Clinical and Electrophysiological Features in Charcot-Marie-Tooth Disease With Mutations in the NEFL Gene

Background  To date, 13 different neurofilament light-chain polypeptide gene (NEFL) mutations have been identified in 55 patients with Charcot-Marie-Tooth disease (CMT) from 16 families. NEFL mutations were found to be associated with axonal and demyelinating variants of CMT.

Objectives  To describe the clinical features of 11 patients with CMT and NEFL mutations and to explore possible genotype-phenotype correlations.

Design  Standardized neuromuscular and nerve conduction studies were performed, and the coding regions of the peripheral myelin protein 22 (PMP22), myelin protein zero (MPZ), gap junction β-1 protein (GJB1), and NEFL genes were analyzed by direct DNA sequencing.

Setting  Two university hospitals in Austria (referral centers for neuromuscular disorders).

Patients  Eleven patients with CMT and NEFL mutations.

Main Outcome Measure  We genotyped NEFL in all of the patients and healthy relatives and correlated the genotype with the phenotype.

Results  A novel NEFL mutation (p.L93P) was detected in 1 family with 4 affected individuals exhibiting a severe CMT phenotype. Nerve conduction velocities were intermediately slowed to a range of 35 to 39 m/s. In a second family and in a sporadic patient, a p.P8R mutation was identified with intermediate and severe nerve conduction slowing.

Conclusion  The results argue against an obvious genotype-phenotype correlation regarding disease onset, degree of muscle weakness, and nerve conduction slowing caused by NEFL mutations.

Charcot-Marie-Tooth disease (CMT) (Online Mendelian Inheritance in Man #118220), also called hereditary motor and sensory neuropathy, has been subdivided into 2 main groups based on the underlying pathological findings.^1-5 The demyelinating variant, CMT1, is electrophysiologically characterized by reduced motor nerve conduction velocities (NCVs) of the median nerve (NCV < 38 m/s) as well as segmental demyelination, segmental remyelination, and hyperplasia of Schwann cells causing onion bulb formations on histopathological examination. In the axonal variant, CMT2, NCVs are almost normal, but amplitudes of compound motor evoked potentials are reduced and nerve pathological findings show axonal loss and regenerative sprouting.The additional classification of intermediate CMT has been suggested in cases with motor NCVs between 25 and 45 m/s.^6,7

Charcot-Marie-Tooth disease is genetically very heterogeneous, and mutations in 36 genes have been described so far (Inherited Peripheral Neuropathies Mutation Database, http://www.molgen.ua.ac.be/CMTMutations/). A large study of 323 patients with CMT by Jordanova et al⁸ found neurofilament light-chain polypeptide gene (NEFL) mutations in 2% of the cases. Thirteen disease-causing NEFL mutations and clinical features of 55 patients have been described, providing a limited set of data for genotype-phenotype analysis.^8-17

Here we describe the results of clinical, electrophysiological, and genetic studies in 2 Austrian families and 1 single patient exhibiting a severe clinical CMT phenotype and a novel NEFL mutation.

Two Austrian families (family 1 and family 2) and 1 sporadic patient with classical CMT were studied. Neurological examinations and standard nerve conduction studies were performed in 9 individuals (4 affected and 5 at risk) of family 1, in 10 individuals (6 affected and 4 at risk) of family 2, and in the sporadic patient. Written informed consent was obtained from all of the participants according to the Declaration of Helsinki, and the study was approved by the local ethics committee.

Participants' DNA was extracted from peripheral blood samples using an automated extractor according to the manufacturer's protocols (GenoM 48; Qiagen, Vienna, Austria). Mutations in the most common genes (peripheral myelin protein 22 gene [PMP22] duplication and deletion, PMP22 point mutations, and mutations in myelin protein zero [MPZ] and gap junction β-1 protein [GJB1] genes) were excluded first. Intronic primers were designed to flank each of the 4 NEFL (GenBank accession number NM_006158) exons encoding 542 amino acids. Each polymerase chain reaction fragment was bidirectionally sequenced in each index case (ABI 3100 DNA sequencer with BigDye terminator mix and SeqScape Software version 2.0; Applied Biosystems, Vienna, Austria). Eight further individuals from family 1 and 9 individuals from family 2 were investigated for the presence of the respective sequence change identified in the index case. Sequencing of these 2 positions was performed in 200 normal chromosomes derived from a reference set of ethnically matched individuals.

In family 1, 56 individuals were identified (Figure 1). In all of the affected individuals, disease onset was during the second decade of life related to plantar extensor weakness and foot deformity, ie, pes cavus, that resulted in ankle arthrodesis in 2 patients at ages 14 (V:10) and 17 (IV:18) years. Distal lower limb atrophy and weakness progressed over years and involved distal upper limbs at older ages. Two patients became wheelchair bound. Sensory loss predominantly affected the lower limbs, and large fiber function and tendon reflexes were absent in the lower limbs (Table 1).

In family 2, disease onset was before age 15 years in all of the patients except for 1 in whom gait problems were not reported until age 25 years. Symptoms were slowly progressive, resulting in a severe and disabling CMT phenotype. Interestingly, the oldest patient was initially diagnosed with spinocerebellar ataxia, as she presented with markedly dysarthric speech, nystagmus, and ataxia (Table 1).

A severe CMT phenotype without signs of ataxia was observed in the single patient who had no reported family history (Table 1).

Electrophysiological studies (Table 1) of families 1 and 2 demonstrated intermediate motor NCVs and absent median and sural nerve sensory action potentials. Compound motor action potentials were reduced to variable degrees. All of the clinically unaffected individuals who were tested had normal nerve conduction study results. The sporadic patient (patient 33) had severely reduced NCVs and compound motor action potentials.

The index patient of family 1 (IV:25) showed a heterozygous single nucleotide exchange (c.278T>C) resulting in a leucine-to-proline amino acid change at codon 93 (p.L93P) in the coil 1a domain of the NEFL protein. The leucine-93 residue is highly conserved (Figure 2). No other nucleotide changes were detected. All of the affected family members but none of the 5 unaffected family members carried the mutation (Figure 1), which was also absent in 100 controls.

In the affected patients of family 2 as well as in the sporadic patient, we found a heterozygous C-to-G nucleotide change in the first coding exon of NEFL (c.23C>G) resulting in a proline-to-arginine substitution at codon 8 (p.P8R) located in the head domain of the NEFL protein. The p.P8R change was absent in clinically unaffected individuals of family 2. It was absent in the unaffected parents of the sporadic case, suggesting that this mutation arose de novo (Figure 2 and Table 1).

To date, 13 different NEFL mutations have been identified in 55 patients with CMT from 16 families.We describe phenotypic features of 11 further patients with CMT from 3 families with NEFL mutations, and we also describe a novel NEFL mutation, p.L93P. Nerve conduction studies suggested that this mutation resulted in an axonopathy with secondary demyelination in all of the affected individuals, which was also seen in family 2 associated with a p.P8R mutation. Indeed, neuropathological studies in 2 cases demonstrated predominantly axonal atrophy in intermediate CMT, but axonal swelling, paranodal abnormalities, and onion bulb formations have also been observed.^12,15

So far, 22 patients with a mutation involving the proline residue at codon 8^8,10 have been described and at least 5 different mutational events can be inferred, suggesting that codon 8 is a hot spot for NEFL mutations. Of 13 patients with this mutation, 1 was classified as having intermediate CMT and the others were classified as having CMT1, with NCVs ranging from 13 to 39 m/s. However, the classification based on nerve conduction studies alone becomes difficult in cases where the compound motor action potential amplitude is reduced, as in our single patient. Despite this, phenotypic variability can be seen with codon 8 mutations not only with NCVs but also with disease onset, which varied between ages 2 and 25 years.^8,10

The phenotypic variability found for codon 8 mutations can also be observed for mutations at codon 22 in the head domain of the NEFL protein^13-15 or at codon 397 in the coil 2b domain^11,12 (Figure 2).

This indicates that no simple genotype-phenotype correlation exists and suggests the existence of modifiers in NEFL-related CMT. However, it appears that mutations localizing to the NEFL protein head domain can cause more severe motor and sensory NCV slowing than mutations in the coil 2B domain (Figure 2 and Table 2),^8,12 while the intrafamilial phenotype appears to “run true.”

In addition, the phenotypic variability of the NEFL mutation seems to also include possible central nervous system involvement, as 1 patient of family 2 showed signs of cerebellar dysfunction and was initially diagnosed with spinocerebellar ataxia. Of previously described patients with NEFL mutations, 1 was initially diagnosed with Friedreich ataxia¹² and another had a complex phenotype with nystagmus and global developmental delay.⁸ Although the pathophysiological abnormalities underlying central nervous system dysfunction remain unclear, their occurrence is suggested by the widespread presence of neurofilament light-chain polypeptide in the central nervous system. However, concomitant mutations in NEFL and other genes cannot be excluded.¹⁸

Since the manuscript was accepted for publication, 3 additional mutations in the NEFL gene have been reported, highlighting the phenotypic variability in CMT with NEFL mutations.^19,20 The NEFL reference sequence has recently been changed; therefore, the new mutation reported in this article is now described as c. 281T>C, p.L94P.

Correspondence: Wolfgang N. Löscher, MD, PhD, Clinical Department of Neurology, Innsbruck Medical University, Anichstrasse 35, A-6020 Innsbruck, Austria (wolfgang.loescher@i-med.ac.at).

Accepted for Publication: August 9, 2006.

Author Contributions:Study concept and design: Janecke, Wanschitz, and Löscher. Acquisition of data: Miltenberger-Miltenyi, Janecke, Wanschitz, Timmerman, Windpassinger, Auer-Grumbach, and Löscher. Analysis and interpretation of data: Miltenberger-Miltenyi, Janecke, Timmerman, Windpassinger, Auer-Grumbach, and Löscher. Drafting of the manuscript: Miltenberger-Miltenyi, Janecke, Windpassinger, Auer-Grumbach, and Löscher. Critical revision of the manuscript for important intellectual content: Janecke, Wanschitz, Timmerman, Windpassinger, Auer-Grumbach, and Löscher. Obtained funding: Janecke and Timmerman. Administrative, technical, and material support: Miltenberger-Miltenyi, Janecke, Windpassinger, Auer-Grumbach, and Löscher. Study supervision: Wanschitz.

Financial Disclosure: None reported.

Funding/Support: This work was supported in part by the Fund for Scientific Research–Flanders, a special research fund of the University of Antwerp, the Medical Foundation Queen Elisabeth, the Association Belge contre les maladies Neuromusculaires, the Interuniversity Attraction Poles Program P5/19 of the Belgian Federal Science Office, and grant P18470-B5 from the Austrian Science Fund.

Additional Contributions: Els De Vriendt and Veerle Van Gerwen provided technical assistance with the mutation analyses in some patients. We thank the families for their participation in this study.