METHODS

Patients

Figure 1 shows the family pedigree. The probands–three affected sisters (IV-12, IV-13, and IV-14)–were referred to the Human Genome Research Center at the Department of Biology, University of São Paulo, with a diagnosis of motor neurone disease. Extended pedigree analysis showed 26 members were affected (10 men and 16 women); 15 of these already were deceased. Family members reported that patients III-26, IV-10, IV-12, IV-31, and IV-38 died of respiratory failure, although no postmortem confirmation of amyotrophic lateral sclerosis was done. The mean age of death was 49.8 (SD 8.1) years, and the mean age at onset was 38 (SD 6) years.

The diagnosis of amyotrophic lateral sclerosis was based on El Escorial revised criteria.¹⁵ All studies were performed after informed consent. Demographic and clinical data were taken and followed by a complete clinical and neurological examination. Serum levels of creatine kinase, magnetic resonance imaging, nerve conduction study (motor and sensory), and electromyography were performed in the standard way. Muscle biopsies were taken from the biceps of three affected patients (IV-7, IV-14, and IV-37).

We obtained DNA from the peripheral blood of 25 family members (11 affected and 14 unaffected) according to standard procedures. As the disorder has late onset, we included in the linkage analysis only affected patients, older unaffected siblings, and unrelated unaffected spouses. We performed haplotype reconstruction, on the basis of haplotype analysis of members from generation IV, for three patients (III-12, III-20, and III-22), as well as for member III-19.

Linkage analysis

We excluded all previously known loci for amyotrophic lateral sclerosis. We performed scan analysis with distant markers about 10 cM in length from the ABI Prism Linkage Mapping Set kit (version 2; Applied Biosystems, Foster City, CA, USA). Additional polymorphic markers were included in this analysis to refine the region. The PCR products were analysed in MegaBace 1000 DNA Sequencers (Amersham Biosciences, Little Chalfont, UK). The order of the markers was based on different genetic maps of the Marshfield Medical Research Foundation database and the National Center for Biotechnology Information (NCBI) database. All information about sequence tagged sites was obtained from UniSTS of NCBI and the genome database website. As the allele frequency varies according to the population, we analysed at least 30 chromosomes from normal Brazilian controls. For linkage analysis, we used the MLink program of the Fastlink package (Columbia University, New York, NY, USA) for two point analysis, and we assumed autosomal dominant inheritance with penetrance 1 until age 45 years (as the onset in all affected patients occurs before age 45 years), equal male and female recombination rates, and a gene frequency of 0.0001 for parametric analysis. Multipoint analysis was performed with SimWalk (version 2.86).^16–18

RESULTS

Clinical features

Table 1 shows the affected family members’ clinical and neurological features, as well as the results of complementary examinations.

All patients had lower motor neurone symptoms, with signs in the four limbs, and one also had upper neurone signs. Five patients (IV-23, IV-32, IV-33, IV-34, and IV-37) also had bulbar involvement. Postural tremor, one of the first symptoms to appear, was seen in eight patients, but it was stable during disease progression (IV-1, IV-16, IV-17, IV-23, IV-32, IV-33, IV-34, and IV-37).

The symptom of painful cramps was prominent, long standing, and easily obtained, and had a disabling pattern. As lower motor neurone findings confirmed by electrophysiological studies were the most conspicuous and uniform symptom in all patients, such patients first were classified as having motor neurone disease. The presence of pyramidal signs in one patient (IV-1), who could be classified as having clinically probable, laboratory supported amyotrophic lateral sclerosis, subsequently directed the investigation to the amyotrophic lateral sclerosis/motor neurone disease (AML/MND) group.

Serum creatinine kinase was normal or slightly elevated.

Electromyography in eight patients showed a neurogenic pattern compatible with motor neurone disease. Needle electromyography showed abnormal spontaneous activity such as fasciculations, fibrillations, and positive waves in the tongue and muscles of the upper and lower limbs. On effort, polyphasic and giant motor unit action potentials with reduced recruitment were seen. Sensory and motor nerve conduction studies were normal.

Analyses of muscle biopsies showed a neurogenic pattern, with groups of large and small angulated fibres and fibre type grouping.

Linkage data

We detected a significant linkage at region 20q13.3 with the flanking markers D20S178, D20S196, D20S120, D20S100, D20S102, D20S171, and D20S173. Recombination events in two patients (IV-1 and IV-34), as well as the analysis of 12 additional polymorphic markers (D20S857, D20S1083, D20S839, D20S833, D20S606, D20S183, D20S840, D20S211, D20S1148, D20S430, D20S164, and D20S93), allowed us to locate the candidate gene in a region of about 2.7 Mb between the markers D20S430 and D20S173. According to NCBI MapView, this region contains 17 known genes, one pseudogene, and six predicted genes. Table 2 shows the logarithm of odds (LOD) scores. Two point LOD score analysis showed LOD scores >3 for several markers, with a maximum value of 6.02 at θ = 0.0 for the marker D20S171, while multipoint analysis showed a maximum LOD score of 7.45 for the marker D20S164 (fig 2).

We are currently screening functional candidates genes in the region. Until now, the tubulin beta-1 (TUBB-1), cathepsin Z (CTSZ), and ATP synthase–epsilon subunit (ATP5E) genes were screened through single strand conformation polymorphism followed by direct sequencing of all amplicons, but no pathogenic mutation was identified.

DISCUSSION

We have identified a novel locus for an autosomal dominant late onset ALS/MND at 20q13.33 in a large Brazilian Caucasian family. Recently, a new amyotrophic lateral sclerosis locus was also mapped on chromosome 20.¹⁴ This locus is located in the short arm of chromosome 20 (at 20p13), however; while ours is in the long arm.

The disease seems to affect both sexes equally, with no evidence of clinical anticipation. Clinical onset in affected patients occurs earlier than in familial amyotrophic lateral sclerosis or classic amyotrophic lateral sclerosis (mean age 38 (SD 6) years v 46 years and 56 years, respectively). In a previous study of Brazilian patients with sporadic amyotrophic lateral sclerosis, the mean age of onset was 52 years, although the first symptoms presented before age 40 years in 18.1% of patients.¹⁹ The progression of the disease in patients from the present family, however, was slower than in classic amyotrophic lateral sclerosis. In addition, other signs, such as postural tremor and disabling cramps, are atypical, although postural tremor because of increased physiological tremor may be seen in patients with weakness.^20,21 On the other hand, the presence of atypical signs has prompted us to search for a differential diagnosis from other causes of motor neurone disease associated with postural tremor, such as extrapyramidal degeneration, distal hereditary motor neuropathy, adult onset spinal muscular atrophy, or even an occasional association of motor neurone disease and essential tremor.

Sporadic cases of motor neurone disease with tremor or choreiform movements, or both, have been described.²² These patients, however, did not show the classical findings of amyotrophic lateral sclerosis–dementia–parkinsonism, which are typical of Guam complex. In these cases, autopsy studies showed degeneration in the lower motor neurones and neostriato–pallido–nigral system or in the upper and lower motor neurones, pallido–luysio–nigral system, and brainstem tegmentum.²³

The diagnosis of distal hereditary motor neuropathy was excluded because none of our patients had pes cavus or nerve conduction abnormalities.

Many years ago, we reported a Brazilian family with seven members affected by late onset autosomal dominant spinal muscular atrophy and symptoms that overlapped with amyotrophic lateral sclerosis.²⁴ In this family, however, all patients showed rapid progression, with death occurring 2–3 years after onset.

We propose classifying the present family as having late onset amyotrophic lateral sclerosis/motor neurone disease with atypical signs. Recently, a family with atypical autosomal dominant amyotrophic lateral sclerosis with normal life expectancy, absence of bulbar involvement, and symmetrical distal distribution of atrophy and weakness was mapped at 9q34.⁶ The authors suggested that this form and distal hereditary motor neuropathy with pyramidal signs could be the same disorder. We agree with those authors’ comment that it is not easy to classify these neuromuscular conditions on the basis of clinical and neurological signs.

On the other hand, genetic findings and clinical heterogeneity with atypical findings have been common observations, and other families with this intriguing phenotype may exist. Two large Brazilian genealogies with 80 affected members with late onset (mean age 48.8 years) spinal muscular atrophy (MIM 182980) were reported several years ago.²⁵ Patients from these families showed slow loss of muscle strength, with progressive proximal atrophy, hypoactive or absent deep tendon reflexes, and fasciculations. It will be of interest to verify if they have the disorder we describe.

Intrafamilial clinical variability in patients who carry the same pathogenic mutation has been reported, however, for many gene related diseases, such as autosomal recessive limb girdle muscular dystrophies.²⁶ Furthermore, recent studies showed that modifier genes also could alter clinical phenotype. Faster progression of disease has been reported in patients with a mutation in the ciliary neurotrophic factor (CNTF) gene associated with the SOD1 gene compared than in those who lack the CNTF mutation.²⁷ The authors concluded that CNTF is a modifier gene that could modulate the progression of the disease, but this finding was not confirmed by other investigators.²⁸

On the other hand, the vascular endothelial growth factor (VEGF) gene is considered a risk factor for sporadic amyotrophic lateral sclerosis. Patients homozygous for −2578A/−1154A/−634G or −2578A/−1154G/−634G in the VEGF promoter/leader sequence were shown to be at 1.8 times greater risk of amyotrophic lateral sclerosis.²⁹

Conclusion

We mapped a novel gene for an ALS/MND variant at 20q13.33 and are currently screening functional candidate genes in the region for mutations. According to the NCBI Map Viewer, the flanking markers D20S430–D20S173 span a region of about 2.7 Mb that contains 17 known genes, six predicted genes, and one pseudogene.

The TUBB-1 gene was considered to be a strong candidate gene, because it is the major component of microtubules and, recently, alterations of the axonal transport and microtubule network have been shown to be potential causes of motor neurodegeneration in amyotrophic lateral sclerosis model mice.³⁰ No mutation was found, however, in this gene or in the CTSZ or ATP5E genes, although the possibility of an alternative splicing or other post-transcriptional mechanism has not been ruled out yet.

Identification of the causative gene will be very important for enhancing our understanding of the underlying pathological mechanisms in this heterogeneous group of disorders and hopefully will contribute to the opening of new avenues for future therapies.