Download Supplementary Information 1 (doc 76K)

1
SOM Results
The Nine Additional Regions of Association
In addition to the four regions of association described in the text, nine more regions (B-2, B3, D-1, D-3 to D-6, D-8 and D-9) were identified and delineated based on support from
significant (i.e. meeting the region-wide Nyholt-corrected thresholds) individual, but not
global, haplotypes (region B: Pg>0.0005, Pi≤0.0005; region D: Pg>0.0003, Pi≤0.0003). The
computational demands of permutation analysis impeded multiple-testing correction for
these haplotypes and, thus, a more accurate assessment of their significance. These
regions and the genes they contain are described and discussed below.
Region B
Cluster B-2: A single five-SNP haplotype (SNP 75 to 79) was associated in the full sample of
SCZ individuals (Pi=0.00046) (SOM Table S3), and it conferred an increase in risk to SCZ
(OR=2.2, 95%CI: 1.3-3.6). Following the individual sliding-window haplotypes in both
directions, while Pi≤0.01, we identified a 53kb region of association. No Known/RefSeq or
predicted genes overlapped with B-2. The nearest gene, heparan sulfate (glucosamine) 3-Osulfotransferase 1 (HS3ST1), is located 77kb centromeric of B-2 (SOM Table S4).
Cluster B-3: A rare 3-SNP haplotype (frequency: cases=0.04, controls=0.00), spanning
SNPs 123 to 125, was associated in BP males (Pi=0.00020) (SOM Table S3). Following the
individual haplotypes in both directions, while Pi≤0.01, revealed a 221kb region of
association. No Known/RefSeq genes were found in the region, but one predicted gene was
identified. The HS3ST1 gene is the nearest gene, located 577kb telomeric of B-3 (SOM
Table S4).
Region D
2
Cluster D-1: Two nested haplotypes were associated in all affected females, with the more
significant haplotype being the 4-SNP haplotype covering SNPs 14 to 18 (Pi=0.00011) (SOM
Table S5). Both haplotypes confer a protective effect against BP and SCZ to females
(OR=0.4, 95% CI: 0.3-0.6). Following the haplotypes in both directions identified a 94kb
region of association, in which the G protein-coupled receptor 125 (GPR125) and a
predicted gene are located (SOM Table S4).
Cluster D-3: A single 5-SNP haplotype (SNP 59 to 63) was associated in all affected males
(Pi=0.00023; OR=3.2, 95% CI: 1.8-5.9) (SOM Table S5). Upon extending the individual
haplotype in both directions, while Pi≤0.01, an 83kb region of association was identified. This
region contains the peroxisome proliferator activated receptor gamma coactivator 1 alpha
(PPARGC1A) gene, as well as five predicted transcripts (SOM Table S4).
Cluster D-4: A 5-SNP haplotype (SNP 148 to 152) was associated in all affected males
(Pi=0.00030; OR=0.6, 95% CI: 0.5-0.8) (SOM Table S5). Following the individual haplotype
in the 5’ and 3’ directions, we identified a 180kb region of association, in which the
hypothetical DKFZp761B107 gene is located. Two predicted genes also maps to D-4.
Cluster D-5: A rare 5-SNP haplotype (frequency: cases=0.03, controls=0.00), spanning
SNPs 182 to 186, was associated in the full sample of BP individuals (Pi=0.00014) (SOM
Table S5). The region of association was found to span 34kb, which only contains one
predicted gene. The nearest gene is the solute carrier family 34 (sodium phosphate),
member 2 (SLC34A2) gene, which is located 62kb centromeric of D-5.
Cluster D-6: A rare 5-SNP haplotype (frequency: cases=0.03, controls=0.00) in the full
sample of SCZ individuals (Pi=0.00024) (SOM Table S5), and it spans SNPs 202 to 206.
The delineated region of association extends across 185kb and overlaps with the
hypothetical KIAA0746 gene and seven predicted genes (SOM Table S4).
3
Cluster D-8: A single 5-SNP haplotype was associated in BP males (Pi=0.00016) and
confers an increase in risk (OR=3.2, 95%CI: 1.7-5.9) (SOM Table S5). Following the
haplotype in both directions, while Pi≤0.01, revealed an 87kb region of association, in which
no Known/RefSeq or predicted genes were located. The closest gene is recombining binding
protein suppressor of hairless (Drosophila) (RBPSUH), which falls 48kb centromeric of D-8
(SOM Table S4).
Cluster D-9: Two nested, rare haplotypes (frequency of both: cases=0.05, controls=0.00)
were associated in SCZ females. The more significant of these was the 4-SNP haplotype,
spanning from SNP 250 to 253 (Pi=0.00024) (SOM Table S5). Following the haplotypes in
both directions identified a 109kb region of association that overlapped with the
cholecystokinin A receptor (CCKAR) gene (SOM Table S4).
4
SOM Discussion
Genes near Associated Regions B-1, B-4, D-2 and D-7
The known genes KIAA0746, PPARGC1A, MIST and HS3ST1 map 17, 25, 239 and 577kb,
respectively, from a main region of association, suggesting the possibility that the functional
variant is disrupting a regulatory element located in non-coding sequence. Their candidacy
as susceptibility genes for psychiatric illness is discussed below.
The brain-expressed gene KIAA0746 encodes the hypothetical protein LOC23231. It is
conserved in the mouse (87.2% identity) and contains Sel-1-like repeats (IPR006597).
These are tetratricopeptide repeat sequences originally identified in a C. elegans receptor
molecule - a key negative regulator of the Notch pathway.1 Associated region D-7 is located
17kb upstream of the transcription start site.
The peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PPARGC1A)
gene encodes a transcriptional coactivator that regulates genes involved in energy
metabolism, mitochondrial biogenesis, and regulation of muscle fiber type determination.
PPARGC1A may also be involved in controlling blood pressure, regulating cellular
cholesterol homoeostasis and the development of obesity. In rat liver, the expression of
PPARGC1A has been shown to be significantly affected by antipsychotic treatment (Johan
Ferno and Vidar Steen, personal communication), suggesting a possible role for
PPARGC1A in psychiatric illness. Associated region D-2 is located 25kb downstream of the
coding region.
The mast cell immunoreceptor signal transducer (MIST) gene is thought to be an important
part of the high-affinity IgE receptor (type I, Fc RI) intracellular signalling cascade, which is
involved in the development of an array of acute and chronic allergic reactions.2 Associated
region B-1 lies 239kb upstream of the transcription start site.
5
The heparan sulfate (glucosamine) 3-O-sulfotransferase 1 (HS3ST1) gene is highly
expressed in the brain and kidney and weakly expressed in the heart, lung and placenta. It
possesses both heparan sulfate glucosaminyl 3-O-sulfotransferase activity and
anticoagulant heparan sulfate conversion activity, is a rate limiting enzyme for synthesis of
anticoagulant heparin and is an intraluminal Golgi resident protein. However, Hs3st1deficient mice did not show the expected anticoagulant phenotype. Instead, they developed
unexpected non-thrombotic, perinatal phenotypes, including eye degeneration and postnatal
lethality.3 This suggests that the HS3ST1 enzyme might have additional or alternative
biological roles. Associated region B-4 is positioned 577kb upstream of the coding region.
Additional Genes
In addition to the four regions of association described in the text, nine more regions were
identified and delineated based on support from significant individual, but not global,
haplotypes (region B: Pi≤0.0005, SOM Tables S3 and S4a; region D: Pi≤0.0003, SOM
Tables S5 and S4b). These nine additional regions of association contained five known
genes, in addition to several predicted genes (SOM Table S4). The known genes are
GPR125 (D-1), PPARGC1A (D-3, described in manuscript) and CCKAR (D-9) and
hypothetical genes DKFZp761B107 (D-4) and KIAA0746 (D-6, described in manuscript).
The G protein-coupled receptor 125 (GPR125) gene encodes a member of a family of
integral membrane proteins that regulate gene transcription associated with cell proliferation,
differentiation and dynamic cellular processes. The processes include transduction of
extracellular signals from neurotransmitters, transactivation of receptor tyrosine kinases and
activation of mitogen activated protein kinase cascades. The signalling mechanisms can
modulate synaptic activity in the central nervous system. Therefore, GPR activity may be
important to maintain the brain’s homeostatic response to neurotransmitter signalling and
thus variation in this gene family may contribute to psychiatric illness.4 GPR125 is expressed
in a variety of tissues, including brain.5
6
The cholecystokinin A receptor (CCKAR) gene encodes a G protein-coupled receptor that
binds non-sulfated members of the cholecystokinin (CCK) family of peptide hormones. This
receptor is a major physiologic mediator of pancreatic enzyme secretion and smooth muscle
contraction of the gallbladder and stomach. In the central and peripheral nervous system, it
regulates satiety and the release of beta-endorphin and dopamine. Polymorphisms in
CCKAR have been associated with several psychiatric and neurologic disorders in the
Japanese population (schizophrenia,6 Parkinson,7 alcohol dependence8 and panic
disorder9).
The gene DKFZp761B107 encodes the hypothetical protein LOC91050. It shares 94%
identity with the mouse hypothetical protein XP_144338 and 99% identity with the Bos
taurus (cattle) predicted protein XP_868963. This protein contains a smc conserved domain
(COG1196.1) identified from the NCBI Conserved Domain Database.10 As a chromosome
segregation ATPase, this domain is potentially involved in cell division and chromosome
partitioning.
7
SOM References
1.
Grant B, Greenwald I. The Caenorhabditis elegans sel-1 gene, a negative
regulator of lin-12 and glp-1, encodes a predicted extracellular protein. Genetics 1996
May; 143(1): 237-247.
2.
Fujii Y, Wakahara S, Nakao T, Hara T, Ohtake H, Komurasaki T et al. Targeting
of MIST to Src-family kinases via SKAP55-SLAP-130 adaptor complex in mast cells.
FEBS Lett 2003 Apr 10; 540(1-3): 111-116.
3.
Shworak NW, HajMohammadi S, de Agostini AI, Rosenberg RD. Mice deficient
in heparan sulfate 3-O-sulfotransferase-1: normal hemostasis with unexpected
perinatal phenotypes. Glycoconj J 2002 May-Jun; 19(4-5): 355-361.
4.
Ferguson SS. Receptor tyrosine kinase transactivation: fine-tuning synaptic
transmission. Trends Neurosci 2003 Mar; 26(3): 119-122.
5.
Fredriksson R, Gloriam DE, Hoglund PJ, Lagerstrom MC, Schioth HB. There
exist at least 30 human G-protein-coupled receptors with long Ser/Thr-rich N-termini.
Biochem Biophys Res Commun 2003 Feb 14; 301(3): 725-734.
6.
Lu WT, Zhang X, Zhang M, Gong SL, Wei J. Association analysis of the
cholecystokinin type A receptor gene in schizophrenia. Chin Med J (Engl) 2004 Apr;
117(4): 627-629.
7.
Wang J, Si YM, Liu ZL, Yu L. Cholecystokinin, cholecystokinin-A receptor and
cholecystokinin-B receptor gene polymorphisms in Parkinson's disease.
Pharmacogenetics 2003 Jun; 13(6): 365-369.
8
8.
Miyasaka K, Yoshida Y, Matsushita S, Higuchi S, Maruyama K, Niino N et al.
Association of cholecystokinin-A receptor gene polymorphism with alcohol
dependence in a Japanese population. Alcohol Alcohol 2004 Jan-Feb; 39(1): 25-28.
9.
Ise K, Akiyoshi J, Horinouchi Y, Tsutsumi T, Isogawa K, Nagayama H.
Association between the CCK-A receptor gene and panic disorder. Am J Med Genet B
Neuropsychiatr Genet 2003 Apr 1; 118(1): 29-31.
10.
Marchler-Bauer A, Anderson JB, Cherukuri PF, DeWeese-Scott C, Geer LY,
Gwadz M et al. CDD: a Conserved Domain Database for protein classification. Nucleic
Acids Res 2005 Jan 1; 33(Database issue): D192-196.