Schematic diagram showing location of variants on the ALS2 gene. Exons are shown as a series of dark gray boxes along a horizontal line and are numbered underneath. The location of each variant is indicated: coding sequence and untranslated region variants are shown above the exons, and intronic variants are shown below. The labels indicating variants not detected in a panel of normal control subjects are enclosed in a pale gray–shaded box. The labels indicating potentially functional variants (coding single nucleotide polymorphisms [SNPs], regulatory SNPs, and intronic SNPs <10 base pairs from a splice site) are shown underlined and in open boxes.
Hand CK, Devon RS, Gros-Louis F, Rochefort D, Khoris J, Meininger V, Bouchard J, Camu W, Hayden MR, Rouleau GA. Mutation Screening of the ALS2 Gene in Sporadic and Familial Amyotrophic Lateral Sclerosis. Arch Neurol. 2003;60(12):1768–1771. doi:10.1001/archneur.60.12.1768
Mutations in the ALS2 gene cause juvenile-onset autosomal recessive amyotrophic lateral sclerosis (ALS) and hereditary spastic paraplegia.
To assess the role of ALS2 among more common forms of ALS.
DNA from 95 unrelated familial, 95 unrelated sporadic, and 11 early-onset ALS patients was screened for mutations in ALS2 by denaturing high-performance liquid chromatography and direct sequencing of polymerase chain reaction–amplified fragments. Each variant identified was also analyzed among control subjects. All 34 exons of ALS2 plus the 5′ and 3′ untranslated region were screened.
We detected 23 novel sequence variants; however, none is disease-associated.
Mutations of ALS2 are not a common cause of ALS.
AMYOTROPHIC LATERAL sclerosis (ALS) is an adult-onset neurodegenerative disorder characterized by the death of motor neurons in the cortex, brainstem, and spinal cord, resulting in progressive paralysis and death, usually within 3 to 5 years of symptom onset.1,2 Most ALS cases are sporadic, without any family history. Approximately 15% to 20% of cases are familial, and within these lie several distinct forms of the disease.
The most common form of familial ALS is adult-onset, autosomal dominant ALS, for which to date 2 loci have been identified. Approximately 10% to 20% of cases of ALS1 are caused by mutations in the superoxide dismutase 1 (SOD1) gene on chromosome 21q3 (Online Mendelian Inheritance in Man [OMIM] 105400). The causative gene for ALS3 (OMIM 606640), which maps to chromosome 18q, is as yet unidentified.4
Recently, 2 groups of investigators reported the genetic basis of a rare autosomal recessive, juvenile-onset form, ALS2 (OMIM 205100).5,6 The ALS2 gene on chromosome 2q encodes alsin, a putative GTPase regulator. ALS2 mutations were originally detected in 2 consanguineous ALS2 families of North African and Middle Eastern origin. Both mutations were homozygous short (1– to 2–base pair [bp]) deletions, predicted to result in a frameshift and truncated protein, supporting a loss of functional mechanism for the disease.
Six additional ALS2 mutations have been described in 2 clinically related conditions, primary lateral sclerosis6 (PLS) (OMIM 606353) and infantile-onset hereditary spastic paraplegia (HSP)7,8 (OMIM 607225). In ALS, upper and lower motor neurons are affected, whereas in PLS and HSP, the neurodegeneration is restricted to the upper motor neurons. In the single PLS family and the 5 HSP families in which ALS2 mutations were detected, the mode of inheritance was autosomal recessive and the disease had juvenile onset. All mutations described were homozygous deletions of 1, 2, or 10 bp, again predicted to result in a truncated protein. The origins of these families were Saudi Arabian, French, Italian, Algerian, and Pakistani, and all but 2 exhibit consanguinity.
ALS2 comprises 34 exons spanning an 83-kilobases region of chromosome 2q33. The full-length alsin protein is 1657 amino acids long, and alternate splicing after exon 4 yields an additional shorter transcript of 396 amino acids. Both forms are expressed in a wide range of tissues, with maximal expression in the brain.6 The protein sequence of alsin shows similarity to that of 3 domains (regulator of chromosome condensation [RCC1], Pleckstrin-Dbl, and VPS9) and to membrane occupation and recognition nexus (MORN) repeat motifs. These 3 domains are characteristic of guanine exchange factors for Ran, Rho, and Rab, respectively. The predicted protein function suggests a role in cell signaling pathways involving these GTPases.
The finding of ALS2 mutations in families with ALS2, PLS, and HSP suggests a broad range of resulting phenotypes or a common pathway in each disorder, implicating the importance of the ALS2 product in the process of neurodegeneration. An important question is whether alsin mutations contribute to more common forms of ALS and have a role in other disease populations. We therefore screened the gene for mutations among familial and sporadic ALS patients of North American and European origin, with the hypothesis that a single dominant mutation may result in a later-onset disease.
Ninety-five unrelated, non-SOD1 familial ALS patients, 95 unrelated sporadic ALS patients, and 11 early-onset ALS patients were examined. All patients were assessed by expert clinicians and gave written informed consent. The familial ALS cohort comprised French (n = 57), French Canadian (n = 18), and North American (n = 20) patients. The sporadic ALS cohort comprised 78 French, 14 French Canadian, and 3 North American patients. The early-onset ALS cohort exhibited symptom onset from ages 13 to 24 years; 10 were French and 1 was French Canadian. An initial group of 93 controls comprised 65 non-ALS French subjects and spouses from ALS families (19 French, 6 French Canadian, and 3 North American). Additional controls comprised non-ALS French, French Canadian, and Dutch subjects.
DNA was extracted from whole blood using standard procedures. Primers were designed using Primer39 and were synthesized by Invitrogen Canada Inc, Burlington, Ontario. Thirty-nine primer pairs were designed from genomic DNA to amplify each exon of ALS2, plus the flanking splice sites. Products were polymerase chain reaction (PCR)–amplified, checked on agarose gels, and then pooled into pairs. Pools were analyzed for sequence variation by high-performance liquid chromatography (Transgenomic WAVE system; Transgenomic Inc, Omaha, Neb). Representative samples in which variants were detected were sequenced to confirm and identify the sequence change.
Genotyping for the 30G>C, 598G>A, and 5544G>A variants was performed by restriction enzyme digestion analysis. To identify the 30G>C variant, amplified exon 1 fragments were digested with MspI; the presence of the variant removes 1 of 4 restriction sites. The 598G>A analysis was performed by MnlI digestion of amplified exon 4 fragments; the presence of the variant removes 1 of 3 restriction sites. Detection of the 5544G>A variant was achieved by RsaI digestion of amplified exon 34; the variant remains uncut.
Reverse transcriptase–PCR was carried out to investigate the possibility of altered splicing as a result of variant IVS7 + 3 A>G (intron 7). First-strand complementary DNA was synthesized using total RNA isolated from human lymphoblastoid cells and primed with random hexamers.10 The sequence of the forward primer (5′-TGCACGGGTGAAAACGAGGACAGT-3′) lies within exon 6, and the reverse primer (5′-AAAATCGGAATGCCCAAGTTGACC-3′) lies within exon 8. The expected amplified product was 320 bp in size.
The frequency of each variant was compared among ALS patients vs controls using the conditional form of the 2-tailed Fisher exact test, assuming independent inheritance of each variant.
Twenty-six sequence variants were detected within ALS2 in ALS patients. Of these, 9 were within the coding sequence, 1 in the 5' untranslated region, and 4 in the 3' untranslated region. The remaining 12 were in the intron immediately flanking an exon; 3 of these were within 10 bp of the splice site and therefore may affect splicing. Twenty-four of the variants were single nucleotide polymorphisms (SNPs), 1 was a deletion of a single thymine, and 1 was an insertion of a single thymine. Of these 26 variants, 3 were already present in the SNP database (available at http://www.ncbi.nlm.nih.gov/SNP/), and the remaining 23 have been submitted with the accession numbers ss 4480563 to 4480588. The details, location, and genomic context of each variant are shown in Table 1 and Figure 1.
Of the 9 variants in the coding sequence, 4 occurred in the first codon position and 3 are predicted to result in an altered amino acid (coding SNPs), all within exon 4. These are SNPs 403A>G (Iso94Val), 598G>A (Glu159Lys), and 1225A>G (Met368Val). Methionine to valine is a conservative change; this residue does not occur within a predicted functional domain and is not conserved in the mouse alsin protein, so it is unlikely that this variant is pathogenic. Isoleucine and valine are both hydrophobic, and although residue 94 resides within an RCC1 domain, this residue is not part of the critical RCC1 motif. The most interesting amino acid change is the glutamic acid to lysine substitution at amino acid 159. This is a nonconservative substitution (a positively charged, polar residue replaces an aliphatic, hydrophobic residue) and occurs at a highly conserved position within the first RCC1 domain. This position forms part of a loop between 2 blades of a 7-bladed propeller structure. The loops are probably essential for the guanine nucleotide exchange activity of RCC1; indeed, there is a histidine residue located only 2 amino acids downstream of the Glu159Lys variant that is believed to be important in the interaction with Ran. Replacement of this charged amino acid with a hydrophobic residue may well affect interaction with target GTPases and thus be pathogenic.12
To test whether any of these variants may be causative mutations or susceptibility factors for ALS, we compared the frequency of each variant among patients vs normal controls. The variant frequencies and results of statistical analysis are shown in Table 1. One variant was signifi cantly increased in ALS patients (IVS30 − 69A>T, P = .002); however, this did not remain statistically significant after correction for multiple testing. This variant is a single nucleotide substitution within a nonconserved intronic sequence, 69 bp from the nearest exon, so it is unlikely that it is a susceptibility factor for ALS.
Five variants were not present at all among the control population. Of these, 2 are potentially functional SNPs: IVS7 + 3A>G, which may affect the intron 7–splice donor site, and 5544G>A in the 3' untranslated region, which may affect transcript regulation. We tested for aberrant splicing of exon 7 in the single ALS patient bearing IVS7 + 3A>G by reverse transcriptase–PCR from a patient-derived lymphoblastoid cell line, but did not detect any abnormalities.
There is as yet no assay for alsin function, so to investigate whether some of these variants may be pathogenic, we tested whether any of the 3 variants that cause amino acid changes cosegregate with disease in families. For variant 1225A>G, the affected sibling of an individual carrying this variant was screened and found to be negative for the variant, indicating that 1225A>G does not segregate with disease in this family. Because this amino acid change is conservative, and the variant is found commonly within the control population (see the "Comment" section), we did not study its pattern of inheritance in other families. Unfortunately, because of the lack of affected relatives in the appropriate families, this analysis was not informative for variants 403A>G or 598G>A.
We identified 23 novel variants within ALS2, a gene recently described as causing autosomal recessive juvenile-onset ALS, PLS, and HSP. Of these variants, 3 are predicted to cause amino acid changes; 1 of the 3 occurs at a highly conserved residue within the RCC1 domain of the alsin protein and is likely to form part of a loop structure that is critical for its guanine nucleotide exchange activity. To assess the involvement of this and the other variants in predisposition to ALS, we compared the allelic frequency of each variant in populations of ALS patients vs controls and tested for cosegregation of the variants with disease status in families.
We did not detect a statistically significant frequency difference for any of the variants and did not observe cosegregation with the disease phenotype. It is likely therefore that these variants represent polymorphisms that do not predispose to ALS in the individuals tested.
The data presented herein suggest that ALS2 mutations are not a common cause of adult-onset ALS among mostly European-derived populations. It is perhaps surprising that we did not detect mutations in an 11-sample cohort of juvenile-onset subjects, although none of these samples came from families with demonstrated autosomal recessive inheritance, which may be a prerequisite for ALS2 mutation. In addition, the age at onset in these samples (13-24 years), although juvenile, is older than in all known cases of ALS2 mutation (1-10 years).5- 8
We conclude that ALS2 mutations are confined to rare, autosomal recessive forms of hereditary neurodegeneration, including a clinical spectrum from HSP to ALS. Nevertheless, further study of the ALS2 gene should provide invaluable insight into the pathogenesis of ALS and other neurodegenerative diseases.
Corresponding author and reprints: Guy A. Rouleau, MD, PhD, Montréal General Hospital Research Institute, 1650 Cedar Ave, Room L7-224, Montréal, Quebec, H3G 1A4, Canada (e-mail: Guy.Rouleau@mcgill.ca).
Accepted for publication July 21, 2003.
Author contributions: Study concept and design (Drs Hand, Devon, Hayden, and Rouleau, and Mr Gros-Louis); acquisition of data (Drs Hand, Devon, Khoris, Meininger, Bouchard, Camu, and Rouleau, and Messrs Gros-Louis and Rochefort); analysis and interpretation of data (Drs Hand, Devon, Hayden, and Rouleau, and Messrs Gros-Louis and Rochefort); drafting of the manuscript (Drs Hand, Devon, Khoris, Bouchard, Camu, and Rouleau); critical revision of the manuscript for important intellectual content (Drs Hand, Devon, Meininger, Hayden, and Rouleau, and Messrs Gros-Louis and Rochefort); statistical expertise (Dr Devon); obtained funding (Drs Hayden and Rouleau and Mr Gros-Louis); administrative, technical, and material support (Drs Hand, Khoris, and Camu, and Messrs Gros-Louis and Rochefort); study supervision (Drs Hand, Devon, Bouchard, Hayden, and Rouleau).
This study was supported by the Muscular Dystrophy Association, ALS Association, Association pour la Recherche sur la Sclérose Latérale Amyotrophique, and Association Francaise contre les Myopathies. It was also supported by the Canadian Institute for Health Research (Dr Rouleau and Mr Gros-Louis). Dr Devon is the recipient of Wellcome International Prize Travelling Research Fellowship 060161.
We thank the families and collecting physicians for their cooperation. We acknowledge the technical support of Sandra Laurent.