Patients were recruited from the Second Affiliated Hospital of Zhejiang University School of Medicine from May 2017 to April 2021. A total of 114 patients with SALS and 15 with FALS were screened by WES. In addition, we performed the C9orf72 gene test, and all the hexanucleotide expansions of patients except one were within 30 repeats. For further phenotype–genotype analysis, we included previously reported SOD1-mutated probands in our centers from December 2007 to April 2017. All patients were diagnosed by at least two senior neurologists as having ALS according to the Revised El Escorial criteria (Brooks et al., 2000). All participants were of Han Chinese descent and came from southeastern China. The ethics committee of each participating center approved the study, and written consents were obtained from all participants during their first hospital visit. The clinical data of ALS patients, including gender, family history, age at onset (AAO), diagnostic delay, site of onset, side of onset, disease duration, and clinical manifestations were collected upon the first visit to the hospital.

Genomic DNA was extracted from blood using a DNA Extraction Kit (Qiagen, Hilden, Germany). The samples were captured by the Agilent Sure Select Human All Exon V6 products and sequenced on the Illumina HiSeq X Ten platform (XY Biotechnology Co. Ltd., Hangzhou, China). Further bioinformatic analysis has been performed according to our previously reported protocol (Dong et al., 2021). Briefly, all variants were annotated by ANNOVAR. SIFT and PolyPhen-2 software were used to predict the functional changes in proteins caused by the variants. Single Nucleotide Polymorphism (dbSNP) Database, the 1,000 Genomes Project, and the ExAC database were used to check the frequency in the general population. Variants were finally classified according to the American College of Medical Genetics and Genomics (ACMG) standards and guidelines. Sanger sequencing was performed to validate the potential variants. All primers covering the five exons in SOD1 were seen in our previous report (Niu et al., 2011).

The coding sequence of human wild-type (WT) SOD1 gene (RefSeq NM_000454.5) was cloned into the pFlag-CMV-4 vector using a ClonExpress II One Step Cloning Kit (Vazyme). Plasmids with mutant SOD1 (c.400G > T and c.49_50del insCA) were created by PCR mutagenesis (Toyobo, Osaka, Japan) and were verified by Sanger sequencing. Four plasmids expressing Vector, WT, p.G17H, and p.E134* were constructed.

HEK293T cells were maintained in Dulbecco’s modified essential medium DMEM-supplemented 10% fetal bovine serum stored in a humidified incubator under 5% CO₂ at 37°C. Transient transfection was performed using Lipofectamine 3,000 (Invitrogen, Life Technologies, Grand Island, NY, United States ).

HEK293T cells were seeded on poly-D-lysine (PDL)-treated coverslips (NEST, China) and transiently transfected with various expression plasmids (pFLAG-SOD1-WT, pFLAG-SOD1-G17H, pFLAG-SOD1-E134*). Twenty-4 hours later, the cells were rinsed with 1 × phosphate-buffered saline (PBS), fixed with 4% paraformaldehyde for 8 min, and then permeabilized with 1 × PBS supplemented with 0.01% Triton X100. The cells were blocked with 5% donkey serum and 1% bovine serum albumin (BSA; sigma, St. Louis, MO) in PBS for 1 h at room temperature. The primary antibody was rabbit antibody anti-flag (1:800, Cell Signaling Technology, 14793S). The secondary antibody was Alexa Fluor 488 donkey anti-rabbit. The cells were incubated with NucBlue Live Reagent (Hoechst 33342; Thermo Fisher Scientific, Oregon, United States) and visualized using Olympus FluoView FV3000 confocal microscopy with a ×63 objective.

Forty-eight hours after transfection, HEK293T cells over-expressing pFlag-SOD1-WT, pFlag-SOD1 mutants, or the empty vector were lysed and harvested. The protein samples were resolved by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), transferred to polyvinylidene difluoride (PVDF) membrane, and blotted with the 5% non-fat milk. The antibodies against flag (1:1,000, CST) and β-tubulin (1:5,000) were used. All immunoblotting images were acquired using a BioRad system.

Total RNA was extracted using Trizol reagent (Takara, Kusatsu, Japan), which was then reverse transcribed to cDNA by the PrimeScript RT reagent Kit (Takara, Kusatsu, Japan) in accordance with the manufacturer’s instructions. Reverse transcription (RT)-PCR was further performed using a SYBR Premix Ex Taq Kit (Takara, Kusatsu, Japan). The PCR conditions are listed as follows: incubation for 3 min at 50°C followed by incubation for 3 min at 95°C, and finally 40 cycles at a duration of 10 s each while at 95°C and then 30 s at 30°C. The following primers were used: target SOD1 and human GAPDH. Detection and data analysis were conducted using an ABI StepOnePlus sequence detection system (Thermo Fisher Scientific, Oregon, United States ), and endogenous glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal control. Expression levels were quantified by threshold cycle values.

Descriptive statistics were provided for the site of onset, heredity of disease (FALS or SALS), gender, AAO, diagnostic delay, and survival time. Besides, continuous data were compared using Student’s t-test or the Mann–Whitney test, while dichotomous variables, such as gender and site of onset, were analyzed using either the standard chi-square test or Fisher’s exact test. Survival duration was determined by Kaplan-Meier analysis, and differences were determined by log-rank testing. A two-tailed p < 0.05 was considered statistically significant. All analyses were performed using GraphPad Prism 8.0 (GraphPad Software, CA, United States).

From 2017 to 2021, a total of eight different variants in SOD1 were found, including six known pathogenic variants (p.G38R, p.V48A, p.N87S, p.C112Y, p.I114T, and p.L145S) and two novel variants (p.G17H and p.E134*). Two heterozygous variants, c.49_50del insCA (p.G17H) and c.400G > T (p.E134*), were detected in two families with a positive family history (Figures 1A,B), which was confirmed with Sanger sequencing (Figures 1C,D). Both novel variants were absent in ESP6500 and ExAC database and were highly conserved from chimpanzee to zebrafish (Figures 1E,F). The nonsense variant, p.E134*, generates the truncated protein caused by premature termination. However, the nonsense variant was not regarded as very strong evidence of pathogenicity because loss of function is not the primary mechanism in SOD1-related ALS. In addition, p.E134* was predicted to be detrimental by Mutation Taster and CADD. The variants p.G17H and p.E134* were classified as likely pathogenic and pathogenic, respectively, according to ACMG criteria obtained using Varsome.

To further elucidate the biological effects of these two variants, SOD1 protein and mRNA levels of mutants were markedly reduced compared to WT (Figures 1G,H), indicating that the reduced protein expression was caused by reduced mRNA synthesis. The in vitro results showed that the mRNA levels of mutant SOD1 were degraded more rapidly than WT, suggesting a loss-of-function effect. The aggregation propensity assay showed that the cells overexpressing WT plasmid showed diffuse cytoplasmic SOD1 protein, while the misfolded aggregates were seen in cells transfected with p.G17H and p.E134* mutants (Figure 1I). The soluble physiological SOD1 protein had a strong tendency to become toxic aggregates due to mutation, implying a gain-of-function effect.

In this study, a total of 114 patients with SALS and 15 patients with FALS were screened by WES. Among them, 12 subjects were genetically identified as having SOD1 variants, 10 (10/15, 66.7%) of whom were FALS patients and 2 (2/114, 1.8%) of whom were apparently SALS patients. We previously reported 20 probands with SOD1 mutations in 36 unrelated FALS patients (Niu et al., 2011; Liu et al., 2014; Lin et al., 2019; Liu et al., 2019). When integrating these results into the current study, SOD1 mutations account for 58.9% (32/51) of our FALS cases.

A total of 21 SOD1 mutations, including 1 nonsense and 1 deletion/insertion mutations, were found spanning all five exons, but only one mutation was in exon 3 (Table 1). The most frequent SOD1 mutation was p.V48A (4/32, 12.5%) in four FALS probands, followed by p.H47R, p.C112Y, and p.G148D [each were found in three probands (3/32, 9.4%)], as well as p.H121Q and p.L145S [each were found in two probands (2/32, 6.3%)], and the remaining mutations were each found in one proband (Figure 1J). Among all the mutations, p.L145S, p.P75S, and p.I114T are found in SALS patients (Figure 1J).

Combining our previously reported SOD1-mutated patients in our centers, a total of 32 probands were genetically identified with SOD1 mutations (Table 1). Twenty-nine (29/32, 90.6%) of the probands were FALS patients, and three (3/32, 9.4%) were apparently SALS patients. The gender ratio (M:F) was 1:1. With the exception of 1 case with missing data, only 1 patient among 31 available probands exhibited bulbar onset. The majority of them (18/31, 58%) presented with lower limb onset, but some (10/31, 33.3%) presented with upper limb onset. Sixteen probands presented with a predominantly lower motor neuron phenotype, while another 16 probands were the classical ALS phenotype.

With the exception of the patient with missing data, 31 patients had available AAO data for analysis. The mean (SD) AAO was 46 ± 11.4 years, and the median AAO was 47 years (Figure 2A). The mean AAO of patients carrying mutations in exon 1 [n = 5, 34.6 (12.4)] was statistically significantly earlier than those with mutations in exon 2 [n = 8, 51.4 (8.2)] (p = 0.038) (Figure 2B). There was no difference in the mean (SD) AAO between male and female patients [47.6 (12.2) vs. 45.25 (11.2), p = 0.58] (Figure 2C). The onset of FALS in patients is earlier than that of SALS, albeit without statistical significance [45.4 (11.4) vs. 55.7 (10.2), p = 0.15] (Figure 2D). The AAOs of ALS patients were highly heterogenous with the 25th and 75th percentiles at 38 and 53 years, respectively. Six mutations (p.A5S, p.G11V, p.G17H, p.D84G, p.G148D, and p.I150T) showed relatively younger AAOs, which was lower than the first quartile (25%, <38 years), and another six mutations (p.H47R, p.V48A, p.P75S, p.N87S, p.I114T, and p.H121Q) presented with relatively older AAOs, which were higher than the third quartile (75%, >53 years). The youngest and oldest AAOs are p.G17H and p.N87S, which were found in patients in their twenties and seventies, respectively.

Twenty-nine probands patients were accessible for data regarding diagnostic delay. With the exception of one case with an extremely long diagnostic delay of 84 months, the mean (SD) diagnostic delay of the remaining was 10.7 ± 6.2 months (Figure 3A). The diagnostic delay showed no differences in different exons (Figure 3B). The time to diagnosis of male patients exhibited no difference compared to female patients [10.4 (5.9) vs. 11.0 (6.7) months, p = 0.82] (Figure 3C). It is worth noting that the mean of the diagnostic delay of FALS patients is less than that of SALS patients and was found to be statistically significant [9.5 (4.8) vs. 20.3 (9.3) months, p = 0.0026] (Figure 3D). Diagnostic delay varied greatly across different mutations. Four mutations of four probands, namely, p.A5S, p.L107V, p.C147R, and p.G148D, showed a diagnostic delay less than the first quartile (25%, <6 months). Another five mutations of eight probands, namely, p.H47R, p.V48A, p.P75S, p.C112Y, and p.L145S, showed a diagnostic delay more than the third quartile (75%, >12 months).

Disease duration data were available for 31 probands, and 9 patients still survive at the censoring date. The median survival time was 40.0 months, and the 5-years survival rate was 30.7% for all subjects (Figure 4A). The survival time showed no difference between all male and female patients (Figure 4B). Given the extremely long survival time in mutation of p.H47R, we excluded the p.H47R mutation in both male and female patients for statistical analysis. Strikingly, male patients showed significantly longer survival time than female patients (40 vs. 16 months, p = 0.05) (Figure 4C).

The SOD1-mutant ALS patients in this cohort all come from southeastern China; thus, it should be mentioned that these data are not representative of all Chinese populations. Given the limited sample size, the frequency of SOD1 for FALS probably was overestimated. Moreover, our center is still working on the enrollment of the ALS patients and screening for the genetic causes. In addition, for those patients with family history but negative genetic results, we would need a larger scale sequence such as whole genome sequencing (WGS) to define the genetic information of Chinese ALS cases.