G72/G30 in schizophrenia and bipolar disorder: review and meta-analysis.
ABSTRACT Association of the G72/G30 locus with schizophrenia and bipolar disorder has now been reported in several studies. The G72/G30 locus may be one of several that account for the evidence of linkage that spans a broad region of chromosome 13q. However, the story of G72/G30 is complex. Our meta-analysis of published association studies shows highly significant evidence of association between nucleotide variations in the G72/G30 region and schizophrenia, along with compelling evidence of association with bipolar disorder. But the associated alleles and haplotypes are not identical across studies, and some strongly associated variants are located approximately 50 kb telomeric of G72. Interestingly, G72 and G30 are transcribed in opposite directions; hence, their transcripts could cross-regulate translation. A functional native protein and functional motifs for G72 or G30 remain to be demonstrated. The interaction of G72 with d-amino acid oxidase, itself of interest as a modulator of N-methyl-d-aspartate receptors through regulation of d-serine levels, has been reported in one study and could be a key functional link that deserves further investigation. The association findings in the G72/G30 region, among the most compelling in psychiatry, may expose an important molecular pathway involved in susceptibility to schizophrenia and bipolar disorder.
- SourceAvailable from: Ole A Andreassen[Show abstract] [Hide abstract]
ABSTRACT: Several lines of evidence suggest that genome-wide association studies (GWAS) have the potential to explain more of the "missing heritability" of common complex phenotypes. However, reliable methods to identify a larger proportion of single nucleotide polymorphisms (SNPs) that impact disease risk are currently lacking. Here, we use a genetic pleiotropy-informed conditional false discovery rate (FDR) method on GWAS summary statistics data to identify new loci associated with schizophrenia (SCZ) and bipolar disorders (BD), two highly heritable disorders with significant missing heritability. Epidemiological and clinical evidence suggest similar disease characteristics and overlapping genes between SCZ and BD. Here, we computed conditional Q-Q curves of data from the Psychiatric Genome Consortium (SCZ; n = 9,379 cases and n = 7,736 controls; BD: n = 6,990 cases and n = 4,820 controls) to show enrichment of SNPs associated with SCZ as a function of association with BD and vice versa with a corresponding reduction in FDR. Applying the conditional FDR method, we identified 58 loci associated with SCZ and 35 loci associated with BD below the conditional FDR level of 0.05. Of these, 14 loci were associated with both SCZ and BD (conjunction FDR). Together, these findings show the feasibility of genetic pleiotropy-informed methods to improve gene discovery in SCZ and BD and indicate overlapping genetic mechanisms between these two disorders.PLoS Genetics 04/2013; 9(4):e1003455. · 8.17 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Ion channels are crucial components of cellular excitability and are involved in many neurological diseases. This review focuses on the sodium leak, G protein-coupled receptors (GPCRs)-activated NALCN channel that is predominantly expressed in neurons where it regulates the resting membrane potential and neuronal excitability. NALCN is part of a complex that includes not only GPCRs, but also UNC-79, UNC-80, NLF-1 and src family of Tyrosine kinases (SFKs). There is growing evidence that the NALCN channelosome critically regulates its ion conduction. Both in mammals and invertebrates, animal models revealed an involvement in many processes such as locomotor behaviors, sensitivity to volatile anesthetics, and respiratory rhythms. There is also evidence that alteration in this NALCN channelosome can cause a wide variety of diseases. Indeed, mutations in the NALCN gene were identified in Infantile Neuroaxonal Dystrophy (INAD) patients, as well as in patients with an Autosomal Recessive Syndrome with severe hypotonia, speech impairment, and cognitive delay. Deletions in NALCN gene were also reported in diseases such as 13q syndrome. In addition, genes encoding NALCN, NLF- 1, UNC-79, and UNC-80 proteins may be susceptibility loci for several diseases including bipolar disorder, schizophrenia, Alzheimer's disease, autism, epilepsy, alcoholism, cardiac diseases and cancer. Although the physiological role of the NALCN channelosome is poorly understood, its involvement in human diseases should foster interest for drug development in the near future. Toward this goal, we review here the current knowledge on the NALCN channelosome in physiology and diseases.Frontiers in Cellular Neuroscience 05/2014; 8:132. · 4.18 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Introduction: Bipolar disorder (BPD) is a severe illness with few treatments available. Understanding BPD pathophysiology and identifying potential relevant targets could prove useful for developing new treatments. Remarkably, subtle impairments of mitochondrial function may play an important role in BPD pathophysiology. Areas covered: This article focuses on human studies and reviews evidence of mitochondrial dysfunction in BPD as a promising target for the development of new, improved treatments. Mitochondria are crucial for energy production, generated mainly through the electron transport chain (ETC) and play an important role in regulating apoptosis and calcium (Ca(2+)) signaling as well as synaptic plasticity. Mitochondria move throughout the neurons to provide energy for intracellular signaling. Studies showed polymorphisms of mitochondria-related genes as risk factors for BPD. Postmortem studies in BPD also show decreased ETC activity/expression and increased nitrosative and oxidative stress (OxS) in patient brains. BPD has been also associated with increased OxS, Ca(2+) dysregulation and increased proapoptotic signaling in peripheral blood. Neuroimaging studies consistently show decreased energy levels and pH in brains of BPD patients. Expert opinion: Targeting mitochondrial function, and their role in energy metabolism, synaptic plasticity and cell survival, may be an important avenue for development of new mood-stabilizing agents.Expert Opinion on Therapeutic Targets 07/2014; · 4.90 Impact Factor
G72/G30 in Schizophrenia and Bipolar Disorder:
Review and Meta-analysis
Sevilla D. Detera-Wadleigh and Francis J. McMahon
Association of the G72/G30 locus with schizophrenia and bipolar disorder has now been reported in several studies. The G72/G30 locus
may be one of several that account for the evidence of linkage that spans a broad region of chromosome 13q. However, the story of
G72/G30 is complex. Our meta-analysis of published association studies shows highly significant evidence of association between
nucleotide variations in the G72/G30 region and schizophrenia, along with compelling evidence of association with bipolar disorder.
But the associated alleles and haplotypes are not identical across studies, and some strongly associated variants are located ?50 kb
telomeric of G72. Interestingly, G72 and G30 are transcribed in opposite directions; hence, their transcripts could cross-regulate
translation. A functional native protein and functional motifs for G72 or G30 remain to be demonstrated. The interaction of G72 with
D-amino acid oxidase, itself of interest as a modulator of N-methyl-D-aspartate receptors through regulation of D-serine levels, has been
reported in one study and could be a key functional link that deserves further investigation. The association findings in the G72/G30
region, among the most compelling in psychiatry, may expose an important molecular pathway involved in susceptibility to
schizophrenia and bipolar disorder.
Key Words: DAAO, DAO, G72, G30, gene
years. However, the story of G72/G30 in this field is complex. In
this article, we will review the genetics, genomics, and biology of
G72/G30, attempting to place it into the context of what we
already know about these illnesses. We will present a meta-
analysis of the published association studies that shows highly
significant evidence of association between several markers near
G72/G30 and schizophrenia, along with compelling, albeit less
significant, evidence of association with bipolar disorder. We
also will outline some of the key unanswered questions about
this locus, its biology, and its potential role in major mental
illness. We begin with a review of the genetic linkage evidence
that led to the discovery of G72/G30 in 2002.
he genetic association between alleles near the G72/G30
locus and schizophrenia and bipolar disorder is one of the
major developments in psychiatric genetics in recent
Linkage to Chromosome 13q
Linkage analysis on diverse sample collections has provided
accumulating evidence that a large portion of chromosome 13q,
spanning some 68 Mb from 13q12 to 13q34, may be involved in
susceptibility to schizophrenia (Figure 1; Abecasis et al 2004;
Blouin et al 1998; Brzustowicz et al 1999; Camp et al 2001;
Cardno et al 2001; Faraone et al 2002; Lin et al 1997; Shaw et al
1998; Wijsman et al 2003). Consistent with the results of these
studies, a meta-analysis on genome scans found strong evidence
for linkage to schizophrenia on chromosome 13q (Badner and
Gershon 2002), although the rank-order method has not sup-
ported this finding (Lewis et al 2003). Thus, at least on the basis
of the linkage evidence, chromosome 13q could harbor one or
more genes involved in schizophrenia.
Genetic studies of bipolar disorder recapitulate the breadth
and variability of the weight of linkage evidence on chromosome
13q in schizophrenia (Figure 1). Several studies have shown
evidence of linkage to distal 13q (Detera-Wadleigh et al 1999;
Kelsoe et al 2001; Liu C et al 2001; Liu J et al 2003; Park et al 2004;
Shaw et al 2003). Others have reported linkage signals on the
proximal region of 13q (Badenhop et al 2001; Cichon et al 2001;
Ginns et al 1996; McInnis et al 2003a, 2003b; Potash et al 2003).
Of the three large meta-analyses of bipolar disorder linkage
scans published to date, genomewide significant linkage to chro-
mosome 13q was demonstrated in one study (Badner and Gershon
2002) but not in the other two studies (McQueen et al 2005;
Segurado et al 2003). If more than one susceptibility locus exists on
13q and if each locus is relatively uncommon, the consequent loss
of power in the linkage analysis could make individual loci very
difficult to detect consistently across studies, even by meta-analysis.
Other Mental Disorders
Studies in other psychiatric diseases have detected suggestive
evidence of linkage to 13q in autism (Barrett et al 1999), anorexia
nervosa (Devlin et al 2002), panic disorder (Hamilton et al 2003),
Tourette syndrome (Curtis et al 2004), language impairment
(Bartlett et al 2004), and recurrent major depression (McGuffin et
al 2005). Recently, association of variants on SLITRK1 on 13q31.1
with Tourette syndrome has been reported (Abelson et al 2005).
It is possible that one or more of the inferred loci on 13q underlie
genetic risk to multiple mental abnormalities.
Discovery of G72/G30
Despite multiple linkage reports extending over almost the
entire long arm of chromosome 13, to our knowledge only one
published study has so far systematically examined a 13q-linkage
region (Chumakov et al 2002). A single-nucleotide polymor-
phism (SNP) search and a case–control association study in
French Canadian probands with schizophrenia targeted an
?5-Mb region on 13q33. Two distinct clusters of SNPs detected
evidence of association, at significance levels that appeared to
exceed what would be expected by chance alone, even after
adjustment for multiple testing. Selected SNPs from both clusters
were genotyped in a second sample of cases, this time from
Russia. Several SNPs in the more distal cluster showed significant
From the Genetic Basis of Mood and Anxiety Disorders, Mood and Anxiety
Program, National Institute of Mental Health Intramural Research Pro-
gram, National Institutes of Health, U.S. Department of Health and Hu-
man Services, Bethesda, Maryland.
Convent Drive, Building 35, Room 1A/203, Bethesda, MD 20892-3719;
Received June 13, 2005; revised January 19, 2006; accepted January 28,
BIOL PSYCHIATRY 2006;60:106–114
© 2006 Society of Biological Psychiatry
allelic and haplotypic differences in cases compared with con-
trols, but these SNPs mapped to a region devoid of known genes.
Sequencing of bacterial artificial chromosome (BAC) clones in
the region and sequence annotation yielded two novel, overlap-
ping genes transcribed in opposite orientations, referred to as
G72 and G30 (University of California Santa Cruz [UCSC] genome
browser, www.genome.ucsc.edu, May 2004).
The provisional reference gene sequence (National Center for
Biotechnology Information (NCBI) Reference Sequences database
[RefSeq]) NM_172370 for G72 encodes a 742-bp mRNA. The gene
coding for this transcript is relatively small (?25 kb), consisting of
five exons (Chumakov et al 2002; UCSC genome browser, May
2004; Ensembl genome browser, http://www.ensembl.org/; Figure
2). Up to five other alternate transcripts are supported by existing
cDNA clones (Chumakov et al 2002; Hattori et al 2003; UCSC
genome browser, May 2004; Ensembl genome browser). Tran-
scripts of G72 have been demonstrated only in primates (Chumakov
et al 2002). In silico examination of available sequences from other
organisms confirms the absence of G72 orthologs, except in dog, in
which sequences that align with the five exons of the human G72
RefSeq gene are present but lack intron sequences (UCSC genome
browser, May 2004). It is unclear whether the dog G72 sequences
represent a functional gene, and there are no published studies
investigating G72 in dog.
Reverse-transcription PCR revealed the expression of G72 in
the amygdala, caudate nucleus, spinal cord, and testis (Chuma-
kov et al 2002). An in vitro transcription and translation assay
produced the G72, but not G30, protein, suggesting that only the
G72 gene is actively translated. The apparent absence of a
protein product for G30, which is predicted to be a ?47-kb gene
with seven exons (UCSC genome browser, May 2004), has
hindered further work on this gene. A later study using real-time
PCR quantification of mRNA showed increased transcripts of
G72, but not G30, in the dorsolateral prefrontal cortex of
postmortem brains from about 40 cases with schizophrenia,
compared with controls (Korostishevsky et al 2004). Thus, at
least at the level of transcription, G72 may have greater relevance
than G30 in schizophrenia.
G72 was a novel gene with no recognizable functional motifs;
hence, it was important to explore the interaction properties
of the protein. Yeast two-hybrid screening using the carboxy-
terminal 89-amino acid peptide as “bait” captured a clone
encoding D-amino acid oxidase (DAO; Chumakov et al 2002), an
important modulator of N-methyl D-aspartate (NMDA) receptors
via regulation of D-serine levels. Physical binding between G72
and DAO was confirmed by column binding and elution. Fur-
ther, it was shown that addition of increasing amounts of G72
stimulated an increase in the oxidation of D-serine by DAO.
Subsequently, G72 was given the designation D-amino acid
oxidase activator (UCSC genome browser; Ensembl genome
browser; National Center for Biotechnology Information [NCBI],
The Chumakov et al (2002) study was a landmark in several
ways. It was the first study to detect robust evidence of genetic
association within the 13q linkage region for schizophrenia. It was
the first study to describe G72/G30 and the first to demonstrate its
line) and schizophrenia (bottom horizontal line) are numbered and represented by vertical bars. Full vertical bars correspond to maximum LOD or NPL ?2, and
1, Cichon et al (2001); 2, McInnis et al (2003a); 3, Ginns et al (1996); 4, Maziade et al (2004); 5, Badenhop et al (2001); 6, McInnis et al (2003b); 7, Potash et al (2003); 8, Stine
et al (1997); 9, Kelsoe et al (2001); 10, Shaw et al (2003); 11, Detera-Wadleigh et al (1999); 12, Liu et al (2001); 13, Liu et al (2003); and 14, Park et al (2004). Studies in
schizophrenia: 1, Camp et al (2001); 2, Shaw et al (1998); 3, Wijsman et al (2003); 4, Lin et al (1997); 5, Blouin et al (1998); 6, Brzustowicz et al (1999); 7, Faraone et al (2002);
8, Cardno et al (2001); and 9, Abecasis et al (2004). LOD, logarithm of odds; NPL, nonparametric LOD.
S.D. Detera-Wadleigh and F.J. McMahon
BIOL PSYCHIATRY 2006;60:106–114 107
expression in brain. It also was the first study to report evidence of
through which variation within G72 could affect NMDA signaling.
A close review of the Chumakov et al study (2002) also raises
several questions that subsequent studies have not yet been able
to address. Single-nucleotide polymorphisms that detected asso-
ciation with schizophrenia map to varied locations covering ?95
kb and extending 40–50 kb downstream from the predicted
coding region of G72/G30. Western blots that used lysates from
either nontransfected cells or primate tissues were not presented.
Such data could have demonstrated the existence of a native G72
protein. This is an important issue, because one recent study
cited G72 as a potential example of genes consisting entirely of
repetitive elements that encode RNAs but not proteins (Britten
2004). Antiserum raised against a G72 peptide generated by
plasmid overexpression consistently recognized a protein band
in G72-transfected COS7 cells but the molecular weight did not
match the expected size. In addition, a distinctive feature of the
G72 gene is the variability of potential splice acceptors in several
exons that could underlie the complexity of mRNA species and
may partly account for the uncertainty in the functional protein
products (Hattori et al 2003). Finally, the proposed interaction
between G72 and DAO that is the crucial link between G72 and
a well-studied signaling pathway needs to be replicated and
more fully characterized.
Despite these questions, findings from several subsequent
studies tend to advance a role for G72/G30 in the overall risk for
schizophrenia and bipolar disorder. We will now turn to a review
of those studies.
Meta-analysis of Published Associations with G72/G30
To assess the total evidence of association between markers
near G72 and various phenotypes, we performed a meta-analysis
of published studies. A total of 11 genetic-association studies
were located in PubMed (Addington et al 2004; Chen et al 2004;
Chumakov et al 2002; Hall et al 2004; Hattori et al 2003;
Korostishevsky et al 2004; Mulle et al 2005; Schumacher et al
2004, 2005; Wang et al 2004; Zou et al 2005). One study was
excluded because alleles were not identified in the individual
marker analyses (Hall et al 2004). The remaining studies are
shown in Table 1. Seven studies focused on schizophrenia or
psychosis not otherwise specified, two studies focused on bipo-
lar affective disorder, one study focused on both, and one study
focused on panic disorder.
Upon tabulating the results, it became clear that the associated
alleles varied across studies at every marker except rs1421292.
These discrepancies were not fully resolvable by accounting for the
phenotype tested, the study design (transmission–disequilibrium
test vs. case–control), or ethnicity. Because it has been argued that
genetic-association studies should be compared at the level of the
gene (Neale and Sham 2004) and because differing alleles can be in
linkage disequilibrium with the same disease allele in differing
have been studied. The gene shown above is the designated reference sequence (National Center for Biotechnology Information Reference Sequences
database; University of California Santa Cruz genome browser). The haplotype block structure was generated by using the default method (Gabriel et al 2002)
applied to HapMap SNPs spanning a ?97.5-kb region. The display corresponds to the extent of intermarker linkage disequilibrium and graded shading
intensity on the squares, in other words, from white to red corresponds toD=values from 0 to 1, respectively. The locations of SNPs that have been analyzed
in various studies are indicated by arrows, and the SNP position in the haplotype blocks is shown. The sizes of exons and introns are not drawn to scale. The
equivalent “rs” designations for the M-SNPs are the following: M12 (rs3916965), M13 (rs3916966), M14 (rs3916967), M15 (rs2391191), M16 (rs3918341), M18
(rs947267), M19 (rs778294), M20 (rs3916970), M21 (rs3916971), M22 (rs778293), M23(rs3918342), and M24 (rs1421292).
108 BIOL PSYCHIATRY 2006;60:106–114
S.D. Detera-Wadleigh and F.J. McMahon
populations, we decided for the purposes of the meta-analysis to
combine results without regard to allele identity. The association of
differing marker alleles across samples has generated controversy in
the field of complex genetics because the phenomenon character-
izes many of the most prominent findings (Hirschhorn at al 2002;
Owen et al 2005; Tusie-Luna 2005). A complete discussion of this
issue is beyond the scope of this article.
We could not perform a complete test of the gene-based
hypothesis as recommended in Neale and Sham (2004), because
that would have required that we consider jointly all common
variation within the candidate gene. To date, we are aware of no
comprehensive studies of the G72 region. Instead, the published
studies have evaluated between 3 and 16 markers spanning 50 MB.
To accommodate the great variation across studies in the test
statistics used and the way in which results were reported, the
meta-analysis was performed with the Fisher method, which can
combine p values across studies without need of the primary data
(Fisher 1954). This method does not take into account any
publication bias, which is difficult to rule out in any meta-
analysis. A funnel-plot analysis of the published studies did not
indicate publication bias (data not shown).
Sample sizes of the published studies have been modest,
comprising fewer than 500 probands or trios. Despite the lack of
power implied by such study designs, most studies report one or
more significant associations. The results do not appear to be
attributable to weaknesses of either the TDT (Gordon et al 2001;
Spielman et al 1993) or case– control designs (Cardon and Palmer
2003), because significant results are seen in multiple samples
and with both study designs.
The meta-analysis revealed highly significant evidence of
association between several markers near G72/G30 and schizo-
phrenia, along with compelling, albeit less significant, evidence
of association with bipolar disorder. Combined results in schizo-
phrenia are significant for several markers and are highly signif-
icant (p ? .001) for three markers spanning more than 82 kb.
Only three studies have been published in bipolar-disorder
samples, but significant (p ? .05) combined results are seen for
three markers spanning more than 74 kb. When results are
combined across studies and phenotypes, eight markers show
significant combined p values, including three consecutive mark-
ers near rs3918342 (M23), all of which have combined p values of
?.001. Interestingly, these markers are located more than 50 kb
distal to the predicted coding region of G72, with no other
known human gene in sight. Most of these markers remain
Table 1. Meta-analysis of Published Association Studies
dbSNP Accession/Alias/Position/Forward Alleles
.04 .5 .5.5 .5
Total Studies 10236333692
aA number of cases (CC) or trios (TDT).
cOver transmitted allele.
S.D. Detera-Wadleigh and F.J. McMahon
BIOL PSYCHIATRY 2006;60:106–114 109
significant at the meta-analysis level even after removal of the
initial study (Chumakov et al 2002) from the analysis. Many
studies also performed haplotype analyses, with several highly
significant results, but because of differences in window size,
alleles, and haplotype estimation methods, we did not attempt a
meta-analysis of the haplotype findings in G72.
In summary, this meta-analysis supports significant associa-
tion of markers in the G72 region with both schizophrenia and
bipolar affective disorder. It remains unclear why the associated
alleles vary across studies. It is also unclear whether the associ-
ated markers all reflect the same functional variation or whether
multiple functionally important variants actually are present in a
large region encompassing G72. Identification of functional
variants will probably require biological as well as additional
Linkage Disequilibrium at G72/G30
We evaluated the patterns of linkage disequilibrium (LD) at
the G72/G30 locus covering the ?97.5-kb interval between
rs7331194 (?3 kb upstream of M12) and rs1421292 (M24) that
includes the 5= and 3= flanking regions of the gene (Figure 2A).
We used data from HapMap (release 19/phaseII Oct05; http://
www.hapmap.org) derived from trios of European origin (CEU),
analyzed with Haploview (version 3.2; Barrett et al 2005) on
SNPs that had minor allele frequencies (MAF) of ?.1, because
rare SNPs may display patterns of LD that are very sensitive to
sample size. The locus is defined by two major regions of high
LD (Figure 2A), suggesting that the locus could have arisen by
accretion, one ancestral segment gaining another during evolu-
tion. Each region is divided into four smaller subregions, and
within each of these, LD (r2or D=) is variable. M20 maps to a site
of little or no LD. A 29-kb LD region contains M22, M23, and M24,
SNPs that showed strong association with schizophrenia in
published studies and also in our meta-analysis (see previous
section of this article). Notably, there apparently is no LD
between the gene regions that contain SNPs that show asso-
ciation in various published studies. This suggests that there
may be more than one functional allele contributing to disease
susceptibility in the G72 region and may help explain the
inconsistency between studies in the associated marker al-
To explore possible differences in LD between different
continental populations, we examined the haplotype block
structures in Han Chinese (CHB) and Yorubans of the Ibadan
tribe in Nigeria (YRI) by using HapMap genotypes. The LD
patterns generated in CHB were similar to the CEU-derived
profile, but LD was generally less (Supplement 1). In contrast to
the CEU pattern, in the CHB sample, the region was represented
by nine clusters of markers in high LD, with rs1935062 residing in
a different cluster than M22, which itself resides in a cluster that
is separate from that containing M23 and M24. In both CEU and
CHB, rs2391191, which gives rise to a missense mutation, maps
Table 1. (continued)
dbSNP Accession/Alias/Position/Forward Alleles
et al (2002)
Hattori et al
Chen et al
et al (2005)
et al (2004)
et al (2004)
et al, (2004)
Wang et al,
et al, (2004)
Zou et al
Mulle et al
110 BIOL PSYCHIATRY 2006;60:106–114
S.D. Detera-Wadleigh and F.J. McMahon
to block 1. The excess allele in French Canadian and mixed U.S.
cases is the major allele G (0.658 in CEU, HapMap), and in Han
Chinese, it is the major allele A (0.522, CHB, HapMap; Table 1).
It is interesting that in both continental groups, it is the major
allele that is overrepresented in schizophrenia cases, even
though the actual major allele differs between the groups. In the
German sample, it is the minor allele that was reported to be in
excess in schizophrenics (Table 1). We caution that the current
HapMap allele frequency estimates may change when larger
population samples are analyzed.
In the YRI sample, as expected, diversity within and between
sets of markers is much more prominent than in CEU and CHB
(Supplement 2). M20 and M21 are located in a region of no LD
that divides the locus into two major regions. M22 and M23 are in
separate regions that both include clusters of common SNPs. M24
(?1 kb distal to the last SNP on block 10) was one of the many
SNPs that had MAF of ?.1 and is therefore not included in the
haplotype structure we estimated. Although LD is variable, the
haplotype grouping of M12, M13, and rs1935058 appears to have
been preserved in YRI. In our meta-analysis, the most significant
association for bipolar disorder is at rs1935062, ?8.7 kb distal to
rs2391191 (M15), an exonic SNP that displayed association with
schizophrenia but not with bipolar disorder. In YRI, these SNPs are
predicted to be in a region with little or no LD (Supplement 2).
D-Amino Acid Oxidase
As noted above, interaction between G72/G30 and DAO first
was proposed by Chumakov et al (2002) on the basis of yeast
two-hybrid and co-immunoprecipitation assays. Although inter-
action of G72 and DAO remains to be established in vivo, DAO
is of interest by itself because of its known upstream effect on
NMDA receptors. Hypofunction of NMDA-mediated neurotrans-
mission has been postulated to be involved in the pathophysi-
ology of schizophrenia (Coyle 2004), and many drugs used for
the treatment of schizophrenia impinge on this receptor subtype
(Javitt 2004). It appears intuitive that similar mechanisms underlie
N-methyl D-aspartate receptors are ionotropic, voltage-gated
receptors that are composed of at least four subunits, with at least
one NR1 (GluR?1) and one or more NR2A-D (GluRε1–ε4)
subunits that include binding sites for glutamate and glycine
(Javitt 2004). In the presence of glutamate, NMDA receptor
activity is modulated by glycine and D-serine ligands, which both
occupy the glycine recognition site (Mothet et al 2000). D-serine,
a potent activator of NMDA receptors, is abundant in brain and
enriched in astrocytes (Mothet et al 2000). D-serine is formed by
the racemization of L-serine by glial serine racemase (Wolosker et
al 1999). Decreased serum levels of D-serine have been reported
in Japanese subjects with schizophrenia (Hashimoto et al 2003),
Figure 3. Gene structure of DAO, haplotype block structure in the gene and flanking regions, and single-nucleotide polymorphisms (SNPs) that have been
disequilibrium and graded shading intensity on the squares; in other words, from white to black corresponds to D= values from 0 to 1, respectively. The
S.D. Detera-Wadleigh and F.J. McMahon
BIOL PSYCHIATRY 2006;60:106–114 111
and D-serine supplementation of standard anti-psychotic therapy
appeared to improve symptoms in some studies (Tsai et al 1999;
Tuominen et al 2005).
D-amino acid oxidase is a flavoenzyme that regulates the
concentration of D-serine by catalyzing its oxidation. D-amino
acid oxidase maps to 12q24.11, a linkage region for bipolar
disorder (Craddock et al 1994; Curtis et al 2004; Ewald et al 1998;
Green et al 2005; Morrisette et al 1999; Shink et al 2004). To our
knowledge, only one published report has presented evidence
linking schizophrenia to chromosome 12q, and the evidence is
modest (Wilcox et al 2002).
To date, few association studies have tested DAO in psychi-
atric samples. In the original Chumakov et al (2002) study,
case–control analysis showed allelic association between four
DAO SNPs and schizophrenia. The evidence was less statistically
significant than that for G72, and there was little evidence of
association between DAO and G72 by logistic regression analysis
(Chumakov et al 2002). A study in German cases and controls
appeared to support an association between DAO SNPs and
schizophrenia but not bipolar disorder (Schumacher et al 2004).
Analysis of more than 500 Chinese cases with schizophrenia and
matched controls found significant evidence of association with
rs3741775 (MDAAO-6), and haplotypes from six SNPs also
yielded evidence for association (Liu et al 2004; Figure 3). These
reports suggest that sequence variation in DAO may contribute to
schizophrenia susceptibility in at least two different ethnic
groups, but more studies are needed.
The absence of a known rodent homologue of G72 has
hindered work on the biology of this gene. If the G72 exonic
sequences in dog represent a functional gene, the dog could be
used as an animal model to investigate detailed aspects of gene
A recent study of mutant mice lacking DAO showed that
homozygous DAO?/?mice had high levels of D-serine and
significantly reduced stereotypy and rotational activity after
administration of NMDA receptor antagonists than did wild-type
and DAO?/? mice (Hashimoto et al 2003). This suggests that
DAO activity is somehow increased in schizophrenia, leading to
increased oxidation of D-serine in the disease. This, in turn, might
reduce the available effective concentration of D-serine, eliciting
a decline in NMDA receptor potentiation. Consistent with this
idea is the observation that DAO?/?mice have indiscernible
levels of D-serine in the cerebellum but that DAO?/?mice
display high levels of the co-agonist (Hashimoto et al 2005).
The G72/G30 gene region appears to be a common locus for
schizophrenia and bipolar disorder, but the associated alleles are
not consistent across studies. This suggests that there may not
exist any distinctive haplotype that correlates with susceptibility.
Possible causes for this phenomenon include the following:
(1) allelic heterogeneity; (2) phenotypic heterogeneity; (3) geno-
typing errors; (4) analytical errors, particularly when haplotypes
are derived from cases and controls; and (5) false-positive
findings in some studies. Another issue of interest is that the
distal SNPs that show association in multiple studies (M22, M23,
and M24) map to a region devoid of known genes. This poses a
challenge in identifying risk variants. It may require examination
of functional consequences of the sequence alterations such as
effects on gene regulation or mRNA stability. As discussed, other
associated SNPs are located proximal to and within G72. Hence,
the possibility that the associated variants are in linkage
disequilibrium with exonic alterations can not be ruled out.
The functional G72 protein may be expressed at very low
levels (if at all), or it could have a short half-life. Alternatively, if
both G72 and G30 are transcribed in vivo, the two complemen-
tary transcripts could act as mutually interfering RNAs, inhibiting
each other’s translation. In this case, it may well be the relative
stoichiometric concentrations of G72 and G30 that actually are
altered in disease states. Finally, as illustrated above, the linkage
signals mark almost the entire chromosome 13q. The G72/G30
association may account for only a part of this linkage. Therefore,
we speculate that at least one other gene on chromosome 13q
will be found to contain disease-predisposing variants.
The few reports that show association between DAO variants
and schizophrenia merit verification in additional independent,
preferably large, samples. Most of the published studies have
focused on cases and controls, which are subject to false results
caused by population stratification. D-amino acid oxidase is
encoded by a small gene; nevertheless, analysis with a tight grid
of SNPs that includes flanking sequences, particularly the pro-
moter region, would be warranted (Figure 3).
Although more work is needed to explain the differences in
alleles and haplotypes across studies, the overall evidence
strongly supports association of the G72/G30 locus with both
schizophrenia and bipolar disorder. The function of the G72/G30
locus; the existence of a native protein product; and the possi-
bility that at least some of the association evidence points to
another, as-yet undiscovered gene remain unsettled issues. The
putative interaction between G72 and DAO, along with the
latter’s upstream effect on NMDA neurotransmission, exposes a
key molecular pathway in susceptibility to schizophrenia and
bipolar disorder. We still need to learn much more about the
biology of the G72/G30 locus before we can begin to turn the
data into an understanding of genes and disease.
This work was supported by the National Institute of Mental
Health Intramural Research Program. Data used in the linkage
disequilibrium analysis were produced by the International
Supplementary material cited in this article is available
Abecasis GR, Burt RA, Hall D, Bochum S, Doheny KF, Lundy SL, et al (2004):
ulation of Afrikaners reveals evidence for linkage and uniparental dis-
omy on chromosome 1.Am J Hum Genet74:403–417.
Abelson JF, Kwan KY, O’Roak BJ, Baek DY, Stillman AA, Morgan TM, et al
drome. Science 310:317–320.
childhood-onset schizophrenia and psychosis not otherwise specified.
Badenhop RF, Moses MJ, Scimone A, Mitchell PB, Ewen KR, Rosso A, et al
(2001): A genome screen of a large bipolar affective disorder pedigree
supports evidence for a susceptibility locus on chromosome 13q. Mol
of bipolar disorder and schizophrenia. Mol Psychiatry 7:405–411.
Barrett JC, Fry B, Maller J, Daly MJ (2005): Haploview: Analysis and visualiza-
tion of LD and haplotype maps. Bioinformatics 21:263–265.
autosomal genomic screen for autism. Collaborative Linkage Study of
112 BIOL PSYCHIATRY 2006;60:106–114
S.D. Detera-Wadleigh and F.J. McMahon
(1998): Schizophrenia susceptibility loci on chromosomes 13q32 and
8p21. Nat Genet 20:70–73.
Britten R (2004): Coding sequences of functioning human genes derived
entirely from mobile element sequences. Proc Natl Acad Sci U S A 101:
(1999): Linkage of familial schizophrenia to chromosome 13q32. Am
Camp NJ, Neuhausen SL, Tiobech J, Polloi A, Coon H, Myles-Worsley M
(2001): Genomewide multipoint linkage analysis of seven extended Pa-
lauan pedigrees with schizophrenia, by a Markov-chain Monte Carlo
Cardon LR, Palmer LJ (2003): Population stratification and spurious allelic
association. Lancet 361:598–604.
(2004): Findings in an independent sample support an association be-
13q33. Mol Psychiatry 9:87–92.
Chumakov I, Blumenfeld M, Guerassimenko O, Cavarec L, Palicio M, Abder-
human gene G72 and the gene for D-amino acid oxidase in schizophre-
al (2001): A possible susceptibility locus for bipolar affective disorder in
chromosomal region 10q25–q26. Mol Psychiatry 6:342–349.
the proximate cause? Biochem Pharmacol 68:1507–1514.
Craddock N, McGuffin P, Owen M (1994): Darier’s disease cosegregating
with affective disorder. Br J Psychiatry 165:272.
Curtis D, Brett P, Dearlove AM, McQuillin A, Kalsi G, Robertson MM, et al
(2004): Genome scan of Tourette syndrome in a single large pedigree
Detera-Wadleigh SD, Badner JA, Berrettini WH, Yoshikawa T, Goldin LR,
Turner G, et al (1999): A high-density genome scan detects evidence for
Genome-wide multipoint linkage analyses of multiplex schizophre-
Ewald H, Degn B, Mors O, Kruse TA (1998): Significant linkage between
bipolar affective disorder and chromosome 12q24. Psychiatr Genet 8:
ans Affairs cooperative study sample. Am J Med Genet 114:598–604.
United Kingdom: Oliver & Boyd.
Gabriel SB, Schaffner SF, Nguyen H, Moore JM, Roy J, Blumenstiel B, et al
Ginns EI, Ott J, Egeland JA, Allen CR, Fann CS, Pauls DL, et al (1996): A
genome-wide search for chromosomal loci linked to bipolar affective
disorder in the Old Order Amish. Nat Genet 12:431–435.
Gordon D, Heath SC, Liu X, Ott J (2001): A transmission/disequilibrium test
that allows for genotyping errors in the analysis of single-nucleotide
of the chromosome 12q23-q24 region in two pedigrees with bipolar
disorder and Darier’s disease. Am J Psychiatry 162:35–42.
Hall D, Gogos JA, Karayiorgou M (2004): The contribution of three strong
candidate schizophrenia susceptibility genes in demographically dis-
Further genetic evidence for a panic disorder syndrome mapping to
Hashimoto K, Fukushima T, Shimizu E, Komatsu N, Watanabe H, Shinoda N,
et al (2003): Decreased serum levels of D-serine in patients with schizo-
Hattori E, Liu C, Badner JA, Bonner TI, Christian SL, Maheshwari M, et al
(2003): Polymorphisms at the G72/G30 gene locus, on 13q33, are asso-
ciated with bipolar disorder in two independent pedigree series. Am J
review of genetic association studies. Genet Med 4:45–61.
Javitt DC (2004): Glutamate as a therapeutic target in psychiatric disorders.
(2001): A genome survey indicates a possible susceptibility locus for
Korostishevsky M, Kaganovich M, Cholostoy A, Ashkenazi M, Ratner Y, Da-
Single nucleotide polymorphisms, haplotypes, and gene expression
analysis. Biol Psychiatry 56:169–176.
Lewis CM, Levinson DF, Wise LH, DeLisi LE, Straub RE, Hovatta I, et al (2003):
Genome scan meta-analysis of schizophrenia and bipolar disorder, part
Lin MW, Sham P, Hwu HG, Collier D, Murray R, Powell JF (1997): Suggestive
evidence for linkage of schizophrenia to markers on chromosome 13 in
Caucasian but not Oriental populations. Hum Genet 99:417–420.
Liu C, Badner JA, Christian SL, Guroff JJ, Detera-Wadleigh SD, Gershon ES
(2001): Fine mapping supports previous linkage evidence for a bipolar
disorder susceptibility locus on 13q32. Am J Med Genet 105:375–380.
putative bipolar disorder locus on 2p13-16 and other potential loci on
4q31, 7q34, 8q13, 9q31, 10q21–24, 13q32, 14q21 and 17q11–12. Mol
with schizophrenia in the Chinese population. Neurosci Lett 369:228–
Maziade M, Roy M-A, Rouillard É, Bissonnette L, Fournier J-P, Roy A, et al
(2001): A search for specific and common susceptibility loci for schizo-
Whole genome linkage scan of recurrent depressive disorder from the
depression network study. Hum Mol Genet 14:3337–3345.
SG, et al (2003a): Genome-wide and conditional analysis in bipolar dis-
order: Evidence for gene-gene interaction in the NIMH Genetics Initia-
tive bipolar pedigrees. Biol Psychiatr 54:1265–1273.
al (2003b): Genome-wide scan of bipolar disorder in 65 pedigrees: Sup-
(2005): Combined analysis from eleven linkage studies of bipolar disor-
der provides strong evidence of susceptibility loci on chromosomes 6q
points to a locus of major effect on chromosome 12q23-q24. Am J Med
Mothet J-P, Parent AT, Wolosker H, Brady RO Jr, Linden DJ, Ferris CD, et al
(2000): D-serine is an endogenous ligand for the glycine site of the
association to the G72/G30 locus in an independent sample of schizo-
phrenia families. Mol Psychiatry 10:431–433.
Neale BM, Sham PC (2004): The future of association studies: Gene-based
disorder and shared susceptibility with schizophrenia. Mol Psychiatry
S.D. Detera-Wadleigh and F.J. McMahon
BIOL PSYCHIATRY 2006;60:106–114 113
Suggestive linkage to chromosomal regions 13q31 and 22q12 in fami-
lies with psychotic bipolar disorder. Am J Psychiatry 160:680–686.
Schumacher J, Abou JR, Becker T, Klopp N, Franke P, Jacob C, et al (2005):
Investigation of the DAOA/G30 locus in panic disorder. Mol Psychiatry
Schumacher J, Jamra RA, Freudenberg J, Becker T, Ohlraun S, Otte AC, et al
(2004): Examination of G72 and D-amino-acid oxidase as genetic risk
factors for schizophrenia and bipolar affective disorder. Mol Psychiatry
JI Jr, et al (2003): Genome scan meta-analysis of schizophrenia and
bipolar disorder, part III: Bipolar disorder. Am J Hum Genet 73:49–62.
Shaw SH, Kelly M, Smith AB, Shields G, Hopkins PJ, Loftus J, et al (1998): A
genome-wide search for schizophrenia susceptibility genes. Am J Med
Shaw SH, Mroczkowski-Parker Z, Shekhtman T, Alexander M, Remick RA,
Sadovnick AD, et al (2003): Linkage of a bipolar disorder susceptibility
locus to human chromosome 13q32 in a new pedigree series. Mol Psy-
Shink E, Morissette J, Sherrington R, Barden N (2005): A genome-wide scan
points to a susceptibility locus for bipolar disorder on chromosome 12.
Spielman RS, McGinnis RE, Ewens WJ (1993): Transmission test for linkage
Stine OC, McMahon FJ, Chen L, Xu J, Meyers DA, MacKinnon DF, et al (1997):
Initial genome screen for bipolar disorder in the NIMH genetics initiative
to clozapine for the treatment of schizophrenia. Am J Psychiatry 156:
Tuominen HJ, Tiihonen J, Wahlbeck K (2005): Glutamatergic drugs for schizo-
Tusie-Luna MT (2005): Genes and type 2 diabetes mellitus. Arch Med Res
Wang X, He G, Gu N, Yang J, Tang J, Chen Q, et al (2004): Association of
Wijsman EM, Rosenthal EA, Hall D, Blundell ML, Sobin C, Heath SC, et al
(2003): Genome-wide scan in a large complex pedigree with predomi-
nantly male schizophrenics from the island of Kosrae: Evidence for link-
age to chromosome 2q. Mol Psychiatry 8:695–705, 643.
Wilcox MA, Faraone SV, Su J, Van Eerdewegh P, Tsuang MT (2002): Genome
scan of three quantitative traits in schizophrenia pedigrees. Biol Psychi-
synthesizing D-serine to regulate glutamate-N-methyl-D-aspartate neu-
Chinese population. Schizophr Res 73:257–261.
114 BIOL PSYCHIATRY 2006;60:106–114
S.D. Detera-Wadleigh and F.J. McMahon