The G526R glycyl-tRNA synthetase gene mutation in distal hereditary motor neuropathy type V

Distal hereditary motor neuropathy (dHMN) or distal spinal muscular atrophy (dSMA) is a heterogeneous group of disorders characterized by an almost exclusive degeneration of motor nerve fibers, predominantly in the distal part of the limbs. In 1993, Harding proposed a classification of dHMN based on mode of inheritance, age at onset, topography of weakness, and associated signs and divided these disorders into seven subtypes (dHMN-I to VII).¹ Distal HMN type V (dHMN-V) is an autosomal dominant form with predominant involvement of the hands. Both sporadic and familial cases have been reported with similar phenotypes.^2–6 Pyramidal features were present in some of the families.^3,5,6

A first locus was assigned to chromosome 7p14 in a large Bulgarian family.⁷ In 1996, a related phenotype was described in a North American family that mapped to chromosome 7p but was classified as an axonal form of Charcot–Marie–Tooth (CMT) disease or CMT2 as CMT2D, because sensory loss was common in this family.⁸ An additional North American family with CMT2D was then reported.⁹ Finally, linkage analysis in a large Mongolian family in which some patients presented with CMT2D but others with dHMN-V suggested that the disorders were allelic.¹⁰ In 2003, four disease-associated missense mutations in the glycyl-tRNA synthetase (GARS) gene (OMIM 600287) in the four families previously linked to chromosome 7p and an additional family, classified as dSMA-V, were identified.¹¹ The phenotype of these families was recently reported.¹² The genetic heterogeneity of dHMN-V was demonstrated by the identification of mutations in a second gene, BSCL2 (Berardinelli–Seip congenital lipodystrophy; OMIM 269700), giving rise to a related phenotype.¹³

We screened eight families with dHMN, in which at least one patient had inaugural and predominant weakness in distal upper limbs, for the presence of mutations within the GARS gene. We identified G526R missense mutations in exon 14 of the GARS gene in 16 individuals from three French families of Sephardic Jewish origin and excluded a mutation in the BSCL2 gene in the five remaining families. We report the phenotypic characteristics of patients in the identified families with the G526R GARS gene mutation and present evidence for a founder effect.

Methods.

Results.

Genetic findings.

Eight families were included in this study if at least one member had the clinical and electrophysiologic criteria of dHMN (or dSMA) -V, regardless of the presumptive mode of inheritance. The coding sequence of the GARS gene was sequenced directly in the index cases. No sequence variants were identified in five of the patients, but a c. 2094 A→G nucleotide change leading to the G526R missense mutation was detected in three families (HMN-1054, HMN-523, and HMN-1189) (figure). This mutation has already been reported.¹¹ No mutations in the coding sequence of the BSCL2 gene were found in the five remaining families.

• Download figure
• Open in new tab
• Download powerpoint

Figure. Pedigrees of the three families with the G526R GARS gene mutation. Squares = males; circles = females; black symbols = affected; hatched symbols = affected by history; gray symbols = asymptomatic; asterisks = examined by authors; wt/m = G526R mutation present; wt/wt = G526R mutation absent.

As our families with the G526R mutation in the GARS gene were from Algeria with a common Sephardic ethnic background, a founder effect was suspected. Family members were genotyped for 11 microsatellite markers (D7S628 to D7S2252) on 7p14.3 located very close to the GARS gene. A common haplotype (127-198-219-181-89-274-163-219) for D7S435, D7S2848, D7S2496, D7S2492, AFMa246zc9, AFMb50yd1, D7S2491, and D7S632, spanning 1.56 Mb, was associated with the G526R mutation in the three families.

Discussion.

In this study, we analyzed eight families with dHMN, selected on the basis of an inaugural and predominant involvement of the distal upper limbs in at least one family member, for mutations in the GARS gene. Three of the families, HMN-1054, HMN-533, and HMN-1189 (16 patients), had the same missense G526R mutation, a frequency of 37% in this set of eight pedigrees with a selected phenotype and 50% of the familial cases. This mutation was also shown to cause disease in two patients from an Algerian Sephardic Jewish family.¹¹ Interestingly, our three families were from the same ethnic group. The same 127-198-219-181-89-274-219-163 haplotype for D7S435, D7S2848, D7S2496, D7S2492, AFMa246zc9, AFMb50yd1, D7S2491, and D7S632 was associated with mutation in the three families, suggesting the existence of a common ancestor. However, additional families with the same mutation and the same genetic background should be tested to validate this hypothesis.

Owing to the relatively large number of patients (n = 16) in the three families, we were able to study the phenotype associated with the mutation. Age at onset, which occurred between the second and fourth decades, was quite variable in this group of genetically homogeneous patients. The identification of four asymptomatic mutation carriers, two of whom were already age 49, suggests that the G526R mutation may have incomplete penetrance. Some patients with the G526R mutation may therefore appear to be isolated cases. The site of onset was in the distal upper limbs in most patients, but may also be in distal lower limbs (one patient) or both distal upper and lower limbs (two patients). Thus, when seen individually, patients with disease onset in the feet or both the feet and hands would be diagnosed as dHMN type I or II, according to Harding's criteria.¹ It is therefore important to carefully evaluate the family history and examine several family members to identify the main presenting features and correctly classify the disease.¹⁶ At the time of examination, one-half of the patients had distal deficit in both upper and lower limbs. The same observation was made in four families with dSMA-V and different GARS gene missense mutations, including two patients with the G526R mutation.¹² However, the disease remained restricted to the hands 53 years after onset in one of our patients. The evolution of the disease was slow, as for the other families with GARS gene mutations.¹² Proximal muscles were never involved, even in late stages of the disease, so that the older patients remained ambulant.

Electrophysiologic examination confirmed that the G526R mutation was associated with a purely motor neuropathy in our families. Indeed, SNAPs were normal in both upper and lower limbs, even in older patients. In upper limbs, motor fibers of the median nerve were more severely affected than those of the ulnar nerve, which correlates with the clinical phenotype. CMAPs could not be elicited in the abductor pollicis brevis in six patients, whereas CMAPs in the abductor digiti minimi were normal in three of them and only moderately reduced in one. This has also been reported in other families with dHMN-V¹² as in X-linked CMT and ALS.^17,18 The predominance of signs in this territory cannot be explained by an associated carpal tunnel syndrome, as the amplitudes of median SNAPs were preserved. Thus, selective involvement of motor fibers is probable.

In two families with other GARS mutations, G240R and E71G,^8,10,12 the observation of sensory abnormalities on clinical examination led to the diagnosis of CMT2D (table 3). The phenotype also differs from that of dHMN-V families for motor symptoms. All CMT2D patients, after the disease started in the hands, had a relatively quick progression to peroneal and foot involvement. The authors reported that sensory abnormalities were not related to disease duration, but merely to the severity of the disease.¹² The fact that the E71G mutation was found in a Mongolian kindred in which some patients were diagnosed as having dHMN-V and others as CMT2D shows that the sensory involvement is not exclusively related to the type of mutation. SNAP amplitudes were normal or moderately decreased in the patients with CMT2D, owing to predominant involvement of axons in small- and medium-sized myelinated fibers, as shown on one nerve biopsy, and regenerative clusters were present. Similar but milder changes were observed on nerve biopsies of two patients diagnosed as having dHMN-V, which shows that some subclinical sensory abnormalities could be present in this condition.

Evidence for genetic heterogeneity in dHMN-V was provided by the identification of an Austrian family with a closely related phenotype but no linkage to chromosome 7p14.¹⁶ In this family, onset occurred in the second decade in most patients, but the clinical presentation was quite variable. The majority of patients started with amyotrophy and weakness in the thenar and first dorsal interosseous muscles, the others with foot deformities and difficulty walking. Asymmetric involvement of the hands was noted in most patients. Progression was very slow. Although clinically limited to brisk tendon reflexes and elevated muscle tone in the lower limbs, prolonged central motor conduction times indicated additional involvement of central motor pathways in two-thirds of patients. Heterozygous missense mutations N88S and S90L in BSCL2 gene were identified in this family and in others with either dHMN-V or Silver syndrome phenotypes, characterized by amyotrophy and weakness in the hands and various degrees of spasticity in the legs.¹³ The phenotype associated with the N88S mutation was extended in a study of 90 patients from one large Austrian and two German families. In addition to dHMN-V and Silver syndrome phenotypes, the patients may present with 1) a CMT-like phenotype with weakness and wasting predominating in distal lower limbs, with or without sensory abnormalities; or 2) a hereditary spastic paraplegia phenotype, either pure or complicated by motor or motor and sensory signs in distal legs.¹⁹ Finally, a severe phenotype with spastic paraplegia and distal amyotrophy in the hands and legs was reported in a Belgian family with the S90L mutation.²⁰ (table 3). None of our families had mutations in the coding sequence of the BSCL2 gene.

Because some patients with the G526R mutation, and possibly others, may appear to be isolated cases, the diagnosis of a patient with a recent history of distal amyotrophy and weakness of the upper limbs without sensory signs may be complicated. This condition may evoke ALS, motor neuropathy with persistent conduction blocks, or Hirayama disease. As tendon reflexes were often preserved in our patients as well as in others with a GARS gene mutation and a dHMN phenotype,¹² it may be difficult to exclude early-stage ALS by clinical examination. The presence of a Babinski sign that has sometimes been reported,⁷ but was observed in only one of our patients, may increase the confusion. The differential diagnosis between both diseases is important, given the poor prognosis of ALS. It may also be difficult to distinguish patients with GARS mutations from those with Hirayama disease, as asymmetry is frequent. Asymmetry was found in 4 of our 16 patients. Clinical and electrophysiologic examinations of the parents may help to discern a genetic disease in these patients. The fact that the obligate carrier (HMN-1189-II1) did not have any abnormalities on clinical examination despite obvious signs of denervation in distal limb muscles shows that this strategy may be very effective. In the absence of a clear diagnosis, screening of the GARS gene could be included in investigations when purely motor distal upper limb involvement occurs in adolescence or early adulthood.