Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disorder predominantly involving motor neurons leading to paralysis and death within 3–5 years from symptom onset. The pathogenic mechanism leading to motor neuron degeneration is unknown in the majority of cases. Frontotemporal dementia (FTD) is a degenerative disorder of the frontal and anterior temporal lobes.¹ Clinical, pathological and genetic data suggest that ALS and FTD form a spectrum of disease.² Approximately 5% of ALS patients have FTD (ALS–FTD)³ and approximately half of patients with “classical” ALS have subtle frontal and temporal lobe cognitive impairment.⁴ Many FTD cases similarly develop symptoms of motor neuron involvement during the course of their illness⁵ and up to one third of FTD patients without overt motor symptoms have loss of anterior horn cells with characteristic ubiquitin inclusions in surviving motor neurons on autopsy.⁶

Recently, mutations in the progranulin (PGRN) gene have been described as the cause of ubiquitin positive FTD.^7,8 The overlap between ALS and FTD prompted us to screen 272 cases of sporadic ALS, 40 familial ALS cases and 49 patients diagnosed with ALS–FTD for mutations in this gene.

METHODS

Subjects

Genetic studies were approved by local research ethics committees (NIA IRB protocol #2003-081). Diagnosis of ALS was based on the El Escorial diagnostic criteria.⁹ Samples used are outlined in table 1. The samples from the Coriell NINDS DNA repository consisted of 45 Caucasian women and 89 Caucasian men, with an average age of symptom onset of 54.3 years (range 26–81). Of these, 34 (25.4%) had bulbar onset disease, 98 (73.1%) had limb onset weakness and the remaining 2 (1.5%) cases presented with respiratory symptoms. Three sporadic ALS patients were reported to have cognitive changes, but did not have a formal diagnosis of FTD or dementia.

Of the 48 Irish cases, 20 were women and 28 were men. Average age of symptom onset was 58.9 years (range 29–79). Eight (16.7%) patients initially manifested symptoms in bulbar musculature, 38 (79.2%) cases reported limb onset symptoms and the site of onset was unspecified in the remaining 2 (4.1%) cases.

Samples obtained from the brain banks at Johns Hopkins University, Columbia University and the University of Miami, consisted of 26 women and 40 men (missing data = 15), with an average onset of 65.5 years of age (range 27–88, missing data = 16).

Of the 96 Swedish patients, 49 were males and 47 females. Thirty-one had bulbar onset motor neuron disease and 63 had spinal onset motor neuron disease. Two presented with cognitive symptoms. The mean age of onset of first symptom was 56.2 years (range 16–85). Diagnosis of FTD was based on the consensus Manchester–Lund clinical diagnostic criteria,¹⁰ supported by ancillary investigations. Of the 36 Swedish ALS–FTD cases, 30 patients had ALS with FTD, four manifested ALS with language disorders characteristic of the FTD spectrum (ie, semantic dementia or progressive non-fluent aphasia) and two cases had ante mortem diagnoses of “pure” FTD with loss of anterior horn cells evident on autopsy (ie, FTD–motor neuron disease). Both of these patients had siblings or parents with ALS. North American control DNA samples were obtained from the Coriell Institute for Medical Research (n = 159, NDPT002, NDPT006, NDPT009) together with 250 additional North American controls ascertained at the Mayo Clinic in Jacksonville, Florida and Scottsdale, Arizona.

PCR amplification and sequencing of PGRN

DNA was extracted from blood using the Wizard purification kit (Promega Corp., Wisconsin, USA) or from brain tissue using the DNeasy kit (Qiagen, Inc., California, USA). The 12 coding exons of PGRN and at least 30 bp of flanking intronic sequence were PCR amplified using primer pairs listed in appendix 1 (see http://www.jnnp.com/supplemental) and Roche FastStart PCR Master Mix polymerase (Roche Diagnostics Corp., Indiana, USA). Each product was sequenced using forward and reverse primers with Applied Biosystems BigDye terminator v3.1 sequencing chemistry.

RESULTS

Sequence analysis of the PGRN gene identified novel heterozygous genetic variants in a single patient diagnosed with ALS–FTD (2.0%, 1 of 49) and in a single case of limb onset sporadic ALS (0.4%, 1 of 272 sporadic cases). Mutations were not present in the 40 familial ALS cases (see appendix 2 at http://www.jnnp.com/supplemental).

A single nucleotide change, c.578C>A, was identified in an ALS–FTD patient (New York Brain Bank sample T-51) predicting a serine to tyrosine substitution (p.S120Y, RefSeq NM_002087.2). A different nucleotide change, c.764C>T, was identified in a patient diagnosed with sporadic ALS (Coriell ALS repository ND10418) predicting a threonine to methionine substitution at residue 182 (p.T182M). S120Y and T182M were not identified in 818 North American control chromosomes, in public databases of single nucleotide polymorphisms or in any of the other 361 ALS/ALS–FTD samples screened as part of this study, or in 523 previously published FTD samples.^7,8,11 A c.1395 A>C base pair change was present in a single Swedish ALS case. Although this variant was also not found in 818 control chromosomes, it is unlikely to be pathogenic as it does not alter the amino acid at this codon (ie, synonymous mutation, p.Pro392Pro). Autopsy of sample T-51 revealed prominent frontal atrophy with marked neuronal loss and reactive gliosis involving the motor cortex, hypoglossal nucleus and anterior horn cells (see supplementary fig 1 at http://www.jnnp.com/supplemental). Rare motor neurons showed Bunina bodies. Myelin loss was marked within the lateral and ventral corticospinal tracts.

DISCUSSION

We found variants in the PGRN gene in a single case of ALS–FTD and in a single case of limb onset sporadic ALS, but mutations were not found in the other 271 sporadic ALS, 40 familial ALS or in the 48 additional ALS–FTD samples screened as part of this study. These findings suggest that PGRN mutations are not a common cause of motor neuron degeneration, although the data do not exclude the possibility that PGRN mutations may be relevant in other populations/ethnicities. Furthermore, the possibility of genomic insertion/deletion mutations in ALS patients has not been excluded by this study.

Our findings agree with a previous report that failed to find PGRN mutations in 48 ALS patients (29 sporadic cases and 19 familial cases).¹¹ To date, only one other individual with a PGRN mutation (family UBC-17, individual 60) fulfilled the El Escorial criteria for ALS and it appears that mutations in this gene are most commonly associated with a behavioural FTD syndrome with possible parkinsonism features in the later stages of illness.^7,8,11 The paucity of PGRN mutations in ALS patients is surprising given the clinicopathological evidence indicating that these two ubiquitionopathy syndromes overlap. However, available data indicate that the pathogenic mechanisms of ALS and FTD do not overlap completely and each individual FTD or ALS causing gene will display its own phenotype pattern.

The nature of the S120Y and T182M variants suggests that they are not pathogenic. All previously described PGRN mutations are truncating mutations that are thought to lead to nonsense mediated decay of the mutant mRNA and consequently haploinsufficiency due to loss of functional PGRN protein.^7,8,11 The only exception is a c.26C>A point mutation predicting a p.A9D mutation in the signal peptide of the PGRN protein that may lead to incorrect localisation of the PGRN peptide within the neuron.^11,12 Neither the S120Y nor the T182M mutations are within the signalling domain or located within any of the tandem repeats of 12 cysteines, and these residues are not highly conserved across species. Furthermore, it was not possible to demonstrate segregation of the S120Y mutation with disease, as additional DNA samples were not available from this family. However, S120Y and T182M were not found in 818 control chromosomes, indicating that they are not common polymorphisms.

In summary, we identified missense nucleotide variants in the PGRN gene in a single case of FTD associated with motor neuron dysfunction (S120Y) and in a single case of limb onset sporadic ALS (T182M), although the pathogenicity of these variants remains unclear. We conclude that PGRN mutations are not a common cause of ALS phenotypes.