Download Examination of G72 and D-amino-acid oxidase

Molecular Psychiatry (2004) 9, 203–207
& 2004 Nature Publishing Group All rights reserved 1359-4184/04 $25.00
www.nature.com/mp
ORIGINAL RESEARCH ARTICLE
Examination of G72 and D-amino-acid oxidase as genetic
risk factors for schizophrenia and bipolar affective
disorder
J Schumacher1, R Abon Jamra1, J Freudenberg1, T Becker2, S Ohlraun3,4, ACJ Otte1, M Tullius4,
S Kovalenko4, A Van Den Bogaert5, W Maier4, M Rietschel3,4, P Propping1, MM Nöthen5
and S Cichon5
1
Institute of Human Genetics, University of Bonn, Germany; 2Institute for Medical Biometry, Informatics and Epidemiology,
University of Bonn, Germany; 3Central Institute of Mental Health, Mannheim, Germany; 4Department of Psychiatry, University
of Bonn, Germany; 5Department of Medical Genetics, University of Antwerp, Belgium
Keywords: schizophrenia; bipolar affective disorder; G72;
DAAO
A recent study has suggested that the brain-expressed
genes for G72 and D-amino-acid oxidase (DAAO) exert
an influence on susceptibility to schizophrenia. Our aim
was to replicate this finding in German schizophrenic
patients and to assess whether G72 and DAAO might
also contribute to the development of bipolar affective
disorder. We genotyped seven single-nucleotide polymorphisms (SNPs) in the G72 gene and three in the
DAAO gene in 599 patients (299 schizophrenic, 300
bipolar) and 300 controls. At G72, individual SNPs and a
four-marker haplotype were associated with schizophrenia. The most significant SNP as well as the haplotype
were also associated with bipolar affective disorder
(BPAD). DAAO was associated with schizophrenia, but
not with BPAD. The association of variation at G72 with
schizophrenia as well as BPAD provides molecular
support for the hypothesis that these two major
psychiatric disorders share some of their etiologic
background.
Molecular Psychiatry (2004) 9, 203–207. doi:10.1038/
sj.mp.4001421
Over the past decade, inconsistency in linkage
findings as well as a host of nonreplications of
‘candidate gene’ association studies seemed to be
the inevitable fate of genetic research in psychiatry,
preventing any major breakthrough. Recently, however, the progress of the human genome project and
the development of high-throughput genotyping
technology have laid the foundations for a more
efficient search for the involved genes. As a result of
this development, systematic fine mapping in wellestablished linkage regions and replications of association findings have identified Dysbindin and
Neuregulin-1 as genetic factors contributing to the
etiology of schizophrenia.1–4
Recently, an association of the newly identified brain-expressed protein G72 and its likely
interaction partner D-amino-acid oxidase (DAAO)
with schizophrenia was suggested by the analysis
of single-nucleotide polymorphisms (SNPs) in
the respective genes.5 Both genes are located in
chromosomal regions showing evidence for linkage
with the disorder (G72 on 13q32–33; DAAO on
12q24).6
Given the importance of independent observation
of association findings in genetically complex diseases such as schizophrenia, we aimed at replicating
these findings in an independent sample of schizophrenic patients and controls from Germany. Furthermore, we hypothesized that G72 and DAAO might
also contribute to the development of bipolar affective
disorder (BPAD) for two reasons: first, although
schizophrenia and BPAD are characterized by distinct
clinical patterns of symptoms, a variety of psychopathological features are shared between these two
disease entities. This has brought forward the hypothesis of a shared etiologic background.7 Second,
linkage findings in the chromosomal regions harboring the genes for G72 and DAAO have also been
reported for BPAD.8–11
We genotyped seven SNPs in the G72 gene and
three SNPs in the DAAO gene. Our samples comprised 299 patients suffering from schizophrenia, 300
patients with BPAD, and a population-based sample
of 300 healthy controls.
Of the seven G72 variants typed, four SNPs (M12,
M15, M23 and M24) proved to be in significant
linkage disequilibrium with schizophrenia, the strongest evidence being observed for M23 (P¼0.033 for
allele frequencies) (Table 1). The same SNP was
significantly associated with BPAD (P¼0.013 for
allele frequencies) (Table 1). Haplotype analysis
strengthened the statistical significance for schizophrenia and—to a lesser extent—also for BPAD. We
tested the four-locus haplotype M12–M15–M23–M24
and found that it was associated with schizophrenia
(global, P¼0.032) (Table 2). Individual comparison of
the common haplotypes showed that haplotype
M12–M15–M23–M24 (C-G-T-A) was less frequent in
schizophrenic patients (P¼0.00023). The same haplotype was also significantly less frequent in patients
204
Molecular Psychiatry
Table 1
Statistical evaluation of allele and genotype frequencies for SNPs at the G72 and DAAO loci
SNPa
G72
M12
# rs1341402
# M15 (rs2391191)
# rs1935062
# M19 (rs778294)
# M23
M24
DAAO
MDAAO4
MDAAO5
MDAAO6
Distance Polymorphisma Controls Schizophrenia P-valueb
inter-SNP
(odds ratio)c
(kb)
0
12
3
8
14
43
12
C/T
T/C
G/A
A/C
C/T
T/C
A/T
0.64
0.75
0.64
0.67
0.69
0.54
0.50
(C)
(T)
(G)
(A)
(C)
(T)
(A)
0
3
2
A/C
C/T
T/G
0.66 (A)
0.72 (C)
0.61 (T)
0.59
0.77
0.59
0.65
0.72
0.47
0.44
(C)
(T)
(G)
(A)
(C)
(T)
(A)
0.72 (A)
0.78 (C)
0.53 (T)
0.048
0.352
0.037
0.494
0.270
0.033
0.036
(1.25)
(1.14)
(1.28)
(1.09)
(1.15)
(1.28)
(1.27)
Genotypic
testsd
BPAD
0.61
0.77
0.61
0.66
0.72
0.47
0.45
(C)
(T)
(G)
(A)
(C)
(T)
(A)
P-valueb
(odds ratio)c
0.225
0.437
0.222
0.715
0.337
0.013
0.064
(1.16)
(1.11)
(1.16)
(1.05)
(1.13)
(1.33)
(1.24)
Controls
Schizophrenia
1–1
1–2
2–2
1–1
1–2
2–2
0.39
0.56
0.40
0.44
0.47
0.29
0.25
0.50
0.38
0.48
0.46
0.45
0.49
0.50
0.11
0.06
0.12
0.10
0.08
0.22
0.25
0.32
0.59
0.33
0.40
0.52
0.23
0.29
0.53
0.37
0.52
0.51
0.41
0.49
0.50
0.15
0.04
0.15
0.09
0.07
0.28
0.18
P-valueb
(odds ratio)e
0.142
0.550
0.108
0.450
0.504
0.102
0.068
(1.34)
(1.48)
(1.41)
(1.21)
(1.21)
(1.42)
(1.59)
P-valueb
(odds ratio)e
BPAD
1–2
1–2
2–2
0.37
0.60
0.38
0.47
0.51
0.20
0.28
0.47
0.35
0.46
0.38
0.41
0.54
0.55
0.16
0.05
0.16
0.15
0.08
0.26
0.17
0.234
0.629
0.269
0.085
0.587
0.025
0.069
(1.53)
(1.17)
(1.48)
(1.57)
(1.19)
(1.68)
(1.61)
0.026 (1.33) 0.71 (A) 0.071 (1.25) 0.44 0.45 0.11 0.52 0.40 0.08 0.085 (1.40) 0.50 0.42 0.08 0.189 (1.28)
0.019 (1.37) 0.75 (C) 0.264 (1.16) 0.51 0.43 0.06 0.62 0.33 0.05 0.041 (1.52) 0.56 0.38 0.06 0.480 (1.22)
0.021 (1.31) 0.56 (T) 0.064 (1.26) 0.35 0.51 0.14 0.29 0.50 0.21 0.058 (1.60) 0.30 0.49 0.21 0.114 (1.53)
Second allele is the rare allele.
a
All SNPs were taken from Chumakov et al5; SNPs indicated by a # were also investigated by Hattori et al.12
b
P-values are not corrected for multiple testing.
c
Odds ratios are calculated using the more frequent alleles in affected individuals compared to controls.
d
1 is the frequent allele, 2 is the rare allele.
e
Odds ratios are given for the best-fitting model. The underlined genotype(s) are compared to the remaining genotype(s) in controls.
P-values are given in italic numerals. In addition, significant P-values are given in bold numerals.
Examination of G72 and DAAO as genetic risk factors
J Schumacher et al
Allelic tests
Examination of G72 and DAAO as genetic risk factors
J Schumacher et al
Table 2
205
Estimated haplotype frequencies and association significance
Frequencies and P-values
Haplotypea
Controls
Schizophrenia
P-valueb (odds ratio)
BPAD
P-valueb (odds ratio)
G72
M12–M15–M23–M24
T-A-C-T
0.17
T-A-T-A
0.19
C-G-C-T
0.29
C-G-T-A
0.31
0.18
0.24
0.35
0.20
0.7
0.093
0.07
0.00023
0.032
(1.07)
(1.31)
(1.29)
(0.57)
Global
0.19
0.20
0.35
0.24
0.497
0.815
0.069
0.032
0.08
(1.16)
(1.04)
(1.29)
(0.73)
Global
DAAO
MDAAO4–MDAAO5–MDAAO6
A-C-G
0.34
A-C-T
0.32
C-C-G
0.06
C-T-T
0.27
0.40
0.32
0.07
0.21
0.045
0.903
0.57
0.019
0.042
(1.29)
(1.00)
(1.18)
(0.72)
Global
0.40
0.30
0.05
0.24
0.032
0.459
0.471
0.213
0.325
(1.29)
(0.91)
(0.82)
(0.85)
Global
a
Only haplotypes with a frequency of Z5% in controls are given.
P-values are not corrected for multiple testing.
b
with BPAD (P¼0.032), but the global P-value did not
reach significance (P¼0.08) (Table 2). It should be
noted that M12 and M15 as well as M23 and M24
show a high degree of intermarker LD (Table 3) in our
study.
The three SNPs in the DAAO gene were associated
with schizophrenia, but failed to reach statistical
significance in the BPAD sample (Table 1). At the
haplotype level, the three-locus haplotype MDAAO4–
MDAAO5–MDAAO6 gave a significant global P-value
in schizophrenia (global, P¼0.042), but not BPAD
(global, P¼0.325) (Table 2).
It should be noted that the schizophrenia-associated haplotype that we observed for G72 is different
from that reported by Chumakov et al.5 We confirmed
the identity of the associated haplotypes in our
sample by sequencing homozygous genotypes at the
10-marker loci. A comparison of our results to the
study published by Hattori et al.12 shows that five
identical SNPs were analyzed in the two studies:
rs1341402, rs2391191 [M15], rs1935062, rs778294
[M19] and M23. Of these, rs1341402, rs1935062
and rs778294 [M19] produced significant results
in the single-marker analysis of Hattori et al,12 but
not in our study. On the other hand, rs2391191
[M15] and M23 were significant in our analysis, but
not in their study. A possible explanation for this
observation could be the presence of different risk
haplotypes in the populations under study, as has
been reported for Dysbindin and schizophrenia.1,2
It is also worth noting that the association of
haplotype M12–M15–M23–M24 with schizophrenia
and BPAD is mainly due to a highly significant underrepresentation of the very same haplotype (C-G-T-A)
in affected individuals as compared to controls. This
might be explained by a protective effect of this
haplotype.
Comparison of our results for the MDAAO-SNPs
with the data reported by Chumakov et al5 reveals that
the association of MDAAO4 and MDAAO5 is in the
opposite direction. While we find a significant
increase of the more frequent SNP alleles, Chumakov
et al5 observed an increase of the rarer SNP alleles.
For MDAAO6, both studies find a significant increase
of the rarer allele.
Confirming functional molecular data, Chumakov
et al5 found statistical evidence for an interaction
between G72 and DAAO. Using logistic regression
analysis, we tested different models of interaction for
schizophrenia as well as for BPAD. The obtained odds
ratios, however, did not favor a multiplicative/interactive effect over an additive effect (data not shown).
In conclusion, our results suggest that variability in
the G72 and DAAO genes is involved in the etiology
of schizophrenia in the German population. Our
findings independently confirm the report of an
association of G72 with schizophrenia in the French
Canadian and Russian population.5 Moreover, our
study provides evidence that variability in G72
contributes to the development of BPAD. This is in
accordance with a study by Hattori et al.,12 who report
an association of G72 with BPAD in two independent
samples of US-American descent. There is no strong
evidence for an involvement of DAAO in the etiology
of BPAD in our sample, although a small effect cannot
be excluded. Definitive conclusions will require the
investigation of larger samples or the investigation of
samples that could represent genetically more homogeneous subgroups of the disorder (eg early-onset
bipolar disorder13).
Based on our results, G72 seems to contribute to the
risk for schizophrenia and BPAD via the same
pathological pathway. Further studies will explore
the correlation between risk haplotypes and phenoMolecular Psychiatry
Examination of G72 and DAAO as genetic risk factors
J Schumacher et al
206
Table 3
D0 and r2 values for the investigated SNPs in the G72 and DAAO genes
G72
r2
M12
rs1341402
M15
rs1935062
M19
M23
M24
0.17
0.97
0.19
0.45
0.15
0.46
0.22
0.69
0.24
0.21
0.00
0.00
0.00
0.02
0.00
0.00
0.01
0.00
0.04
0.02
0.86
D0
M12
rs1341402
M15
rs1935062
M19
M23
M24
0.94
0.99
0.72
0.95
0.01
0.08
1.00
0.95
0.95
0.03
0.14
0.72
0.98
0.02
0.09
MDAAO5
MDAAO6
0.74
0.11
0.25
0.98
0.23
0.28
0.08
0.19
0.86
DAAO
MDAAO4
D0
MDAAO4
MDAAO5
MDAAO6
1.00
0.57
1.00
type in patients with different clinical features of
schizophrenia or BPAD. These findings may open a
door to a more precise classification of these two
major disorders.
Materials and methods
Subjects
All individuals included in the study were of German
descent and recruited in the same geographical area.
Patients were systematically ascertained at the Department of Psychiatry at the University of Bonn.
Written informed consent was obtained from all
patients and controls. All patients had been interviewed by experienced psychiatrists using the Structured Clinical Interview for DSM-IV Disorders.14
Lifetime ‘best estimate’ diagnoses according to DSMIV criteria15 were based on multiple sources of
information, including personal structured interview
(SCID I), medical records, and family history method.
Consensus diagnoses were performed by two psychiatrists, and whenever necessary, more psychiatrists were included in the decision process. The 299
schizophrenic patients comprised 151 males and 148
females (mean age of 38.6711.9 years); the 300 BPAD
cases consisted of 138 males and 162 females (mean
age of 42.3713.2 years). The 300 controls comprised
121 males and 179 females (mean age of 47.1715.2
years).
Genotyping and statistical analysis
Seven SNPs were genotyped in the G72 gene (M12
rs1341402, M15 (rs2391191), rs1935062, M19
(rs778294), M23, M24),5,12 and three in the DAAO
gene (MDAAO4, MDAAO5, MDAAO6).5 Genotypes
were determined by Masscodet Technology (QIAGEN
Molecular Psychiatry
Genomics). Genotype frequencies for cases and controls were in Hardy–Weinberg equilibrium. Singlemarker and haplotype-based analyses were performed. For the single-marker analysis, we used the
Armitage trend test (allelewise analysis) and the
standard w2 test (genotypewise analysis). For the
comparison of haplotype frequencies between cases
and controls, we used the computer program FAMHAP.16 FAMHAP estimates haplotype frequencies
using an expectation-maximization algorithm. Haplotype frequencies in cases and controls are then
compared in a global-likelihood-ratio w2 test. To
overcome the uncertainty of the w2 approximation,
we verified the validity of the P-values with the
simulation feature of the program. Pairwise LD
measures D0 and r2 were calculated from the estimated
haplotype frequencies.17
Due to the high degree of intermarker linkage
disequilibrium, a Bonferroni correction for multiple
testing would be too conservative. Therefore, uncorrected P-values are given.
Acknowledgements
This study was supported by the National Genomic
Network (NGFN) of the German Ministry of Education
and Research, the SFB 400 of the German Research
Council (DFG), the Fund for Scientific Research
Flanders (FWO, Grants G.0425.02 and G0438.03),
the Belgium Interuniversity Attraction Pole (IUAP
‘Molecular Genetics and Cell Biology’) and a Concerted Research Project (GOA) by the University of
Antwerp. AVDB holds a predoctoral position with the
Institute for the Promotion of Innovation by Science
and Technology in Flanders (IWT). We thank Dr E
Examination of G72 and DAAO as genetic risk factors
J Schumacher et al
Hattori for communicating primer sequences for SNPs
rs1341402 and rs1935062.
References
1 Straub RE, Jiang Y, MacLean CJ, Ma Y, Webb BT, Myakishev MV.
Genetic variation in the 6p22.3 gene DTNBP1, the human ortholog
of the mouse dysbindin gene, is associated with schizophrenia.
Am J Hum Genet 2002; 71: 337–348.
2 Schwab SG, Knapp M, Mondabon S, Hallmayer J, BorrmannHassenbach M, Albus M. Support for association of schizophrenia
with genetic variation in the 6p22.3 gene, dysbindin, in sib-pair
families with linkage and in an additional sample of triad families.
Am J Hum Genet 2003; 72: 185–190.
3 Stefansson H, Sigurdsson E, Steinthorsdottir V, Bjornsdottir S,
Sigmundsson T, Ghosh S. Neuregulin 1 and susceptibility to
schizophrenia. Am J Hum Genet 2002; 71: 877–892.
4 Stefansson H, Sarginson J, Kong A, Yates P, Steinthorsdottir V,
Gudfinnsson E. Association of neuregulin 1 with schizophrenia
confirmed in a Scottish population. Am J Hum Genet 2003; 72:
83–87.
5 Chumakov I, Blumenfeld M, Guerassimenko O, Cavarec L, Palicio
M, Abderrahim H. Genetic and physiological data implicating the
new human gene G72 and the gene for D-amino acid oxidase in
schizophrenia. Proc Natl Acad Sci USA 2002; 99: 13675–13680.
6 Sklar P. Linkage analysis in psychiatric disorders: the emerging
picture. Annu Rev Genomics Hum Genet 2002; 3: 371–413.
7 Maier W, Lichtermann D, Minges J, Hallmayer J, Heun R,
Benkert O. Continuity and discontinuity of affective disorders
and schizophrenia. Results of a controlled family study. Arch Gen
Psychiatry 1993; 50: 871–883.
8 Dawson E, Parfitt L, Roberts Q, Daniels J, Lim L, Sham P. Linkage
studies of bipolar disorder in the region of the Darier’s disease
gene on chromosome 12q23–24.1. Am J Med Genet 1995; 60:
94–102.
9 Jones I, Owen MJ, Craddock N. Further evidence of familial
cosegregation of major affective disorder and Darier’s disease. Mol
Psychiatry 1999; 4: 73.
10 Detera-Wadleigh SD, Badner JA, Berrettini WH, Yoshikawa T,
Goldin LR, Turner G. A high-density genome scan detects
evidence for a bipolar-disorder susceptibility locus on 13q32 and
other potential loci on 1q32 and 18p11.2. Proc Natl Acad Sci USA
1999; 96: 5604–5609.
11 Kelsoe JR, Spence MA, Loetscher E, Foguet M, Sadovnick AD,
Remick RA. A genome survey indicates a possible susceptibility
locus for bipolar disorder on chromosome 22. Proc Natl Acad Sci
USA 2001; 98: 585–590.
12 Hattori E, Liu C, Badner JA, Bonner TI, Christian SL, Maheshwari
M. Polymorphisms at the G72/G30 gene locus, on 13q33, are
associated with bipolar disorder in two independent pedigree
series. Am J Hum Genet 2003; 72: 1131–1140.
13 Grigoroiu-Serbanescu M, Martinez M, Nöthen MM, Grinberg M,
Sima D, Propping P et al . Different familial transmission patterns
in bipolar I disorder with onset before and after age 25. Am J Med
Genet 2001; 105: 765–773.
14 First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical
Interview for DSM-IV Axis I Disorders, Patient edn. New York:
Biometrics Research Department New York State Psychiatric
Institute, 1997.
15 American Psychiatric Association. Diagnostic and Statistical
Manual of Mental Disorders, 4th edn Washington, DC: American
Psychiatric Association, 1994.
16 Becker T, Knapp M. Maximum-likelihood estimation of
haplotype frequencies in nuclear families. (submitted for
publication).
17 Devlin B, Risch N. A comparison of linkage disequilibrium
measures for fine-scale mapping. Genomics 1995; 29:
311–322.
207
Correspondence: Dr J Schumacher, MD, Institute of Human
Genetics, University of Bonn, Wilhelmstr. 31, Bonn D-53111,
Germany. E-mail: [email protected]
Received 10 March 2003; revised 11 July 2003; accepted 21 July
2003
Molecular Psychiatry