FUS-associated ALS in Taiwan: genetic spectrum, clinical features, and a founder haplotype of p.H517D

Abstract

Objective

To characterize the genetic spectrum and clinical features of FUS-associated amyotrophic lateral sclerosis (ALS) in a Taiwanese cohort and to investigate whether the recurrent p.H517D variant represents a founder mutation.

Methods

All coding exons and flanking intronic regions of FUS were analyzed by Sanger sequencing in 650 unrelated Taiwanese patients with ALS. Clinical characteristics of patients carrying FUS variants were evaluated. Haplotype analysis using polymorphic microsatellite markers flanking FUS was performed to assess a potential founder effect of the p.H517D variant.

Results

Eight distinct heterozygous pathogenic FUS variants were identified in 11 probands and five affected relatives, including six missense and two frameshift variants. The most frequent variant was p.H517D, detected in four probands. A novel frameshift variant, p.G499Vfs*30, was identified as a de novo mutation in a juvenile-onset ALS patient. Compared with the non FUS-associated ALS cohort, patients with FUS-associated ALS had a significantly younger mean age at onset (40.1 vs 56.6 years) and more frequent bulbar onset (50% vs 19%). Haplotype analysis suggested a common founder for the p.H517D variant.

Conclusions

FUS mutations accounted for 1.7% of ALS cases in this Taiwanese cohort. The recurrent p.H517D variant appears to represent a population-specific founder mutation. Patients with FUS variants presented with earlier disease onset and heterogeneous clinical phenotypes, and de novo variants contributed to juvenile-onset disease.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00415-026-13929-4.

Introduction

Amyotrophic lateral sclerosis (ALS) is a progressive and fatal neurodegenerative disorder characterized by degeneration of upper and lower motor neurons. ALS typically presents with limb or bulbar weakness and subsequently spreads to other body regions, ultimately leading to respiratory failure and death. The survival from symptom onset is typically two to five years, and only a minor group of patients survive longer than 5 years [1]. Despite substantial advances in understanding disease biology, therapeutic options remain limited, with riluzole and edaravone proving only modest delays in disease progression and no functional improvement [2, 3].

Genetic factors play an important role in ALS pathogenesis. Disease-associated genetic variants are identified in approximately 10%–15% of patients, encompassing both familial and apparently sporadic cases. [1]. Since the identification of SOD1 as the first ALS-associated gene in 1993 [4], more than 50 genes have been implicated, highlighting the marked genetic heterogeneity of ALS [5]. The distribution of ALS-associated mutations varies substantially across populations. Hexanucleotide repeat expansions in C9ORF72 represent the most common genetic cause of ALS in European and North American cohorts [6], whereas SOD1 variants are more prevalent in several Asian populations, including those from China, Japan and Korea [7]. These population differences emphasize the need for region-specific genetic studies to accurately define mutation spectra and to guide diagnostic strategies.

Accurate genetic characterization has become increasingly relevant to ALS care and research. Identification of causative genes informs diagnosis, enables family counseling, and provides entry points for mechanism-based therapies. Recent progress in gene-targeted approaches, exemplified by the approval of antisense oligonucleotide (ASO) therapy for SOD1-associated ALS and ongoing trials targeting FUS [8, 9], further highlights the importance of precise molecular diagnosis in current ALS practice.

Among ALS-associated genes, FUS encodes an RNA/DNA-binding protein involved in RNA processing, transport, and cellular stress responses.[10, 11]. Pathogenic FUS variants were firstly linked to familial ALS in 2009 [12]. Most disease-associated variants cluster within the C-terminal nuclear localization signal (NLS) region, leading to impaired nucleocytoplasmic transport, cytoplasmic mislocalization, and disruption of RNA homeostasis [13]. Mutant FUS can also disrupt stress granule dynamics and promote aberrant protein aggregation [14].

FUS mutations are a rare genetic cause of ALS, accounting for approximately 2.8% of familial ALS and about 0.3% of sporadic ALS in Western populations [15]. A higher prevalence has been reported in Asian cohorts, including 6.4% of familial ALS and 0.9% of sporadic ALS overall [7], with even greater frequencies in China (11.4% of familial and 1.5% of sporadic ALS) [16]. Notably, FUS is the most common causative gene of juvenile-onset ALS mutation in East Asia. [7, 17–19] Clinically, FUS-associated ALS is frequently characterized by early age at onset, frequent bulbar involvement, and rapid progression, and short survival time [15].

Although FUS-associated ALS has been described in several cohorts, the genetic spectrum and clinical characteristics across different populations remain incompletely defined, particularly with regard to recurrent variants and potential founder effects. We therefore investigated the genetic and clinical features of FUS-associated ALS in a Taiwanese cohort. In addition, we performed haplotype analysis of individuals carrying the recurrent p.H517D variant to evaluate evidence for a founder effect.

Methods

Study subjects

A consecutive cohort of 650 unrelated patients with ALS were recruited at Taipei Veterans General Hospital between 2000 and 2024. All participants were of Han Chinese ancestry. Demographic and clinical information, including age at symptom onset, sex, site of onset, neurological findings, and family history, was obtained through medical record review and patient interviews. The cohort comprised 388 men and 262 women (male-to-female ratio, 1.48:1). Among them, 80.2% had spinal-onset ALS and 19.8% presented with bulbar-onset disease. The mean age at symptom onset was 56 ± 13 years. ALS was diagnosed according to the revised El Escorial criteria prior to 2020 [20] and the Gold Coast criteria from 2021 onward [21]. Written informed consent was obtained from all participants. The study has been approved by the Institutional Review Board of Taipei Veterans General Hospital and has therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments.

Genetic analysis of FUS variants

Genomic DNA was extracted from peripheral blood leukocytes. All coding exons and flanking intronic regions of FUS were analyzed by Sanger sequencing. PCR amplification was performed with intronic primers, and sequencing was conducted using the BigDye Terminator v3.1 Cycle Sequencing Kit (Thermo Fisher Scientific, Waltham, MA, USA). Sequencing products were analyzed on an ABI Prism 3700 Genetic Analyzer (Thermo Fisher Scientific). Sequence reads were aligned to the FUS reference transcript (NM_000.4), and all identified variants were confirmed by bidirectional sequencing.

Population allele frequencies of nonsynonymous FUS variants were queried in the Taiwan Biobank database, which comprises whole-genome sequencing data from 1,492 unrelated healthy Taiwanese individuals (https://taiwanview.twbiobank.org.tw/) [22], and the Genome Aggregation Database (gnomAD v4.1.0; https://gnomad.broadinstitute.org), which includes more than 800,000 exome and genome sequences from diverse populations [23].

Haplotype analysis

Haplotype analysis was performed in seven individuals carrying the FUS p.H517D variant. Seven polymorphic microsatellite markers spanning approximately 28 Mb around the FUS locus were genotyped. These markers are D16S3068, D16S3100, D16S3022, D16S753, D16S3080, D16S3136 and D16S415 (Supplementary Table S1). The first three (D16S3068, D16S3100, D16S3022) are located telomeric to FUS on chromosome 16p, D16S753 lies centromeric to the gene, and the remaining markers (D16S3080, D16S3136, D16S415) are located on chromosome 16q.

Statistical analysis

Statistical analyses were performed using SPSS version 22.0 (IBM Corp., Armonk, NY, USA). Because of unequal variances between the two groups, age at symptom onset were compared between patients carrying pathogenic FUS variants and all ALS patients without FUS variants in our cohort using Welch’s T test. Fisher's exact test was used to compare the frequencies of categorical variables (gender and site of symptom onset) between patients with pathogenic FUS variants and those without FUS variants. Similar analyses were also performed between patients with FUS p.H517D and ALS patients without FUS variants. Statistical significance was considered as a two-sided p value < 0.05.

Results

Genetic analysis

Among the 650 unrelated ALS patients, eight distinct heterozygous pathogenic FUS variants were identified in 11 probands (1.7%; 11/650), including six missense and two frameshift variants (Fig. 1). The most frequent variant was p.H517D, identified in four probands, whereas each of the remaining variants was observed in a single proband. All six missense variants and the p.G504Wfs*12 variant have been previously reported as pathogenic [10, 12, 16, 24–26]. The p.G499Vfs*30 variant was novel. It was absent from both gnomAD and Taiwan Biobank database and was confirmed as a de novo variant by parental DNA sequencing, which showed wild-type alleles in both parents. According to American College of Medical Genetics and Genomics (ACMG) guidelines, this variant was classified as pathogenic based on de novo inheritance (PS2), a frameshift predicted to disrupt the C-terminal NLS, a recognized mutational hotspot in FUS- associated ALS (PM1), absence from population databases (PM2), and predicted protein length alteration (PM4) [27].

Fig. 1.

Representative sequencing traces of the eight FUS variants identified in the study. The altered nucleotides and corresponding amino acid changes are highlighted in red

Clinical characteristics

In addition to the probands, another five affected relatives were recruited, giving a total of 16 patients from 11 families with FUS-associated ALS (Table 1). Six of the eleven probands had a family history of ALS and the remaining five were apparently sporadic. The frequency of positive family history was significantly higher in FUS-ALS compared to sporadic ALS. Pedigrees with FUS variants are shown in Fig. 2. Among the 16 patients with FUS-associated ALS, seven (44%) were male and nine (56%) were female. The mean age at onset was 40.1 ± 12.4 years (range, 16–71 years), which was significantly younger than that of non-FUS ALS patients (p < 0.01). Eight patients (50%) presented with bulbar-onset disease. Eight patients presented with both upper and lower motor neuron involvement, and the remaining eight patients presented with lower motor neuron predominant phenotype. The median survival time from symptom onset was 3 years (range, 1–6 years). Three patients were alive at the time of analysis. None of the FUS-ALS patient presented with dementia.

Table 1.

Clinical characteristics of the patients with FUS variants in this study

Pedigree	Individual	Mutation	Sex	Age of onset (yr)	Age of evaluation (yr)	Survival time (yr)	Site of onset	UMN/LMN	Dementia	Family history
X653	II-1	p.G499Vfs*30	F	16	18	5	Bulbar	U/LMN	N	N
P925	III-4	p.G504Wfs*12	M	36	37	2	Bulbar	LMN	N	N
C593	II-1	p.H517D	M	42	43	3	Spinal (U)	U/LMN	N	Y
	II-5		M	32	34	3	Bulbar	LMN	N	Y
H259	II-1	p.H517D	F	39	42	3	Bulbar	U/LMN	N	Y
	II-5		M	35	36	2	Bulbar	LMN	N	Y
M391	II-4	p.H517D	F	35	36	2	Bulbar	U/LMN	N	Y
	II-2		M	40	42	4	Spinal (L)	LMN	N	Y
X121	II-5	p.H517D	M	54	54	(4)	Spinal (U)	U/LMN	N	Y
	III-2		F	26	27	3	Spinal (U)	LMN	N	Y
1A553	II-3	p.H517Q	M	71	73	(2)	Spinal (U)	LMN	N	Y
1A643	III-4	p.Q519E	F	49	58	(11)	Spinal (L)	U/LMN	N	Y
H659	III-3	p.R521C	F	37	39	6	Bulbar	LMN	N	N
7355	II-5	p.R521H	M	44	46	3	Bulbar	U/LMN	N	Y
	II-3		F	55	56	3	Spinal (U)	LMN	N	Y
Q603	II-3	p.P525L	M	34	34	1	Spinal (U)	U/LMN	N	N

yr years, B bulbar, S spinal, U upper limb, L lower limb, UMN upper motor neuron signs, LMN lower motor neuron signs, U/LMN both upper and lower motor neuron signs

Parentheses indicate survival duration to date for patients alive at manuscript writing

Fig. 2.

Pedigrees of the FUS variant carriers and Sanger sequencing traces. A Schematic representation of the amino acid sequence of the C-terminal region of the FUS protein, with the positions of the eight variants identified in this study. The red bar indicates the proline/tyrosine-nuclear localization signal region, and blue bar depicts part of the arginine-glycine-rich domain. B The 16 pedigrees with pathogenic FUS variants in the study. The arrows represent the proband. The filled and open symbols indicate the affected and unaffected individuals. The slashed symbols indicate the deceased subjects

Eight patients from four pedigrees carried the p.H517D variant, including three males and five females. Three patients presented with bulbar-onset disease. The mean age at onset were 37.5 ± 8.2 years (range, 25–54 years), which was not significantly different from that of patients carrying other FUS variants. In contrast, the remaining non-H517D FUS variants were each identified in a single proband. A family history was reported only in the proband carrying the p.R521H variant (pedigree 7355). The two individuals with p.R521H developed symptoms at ages 44 and 55 years, respectively, and both experienced rapidly progressive disease and died within three years of symptom onset. Among the apparently sporadic cases, age at onset ranged from 16 to 71 years. Overall, the clinical phenotypes were heterogeneous with respect to age at onset, site of onset, and disease progression (Table 1). Compared with the ALS patients without FUS variants in our cohort (n = 634), the overall FUS-ALS patients and the patients carrying FUS p.H517D both had significant younger ages at symptom onset (p-value = 0.00008 and 0.0002, respectively). In addition, a significantly higher proportion of FUS-ALS patients exhibited bulbar onset than ALS patients without FUS mutations in our cohort (p-value = 0.006) (Table 2).

Table 2.

Clinical characteristics of ALS patients, including those with pathogenic FUS variants

	All ALS patients	Non FUS-ALS patients	All FUS-ALS patients	Patients with FUS p.H517D
Number	650	634	16	8
Gender (M:F)	388: 262	379:255	9:7	5:3
Age of onset (yr)	56.2 ± 12.8	56.6 ± 12.5	40.1 ± 12.4*	37.5 ± 7.90*
Family history	21(3.2%)	9 (1.4%)	12 (75%)*	8 (100%)*
Site of onset
Bulbar	129 (20%)	122(19%)	8 (50%)*	3 (37.5%)
Upper limb	328 (51%)	321(51%)	6 (37.5%)	4 (50%)
Lower limb	193 (29%)	191(30%)	2 (12.5%)	1 (12.5%)

M male, F female, yr years

^*p < 0.05 when compared to non FUS- ALS patients using Welch’s t-test or Fisher’s exact test

The proband (X653) carrying the novel FUS variant, p.G499Vfs*30, presented with a very early disease onset at 16 years and a rapidly progressive disease course. Initial symptoms included right facial palsy, followed by upper limb weakness, dysarthria, and dysphagia within six months. Lower limb weakness developed one year after onset, resulting in frequent falls, and respiratory function deteriorated by the second year. The patient died at age 21 years due to respiratory failure complicated by recurrent aspiration pneumonia and severe bulbar dysfunction.

Haplotype analysis

Haplotype analysis was performed in 7 individuals from 5 pedigrees carrying the p.H517D variant. Seven microsatellite markers flanking a region of 28 Mb around the FUS gene were genotyped (Fig. 3). All carriers shared a common haplotype between markers D16S3022 and D16S3136, indicating a conserved region of approximately 22.4 Mb and supporting the presence of a shared ancestral haplotype. (Fig. 3).

Fig. 3.

Haplotype analysis using the polymorphic microsatellite markers flanking the FUS gene in pedigrees carrying the FUS p.H517D variants. The squares represent males and the circles represent females. The filled and open symbols indicate the affected and unaffected individuals. The slashed symbols indicate the deceased subjects

Discussion

In this study, we characterized the genetic spectrum and clinical features of FUS-associated ALS in a Taiwanese cohort. Eight distinct pathogenic FUS variants were identified in 11 of the 650 unrelated ALS patients (1.7%), confirming that FUS represents an important genetic contributor to ALS in Taiwan. The recurrent p.H517D variant accounted for 36% (4/11) of FUS-associated ALS probands in our cohort, in contrast to studies from other populations in which p.R521H is most frequently reported [28]. Haplotype analysis demonstrated a shared ancestral segment among p.H517D carriers, providing evidence for a founder haplotype for FUS p.H517D in this population.

The distribution of variants in our cohort aligns with the established mutation spectrum of FUS [28]. All missense variants were located within the C-terminal NLS region, the predominant mutational hotspot in FUS-associated ALS. Disruption of the NLS region impairs nucleocytoplasmic transport of FUS protein and is considered a key pathogenic mechanism [29–31]. The two frameshift variants identified were located in proximity to the NLS and were predicted to alter this critical region. One of these, p.G499Vfs*30, was confirmed to be de novo, further supporting its pathogenicity.

Clinically, patients with FUS variants in our cohort showed a significantly younger age at onset compared with non-FUS ALS, and half presented with bulbar-onset disease. Substantial clinical heterogeneity was observed in age at onset, site of onset, and disease progression. This variability was evident even among individuals carrying the same variants. For example, onset ages in p.H517D carriers ranged from 26 to 54 years. In one p.H517D family (Pedigree* X121, Fig. 2), two obligate relatives remained clinically unaffected despite having affected offspring, suggesting possible reduced penetrance. However, these individuals declined genetic testing, and carrier status could not be confirmed. These findings indicate that, although early onset is common in FUS-associated ALS, the clinical course is not uniformly more aggressive.

The juvenile-onset case with the novel p.G499Vfs*30 variant further expands the mutational spectrum of FUS-associated ALS. This patient presented with early disease onset at 16 years and a rapidly progressive course, a phenotype consistent with reports of other truncating variants located adjacent to the nuclear localization signal. Frameshift variants involving nearby residues, such as p.G496Gfs*31 [32], p.R498Afs*32 [33], and p.D502Tfs*27 [34], have been associated with juvenile or young-adult onset and short survival (Supplementary Table 2). The de novo origin of p.G499Vfs30, together with its absence from population databases and predicted disruption of the C-terminal NLS region, provides strong genetic evidence for pathogenicity. These observations support the concept that alterations affecting the whole NLS region are associated with earlier onset and severe phenotypes, although individual variability remains considerable. Our case further emphasizes that de novo FUS variants should be considered in adolescents and young adults with sporadic ALS.

Our findings have several implications. The identification of a shared haplotype among p.H517D carriers suggests a population-specific founder effect contributing to the FUS variant landscape in Taiwan. In addition, the presence of de novo pathogenic variants highlights the importance of genetic testing in early-onset or apparently sporadic ALS. The marked clinical heterogeneity observed in our cohort further indicates that FUS-associated ALS does not follow a uniform clinical course, emphasizing the need for cautious genotype–phenotype interpretation and individualized counseling.

This study has several limitations. The number of patients with FUS-associated ALS was modest, limiting detailed genotype–phenotype analyses. Survival comparisons with non-FUS ALS were derived from retrospective data at a single center. In addition, haplotype mapping relied on microsatellite markers rather than high-density SNP data, and the shared haplotype interval may therefore represent a minimal estimate of the ancestral segment.

In summary, this study defines the genetic spectrum and clinical characteristics of FUS-associated ALS in a Taiwanese cohort. FUS mutations accounted for 1.7% of ALS cases, with p.H517D emerging as a recurrent variant supported by evidence of a founder haplotype. Patients with FUS variants showed younger onset and heterogeneous clinical courses, and de novo mutations contributed to early-onset disease. These findings expand the understanding of FUS-associated ALS in East Asia and support the role of comprehensive genetic testing in the evaluation of ALS patients.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Study concept and design: Yi-Chung Lee. Acquisition of data: Kang-Yang Jih, Yu-Sheun Tsai, Shih-Yu Fang, Yi-Chu Liao, Yi-Chung Lee. Analysis and interpretation of data: Hou-Ping Sytwu, Kang-Yang Jih, Yu-Sheun Tsai, Yi-Chu Liao, Yi-Chung Lee. Drafting of the manuscript: Hou-Ping Sytwu. Critical revision of the manuscript for important intellectual content: Kang-Yang Jih, Yu-Sheun Tsai, Shih-Yu Fang, Yi-Chu Liao, Yi-Chung Lee. Study supervision: Yi-Chung Lee.

Funding

Open access funding provided by Taipei Veterans General Hospital. This work was supported by grants from National Science and Technology Council, Taiwan (112–2314-B-075–034-MY3, 113–2314-B-075–018-MY3, and 114–2314-B-075–021-MY3), and supported by the "Center for Intelligent Drug Systems and Smart Bio-devices (IDS2B)" and “Brain Research Center”, National Yang Ming Chiao Tung University, from The Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education (MOE) in Taiwan.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Conflicts of interest

The authors have no competing interests to declare that are relevant to the content of this article.