Introduction

A PARK2 mutation was first described in a Japanese family with an autosomal recessive juvenile Parkinson’s disease (PD) (Kitada et al. 1998) and 213 different mutations have since been identified throughout PARK2, including large deletions or amplifications, small deletions or insertions, as well as missense mutations (The Parkinson Disease Mutation Database, accessed July 2015). Studies have shown PARK2 mutations in about 40-50 % of early onset FPD, and about 1.3–20 % of SPD patients (Choi et al. 2008; Hedrich et al. 2004; Kann et al. 2002; Mellick et al. 2009; Oczkowska et al. 2013; Sironi et al. 2008), suggesting that PARK2 dysfunction by mutations might be involved in the pathogenesis of both FPD and SPD.

PARK2, an E3 ubiquitin ligase regulates a variety of processes, including receptor trafficking and mitochondrial quality control, via mono- or poly-ubiquitinations of PARK2 substrates (Dawson and Dawson 2014). Several distinct pathomechanisms by PARK2 mutations have been reported. Many missense mutations reduce the enzymatic activity of PARK2, leading to the abnormal accumulation of toxic proteins and neurodegeneration (Corti et al. 2011; Dawson and Dawson 2010; Sul et al. 2013). The mutations in PARK2 gene also change the general physical characteristics of the protein such as stability and/or solubility, promoting aggregate formation and impairing mitochondrial maintenance, collectively resulting in increased cellular toxicity (Gaweda-Walerych and Zekanowski 2013; Hampe et al. 2006; Oczkowska et al. 2013; Sriram et al. 2005). These results suggest that PARK2 mutants might influence the risk of developing PD.

The frequency of PARK2 mutation has been reported in the 5.5–15.8 % of the Korean population with early onset PD (EOPD) (Choi et al. 2008; Chu et al. 2014; Kim et al. 2012). However, the frequency of PARK2 mutation has not been examined in Korean PD patients in other age at onset (AAO) groups. Therefore, this is the first analysis of frequency of PARK2 variants in Korean PD patients with AAO ≥ 50 years.

Materials and methods

Subjects

A total of 83 PD patients were included in the study. To analyze the relationship between AAO and variants, we divided patients into three groups based on their AAO of the disease (Mehanna et al. 2014): those younger than 49 years (young onset PD, YOPD), those between 50 and 69 years (middle onset PD, MOPD), and those older than 70 years (late onset PD, LOPD).

Mutation analysis

Genomic DNA was extracted from peripheral blood by standard protocol. The exons 1–12 of the PARK2 gene were amplified from genomic DNA using intronic primers (Table 1). The reaction mixture for PCR contained 50 ng of genomic DNA and 10 pmol of each primer. The PCR reactions were denatured at 94 °C for 5 min followed by 30 cycles of 94 °C for 45 s, 50–60 °C for 45 s, and 72 °C for 45 s with a final extension of 7 min at 72 °C. The PCR products were purified and sequenced directly using the ABI 3730XL DNA analyzer (Applied Biosystems).

Table 1 PCR primers and annealing temperatures
Full size table

Gene dosage analysis

Exon rearrangements were analyzed with a multiplex ligation-dependent probe amplification (MLPA) assay using commercially available probes (SALSA P051 Parkinson MLPA kits, MRC Holland). Sequence-specific probes enclosed in this kit are against all exons of PARK2 and PINK1, and specific exons of SNCA (exons1, 3–6); PARK7 (DJ-1, exons 1, 3, 5, 7); TNFRSF9 (exon 3); ATP13A2 (exon 2, 9); LPA (6q26) and two-point mutations (A30P in SNCA and G2019S in LRRK2). Experimental procedures were conducted for all 83 samples according to the manufacturer protocol. MLPA peaks analysis, normalization, and calculation of dosage ratio were performed with the GeneMarker software (SoftGenetics LLC) with standard parameters of analysis.

Statistical evaluation

The paired t test was used to compare mean AAOs among PD patients with no variant, with one variant, and with ≥2 variants. Values were expressed as mean ± SD and P < 0.05 was considered statistically significant.

Results

We sequenced all of the 12 exons of the PARK2 gene from 15 patients selected randomly from our PD samples in order to see a preliminary profile of the PARK2 gene of our PD patients. Sequence analysis of 15 patients disclosed three different missense mutations: S167N (exon 4), V380L (exon 10), and M458L (exon 12). We did not detect novel mutations other than the reported variations in exon 4, 10, and 12. Thereafter, we screened sequence variations in these three exons (exon 4, 10 and 12) of other patient samples, based on the above data. In addition, we also screened exon 2 encoding the ubiquitin-like (Ubl) domain, and exon 6 and 7 encoding RING1 domain, because mutations in these domains tend to affect the enzymatic function of PARK2 (Caulfield et al. 2014).

The demographic characteristics of the 83 patients are presented in Table 2 (YOPD, N = 2; MOPD, N = 32; LOPD, N = 49). Five variants identified in this study are S167N (c.500G > A), Y267H (c.799T > C), L272I (c.814C > A), V380L (c.1138G > C), and M458L (c.1372A > C) (Table 3). S167N polymorphism in exon 4 was the commonest variant (allele frequency of 40.4 %) when compared to the other polymorphisms. L272I and V380L alleles were found relatively infrequently (0.6 and 4.2 % of the allele frequencies in the PD patients, respectively).

Table 2 Demographics of PD patients with one, with ≥ two variants, and without variant
Full size table
Table 3 Allele and genotype frequencies of PARK2 variants in PD patients
Full size table

In order to detect exon dosage changes caused by genomic rearrangements of the known PD genes, we used the SALSA P051-B1 kit. The kit consists of probes for specific exons of SNCA, PARK7, TNFRSF9, ATP13A2, and LPA, as well as point mutations of SNCA A30P and LRRK2 G2019S. Three out of 83 patients had genomic duplications in the PARK2 and PINK1 loci (Fig. 1; Supplementary Table 1) and other major PD gene rearrangements were not detected: two patients had duplication of the PARK2 gene and one patient had duplication of the PINK1 gene.

Fig. 1
figure 1

MLPA analysis results of three patients with gene dosage alteration

Full size image

When point mutations and exonic rearrangements were taken together, PARK2 variants were found in 54 (65 %) of 83 PD patients (Table 4): 23 patients carried compound heterozygous or homozygous variants, and 31 patients carried a single heterozygous variant. One patient with PINK1 mutation did not have PARK2 variant.

Table 4 Detected variants in this study
Full size table

Discussion

PARK2 is a well-known risk factor for juvenile and early onset occurrence of PD (Kitada et al. 1998). Moreover, the importance of PARK2 mutations in sporadic PD have also been reported (Foroud et al. 2003; Oczkowska et al. 2013). In the Korean population, PARK2 mutations were reported in about 5.5–18.5 % of EOPD and FPD (Choi et al. 2008; Chu et al. 2014; Kim et al. 2011, 2012). However, genetic study of late onset PD has not yet been reported. Therefore, this study assessed the frequency of PARK2 mutation in Korean PD patients, with AAO after 50 years.

We found that S167N polymorphism was the commonest with high frequency (40.4 %) in our study. Studies regarding the allele frequency of S167N vary greatly among populations with different ethnic origins ((Li et al. 2005; Martinez et al. 2010; Mellick et al. 2001; Sakai et al. 2010; Sinha et al. 2005). In the Asian SPD populations (Japanese, Chinese, and Taiwanese), the frequency of S167N polymorphism was 38.6–46.6 % (Supplementary Table 2) and that of Korean SPD in this study also fits in this range. On the other hand, S167N allele was found in only 0.2–2.5 % of a Caucasian population. The frequency of S167N polymorphism was also higher in the Asian EOPD population (around 40 %) than in the Caucasian EOPD population (2.6–12.9 %) (Supplementary Table 3). Taken together, S167N polymorphism may be a susceptibility factor to PD in Asian populations.

The study by Ghione et al.(Ghione et al. 2007) reported that the combination of PARK2 polymorphisms including S167N (14 % of cases, heterozygous) and environmental factors including pesticides, organic solvents, and rural living, strongly affects lowering the AAO of PD. S167N is located in exon 4 encoding the RING0/unique Parkin domain (UPD), a PINK1 interacting domain. PINK1 kinase activates PARK2 via phosphorylation of PARK2 on residue Ser65 (Iguchi et al. 2013; Kondapalli et al. 2012; Okatsu et al. 2012; Seirafi et al. 2015). It would be interesting to know whether S167N polymorphism interferes with the binding of PARK2 to PINK1 when an environmental risk factor is added. It is also probable that S167N polymorphism may promote aggregation or impair mitochondrial homeostasis by environmental and/or other genetic factors. Further functional studies of PARK2 polymorphisms would be useful for diagnostic and prognostic processes related to PD.

In the Korean LOPD patients, we found PARK2 exon duplication in only two patients: one had exon 5 duplication and the other had multiple duplications (exon 2, 3, 5–11) (Fig. 1 and Supplementary Table 1). On the other hand, a larger case study showed duplication of PARK2 on exons 2–7 in 2,091 PD patients (mean AAO 58.3 ± 12.1 years), and duplication of exon 1, 2, and 3 in 1,686 controls (mean age 66.1 ± 13.1 years) (Kay et al. 2010). Interestingly, only duplication of exon 4 was reported in American, Polish, and French EOPD patients (Ambroziak et al. 2015; Hedrich et al. 2001; Kay et al. 2010; Periquet et al. 2003). Therefore, it seems worthy to further investigate the role of exon 4 of PARK2 in the pathomechanism of EOPD.

To assess the relevance between variants and AAO, we performed a subgroup analysis (Mehanna et al. 2014): YOPD, MOPD and LOPD (Table 2). In MOPD, the average AAO was lower in patients with ≥2 variants than in patients with one variant, or without variant of statistical significance. We also observed lowered average AAO in LOPD patients with ≥2 variants; however, this was without statistical significance.

In this study, two variants (Y267H and L272I) in exon 7 encoding the RING1 domain were found in LOPD, while two variants (V380L and M458L) in exon 10 and 12 encoding c-terminus, were observed in MOPD (Table 3). RING1 recruits the E2 enzyme, but loss of the RING1 domain does not result in a loss of ligase activity (Chew et al. 2011; Matsuda et al. 2006; Spratt et al. 2013). This suggests that PARK2 is an E2-independent ubiquitin ligase, a RING between RING (RBR) E3 ligase. RING2 in the c-terminus of PARK2 is the true catalytic domain, and structural study of RING2 has shown that it is required for E2 recruitment, catalysis, and trans autoubiquitination. Consistent with this, AAOs of patients with c-terminal variant (V380L and M458L) were lowered when compared to those of patients with RING1 variant in our study (Table 3). Our data also suggest that the c-terminus of PARK2 might play a more important role than the RING1 region, in PD pathogenesis.

In conclusion, we have for the first time, presented a genetic analysis of the PARK2 gene in relation to Korean MOPD and LOPD. A high frequency of PARK2 variants in PD patients was found, suggesting that PARK2 variants might be a genetic risk factor for PD in Korean populations.