/plantar flexion at all. The motor power of flexion/extension in fingers and wrists and knee extension were four as the Medical Research Council (MRC) grade. The MRC grade of knee flexion was three. The power of the shoulder and hip joints remained normal. Strikingly, the atrophy was severe on the lower legs, in which the circumference was 23 cm at the point of 15 cm below the patella (Fig. 1A). Overall deep tendon reflexes were decreased. Vibration sense was disturbed below both ankles. The hammertoes on both feet were revealed, but pes cavus was not evident. The nerve conduction studies (NCS) demonstrated severely decreased amplitude of compound muscle action potential (CMAP) ranging from 0.2 to 4.5 mV in upper extremities and sensory nerve action potentials ranging from 2.0 to 3.6 mV in lower extremities. The CMAPs were not elucidated in the lower extremities. In contrast, the conduction velocity remained more than 30.4 m/s (Table 1). According to the findings of history, examination, and laboratory, this patient was graded as 14 by the CMT neuropathy score (CMTNS). Based on the results of NCS, we proceeded directly to whole-exome sequencing. To identify the pathogenic variant, we uploaded the variant call format file to the MutationDistiller website (https://www.mutationdistiller.org). Once uploaded, we queried “Charcot-Marie-Tooth” in the phenotype inquiry. The MutationDistiller system identified heterozygous variant c.2228G>T (p.Ser743Ile, NM_014874.3, Fig. 1B) of MFN2 in the proband, but is not listed in the Genome Aggregation Database (gnomAD) or the Exome Aggregation Consortium. And this variant was predicted to be “damaging” with a score of 0.018 by SIFT, “deleterious” with a score of -3.18 by PROVEAN, and “possibly damaging” with a score of 0.945 by PolyPhen-2 systems. Also, this variant is categorized as “likely pathogenic” satisfying two moderate and two supporting evidences based on ACMG/AMP guidelines [3].

We recognized axonal CMT based on the electrophysiological findings, and the diagnosis was confirmed by detecting a novel mutation in MFN2 c.2228G>T (p.Ser743Ile). Moreover, we have discovered the pathogenic mutation efficiently through MutationDistiller [4]. Also, different amino acid substitutions at the same position (c.2229T>A, p.Ser743Arg) had been reported to be pathogenic, which can support the pathogenicity of a variant in our study [1].

WES is currently considered to be a powerful tool for genetic analysis. However, the more we perform WES, the more we get tons of variants to be sifted. To determine the pathogenic mutation from the results of WES, we have filtered allele frequency and the score of in-silico analysis such as SIFT, Proven, and Polyphen2 manually. MutationDistiller is a recently introduced online tool for the analysis of WES [4]. It combines MutationTaster’s pathogenic predictions with a phenotype database of the Human Phenotype Ontology and provides a list of potential disease mutations [4]. Therefore, this tool can reduce the burden of identifying possible pathogenic variants.

Human MFN2 protein consists of GTPase, first heptad repeat (HR1), transmembrane, and HR2 domains [2]. HR2 is an essential part of the initiation of mitochondrial fusion by tethering each other. And the dimerization of the GTPase domains is responsible for pulling the mitochondrial membrane [2]. The loss-of-function of a novel variant can lead to the failure of mitochondrial fusion by interferences of tethering.

In terms of phenotype-genotype correlation in CMT2A, the type or location of mutation was not associated with severe or mild phenotype [5,6]. The previous study could not reveal any relationship between the site or type of mutation and the age at onset, clinical severity, or additional features [6,7]. However, early-onset patients seem to present with severe phenotype with higher score of CMTNS [5,7].

In conclusion, this study demonstrated the CMT2A patient with a novel mutation c.2228G>T in MFN2, analyzed by an efficient tool of MutationDistiller.

The authors declare that they do not have any conflicts of interest.