Abstract
Background: Mutations in GBA and LRRK2 genes have been implicated in Parkinson disease (PD), particularly in Ashkenazi Jews.
Methods: An Israeli Ashkenazi cohort of 420 patients with PD, 333 elderly controls, and 3,805 young controls was screened for eight GBA mutations, which are associated with mild (N370S, R496H) and severe (84GG, IVS2 + 1, V394L, D409H, L444P, RecTL) Gaucher disease. Patients with PD and elderly controls were also genotyped for LRRK2 G2019S.
Results: GBA carrier frequency was 17.9% in patients with PD compared to 4.2% in elderly and 6.35% in young controls. The proportion of severe mutation carriers among PD patient GBA carriers was 29% compared to 7% among young controls. Severe and mild GBA mutations increased the risk of developing PD by 13.6- and 2.2-fold, and affected the average age at PD onset (AAO), 55.7 and 57.9 years, compared to 60.7 years in patients without known GBA or LRRK2 mutations.
Conclusions: These data demonstrate genotype-phenotype correlations between different GBA mutations and Parkinson disease (PD) risk and AAO in Ashkenazi Jews. Additionally, an earlier AAO was observed in LRRK2 G2019S carrier PD patients. Finally, these data demonstrate that a surprisingly high frequency, more than one third of our patient population, carried a mutation in GBA or LRRK2.
Glossary
- AAO=
- age at onset;
- ANOVA=
- analysis of variance;
- GD=
- Gaucher disease;
- HWE=
- Hardy-Weinberg equilibrium;
- PD=
- Parkinson disease.
Clinical observations, neuropathologic evidence, and genetic studies in the last decade have implicated mutations in the β-glucocerebrosidase (GBA) gene in parkinsonian phenotypes and in Parkinson disease (PD) susceptibility. Mutations and rearrangements in GBA cause Gaucher disease (GD), a recessively inherited deficiency of the lysosomal enzyme, glucocerebrosidase. GD is most common in the Ashkenazi Jewish population, where the 1226A > G (N370S) mutation predominates. Common GBA mutations have been classified as null, severe, or mild, based on their phenotypic effect. Severe and null mutations cause neuronopathic forms of GD, while mild mutations are conventionally associated with the nonneuronopathic form of disease.1,2
Parkinsonian manifestations reported in genotypically heterogeneous patients with GD,3–11 together with family studies revealing a significant frequency of parkinsonian symptoms in obligate or confirmed GBA mutation carrier relatives of patients with GD,6,9,12 suggested an association between GD and PD. Moreover, brain samples from autopsy-confirmed PD cases revealed significantly higher carrier frequencies than the estimated GBA mutation carrier frequency in the general population,13,14 supporting an association between GBA mutations, in both homozygotes and heterozygotes, and PD susceptibility.
The high carrier frequency of GBA mutations in Ashkenazi Jews compared to the population at large suggests the importance of this population for the study of the relationship between GBA mutations and PD. Previous publications reported GBA mutation carrier frequencies of 31.3% and 10.6% in 99 and 150 Israeli Ashkenazi patients with PD,15,16 and a carrier frequency for N370S of 10.7% in 160 Jewish American patients with PD,17 two- to fivefold greater than frequencies observed in their ethnically matched controls. Subsequent independent studies in various ethnic populations also supported the association between GBA mutations and increased PD risk.18–21 Two additional studies found no such association.22,23
Herein, using the largest cohort of Jewish Ashkenazi patients with PD analyzed to date, we aimed to validate the association between GBA mutations and PD, and explore the effects of mild and severe GBA mutations on PD risk and phenotype. Additionally, we explored the potential relationship between GBA mutations and the LRRK2 G2019S mutation in patients with PD of Ashkenazi origin.
METHODS
Patient population.
The study population included 420 unrelated patients with PD of Jewish Ashkenazi ancestry, 62.4% men (n = 262) and 37.6% women (n = 158), treated at the Movement Disorders Unit, Tel-Aviv Sourasky Medical Center. All patients were diagnosed by a movement disorder specialist, according to UK PD Society Brain Bank criteria.24 Patients were enrolled consecutively between September 2005 and January 2007. The mean age at enrollment was 67.96 years (SD ± 10.2, range 41–96), with an average age at enrollment of 67.7 (±10.3) and 68.4 (±10.0) years for men and women.
All patients underwent a detailed interview to ascertain ancestry, family history of PD or other movement disorders, presenting symptoms, and age at onset of motor symptoms that retrospectively could be associated with PD onset. The mean age at onset (AAO) of patients was 59.4 (±11.3; range 21–94) years. Eighty patients (19.3%) had symptoms of PD before the age of 50 and are considered as having early onset PD. The mean AAO for men and women did not differ: 59.3 (±11.4) and 59.6 (±11.3) years. The mean age at PD diagnosis was 60.7 years (±10.9; range 26–94).
Family pedigrees were constructed for all patients. None were from consanguineous families. Familial PD, defined as having at least one first- or second-degree relative with a diagnosis of PD, was noted in 26.7% (111/416) of patients with information regarding family history, 16.8% of which had an affected first-degree relative.
Our cohort was also tested for the LRRK2 G2019S (6055G>A) mutation but not for mutations in the recessively inherited Parkin, PINK1, or DJ1 genes. The status of LRRK2 G2019S in part of this cohort was previously reported.25
Control population.
A total of 4,138 Ashkenazi control individuals were analyzed. These included 3,805 anonymous young individuals, aged 20–45, mostly women, who underwent routine genetic screening tests for GD at the Genetic Institute, Tel-Aviv Sourasky Medical Center, between January 2000 and January 2007. These individuals represent an average risk population for carrying GBA mutant alleles in Israel.
An additional 333 older Ashkenazi controls were tested (mean age at enrollment 65.3 years [±10.2], range 38–89 years, data were not available for one individual), including 173 who were interviewed, mostly spouses of patients with PD, and 160 anonymous healthy controls, whose DNA samples were purchased from the National Laboratory for the Genetics of Israeli Populations (described in detail in the NLGIP Web site). Neurologic examinations were not performed on control individuals; however, controls reporting tremor, balance problems, or movement and gait disturbances were excluded. Of the 333 older Ashkenazi controls, 204 were randomly selected to match in sex and age to the PD group. This age- and sex-matched control group included 58.3% men and 41.7% women with an average age of 66.8 years (±10.6) at study enrollment.
All patients and spouse controls signed an informed consent before entering the study. Patients and elderly controls with clinically significant dementia who could not sign an informed consent were excluded from our study. All DNA samples were coded and tested in an anonymous manner. The Institutional and National Supreme Helsinki Committees for Genetic Studies approved the study protocols and the informed consents.
GBA and LRRK2 mutation analyses.
Genomic DNA was isolated from peripheral blood using standard protocols, or from saliva according to manufacturer's instruction (Oragene, Ottawa, Canada). Patients and controls were tested for the 84GG, IVS2 + 1, N370S, D409H, L444P, V394L, and RecTL GBA mutations, using PRONTO Gaucher kit (Pronto Diagnostics, Rehovot, Israel,26) and for the R496H mutation, using PRONTO Gaucher 496 MUT-only kit. The 3,805 anonymous young controls were not tested for the R496H mutation, which is not included in the GBA-mutation screening panel recommended by the Israeli Society of Medical Geneticists.
Alleles separation in RecTL carriers was performed using specific primers (primers p1 and 271, supplied by Dr. Nir Navot, Pronto Diagnostics, Rehovot, Israel), in order to delineate whether the N370S mutation is in cis (RecTL,N370S/+) or trans (N370S/RecTL) with D409H and L444P.
All samples carrying GBA mutations were tested by another method to confirm the results of the analysis above. Validations of the 84GG, IVS2 + 1, L444P, V394L, and R496H mutations were performed by sequence analysis (Applied Biosystems, Foster City, CA). Validation of the N370S mutation was performed using XhoI restriction enzyme analysis (New England Biolabs, Beverly, MA). All PCR reactions for validation were performed using Sigma Taq Polymerase (Sigma Aldrich, St. Louis, MO) and Biometra PCR systems (Biometra GmbH, Gottingen, Germany). All specific primer pairs used are detailed in table e-1 on the Neurology® Web site at www.neurology.org.
All patients with PD were screened for the LRRK2 6055G>A (G2019S) mutation as described previously.25
Statistical analysis.
Data are presented as mean (±SD) for continuous variables. Clinical and demographic categorical variables are presented in percentage whereas allele frequency is presented on a range of 0–1. Any differences among groups in continuous variables were tested using analysis of variance (ANOVA) and χ2 or Fisher exact test were used for comparison of categorical variables. Goodness of fit test with one degree of freedom was applied to look for any deviation from the Hardy-Weinberg equilibrium (HWE) among the 3,805 young controls and among patients with PD who were screened for GBA mutations. Logistic regression model was applied with the mutation status of the individual as the predictive variable and the disease status (PD/non PD) as the dependent variable. With this model, the Exp(β) was used to determine the calculated OR and the 95% CI around it. When comparing to the group of 3,805 young controls, an online calculator was used to determine the OR and CI (DJR Hutchon Calculator). SPSS software V. 15 (SPSS Inc., Chicago, IL) was used for all data analysis unless otherwise mentioned.
RESULTS
GBA mutations are more frequent in Ashkenazi patients with PD, with an over-representation of severe GBA mutations.
Of the 420 Jewish Ashkenazi patients with PD, 75 (17.9%) were carriers of one of the GBA mutations tested compared to 4.2% and 6.35% carriers among the elderly and young control populations (table 1, p < 0.0001). Six patients (1.4%) were either homozygous or compound heterozygous for GBA mutations. Sixty patients carried the LRRK2 G2019S mutation (14.3%), four of whom also carried the GBA N370S mutation. These four patients were excluded from additional statistical analyses to enable direct comparison among the three groups of patients with PD: those who carried either GBA (n = 71) or LRRK2 G2019S (n = 56) mutations and patients who were not detected as mutation carriers (n = 283, table 2).
Table 1 Frequencies of GBA gene mutations and alleles among Ashkenazi patients with Parkinson disease (PD) and control populations
Table 2 Comparison between Ashkenazi patients with Parkinson disease (PD) who are carriers of either GBA mutations or LRRK2 G2019S mutation and noncarrier patients
GBA mutations tested here are classified as mild (N370S and R496H) or severe (84GG, IVS2 + 1, V394L, D409H, L444P, and RecTL), depending on their expected and observed phenotypic effects in GD.1,2 The frequency of N370S carriers among patients with PD was higher than among the elderly and young control populations (10.95%, 3.30%, and 5.89%, p < 0.001). However, the frequency of severe GBA mutation carriers among patients with PD was much higher than the frequency of these mutation carriers among the elderly and young control populations (5.24%, 0.6%, and 0.45%, p < 0.001). Furthermore, while N370S carriers accounted for 93% of all GBA mutation carriers detected among the 3,804 young control individuals, they accounted for only 61% of PD patient mutation carriers (table 1, p < 0.001), demonstrating an over-representation of severe GBA mutations among the patients with PD. No deviation from HWE was found regarding the number of homozygous and compound heterozygous individuals among patients with PD and the young controls.
Carriers of severe GBA mutations have a particularly increased risk for developing PD.
The risk of heterozygous GBA mutation carriers to develop PD was calculated relative to an elderly control group (age 38–89 years) and a young control group (age 20–45 years). When PD patient carriers of GBA mutations were compared to the elderly controls using a logistic regression model (not including the LRRK2 G2019S carriers and individuals homozygous or compound heterozygous for GBA mutations), the Exp(β) was 5.6 (CI 3.1–10.1) when comparing the 354 patients with PD to 324 elderly controls, and 8.1 (CI 3.4–18.9) when comparing these patients to the 199 age- and sex-matched controls only. For an additional estimation of this risk, we calculated an OR of 3.7 (CI 2.8–5.0, table 3) for a mutation carrier to develop PD using the young control population as a reference, possibly reflecting the influence of aging.
Table 3 The average age at Parkinson disease motor symptoms onset and the OR depend on the type of GBA mutation
Since only about 1/250 Ashkenazis carry a severe GBA mutation, the number of mutation carriers among the 330 elderly controls would not have been sufficient for estimating the OR of a specific GBA mutation carrier to develop PD. Therefore, we performed this analysis relative to the 3,804 young controls (table 3). The OR for carriers of all five severe GBA mutations tested was more than sixfold higher compared to the OR of N370S carriers (13.6 [CI 7.2–25.9] and 2.2 [CI 1.5–3.1]).
Earlier age at PD onset in GBA mutation and LRRK2 G2019S carriers.
The age at PD motor symptoms onset (AAO) differed among the four groups presented in table 2 [one-way ANOVA, F(3,406) = 3.99, p = 0.008]. Post hoc analysis revealed that the source of significance was the younger AAO in the three groups of patients with mutations compared to patients without mutations (p = 0.021, p = 0.027, and p = 0.042). While the mean AAO of noncarriers was 60.7 years, it was 3.5, 3.7, and 9.5 years younger in these three groups of PD patients with mutations.
Furthermore, while the AAO in patients with PD carrying mild GBA mutations (n = 49) was 57.9 years (±10.9), it was 55.7 years (±8.9) in patients with severe GBA mutations [one-way ANOVA, F (2,347) = 2.97, p = 0.05]. Post hoc analysis revealed that the source of significance was the earlier age at PD onset (5.0 years younger) in the 22 patient carriers of severe GBA mutations compared to the 279 patients who did not carry GBA or LRRK2 mutations (p = 0.047). Table 3 further details the different AAOs of PD according to the specific GBA mutation.
Are there distinctive clinical characteristics for PD patient carriers of GBA mutations?
The predictive value of a family history (first- or second-degree relative with PD) for the development of PD was calculated using a binary regression model with the status of the subject (patient or control) as the dependent variable. When comparing all Ashkenazi patients with PD to elderly controls (data were available for 416 and 169 individuals), the OR for a family history was 4.8 (CI 2.5–8.9). Additionally, the frequency of a family history of first- or second-degree relatives with PD in the three groups of patients with mutations differed (25.7%, 43.6%, and 22.8%, χ2 = 15.216, df = 4, p = 0.004, table 2) when compared to noncarrier patients, similar to the proportions of patients with PD with only first-degree relatives (15.7%, 33.9%, and 13.2%, χ2 = 15.842, df = 4, p = 0.003, table 2). While the ratio of first- or second-degree relatives among patients who carried the LRRK2 G2019S mutation was higher, as previously reported,25 there was no difference in the frequency of a family history of PD between patients with GBA mutations and patients who did not carry mutations.
When analyzing the presenting symptoms of PD (table 4), a tendency toward a higher frequency of weakness (χ2 = 6.51, df = 1, p = 0.01) and a lower frequency of rigidity (χ2 = 3.99, df = 1, p = 0.046) was noted in GBA mutation carriers. However, these results should be interpreted with caution, as applying Bonferroni correction for multiple comparison sets the cutoff p value from the a priori p = 0.05 to p = 0.006. Therefore, the over-representation of weakness and under-representation of rigidity as presenting symptoms in GBA mutation carriers is only of borderline significance.
Table 4 Presenting symptoms of Parkinson disease in patient carriers of GBA mutations and noncarrier patients
DISCUSSION
The well-defined Ashkenazi population, which carries a significantly high frequency of mutant GBA alleles, is extremely valuable for validation and further elucidation of the relationship between GBA mutations and PD risk. The first association study, reporting a 31.3% GBA carrier frequency in Jewish Ashkenazi patients with PD compared to 4–6% in ethnically matched controls,15 generated considerable interest. However, the discrepancy between these results and threefold lower frequencies in two ethnically similar cohorts, 10.6% in a multicenter Israeli study of 150 Jewish Ashkenazi patients with PD16 and 10.7% in a Jewish American cohort of 160 patients with PD,17 raised a number of questions regarding the validity of the association. Our study resolves this discrepancy, confirming that genetic variance in GBA is a risk factor for PD.
Our cohort afforded a sufficient frequency of variant GBA alleles to permit the comparison of carriers of mild vs severe GBA mutations. Due to the increased representation of severe mutations in the PD group, the proportion of mild vs severe mutations varied significantly between GBA carriers in the young control and patient populations. We suggest that the reduced proportion of N370S among PD patient GBA carriers might explain the 10.7% GBA carrier frequency reported previously in a Jewish PD cohort screened only for the N370S mutation.17
The marked overrepresentation of severe GBA mutations among patients with PD detected here allowed us to examine the possible clinical implications of carrying mild vs severe mutations. Carriers of severe mutations had a substantially increased disease risk and decreased AAO, compared to carriers of mild mutations, while patients homozygous or compound heterozygous for GBA mutations had the earliest AAO. Of note, previous studies in Ashkenazi patients,15 ethnic Chinese patients,21 and autopsy samples from American PD patients13 reported the influence of GBA carriage on early age at onset, and the development of parkinsonian symptoms in patients with GD has been reported mainly in patients' fourth and fifth decades.3–11 Moreover, when calculating the average AAO of all homozygous and compound heterozygous GBA carriers published in these studies, together with the six patients from our cohort (a total of 37 patients), the average age at PD onset was 49.0 years, approximately 10 years earlier than the average onset of PD.
The high frequency of GBA mutant alleles in the Ashkenazi population together with observations of increased PD risk and early AAO among Ashkenazi carriers suggest that GBA-associated PD might represent a significant health risk in this population. Interestingly, while epidemiologic research has confirmed that PD occurs worldwide, with a higher prevalence rate among Caucasian populations, individuals of Jewish Ashkenazi ancestry have not been defined as a particularly high risk group for disease.27 While our observations suggest that carriers of severe mutations have a significantly increased risk for disease, only about 1/250 Ashkenazis carry these mutations, therefore challenging their impact on PD frequency at the population level. On the other hand, carriers of the mild GBA mutations are more common, about 1/17, but have only a twofold increased disease risk. Furthermore, since PD is a disease of advanced age, carriers might have other lethal diseases prior to the onset of PD. Additionally, since the vast majority of GBA carriers and individuals with homozygous or compound heterozygous GBA mutations never develop PD,28–30 it is likely that disease risk is influenced by additional genetic modifiers and environmental factors. These issues must be carefully considered, together with the absence of proven presymptomatic medical or behavioral interventions that can modify the natural history of PD, before meaningful genetic counseling can be offered to thousands of young Jewish individuals who perform genetic screening tests for GBA mutations.
Mutations in the LRRK2 gene are the most common genetic determinant of PD identified to date.31 The significant frequency of the common LRRK2 G2019S mutation in Ashkenazi Jewish patients with PD25,32,33 suggests a possible relationship between LRRK2 G2019S and GBA variants in PD risk in Ashkenazi Jews. Given the frequency of LRRK2 G2019S in our patient population,25 12 of the 81 GBA mutation carriers detected herein were also expected to harbor LRRK2 G2019S. The detection of only four such patients (two tailed χ2 = 5.49, df = 1, p = 0.04), who lacked remarkable clinical symptoms or an earlier AAO, may suggest that these founder mutations occurred in distinct Ashkenazi Jewish subpopulations, who due to sociogeographic factors have not been fully mixed. Interestingly, the separate analysis of LRRK2 G2019S carrier, GBA carrier, and noncarrier patients with PD in our cohort revealed an earlier age at disease onset in LRRK2 G2019S carriers, compared to noncarriers, a finding that was masked in our previous analysis.25
Our data demonstrate that more than one third of our Ashkenazi patients with PD carried a mutation in genes that are associated with or contribute to the development of PD, GBA or LRRK2. However, since a substantial percentage of our cohort reporting a familial history of PD did not carry these mutations, it is likely that additional genetic factors influence PD susceptibility in this population. Genetic studies have provided tremendous insight into the molecular mechanisms underlying PD pathogenesis, suggesting roles for genes involved in putative pathogenic pathways including lipid and vesicle dynamics, the accumulation of aberrant or misfolded protein, impaired ubiquitin-proteosomal system, mitochondrial dysfunction, and the impairment of mechanisms protecting from oxidative stress and apoptosis.31,34–37 We therefore suggest that additional studies of patients with PD of Ashkenazi origin could be of great importance for further understanding of the proposed PD-causing and PD-risk genes, and the genetic contribution to PD.
ACKNOWLEDGMENT
The authors thank the patients and their spouses for their participation in this study. The assistance of Dr. Helena Yagev-More, Ilana Elroi, Drs. Balash, Merims, and Shabtai, and Orna Moore is acknowledged.
Footnotes
-
Supplemental data at www.neurology.org
Editoral, page 2272
e-Pub ahead of print on April 23, 2008, at www.neurology.org.
Supported by grants from M.K. Humanitarian Fund, National Parkinson Foundation, Miami, FL, Tel Aviv Sourasky Medical Center Grant of Excellence, and Wolfson and Kahn Foundations.
Disclosure: The authors report no disclosures.
Received August 15, 2007. Accepted in final form October 22, 2007.
REFERENCES
Disputes & Debates: Rapid online correspondence
NOTE: All authors' disclosures must be entered and current in our database before comments can be posted. Enter and update disclosures at http://submit.neurology.org. Exception: replies to comments concerning an article you originally authored do not require updated disclosures.
- Stay timely. Submit only on articles published within the last 8 weeks.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- 200 words maximum.
- 5 references maximum. Reference 1 must be the article on which you are commenting.
- 5 authors maximum. Exception: replies can include all original authors of the article.
- Submitted comments are subject to editing and editor review prior to posting.