Genetic contribution of FUS to frontotemporal lobar degeneration

Frontotemporal lobar degeneration (FTLD) and amyotrophic lateral sclerosis (ALS) are 2 related, ultimately fatal neurodegenerative disorders. FTLD is a focal dementia syndrome affecting primarily the frontal and temporal lobes of the brain whereas ALS is predominantly a disease of the lower and upper motor neurons. Although distinct identities, accumulating evidence supports the view that both diseases are part of a clinical, pathologic, and genetic spectrum. In up to 50% of patients with ALS, mild disturbances of executive functions can be observed related to frontal lobe dysfunction, and in ∼10% cognitive and behavioral changes are present that meet the criteria of FTLD.^1,2 Conversely, a proportion of patients with FTLD will develop muscle wasting and spasticity in later stages of the disease.³ At neuropathologic examination, over half of FTLD patients' brains show pathologic aggregates of ubiquitinated TAR DNA-binding protein 43 (TDP-43) in cortical neurons of affected regions (FTLD-TDP).^4–6 Similarly, for the majority (∼90%) of patients with ALS, inclusions of TDP-43 were detected in degenerating motor neurons.^5,7 The link between FTLD and ALS is also supported by genetic evidence. Several families have been reported in which affected family members manifest FTLD, ALS, or both, and showed linkage to a region on chromosome 9p13-21.^8–10

Recently a new causal gene was identified for ALS. Mutations in the Fused in Sarcoma gene (FUS) were found segregating with ALS in families linked to chromosome 16. Two independent articles described a total of 15 different FUS mutations in 26 unrelated ALS families (∼4% of familial ALS).^11,12 Twelve were missense mutations, clustering in the C-terminal region of the protein encoded by exons 14 and 15. In addition, 1 missense mutation and a deletion and insertion of 2 glycine (G) residues were identified in the upstream G-rich region of FUS in exons 5 and 6 (figure). The clinical presentation of FUS mutations was classic ALS, with a mean age at onset of 46 years and mean disease duration of 33 months.

The FUS protein shows remarkable functional and pathologic similarities with TDP-43.^13,14 Both FUS and TDP-43 are nuclear proteins involved in RNA and DNA metabolism. The neuropathologic examination of FUS mutation carriers revealed inclusions of FUS protein within the cytoplasm of affected motor neurons. Cellular expression studies further showed subcellular mislocalization and cytoplasmic retention of mutant FUS protein.^11,12 These observations are similar to the aberrant TDP-43 processing seen in TDP-43 proteinopathies.^5,6 Following the identification of TDP-43 as a major pathologic protein in both ALS and FTLD, pathogenic mutations were detected in the encoding TARDBP gene in patients with ALS,^15,16 establishing a primary genetic role for TARDBP in ALS. Whereas the initial mutation screens of TARDBP in FTLD were negative,^17,18 rare TARDBP variations have now also been identified in patients with FTLD-ALS and FTLD in the absence of ALS^19–21 (AD & FTD Mutation Database, http://www.molgen.ua.ac.be/FTDMutations).

The discovery of FUS as an important disease gene in ALS prompted us to investigate a possible role for FUS in FTLD and within the FTLD-ALS spectrum. We set up a mutation screen of FUS in a series of patients with FTLD (n = 122), ALS (n = 47), and concurrent FTLD-ALS (n = 15).

METHODS

RESULTS

We performed a systematic mutation analysis of all FUS exons on genomic DNA of 122 patients with FTLD and 15 patients with FTLD-ALS. In addition, we screened 47 patients with pure ALS. Two patient-specific missense mutations were detected, 1 in a patient with FTLD and 1 in a patient with ALS (figure and table 2). Further, our analyses identified in 2 neurologically healthy control individuals a GG deletion which was previously reported as pathogenic.

• Download figure
• Open in new tab
• Download powerpoint

Figure FUS mutations in amyotrophic lateral sclerosis and frontotemporal lobar degeneration

(A) Schematic representation of reported FUS mutations relative to the gene structure and the protein's functional domains. Mutations identified in the present study are highlighted in bold. QGSY-rich = glutamine, glycine, serine, tyrosine–rich region; G-rich = glycine-rich region; NES = nuclear export signal; RRM = RNA recognition motifs; RGG-rich = arginine, glycine–rich region; Zn F = zinc finger. (B) Sequence chromatograms of FUS mutations identified in a Belgian patient with frontotemporal lobar degeneration (c.760A>G, M254V) and a Belgian patient with amyotrophic lateral sclerosis (c.1562G>A, R521H) and alignment of FUS homologues displaying evolutionary conservation of residues M254 and R521 across species.

DISCUSSION

Recently, a new player was introduced in the FTLD-ALS disease spectrum with the identification of FUS mutations in patients with familial ALS.^11,12 To determine whether this newly identified ALS gene could have a role in FTLD etiology, we performed a genetic mutation screen of FUS in 122 patients with FTLD and 15 patients with FTLD and ALS.

We detected 1 patient with FTLD with a novel FUS missense mutation, M254V, located in exon 6 of FUS. The patient was diagnosed with behavioral variant FTD and further clinical follow-up showed no signs of lower motor neuron disease. In the initial FUS reports, the majority of the missense mutations (12/13) were clustering in the C-terminus of the protein except for 1 in exon 6, R244C. Therefore, the question arises whether these rarer exon 6 variants have biologic relevance to the disease. There is, however, genetic evidence that supports a potential pathogenic role for M254V. First, this mutation was absent in a 638 control individuals (mutant allele frequency <0.078%). Second, multiple alignments of FUS homologues and paralogues showed that the M-residue at codon 254 is strongly conserved throughout evolution. Third, in silico analysis predicted a likely pathogenic effect of M254V on FUS functioning. Finally, mutations in known FTLD genes—MAPT, PGRN, VCP, and CHMP2B—were excluded in the FUS M254V mutation carrier. On the other hand, the family history data that could be collected from this patient were not compelling. One maternal uncle developed dementia at a later age; however, both parents are currently still alive and asymptomatic. DNA of the parents for mutation testing could not be obtained. If pathogenic, this mutation must either have arisen de novo or is not fully penetrant. Reduced penetrance was demonstrated in an ALS family segregating FUS R521G.¹¹ Possibly, some FUS mutations act as a susceptibility factor rather than a fully penetrant mutation in developing neurodegenerative disease, including FTLD.^30,31 Immunohistochemistry will be needed to allow a definite classification of the brain pathology in this patient with FTLD and will provide more insight in the biologic relevance of FUS M254V.

We also included 47 patients with a classic ALS phenotype in the mutational analysis, of which 18 patients had a positive family history, and identified the known R521H mutation in 1 familial ALS patient (1/18 or 5.6%). Familial history was conforming to an autosomal dominant inheritance of the disease. We noted variable ages of onset of 33, 50, and 60 years, indicating that other genetic or environmental modifying factors could be implicated in the disease development. There was no indication of cognitive deficits in any of the patients in this family. In the group of patients with FTLD-ALS (n = 15) we did not find any mutations in FUS, although admittedly the study population was small.

The majority of FUS mutations linked to ALS were missense mutations, except for 2 insertion/deletion mutations in the G-rich region of FUS.¹¹ The authors found a GG deletion in 3 related patients with ALS as well as a GG insertion within the same glycine stretch in 1 patient with familial ALS. The GG insertion and deletion were excluded in 176 control individuals. In our study, however, sequencing of 638 control individuals revealed the GG deletion in 2 healthy individuals without clinical evidence of neurodegenerative disease. These GG-insertions/deletions are located in a G₁₀-stretch of the G-rich region of FUS, a region that is not strongly conserved in homologues of closely related organisms. Moreover, we observed similar insertions/deletions of G-residues in the G-rich domain in 4 other control individuals. Together, our findings suggest that small G-insertions/deletions in this region may well be tolerated.

At this point it is difficult to estimate the exact contribution of FUS in FTLD genetic etiology, as uncertainty still exists about the true pathogenicity of M254V identified here. As more FTLD cohorts will be analyzed, the impact of FUS in FTLD will become clearer. Considering the structural and functional parallel that exists between TDP-43 and FUS,^13,14 it is noteworthy that although TDP-43 inclusions are a major pathologic hallmark in patients with ALS and patients with FTLD,^4–7 dominant mutations in TARDBP are primarily observed in pure ALS at one end of the FTLD-ALS spectrum.^15–18 Recently, a few publications reported that the phenotype of TARDBP mutations may occasionally also encompass FTLD.^19–21 The same might possibly hold for FUS. It will be interesting to see if the fraction of patients with FTLD with ubiquitin-positive but TDP-43-negative brain pathology (denoted FTLD-UPS and about 10%–20% of FTLD-U^32–35) will display similar FUS neuropathology as observed in the FUS-positive patients with ALS. Our series included 2 patients with FTLD-UPS pathology but they were negative for FUS mutations.

ACKNOWLEDGMENT

The authors thank the patients for their cooperation in this study, the personnel of the Genetic Service Facility of the VIB–Department of Molecular Genetics (http://www.vibgeneticservicefacility.be), and the Antwerp Biobank of the Institute Born-Bunge (IBB).

DISCLOSURE

Dr. Van Langenhove, Dr. van der Zee, and Dr. Sleegers report no disclosures. Dr. Engelborghs serves on a scientific advisory board of Janssen-Cilag and UCB; and serves on the editorial advisory boards of Clinical Neurology and Neurosurgery and Current Medical Literature–Neurology. Dr. Vandenberghe serves on the editorial board of Frontiers in Neuroscience and receives research support from GE Healthcare, Wyeth, Pfizer Inc., Eli Lilly and Company, and Medivation, Inc. Dr. Gijselinck, M. Van den Broeck, M. Mattheijssens, K. Peeters, Dr. De Deyn, Dr. Cruts, and Dr. Van Broeckhoven report no disclosures.