Article Text

PDF

Research paper
Burden of rare variants in causative genes for amyotrophic lateral sclerosis (ALS) accelerates age at onset of ALS
  1. Hiroya Naruse1,
  2. Hiroyuki Ishiura1,
  3. Jun Mitsui1,2,
  4. Yuji Takahashi3,
  5. Takashi Matsukawa1,2,
  6. Masaki Tanaka1,
  7. Koichiro Doi4,
  8. Jun Yoshimura4,
  9. Shinichi Morishita4,
  10. Jun Goto5,
  11. Tatsushi Toda1,
  12. Shoji Tsuji2,6
  1. 1 Department of Neurology, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan
  2. 2 Department of Molecular Neurology, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan
  3. 3 Department of Neurology, National Center Hospital, National Center of Neurology and Psychiatry, Tokyo, Japan
  4. 4 Department of Computational Biology and Medical Sciences, Graduate School of Frontier Sciences, The University of Tokyo, Tokyo, Japan
  5. 5 Department of Neurology, International University of Health and Welfare Mita Hospital, Tokyo, Japan
  6. 6 Institute of Medical Genomics, International University of Health and Welfare, Chiba, Japan
  1. Correspondence to Dr Shoji Tsuji, Department of Molecular Neurology, Graduate School of Medicine, The University of Tokyo, Tokyo 113-8655, Japan; tsuji{at}m.u-tokyo.ac.jp

Abstract

Objectives To evaluate the burden of rare variants in the causative genes for amyotrophic lateral sclerosis (ALS) on the age at onset of ALS in a Japanese case series.

Methods We conducted whole-exome sequencing analysis of 89 families with familial ALS (FALS) and 410 patients with sporadic ALS (SALS) to identify known pathogenic mutations or rare functionally predicted deleterious variants in the causative genes for ALS. Rare variants (minor allele frequency <1%) with scaled Combined Annotation-Dependent Depletion score >20 were defined as rare functionally predicted deleterious variants. The patients with ALS were classified on the basis of the number of pathogenic and/or rare functionally predicted deleterious variants, and the age at onset was compared among the classified groups.

Results Whole-exome sequencing analysis revealed known pathogenic mutations or rare functionally predicted deleterious variants in causative genes for ALS in 56 families with FALS (62.9%) and 87 patients with SALS (21.2%). Such variants in multiple genes were identified in seven probands with FALS and eight patients with SALS. The ages at onset in the patients with ALS with multiple variants were significantly earlier than those in other patients with ALS. Even when the patients with known pathogenic mutations were excluded, a significantly earlier onset of the disease was still observed in patients with multiple rare functionally predicted deleterious variants.

Conclusions A substantial number of patients carried rare variants in multiple genes, and the burden of rare variants in the known causative genes for ALS affects the age at onset in the Japanese ALS series.

Statistics from Altmetric.com

Introduction

Amyotrophic lateral sclerosis (ALS) is a devastating neurological disorder characterised by progressive degeneration of upper and lower motor neurons. Only 5%–10% of ALS cases are familial (FALS), whereas the majority of cases are sporadic (SALS).1 Currently, more than 20 genes have been identified to be causative genes for FALS,2 and a report stated that mutations in causative genes for ALS, including C9ORF72 repeat expansion, collectively account for approximately 68% of FALS and 11% of SALS cases in populations of European ancestry.3 In our recent study, on the basis of whole-exome sequencing analysis, pathogenic mutations were identified in 41 (60.3%) of the 68 FALS families.4 In our Japanese FALS series, mutations in SOD1 are the most frequent, occurring in 39.7% of the families. This study also revealed that the frequency of patients carrying the repeat expansion mutations in C9ORF72 in the Japanese population is considerably lower than those in the European populations,5 with the exception of the concentration of the C9ORF72 repeat expansion in the patients with ALS in the Kii Peninsula.6

The molecular basis of SALS, however, was less fully elucidated than that of FALS. Of note, repeat expansions in C9ORF72 have been found in 7% of patients with SALS in the European population.3 Moreover, this repeat expansion was found in 21.1% of patients with SALS in the Finnish population.7 Besides expansion mutations in C9ORF72, pathogenic mutations in the ALS causative genes were reported to collectively account for 4.4% and 3.0% of patients with SALS from Chinese or Japanese ancestry, respectively.8–10Thus, molecular basis underlying ALS has not been revealed yet in large populations of patients with SALS. Therefore, it is conceivable that a combination of multiple variants in these genes may contribute to the disease expression of ALS even though individual mutations are not sufficient to cause the disease. In our recent comprehensive genome analysis using next-generation sequencers, we observed that individual patients with ALS carry a considerable number of variants in these genes. From these findings, we hypothesised that a combination of multiple rare variants including known pathogenic mutations for FALS and rare variants in these genes predicted to be functionally deleterious may contribute to the disease expression including the age at onset.

On the basis of this hypothesis, we investigated the possibility that the age at onset of ALS is accelerated in the presence of a combination of multiple variants, including known pathogenic mutations for FALS and the rare functionally predicted deleterious variants in these causative genes for ALS found in our FALS and SALS case series. In this study, we found that the burden of rare variants in the known causative genes for ALS accelerates the age at onset in the Japanese ALS series.

Materials and methods

Subjects: patients with FALS and SALS

All of the genomic DNA samples were obtained from the participants with their written informed consent. Mutational analyses of our case series were previously reported partially, which include microarray-based resequencing,11 Sanger sequencing,12–17 repeat-primed PCR analysis6 and whole-exome sequencing analysis4 in our laboratory. Initially, 509 patients with ALS and 1183 control subjects were recruited to the study. To avoid inclusion of related individuals, we checked all pairwise identity by descent (IBD) values by analysing the whole-exome data using PLINK V.1.90. To calculate IBD values, 6828 single-nucleotide polymorphisms (SNPs) were selected, after removing SNPs with minor allele frequency (MAF) <0.1, average read depth <40, Hardy-Weinberg Equilibrium (HWE) p value <0.05 or linkage disequilibrium metrics r2 >0.2. We identified pairs with PIHAT values greater than 0.2. Based on these findings, we excluded five patients with ALS and 18 control subjects from the current study. To exclude subjects with non-membership of a homogeneous population, we conducted principal components analysis (PCA) using PLINK V.1.90. We extracted 862 938 SNPs that fulfil the conditions of MAF >0.05 and HWE p value >0.05 from the genotyping data of HapMap3 datasets including the 11 global populations (Release 3, NCBI build 36; ftp://ftp.ncbi.nlm.nih.gov/hapmap/). We selected 2999 autosomal SNPs that were present both in the whole-exome data and the HapMap3 datasets and conducted PCA using the data from the patients with ALS (n=504), control subjects (n=1165) and the unrelated individuals from the 11 global HapMap3 (n=924) populations. As a result, five ALS cases with ancestral differences were detected, which was also confirmed by the medical records. We excluded these cases in this study. Two control subjects with ancestral differences were also detected and excluded from this study (online supplementary figure 1). The remaining 499 patients with ALS, including 89 patients with FALS and 410 with SALS, and 1163 control subjects were subjected to the analysis in this study.

Supplementary data

Eighty-nine families with FALS and 410 patients with SALS are all from the Japanese population. Among the 89 families with FALS, 67 showed the autosomal-dominant mode of inheritance, 2 with affected siblings with parental consanguinity, 14 with affected siblings without parental consanguinity and 6 with no detailed information available about the mode of inheritance except for having a family history of ALS. The median age at onset of patients with both FALS and SALS was 60 years (55 years in FALS and 61 years in SALS), and the male:female ratio was 1.52:1. Regarding the site of symptom onset, the proportions of subjects with initial bulbar, upper extremity and lower extremity symptoms were 25.9%, 36.9% and 37.2%, respectively.

Whole-exome sequencing and variant analysis

Whole-exome sequencing analysis was performed for all of the 89 families with FALS and 410 patients with SALS. The genomic DNAs of the patients with ALS and 1163 healthy control subjects in our Japanese series were subjected to enrichment of exonic sequences using SureSelect V5 +UTRs or V6 +UTRs (Agilent). Massively parallel sequencing was accomplished using HiSeq sequencers (Illumina, 100 or 101 bp, paired-end reads). The Burrows-Wheeler Alignment tool18 and SAMtools19 were used with default parameter settings for alignment of raw reads and detection of single-nucleotide variants and short insertion/deletion variants (indels). After annotation with RefSeq (http://www.ncbi.nlm.nih.gov/RefSeq/), 1000 Genomes Project database (http://www.1000genomes.org/) and dbSNP135 (http://www.ncbi.nlm.nih.gov/projects/SNP/), all of the non-synonymous, nonsense, insertion/deletion or canonical splice site variant calls of the ALS causative genes were collected. As the ALS causative genes, 13 causative ALS genes (SOD1, FUS, TARDBP, ALS2, VAPB, SETX, SPG11, OPTN, VCP, UBQLN2, SIGMAR1, PFN1 and HNRNPA1),20 2 recently identified causative ALS genes (ERBB4 and MATR3)15 21 and 2 recently described genes from a large-scale ALS exome study (TBK1 and NEK1)22 were selected. To exclude the variants with low quality of variant call, those with the depth of coverage <10 or the Phred quality score <20 were excluded from the analysis. To focus on rare functionally predicted deleterious variants, variants with MAF ≥0.01 in any of the following databases were excluded from the analysis, which included the Japanese population databases (in-house database (1163 Japanese controls) and HGVD (Human Genetic Variation Database; http://www.hgvd.genome.med.kyoto-u.ac.jp)) and the global databases (ExAC (Exome Aggregation Consortium; http://exac.broadinstitute.org) and 1000 Genomes (http://browser.1000genomes.org)) (last accessed October 2017). In the filtering, 12 and 1 variants were excluded using the in-house database and HGVD, respectively. To detect repeat expansion mutations in C9ORF72, a repeat-primed PCR analysis was carried out as described previously.6

We defined those variants as the known pathogenic mutations if they were previously reported as well-established pathogenic variants in HGMD (Human Gene Mutation Database; http://www.hgmd.org) or if they were novel missense variants where different missense variants involving the same amino acids were reported as pathogenic variants in HGMD. Rare variants in this study were defined as the variants that were rare (MAF <1%) in the in-house database (1163 healthy Japanese control series in our facility) and all of the population databases. We defined rare variants with the scaled CADD score >20 as the functionally predicted deleterious variants in the following analysis.23 All of the known pathogenic mutations newly identified by whole-exome sequencing analysis were confirmed by direct nucleotide sequence analysis.

Statistical analysis

The ages at onset in the patients with ALS carrying known pathogenic mutations and/or rare functionally predicted deleterious variants in multiple genes were compared with those in other patients carrying zero or one variant by the Wilcoxon rank-sum test using R V.3.4.1. We also used Fisher’s exact test and multiple regression analysis to calculate the significance of differences. A p value of less than 0.05 was considered statistically significant.

Results

Identification of known pathogenic mutations or rare functionally predicted deleterious variants in causative genes for ALS

Including our previous reports, we identified 48 (53.9%) of the 89 families with FALS that harboured known pathogenic mutations in the causative genes for ALS, including 32, 8, 3, 1, 2, 1 and 1 in SOD1, FUS, TARDBP, ERBB4, C9ORF72, VCP and HNRNPA1, respectively. Of the 410 patients with SALS, we identified 16 patients (3.9%) carrying known pathogenic mutations in the causative genes for ALS, including 9, 2, 2, 1, 1 and 1 in SOD1, FUS, TARDBP, C9ORF72, VCP and SETX, respectively.

In addition to the pathogenic mutations, we identified 16 rare functionally predicted deleterious variants in patients with FALS and 82 in patients with SALS in the causative genes for ALS. In total, 64 variants in patients with FALS and 98 in patients with SALS were analysed further (table 1).

Table 1

Summary of known causative mutations or rare functionally predicted deleterious variants identified in patients with FALS and SALS

Burden of rare variants in the causative genes for ALS accelerates age at onset of ALS

Of the variants identified in this study, multiple variants (known pathogenic mutations or rare functionally predicted deleterious variants) were identified in seven probands with FALS (7.9%) and eight patients with SALS (2.0%; table 2). In seven of the 89 families, we identified multiple variants including known pathogenic mutations and rare functionally predicted deleterious variants, while these variants were present in 3.27% of the control subjects. We performed a binomial test to evaluate whether the observed frequency of patients with multiple variants was in excess than expected by chance,24 which demonstrated that the frequency of the patients with FALS with multiple variants was significantly higher than one might expect on the basis of chance (p=0.00051). To investigate whether the number of variants carried by the patients accelerates the age at onset of ALS, the subjects with ALS including those with FALS and SALS were classified into three groups on the basis of the number of variants (known pathogenic mutations and/or rare functionally predicted deleterious variants) they carry. The age at onset in the group of patients with ALS carrying multiple variants (N=15; median, 43.0 years; IQR 37.0–50.0) was significantly earlier than that of the group carrying zero or one variant (N=427; median, 61.0 years; IQR 52.0–69.0; p<0.00005), and are also significantly earlier than that of the group with one variant (N=114; median, 60.0 years; IQR 49.0–69.8; p=0.0002) (figure 1 and online supplementary table 1). We performed the multiple regression analysis incorporating subjects with multiple variants, family history and gender as the covariates. Age at onset of ALS is significantly correlated with subjects with multiple variants (p<0.00005) as well as with the family history (p=0.0149) (online supplementary table 2). These results suggest that the burden of multiple rare variants in known causative genes for ALS, which includes known pathogenic mutations and rare functionally predicted deleterious variants, significantly affects the age at onset in the Japanese ALS series.

Supplementary data

Table 2

Subjects with multiple variants in causative genes for ALS*

Figure 1

Boxplots of ages at onset of patients classified into three groups based on the number of known pathogenic mutations and/or rare functionally predicted deleterious variants. Patients were classified into three groups based on the number of known pathogenic mutations and/or rare functionally predicted deleterious variants (0, 1 and multiple (2 or 3) variants). Ages at onset in the three groups are shown in boxplots. The boxes indicate the 25th and 75th percentiles, the horizontal bars within boxes mark the median, the whiskers indicate the 10th and 90th percentiles, the bars indicate the highest and lowest values, and the circles indicate outliers. The ages at onset of patients with ALS with the multiple known pathogenic mutations and/or rare functionally predicted deleterious variants are significantly earlier than those with zero or one such variant (****p<0.00005).

Since the presence of pathogenic mutations for FALS alone may have substantial effects on the age at onset of ALS, we further investigated the potential effect of multiple rare functionally predicted deleterious variants on the age at onset of ALS, excluding patients carrying known pathogenic mutations. In this analysis, rare functionally predicted deleterious variants in multiple genes were identified exclusively in seven patients with SALS. The ages at onset in these patients with ALS (N=7; median, 50.0 years; IQR 41.5–53.0) were significantly earlier than those of all other patients who carried zero or one rare functionally predicted deleterious variant (N=375; median, 62.0 years; IQR 54.0–70.0; p=0.00803) and are also significantly earlier than those of patients with one variant (N=62; median, 65.0 years; IQR 56.3–64.2; p=0.0014) (figure 2 and online supplementary table 3).

Figure 2

Boxplots of ages at onset of patients classified into three groups based on the number of rare functionally predicted deleterious variants. Patients were classified into three groups based on the number of rare functionally predicted deleterious variants (0, 1 and multiple (2 or 3) variants). Ages at onset in the three groups are shown in boxplots. The patients carrying known pathogenic mutations were excluded. The boxes indicate the 25th and 75th percentiles, the horizontal bars within boxes mark the median, the whiskers indicate the 10th and 90th percentiles, the bars indicate the highest and lowest values, and the circles indicate outliers. The ages at onset of patients with amyotrophic lateral sclerosis with the multiple functionally predicted deleterious variants are significantly earlier than those with zero or one such variant (*p<0.05).

Moreover, as the additional analysis, we excluded the patients with ALS carrying known pathogenic mutations in SOD1, which were most frequently identified in our Japanese ALS series. In the remaining patients with ALS, the age at onset in the group of patients with ALS carrying multiple variants (N=9; median, 46.0 years; IQR 37.0–50.0) was significantly earlier than that of the group carrying zero or one variant (N=396; median, 62.0 years; IQR 53.0–69.3; p=0.00109) and was also significantly earlier than that of the group with one variant (N=83; median, 62.0 years; IQR 55.0–70.0; p=0.0017). These results further support the idea that the burden of rare variants in the known causative genes for ALS significantly accelerates the age at onset of ALS, even when known pathogenic mutations are excluded from the analysis.

Discussion

In this study, known pathogenic mutations and/or rare functionally predicted deleterious variants in multiple causative genes for ALS were identified in seven probands with FALS and eight patients with SALS. The ages at onset in those patients with ALS with multiple variants were significantly earlier than those in other patients with ALS. We further confirmed that age at onset of ALS is significantly correlated with subjects with multiple variants as well as with family history by multiple regression analysis. These results suggest that the burden of variants in known causative genes for ALS affects the age at onset in the Japanese ALS series.

Recently, there have been several studies of cases with multiple variants in the causative genes for ALS.10 25–27 Of such studies, targeted sequencing analysis of 17 known ALS genes in 391 patients with ALS from the USA revealed that 3.8% of subjects had rare potentially pathogenic variants in multiple genes, and these subjects had a disease onset 10 years earlier (p=0.0046) than the other subjects.25 Although the method of comprehensive sequencing analysis or the causative genes sequenced were slightly different between this previous study and our study, the same tendency that the burden of variants in known causative genes for ALS accelerates the age at onset was revealed. As described above, the genetic epidemiology of ALS in the Japanese and European patients with ALS differ substantially, particularly in the prevalence of C9ORF72 mutations. Hence, we need to be careful in assessing the burden of these variants on the age at onset of ALS, taking the genetic epidemiology in ethnic backgrounds into account. Nonetheless, the current study reinforced the findings of the recently published study that the burden of rare variants accelerated the age at onset of ALS.25 Furthermore, future replication studies in larger populations of ALS cases and control subjects are important for further extending the current studies.

Other recent studies have also shown that multiple rare variants in ALS-related genes affect disease progression or survival probability.28 29 These findings suggest that the burden of rare variants contribute to the age at onset and also the progression of ALS and survival. Because no analysis of disease duration was performed in our study, the association between survival and multiple variants should be further investigated in the Japanese population.

Among the variants, known pathogenic mutations were identified in 48 families with FALS and 16 patients with SALS. The known pathogenic mutations are considered to have a much greater impact on ALS pathogenesis than other rare functionally predicted deleterious variants because these known causative mutations in a single gene alone are sufficient to cause ALS. Furthermore, the age at onset in the patients with FALS carrying pathogenic mutations in SOD1 (median, 46.5 years; IQR 39.0–58.0) was significantly earlier than that in the other patients with FALS without pathogenic mutations in SOD1 (median, 59.5 years; IQR 49.0–67.5; p=0.0053) in our case series. In another study of 366 faimlies with FALS from the USA, the age at onset in the SOD1 group was also significantly earlier than that in the non-SOD1 group,30 which is consistent with the finding of our Japanese FALS series. In our series, excluding known pathogenic mutations in the causative genes for ALS or in SOD1, however, the ages at onset in the patients with SALS carrying rare functionally predicted deleterious variants in multiple genes were also significantly earlier than those in all other patients who carried zero or one rare functionally predicted deleterious variant. Since this sort of analyses was not mentioned in the recent report in the USA,25 our analysis further emphasises the importance of a thorough investigation of the contribution of multiple rare functionally predicted deleterious variants in addition to known pathogenic mutations. Furthermore, the biological implications of these functionally predicted deleterious variants should be thoroughly investigated to clarify the molecular bases of ALS.

This study further suggests that multiple variants modify the disease expression of ALS. When we excluded the patients with known pathogenic mutations in ALS causative genes, ages at onset in patients with ALS with rare functionally predicted deleterious variants in multiple genes were significantly earlier than those in other patients with ALS, suggesting the role of multiple such variants as a modifier for age at onset of ALS. We emphasise that burden of multiple rare variants accelerated the age at onset of ALS as a modifier of ALS phenotypes. Given the availability of comprehensive genome sequencing technologies, we should further extend our genome analyses to explore variants underlying ALS.

In summary, we have shown that the ages at onset in patients with ALS with pathogenic mutations and/or rare functionally predicted deleterious variants in multiple genes were significantly earlier than those in other patients with ALS. This finding suggests that the burden of rare variants in causative genes for ALS affects the age at onset in the Japanese ALS series. This finding should contribute to the better understanding of the molecular basis of SALS as well as FALS.

Acknowledgments

We thank all the patients and their family members for participating in this study. We also thank all the neurologists who provided samples for this study and the anonymous reviewers for their helpful suggestions for improving our report.

References

View Abstract

Footnotes

  • Contributors HN contributed to the study design, data acquisition, analysis and manuscript preparation. HI, JM, YT, TM, MT and JG contributed to the study design, data acquisition and analysis. KD, JY and SM contributed to data acquisition and analysis. TT and ST contributed to the study design, data acquisition, analysis and manuscript preparation.

  • Funding This work was supported in part by KAKENHI (Grants-in-Aid for Scientific Research on Innovative Areas Nos. 22129001 and 22129002) from the Ministry of Education, Culture, Sports, Science and Technology of Japan, and Grants-in-Aid (H23-Jitsuyoka (Nanbyo)-Ippan-004 and H26-Jitsuyoka (Nanbyo)-Ippan-080) from the Ministry of Health, Welfare and Labour, Japan, and grants (nos. 15ek0109065h0002, 16kk0205001h001, 17kk0205001h0002 and 17ek0109279h0001) from the Japan Agency for Medical Research and Development (AMED) to ST.

  • Competing interests None declared.

  • Patient consent Not required.

  • Ethics approval Ethical approval in this research was obtained from the institutional review board of the University of Tokyo.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.