Introduction

Charcot-Marie-Tooth disease (CMT) represents a genetically heterogeneous group of inherited motor and sensory neuropathies characterised by slowly progressive weakness, muscle wasting and sensory loss, primarily affecting distal leg muscles. Current classification of CMT continues to be based on forearm motor nerve conduction velocity (MNCV) that divides CMT into type 1 (demyelinating; MNCV < 38 m/s) and type 2 (axonal; MNCV > 38 m/s) and on the mode of inheritance. However, patients with X-linked CMT (CMT1X) typically have “intermediate” slowing of nerve conduction velocities, which are faster than in most CMT1 patients and slower than in most CMT2 patients (Nicholson and Nash 1993; Kleopa and Scherer 2006).

The X-linked form of CMT (CMTX) is the second most frequent form of CMT (Ionasescu et al. 1995; Nelis et al. 1996; Mersiyanova et al. 2000; Mostacciuolo et al. 2001; Hattori et al. 2003; Casasnovas et al. 2006) and is associated with a large number of mutations in the gap junction beta 1 (GJB1) gene on chromosome Xq13 encoding the gap junction protein Connexin 32 (Cx32) (Bergoffen et al. 1993; Shy et al. 2007). Cx32 is expressed by Schwann cells and oligodendrocytes. This protein belongs to a family of homologous integral membrane proteins that form functional channels allowing rapid transport of ions and small nutrients between coupled cells. In Schwann cells, Cx32 forms intracellular gap junctions between paranodal loops and Schmidt–Lanterman incisures that allow diffusion of small molecules through the myelin sheath (Balice-Gordon et al. 1998). Normal structure of Cx32 reveals the presence of two extra cellular (EC) loops, four transmembrane (TM) regions, one intracellular (IC) loop and two IC ends: carboxyl terminus (C) and amino terminus (N). Cx32 is also expressed in the central nervous system (CNS), in cell bodies and in processes of oligodendrocytes (Scherer et al. 1995). This is of interest because subclinical CNS involvement shown by changes in visual evoked potentials (VEPs) and brainstem auditory evoked potentials (BAEPs) has been described (Nicholson and Corbett 1996; Bahr et al. 1999; Seeman et al. 2001; Taylor et al. 2003). The hypothesis is that GJB1 mutations lead to a loss of normal cellular communication, which in turn may lead to myelinating Schwann cell dysfunction and peripheral neuropathy.

The clinical manifestations of CMTX vary. Onset may be congenital or delayed until the third decade of life (Ionasescu et al. 1996). Muscle weakness and wasting of the hand muscles may be severe. Clinical electrophysiological abnormalities are well characterised. Nerve conduction velocities are reduced, but the reductions are less severe than those seen in other forms of demyelinating CMT (30–40 m/s in males; 30–50 m/s in females) (Nicholson and Nash 1993; Kleopa and Scherer 2006). Men usually show an earlier onset than women and are more severely affected. Manifestations in female heterozygotes are variable, probably because of random X-chromosome inactivation (Lyonisation), as directly demonstrated in mice (Scherer et al. 1998). In some cases, involvement of female hemizygotes is severe, but overall, they are affected to a degree that lies between normal and affected male subjects (Nicholson and Nash 1993; Gutierrez et al. 2000; Lewis and Shy 1999; Lewis et al. 2000).

More than 294 different mutations in GJB1 have been reported in CMTX patients, which affect both the 5′-untranslated region (UTR) region as well as the coding region of GJB1. These include missense, frameshift, deletion and nonsense mutations. However, missense mutations predominate by far (Inherited Peripheral Neuropathies Mutation Database, http://www.molgen.ua.ac.be/CMTMutations).

To evaluate the frequency of mutations of the GJB1 gene in a series of Italian patients, we tested 114 patients affected by sensory and motor peripheral neuropathy with a possible X-linked transmission or sporadic pattern and absence of the 17p11.2 duplication and CMT2 most commonly associated mutations.

Subjects and methods

Between 1997 and 2006, 684 consecutive blood samples from patients with peripheral motor and sensory neuropathy were seen at our Service of Medical Genetics for molecular diagnosis of CMT. The clinical phenotype was retrospectively defined based on the clinical and electrophysiological data and, when available, on sural nerve biopsy evaluation, reported by highly specialised Italian neurological centres. A CMT1 was diagnosed when the MNCV, recorded from median/ulnar nerves, was <38 m/s, whereas CMT2 was diagnosed when upper-limb MNCV was >38 m/s with reduced compound muscle action potential (CMAP).

The presence of the 17p11.2 duplication was previously ruled out in 359 patients with possible CMT1 by means of pulse-field gel electrophoresis analysis and microsatellite analysis of seven polymorphic markers (http://www.molgen.ua.ac.be/CMTMutations; Latour et al. 2001). One hundred and fifty-eight patients (158/359, 44%) had 17p11.2 duplication. Mutational analysis of the most common genes involved in axonal neuropathies (mitofusin-2, myelin protein zero, neurofilament light gene and heat-shock proteins) was performed in 325 CMT2 patients. Seventy-six CMT1 patients who did not carry the CMT1A duplication and 38 CMT2 patients without mutations in the above-mentioned genes, with a possible X-linked transmission or sporadic pattern, were analysed for mutations in the GJB1 gene. Two hundred unrelated healthy Italian subjects were used as normal controls. Informed consent was obtained from all subjects included in the study.

Genomic DNA was extracted from peripheral blood samples according to a standard protocol. The coding sequence of GJB1, including exon–intron boundaries and the 5′-UTR region, was amplified by polymerase chain reaction (PCR). PCR products were analysed for mutations by direct sequencing on automated sequencer 3100—Avant Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).

Results

Molecular analysis identified 22 GJB1 mutations in 114 patients affected by sensory-motor peripheral neuropathy. Among patients carrying GJB1 mutations, 20 had a positive family history compatible with X-linked inheritance. In 13 families, mutation occurrence was confirmed in other family members by molecular analysis. Out of 22 patients with mutations, one had a mutation in the 5′-UTR region (−459C > T), 16 had a missense mutation, three carried a frame-shift mutation that leads to a premature stop codon, one had a small deletion and another a small insertion (Table 1). Moreover, the c.490C > T mutation was found in three families, and the c.64C > T was detected in two families. Two different mutations (c.545C > T, c.545C > G) affected the same nucleotide in two families. Eight out of these 22 mutations were not previously described: six point mutations (c.50C > G, c.107T > A, c.545C > T, c.545C  > G, c.548G > C, c.791G > T) and two deletions (c.84delC, c.573_581delCGTCTTCAT). Further sequencing analysis did not identified any of these new mutations in 200 normal controls.

Table 1 Gap junction beta 1 (GJB1) gene mutations in Italian patients with X-linked Charcot-Marie-Tooth (CMT) disease
Full size table

Discussion

Here we report the results of a mutation screening in the GJB1 gene involved in CMTX carried out on 114 patients referred to our centre with a diagnosis of motor and sensory peripheral neuropathy in which a male-to-male inheritance and the 17p11.2 duplication were previously ruled out. Twenty-two mutations have been identified, including eight novel mutations. The EC2 domain was the most affected, as also shown in a large study that described 34 Spanish families with mutations in the GJB1 gene (31.8 vs. 36.8%) (Casasnovas et al. 2006).

In two families, the probands (CMT-37 and CMT-51) were female subjects with severe neuropathy with distal paresis requiring ankle–foot orthesis in the second decade of life. In the CMT-37 patient, the ulnar nerve MNCVs was 35 m/s with normal CMAP and distal latency, whereas in the CMT-51 patient, MNCVs were in the axonal range.

In CMT-7 and CMT-51 pedigrees, in which the same aminoacidic residue was differently mutated (Ser182Cys and Ser182Phe), all were female subjects affected by axonal neuropathy (CMT-7: mother and daughter; CMT-51: mother and two daughters). However, a significant intrafamiliar phenotypic variability, ranging from a severe, early onset neuropathy to a mild, late-onset peripheral nervous disease, was present.

The clinical variability, particularly in female subjects, has already been reported and may be explained by the process of Lyonisation, which leads to a mosaic of wild-type and mutant myelin in heterozygous female subjects (Gutierrez et al. 2000). The likelihood that a female patient with a severe neuropathy could be affected by CMT1X must be therefore taken into account. These findings suggest that GJB1 mutation analysis should be performed in female subjects with axonal neuropathy, regardless the severity of peripheral neuropathy. Furthermore, this gene should be investigated in female subjects with demyelinating neuropathy with possible X-linked inheritance or sporadic cases.

The GJB1 mutation frequency in our series of CMT1 patients was 6.1% (22/359) and increased to 19.3% (22/114) in patients without 17p11.2 duplication and autosomal dominant inheritance, therefore representing the second most frequent cause of CMT. Mostacciuolo et al. (2001) described 12 mutations in the GJB1 gene and reported a similar mutation frequency (20.3%). Four of the mutations identified (Tyr7Cys, Thr8Pro, Arg164Trp and Phe193_Met19insTVF) were already reported in our population (Mostacciuolo et al. 2001; Vazza et al. 2006). Three aminoacidic residues (Arg22, Ser182 and Arg164) were affected in more than one family, confirming that they are hot spots for mutations.

To our knowledge, this is the largest cohort of CMT patients with GJB1 mutations reported in the Italian population. The high frequency of GJB1 mutations and the variability of clinical phenotype suggest that all CMT1 patients, negative for 17p11.2 duplication/deletion and without male-to-male inheritance, regardless of the gender of the proband, should be tested for mutations of the GJB1 gene, as should CMT2 patients with possible X-linked inheritance. A careful collection of the family history is therefore relevant for addressing molecular tests, offering accurate genetic counselling and predicting recurrence risk within the family.