?-Synuclein and Parkinson disease
S. Winkler, BS
J. Hagenah, MD
S. Lincoln, BS
M. Heckman, MS
K. Haugarvoll, MD
V. Kostic, MD
M. Farrer, PhD
C. Klein, MD
Background: Mutations in the ?-synuclein (SNCA) gene have been shown to be responsible for a
rare familial form of Parkinson disease (PD). Furthermore, polymorphic variants in multiple regions
of the gene have been associated with susceptibility to idiopathic PD in different populations.
Objective: To evaluate and to confirm the role of SNCA variants in PD pathogenesis.
Methods: We included 667 subjects (397 cases with idiopathic PD and 270 healthy, ethnically
matched controls) of Northern Central and Southeastern European origin. We analyzed geno-
types at 14 markers spanning the SNCA locus and its major haplotype blocks and conducted a
haplotype analysis for four promoter markers including the microsatellite marker Rep1.
Results: The three single nucleotide polymorphisms (SNPs) of the promoter region (rs2583988,
rs2619364, rs2619363) and a SNP in the 3=UTR (rs356165) of the SNCA gene showed the
greatest evidence for an association with PD (p ? 0.003), with significant pairwise values for
linkage disequilibrium (D= ? 0.74, r2? 0.29). The promoter haplotype “261-T-G-T” (Rep1-
rs2583988-rs2619364-rs2619363) was associated with disease (p ? 0.032). The most sig-
nificant association with PD was generated by excluding Rep1 (p ? 0.008). This association
remained significant when analyzing the Serbian patients separately and was of borderline signif-
icance for the German patients.
Conclusions: Our findings confirm that genetic variability within the SNCA locus is associated
with susceptibility to idiopathic Parkinson disease (PD). We found evidence for disease associa-
tion with single nucleotide polymorphisms at both the 5= and the 3= end of the gene with pairwise
linkage disequilibrium between them. The association was independent of the Rep1 status, and
one major SNCA promoter haplotype class seems to be associated with PD susceptibility.
SNCA ? ?-synuclein; CI ? confidence interval; HT ? haplotype-tagging; LD ? linkage disequilibrium; OR ? odds ratio; PD ?
Parkinson disease; SNP ? single nucleotide polymorphism.
Idiopathic Parkinson disease (PD) is the most frequent form of parkinsonism (approxi-
mately 75%) and usually refers to a syndrome characterized by late-onset, largely non-
heritable, classic parkinsonism, i.e., the symptom triad of bradykinesia, rigidity, and rest
tremor (not all features mandatory) with a therapeutic response to levodopa, and the
frequent development of motor complications.1The term parkinsonism is also often used
for hereditary forms of the disorder, of which dominantly inherited missense mutations
and multiplications of the ?-synuclein (SNCA) gene are an infrequent cause.2Neverthe-
less, ?-synuclein protein was identified as a major component of Lewy body and Lewy
neuritic inclusions, the pathologic hallmark of idiopathic PD.3
e-Pub ahead of print on September 12, 2007, at www.neurology.org.
From the Departments of Neurology (S.W., J.H., K.L.-H., C.K.) and Human Genetics (S.W., K.L.-H., C.K.), University of Lu ¨beck, Lu ¨beck,
Germany; Department of Neuroscience, Mayo Clinic, Jacksonville, FL (S.L., M.H., K.H., M.F.); and Department of Neurology, University of
Belgrade, Belgrade, Serbia and Montenegro (V.K.).
Supported by a grant from the Deutsche Forschungsgemeinschaft, the PD Foundation/National Parkinson Foundation, the
Bundesministerium fu ¨r Bildung und Forschung, and the University Lu ¨beck (all to C.K.). C.K. is supported by the VolkswagenStiftung. The
Morris K. Udall Center for Excellence (NIH P50 NS40256) supports Parkinson disease research at Mayo Clinic.
Disclosure: The authors report no conflicts of interest.
Supplemental data at
Address correspondence and
reprint requests to Dr. Christine
Klein, Department of
Neurology, University of
Lu ¨beck; Ratzeburger Allee 160,
23538 Lu ¨beck, Germany
Copyright © 2007 by AAN Enterprises, Inc.
Early association studies suggested poly-
morphic variants at the SNCA locus to
contribute to the risk of developing com-
mon, idiopathic PD, although much of the
literature has remained equivocal.4-13The
best examined genetic variant is Rep1
(D4S3481), a mixed dinucleotide repeat lo-
cated approximately 10 kb upstream of the
translational start of SNCA.14A recent,
collaborative pooled analysis of Rep1 data
from 2,692 PD cases and 2,652 unrelated
controls confirmed that the Rep1 allele
length contributes to disease risk.9
However, within defined populations,
other genetic variability in SNCA has been
implicated in PD susceptibility and age at
disease onset. For example, multimarker
haplotype analyses in a Flemish-Belgian
sample showed that genetic variability in
the SNCA promoter, other than Rep1, may
explain the association with PD.12Addi-
tional studies using a similar approach in
Southern German or Japanese samples
confirmed the SNCA genetic variability to
be associated with disease but highlighted
sequence variants at the 3= end of the
gene.10,11The haplotype block structure
across the SNCA locus in Caucasian and
Asian studies is strikingly different, which
suggests that these disparate results are un-
likely to be due to linkage disequilibrium
and rather based on independent muta-
To further evaluate and to confirm the
role of SNCA variants in PD pathogenesis,
we have analyzed genotypes from a sample
drawn from both Central and Southeastern
European populations. We focused on ge-
netic variability implicated in previous
studies, spanning the SNCA locus and its
major haplotype blocks.
METHODS Study subjects. A total of 397 cases with
idiopathic PD and 270 healthy, ethnically matched control
subjects were included in this study. Subjects either origi-
nated in Northern Germany (Northern Central Europe,
Group A: 234 PD patients and 160 control subjects) or in
Serbia (Southeastern Europe, Group B: 163 PD patients and
110 control subjects). A diagnosis of PD was established ac-
cording to the UK Brain Bank diagnostic criteria.15The Ger-
man control subjects underwent a detailed neurologic
examination, whereas the Serbian control subjects were not
the study by the local ethics committees, DNA was extracted
from venous blood samples using standard techniques.
Genotyping. Genotyping was performed for 14 markers
spanning haplotype blocks across the SNCA gene, using
linkage disequilibrium (LD) values previously defined11(fig-
ure). Markers were selected based on past reports, including
three single nucleotide polymorphisms (SNPs) and a micro-
satellite repeat Rep1 in the SNCA promoter,12and SNP
rs365165 toward the 3= end of the gene. The remaining 9
haplotype-tagging (ht) SNPs were distributed across the
120-kb SNCA locus.9,11,12HtSNP primers were ordered di-
rectly from Applied Biosystems or were designed using pub-
lic databases. After amplifying genomic DNA template by
PCR using an ABI Taqman PCR Mastermix, all products
were analyzed on an ABI7900 automated sequencer with the
SDS 2.02 allelic discrimination software (Applied Biosys-
tems). The microsatellite marker Rep1 was amplified as pub-
lished.4Marker positions are given with reference to the
forward, sense strand in the Human Genome Build 36.1 and
the dbSNP Build 126 (http://www.ncbi.nlm.nih.gov/SNP).
Statistical analysis. Haplotype analysis was performed
using S-Plus score tests for association,16with adjustments
made for age and sex. P values were obtained from the as-
ymptotic ?2distribution of the score statistic. Haplotypes
with an overall frequency of less than 1% were not consid-
ered. Individual associations between PD and each marker
were measured by odds ratios (ORs), and corresponding
95% CIs were obtained from logistic regression models ad-
justed for age and sex. In PD cases, linear regression models
adjusted for sex were used to examine individual associa-
tions between age at onset and each marker. All models in-
volving patients from both Germany and Serbia were
additionally adjusted for sample location. LD between
markers in study controls was defined by pairwise D= values
or not for LD analysis) as implemented in the Haploview
software.17Values of r2were also estimated as a measure of
LD. In addition, an LD unit map was constructed.18Conser-
vative values provided in the text were comparable between
patient and controls groups, between separate populations
and combined. Statistical significance was determined at the
5% level. For each family of statistical tests, multiple testing
was adjusted for by using the single-step minP procedure,19
with 10,000 permutations of case and control labels, to de-
termine the level of significance that controls the family-wise
error rate at 5%.19P values less than or equal to this level are
RESULTS Of a total of 667 study subjects, there
were 397 patients with idiopathic PD (40% wom-
en; mean age at onset 47 ? 11 years; range 30 to
80 years; mean age 55 ? 11 years; range 34 to 85
years) and 270 controls (50% women; mean age
51 ? 12 years; range 31 to 90 years).
The figure highlights the genomic positions of
the markers used, and all pairwise values of LD,
and illustrates the haplotype block structure
across the SNCA locus. Table 1 lists the estimated
ORs and associated 95% CIs between each of the
14 markers and PD. Thereby, SNPs rs2619363,
rs2619364, and rs2583988 located in the promoter
Neurology 69 October 30, 2007
region and SNP rs356165 more 3= showed the
greatest evidence for an association with disease
(p ? 0.003). These results were consistent when
considering the two major ethnic groups sepa-
rately. Pairwise values for LD between 5= pro-
moter SNPs and rs365165 were significant (D= ?
0.74, r2? 0.29). Many markers throughout the
gene showed evidence for linkage disequilibrium;
minor allele frequencies and pairwise values for
D= and r2were closely comparable for all mark-
ers, between populations, between patient and
control groups, and combined (data available on
request). None of the SNPs considered showed
evidence for an association with age at disease on-
set (data not shown).
Table 2 shows the results of the haplotype
analysis considering all four markers within the
SNCA promoter; p values given are for score tests
of overall association. Additional promoter hap-
lotypes were then constructed, dropping each
marker in a reiterative fashion. The haplotype
with the most significant association with PD was
generated by excluding Rep1 (p ? 0.008).
specific haplotypes in patients and control sub-
jects when all four markers were considered. The
haplotype “261-T-G-T” for (Rep1-rs2583988-
rs2619364-rs2619363) showed the greatest evi-
of 30% in patients vs 22% in controls (p ? 0.001).
Linkage disequilibrium map and haplotype structure of the SNCA locus, constructed using genotypes
from control subjects
From the top, a metric linkage-disequilibrium unit (LDU) map is shown spanning the SNCA locus on chromosome 4q21, in its
centromeric to telomeric orientation. Slopes highlight regions with ancestral recombination; plateaus indicate intervening
haplotype blocks. The middle panel shows the order of the 14 markers genotyped from 3= (left) to 5= (right). The coding region
of the gene is within rs2301134 and rs365165. The lower panel shows all pairwise linkage disequilibrium (LD) relationships
using the r2method (control groups, combined): Black and dark gray cells indicate strong LD, gray cells have intermediate LD,
and light gray and white cells have evidence for ancestral recombination.
Neurology 69 October 30, 2007
In subsequent analyses, we assessed the two ma-
jor ethnic groups separately, and similar results
were obtained (table 3). No deviations of the ge-
notype frequencies from Hardy–Weinberg equi-
librium were detected in either cases or controls.
The exact allele and genotype distributions in the
different samples for each of the polymorphisms
genotyped are given in tables E-1 through E-3 on
the Neurology®Web site at www.neurology.org.
DISCUSSION The samples we have assessed
originate in Northern Germany and Serbia, and
complement previous population studies in Eu-
rope, the United States, and Asia. Overall, our
findings corroborate that genetic variability
within the SNCA locus is associated with suscep-
tibility to idiopathic PD. Nevertheless, we recog-
nize that the use of “convenience” samples in
association studies is not ideal and may lead to
spurious findings.20Although similar trends were
observed in German and Serbian subgroups, our
power to detect an association was compromised
in stratified analyses. Large population series are
required, especially for subgroup analyses, and
our results and discussion are tempered accord-
ingly. However, the minor allele frequencies and
pairwise values for D= and r2we observed were
closely comparable for all markers, between pop-
ulations, and between patient and control groups,
and thus conclusions from combined analyses are
Single SNP associations with PD: Estimated odds ratios and 95% CIs
Overall (n ? 667) German samples (n ? 394) Serbian samples (n ? 273)
90986062–90986328 1.02 (0.83–1.26)0.82 1.19 (0.89–1.58)0.24 0.86 (0.62–1.18)0.34
90979851 1.47 (1.16–1.89)0.002 1.47 (1.06–2.05)0.023 1.49 (1.04–2.16) 0.035
909789101.46 (1.14–1.87)0.0031.47 (1.06–2.05) 0.023 1.45 (1.01–2.11)0.048
909780701.45 (1.14–1.86) 0.0031.46 (1.06–2.04)0.024 1.45 (1.01–2.10)0.047
909779681.24 (0.99–1.55) 0.059 1.15 (0.85–1.55)0.371.38 (0.99–1.94) 0.062
90959901 0.94 (0.72–1.23)0.660.84 (0.59–1.20) 0.34 1.10 (0.73–1.69)0.65
909418940.80 (0.64–0.99)0.045 0.87 (0.65–1.16)0.330.71 (0.50–1.01) 0.061
90924387 0.81 (0.61–1.08) 0.160.73 (0.49–1.08)0.11 0.91 (0.59–1.41)0.67
90912499 0.78 (0.55–1.11)0.16 0.72 (0.44–1.16)0.180.84 (0.50–1.44)0.52
909033011.00 (0.64–1.56) 0.981.24 (0.71–2.20) 0.460.66 (0.30–1.43) 0.28
908659091.48 (1.17–1.86)0.001 1.47 (1.09–2.01)0.013 1.46 (1.03–2.09)0.038
908614871.30 (1.04–1.62) 0.021 1.28 (0.96–1.71)0.0961.30 (0.92–1.85) 0.14
908585381.23 (0.80–1.94) 0.35 1.49 (0.86–2.67)0.160.85 (0.40–1.83) 0.67
90854043 1.39 (1.10–1.75)0.006 1.36 (1.00–1.86)0.051 1.41 (0.99–2.02)0.058
Odds ratios (ORs) correspond to a one-disease-associated allele increase for all markers except Rep1, for which the OR corresponds to a 2-bp increase in
combined allele length. P values result from univariable tests of association in logistic regression models adjusted for age, sex, and sample location. After a
permutation adjustment for multiple testing, p values ? 0.0060 (overall), 0.0061 (Germans), and 0.0080 (Serbians) are considered statistically significant.
SNP ? single nucleotide polymorphism; PD ? Parkinson disease.
Overall haplotype analysis of the SNCA promoter
Haplotype Overall (N?667)German samples (n ? 394) Serbian samples (n ? 273)
All four promoter markers
0.032 0.15 0.17
P values result from score tests of overall association adjusted for age and sex. Sample location was additionally adjusted for
in the overall group. After a permutation adjustment for multiple testing, p values ? 0.034 (overall), 0.039 (Germans), and
0.041 (Serbians) are considered statistically significant.
Neurology 69October 30, 2007
Most previous studies, including the large re-
cent meta-analysis, focused on the Rep1 and
other promoter variants. In contrast, we investi-
gated markers across the entire SNCA locus. Af-
ter applying a permutation method to adjust for
multiple testing, we found evidence for a disease
association at both 5= and 3= ends of the gene.
Although SNPs associated with disease are more
than 100 kb apart and in different haplotype
blocks, there was evidence for pairwise LD be-
tween them. The magnitude of the association at
both ends of the gene was also similar, with com-
parable ORs of approximately 1.5. However, it
was not possible to discern whether one popula-
tion sample contributed more to the 5= or 3= asso-
rs2619363 within the SNCA promoter are in al-
most complete LD (D= ? 0.99, r2? 0.97) with
each other and are associated with disease. It is
conceivable that their allelic variability may have
a direct effect on the quantitative, regional, and
temporal regulation of gene expression. In the
SNCA promoter, the major haplotype associated
with PD was “261-T-G-T,” corresponding to
marker alleles at Rep1, rs2583988, rs2619364, and
rs2619363. An association with this haplotype
had previously been observed in a Flemish-
Belgian series, and similar frequencies (29% in
patients vs 20% in control subjects) have been re-
ported.12However, these authors used a “gene-
based” metric to describe SNP markers in the
SNCA locus and labeled the associated haplotype
-“261-A-G-A”-; SNCA is transcribed from the an-
tisense strand in a telomeric to centromeric orien-
tation, and SNP entries published were kept
consistent with the public database. Herein, we
used the updated nomenclature from dbSNP con-
sistent with the chromosome 4q centromeric to
telomeric physical map (Build 36.1) and a sense
Rep1 variability has been reproducibly associ-
ated with PD. It may have a direct functional ef-
fect on transcription factor binding and may be in
LD with other disease-associated variability.
However, in this and previous studies, three ob-
servations are worthy of note: 1) promoter SNPs
are independently associated with PD, regardless
of the Rep1 status; 2) promoter SNPs are not in
significant LD with Rep1 alleles; and 3) one major
SNCA promoter haplotype class seems to be asso-
ciated with PD susceptibility. Hence, sequencing
and functional studies of this specific promoter
fragment may be worthwhile.
This study also implicates SNP rs356165 in the
3=UTR in PD susceptibility and suggests that car-
riers of a “G” allele are more likely to have PD.
rs356165 is in significant LD with the promoter
SNPs associated with disease (D= ? 0.68, r2?
0.29), but not with Rep1. The same marker/allele
has been associated with PD in both Caucasian
and Asian studies.10,11Preliminary studies on
SNCA gene expression, in relation to SNCA sus-
ceptibility genotypes, suggest that disease-
expression.10However, in these reports, several
adjacent SNPs from intron 4 to the 3=UTR pro-
vide as much evidence for association. Because
the SNP variability is noncoding and not within a
region of species-conserved sequence identity or a
miRNA binding site, the biologic mechanism re-
Gene expression may also be influenced by
epigenetic mechanisms, including methylation re-
cently implicated in the downregulation of SNCA
gene expression,21which may warrant further
With the exception of a single previous study,5
it is curious that SNCA variability does not seem
to be associated with age at disease onset. In con-
trast, in families with SNCA multiplications,
copy number and SNCA expression seem to be
directly related to the onset of motor symptoms,
disease severity, and duration to death.22,23One
explanation for this discrepancy could be that
data collected in heterogeneous, cross-sectional
Estimated frequencies for specific haplotypes within the SNCA promoter
Frequency in cases, %Frequency in controls, %
Haplotype*OverallGermans Serbians OverallGermans Serbians Overall
4042 37 45 47 42 0.035
242226 26 2726 0.47
3028 3322 21 250.001
* The four-marker haplotype is defined by 5=-Rep1-rs2583988-rs2619364-rs2619363-3=.
After a permutation adjustment for multiple testing, p values ? 0.013 are considered statistically significant.
Neurology 69October 30, 2007
population studies of PD may not be as compre-
hensive or reliable as that from longitudinal fam-
ily studies. Thus, the influence on age at onset of
specific SNPs may be clouded in the former.
There seem to be multiple mechanisms con-
tributing to the association of SNCA with PD,
presumably affecting its neuronal expression:
some may be stochastic, some environmental, and
some genetic or epigenetic. Attempts to correlate
SNCA genotypes with SNCA expression in brain
and peripheral tissues are warranted, because
agents that modulate SNCA expression may be
therapeutic options in both familial and idio-
Received February 15, 2007. Accepted in final form May 21,
1. Galpern WR, Lang AE. Interface between tauopathies
and synucleinopathies: a tale of two proteins. Ann
2. Farrer MJ. Genetics of Parkinson disease: paradigm
shifts and future prospects. Nat Rev Genet 2006;7:306–
3. Spillantini MG, Goedert M. The alpha-synucleinopa-
thies: Parkinson’s disease, dementia with Lewy bodies,
and multiple system atrophy. Ann NY Acad Sci 2000;
4.Farrer M, Maraganore DM, Lockhart P, et al.
?-Synuclein gene haplotypes are associated with Par-
kinson’s disease. Hum Mol Genet 2001;10:1847–1851.
5. Hadjigeorgiou GM, Xiromerisiou G, Gourbali V, et al.
Association of alpha-synuclein Rep1 polymorphism
and Parkinson’s disease: influence of Rep1 on age at
onset. Mov Disord 2006;21:534–539.
6.Holzmann C, Kruger R, Saecker AM, et al. Polymor-
phisms of the alpha-synuclein promoter: expression
analyses and association studies in Parkinson’s disease.
J Neural Transm 2003;110:67–76.
7. Izumi Y, Morino H, Oda M, et al. Genetic studies in
Parkinson’s disease with an alpha-synuclein/NACP
gene polymorphism in Japan. Neurosci Lett 2001;300:
8. Kruger R, Vieira-Saecker AM, Kuhn W, et al. In-
creased susceptibility to sporadic Parkinson’s disease
by a certain combined alpha-synuclein/apolipoprotein
E genotype. Ann Neurol 1999;45:611–617.
9. Maraganore DM, de Andrade M, Elbaz A, et al. Col-
laborative analysis of alpha-synuclein gene promoter
variability and Parkinson disease. JAMA 2006;296:
Mizuta I, Satake W, Nakabayashi Y, et al. Multiple
candidate gene analysis identifies alpha-synuclein as a
susceptibility gene for sporadic Parkinson’s disease.
Hum Mol Genet 2006;15:1151–1158.
Mueller JC, Fuchs J, Hofer A, et al. Multiple regions of
alpha-synuclein are associated with Parkinson’s dis-
ease. Ann Neurol 2005;57:535–541.
Pals P, Lincoln S, Manning J, et al. ?-Synuclein pro-
moter confers susceptibility to Parkinson’s disease.
Ann Neurol 2004;56:591–595.
Tan EK, Chai A, Teo YY, et al. Alpha-synuclein haplo-
types implicated in risk of Parkinson’s disease. Neurol-
Chiba-Falek O, Nussbaum RL. Effect of allelic varia-
tion at the NACP-Rep1 repeat upstream of the alpha-
synuclein gene (SNCA) on transcription in a cell
culture luciferase reporter system. Hum Mol Genet
Gibb WR, Lees AJ. A comparison of clinical and
pathological features of young- and old-onset Parkin-
son’s disease. Neurology 1988;38:1402–1406.
land GA. Score tests for association between traits and
haplotypes when linkage phase is ambiguous. Am J
Hum Genet 2002;70:425–434.
Barrett JC, Fry B, Maller J, Daly MJ. Haploview: anal-
ysis and visualization of LD and haplotype maps.
Maniatis N, Collins A, Xu CF, et al. The first linkage
disequilibrium (LD) maps: delineation of hot and cold
blocks by diplotype analysis. Proc Natl Acad Sci USA
Westfall PH, Young SS. Resampling-based multiple
testing: Examples and methods for p-value adjustment.
New York: John Wiley and Sons, 1993.
Edland SD, Slager S, Farrer M. Genetic association
studies in Alzheimer’s disease research: challenges and
opportunities. Stat Med 2004;23:169–178.
Boensch D, Lenz B, Kornhuber J, Bleich S. DNA hyper-
methylation of the alpha synuclein promoter in pa-
tients with alcoholism. Neuroreport 2005;16:167–170.
Farrer M, Kachergus J, Forno L, et al. Comparison of
kindreds with parkinsonism and alpha-synuclein
genomic multiplications. Ann Neurol 2004;55:174–179.
Fuchs J, Nilsson C, Kachergus J, et al. Phenotypic vari-
ation in a large Swedish pedigree due to SNCA duplica-
tion and triplication. Neurology 2007;68:916–922.
Neurology 69 October 30, 2007