Journal of Movement Disorders 2010;3:1-5
pISSN 2005-940X / eISSN 2093-4939
Human Genetic Variation and Parkinson’s Disease
REVIEW ARTICLE
Sun Ju Chung
Department of Neurology,
Asan Medical Center,
University of Ulsan
College of Medicine,
Seoul, Korea
Received March 9, 2010
Accepted April 10, 2010
Corresponding author
Sun Ju Chung, MD, PhD
Department of Neurology,
Asan Medical Center,
University of Ulsan
College of Medicine,
86 Asanbyeongwon-gil, Songpa-gu,
Seoul 138-736, Korea
Tel +82-2-3010-3440
Fax +82-2-474-4691
E-mail sjchung@amc.seoul.kr
•- The author has no financial conflicts of
interest.
Copyright © 2010 The Korean Movement Disorder Society 1
Parkinson’s disease (PD) is a chronic neurodegenerative disorder with multifactorial etiolo-
gy. In the past decade, the genetic causes of monogenic forms of familial PD have been defined.
However, the etiology and pathogenesis of the majority of sporadic PD cases that occur in out-
bred populations have yet to be clarified. The recent development of resources such as the In-
ternational HapMap Project and technological advances in high-throughput genotyping have
provided new basis for genetic association studies of common complex diseases, including
PD. A new generation of genome-wide association studies will soon offer a potentially powerful
approach for mapping causal genes and will likely change treatment and alter our perception
of the genetic determinants of PD. However, the execution and analysis of such studies will re-
quire great care. Journal of Movement Disorders 2010;3:1-5
Key Words: Parkinson’s disease, Genome, Genetic variation, Genome-wide association study.
In 2001, two reference versions of the human genome were published.1,2 One human ge-
nome sequence was reported by the Human Genome Sequencing Consortium and reflected
the assembly of sequences derived from numerous donors,1 whereas the other genome se-
quence, released by Celera Genomics, was a consensus sequence derived from five individu-
als.2 However, both versions of the genome sequence represented the human genome as a hap-
loid sequence, and generic variation was not annotated. Therefore, many researchers have st-
udied how genetic variants contribute to phenotype diversity and have conducted large-scale
studies to identify and catalogue nucleotides that differ among individuals. Initial studies fo-
cused largely on understanding the range of patterns and frequencies of single nucleotide
polymorphisms (SNPs).3-5 As their prevalence and contribution to human traits and biology
were realized, several consortia were formed, and systemic studies were performed to improve
our understanding of diverse human genomic variants.6,7
The first complete human genome sequence of a single individual, Levy et al.8 was published
in 2007. Shortly thereafter, the second complete genome sequence of an individual, Watson,
determined with next-generation sequencing technology, was published.9 Subsequently, three
additional genomes from anonymous individuals were sequenced: one Han Chinese (Asian),
one Nigerian (African), and one Korean (Asian).10-12 Although these data have rapidly in-
creased our knowledge of the various forms of human genetic variation, our understanding of
the location and frequencies of structural variants across the genome is still limited. Howev-
er, an enormous amount of effort is being expended to identify the common genetic variations
that contribute to the development of common complex diseases.
This review is a general overview of human genetic variation and its contribution to Parkin-
son’s disease (PD).
Classes of Human Genetic Variation
Common vs. rare variants
Human genetic variants are typically referred to as either common or rare to denote the
frequency of the minor allele in the human population. Common variants are synonymous with
polymorphisms, defined as genetic variants with a minor allele frequency of at least 1% in the
online © ML Comm

2Journal of Movement Disorders ▐ 2010;3:1-5
population, whereas rare variants have a minor allele frequen-
cy of less than 1%.
Single nucleotide polymorphisms
A SNP is a single base change in the DNA sequence at a par-
ticular point compared with the “common” or “wild-type” se-
quence. SNPs are the most prevalent class of genetic variation
among individuals. It has been estimated that the human ge-
nome contains at least 11 million SNPs, with about 7 million
of these occurring with minor allele frequencies exceeding
5% and the remaining having minor allele frequencies between
1 and 5%.
Structural variants
Structural variants are defined as all base pairs that differ
between individuals and that are not single nucleotide vari-
ants. These include insertion-deletion variants (indels), block
substitutions, inversions of DNA sequences, and copy number
differences. The technical ability to detect structural variants
in the human genome has only recently emerged.6,13
Genetic Association Studies
in Parkinson’s Disease
Investigators conducting genetic association studies may
target genes for investigation according to the known or pos-
tulated biology and previous results, an approach known as
candidate gene association. As a large-scale candidate gene as-
sociation study, Chung et al. investigated the association of
common variants in PARK loci and related genes with PD sus-
ceptibility and age at onset in an outbred population (unpub-
lished data: correspondence to Dr. Maraganore at NorthShore
University Health System, Chicago, USA). They matched 1,103
PD cases from the upper Midwest, USA, individually with un-
affected siblings (n = 654) or unrelated controls (n = 449) from
the same region. Using a sequencing approach in 25 cases and
25 controls, SNPs in species-conserved regions of PARK loci
and related genes were detected. Additional tag SNPs were
selected from the HapMap. A total of 235 SNPs and two vari-
able-number tandem repeats (VNTRs) in the ATP13A2, DJ1,
LRRK1, LRRK2, MAPT, Omi/HtrA2, PARK2, PINK1, SNCA,
SNCB, SNCG, SPR, and UCHL1 genes were genotyped in
all 2,206 subjects. Case-control analyses were performed to
study the association with PD susceptibility, whereas case-only
analyses were used to study the association with age at onset.
Only MAPT SNP rs2435200 was associated with PD suscep-
tibility after correcting for multiple testing [odds ratio (OR) =
0.74, 95% confidence interval (CI) = 0.64-0.86, uncorrected p
< 0.0001, log additive model]; however, 16 additional MAPT
variants, seven SNCA variants, and one LRRK2, PARK2, and
UCHL1 variant each had significant uncorrected p-values (Ta-
ble 1). No significant associations were found for age at on-
set after correcting for multiple testing. These results con-
firmed the association of the MAPT and SNCA genes with PD
susceptibility, but showed limited association of other PARK
loci and related genes with PD.
Alternatively, we may screen the entire genome for associ-
ation, an approach that has recently transformed the field of
genetic association studies. Such a “genome-wide association
study (GWAS)” is hypothesis-free, as there is no bias or pre-
sumptive list of candidate genes that are being tested. GWAS has
greatly accelerated the pace of discovery of genetic association.
As testing so many potential genes simultaneously carries
the risk of finding many spurious associations, genetic variants
that seem to have strong or suggestive statistical signals in an
initial GWAS need to be tested for replication in other large
data sets or studies.
The boundaries between candidate gene studies and GWAS
can become blurred, and the two types of study are not mutu-
ally exclusive.
Genome-Wide Association Study
in Parkinson’s Disease
Six GWAS of PD have been published (Table 2).14-19 The st-
udy by Maraganore et al. included 775 PD cases and 775 mat-
ched controls. This study genotyped 198,345 informative ge-
nomic SNPs, and found that a SNP within the semaphorin 5A
gene (SEMA5A) had the lowest combined p-value (p = 7.62
× 10-6).14 The authors also reported some suggestive findings
for MAPT and SNCA, as well as other PARK loci and related
genes. However, none of the findings was significant after
correcting for multiple testing. The study by Fung et al.15 ex-
amined more SNP markers (408,000 SNPs), but also failed to
observe an association of any genetic variation with PD sus-
ceptibility after correcting for multiple testing; however, that
study included only 276 PD cases and 276 unmatched con-
trols. The study by Pankratz et al.16 enrolled 857 familial PD
cases and 867 controls, and observed suggestive associations
for the GAK/DGKQ region on chromosome 4 (additive mod-
el: OR = 1.69; p = 3.4 × 10-6), MAPT SNPs (recessive model:
OR = 0.56; p = 2.0 × 10-5), and the SNCA SNPs (additive mo-
del: OR = 1.35; p = 5.5 × 10-5). Despite enriching their sam-
ple for genetic load (familial PD cases), none of the SNPs was
significant after correcting for multiple testing.
Recently, three GWAS confirmed that common variants in
SNCA and MAPT genes increase PD susceptibility.17-19 The
study by Satake et al.17 (2,011 cases and 18,381 controls) re-
ported strong associations at SNCA on 4q22 (rs11931074,
OR = 1.37, p = 7.35 × 10-17), PARK16 on 1q32 (p = 1.52 × 10-12),
BST1 on 4q15, (p = 3.94 × 10-9), and LRRK2 on 12q12 (p = 2.72
× 10-8). The study by Simón-Sánchez et al.18 (5,074 cases and
8,551 controls) observed two strong association signals in
the SNCA gene (rs2736990, OR = 1.23, p = 2.24 × 10-16) and