Abstract
Background: Early onset PD has been associated with different mutations in the Parkin gene, including exon deletions and duplications.
Methods: The authors performed an extensive mutational analysis on 50 probands with onset of PD at younger than 50 years of age. Thirteen probands were ascertained from a registry of familial PD and 37 probands by age at onset at younger than 50 years, blind to family history. Mutational analysis was undertaken on the probands and available family members and included conventional techniques (single strand conformation polymorphism analysis and sequencing) and a newly developed method of quantitative duplex PCR to detect alterations of gene dosage (exon deletions and duplications) in Parkin.
Results: Using this new technique, the authors detected eight alterations of gene dosage in the probands, whereas 12 mutations were found by conventional methods among the probands and another different mutation in an affected family member. In total, the authors identified compound heterozygous mutations in 14%, heterozygous mutations in 12%, and no Parkin mutation in 74% of the 50 probands. We expanded the occurrence of Parkin mutations to another ethnic group (African-American).
Conclusion: The authors systematically screened all 12 Parkin exons by quantitative PCR and conventional methods in 50 probands. Eight mutations were newly reported, 2 of which are localized in exon 1, and 38% of the mutations were gene dosage alterations. These results underline the need to screen all exons and to undertake gene dosage studies. Furthermore, this study reveals a frequency of heterozygous mutation carriers that may signify a unique mode of inheritance and expression of the Parkin gene.
PD is the second most common neurodegenerative disorder with a prevalence of about 2% in persons older than 65 years of age.1 However, PD also can manifest earlier in life, which is referred to as early onset PD (EOPD). Age at onset in EOPD usually is considered to be below 402 to 50 years.3-5⇓⇓ PD is characterized clinically by tremor, rigidity, bradykinesia, and postural instability. In the last few years, three genes and four gene loci have been associated with PD (alpha-synculein,6 Parkin,7 UCH-L1,8 and gene loci on chromosomes 2p13,9 4p14–16,10 1p35–36,11 and 1p3612). Whereas mutations in alpha-synculein and UCH-L1 are rare, mutations in the Parkin gene have been found in many patients with early onset autosomal recessive PD.4,5,7,13⇓⇓⇓ The phenotype associated with Parkin mutations is variable but usually is characterized by an early age at onset, even presenting itself in childhood.4 In addition, disease progression usually is slow, patients respond well to levodopa therapy, and various additional signs may be observed such as diurnal fluctuation of symptoms, sleep benefit, dystonia, early levodopa-induced dyskinesia, and hyperreflexia.4,14⇓
All proteins currently associated with monogenic forms of parkinsonism appear to be involved in the ubiquitin-mediated pathway of protein degradation. It has been shown recently that Parkin acts as an E3–ubiquitin ligase that ubiquitinates itself and promotes its own degradation.15-17⇓⇓ In addition, it is involved in the ubiquitin-dependent degradation of several substrates, for example, O-glycosylated alpha-synuclein,18 the parkin-related endothelin receptor–like receptor (Pael-R),19 the synaptic vesicle-associated protein called CDCrel-1,17 and synphilin-1, a protein interacting with alpha-synuclein.20
Currently, only one study has investigated the frequency of Parkin mutations in EOPD by using a combined approach of conventional testing and semiquantitative gene dosage studies, which, however, did not include screening of exon 1.4 In 100 sporadic and 73 familial cases, the group identified 39 homozygous or compound heterozygous mutation carriers (23%) and 15 patients with heterozygous Parkin mutations (9%). The unexpectedly high percentage of heterozygous mutation carriers remains to be investigated. PET studies in asymptomatic carriers of a heterozygous Parkin mutation with one apparently normal allele reveal a mild but significant decrease of mean FDOPA uptake compared with control subjects in all striatal regions, suggesting that even heterozygous Parkin mutations may be associated with changes in the dopaminergic system.21
In this study, we performed extensive mutational analysis of 50 EOPD probands from a range of ethnic backgrounds and screened all 12 exons of Parkin by a newly developed highly accurate quantitative duplex PCR assay to test for gene dosage alterations5 in conjunction with conventional methods.
Materials and methods.
Patients.
Two groups of patients selected by different criteria were included in the study. Group A patients were taken from a PD genetic research registry based at the Department of Neurology, Columbia University. Multiplex PD families were recruited into this registry between the years 1995 and 1998. The genetics consent form for this registry gave permission for blood samples to be used in future genetic research. Where possible, the following methods of data collection were applied to both affected and unaffected family members: blood sample, clinical examination, video examination, handwriting sample, family history, medical risk factors, and drug questionnaire. Thirteen probands from this registry whose age at onset of PD was younger than 50 years (mean age at onset, 35.0 ± 12.6 years) underwent Parkin gene analysis. Parkin mutations were identified in seven probands of Group A. The DNA of 22 additional family members of six of these seven mutation-positive probands was available for segregation analysis. Investigators involved in the Parkin gene analysis were blind to case status of these relatives. Group B included 37 probands (mean age at onset, 39.0 ± 9.2 years) recruited from the Department of Neurology, Columbia University, between 1999 and 2000 solely because their age at onset of PD was younger than 50 years. Therefore, Group B, unlike Group A, was not ascertained based on family history of PD; rather, this information was gathered retrospectively. However, no family members of probands in Group B were available. All participants in both Groups A and B discussed the research at length with a genetic counselor and signed a consent form.
All probands but one (Family A-9) underwent a detailed neurologic examination or medical record review. The diagnosis of PD was defined by clinical and research criteria.22-24⇓⇓ Individuals with secondary or symptomatic parkinsonism or Parkinson plus syndromes were excluded. Age-at-onset information was obtained by direct interview and corroborated with medical records when available. Information on specific features that have been associated with the Parkin phenotype4 was not systematically collected. However, if atypical features were noted either on examination, standardized video protocol, or medical record review, they were tabulated (tables 1 and 2⇓⇓). All clinical data collection was done by researchers blind to Parkin mutation status.
Demographic and clinical data of Group A
Continued
Demographic and clinical data of mutation-positive individuals of Group B
Molecular analysis.
Conventional mutational analysis.
To detect point mutations and small deletions or insertions, we performed conventional single strand conformation polymorphism (SSCP) analysis after amplification of all 12 exons of Parkin using published intronic primers.7 This was followed by exon sequencing on an automated sequencing machine (LI-COR; Lincoln, NE) in cases of observed band shifts. Sequence changes were investigated in 200 control subjects by SSCP analysis. In addition, all 12 exons and exon-intron boundaries were sequenced in patients with only one detected mutation.
Gene dosage studies.
Gene dosage analysis was performed in a quantitative duplex PCR assay of all 12 exons of Parkin on the LightCycler (Roche Diagnostics, Mannheim, Germany) with a quantitative duplex PCR assay using the fluorescence resonance energy transfer technique. The Beta globin gene was co-amplified with each individual Parkin exon and served as internal standard. Primers, probes, and details of the method are described elsewhere.5 In brief, the following reagents were used for amplification in a 10-μL reaction: 1 μL of Hybridization FastStart Mix (Roche Diagnostics); 2 to 4.5 mmol/L of MgCl2; 0.2 μmol/L of each of the hybridization probes (one pair of 3′-fluorescein and 5′-LightCycler Red 705 for the reference gene Beta globin, and one exon-specific pair of 3′-fluorescein and 5′-LightCycler Red 640 for Parkin); 0.5 to 1.0 μmol/L of each primer, and 1 to 15 ng of DNA. A standard curve was generated using human genomic DNA (Roche Diagnostics) in concentrations of 5 ng/μL, 1.25 ng/μL, and 0.3125 ng/μL, respectively. All standards were amplified in duplicate, and a regression curve was calculated. Based on this regression curve, sample concentrations were inferred and accepted only when within the range of the standard templates. All samples also were measured in duplicate. Using Beta globin as internal control provided a relative ratio: concentration of each Parkin exon to concentration of Beta globin. A ratio between 0.8 and 1.2 was considered as normal (figure 1A), a heterozygous deletion was expected at a ratio between 0.4 and 0.6, and a heterozygous duplication between 1.3 and 1.7. All detected gene dosage variations were confirmed at least twice.
Figure 1. (A) Gene dosage analysis of all samples. Mean Parkin–Beta globin ratios for each exon of Parkin using 50 individuals. Only values that did not suggest an alteration of gene dosage (i.e., ratios from 0.8 to 1.2) were included. Error bars indicate SD. (B) Results of gene dosage studies for Patient II.4 of Family A-1 carrying a heterozygous deletion of exon 3 and a heterozygous duplication of exon 5.
Haplotype analysis.
Haplotype analysis at the PARK2 locus (6q25–27) was performed by PCR using published primers (http://gdbwww.gdb.org) with the following microsatellite DNA markers (location is given in parentheses): D6S1277 (173.33 cM; upstream of Parkin), D6S1599 (169.95 cM; intron 2), D6S980 (167.78 cM; intron 3), D6S305, D6S411 (166.39 cM; both intron 7), D6S1550 (166.39 cM; intron 9), D6S1579 (166.39 cM; intron 10), and D6S1035 (164.78 cM) and D6S2436 (154.64 cM; both downstream of Parkin) (http://research.marshfieldclinic.org/genetics/; http:// www.ncbi.nlm.nih.gov/entrez/[NT_007122 Protein, Homo sapiens chromosome 6 working draft sequence segment, version October 16, 2001]). PCR products were analyzed on an automated sequencer (LI-COR). To ensure accurate sizing of the alleles, two control samples of individuals 1331.1 and 1331.2 from the Centre d′Étude du Polymorphisme Humain underwent genotyping.
Statistical analysis.
Mean age at onset for mutation-positive and mutation-negative individuals and those with heterozygous and compound heterozygous mutations are reported separately for Groups A and B. Because the ascertainment of individuals in each of these two groups was completely different, comparison would not be valid. Based on the positive family history in Group A, younger age at onset in relatives of these individuals might be expected solely because of earlier recognition of disease in these individuals. Four individuals in Group B also had a family history of PD but were not ascertained on this basis.
Results.
Patients.
Fifty probands, and an additional 22 family members (11 definitely or possibly affected and 11 unaffected) of six mutation-positive families of Group A were included in this study (no family members were available for A-7). Subjects came from ethnic backgrounds as diverse as white, Caribbean Hispanic, and African American. The proband (A-7) of African-American descent previously was described elsewhere.25 Clinical data for all families of Group A and the mutation-positive subjects of Group B are summarized in tables 1 and 2⇑⇑. Selected pedigrees are shown in figure 2.
Figure 2. Selected pedigrees with haplotypes at the PARK2 locus. Probands are marked by an arrow. Definitely affected individuals are shaded in black. Gray diamonds symbolize individuals with possible PD. Haplotypes are indicated below the respective individual. The putative shared haplotype associated with each of the three recurrent mutations is in bold font. In cases of unknown chromosomal phase, genotypes are given in parentheses. (A) Pedigrees of Family A-2 and of Patient B-3 are shown. In both, an exon 3–4 deletion was detected. Symbols represent the following mutations: * = 202–203delAG; # = Ex3–4del; ∼ = 56delA. (B) Pedigrees of Family A-4 and of Patients B-1 and B-4, members of whom were carriers of the 924C→T substitution. Mutations are indicated as follows: * = 924C→T; # = 867C→T.C; ∼ = 438–477del. There are two affected individuals in Family A-4 with two different heterozygous mutations. (C) Pedigrees of families with a deletion of exon 3 and the respective additional mutation (A-1 and A-5): * = Ex3del; # = Ex5dupl; ∼ = 1390G→A. Diagram shows the gene dosage of exon 5 of Parkin compared with the reference gene Beta globin. Dosage of exon 5 is represented by a fixed value of 1.0 for Beta globin (black bars) and the respective amount of Parkin (gray bars). One parent and each child carry the duplication of exon 5, demonstrating the segregation of this type of mutation in a family.
Mutational analysis.
We found 17 different Parkin mutations, 8 of which are unreported and 3 of which were found to be recurrent in this study. Among the 50 probands in Groups A and B, we identified 7 compound heterozygous mutation carriers (14%), 6 heterozygous mutation carriers (12%), and no homozygotes.
In Group A, among 13 probands and 11 definitely and possibly affected family members, we identified 8 compound heterozygous mutation carriers (4 probands and 4 family members) and 8 heterozygous mutation carriers (3 probands and 5 family members), whereas a further 8 affected individuals (6 probands and 2 affected family members from mutation-positive families) seem to be mutation negative. Conversely, nine unaffected family members of Group A carried a heterozygous Parkin mutation, but none of the unaffected subjects had a compound heterozygous mutation (see table 1⇑).
Among the 37 probands of Group B, 6 were identified to carry a Parkin mutation (3 compound heterozygotes, 3 heterozygotes).
Taking both groups together, we detected 8 gene dosage alterations among the 50 probands, including heterozygous deletions of exons 1, 2, 3 (twice), 4, and 3–4 (twice), and a heterozygous duplication of exon 5. In addition, we identified 12 mutations among the probands by SSCP analysis and sequencing (6 different missense mutations [81G→T, 226G→C, 675A→C, 924C→T (three times), 1390G→A, 1411C→T], a silent single basepair substitution [884A→G], and three small deletions [202–3delAG, 256delA, 438–477del]) (figure 3). Furthermore, another different missense mutation (867C→T) was found in an affected second-degree relative of the proband of Family A-4 and the unaffected child (see figure 2B). Known polymorphisms in amplicons for exons 2, 4, 8, 10, and 11 were detected by conventional mutation analysis as well (data not shown).
Figure 3. Localization of mutations in Parkin identified in this study. Exon rearrangements (duplications and deletions) are indicated above the sequence, and missense mutations and small deletions, below the sequence. Regions encoding special domains in the predicted protein are highlighted (dotted, ubiquitin-like domain; horizontally striped, RING-finger domain; diagonally striped, in-between–RING domain). Sample numbers of patients carrying the respective mutation are given in parentheses. The numbers of nucleotides indicate the nucleotide position in the complementary DNA sequence as described.7 Newly identified mutations are boxed.
Regarding the type of mutation, five probands (Families A-2, A-5, A-7 and probands B-2, B-3) were found to be compound heterozygotes, having a gene dosage alteration together with a small basepair alteration (see tables 1 and 2⇑⇑). A further proband was identified as compound heterozygous only by sequence analysis (B-1, see table 2), whereas another proband (A-1) was considered mutation negative by SSCP analysis but was found to be a compound heterozygote with alterations of gene dosage by quantitative duplex PCR (see table 1⇑ and figure 1B). In another six probands (Family A-3, A-4, A-6 and probands B-4, B-5, B-6) only one mutation was detected, five by sequencing and one by gene dosage analysis (see tables 1 and 2⇑⇑).
All results of the mutational analysis in family members are shown in tables 1 and 2⇑⇑ and in figure 2.
Three of the mutations were found in multiple probands in this study: Ex3del (n = 2), Ex3–4del (n = 2), and 924C→T (n = 3), and we performed a haplotype analysis in these cases to investigate if they have a common founder. The deletion of exons 3 to 4 was detected in two patients from Puerto Rico (Family A-2, B-3) who probably shared a common haplotype at all markers tested, assuming hemizygosity at the marker D6S980, because of a deletion in the mutation-bearing chromosome. Unfortunately, phasing was possible only for one of the probands. Another recurrent mutation (924C→T in exon 7) was found in three European patients (Family A-4, B-1, B-4) with a putatively shared haplotype at five markers close to exon 7 (D6S305–D6S411–D6S1550–D6S1579–D6S1035). Moreover, the allele of 220bp at D6S305 was found in all three patients but in only 5 of 194 control chromosomes (2.6%). Unfortunately, the disease-bearing chromosome could be identified in one proband only. Another two families of European origin carried a deletion of exon 3 (A-1, A-5). Both showed a common haplotype at markers representing the 5′ part of the Parkin gene (D6S1277–D6S1599–D6S980). Identified haplotypes are shown in figure 2.
In addition, we demonstrated the segregation of some of the mutations in the families by mutational and haplotype analysis (see figure 2).
Compound heterozygotes vs heterozygotes.
We calculated the mean age at onset for definitely affected patients of both groups with respect to their mutational status. For this, all affected subjects (probands and relatives) were included for the calculation in Group A. Among all definitely affected individuals in each group, the onset of the disease was earlier in patients with at least one mutation in Parkin than in patients without any detected mutation (30.8 ± 15.6 years vs 42.6 ± 10.7 years in Group A, and 26.8 ± 10.5 years vs 41.3 ± 6.9 years in Group B, respectively). Furthermore, in both groups, patients with two mutations were younger at onset than patients with only one detected mutation (27.1 ± 16.3 years vs 35.8 ± 14.5 in Group A, and 19 ± 7.0 years vs 34.7 ± 6.5 years in Group B, respectively).
These studies are exploratory and were not subjected to formal statistical analysis because of the small numbers. Also, a comparison of mean age at onset between Groups A and B was not made because of the different origins and methods of ascertainment of these two cohorts.
Discussion.
We describe the results of systematic mutation screening of a large patient cohort comprising familial EOPD and patients ascertained by age at onset alone from a range of different ethnic backgrounds.
We identified 17 different mutations, including 11 different small sequence changes and 6 different gene dosage alterations, which stresses the importance of gene dosage studies as part of the search for mutations in the Parkin gene. Without performing gene dosage studies, six of seven probands who were identified as compound-heterozygous mutation carriers would have been considered heterozygous, and even one, mutation-negative. To our knowledge, eight of the mutations (47%) are described for the first time in this study, including a single basepair substitution (81G→T) in exon 1 and a deletion of exon 1. This demonstrates the necessity of screening all exons of the Parkin gene. In the largest published study of mutation analysis of the Parkin gene,4 screening of exon 1 was not undertaken. In addition, we have identified an undescribed duplication of exon 5 and demonstrated its segregation in a family (see figure 2C). Overall, duplications seem to be a relatively rare finding, as indicated by only 7 detected duplications in 346 (2%) chromosomes4 and confirmed in this investigation (1/100 [1%] chromosomes of the 50 probands). Furthermore, five newly identified small sequence variations in exons 2 to 12 (226G→C, 256delA, 675A→C, 884A→G, and 1411C→T) underline the wide spectrum of mutations in the Parkin gene.
Group A included a proband of African-American background (A-7) whose family members previously screened negative for mutations in the alpha-synuclein gene.25 We identified the proband to be compound heterozygous for mutations in the Parkin gene (Ex2del + 675A→C). This is the first patient of West or central African origin to carry Parkin mutations, illustrating the pan-ethnic occurrence of mutations in the Parkin gene.
Family A-4, of European origin, proved to be interesting because there were two affected heterozygous mutation carriers with two different mutations and four asymptomatic heterozygotes, two of whom were older than the age at onset of affected family members. Genetic results in Family A-2 also were surprising. There were three affected siblings of whom only two were compound heterozygous for Parkin mutations. The cause of the disease in the mutation-negative sibling (II.4) remains to be investigated. Haplotype analysis confirmed that II.4 is a sibling of the unaffected Individual II.2. Similarly, we identified an affected but mutation-negative family member in Family A-6. A conceivable explanation for these findings might be the occurrence of additional, yet to be identified, mutations in Parkin or another gene.
A common founder is the most likely explanation for the deletion of exons 3 to 4, which was identified in two patients from Puerto Rico who share the entire PARK2 haplotype. The substitution 924C→T in exon 7 was detected exclusively in European patients and might result from a more distant founder, as suggested by a rare allele at D6S30526 and the possibility of identical haplotypes at markers close to the mutation-bearing exon 7. A common founder also may be an explanation for a third recurrent mutation (Ex3del), which was detected in two probands of European origin (see figure 2).
A heterozygous mutation was found in 12% of our probands (6/50) and in several definitely or possibly affected family members of Group A despite an extensive mutational analysis.5 The idea that even heterozygous mutations may lead to disease is strengthened by three observations: 1) the changes in FDOPA uptake identified in PET studies21; 2) the function of Parkin as an enzyme,17 which may explain a mechanism of haploinsufficieny for developing the disease; and 3) the similarly high percentage of heterozygous mutation carriers in another large study.4 In addition, the preliminary observation of differences in age at onset between individuals with PD according to their mutational status may add weight to this theory. Compound heterozygotes had the earliest onset; heterozygotes, later; and mutation-negative individuals, the latest age at onset. A single Parkin mutation might cause or increase susceptibility to the disease, possibly resulting in later onset PD, with other environmental or genetic factors precipitating the development of PD. Such genes may encode proteins, which are substrates for Parkin, and possible candidate genes for this theory may be localized at the recently identified PARK6 and PARK7 loci.11,12⇓
Small deletions and exon deletions or duplications in the Parkin gene probably result in the absence of functional gene product because of frameshift or alterations of splicing. In this context, notice that all of these deletions or duplications identified in our patient cohort were localized in the first five exons, encoding proteins most likely lacking the RING1-IBR-RING2 motif. This domain, especially RING2, was shown to be responsible for ubiquitination of target proteins.15-17⇓⇓ In addition, two of the identified missense mutations (1390G→A [Gly430Asp] and 1411C→T [Pro437Leu]) affect RING2, and two mutations (867C→T [Arg256Cys] and 924C→T [Arg275Trp]) are considered to alter the RING1 sequence. Another mutation (226G→C [Arg42Pro]) is localized in the amino-terminal ubiquitin-like domain. The role of the three other single basepair alterations remains to be investigated. The substitution 81G→T is part of the 5′ untranslated region (5′-UTR) and might alter transcription or translation efficiency or may be a rare polymorphism because it was found in another (healthy) individual in the heterozygous state. The 884A→G in exon 7 does not result in an amino acid change (Leu261Leu) but was not found in any control subject or patient. This implicates a rare polymorphism or a not yet understood mutational mechanism that might result in an aberrant protein, possibly because of the activation of a cryptic splice site. Similarly, the pathologic mechanism of the mutation 675A→C (Met192Leu) is unknown but may lead to alterations of the conformation of Parkin.
Taken together, mutations of the Parkin gene are a relatively frequent finding among patients with EOPD, especially in patients with a positive family history. Mutations were found in probands from a range of ethnic backgrounds, and further studies must be undertaken to elucidate the frequency of Parkin mutations in non-whites. The wide spectrum of mutations and the large size of the gene make a molecular diagnosis difficult. It remains to be investigated whether the considerable percentage of mutation-negative patients carry undetected mutations, possibly in the promoter or in the intron sequences, or whether one or more other genes and loci are responsible for the development of the disease. However, we favor the possibility that even a heterozygous Parkin mutation may lead to or increase the susceptibility to develop PD.
Acknowledgments
Supported by the Deutsche Forschungsgemeinschaft (Kl-1134/2-1 - K.H., C.K.), the Parkinson’s Disease Foundation (C.K., T.L., K.M., S.F., H.M.), NIH (K08 to T.L., NS36630 to K.M., RR00645 to K.M.), the Irving Scholarship (T.L.), the Lowenstein Foundation (T.L.), ARDAD (T.L.), NARSAD (T.L.), and a Galen Fellowship (T.L.).
Acknowledgment
The authors thank the patients and family members who participated in this study and the physicians at the Center for Parkinson’s Disease at Columbia University: Paul Greene, MD, Steven Frucht, MD, Cheryl Waters, MD, Blair Ford, MD, and Elan Louis, MD.
- Received September 4, 2001.
- Accepted January 2, 2002.
References
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
