Replication analysis of the PRKN V380L (rs1801582) variant in a Japanese cohort of spinocerebellar ataxia type 3
Authors: Ekaterina Nadbitova, Nobuyuki Takei, Sachiko Hirokawa, Yuya Hatano, Tomohiko Ishihara, Osamu Onodera, Yuka Mitsuhashi Koike
Published: 2026

ABSTRACT
AbstractBackgroundSpinocerebellar ataxia type 3 (SCA3) is caused by CAG repeat expansion in ATXN3, which inversely correlates with age at onset but does not fully account for interindividual variability. A recent study in a mixed European/South and North American cohort suggested that the PRKN V380L variant (rs1801582), particularly in C/C homozygotes, was associated with earlier disease onset. We therefore performed a replication analysis to evaluate the frequency and potential clinical relevance of rs1801582 in a Japanese SCA3 cohort.Methodsrs1801582 genotypes were determined by Sanger sequencing in 228 genetically confirmed SCA3 patients and 260 SCA6 patients as convenience disease controls. Genotype frequencies were additionally compared with East Asian reference populations from the 1000 Genomes Project Phase 3 dataset. Associations with age at onset were evaluated using multivariable linear regression adjusted for expanded ATXN3 CAG repeat length, normal ATXN3 CAG repeat length, and sex.ResultsGenotype frequencies in SCA3 (G/G: 86.0%, G/C: 13.2%, C/C: 0.9%) were similar to those in controls (G/G: 85.8%, G/C: 13.5%, C/C: 0.8%) and closely resembled those reported in East Asian reference populations. Because of the extremely small number of C/C carriers (n = 2), subsequent analyses focused on G/G and G/C carriers. In SCA3, median age at onset was comparable between genotypes (42 vs. 41 years), and median expanded ATXN3 CAG repeat lengths did not differ between groups (71 vs. 71 repeats). Multivariable regression analysis adjusting for expanded ATXN3 CAG repeat length, normal ATXN3 CAG repeat length, and sex demonstrated that rs1801582 was not independently associated with age at onset (β = − 0.42 years, p = 0.84), whereas expanded CAG repeat length remained a strong predictor (β = − 1.68 years per repeat, p < 0.0001).ConclusionsThe rs1801582 C/C genotype is extremely rare in Japan; therefore, the present cohort had limited statistical power to directly evaluate the previously proposed recessive modifier effect. No detectable association between rs1801582 genotype and age at onset was identified among G/G and G/C carriers in this Japanese cohort. These findings highlight the importance of population-specific validation and adequately powered replication studies when evaluating candidate genetic modifiers in SCA3.Keywords: Machado-Joseph disease (MJD), Single nucleotide polymorphism, Japanese population, Age at onset

Background

Spinocerebellar ataxia type 3 (SCA3; Machado–Joseph disease) is an autosomal dominant neurodegenerative disorder resulting in progressive ataxia and pyramidal symptoms [1]. It is characterized by the accumulation of aggregates composed of polyglutamine ataxin-3 (ATXN3) proteins translated from CAG repeat expansions in ATXN3 [2, 3]. Although CAG repeat length is inversely correlated with age at onset, it explains only part of the observed variability [4, 5], suggesting the involvement of additional genetic modifiers. Several studies have investigated genetic modifiers of age at onset in SCA3, including variants in APOE, ATXN2, DNA repair–related genes, and other candidate loci [6–11]. However, the reported effects have not always been consistently replicated across different populations, suggesting that genetic modifier effects in SCA3 may be ancestry- or population-specific.

PRKN encodes the E3 ubiquitin ligase parkin, which is involved in ubiquitin-mediated protein degradation and mitochondrial quality control and has been reported to functionally interact with ATXN3 within the ubiquitin–proteasome system [12–14]. A recent study in a mixed European/South and North American SCA3 cohort reported that the PRKN V380L (rs1801582) variant was associated with an approximately 3-year earlier age at onset, primarily in patients carrying the C/C genotype [15]. However, whether this association is present in East Asian populations remains unclear. Because the frequency of rs1801582 varies substantially across populations, independent replication studies in different ethnic cohorts are necessary to clarify the clinical relevance of this variant. Therefore, we performed a replication analysis of rs1801582 genotypes in a Japanese cohort to evaluate its potential role as a genetic modifier of age at onset.

Methods

Patients

We enrolled 228 and 260 Japanese patients with genetically confirmed SCA3 and SCA6, respectively. Patients with SCA6 served as convenience disease controls as an independently ascertained hereditary ataxia cohort recruited under comparable clinical evaluation protocols. Both SCA3 and SCA6 patients were evaluated using standardized protocols, allowing comparison of rs1801582 genotype frequencies between similarly ascertained ataxia cohorts. Genomic DNA was extracted from peripheral blood, and the rs1801582 genotypes (G/G, G/C, and C/C) were determined by Sanger sequencing. Clinical data, including age at onset and expanded/normal ATXN3 CAG repeat lengths, were obtained from medical records. Age at onset was defined as the age at which the patient first recognized any neurologic symptoms related to spinocerebellar degeneration.

Statistical analysis

Because of the extremely low frequency of the C/C genotype, comparisons were restricted to G/G and G/C carriers. Continuous variables were analyzed using the Mann–Whitney U test, and categorical variables were analyzed using the chi-square test. Hardy–Weinberg equilibrium was assessed in both SCA3 and SCA6 cohorts using a chi-square goodness-of-fit test. Given the very small expected counts for the C/C genotype, Hardy–Weinberg equilibrium results were interpreted cautiously. To evaluate the independent association between rs1801582 genotype and age at onset, multivariable linear regression analysis was performed with age at onset as the dependent variable and expanded ATXN3 CAG repeat length, normal ATXN3 CAG repeat length, sex, and rs1801582 genotype as independent variables. Because the C/C genotype was extremely rare in this cohort, regression analyses compared G/G and G/C carriers. The rs1801582 genotype was coded as a binary variable (G/G = 0, G/C = 1) and sex was coded as male = 0 and female = 1 in the regression model. Statistical analyses were performed using GraphPad Prism version 10.2.3 (GraphPad Software). Genotype frequencies in East Asian reference populations were obtained from the 1000 Genomes Project Phase 3 dataset available through Ensembl [16]. Statistical significance was set at p < 0.05.

Results

Clinical and genetic characteristics of the Japanese SCA3 cohort

Clinical and genetic characteristics of the Japanese SCA3 cohort are summarized in Table 1. The cohort included 228 genetically confirmed Japanese patients with SCA3, including 118 male and 110 female. Median age at onset was 42 years, and the median expanded ATXN3 CAG repeat length was 71.

Table 1Clinical and genetic characteristics of Japanese SCA3 patientsVariableJapanese SCA3 patients (n = 228)Male/Female118/110Age at onset, years42 [31–52]Expanded ATXN3 CAG repeat length71 [68–74]Normal ATXN3 CAG repeat length20 [14–28]rs1801582 genotype, number (%) G/G196 (86.0%) G/C30 (13.2%) C/C2 (0.9%)Data are presented as median [interquartile range] or number (%)

Clinical and genetic characteristics of Japanese SCA3 patients

Data are presented as median [interquartile range] or number (%)

Comparison of rs1801582 genotype frequencies across cohorts

In SCA3 patients, the rs1801582 genotype frequencies were as follows: 196 G/G (86.0%), 30 G/C (13.2%), and 2 C/C (0.9%) (Table 1). Frequencies in SCA6 patients were similar (223 G/G [85.8%], 35 G/C [13.5%], and 2 C/C [0.8%]). Genotype frequencies were consistent with Hardy–Weinberg equilibrium in both SCA3 (χ2 = 0.47, p = 0.49) and SCA6 (χ2 = 0.22, p = 0.64) cohorts. A two-group χ2 test comparing G/G and G/C carriers showed no significant difference between SCA3 and SCA6 (χ2 = 0.026, df = 1, p = 0.99).

The genotype frequencies observed in the Japanese SCA3 and SCA6 cohorts closely resembled those reported in East Asian populations from the 1000 Genomes Project Phase 3 dataset [16], whereas the frequency of the C/C genotype was substantially lower than that reported in the mixed European/South and North American SCA3 cohort described by Weber et al. [15] (Fig. 1).

Comparison of rs1801582 genotype frequencies across cohorts. Genotype frequencies of rs1801582 in Japanese patients with SCA3, Japanese patients with SCA6, East Asian populations from the 1000 Genomes Project Phase 3 dataset [16], and the mixed European/South and North American cohort reported by Weber et al. [15] are shown. The frequency of the C/C genotype was substantially lower in the Japanese and East Asian cohorts than in the Weber et al. cohort [15]

Association between rs1801582 genotype and age at onset in SCA3

Because of the low prevalence of C/C carriers, subsequent analyses included only G/G and G/C carriers. Among SCA3 patients, median age at onset was 42 years for G/G carriers and 41 years for G/C carriers (p = 0.59). Median expanded ATXN3 CAG repeat lengths were identical between groups (71 vs. 71; p > 0.05).

Scatter plot analysis demonstrated a strong inverse relationship between expanded ATXN3 CAG repeat length and age at onset, whereas G/G and G/C carriers showed largely overlapping distributions (Fig. 2). Expanded ATXN3 CAG repeat length explained approximately 43% of the variance in age at onset (R2 = 0.429).

Relationship between expanded ATXN3 CAG repeat length and age at onset in Japanese patients with SCA3. Each point represents an individual patient and is color-coded according to rs1801582 genotype. The regression line (black line) indicates the overall inverse relationship between expanded ATXN3 CAG repeat length and age at onset. G/G and G/C carriers showed substantially overlapping distributions, whereas only two patients carried the C/C genotype

Multivariable regression analysis of age at onset in SCA3

To further evaluate whether rs1801582 genotype independently modifies age at onset in SCA3, we performed multivariable linear regression analysis including expanded ATXN3 CAG repeat length, normal ATXN3 CAG repeat length, sex, and rs1801582 genotype (G/G vs G/C).

Expanded ATXN3 CAG repeat length was strongly associated with earlier age at onset (β = − 1.68 years per repeat, 95% CI − 1.935 to − 1.422, p < 0.0001). In contrast, rs1801582 genotype showed no independent association with age at onset (β = − 0.42 years, 95% CI − 4.538 to 3.693, p = 0.84). Neither sex nor normal ATXN3 CAG repeat length showed significant associations with age at onset (Table 2). The multivariable regression model explained approximately 43% of the variance in age at onset (R2 = 0.434).

Table 2Multivariable linear regression analysis of age at onset in Japanese SCA3Variableβ coefficient95% CIp-valueSex (male vs female)1.779−1.012 to 4.5710.210Expanded ATXN3 CAG repeat length−1.678−1.935 to −1.422< 0.0001Normal ATXN3 CAG repeat length0.08381−0.1269 to 0.29450.434rs1801582 genotype (G/G vs G/C)−0.423−4.538 to 3.6930.840Sex was coded as 0 = male and 1 = femaleGenotype was coded as 0 = G/G and 1 = G/CPatients with the C/C genotype were excluded because only two patients carried the C/C genotypeModel R2 = 0.434

Multivariable linear regression analysis of age at onset in Japanese SCA3

Sex was coded as 0 = male and 1 = female

Genotype was coded as 0 = G/G and 1 = G/C

Patients with the C/C genotype were excluded because only two patients carried the C/C genotype

Model R2 = 0.434

Two SCA3 patients carrying the C/C genotype had ages at onset of 25 and 60 years, with CAG repeat lengths of 75 and 69, respectively. Because only two patients carried the C/C genotype, the present cohort lacked sufficient statistical power to directly evaluate the recessive effect previously proposed.

Discussion

The PRKN V380L (rs1801582 C/C) genotype was exceptionally rare in this Japanese SCA3 cohort. Only two patients carried the C/C genotype; therefore, the present study had limited statistical power to directly evaluate the recessive modifier effect previously proposed by Weber et al. [15]. Importantly, multivariable linear regression analysis adjusted for expanded ATXN3 CAG repeat length, normal ATXN3 allele length, and sex demonstrated no independent association between rs1801582 genotype and age at onset, whereas expanded CAG repeat length remained a strong predictor, explaining approximately 43% of the variance in age at onset. This finding is consistent with those of previous studies showing that repeat length is a major, but incomplete, determinant of disease onset in SCA3 [4, 5]. Together with the extremely low frequency of C/C carriers, no clear evidence of a modifying effect among heterozygous C-allele carriers (G/C vs G/G) was observed in this Japanese SCA3 cohort. However, our findings should not be interpreted as evidence against the role of rs1801582 as a modifier in all SCA3 populations. Rather, the present results indicate that the previously proposed recessive effect is difficult to robustly evaluate in Japanese cohorts because of the rarity of the C/C genotype. The genotype frequencies observed in our Japanese SCA3 and SCA6 cohorts closely resembled those reported in East Asian populations from the 1000 Genomes Project Phase 3 dataset [16]. In contrast, the frequency of C-allele carriers was substantially lower than that observed in the mixed European/South and North American cohort reported by Weber et al. [15], in which G/C and C/C genotypes accounted for 30.6% and 3.1% of patients, respectively. Indeed, Weber et al. similarly observed that the association with earlier age at onset was driven primarily by C/C homozygous carriers, whereas G/C heterozygotes showed only modest differences from G/G carriers [15]. These findings suggest that the clinical relevance of PRKN V380L may depend on ancestry-specific genotype distributions and population background.

Biologically, PRKN and ATXN3 interact within ubiquitination and mitophagy pathways [8–10], raising the possibility that functional variation in PRKN could influence disease onset in SCA3. Moreover, recent functional analysis in an SCA3 cell model demonstrated that PRKN V380L reduced the PRKN-ATXN3 interaction and impaired mitophagy, thereby compromising cell viability [11]. Although these findings support a potential biological role of PRKN V380L, the clinical impact of this variant may depend on ancestry-related genetic background and population structure. Previous studies investigating genetic modifiers in SCA3 have similarly reported incomplete replication across cohorts, including studies evaluating APOE, ATXN2, DNA repair–related genes, and other modifier loci [6–11]. Collectively, these observations support the concept that genetic modifier effects in SCA3 may vary across populations.

This population-specific replication study has practical implications. Given its extremely low frequency and the absence of a detectable effect, rs1801582 genotyping is unlikely to substantially improve prediction of age at onset beyond expanded ATXN3 CAG repeat length in Japanese SCA3 patients. Routine testing of rs1801582 is therefore unlikely to influence genetic counseling regarding expected age at onset in the Japanese population. This study also highlights broader considerations in genetic modifier research for polyglutamine disorders. Whereas the pathogenic CAG repeat constitutes a major determinant of onset [4, 5], the residual unexplained variance suggests a polygenic landscape rather than being explained solely by single variants.

Several limitations should be acknowledged. First, the PRKN V380L genotype was extremely rare, limiting statistical power to detect recessive effects. In addition, the very small number of homozygous C/C carriers resulted in low expected genotype counts, and Hardy–Weinberg equilibrium testing should therefore be interpreted cautiously. Second, SCA6 served as a convenience disease control group; while their recruitment under similar protocols minimizes ascertainment bias, their allele frequencies may not fully represent the general Japanese population. However, the genotype frequencies observed in the SCA6 cohort were comparable to those in East Asian reference populations, supporting their use as an additional Japanese frequency reference cohort. Because the present study was specifically designed to evaluate the previously proposed association between PRKN V380L and SCA3 age at onset, subsequent modifier analyses were restricted to the SCA3 cohort. Third, our analysis focused on a single candidate variant in PRKN. It remains possible that PRKN harbors other functional variants, rare coding mutations, or structural alterations that contribute to SCA3 heterogeneity. In addition, family structure, parental origin, anticipation, and detailed longitudinal clinical measures, including disease duration, SARA scores, non-ataxic manifestations, and disease progression, were not systematically available in this retrospective cohort.

Future large international multicenter studies incorporating polygenic approaches will be essential for identifying population-relevant modifiers that better explain the variability in age at onset of SCA3. Larger East Asian cohorts will be particularly important for clarifying whether the recessive effect previously proposed can be detected in populations in which the C/C genotype is rare. Given that PRKN contains multiple functional domains and may harbor additional variants beyond V380L, functional assays investigating PRKN-ATXN3 interactions across different haplotypes—not only PRKN V380L—may help clarify whether certain genetic backgrounds potentiate biological effects.

Conclusion

In this Japanese cohort, no detectable association between rs1801582 genotype and age at onset was identified among G/G and G/C carriers after adjustment for expanded ATXN3 CAG repeat length, normal ATXN3 CAG repeat length, and sex. The PRKN V380L (rs1801582) variant was exceedingly rare; therefore, the present study could not robustly evaluate the previously proposed recessive modifier effect. These findings contrast with those of a previous report in the mixed European/South and North American cohort and suggest that the modifying effect of candidate variants may be population- or ancestry-dependent. Careful replication across diverse ethnic backgrounds is essential before attributing clinical relevance to putative genetic modifiers.

Authors’ contributions

Ekaterina Nadbitova performed the investigation, formal analysis, and visualization, acquired funding, and wrote the original draft of the manuscript. Nobuyuki Takei performed the investigation and contributed to manuscript review and editing. Sachiko Hirokawa, Yuya Hatano, and Tomohiko Ishihara contributed resources and manuscript review and editing. Osamu Onodera conceptualized and supervised the study, acquired funding, and contributed to manuscript review and editing. Yuka Mitsuhashi Koike conceptualized the study, performed formal analysis and visualization, acquired funding, wrote the original draft, and contributed to manuscript review and editing. All authors read and approved the final manuscript.

Funding

This work was supported by grants (23ek0109641h0001) from the Japan Agency for Medical Research and Development (AMED), Grants-in-Aid for Scientific Research (C) (24K10635) from the Japan Society for the Promotion of Science (JSPS), and a grant (PRESTO, Sakigake)(JPMJPR248A) for the Fundamental Understanding of Age-Related Organismal Transformations from the Japan Science and Technology Agency (JST), Health and Labor Sciences Research Grant from the Ministry of Health, Labor, and Welfare (JPMH: 23FC1010) and Kyowakai Medical Research Grants.

Data availability

The datasets generated and/or analyzed during the current study contain individual-level genetic and clinical data and are therefore not publicly available due to privacy and ethical restrictions. The genotype and phenotype data of rs1801582 in SCA3 and SCA6 patients are publicly available in the Zenodo repository and can be accessed via the following link: https://doi.org/10.5281/zenodo.19270651. Additional de-identified data can be obtained from the corresponding author upon reasonable request and with approval from the institutional review board of Niigata University.

Declarations

Ethics approval and consent to participate

The study was approved by the institutional review board of Niigata University (approval number: G2021-0010, approved on September 29, 2021) and was conducted in accordance with the Declaration of Helsinki. The requirement for additional informed consent was waived due to the retrospective design.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

Associated Data

Data Availability Statement

The datasets generated and/or analyzed during the current study contain individual-level genetic and clinical data and are therefore not publicly available due to privacy and ethical restrictions. The genotype and phenotype data of rs1801582 in SCA3 and SCA6 patients are publicly available in the Zenodo repository and can be accessed via the following link: https://doi.org/10.5281/zenodo.19270651. Additional de-identified data can be obtained from the corresponding author upon reasonable request and with approval from the institutional review board of Niigata University.