Article Text

Download PDFPDF

Rare heterozygous parkin variants in French early-onset Parkinson disease patients and controls
  1. S Lesage1,
  2. E Lohmann1,2,
  3. F Tison5,
  4. F Durif6,
  5. A Dürr1,2,3,
  6. A Brice1,2,3,4,
  7. for the French Parkinson’s Disease Genetics Study Group*
  1. 1
    INSERM UMR S679, Paris, France
  2. 2
    AP-HP, Groupe Pitié-Salpêtrière, Fédération des Maladies du Système Nerveux, Paris, France
  3. 3
    AP-HP, Groupe Pitié-Salpêtrière, Département de Génétique et Cytogénétique, Paris, France
  4. 4
    Université Pierre et Marie Curie-Paris VI, UFR, Groupe Pitié-Salpêtrière, Paris, France
  5. 5
    Service de Neurologie, Hôpital Du Haut-Lévêque, Pessac, France
  6. 6
    Service de Neurologie, Hôpital Gabriel Montpied, Clermont-Ferrand, France
  1. Dr A Brice, INSERM UMR 679, Hôpital de la Salpêtrière, 47, Boulevard de l’Hôpital, 75651 Paris cedex 13, France; brice{at}


  • * The French Parkinson’s Disease Genetics Study Group members: Y Agid, A-M Bonnet, M Borg, A Brice, E Broussolle, P Damier, A Destée, A Dürr, F Durif, E Lohmann, M Martinez, C Penet, P Pollak, O Rascol, F Tison, C Tranchant, M Vérin, F Viallet, M Vidailhet

  • Competing interests: None declared.

View Full Text

Statistics from

Autosomal recessive mutations in the parkin gene in the PARK2 locus (OMIM 600116; OMIM 602544) are common causes of familial and sporadic early-onset parkinsonism.13 However, other studies suggest that parkin variants play a role in common late-onset forms of Parkinson disease (PD, OMIM 168600) (age of onset >45 years).4 5 Parkin mutations associated with PD affect all exons, and include point mutations, small insertions/deletions and much larger deletions, and exon duplications and triplications.6 These mutations are present not only in the homozygous or compound heterozygous state, compatible with recessive transmission, but also, in a substantial number of cases, as single heterozygous mutations. It is unknown whether single heterozygous parkin mutations are themselves pathogenic, whether a second mutation may have been missed during screening, or whether there are epistatic interactions with mutations in other causative genes.710 The pathogenic role of heterozygous mutations remains, therefore, a matter of debate. It is not clear whether such mutations influence the risk of PD or the age of onset.4 5 11 12 Kay et al have recently reported similar frequencies of sequence variants in the parkin gene in a large population of patients with PD and control subjects.13 We analysed rare heterozygous exonic parkin variants in 172 Caucasian French patients with sporadic PD and 170 ethnically matched controls.


Patients and control subjects

This study was approved by the ethics committees of the institutions involved, and written informed consent was obtained from all subjects. The PD cohort (172 patients without known family history of PD), was recruited through the French Network for the Study of Parkinson’s Disease Genetics; 94 of them have already been reported.13 The patients fulfilled the diagnostic criteria for clinically defined PD: at least three of the mandatory criteria (akinesia, rigidity, resting tremour, asymmetrical onset or >30% improvement with levodopa), and the absence of exclusion criteria.

The 172 patients comprised 99 male and 73 female patients (mean (SD) age at examination 49.6 (11.8) years (range 16 to 75); age at onset 36.8 (9.3) years (range 6 to 58); duration of disease 10.3 (7.3) years (range 1 to 29)). The 170 ethnically-matched control subjects, mostly spouses of patients without family history of PD, comprised 94 male and 76 female patients. Their mean (SD) mean age at examination of 62.9 (9.6) years (range 27 to 84) was later than that of the 172 patients (p<0.05). All 170 control subjects and 153 of the patients with PD had previously been screened for mutations in exon 41 of the leucine rich-repeat kinase 2 (LRRK2, OMIM 609007).14 One control subject carried the LRRK2 G2019S mutation and two patients with sporadic PD were heterozygous for the G2019S or Y2006H mutation, respectively. No parkin mutations were found in these three subjects.

Mutation analyses

Genomic DNA was extracted from peripheral blood samples collected on EDTA by standard phenol–chloroform procedure.

Parkin mutation analyses

The promoter and all 12 exons of parkin and their intron–exon boundaries were screened for point mutations and exonic rearrangements by denaturing high-performance liquid chromatography (DHPLC) and sequencing and by semi-quantitative multiplex PCR, respectively, as described.15

PINK1 and DJ-1 mutation analyses

The eight exons of PTEN-induced putative kinase 1 (PINK1, OMIM 608309) and the six coding exons of the DJ-1 (OMIM 602533) genes, and their intron–exon boundaries were screened by direct sequencing.16 17

Statistical analyses

Comparisons of parkin allele frequencies between groups and mean age at onset of patients with two mutations versus those with one were computed by using the χ2 and Student t test, respectively. Confidence intervals (CI) were calculated by the logit method.


Although we found 12 common intronic polymorphisms in parkin in this study, we focused on exonic or splicing variants, which are suspected to be pathogenic or risk factors for PD. Of the 172 patients, 18 (10%) had either two mutations (n = 13; mean (SD) age at onset 24.2 (9.5) years, range 12 to 45) or consecutive exon rearrangements but an unknown parental phase (n = 5; mean (SD) age at onset 31.0 (16.4) years; range 17 to 38) (table 1). Two patients with 2 mutations were homozygous and 11 were compound heterozygous. The phase of the latter could be determined for five patients by reverse transcriptase PCR (SPD-166-06, SPD-169-03) or by sequencing available relatives (SPD-134-010, SPD-145-012, SPD-372-001). For the remaining patients, we presumed that the two mutations were on different chromosomes, as the point mutations identified in these patients have already been reported as a second mutation in trans with an exon deletion.6 No truncating mutations or exonic rearrangements were found in the 170 control subjects. In the remaining 154 patients and the 170 control subjects, we found 10 exonic sequence changes (tables 1, 2). Three variants (p.S167N, p.D394N, p.V380L), found in both controls and patients, were known common polymorphisms (minor allele frequency ⩾0.01). The allele frequency of two of the variants, p.S167N and p.D394N, was similar in patients and control subjects. The p.V380L polymorphism, however, which is in Hardy–Weinberg equilibrium, was significantly less common in patients than in controls (9.6% vs 17.6%, p = 0.002) (table 2).

Table 1 Characteristics of patients and control subjects with parkin variants
Table 2 Allele frequencies of parkin exonic coding variants in Patients with PD and control subjects

The seven remaining variants were rare (minor allele frequency <0.01), and all were heterozygous, including two novel amino acid substitutions (p.N273S/c.919 A→G; p.I298S/c.995 T→G). Both variants were predicted by the SIFT program ( to affect protein function. However, they were not evolutionarily conserved among species. The synonymous p.L261L variant was only found in two control subjects. Six non-synonymous variants, including two newly identified sequence variations, were found exclusively in 10 patients (p.A82G, p.R256C, p.N273S, p.R275W, p.I298S, p.G328E). The mean (SD) age at onset of these 10 heterozygous parkin cases was 38.1 (10.0) years (range 18 to 50). To investigate a possible synergy between parkin and other PD-associated genes, the 10 patients with single parkin variants, none of whom had LRRK2 exon 41 mutations, were also screened for PINK1 and DJ-1 mutations. Only one patient (SPD-100-003) with a single parkin variant had the known DJ-1 p.R98Q polymorphism.18 19

This study shows, therefore, that there is a significant difference in the allele frequencies of rare coding variants in patients and controls (2.9% vs 0.6%, p = 0.02; OR = 5.3, 95% CI 1.1 to 23.3) (table 2), and that patients with a single heterozygous parkin mutation had later ages at onset (38.1 (10.0) years) than those with two mutations (24.2 (9.5) years) (p = 0.003).


This study addressed the frequency of rare parkin variants in a series of isolated patients with PD and in controls. In addition to homozygous or compound heterozygous parkin mutations, found exclusively in patients, we identified 10 exonic changes, including 2 novel variants. However, it is possible that the number of heterozygous mutation carriers in our series could have been underestimated, as we did not include in this group patients with consecutive exon rearrangements, or those with compound heterozygous mutations, for whom we could not determine the phase. Of the 10 variants, 2 were novel and located within or close to the ring finger 1 domain that is thought to interact with parkin substrates, and 3 were known common polymorphisms. Interestingly, we confirmed a possible protective effect of the p.L380 allele.20 21 In addition, we identified seven rare sequence variants found either exclusively in patients (n = 10) or exclusively in controls (n = 2). In contrast to previous studies,13 21 we found a significantly higher frequency of rare parkin variants in patients with sporadic PD compared with controls, perhaps because of the much earlier age at onset in our patients with PD compared with those in previous studies (36.8 vs 56.713 or 64 years21). Similar results were also observed by Clark et al,12 who reported a higher allele frequency of heterozygous parkin mutations (5.4%) among patients with an age at onset of 41.12 (7.2) years, compared with controls (1%). Indeed, the age at onset in our 10 heterozygous parkin carriers was similar to that found in typical patients with 2 mutations in parkin. However, the patients in our study who were heterozygous for parkin had a later age at onset than those with two mutations (38.1 vs 24.2 years, p = 0.003). Interestingly, all sequence variants found exclusively in patients were non-synonymous, in contrast to other studies that have reported a variety of heterozygous parkin sequence changes, including missense mutations, exon rearrangements and rare intronic variants in both cases and controls.1113 21 Their role in the pathogenesis of sporadic PD remains to be determined. However, four of them may be pathogenic, as they were found with a second mutation in previous studies.2 3 5 21 In addition, a cellular study demonstrated that the two most common variants in our heterozygous patients, p. R256C and p. R275W, are located in the ring finger 1 domain and are dominant gain-of-function mutations.22 Our study extends previous results12 21 to include exon rearrangement detection in both patients and control groups. The significantly greater frequency of non-synonymous changes in patients than in control subjects could be due to the presence of an undetected second mutation in parkin or another gene, or an increased risk of PD due to the presence of a single mutation. To address the first hypothesis, we screened all patients and controls for rearrangements of the parkin promoter and all exons. This technique has recently allowed the identification of a novel deletion of the common promoter and exon 1 of both parkin and the adjacent parkin coregulated gene (PACRG, OMIM 608427), in 1 family out of 23 with members who were heterozygous parkin carriers.15 No promoter rearrangements were detected in the present series, but it is still possible that a second disease-causing mutation in the very large parkin gene has escaped detection. Recently, individuals with both LRRK2 and parkin7 8 or PINK1 and DJ-19 mutations have been reported. No potential epistatic interaction between parkin and other known PD-associated genes, such as PINK1, DJ-1, and the most common LRRK2 G2019S mutation, was found in our patients. Alternatively, it has been speculated that single parkin mutations increase the risk of PD.11 12 Our results are compatible with this hypothesis, but the possibility of a second undetected parkin mutation cannot yet be excluded. As a whole, the causative role of the single heterozygous mutations not only in the parkin gene but also other causative genes associated with autosomal recessive parkinsonism, such as PINK12326 or more rarely DJ-1,19 remains a matter of debate. The presence of these single mutations might be fortuitous and unrelated to their disease, but this hypothesis cannot be confirmed without knowledge of the frequency of carriers of the different variants in the general population.

Key points

  • In total, 172 isolated cases with early-onset parkinsonism and 170 ethnically matched controls were screened for both parkin point mutations and exon rearrangements.

  • We identified 10 exonic sequence variants, including 3 common polymorphisms and 7 rare heterozygous variants.

  • There was a significant difference in the allele frequencies of rare parkin coding variants in patients and controls.


We are grateful to the patients and their families. We thank M Ruberg for critical reading of the manuscript, S Laine, B Debarges, S Janin and L Leclere for their technical support, and the DNA and Cell Bank of the IFR 070 for sample preparation. This work was supported by Cohortes et Collections 2001 INSERM/the French Ministry of Research and Technology (contract 4CH03G), the European commission (EU Contract No.LSHM-CT-2003-503330/APOPIS), the Association France-Parkinson, and the National Institutes of Health grant NS41723-01A1.


View Abstract


  • * The French Parkinson’s Disease Genetics Study Group members: Y Agid, A-M Bonnet, M Borg, A Brice, E Broussolle, P Damier, A Destée, A Dürr, F Durif, E Lohmann, M Martinez, C Penet, P Pollak, O Rascol, F Tison, C Tranchant, M Vérin, F Viallet, M Vidailhet

  • Competing interests: None declared.

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.