|Home | About | Journals | Submit | Contact Us | Français|
To assess the frequency and clinical characteristics of carriers of previously identified mutations in six genes associated with early onset Parkinson disease (EOPD) and provide empirical data that can be used to inform genetic counseling.
Mutations in SNCA, PRKN, PINK1, DJ1, LRRK2 and GBA were assessed in 953 individuals with EOPD ascertained based on age at onset (AAO) ≤50 years. Participants included 77 Hispanics and 139 of Jewish ancestry. A validated family history interview and the Unified Parkinson’s Disease Rating Scale (UPDRS) were administered. Demographic and phenotypic characteristics were compared among groups defined by mutation status.
One hundred and fifty eight (16.6%) had mutations including 64 (6.7%) PRKN, 35 (3.6%) LRRK2 G2019S, 64 (6.7%) GBA and one (0.2%) DJ1. Mutation carriers were more frequent among cases with AAO ≤30 than among cases with AAO between 31 and 50 (40.6% vs. 14.6% p<0.001), Jews compared to non-Jews (32.4% vs. 13.7% p<0.001) and those reporting a first degree family history of PD than among those who did not (23.9% versus 15.1% p=0.012). Hispanics were more likely to be PRKN carriers than non-Hispanics (15.6% versus 5.9% p=0.003). The GBA L444P mutation was associated with a higher mean UPDRS-III score after adjustment for covariates.
EOPD individuals of Jewish or Hispanic ancestry, those with AAO ≤ 30, and those with a family history of PD in a first-degree relative may benefit from genetic counseling.
Early onset Parkinson Disease (EOPD) may be defined as disease onset before age 40 or 50.1–3 Mutations in several genes have been associated with EOPD, including alpha-synuclein (SNCA), Parkin (PRKN), PTEN induced putative kinase 1 (PINK1), DJ1, leucine-rich repeat kinase 2 (LRRK2) and glucocerebrosidase (GBA).4 Study inclusion characteristics such as age at onset (AAO), family history of PD (FHPD), and the ethnicity of participants influence the prevalence of specific mutations.5 Ideally, the frequency of specific mutations in EOPD would be assessed in a population-based sample. The annual incidence rate of EOPD is estimated at 3 per 100.000 person year,6 making a population based genetic investigation impractical and there are no population based studies of genetic contributions to EOPD. All but three7–9 published studies of the genetic contributions to EOPD have focused on familial PD10–13 or were studies of sporadic PD focusing only on a single gene: PRKN,14 PINK1,15 DJ116, LRRK217 and GBA.18 A recent Dutch EOPD study evaluated 187 subjects with AAO ≤50 for mutations in SNCA, PRKN, PINK1, DJ1 and LRRK2 and found 4% (n=7) who carried an identified mutation;8 however GBA was not evaluated in that study. The goal of the present study was to examine carriers of genetic mutations in EOPD recruited without regard to family history of PD. Here, we present the frequency and the clinical characteristics of carriers of known mutations in 953 subjects, evaluated as part of the Consortium of Risk for Early Onset PD study (CORE-PD), a multicenter study of PD cases ascertained solely because their AAO was ≤50 rather than based on FHPD, all of whom were clinically evaluated.
Cases with AAO ≤50 (n = 953) included consecutive EOPD patients who met research criteria for PD (UK brain bank),19 in 13 cites participating in the CORE-PD study.20 A Mini Mental State Exam (MMSE)21 >23 was required for study inclusion so that accurate self-report data could be obtained. Institutional review boards at all participating sites approved the protocols and consent procedures. Data collected included demographic information, AAO, the Unified Parkinson’s Disease Rating Scale (UPDRS)22 in the “on” state, completed by a movement disorders specialist, MMSE, and a FHPD using a previously validated interview23 however, FHPD was not an inclusion criterion. Ethnic group was documented by self-report using the format of the 1990 United States Census.24 A blood sample - from which DNA was isolated - was sent to the NINDS Human Genetics Resource Center DNA and Cell Line Repository (http://ccr.coriell.org). Examiners were unaware of cases’ genetic status. Cases were subsequently divided into mutation groups based on their genetic status. Cases were classified into motor subtypes based on previously described methodology: tremor dominant (TD), postural instability and gait difficulty (PIGD), or intermediate.25
PRKN: cases were screened for PRKN mutations by denaturing high performance liquid chromatography (DHPLC) (WAVE Transgenomic), as previously described.26 Amplicons were either directly sequenced (n=126) or analyzed (n=827) using a PRKN genotyping array27 in DNA samples with abnormal DHPLC elution profiles. Cycle sequencing was performed on the purified PCR product as per the manufacturer’s instructions (BigDye, Applied Biosystems). Products were processed on an ABI3700 or ABI3730 genetic analyzer. Chromatograms were viewed using Sequencher (Genecodes) and sequence variants determined. To identify genomic deletions and exon rearrangements in PRKN, semi-quantitative multiplex PCR was performed as previously described.28 Further details are described elsewhere.29
LRRK2: For analysis of LRRK2 mutations, DNA samples from the majority of the cases (n=515) were analyzed using a previously described genotyping array 27 (Asper Biotech, Tartu, Estonia). For the remaining 438 cases, genotyping was performed using Matrix-Assisted Laser Desorption/Ionization Time of Flight (MALDI-TOF) mass spectrometry (Sequenom) as previously described.5 All samples analyzed by either the MALDI-TOF or the genotyping array were examined for the mutations G2019S, R1441C, I2020T and Y1699C. Samples analyzed with the genotyping array were also tested for the G2385R variant.
GBA: 90 cases were previously reported and complete sequencing of the GBA gene was performed.20 The remaining cases (863) were analyzed for L444P and N370S mutations: 515 cases were genotyped using the genotyping array, and the remaining cases (n=348) were analyzed by direct sequencing using methods previously described.20
SNCA DJ1 and PINK1: A subset of 515 cases was analyzed using the genotyping array for the point mutations: A157T, A88P and E136K in SNCA. L166P, M26I, D149A and A104T in DJ1 and W437X and G309D in PINK1.
Cases were divided into six groups based on mutation status: (1) non-carriers of mutations of any of the genes, (2) PRKN heterozygotes, (3) PRKN homozygotes and compound heterozygotes combined, (4) LRRK2 G2019S carriers, (5) GBA N370S carriers and (6) GBA L444P carriers. DJ1 mutation carriers, LRRK2 G2385R carriers and carriers of mutations in more than one gene were too rare to be analyzed. Instead, carriers of these mutations are described clinically. Demographic and disease characteristics of the six groups were compared using analysis of variance and chi square tests as appropriate. In order to test whether mutation status (e.g. PRKN) is associated with severity of motor impairment, we used a linear regression model where UPDRS-III score was the dependent variable and several mutations were independent predictors (as compared to non-carriers), adjusting for disease duration, gender, age, daily levodopa dose and history of surgical intervention for PD. Mutation carrier frequency was then assessed in strata defined by decade of AAO (≤20, 21–30, 31–40, 41–50 years), and by Jewish ancestry (defined as at least one Jewish grandparent, given the high LRRK2 and GBA mutation frequency previously reported among Jews 30).
One hundred and fifty eight subjects (16.6%) carried a genetic mutation: 33 were PRKN heterozygotes, and 27 were PRKN homozygotes/compound heterozygotes.29 Thirty one were LRRK2 G2019S carriers (30 heterozygotes, one homozygote). One subject carried the LRRK2 G2385R variant; no other pathogenic LRRK2 mutations were found. Fifty eight were GBA mutation carriers: 18 L444P heterozygotes, 38 N370S heterozygotes and two N370S homozygotes. One subject (previously described16) had a single A104T DJ1 mutation. No carriers of the tested SNCA or PINK1 mutations were identified.
Demographic and disease characteristics of the carriers are described in Table 1. Carriers of PRKN mutations were more likely to be Hispanic, and to have an earlier AAO than non-carriers. Motor phenotype was computed on 707 cases who completed the UPDRS-II and UPDRS-III. LRRK2 G2019S carriers were more likely to manifest the PIGD motor phenotype,25 and GBA L444P carriers had higher scores on the UPDRS-III, indicative of more significant motor impairment, when compared to non-carriers of L444P. In a linear regression model, L444P mutation status was associated with worse motor performance, as represented in a higher UPDRS-III score (OR 7.4, 95%CI 2.0–12.8, p=0.007), after adjustment for age, gender and disease duration, daily levodopa dose and history of surgical intervention for PD (analysis included 738 cases on whom data on all covariates were available). There was no significant difference in the presenting symptom of PD among the groups; 44% presented with rest tremor,13% with stiffness and 7% with action tremor. Six percent reported symmetric findings at presentation.
Demographic and disease characteristics of seven individuals who carried more than one mutation are described in Table 2. None of these subjects reported a FHPD in a first-degree relative.
The distribution of mutation carriers by AAO is presented in Table 3. Overall, mutation frequency increased with younger AAO (Table 3, p<0.001). Mutation carriers were more frequent among cases with AAO ≤30 than among cases with AAO between 31 and 50 (40.6% vs. 14.6% p<0.001, Fisher exact test). AAO was significantly younger among PRKN mutation carriers than among carriers of all other mutations (p <0.001).
Hispanics were more likely to be PRKN carriers than non-Hispanics (15.6% versus 5.9% p=0.003, Fisher exact test). One hundred and thirty nine subjects (14.6%) reported Jewish ancestry (126 reported all four grandparents were Jewish, 2 reported three grandparents, 9 reported two grandparents and 2 reported one grandparent). Overall, mutations were more common in individuals with Jewish ancestry than in those without (32.4% versus 13.7%, p<0.001 Fisher exact test). The association was driven by LRRK2 G2019S carriers and GBA N370S carriers who were more likely to be Jewish than non-Jewish (Table 4). Of the 15 G2019S carriers who did not report Jewish ancestry, data on grandparents’ country of origin was available on 11. Four reported Southern European decent (Italy or Portugal) and 4 reported mixed European decent (including grandparents from Germany, Hungary, Ireland, Scotland and England). One reported Eastern European ancestry (Russia and Latvia) and two carriers reported Puerto-Rican ancestry.
In the entire cohort, mutations were more common in subjects with a FHPD than in those without a FHPD (23.9% versus 15.1%, p=0.012, Fisher exact test) (Table 4). However, this association was much weaker in Jews, where 31.8% of those with FHPD and 32.2% of those without a FHPD carried an identified mutation (p=1.0, Fisher exact test).
This large, uniformly clinically characterized sample of PD patients provides an estimate of the frequency of mutation carriers at movement disorders centers in tertiary referral settings. Although these data were not derived from a population-based sample, the results may guide clinicians in their assessment of the need for genetic counseling, particularly in specific race/ethnic groups. We have provided reference tables, stratifying mutations frequency by AAO, FHPD and Jewish ancestry. PRKN carriers were more likely to report a FHPD than LRRK2 carriers, despite the fact that LRRK2 is autosomal dominant and PRKN is considered a recessive disorder. This finding could be explained by siblings of PRKN carriers who may develop PD at a younger age than siblings of other mutation carriers (i.e. LRRK2 siblings may develop the disease at a later age) and by incomplete penetrance of the G2019S mutation.17 Also, the high frequency of PRKN heterozygotes may support the notion that even a single PRKN mutation is a risk of EOPD.29 In addition to demographic differences among mutation carriers, we found that LRRK2 G2019S carriers were more likely to manifest the PIGD motor phenotype31 and GBA L444P carrier status was associated with a higher UPDRS-III score (implying worse motor function) than non-carriers. While the odds of developing PD may be greater in L444P carriers than in N370S carriers,30 this is the first report of a difference in severity of motor signs among PD cases with different GBA mutations. Larger studies, including exams in the “off” state, are required to assess whether this statistical difference is a true clinical phenotype or a consequence of multiple comparisons.
The higher frequency of mutations detected here, 16.6%, compared to a previous Dutch study,8 4%, may be explained by the inclusion of GBA genotyping (which accounts for 40% of the mutations detected here), and by the higher proportion of subjects of Jewish and Hispanic origin in our study. These populations are known to have a higher risk of GBA and LRRK2 (Jewish and Mediterranean Basin) and PRKN (Hispanic) mutations.29 Previously, when a subset (n=90) of our EOPD sample (AAO≤50) was compared to late onset PD (n=185, AAO >50), GBA mutations were more prevalent among EOPD.5, 20
The high proportion of subjects with identifiable genetic mutations in GBA, LRRK2, and PRKN emphasizes the importance of these genetic factors in the etiology of EOPD. Despite the growing evidence that EOPD frequently has a genetic basis, the role of clinical testing remains controversial.32 Clinical testing may help to explain the etiology of PD in an individual with EOPD, and may be particularly useful for that purpose in patients of Jewish or Hispanic ancestry, those with onset under age 30, and those with a family history of PD in a first-degree relative. Finally, although the probability that an identified genetic mutation has been transmitted by an affected parent to a child can be calculated, the probability that that child will one day develop Parkinson disease (penetrance) remains unknown for any of these mutations. Therefore, there is little or no role for genetic testing for PD in family planning, predictive, or prenatal testing at this time. There has been a single study on attitudes toward genetic testing in PD33 and no studies on the influence of PD mutation carrier status on reproductive choices.
We have previously reported the frequency of LRRK231, PRKN29 and a subset of GBA carriers separately. Simultaneous comparison among carriers and GBA, SNCA and PINK-1 genetic data is unique to this study. A limitation of this study is the lack of inclusion of cognitively impaired cases, since we only included patients with MMSE scores >23. Each gene was not completely sequenced. The PRKN gene was completely sequenced in 126 of 953 cases, and only selected mutations were examined in LRRK2, GBA, PINK1, SNCA and DJ1. Therefore, we may have underestimated the proportion of mutation carriers in the genes we studied, particularly GBA in non-Jews.34 We did not examine the ATP13A2 (PARK9); however it is rarely associated with EOPD.35
In summary, EOPD individuals of Jewish or Hispanic ancestry, those with AAO ≤ 30, and those with a family history of PD in a first-degree relative may benefit from genetic counseling. Because the vast majority of individuals did not carry a known mutation, further studies are needed to identify additional genetic and environmental risk factors in this population.
This study was funded by NIH NS36630, UL1 RR024156 (KM), NS050487, NS060113 (LNC), P50 NS039764 (WKS) and the Parkinson’s Disease Foundation (KM, SF and LNC). RNA was supported by a Parkinson’s Disease Foundation H. Houston Merritt Fellowship in Movement Disorders. The authors thank Paul Greene MD and Diana Ruiz BS for their assistance.
Disclosure: The authors report no conflicts of interest.