RESEARCH ARTICLE

Lack of Association Between p.Ser167Asn Variant ofParkin andParkinson’sDisease: AMeta-Analysis of 15Studies Involving 2,280 Cases and 2,459 ControlsYi Zhang,1 Zhen-Zhen Wang,2 and Hong-Mei Sun1*1Department of Anatomy, School of Preclinical Medicine, Beijing University of Chinese Medicine, Beijing, China2Beijing Institute of Pharmacology and Toxicology, Beijing, China

Received 13 April 2011; Accepted 18 October 2011

Previous clinical trials have evaluated the association between

Parkin p.Ser167Asn (c.601G>A) variant andParkinson’s disease

(PD) risk. However, the results remain conflicting rather than

conclusive. Therefore, we performed thismeta-analysis to assess

whether pooled results show the association. We performed

structured literature searches for studies addressing the associ-

ation between the Parkin p.Ser167Asn variant and PD risk. We

conducted analyses of study characteristics, heterogeneity, and

funnel plot asymmetry in analyses analogous to additive,

dominant, recessive, and general genetic models with the odds

ratio (OR) as the measure of association. When 15 eligible

studies (n¼ 4,739 subjects) were pooled into the meta-analysis,

there was no evidence for significant association in additive

genetic model between Parkin p. Ser167Asn variant and PD

risk (OR¼ 1.02, 95% confidence interval (CI)¼ 0.83–1.25;P¼ 0.866). The OR for the dominant model was 1.06 (95%

CI¼ 0.80–1.41) while the OR for the recessive model was 0.90

(95% CI¼ 0.71–1.14). The OR for the heterozygous was 1.07

(95%CI¼ 0.80–1.43)while theOR for the homozygoteswas 1.19

(95% CI¼ 0.81–1.74). In the subgroup analysis by ethnicity, no

significant association was found in any genetic model. Begg’s

funnel plot and Egger’s test provided visual and statistical

evidences for funnel plot symmetry, suggesting no presence of

publication bias. In summary, the meta-analysis strongly sug-

gests that Parkin p. Ser167Asn variant is not associated with PD

risk. � 2011 Wiley Periodicals, Inc.

Key words: Parkinson’s disease; Parkin; p.Ser167Asn variant;

association analysis; meta-analysis

INTRODUCTION

Parkinson’s disease (PD; Online Mendelian Inheritance in Man,

OMIM #168600), the second most common neurodegenerative

disorder, is characterized clinically by resting tremor, rigidity,

bradykinesia, and postural instability [de Lau and Breteler,

2006]. The etiology of PD has not been fully elucidated, even

though a considerable amount of evidence has manifested that

environmental or endogenous toxins may be causative agents in

genetically susceptible subpopulations [Greenamyre and Hastings,

2004]. In the past decade, the accumulated data indicated the

dominant role of genetic factors, especially in early-onset PD

(EOPD) patients [Kitada et al., 1998].

Themajority of PD cases are sporadic, but the discovery of genes

linked to rare familial forms of disease encoding alpha-synuclein,

Parkin, DJ1, PINK1 (PTEN-induced putative kinase 1), LRRK2

(leucine-rich repeat kinase 2), and UCHL-1 (ubiquitin carboxyhy-

drolase L1) along with studies from experimental animal models

has provided crucial insights intomolecularmechanisms in disease

pathogenesis and identified probable targets for therapeutic inter-

vention [Gandhi and Wood, 2005; Schrag and Schott, 2006].

Mutations in these genes have been identified to segregate with

PD in a Mendelian fashion. Moreover, mutations in Parkin

(PARK2, OMIM *602544) are typically found in EOPD [de Lau

and Breteler, 2006].

Parkin is an E3 ubiquitin ligase that degrades specific aggrega-

tion-prone substrates and regulates the mitochondrial transcrip-

tion [DodsonandGuo, 2007].ThemutationofParkinmight lead to

mitochondrial dysfunction, another important factor strongly

implicated in the pathogenesis of PD [Greene et al., 2003]. In

addition, point mutations and exon rearrangements in Parkin are

Grant sponsor: National Natural Science Foundation of China; Grant

number: 30873335.

*Correspondence to:

Prof.Hong-Mei Sun, Ph.D.,Department ofAnatomy, School of Preclinical

Medicine, Beijing University of Chinese Medicine, No. 11 N. 3rd Ring

Eastern Road, Beijing 100029, China. E-mail: hongmei.sun@yahoo.cn

Published online 16 November 2011 in Wiley Online Library

(wileyonlinelibrary.com).

DOI 10.1002/ajmg.b.31250

How to Cite this Article:Zhang Y, Wang Z-Z, Sun H-M. 2012. Lack of

association between p.Ser167Asn Variant of

Parkin and Parkinson’s disease: A meta-

analysis of 15 studies involving 2,280 cases

and 2,459 controls.

Am J Med Genet Part B 159B:38–47.

� 2011 Wiley Periodicals, Inc. 38

Neuropsychiatric Genetics

the most common genetic causes of EOPD; approximately 18% of

sporadic and 50% of all familial early-onset cases have been

attributed to Parkin mutations [Lucking et al., 2000].

The frequency and spectrum of Parkin mutations have been

extensively researched in large, ethnically mixed patient popula-

tions by different study groups [Lucking et al., 2000; Oliveira et al.,

2003; Poorkaj et al., 2004]. Recently, a missense variant of Parkin,

p.Ser167Asn (dbSNPAccessionNo. rs1801474, namely c.601G>A,

http://www.ncbi.nlm.nih.gov/projects/SNP/

snp_ref.cgi?rs¼1801474) has been focused to associate with PD

risk. This variant was originally identified in Japanese with familial

PD [Wang et al., 1999]. It results from a G to A substitution at

nucleotide 601, which causes the serine-to-asparagine amino acid

change in codon 167 of Parkin protein.

Various studies have analyzed the association between PD and

the p.Ser167Asn variant of Parkin, by reporting different results for

this variant as risk factor, and even negative result in different

populations. However, the results remain conflicting rather than

conclusive. Two studies reported that the p.Ser167Asn variant was

significantly associated with an increased PD risk of Asians [Satoh

and Kuroda, 1999; Wu et al., 2010], while one study reported that

this variant was not the risk-increasing variant in Caucasians

[Koziorowski et al., 2010]. Therefore, we performed this meta-

analysis to assess whether pooled result shows the association

between Parkin p.Ser167Asn variant and PD risk.

MATERIALS AND METHODS

During all phases of the design and implementation of the present

analysis, we strictly followed the standards of Quality of Reporting

ofMeta-analysis (QUOROM)[Moher et al., 1999]. The description

of sequence variants is according to the recommendation nomen-

clature for genetic research [den Dunnen and Antonarakis, 2001].

Literature CollectionWe conducted a systematic literature search of MEDLINE,

Cochrane Library, and China Academic Journals Database,

using the following keywords ‘‘Parkin,’’ ‘‘PARK2,’’ ‘‘PRNK,’’

and ‘‘Parkinson’s disease,’’ ‘‘Parkinson disease,’’ ‘‘Parkinsons

disease,’’ ‘‘PD’’ in combination with ‘‘variant,’’ ‘‘mutation,’’

‘‘polymorphism,’’ or ‘‘genetics.’’ In addition, a hand-search of

references from clinical trial reports or review articles was per-

formed to identify additional relevant researches. All researches

were limited to English and Chinese language articles published up

to March, 2011. Studies were included if the distribution of

genotypes for the Parkin p.Ser167Asn variant was reported for

patients with clinically diagnosed PD and for a control population

(i.e., without PD).

Data ExtractionTwo investigators (Y. Zhang and Z.Z. Wang) independently eval-

uated all trials with disagreement resolved by consensus and

rereading the original manuscripts of studies included according

to search criteria, and extracted relevant data from each study using

a standardized form. The following data were extracted from each

study included: general information (thefirst author’s name, yearof

publication, and location of origin or country), ethnic group of the

study population, characteristics of PD cases and controls, number

of cases and controls for each Parkin genotype.

Analysis ProceduresAllele frequencies were estimated by gene counting. Sensitivity

analysis was performed excluding studies whose allele frequencies

in controls exhibited significant deviation from the Hardy–Weinberg equilibrium (HWE), given that the deviationmay denote

bias. For the assessment of the deviation from HWE, the appro-

priate goodness-of-fit chi-squared test was performed. Deviation

for HWE was deemed significant for P-value <0.05.

We calculated the overall odds ratio (OR) with the correspond-

ing 95% confidence interval (CI), using the random effects model

[DerSimonian andLaird, 1986]. Presenceof heterogeneity in effects

was assessed using the Cochrane’s Q statistics and quantified using

the I2 test [Higgins et al., 2003]. I2 represents the percentage of the

observed between-study variability due to heterogeneity rather

than to chance. It ranges between 0% and 100%, where a value

of 0% indicates no observed heterogeneity, and larger values

indicate an increasing degree of heterogeneity. The P-value

<0.05 indicated a significant statistical heterogeneity across studies,

allowing for the use of the random effects model.

Currently, there is no conclusive evidence for the genetic model

underlying the potential relation between p.Ser167Asn variant of

Parkin and PD; therefore, we examined four genetic models of

inheritance: additive (Asn/Asn and Ser/Asn vs. Ser/Asn and Ser/

Ser), dominant (Asn/Asn and Ser/Asn vs. Ser/Ser), recessive (Asn/

Asn vs. Ser/Ser and Ser/Asn), and general (Asn/Asn vs. Ser/Ser; Ser/

Asn vs. Ser/Ser). The presence of a mutant allele (homozygote or

heterozygote) was considered the exposure variable. Subgroup

analyses were prespecified before data collection to compare the

results separately of Caucasians, Asians, and others.

Visual inspection of funnel plot and Begg’s test statistics were

used to assess for the presence of publication bias [Begg and

Mazumdar, 1994; Egger et al., 1997], P-value<0.05 was considered

significant. All analyses were performed using statistical software

STATA/SE (version 11.0, STATA Corporation, College Station,

TX).

RESULTS

Literature CollectionOur search yielded 178 potential literature citations. Of those, 114

were excluded through abstract review leaving 64 articles for

publication review. Of those 64 articles, 28 were excluded because

they were either not performed in humans. Thus, 36 articles were

chosen for full publication review. Seventeenof thesewere excluded

because they either did not report the desired outcome data.

Furthermore, 4 studies [Wang et al., 1999; Ning et al., 2003;

Wang et al., 2003;Martinez et al., 2010] were also excluded because

the control individuals of these studies showed evidence of pop-

ulation stratification detectable as a departure of genotype frequen-

cies from HWE (P-value of HWE <0.05, respectively). Finally, 15

studies were finally included in our meta-analysis. These 15

included studies summarized in Table I.

ZHANG ET AL. 39

TABLE

I.Descriptive

andClinicalCharacteristicsof

15Studieson

Parkin

p.Ser167AsnVariantandPD:Cases

andControls

Refs.

Publication

language

Location

Ethnicity

Cases(n

¼2,280)

Controls(n

¼2,459)

Enrollm

ent

source

Type

ofPD

nMale

(%)

Ageat

onset

(years)

Ageat

exam

(years)

Diagnose

criteria

Enrollm

ent

source

nMale

(%)

Ageat

exam

(years)

HWE

P-value

Koziorowski

etal.[2010]

English

Poland

Caucasian

Population-

based

Sporadic,

familial

79

58.2

34.9

NA

UKPDBB

Population-

based

204

38.0

30.4

1.000

Wuet

al.

[2010]

English

China

Taiwan

Asian

Clinic

Sporadic

506

55.9

62.8

NA

Gleb

Community

508

54.3

60.5

0.407

Brookset

al.

[2009]

English

USA

Caucasian

Population-

based

Sporadic,

familial

250

61.2

41.0

54.0

NA

Population-

based

276

27.9

39.0

1.000

Aguiar

etal.

[2008]

English

Brazil

Latin-American

Clinic

Sporadic,

familial

72

68.1

36.5

45.6

UKPDBB

Clinic

72

66.7

38.5

1.000

Lesage

etal.

[2008]

English

France

Caucasian

Population-

based

Sporadic

172

57.6

36.8

49.6

Clinicala

Spouses

170

55.3

62.9

1.000

Okubadejoet

al.

[2008]

English

Nigeria

Black

African

Clinic

Sporadic

57

75.4

58.2

62.3

Clinicalb

Clinic

51

68.6

63.7

1.000

Biswas

etal.

[2007]

English

India

Asian

Population-

based

Sporadic,

familial

138

NA

47.0

NA

Calne

Population-

based

141

NA

55.0

1.000

HongandZhang

[2006]

Chinese

China

Asian

Clinic

Sporadic

100

50.0

NA

64.4

Clinicalc

Clinic

100

52.0

60.3

0.166

Luckinget

al.

[2003]

English

Germany

Caucasian

Population-

based

Sporadic,

familial

194

57.7

NA

53.1

Clinicala

Population-

based

125

60.0

53.6

1.000

Penget

al.

[2003]

English

China

Asian

Clinic

Sporadic

116

64.7

56.3

NA

UKPDBB

Population-

based

124

63.7

57.2

0.696

Zhao

etal.

[2003]

Chinese

China

Asian

Clinic

Sporadic

155

57.4

NA

63.0

Clinicalc

Clinic

192

51.6

62.0

0.082

Eerolaet

al.

[2002]

English

Finland

Caucasian

Clinic

Sporadic

147

59.2

NA

67.2

UKPDBB

Spouses

137

36.5

65.8

1.000

Mataet

al.

[2002]

English

Spain

Caucasian

Population-

based

Sporadic

105

NA

NA

NA

CAPIT

Community

150

NA

50.0

1.000

Oliveriet

al.

[2001]

English

Italy

Caucasian

Population-

based

Sporadic,

familial

118

55.1

60.5d,

61.3e

66.3d,

69.1e

UKPDBB

Population-

based

100

52.0

68.8

1.000

Satohand

Kuroda

[1999]

English

Japan

Asian

Clinic

Sporadic

71

39.4

NA

68.8

CAPIT

Clinic

109

45.0

65.7

0.830

NA,not

available;HWE,Hardy–Weinberg

equilibrium

;UKPDBB,United

Kingdom

Parkinson’sDisease

BrainBank;CAPIT,Core

AssessmentProgram

forIntracerebralTransplantationsCommittee.

aAt

leastthreeof

themandatory

criteria(akinesia,rigidity,restingtrem

or,asym

metricalonsetor

>30%improvem

entwithlevodopa),andtheabsence

ofexclusioncriteria.

bThepresence

ofallfive

of:(a)

atleast2of3cardinalsignsoftrem

or,rigidity,bradykinesia(withorwithoutposturalorgaitabnormality);(b)

anasym

metriconset;(c)noidentifiablesecondarycause(e.g.,repeated

stroke,exposuretomedicationscapableofcausingPD

within6monthsbeforeonset,etc);(d)responsivenesstolevodopa

therapy(applicabletotreatedpatientsonly);(e)absence

ofsignsofmoreextensive

nervous

system

involvem

ent(e.g.,earlyautonom

icfeatures

orcognitiveimpairmentw

ithin2yearsofonset,otherwise

unexplained

corticospinaltractdysfunction,andcerebellarsigns).

c PDDiagnosticCriteriaestablishedby

theFirstNationalExtrapyram

idalDiseasesConference

ofChina.

dSporadiccases.

eFamilialcases.

Description and Characteristics ofIncluded StudiesWe obtained all available data on the association between Parkin

p.Ser167Asn variant and PD, which allowed us to perform this

meta-analysis of data of 4,739 individuals (2,280 cases and 2,459

controls). Study descriptive and clinical characteristics are sum-

marized inTable I. Thefirst studywas published in 1999 and the last

one in 2010. Ethnic groups among these studies were as following: 7

studies involved Caucasians, 6 studies involved Asians, and 2

studies included a mixed population. All genotypes and allele

frequencies in the cases and controls are shown in Table II.

Quantitative Data SynthesisIn these studies, genotype distributions of the Parkin p.Ser167Asn

variant showed no significant departure from HWE when each

control group as well as the pooled control sample were considered

(P> 0.05, Table I).

In a random-effects meta-analysis with additive model, the

combined PDOR for the p.Ser167Asn variant across the 15 studies

was 1.02 (95% CI¼ 0.83–1.25) shown in Table III and Figure 1. In

additive model, there was evidence of a moderate degree of incon-

sistency among these studies (I2¼ 39.9%;P¼ 0.056).When studies

were subgrouped by ethnicity, the OR value was nearer 1 for Asians

(OR¼ 1.07, 95% CI¼ 0.83–1.41; I2¼ 74.1%, P¼ 0.02) than for

Caucasians (OR¼ 0.89, 95% CI¼ 0.56–1.41; I2¼ 0%, P¼ 0.995),

although the difference was not statistically significant (POR¼0.866). Results from subgroup analyses in additive model are

also presented in Table III and Figure 1.

In assessing a dominant effect of Parkin p.Ser167Asn variant, no

significant association was found with PD when compared with

homozygous Ser carriers (OR¼ 1.06; 95% CI¼ 0.79–1.41) withI2¼ 50.3%, P¼ 0.016 (Table III and Fig. 2). When studies were

subgrouped by ethnicity, the OR was nearer 1 for Caucasians

(OR¼ 0.90, 95% CI¼ 0.56–1.43; I2¼ 0%, P¼ 0.986) than for

Asians (OR¼ 1.19, 95% CI¼ 0.79–1.82; I2¼ 76.6%, P¼ 0.001),

although the difference was not statistically significant (POR¼0.698). Similarly, no significant association was observed following

analysis of the Asn/Asn in a recessive genetic model (OR¼ 0.90,

95% CI¼ 0.71–1.14, figure and subgroup data not shown).

When a random-effects model was used, the Ser/Ser genotype

being the reference group, the combined OR were 1.07 (95%

CI¼ 0.80–1.43) for the heterozygotes (Ser/Asn vs. Ser/Ser,

Table III and Fig. 3). When studies were subgrouped by ethnicity,

the OR was nearer 1 for Caucasians (OR¼ 0.90, 95%

CI¼ 0.56–1.46; I2¼ 0%, P¼ 0.986) than for Asians (OR¼ 1.23,

95% CI¼ 0.80–1.88; I2¼ 75.0%, P¼ 0.001), although the differ-

ence was not statistically significant (POR¼ 0.667). The I2 statistic

(I2¼ 48.8%, P¼ 0.020) was consistent with these results, suggest-

ing only ‘‘moderate’’ heterogeneity. Similarly, the value of OR was

1.19 (95% CI¼ 0.81–1.74) for the homozygotes (Asn/Asn vs. Ser/

Ser), figure and subgroup data not shown.

Begg’s funnel plot and Egger’s test were performed to access the

publication bias of this meta-analysis. The shape of the funnel plot

for the overall results seemed symmetrical (Fig. 4), suggesting no

presence of publication bias. Moreover, P-value of the Begg’s test

was 0.907 and the P-value of the Egger’s test was 0.826, providing

statistical evidence for funnel plot symmetry.

DISCUSSION

To our knowledge, the currentmeta-analysis is the first to elucidate

the potential association between Parkin and PD. The principal

message of this meta-analysis is the lack of association between

Parkin p.Ser167Asn variant and PD risk. We identified 3 GWASs

(genome-wide association studies) of PD [Satake et al., 2009;

Simon-Sanchez et al., 2009; Chang et al., 2011] and 3meta-analysis

of GWAS of PD [Evangelou et al., 2007, 2010; Nalls et al., 2011].

None of these studies had investigated Parkin p.Ser167Asn variant,

either because the variant was not included on the genotyping

platform or because it did not meet quality control standards for

inclusion in the analysis. We also considered using rs1801474 as a

proxy for the Parkin p.Ser167Asn variant; however, rs1801474 was

on none of the platforms used for these GWAS.

Parkin-linked PDhas a broad range of clinical phenotypes, some

atypical, but is generally early-onset parkinsonism, with a slow

clinical course, frequent treatment-induced dyskinesias, excellent

response to low doses of levodopa, and no dementia [Lohmann

et al., 2009]. Parkin mutations have been found not only in EOPD

but also in the late-onset PD (LOPD) [Sun et al., 2006]. It has

therefore been proposed that individuals with LOPD should not be

excluded from Parkin mutation screening, especially if they have a

positive family history of PD [Pankratz et al., 2009].

Mutations have been found in the homozygous or compound

heterozygous state, compatible with recessive transmission, but

occasionally as single heterozygous mutations [West et al., 2002].

Homozygousmutations have been reported in autosomal recessive

EOPD cases and heterozygous mutations are observed in LOPD

cases [Foroud et al., 2003]. Increasing evidence indicates a role of

heterozygous pathogenic mutations as a susceptibility factor for

PD.Positron emission tomography imaging studies have reporteda

subclinical dopaminergic dysfunction in heterozygous Parkin

mutation carriers [Khan et al., 2005]. There is an increased fre-

quency of heterozygous mutations in patients with PD compared

with healthy controls [Klein et al., 2007]. However, the literature is

not consistent, as some studies report a similar frequency of

heterozygous mutations in cases and in controls [Lincoln et al.,

2003; Kay et al., 2007]. Some studies support that heterozygous

Parkin point mutations are not associated with PD [Lincoln

et al., 2003; Chien et al., 2006]. Additional studies are therefore

required to investigate the relevance of heterozygous pathogenic

mutations to the pathogenesis of PD. Our meta-analysis

showed that the accumulated evidence of heterozygous of Parkin

p.Ser167Asn variant is insufficient to claim its general clinical

importance for PD.

It is well known that association studies represent a powerful

approach to identification of genetic variants that influence sus-

ceptibility to common diseases [Lohmueller et al., 2003]. However,

agreement exists about the fact that epidemiological evidence for

gene–disease association requires replication, validation, and syn-

thesis [Ioannidis et al., 2005].Meta-analyses are statistical tools that

permit to combine the results of multiple studies that address the

same hypothesis [Lohmueller et al., 2003]. When performed

ZHANG ET AL. 41

TABLE

II.GenotypeandAlleleDistributionsforthep.Ser167AsnVariantof

Parkin

inSubjects

Refs.

Ethnicity

Genotypedistribution

Alleledistribution

Cases,

n(%)

Controls,

n(%)

Cases

(%)

Controls

(%)

Total

Ser/Ser

Ser/Asn

Asn/Asn

Total

Ser/Ser

Ser/Asn

Asn/Asn

Ser

Asn

Ser

Asn

Koziorowskiet

al.[2010]

Caucasian

79

75(94.9)

4(5.1)

0(0.0)

204

190(93.1)

14(6.9)

0(0.0)

97.5

2.5

96.6

3.4

Brookset

al.[2009]

Caucasian

250

243(97.2)

7(2.8)

0(0.0)

276

267(96.7)

9(3.3)

0(0.0)

98.6

1.4

98.4

1.6

Lesage

etal.[2008]

Caucasian

172

168(97.7)

4(2.3)

0(0.0)

170

165(97.1)

5(2.9)

0(0.0)

98.8

1.2

98.5

1.5

Luckinget

al.[2003]

Caucasian

194

186(95.9)

8(4.1)

0(0.0)

125

120(96.0)

5(4.0)

0(0.0)

97.9

2.1

98.0

2.0

Eerolaet

al.[2002]

Caucasian

147

0(0.0)

3(2.0)

144(98.0)

137

0(0.0)

2(1.5)

135(98.5)

1.0

99.0

0.7

99.3

Mataet

al.[2002]

Caucasian

105

100(95.2)

5(4.8)

0(0.0)

150

142(94.7)

8(5.3)

0(0.0)

97.6

2.4

97.3

2.7

Oliveriet

al.[2001]

Caucasian

118

113(95.8)

5(4.2)

0(0.0)

100

97(97.0)

3(3.0)

0(0.0)

97.9

2.1

98.5

1.5

Subtotal

1,065

885(83.1)

36(3.4)

144(13.5)

1,162

981(84.4)

46(4.0)

135(11.6)

84.8

15.2

86.4

13.6

Wuet

al.[2010]

Asian

506

183(36.2)

252(49.8)

71(14.0)

508

186(36.6)

235(46.3)

87(17.1)

61.1

38.9

59.7

40.3

Biswas

etal.[2007]

Asian

138

127(92.0)

10(7.2)

1(0.7)

141

111(78.7)

29(20.6)

1(0.7)

95.7

4.3

89.0

11.0

HongandZhang[2006]

Asian

100

38(38.0)

44(44.0)

18(18.0)

100

50(50.0)

37(37.0)

13(13.0)

60.0

40.0

68.5

31.5

Penget

al.[2003]

Asian

116

28(24.1)

68(58.6)

20(17.2)

124

51(41.1)

59(47.6)

14(11.3)

53.4

46.6

64.9

35.1

Zhao

etal.[2003]

Asian

155

44(28.4)

80(51.6)

31(20.0)

192

60(31.3)

83(43.2)

49(25.5)

54.2

45.8

52.9

47.1

SatohandKuroda[1999]

Asian

71

21(29.6)

44(62.0)

6(8.5)

109

48(44.0)

50(45.9)

11(10.1)

60.6

39.4

67.0

33.0

Subtotal

1,086

441(40.6)

498(45.9)

147(13.5)

1,174

506(43.1)

493(42.0)

175(14.9)

63.5

36.5

64.1

35.9

Aguiar

etal.[2008]

Latin-American

72

68(94.4)

4(5.6)

0(0.0)

72

62(86.1)

10(13.9)

0(0.0)

97.2

2.8

93.1

6.9

Okubadejoet

al.[2008]

Black-African

57

52(91.2)

4(7.0)

1(1.8)

51

47(92.2)

4(7.8)

0(0.0)

94.7

5.3

96.1

3.9

Total

2,280

1446(63.4)

542(23.8)

292(12.8)

2,459

1596(64.9)

553(22.5)

310(12.6)

75.3

24.7

76.1

23.9

TABLE III. ORs and 95% CI for the Parkin p.Ser167Asn Variant and PD Under Different Genetic Models

Genetic model EthnicityRandom effectOR (95% CI)

P-valuefor OR

Weight(%) I2 (%)

Q statistic forheterogeneity

P-value forheterogeneity

Additive Overall 1.02 (0.83–1.25) 0.866 100 39.9 23.29 (14) 0.056Caucasians 0.89 (0.56–1.41) 0.625 17.38 0 0.65 (6) 0.995Asians 1.07 (0.80–1.41) 0.664 77.73 74.1 19.32 (5) 0.002Others 0.70 (0.20–2.43) 0.576 4.89 50.3 2.01 (1) 0.156

Dominant Overall 1.06 (0.80–1.41) 0.698 100 50.3 26.17 (13) 0.016Caucasians 0.90 (0.56–1.46) 0.680 26.26 0 0.63 (5) 0.986Asians 1.20 (0.79–1.82) 0.405 66.01 76.6 21.38 (5) 0.001Others 0.61 (0.20–1.85) 0.383 7.74 31.9 1.47 (1) 0.225

Recessive Overall 0.90 (0.71–1.14) 0.378 100 0 5.85 (7) 0.557Caucasians 0.71 (0.12–4.32) 0.711 1.76 — 0 (0) —Asians 0.91 (0.70–1.18) 0.470 97.68 6.1 5.33 (5) 0.377Others 2.74 (0.11–68.63) 0.541 0.55 — 0 (0) —

Ser/Asn versus Ser/Ser Overall 1.07 (0.80–1.43) 0.667 100 48.8 25.41 (13) 0.020Caucasians 0.90 (0.56–1.46) 0.680 27.10 0 0.63 (5) 0.986Asians 1.23 (0.80–1.88) 0.348 65.18 75.0 19.97 (5) 0.001Others 0.53 (0.21–1.34) 0.180 7.72 0 0.90 (1) 0.344

Asn/Asn versus Ser/Ser Overall 1.19 (0.81–1.74) 0.384 100 31.4 8.74 (6) 0.189Caucasians — — 0 — — —Asians 1.19 (0.79–1.79) 0.410 98.62 40.6 8.41 (5) 0.135Others 2.71 (0.11–68.25) 0.544 1.38 — 0 (0) —

— Indicates the insufficient observations, the number is not sufficient to carry out statistical calculation.

FIG. 1. Results of published studies of association between the p.Ser167Asn variant of Parkin and PD in additive geneticmodel. The size of each box is

proportional to the weight of each study. Also shown are the diamonds of the summary effects based on subgroup and all studies, respectively.

[Color figure can be seen in the online version of this article, available at http://wileyonlinelibrary.com/journal/ajmgb]

appropriately, meta-analysis may enhance the precision of the

estimates of the effects of risk alleles, leading to reduced probability

of false negative results. In present study we preformed a meta-

analysis that evaluated 4,739 subjects in an association between the

p.Ser167Asn variant and PD risk. Despite a biological plausible a

priori hypothesis, our analysis results did not support a role for

p.Ser167Asn variant in the etiology of PD.

Heterogeneity in a meta-analysis is mostly produced by differ-

ences in study design and background characteristics of the sub-

jects, and the extent of heterogeneity might influence the

conclusions [Higgins et al., 2003]. In our study, in the analysis

of sensitivity, heterogeneity disappeared in combined studies after

removing outlier studies, as was explained before. In addition, a

random effect model, where heterogeneity is no longer an issue,

provided valid results. However, a note of caution should be added

because heterogeneity may potentially restrict the interpretation of

the pooled risk estimates. Overall, in our meta-analysis, I2 was

estimated to be 39.9% in additive genetic model (P¼ 0.056). This

represents amoderate level of inconsistency attributable to genuine

variation in gene effect size.

As allele frequency differences between ethnic groups and

potential population stratification, ethnicity may have acted as

an important variable in determining association risk with PD

[Djarmati et al., 2004; Hedrich et al., 2004]. Therefore, we per-

formed the meta-analysis by subgroup studies of Caucasians

(7 studies, n¼ 2,227; 47.0%), Asians (6 studies, n¼ 2,260;

47.7%), and others (2 studies, n¼ 252; 5.3%). In these studies,

we identified (Table II) the frequency of the Parkin p.Ser167Asn

variant (Ser/AsnþAsn/Asn) in cases was observed to be higher

in Asian populations (59.4%) than Caucasian populations

(16.9%), and in controls was also observed to be higher in Asian

populations (56.9%) than Caucasian populations (15.6%). We

observed a slightly higher magnitude of association in Asians

compared with Caucasians in additive model, but the difference

was not statistically significant. The associations between Parkin

p.Ser167Asn variant and PD were not statistically significant in all

four genetic models (Table III). Black, Latino subjects showed no

significant associations; however, very few researches were per-

formed for these populations, and the results should be interpreted

with caution.

The inconsistent results with the p.Ser167Asn variant andPDare

not surprising. Genetic association studies are plagued by conflict-

ing results that can be explained by heterogeneity across study

populations as well as false-positive and -negative results. Approx-

imately two-thirds of genetic associations do not hold up onmeta-

analysis [Lohmueller et al., 2003]. Large sample sizes and pooling of

FIG. 2. Results of published studies of association between the p.Ser167Asn variant of Parkin and PD in dominant geneticmodel. The size of each box

is proportional to the weight of each study. Also shown are the diamonds of the summary effects based on subgroup and all studies, respectively.

[Color figure can be seen in the online version of this article, available at http://wileyonlinelibrary.com/journal/ajmgb]

44 AMERICAN JOURNAL OF MEDICAL GENETICS PART B

data are therefore critical to evaluate the association between a

phenotype and genetic variation with confidence.

Despite the present research being ameta-analysis of all available

publications, the potential limitations to our study should be

acknowledged when interpreting these results. First, as with all

meta-analyses, the possibility of publication bias exists [Colhoun

et al., 2003]. Although our assessments did not generally suggest the

presence of publication bias, it is not possible to be ruled it out

entirely. Second, due to limited data, we were unable to evaluate

some important effectmodifiers for the association between Parkin

and PD, such as sex, age at onset, family history, and dementia

status. Most studies included in our analyses did not distinguish

EOPD from LOPD which may have different genetic etiologies.

Third, bias from population stratification may not be an

issue because each study included was based on one ethnic

population.

In conclusion, our results do not support an association between

Parkin p.Ser167Asn variant andPDrisk. This effect size implies that

large study populations involving several thousands of patients and

controls will be needed to draw reliable conclusions.Notwithstand-

ing we confirm the lack of association between PD and the widely

studied Parkin p.Ser167Asn variant, this does not eliminate the

possibility that other Parkin variations may be involved in PD risk.

Further researches analyzing a broader rangeof variantsmay be able

FIG. 3. Results of published studies of association between the p.Ser167Asn variant of Parkin and PD for the heterozygotes. The size of each box is

proportional to the weight of each study. Also shown are the diamonds of the summary effects based on subgroup and all studies, respectively.

[Color figure can be seen in the online version of this article, available at http://wileyonlinelibrary.com/journal/ajmgb]

FIG. 4. Begg’s funnel plot with pseudo 95% confidence limits in

each study. The size of the circle is proportional to the weight of

the study.

ZHANG ET AL. 45

to better describe the relationship, if any exists, between variation in

Parkin and PD risk.

REFERENCES

Aguiar PC, Lessa PS, Godeiro CJ, Barsottini O, Felicio AC, Borges V, SilvaSM, Saba RA, Ferraz HB, Moreira-Filho CA, et al. 2008. Genetic andenvironmental findings in early-onset Parkinson’s disease Brazilianpatients. Mov Disord 23:1228–1233.

Begg CB, Mazumdar M. 1994. Operating characteristics of a rank corre-lation test for publication bias. Biometrics 50:1088–1101.

Biswas A, Maulik M, Das SK, Ray K, Ray J. 2007. Parkin polymorphisms:Risk for Parkinson’s disease in Indian population. Clin Genet 72:484–486.

Brooks J, Ding J, Simon-Sanchez J, Paisan-Ruiz C, Singleton AB, ScholzSW. 2009. Parkin and PINK1 mutations in early-onset Parkinson’sdisease: Comprehensive screening in publicly available cases and control.J Med Genet 46:375–381.

Chang XL, Mao XY, Li HH, Zhang JH, Li NN, Burgunder JM, Peng R, TanEK. 2011. Association of GWAS loci with PD in China. Am JMed GenetPart B 156B:334–339.

Chien HF, Rohe CF, Costa MD, Breedveld GJ, Oostra BA, Barbosa ER,Bonifati V. 2006. Early-onset Parkinson’s disease caused by a novelParkin mutation in a genetic isolate from north-eastern Brazil. Neuro-genetics 7:13–19.

Colhoun HM, McKeigue PM, Davey SG. 2003. Problems of reportinggenetic associations with complex outcomes. Lancet 361:865–872.

de Lau LM, Breteler MM. 2006. Epidemiology of Parkinson’s disease.Lancet Neurol 5:525–535.

denDunnen JT,Antonarakis SE. 2001.Nomenclature for the descriptionofhuman sequence variations. Hum Genet 109:121–124.

DerSimonianR, LairdN. 1986.Meta-analysis in clinical trials. Control ClinTrials 7:177–188.

DjarmatiA,HedrichK, SvetelM, SchaferN, JuricV,Vukosavic S,HeringR,Riess O, Romac S, Klein C, et al. 2004. Detection of Parkin (PARK2)and DJ1 (PARK7) mutations in early-onset Parkinson disease: Parkinmutation frequency depends on ethnic origin of patients. Hum Mutat23:525.

Dodson MW, Guo M. 2007. Pink1, Parkin, DJ-1 and mitochondrialdysfunction in Parkinson’s disease. Curr Opin Neurobiol 17:331–337.

Eerola J, Launes J, Hellstrom O, Tienari PJ. 2002. Apolipoprotein E(APOE), PARKIN and catechol-O-methyltransferase (COMT) genesand susceptibility to sporadic Parkinson’s disease in Finland. NeurosciLett 330:296–298.

Egger M, Davey SG, Schneider M, Minder C. 1997. Bias in meta-analysisdetected by a simple, graphical test. BMJ 315:629–634.

Evangelou E, Maraganore DM, Ioannidis JP. 2007. Meta-analysis ingenome-wide association datasets: Strategies and application in Parkin-son disease. PLoS ONE 2:e196.

Evangelou E, Maraganore DM, Annesi G, Brighina L, Brice A, Elbaz A,FerrareseC,HadjigeorgiouGM,Krueger R, Lambert JC, et al. 2010.Non-replication of association for six polymorphisms from meta-analysis ofgenome-wide association studies of Parkinson’s disease: Large-scalecollaborative study. Am J Med Genet Part B 153B:220–228.

Foroud T, Uniacke SK, Liu L, Pankratz N, Rudolph A, Halter C, Shults C,Marder K, Conneally PM, Nichols WC. 2003. Heterozygosity for amutation in the Parkin gene leads to later onset Parkinson disease.Neurology 60:796–801.

Gandhi S,WoodNW.2005.Molecular pathogenesis ofParkinson’s disease.Hum Mol Genet 14:2749–2755.

Greenamyre JT, Hastings TG. 2004. Parkinson’s–divergent causes, con-vergent mechanisms. Science 304:1120–1122.

Greene JC, Whitworth AJ, Kuo I, Andrews LA, Feany MB, Pallanck LJ.2003. Mitochondrial pathology and apoptotic muscle degenerationin Drosophila Parkin mutants. Proc Natl Acad Sci USA 100:4078–4083.

Hedrich K, Eskelson C, Wilmot B, Marder K, Harris J, Garrels J, Meija-Santana H, Vieregge P, Jacobs H, Bressman SB, et al. 2004. Distribution,type, and origin of Parkin mutations: Review and case studies. MovDisord 19:1146–1157.

Higgins JP, Thompson SG, Deeks JJ, Altman DG. 2003. Measuringinconsistency in meta-analyses. BMJ 327:557–560.

Hong Y, Zhang SB. 2006. Relationship between genetic polymorphism ofParkin gene and genetic susceptibility to Parkinson’s disease. Chin JContemp Neurol Neurosurg 6:291–294.

Ioannidis JP,Bernstein J, BoffettaP,Danesh J,DolanS,HartgeP,HunterD,Inskip P, Jarvelin MR, Little J, et al. 2005. A network of investigatornetworks in human genome epidemiology. Am J Epidemiol 162:302–304.

KayDM,MoranD,Moses L, Poorkaj P, Zabetian CP,Nutt J, Factor SA, YuCE, Montimurro JS, Keefe RG, et al. 2007. Heterozygous parkin pointmutations are as common in control subjects as in Parkinson’s patients.Ann Neurol 61:47–54.

KhanNL, Scherfler C, Graham E, Bhatia KP, QuinnN, Lees AJ, Brooks DJ,Wood NW, Piccini P. 2005. Dopaminergic dysfunction in unrelated,asymptomatic carriers of a single Parkin mutation. Neurology 64:134–136.

Kitada T, Asakawa S, Hattori N, Matsumine H, Yamamura Y, MinoshimaS, YokochiM,Mizuno Y, Shimizu N. 1998.Mutations in the Parkin genecause autosomal recessive juvenile parkinsonism. Nature 392:605–608.

Klein C, Lohmann-Hedrich K, Rogaeva E, Schlossmacher MG, Lang AE.2007.Deciphering the role of heterozygousmutations in genes associatedwith parkinsonism. Lancet Neurol 6:652–662.

Koziorowski D, Hoffman-Zacharska D, Slawek J, Szirkowiec W, Janik P,Bal J, Friedman A. 2010. Low frequency of the PARK2 gene mutations inPolish patients with the early-onset form of Parkinson disease. Parkin-sonism Relat Disord 16:136–138.

Lesage S, Lohmann E, Tison F, Durif F, Durr A, Brice A. 2008. Rareheterozygous parkin variants in French early-onset Parkinson diseasepatients and controls. J Med Genet 45:43–46.

Lincoln SJ,Maraganore DM, Lesnick TG, Bounds R, de AndradeM, BowerJH, Hardy JA, Farrer MJ. 2003. Parkin variants in North AmericanParkinson’s disease: Cases and controls. Mov Disord 18:1306–1311.

Lohmann E, Thobois S, Lesage S, Broussolle E, duMontcel ST, RibeiroMJ,Remy P, Pelissolo A, Dubois B, Mallet L. et al. 2009. A multidisciplinarystudyof patientswith early-onset PDwith andwithout parkinmutations.Neurology 72:110–116.

LohmuellerKE, PearceCL, PikeM, Lander ES,Hirschhorn JN. 2003.Meta-analysis of genetic association studies supports a contribution ofcommon variants to susceptibility to common disease. Nat Genet 33:177–182.

Lucking CB, Durr A, Bonifati V, Vaughan J, De Michele G, Gasser T,Harhangi BS, Meco G, Denefle P, Wood NW, et al. 2000. Associationbetween early-onset Parkinson’s disease and mutations in the Parkingene. N Engl J Med 342:1560–1567.

Lucking CB, Chesneau V, Lohmann E, Verpillat P, Dulac C, Bonnet AM,Gasparini F, AgidY,DurrA, BriceA. 2003. Coding polymorphisms in the

46 AMERICAN JOURNAL OF MEDICAL GENETICS PART B

Parkin gene and susceptibility to Parkinson disease. Arch Neurol 60:1253–1256.

Martinez HR, Gonzalez-Gonzalez H, Cantu-Martinez L, Rangel-Guerra R,Hernandez-Castillo CD, Vergara-Saavedra JJ, Ramos-Gonzalez MR,Cerda-Flores RM, Morales-Garza MA, Guerrero-Munoz MJ, et al.2010. PARKIN-coding polymorphisms are not associated withParkinson’s disease in a population from northeasternMexico. NeurosciLett 468:264–266.

Mata IF, AlvarezV,Garcia-Moreira V,Guisasola LM,RibacobaR, SalvadorC, Blazquez M, Sarmiento RG, Lahoz CH, Menes BB, et al. 2002. Single-nucleotide polymorphisms in the promoter region of the PARKIN geneand Parkinson’s disease. Neurosci Lett 329:149–152.

Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF. 1999.Improving the quality of reports of meta-analyses of randomised con-trolled trials: The QUOROM statement. Quality of Reporting of Meta-analyses. Lancet 354:1896–1900.

Nalls MA, Plagnol V, Hernandez DG, Sharma M, Sheerin UM, Saad M,Simon-Sanchez J, Schulte C, Lesage S, Sveinbjornsdottir S, et al. 2011.Imputation of sequence variants for identification of genetic risks forParkinson’s disease: Ameta-analysis of genome-wide association studies.Lancet 377:641–649.

Ning YP, Liu ZS, Li BQ, Xu YM. 2003. The study of relationship betweenpolymorphism in Parkin gene and sporadic Parkinson’s disease. J ClinNeurol 16:129–130.

OkubadejoN,BrittonA,CrewsC,AkinyemiR,Hardy J, SingletonA,Bras J.2008. Analysis of Nigerians with apparently sporadic Parkinson diseasefor mutations in LRRK2, PRKN and ATXN3. PLoS ONE 3:e3421.

Oliveira SA, ScottWK,Martin ER,NanceMA,Watts RL,Hubble JP, KollerWC, Pahwa R, Stern MB, Hiner BC, et al. 2003. Parkin mutations andsusceptibility alleles in late-onset Parkinson’s disease. Ann Neurol53:624–629.

Oliveri RL, Zappia M, Annesi G, Bosco D, Annesi F, Spadafora P, PasquaAA, Tomaino C, Nicoletti G, Pirritano D, et al. 2001. The Parkin gene isnot involved in late-onset Parkinson’s disease. Neurology 57:359–362.

Pankratz N, Kissell DK, Pauciulo MW, Halter CA, Rudolph A, Pfeiffer RF,Marder KS, Foroud T, NicholsWC. 2009. Parkin dosage mutations havegreater pathogenicity in familial PD than simple sequence mutations.Neurology 73:279–286.

Peng R, Gou Y, Yuan Q, Li T, Latsoudis H, Yuan G, Luo D, Liu X, CollierDA. 2003.Mutation screening and association analysis of the Parkin gene

in Parkinson’s disease patients from South-West China. Eur Neurol49:85–89.

Poorkaj P,Nutt JG, JamesD,Gancher S, Bird TD, Steinbart E, SchellenbergGD, Payami H. 2004. Parkin mutation analysis in clinic patients withearly-onset Parkinson’s disease. Am J Med Genet Part A 129A:44–50.

SatakeW,Nakabayashi Y,Mizuta I,Hirota Y, ItoC,KuboM,Kawaguchi T,Tsunoda T, Watanabe M, Takeda A, et al. 2009. Genome-wide associ-ation study identifies common variants at four loci as genetic risk factorsfor Parkinson’s disease. Nat Genet 41:1303–1307.

Satoh J, Kuroda Y. 1999. Association of codon 167 Ser/Asn heterozygosityin the Parkin gene with sporadic Parkinson’s disease. Neuroreport10:2735–2739.

SchragA, Schott JM. 2006. Epidemiological, clinical, and genetic character-istics of early-onset parkinsonism. Lancet Neurol 5:355–363.

Simon-Sanchez J, Schulte C, Bras JM, SharmaM,Gibbs JR, BergD, Paisan-Ruiz C, Lichtner P, Scholz SW,HernandezDG, et al. 2009. Genome-wideassociation study reveals genetic risk underlying Parkinson’s disease. NatGenet 41:1308–1312.

Sun M, Latourelle JC, Wooten GF, Lew MF, Klein C, Shill HA, Golbe LI,Mark MH, Racette BA, Perlmutter JS, et al. 2006. Influence of hetero-zygosity for Parkin mutation on onset age in familial Parkinson disease:The GenePD study. Arch Neurol 63:826–832.

Wang M, Hattori N, Matsumine H, Kobayashi T, Yoshino H, Morioka A,Kitada T, Asakawa S, Minoshima S, Shimizu N, et al. 1999. Polymor-phism in the parkin gene in sporadic Parkinson’s disease. Ann Neurol45:655–658.

Wang T, Liang ZH, Hou XM, Sun SG, Tong ET. 2003. Polymorphism ofParkin in Parkinson’s disease. Stroke Nervous Dis 10:87–89.

WestA, PeriquetM,Lincoln S, LuckingCB,NichollD, BonifatiV, RawalN,Gasser T, Lohmann E, Deleuze JF, et al. 2002. Complex relationshipbetween Parkin mutations and Parkinson disease. Am J Med Genet114:584–591.

Wu YR, Wu CH, Chao CY, Kuan CC, Zhang WL, Wang CK, Chang CY,Chang YC, Lee-Chen GJ, Chen CM. 2010. Genetic analysis of Parkin inearly onset Parkinson’s disease (PD): Novel intron 9g>a single nucleo-tide polymorphism and risk of Taiwanese PD. Am J Med Genet Part B153B:229–234.

ZhaoXP,XieHJ, TangGM,ZhaoWW,XuL, Su JJ,HaoYX,RenDM. 2003.The relationship between parkin gene polymorphism and Parkinson’sdisease. Chin J Neurol 36:87–89.

ZHANG ET AL. 47