Association of the putative susceptibility gene transient receptor potential protein melastatin type 2 with bipolar disorder.код для вставкиСкачать
American Journal of Medical Genetics Part B (Neuropsychiatric Genetics) 141B:36 – 43 (2006) Association of the Putative Susceptibility Gene, Transient Receptor Potential Protein Melastatin Type 2, With Bipolar Disorder Chun Xu,1 Fabio Macciardi,2,5,6 Peter P. Li,1,4,5 Il-Sang Yoon,7 Robert G. Cooke,3,5 Bronwen Hughes,1 Sagar V. Parikh,3,5 Roger S. McIntyre,3,5 James L. Kennedy,2,5,6 and Jerry J. Warsh1,3,4,5,6* 1 Laboratory of Cellular and Molecular Pathophysiology, Centre for Addiction and Mental Health, University of Toronto, Toronto, Ontario, Canada 2 Laboratory of Neurogenetics, Centre for Addiction and Mental Health, University of Toronto, Toronto, Ontario, Canada 3 Mood Disorders Program, Centre for Addiction and Mental Health, University of Toronto, Toronto, Ontario, Canada 4 Department of Pharmacology, University of Toronto, Toronto, Ontario, Canada 5 Department of Psychiatry, University of Toronto, Toronto, Ontario, Canada 6 Institute of Medical Science, University of Toronto, Toronto, Ontario, Canada 7 Department of Neurosciences, University of California, San Diego, California Disturbed intracellular calcium (Ca2þ) homeostasis has been implicated in bipolar disorder (BD). Reduced mRNA levels of the transient receptor potential Ca2þ permeable channel melastatin type 2, TRPM2, in B lymphoblast cell lines (BLCL) from bipolar I disorder (BD-I) patients showing elevated basal intracellular Ca2þ ([Ca2þ]B), an index of altered intracellular Ca2þ homeostasis, along with its location within a putative BD susceptibility locus (21q22.3), implicates the involvement of this gene in the Ca2þ abnormalities and the genetic diathesis to BD. We tested this hypothesis by examining the association of selected single nucleotide polymorphisms (SNPs) and their haplotypes, spanning the TRPM2 gene, with BD and BLCL [Ca2þ]B, in a case control design. The 50 TaqMan SNP assay was used to detect selected SNPs. BLCL [Ca2þ]B was determined by ratiometric fluorometry. SNP rs1618355 in intron 18 was significantly associated with BD as a whole (P < 7.0 105; odds ratio (OR) ¼ 2.60), and when stratified into BD-I (P < 7.0 105, OR ¼ 2.48) and BD-II (P ¼ 7.0 105, OR ¼ 2.88) subgroups. In addition, the alleles of the individual SNPs forming a seven marker at-risk haplotype were in excess in BD (12.0% in BD vs. 0.9% in controls; P ¼ 2.3 1012). A weak relationship was also detected between BLCL [Ca2þ]B and TRPM2 SNP rs1612472 in intron 19. These findings suggest genetic variants of the TRPM2 gene increase risk for BD and support the notion that TRPM2 may be involved in the pathophysiology of BD. ß 2005 Wiley-Liss, Inc. Grant sponsor: Canadian Institutes of Health Research; Grant number: MOP 12851; Grant sponsor: Ontario Mental Health Foundation. *Correspondence to: Jerry J. Warsh, M.D., Ph.D., Head, Laboratory of Cellular and Molecular Pathophysiology, Centre for Addiction and Mental Health, 250 College Street, Room R20, Toronto, ON, Canada, M5T 1R8. E-mail: firstname.lastname@example.org Received 24 May 2005; Accepted 22 July 2005 DOI 10.1002/ajmg.b.30239 ß 2005 Wiley-Liss, Inc. KEY WORDS: intracellular calcium homeostasis; calcium permeable channel; single nucleotide polymorphisms (SNPs); haplotype INTRODUCTION Genome-wide linkage scan and candidate gene studies (for review see [Sklar, 2002]), together with meta-analyses [Segurado et al., 2003], have provided important clues as to chromosomal regions that may harbor candidate genes, which contribute to the vulnerability to bipolar disorder (BD), inconsistencies, design and methodological limitations, notwithstanding. Among those genes implicated in genome scanning studies, the transient receptor potential protein of the melastatin type (TRPM2), a calcium (Ca2þ) permeable ion channel, is an attractive candidate in regards to some of the pathophysiological disturbances implicated in BD. Straub et al.  first described linkage to BD at 21q21, obtaining a LOD score of 3.41 with D21S171, a marker only 29.6 kb from TRPM2, in an extended BD pedigree. Confirmation was subsequently reported by a number of independent studies, supporting that the 21q21-22 region harbors BD susceptible loci [Straub et al., 1994; Detera-Wadleigh et al., 1997; Smyth et al., 1997; Aita et al., 1999; Curtis, 1999; Kwok et al., 1999; Berrettini, 2000; Kelsoe et al., 2001; Liu et al., 2001; Ewald et al., 2003]. Building evidence on the physiological functions of TRPM2 and disturbances in its expression in BD also support its potential involvement in the pathophysiology of BD. TRPM2 has been shown to regulate Ca2þ entry in response to several novel intracellular messengers including adenine diphosphate-ribose, nicotinamide adenine dinucleotide, and reactive oxygen species, such as H2O2 [Perraud et al., 2004]. TRPM2 is highly expressed in human brain including cerebral cortex, hippocampus, amygdala, caudate nucleus, putamen [Nagamine et al., 1998], B-lymphoblast cell lines (BLCLs) [Yoon et al., 2001], and other tissues (reviewed in [Perraud et al., 2004]). Of particular interest, TRPM2 expression was found significantly lower in BLCLs from BD-I patients with high basal intracellular Ca2þ concentration ([Ca2þ]B) compared with BD-I patients showing normal [Ca2þ]B, with BD-II patients, with major depressive disorder patients and healthy subjects [Yoon et al., 2001]. Sustained elevations of intracellular Ca2þ levels reduce TRPM2 mRNA levels in BLCLs [Yoon, 2002] and TRPM2 isoforms promote (long form) or suppress Association of TRPM2 With Bipolar Disorder (short form) susceptibility to cell death [Zhang et al., 2003]. These observations further implicate altered expression, structure, and/or function of this protein in BD. Such evidence of positional and biological candidacy prompted us to scrutinize more closely the potential role of TRPM2 variants in modifying susceptibility to BD and in the occurrence of altered intracellular Ca2þ homeostasis also observed in this disorder [Warsh et al., 2004]. Towards this end, we examined the relationship between DSM-IV diagnostic categories of BD-I and II disorder, and allelic and haplotype frequencies in the TRPM2 gene. Seven single nucleotide polymorphisms (SNPs) located in the promoter and selected intronic regions (8, 16, 18, 19, 20, and 27), the positions of which are representative of the overall SNP distribution in the TRPM2 gene, were studied. Association with elevated [Ca2þ]B, an index of altered intracellular Ca2þ homeostasis, which may define a pathophysiologically more homogeneous subgroup of BD (reviewed in [Warsh et al., 2004]), was also examined. We report here evidence in a case-control study of 446 Caucasian subjects of an allelic and haplotype association within the TRPM2 gene and BD. MATERIALS AND METHODS Sample Collection Patients and healthy subjects were recruited from the greater Toronto area as previously described [Emamghoreishi et al., 1997]. Because potential participants were assessed in the context of studies of intracellular Ca2þ homeostasis abnormalities in mood disorders, subjects were excluded from the study if they had a history of recent (<3 months) substance abuse or current physical illness. Psychiatric diagnoses were confirmed using the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I), patient edition [First et al., 1995a], administered by a research psychiatrist or trained psychiatric research assistant, complemented by review of available medical records. Healthy subjects had no personal or family (first degree relatives) history of psychiatric disorder, based on structured interview with the SCID-I non-patient version [First et al., 1995b], and no abnormalities on systems review, or where indicated, physical examination. Demographic information including ethnicity, age of onset of primary diagnosis, duration of illness, number of episodes, and number of hospitalizations were obtained during the interview of the probands, facilitated by life charting and review of available medical records. Ethnicity was determined at interview by systematic questioning of the proband in regards to ancestral ethnic and geographic origins for the prior 1–3 generations, to the extent of the participant’s knowledge of family origins. Ethnicity was classified based on ancestral region of origin as identified within current country boundaries. Controls were matched to cases on the basis of age, gender, and ethnicity. The study was approved by the Research Ethics Board of the Centre for Addiction and Mental Health, and after a complete description of the study, all subjects provided written informed consent. DNA Analysis Blood was drawn from each subject into anticoagulant (ACD) containing tubes and the leukocyte fraction transformed with Epstein–Barr virus as previously described [Emamghoreishi et al., 1997]. BLCLs were frozen at approximately 15– 18 passages and stored over liquid nitrogen in aliquots of approximately 5 106 cells. For genotyping, BLCL aliquots were rapidly thawed and washed in phosphate buffered saline (PBS). After re-suspending in 200 ml PBS, genomic DNA was extracted using a QIAamp1 Blood Kit (Qiagen, Mississauga, Ontario, Canada) according to the manufacturer’s instruc- 37 tions. In a subset of the healthy controls recruited in parallel for other psychiatric genetic studies in this center, DNA was extracted from EDTA anticoagulated blood samples using a high salt method [Lahiri and Nurnberger, 1991]. SNP Selection A total of seven SNPs were chosen to explore the association between the TRPM2 gene and BD in this preliminary study. These were selected using the Assay-On-Demand database (www.allsnps.com, Applied Biosystems, Inc., Foster City, CA), which provides the physical location and allele frequency data in four ethnic populations, the NCBI dbSNP database (www.ncbi.nlm.nih.gov/SNP), and SNPper (http://snpper. chip.org/). Criteria for choice of SNPs used were: (1) minor allele frequencies >19% in the Caucasian population; and (2) location within the promoter region, and exonic and intronic sites that could potentially impact on TRPM2 expression and function. The NetGen2 server service for predicting splice-sites (http://www.cbs.dtu.dk/services/NetGene2/ [Hebsgaard et al., 1996] was used to scrutinize the TRPM2 genomic sequence for SNPs located at or proximal to splice-sites. SNP Analysis and Allele Determination Genotyping was performed using the 50 nuclease allelic discrimination TaqMan assay in a 96-well format. To each well, 5 ml of PCR master mix (TaqMan1 Universal PCR Master Mix), 5–10 ng of DNA template, 0.5 ml of 20 Assay-On-Demand SNP Genotyping Assay Mix (sequence-specific primers and probes) were added. The amplification reaction was carried out on an ABI PRISM1 7300 Sequence Detection System (Applied Biosystems, Inc.) using the following cycle parameters: initial denaturation 958C for 10 min, followed by 40 cycles of denaturing at 928C for 15 sec, and annealing and primer extension at 608C for 1 min. After amplification, end point plate reading to discriminate alleles was performed using the proprietary Sequence Detection System software (Applied Biosystems, Inc.). In addition to positive control samples with known genotyping information, four negative control samples without genomic DNA, as well as randomly selected samples from six individuals, were assayed in duplicate on each 96-well plate. SNP genotyping was performed blind to subject diagnosis, characteristics, and sample replication. Basal Intracellular Calcium Determination The [Ca2þ]B was determined in BLCLs at 13–17 passages using fura2-AM and ratiometric fluorometry, as previously described [Emamghoreishi et al., 1997]. Statistical Analysis Dependent measures including [Ca2þ]B, age, and age of onset of illness are expressed as means SD. Hardy–Weinberg Equilibrium (HWE) analysis, linkage disequilibrium (LD), and haplotype block structure were examined using Haploview, version 3.2 [Barrett et al., 2005] (http:// www.broad.mit.edu/mpg/haploview/) based on the normal control subjects. The D0 for all pairs of SNPs was calculated and the haplotype blocks were estimated using the solid spine of LD method [Barrett et al., 2005]. Association between the haplotypes and diagnostic groups was tested using COCAPHASE within the UNPHASED suite of programs (http:// www.hgmp.mrc.ac.uk/fdudbrid/software/unphased/) [Cordell and Clayton, 2002], as well as PowerMarker3 (version 3.09; http://statgen.ncsu.edu/powermarker/) [Zaykin et al., 2002], which obviates the effects of multiple testing. Disease association of individual SNP allele, genotype or haplotype frequencies with BD patients, as a group, compared 38 Xu et al. with healthy individuals, was assessed using 2 2 or 2 3 contingency tables, two-tailed w2 tests, or Fisher exact tests. Differences in dependent measures ([Ca2þ]B, age, age of onset) were analyzed by univariate ANOVA with genotype or haplotype, and diagnosis as factors, followed by post hoc comparisons of group means using the Tukey’s test. Statistical analyses were performed using the SPSS statistical software package (version 10, SPSS, Inc., Chicago, IL.). Differences with two-tailed probability values of P 0.05 were taken as statistically significant. P-values for tests of allelic and genotype association were conservatively corrected for multiple testing using the Bonferroni method. Assessment of Potential Population Stratification Effects The statistical genetics programs FSTAT (http://www. unil.ch/izea/softwares/fstat.html) and GENEPOP, version 3.3 [Raymond and Rousset, 1995] were used to detect potential genetic substructure using unlinked SNP markers located in different human chromosome regions in this study material. In these analyses, the number of ethnic subgroups in the sample was set for four Caucasian subpopulations (Table IB). The default parameter settings were used for the GENEPOP analyses. Assessment of Statistical Power The adequacy of statistical power was evaluated using two different methods: the first employed binomial power calculations (http://calculators.stat.ucla.edu/powercalc/) to determine the power of the w2 tests. The sample size of 178 BD patients and 268 controls used in this study gave a power of 88% with relative risk of 2.0 (two-sided test). The second method was performed with the genetic power calculator (statgen.iop.kcl.ac.uk http://statgen.iop.kcl.ac.uk/gpc/cc2.html) using a variance components analysis with the following parameters: disease prevalence, 0.01; D0 between disease and SNP alleles, 0.8; alpha, 0.05. The power to detect association was estimated as 98%, based on the total sample of 446 subjects with disease and marker allele frequency (A) ¼ 0.30 and 94%, with A ¼ 0.20, since these frequencies are the minor allele frequencies of the tested SNPs/markers. RESULTS Demographic Characteristics Selected demographic characteristics of the subjects studied are detailed in Table IA and Table IB. There was no difference in age (F ¼ 2.18, df ¼ 2,442, P ¼ 0.114) or gender (w2 ¼ 2.21, df ¼ 2, P ¼ 0.33) among comparison groups. Age of onset was significantly lower in BD-II (16.5 6.0 years) compared with BD-I patients (20.4 8.2 years; t ¼ 3.05, df ¼ 171, P ¼ 0.003). TABLE IA. Demographic Characteristics of Bipolar and Healthy Subjects in Caucasian Population: Sex, Age, and Age of Onset Diagnostic group Control BD-I BD-II a N 268 124 54 Sex (F/M) 159/109 82/42 36/18 Age Age of onset a 38.7 9.1 37.9 10.9 35.5 8.9 NA 20.4 8.2 16.5 6.0b Age and age of onset are expressed as the mean SD. Significantly earlier in BD-II compared with BD-I patients (t ¼ 3.05 df ¼ 171, P ¼ 0.003). b LD, SNP Allele, Haplotype Frequency, Haplotype Blocks Figure 1 shows the physical location of the SNP in the putative promoter region and the six SNPs, which span 87,074 bp and approximately 63% of the TRPM2. The use of SNPs with high minor allele frequencies (mean 0.27) enhances the power to detect LD [Ohashi and Tokunaga, 2002]; the observed frequencies in present study concur with those values reported in the Assay-On-Demand SNP database (www.allsnps.com) (Table II). The genotype frequencies of each of the seven SNPs examined did not deviate significantly from HWE in healthy controls (Table II). As shown in Figure 1, two potentially associated haplotype blocks were identified across the 87-kb region and designated as blocks 1 and 2, respectively (see Fig. 1), based on D0 and r2 in the healthy control population. Moderate LD was evident between promoter and intron 8 SNPs (D0 ¼ 0.88), spanning a 31.4 kb region, and strong LD among four of the SNPs (intron 16, 18, 19, 20), spanning a 14.2 kb region. Of these seven SNPs, strong pairwise LD was observed between intron 16 and 18 SNPs (D0 ¼ 0.96) and between intron 19 and 20 SNPs (D0 ¼ 0.98) (Fig. 1), suggesting that there has been little historic recombination in this region over time. The values of D0 between the SNPs in introns 8 and 16, at a 22.6 kb distance, and between SNPs in introns 20 and 27, at a distance of 18.8 kb, were much lower (D0 ¼ 0.33 and 0.28, respectively). Thus, these SNPs are not in complete LD. Finally, blocks 1 and 2 showed some degree of multivariate D0 , suggesting that a putative recombination event between these blocks should be old, if any. Association Between SNPs and Haplotypes, and Diagnostic Groups and Phenotypes Statistically significant increased allele 2 (minor allele) frequencies were observed for intron 18 (P < 7.0 105) and 19 (P ¼ 0.00022) SNPs in BD as compared with those in healthy controls (44.7 vs. 23.7%, OR ¼ 2.60; 38.2% vs. 25.1%, OR ¼ 1.85, respectively, Table III). Moreover, the allele 2 frequencies of intron 18 and intron 19 SNPs were also significantly increased in both BD-I and BD-II patient groups as compared with controls (Table III). Since the use of haplotypes may provide greater statistical power in association studies of common genetic variation than can be obtained with individual SNPs [Gabriel et al., 2002], a seven-locus-haplotype analysis was performed. A number of susceptible haplotypes were detected: the T-T-T-C-T-T-A haplotype is one among these showing greatest statistical significance (P ¼ 2.3 1012) with a frequency of 12.0% in the BD cohort and 0.9% in controls (Table IV). To pinpoint the susceptibility or protective area more precisely, we also analyzed the block-based haplotypes from adjacent SNPs in a sliding-window fashion. Statistically significant associations were observed between most of the haplotypes and BD (Pvalues range from 9.5 1010 to 2.5 1014) compared with controls (data not shown). The best result was obtained for the 4-locus-haplotype analysis in block 2 with an omnibus test showing a P ¼ 2.3 1015, with the largest difference between cases and controls (1.3% in control vs. 15.5% in BD) observed for the T-C-T-T haplotype. Genotype and allele distribution did not differ significantly for the seven TRPM2 SNPs between females and males in both control and BD groups (data not shown). Tests of Population Stratification No significant differences were found for five unlinked SNPs in different genes among Caucasian subpopulations from the four ethnically defined groups (FSTAT: Fst ¼ 0.002) and Association of TRPM2 With Bipolar Disorder 39 TABLE IB. Demographic Characteristics of Bipolar and Healthy Subjects in Caucasian Population: Case and Control Ethnicity Caucasian sub-groups British Islesa European–Otherb Recent admixture of Europeansc Uncertain European ancestryd BD (as a Group) BD-I BD-II Control Total 51 (28.7%) 26 (14.6%) 67 (37.6%) 34 (19.1%) 36 (29.0%) 23 (18.5%) 38 (30.6%) 27 (21.8%) 15 (27.8%) 3 (5.6%) 29 (53.7%) 7 (13.0%) 69 (25.7%) 63 (23.5%) 69 (25.7%) 67 (25.0%) 120 89 136 101 a ‘‘British Isles’’ includes: British, English, Irish (North and South), Welsh, Scottish or admixtures within these groups. ‘‘European–Other’’ includes: non-admixed North European (4BD, 1HC), West European (10BD, 11HC), East European (5BD, 21HC), Russian (0BD, 3HC), Mediterranean (6BD, 27HC), and French Canadian (1BD, 0HC). c ‘‘Recent admixture of Europeans’’ includes: admixture with each parent of different regional/ethnic European origin. d ‘‘Uncertain European ancestry’’ includes: probands who lack knowledge of one or the other of the maternal and/or paternal grandparent. b (GENEPOP: Fst ¼ 0.002). These results suggest the four categorized Caucasian populations used in the present study show relative homogeneity, at least in the context of the genetic analyses of TRPM2 employed here. BLCL Basal Calcium Concentrations Mean BLCL [Ca2þ]B values differed significantly among diagnostic groups (F ¼ 4.26, df ¼ 2,296, P ¼ 0.015), due to significantly higher [Ca2þ]B in BD-I patients (61.7 9.1 nM, N ¼ 160; P ¼ 0.023, post hoc Tukey test), but not in BD-II patients (59.4 7.5, N ¼ 73) compared with that in the healthy subjects (58.4 7.5, N ¼ 64). Moreover, no statistically significant differences were evident in BLCL [Ca2þ]B values Fig. 1. Genomic structure with seven SNP positions and haplotype blocks identified in the human TRPM2 gene as described by Nagamine et al. (1998). Top: Depicts the distance in kb between selected SNPs. Middle: The 32 exons coding for TRPM2 are represented by filled squares. Circles represent SNP locations. Exon and intron sizes are not drawn to exact scale. Bottom: The locations of seven SNPs and LD structure. Two blocks have among different ethnic groups (F ¼ 1.002, df ¼ 3,185, P ¼ 0.393) (data not shown). In addition, [Ca2þ]B was significantly higher (F ¼ 4.64, df ¼ 2,115, P ¼ 0.012) in BD-I patients with the TRPM2-intron 19 SNP T/T genotype (62.6 9.8 nM) compared to those with T/C (61.2 8.5 nM; P ¼ 0.008) and with C/C (54.4 6.5 nM; P ¼ 0.035) genotypes. However, no association was observed between basal [Ca2þ]B and the other six TRPM2 SNPs or haplotypes studied here in Caucasian BD patients as a whole, or in BD-I or BD-II subgroups. DISCUSSION The results of the SNP allelic and haplotype analysis obtained in this study provide strong support for the candidacy been identified. Block 1 consists of promoter and intron 8 SNPs; Block 2 covers the region between intron 16 and 20 SNPs. The regions of high LD (D0 > 95) are shown in dark gray. Regions with lower LD are shown in light gray with the intensity decreasing with decreased D0 value. Blocks defined by solid spine of LD based on normal controls. SNPs are ordered centromeric to telomeric on chromosome 21q. Major/ minor allele. The nucleotide position of each SNP is relative to the start of exon 1 based on the NCBI location. HWE analysis was performed as described in the Statistical Methods. d Frequencies for each SNP are based on the entire sample of current study and compared with published data shown in italics. http://myscience.appliedbiosystems.com/index.jsp). e Abbreviations: CC, Caucasian; AA, African American; Ch, Chinese; NA, not available. Allelic and genotype frequencies for seven SNPs did not differ significantly between Caucasian and Asian populations. c b 0.21 0.25 0.26 0.32 0.33 0.26 0.21 0.62 0.59 0.19 0.70 0.35 0.18 0.95 5851 T/C 25580 T/C 48210 T/C 52878 A/C 59401 T/C 62421 T/C 81223 A/G 4889 bp upstream of E1 284 bp downstream of E8 14 bp downstream of E16 14 bp upstream of E19 788 bp upstream of E20 1806 bp upstream of E21 206 bp upstream of E28 Promoter Intron-8 Intron-16 Intron-18 Intron-19 Intron-20 Intron-27 rs734336 rs2838556 rs1785437 rs1618355 rs1612472 rs933151 rs749909 a NA 0.09 NA NA NA NA NA NA 0.15 NA NA NA NA NA 0.45 0.36 0.35 0.42 0.46 0.28 0.11 0.20 0.22 0.22 0.23 0.20 0.23 0.19 0.25 0.26 0.24 0.30 0.45 0.22 0.13 0.22 0.25 0.26 0.33 0.32 0.26 0.22 Ch AA Entire sample HWEc P-value Locationa SNP ID Distance (bp) from exon Nucleotide changeb Current study d CCe Asian Frequencies CC Published Japanese Xu et al. TABLE II. Location and Frequencies of SNPs in TRPM2 40 of TRPM2 as a potential vulnerability gene that contributes to the risk for BD. In addition, a statistically significant relationship between a genetic variant (intron 19 SNP) of TRPM2 gene and elevated BLCL [Ca2þ]B in BD-I patients also provides some support for a link between variation in TRPM2 and disturbances of intracellular Ca2þ homeostasis in BD. Increased frequencies were found for several of the TRPM2 SNP alleles examined in the BD population compared with those in healthy controls, including allele 2 of the intron 18 SNP and of the intron 19 SNP. The relatively high OR of 2.60, obtained for the intron 18 SNPs also suggests an important biological effect increasing the risk of developing BD, in the context of a polygenic complex disorder. Haplotype block analysis has been used to successfully map mutations associated with complex diseases [Tishkoff and Verrelli, 2003] as it affords the advantage of being able to test associations between genomic segments encoding multiple genes and traits [Gabriel et al., 2002], in contrast to individual polymorphisms. The two haplotype blocks, identified in the present study, spanned 31.4 kb between promoter and intron 8 SNPs, and 14.2 kb between intron 16 and 20 SNPs, respectively (Fig. 1), the latter region containing the intron 18 putative BD-risk SNP rs1618355. This agrees with the suggested average block length of 22 kb and range in length from <1 to 175 kb as reported by Dawson et al. . Strong LD was observed in the region between intron 16 and intron 20, which corresponds with the haplotype block 2 region in which two significantly associated SNPs (intron 18 and 19 SNPs) were located. The block 2 identified here has similar properties of the block generated by the Applied Biosystems SNPbrowserTM software, version 1.0, which provides a visual representation of putative haplotype block boundaries based on Caucasian, African American, Chinese, and Japanese ethnic populations. Of note, the statistically significant association observed between the risk haplotypes of the TRPM2 gene and BD (Table IV) further strengthens the notion that certain TRPM2 variants confer susceptibility to BD. The statistically significant higher frequency of the T-T-T-C-T-T-A haplotype (BD: 12.0% vs. controls: 0.9%) suggests an important role of this haplotype in conferring susceptibility to BD. In addition, some haplotypes were significantly less common in BD, which may reflect their effect as a putative ‘‘protective’’ factor; for example, the frequency of the T-T-T-A-T-T-A haplotype was significantly more common in control (52%) compared with BD (26%) (P ¼ 5.6 1015, Table IV). However, the nature of ‘‘susceptibility’’ and ‘‘protective’’ haplotypes in the TRPM2 locus needs to be explored further in regard to this gene’s role as a candidate in the pathogenesis of BD. Lack of allele frequency information and the low allele frequency of the exonic SNPs from 390 SNPs reported in the dbSNP database ruled out their use, therefore, the two associated SNPs (intron 18 and 19) were then selected, as they lie proximal to exon–intron boundaries. The intron 18 SNP is only 14 bp upstream from the predicted splice site of exon 19, as estimated with 91% confidence using splice-site prediction program, NetGene2 [Hebsgaard et al., 1996]. This proximity might influence RNA splicing [Bracco and Kearsey, 2003] or serve as a marker for another variant(s) that directly affects the function of the TRPM2 gene. Furthermore, the potential for regulation of gene expression by intronic transcribed noncoding RNAs has recently been recognized [Mattick, 2004]. Finally, these SNPs may be close to and in LD with other yet unidentified functional variants. The use of a continuously distributed phenotype in LD analysis has been shown to be a more powerful strategy for detecting genetic susceptibility to complex diseases than the simple classification of individuals into ‘‘affected’’ and ‘‘nonaffected’’ categories (e.g., [Leckman et al., 2001; Rowe et al., 87 (24.4) 69 (27.8) 18 (11.7) 107 (20.0) 97 (27.2) 75 (30.2) 22 (20.4) 121 (22.7) 109 (30.6) 75 (30.2) 34 (31.5) 119 (22.5) 159 (44.7) 108 (43.5) 51 (48.2) 119 (23.7) 136 (38.2) 90 (36.3) 46 (42.6) 133 (25.1) 113 (31.7) 77 (31.0) 36 (33.3) 118 (22.2) 92 (25.8) 65 (26.2) 27 (25.0) 98 (18.4) 269 (75.6) 179 (72.2) 90 (83.3) 427 (80.0) 259 (72.8) 173 (69.8) 86 (79.6) 413 (77.3) 247 (69.4) 173 (69.8) 74 (68.5) 411 (77.5) 197 (55.3) 140 (56.5) 57 (52.8) 383 (76.3) 220 (61.8) 158 (63.7) 62 (57.4) 397 (74.9) 243 (68.3) 171 (69.0) 72 (66.7) 414 (77.8) 264 (74.2) 183 (73.8) 81 (75.0) 434 (81.6) Allele 2 N (%) 1.54 (1.12–2.13) 1.57 (1.10–2.25) 1.48 (0.91–2.40) 1.92 (1.43–2.58) 1.58 (1.13–2.21) 1.75 (1.12–2.75) 1.85 (1.38–2.47) 1.70 (1.23–2.35) 2.21 (1.44–3.40) 2.60 (1.94–3.48) 2.48 (1.79–3.43) 2.88 (1.87–4.743) 1.52 (1.12–2.07) 1.50 (1.07–2.10) 1.59 (1.01–2.50) 1.28 (0.94–1.74) 1.48 (1.05–2.08) 0.87 (0.52–1.45) 1.29 (0.94–1.78) 1.54 (1.08–2.18) 0.80 (0.46–1.38) OR (95% CI) a Allele 1 is the major allele and allele 2 is the minor allele. *Bonferroni corrected P-values for the number of SNPs (N ¼ 7) tested. Promoter-SNP BD (N ¼ 178) BD-I (N ¼ 124) BD-II (N ¼ 54) Control (N ¼ 267) I-8-SNP BD (N ¼ 178) BD-I (N ¼ 124) BD-II (N ¼ 54) Control (N ¼ 267) I-16-SNP BD (N ¼ 178) BD-I (N ¼ 124) BD-II (N ¼ 54) Control (N ¼ 265) I-18-SNP BD (N ¼ 178) BD-I (N ¼ 124) BD-II (N ¼ 54) Control (N ¼ 254) I-19-SNP BD (N ¼ 178) BD-I (N ¼ 124) BD-II (N ¼ 54) Control (N ¼ 265) I-20-SNP BD (N ¼ 178) BD-I (N ¼ 124) BD-II (N ¼ 54) Control (N ¼ 266) I-27-SNP BD (N ¼ 178) BD-I (N ¼ 124) BD-II (N ¼ 54) Control (N ¼ 266) Diagnosis Allele 1a N (%) 6.19 5.48 2.07 10.13 7.09 6.11 17.31 10.36 13.61 41.77 30.97 24.45 7.43 5.48 4.01 2.43 5.18 0.27 2.43 5.88 0.65 w2 2.2 104 9.0 103 1.6 103 3.0 105 1.3 103 2.3 104 0.013 0.019 0.151 NS NS NS 0.011 0.055 NS <7.0 105 <7.0 105 7.0 105 <1.0 105 <1.0 105 1.0 105 0.002 0.008 0.014 0.042 NS NS NS NS NS NS NS NS Corrected P-value* 0.006 0.019 0.045 0.119 0.023 0.602 0.119 0.015 0.420 P-value 102 (57.31) 73 (58.9) 29 (53.7) 177 (66.5) 82 (46.1) 56 (45.2) 26 (48.1) 157 (59.0) 71 (39.9) 50 (40.3) 21 (38.9) 151 (57.0) 58 (32.6) 40 (32.3) 18 (33.3) 148 (58.3) 82 (46.1) 56 (45.2) 26 (48.1) 155 (58.5) 97 (54.5) 62 (50.0) 35 (64.8) 159 (59.6) 101 (56.7) 63 (50.8) 38 (70.4) 170 (63.7) 1/1 N (%) 60 (33.7) 37 (29.8) 23 (42.6) 80 (30.1) 79 (44.4) 59 (47.6) 20 (37.0) 100 (37.6) 78 (43.8) 58 (46.8) 20 (37.0) 95 (35.8) 81 (45.5) 60 (48.4) 21 (38.9) 93 (36.6) 83 (46.6) 61 (49.2) 22 (40.7) 101 (38.1) 65 (36.5) 49 (39.5) 16 (29.6) 95 (35.6) 67 (37.6) 53 (42.7) 14 (25.9) 87 (32.6) 1/2 N (%) 16 (9.0) 14 (11.3) 2 (3.7) 9 (3.4) 17 (9.6) 9 (7.3) 8 (14.8) 9 (3.4) 29 (16.3) 16 (12.9) 13 (24.1) 19 (7.2) 39 (21.9) 24 (19.4) 15 (27.8) 13 (5.1) 13 (7.3) 7 (5.6) 6 (11.1) 9 (3.4) 16 (9.4) 13 (10.5) 3 (5.6) 13 (4.9) 10 (5.6) 8 (6.5) 2 (3.7) 10 (3.7) 2/2 N (%) 7.85 9.74 3.35 11.47 7.80 11.97 16.12 10.19 15.52 41.05 31.44 30.35 8.20 6.30 6.60 3.28 5.74 0.71 2.46 6.14 0.94 w2 0.020 0.008 0.188 0.003 0.020 0.003 3.2 103 6.0 103 4.3 103 1.2 109 1.5 107 2.6 107 0.017 0.043 0.037 0.194 0.057 0.702 0.292 0.046 0.624 P-value TABLE III. Genotype and Allelic Association of TRPM2 SNPs Between Bipolar Patients (BD-I and BD-II) and Healthy Controls in Caucasian Population NS NS NS 0.063 NS 0.042 7 103 NS 7 103 2.0 108 3.0 106 3.6 106 NS NS NS NS NS NS NS NS NS Corrected P-value* Association of TRPM2 With Bipolar Disorder 41 42 Xu et al. TABLE IV. Association of Haplotypes With BD in Caucasian Population Haplotype T-T-T-C-T-T-A T-T-T-C-T-T-G C-C-C-G-C-C-A T-T-T-A-C-T-A T-T-C-C-C-C-A T-T-T-A-T-T-G T-T-C-C-C-C-G T-T-T-A-T-T-A Remaining hapa BD (%) (2N ¼ 356) 12.0 2.9 6.4 5.2 8.8 6.4 2.2 26.3 29.7 Control (%) (2N ¼ 536) 0.9 1.0 1.8 2.6 8.9 8.5 0.8 52.3 23.3 w2 49.24 14.39 6.57 3.94 0.26 1.60 0.38 61.04 — P-value* 12 2.3 10 1.5 103 0.010 0.047 0.610 0.206 0.538 5.6 1015 — Effect Susceptible Susceptible Susceptible Susceptible Protective Global P-value of 1.5 1017 (1.05 1016 after correcting for multiple testing of the number of SNPs [N ¼ 7]). *P-values were calculated using COCAPHASE as described in the Materials and Methods. a Haplotypes with an estimated frequency <0.8% in control are designed as ‘‘remaining haplotypes.’’ 2001]). This strategy was applied in this study, examining the association of TRPM2 SNP alleles and haplotypes in BD with [Ca2þ]B, a putative index of intracellular Ca2þ homeostasis, which has been found elevated in peripheral blood cells and BLCLs from BD patients [Emamghoreishi et al., 1997; Warsh et al., 2004]. Although no association was evident between the individual TRPM2 SNP genotypes evaluated and BLCL [Ca2þ]B, significantly higher BLCL [Ca2þ]B was found in BDI patients stratified on TRPM2-intron 19 SNP genotype. This suggests the possibility of a putative link between TRPM2 genetic variation and the alterations in Ca2þ homeostasis implied by elevated BLCL [Ca2þ]B, although the relationship may typify only a discrete subpopulation of BD-I patients within this heterogeneous disorder. Genetic variation in TRPM2 that is filtered through a network of gene/environment interactions [Petronis, 2003] and complex proteomic networks to modulate intracellular Ca2þ dynamics, could also confound detection of stronger associations with the latter. In order to ensure the quality of this case-control study of BD, we followed suggested standards for population-based association investigations of complex disorders [Schulze et al., 2003]. Original samples were used along with a quality control that included replicated samples in each 96-well plate. Moreover, a nucleotide-specific genotyping method was used based on TaqMan1 probes and 50 nuclease chemistry, a method that provides improved mismatch discrimination [Livak, 1999]. Genotype frequencies were in HWE and two-tailed P-values were estimated. To rule out potential population stratification and to minimize the risk of false-positive results in the present population-based association study, we subdivided Caucasian populations into four subgroups (British Isles, European– Other, Recent admixture of Europeans, and Uncertain European ancestry, Table IB) based on geographical patterns of origin and ancestry (e.g., Northern, Eastern, Southern European descendants). No statistical differences in allele, genotype, and haplotype distributions for seven TRPM2 SNPs were observed (data not shown). Notwithstanding the limited number of unlinked SNPs used, the lack of statistical support for population stratification in the Caucasian population investigated here, as tested with FASTAT and GENEPOP, argues against a major contribution of this phenomenon to the statistically significant differences in frequency of the intron 18 and intron 19 SNPs, and their haplotypes, between patients and controls. However, the possibility cannot be completely excluded. Secondly, possible interactive effects of environmental factor(s) or other gene(s) with TRPM2 that impact the pathophysiology of BD, cannot be excluded, although heritability is uninformative in distinguishing group differences related to interactive effects that influence the expression of a complex trait [Mountain and Risch, 2004]. In summary, we have conducted the first systematic screen of SNPs within the TRPM2 gene in a population based association study of BD and comparison healthy subjects. SNPs were used to capture haplotype diversity within the gene of interest. Potential susceptibility alleles were identified for several SNPs and the haplotypes of the TRPM2 gene, for BD as a whole, as well as for BD-I and BD-II patient subgroups. Moreover, a relationship was also suggested between the specific TRPM2-intron 19 genotypes and elevated [Ca2þ]B in BLCLs from BD-I patients. The present results clearly support the importance of more intensive investigations of TRPM2 and its potential role in modifying risk for BD and disturbed intracellular Ca2þ homeostasis abnormal in BD-I patients. ACKNOWLEDGMENTS Kin Po Siu, Arvind Kamble, and Karen Woo assisted with the transformation and culture of cell lines. Wendy Hiscox and Hester Tims assisted with the structured interview of patients and Damodar Godse with database development. The facilitation of subject recruitment by the staff of the Mood Disorders Program of CAMH and the Department of Psychiatry, University of Toronto is gratefully acknowledged. Supported, in part, by grants from the Canadian Institutes of Health Research, MOP 12851 and Ontario Mental Health Foundation to J.J.W. REFERENCES Aita VM, Liu J, Knowles JA, Terwilliger JD, Baltazar R, Grunn A, et al. 1999. A comprehensive linkage analysis of chromosome 21q22 supports prior evidence for a putative bipolar affective disorder locus. Am J Hum Genet 64:210–217. Barrett JC, Fry B, Maller J, Daly MJ. 2005. Haploview: Analysis and visualization of LD and haplotype maps. Bioinformatics 21:263–265. Berrettini WH. 2000. Genetics of psychiatric disease. Annu Rev Med 51: 465–479. Bracco L, Kearsey J. 2003. The relevance of alternative RNA splicing to pharmacogenomics. Trends Biotechnol 21:346–353. Cordell HJ, Clayton DG. 2002. A unified stepwise regression procedure for evaluating the relative effects of polymorphisms within a gene using case/control or family data: Application to HLA in type 1 diabetes. Am J Hum Genet 70:124–141. Curtis D. 1999. Chromosome 21 workshop. Am J Med Genet 88:272–275. Dawson E, Abecasis GR, Bumpstead S, Chen Y, Hunt S, Beare DM, et al. 2002. A first-generation linkage disequilibrium map of human chromosome 22. Nature 418:544–548. Detera-Wadleigh SD, Badner JA, Yoshikawa T, Sanders AR, Goldin LR, Turner G, et al. 1997. Initial genome scan of the NIMH genetics initiative bipolar pedigrees: Chromosomes 4, 7, 9, 18, 19, 20, and 21q. Am J Med Genet 74:254–262. Association of TRPM2 With Bipolar Disorder 43 Emamghoreishi M, Schlichter L, Li PP, Parikh S, Sen J, Kamble A, et al. 1997. High intracellular calcium concentrations in transformed lymphoblasts from subjects with bipolar I disorder. Am J Psychiatry 154: 976–982. Ohashi J, Tokunaga K. 2002. The expected power of genome-wide linkage disequilibrium testing using single nucleotide polymorphism markers for detecting a low-frequency disease variant. Ann Hum Genet 66: 297–306. Ewald H, Kruse TA, Mors O. 2003. Genome wide scan using homozygosity mapping and linkage analyses of a single pedigree with affective disorder suggests oligogenic inheritance. Am J Med Genet B Neuropsychiatr Genet 120:63–71. Perraud AL, Knowles HM, Schmitz C. 2004. Novel aspects of signaling and ion-homeostasis regulation in immunocytes; The TRPM ion channels and their potential role in modulating the immune response. Mol Immunol 41:657–673. First M, Spitzer R, Gibbon M, Williams JBW. 1995a. Structured clinical interview for DSM-IV Axis I disorders-patient edition (SCID-I/P) version 2.0. New York, NY: Biometrics Research Department, New York Psychiatric Institute. Petronis A. 2003. Epigenetics and bipolar disorder: New opportunities and challenges. Am J Med Genet 123C:65–75. First M, Spitzer R, Gibbon M, Willians JBW. 1995b. Structured clinical interview for DSM-IV Axis I disorders- Nonpatient edition (SCID-I/NP), version 2.0. New York, NY: Biometrics Research Department, New York Psychiatric Institute. Rowe DC, Stever C, Chase D, Sherman S, Abramowitz A, Waldman ID. 2001. Two dopamine genes related to reports of childhood retrospective inattention and conduct disorder symptoms. Mol Psychiatry 6:429–433. Gabriel SB, Schaffner SF, Nguyen H, Moore JM, Roy J, Blumenstiel B, et al. 2002. The structure of haplotype blocks in the human genome. Science 296:2225–2229. Hebsgaard SM, Korning PG, Tolstrup N, Engelbrecht J, Rouze P, Brunak S. 1996. Splice site prediction in Arabidopsis thaliana pre-mRNA by combining local and global sequence information. Nucleic Acids Res 24: 3439–3452. Raymond M, Rousset F. 1995. GENEPOP (Version 1.2) population genetics software for exact tests and ecumenicisim. J Hered 86:248–249. Schulze TG, Hardy J, McMahon FJ. 2003. Inconsistent designs of association studies: A missed opportunity. Mol Psychiatry 8:770–772. Segurado R, Detera-Wadleigh SD, Levinson DF, Lewis CM, Gill M, Nurnberger 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. Sklar P. 2002. Linkage analysis in psychiatric disorders: The emerging picture. Annu Rev Genomics Hum Genet 3:371–413. Kelsoe JR, Spence MA, Loetscher E, Foguet M, Sadovnick AD, Remick RA, et al. 2001. A genome survey indicates a possible susceptibility locus for bipolar disorder on chromosome 22. Proc Natl Acad Sci USA 98: 585–590. Smyth C, Kalsi G, Curtis D, Brynjolfsson J, O’Neill J, Rifkin L, et al. 1997. Two-locus admixture linkage analysis of bipolar and unipolar affective disorder supports the presence of susceptibility loci on chromosomes 11p15 and 21q22. Genomics 39:271–278. Kwok JB, Adams LJ, Salmon JA, Donald JA, Mitchell PB, Schofield PR. 1999. Nonparametric simulation-based statistical analyses for bipolar affective disorder locus on chromosome 21q22.3. Am J Med Genet 88: 99–102. Straub RE, Lehner T, Luo Y, Loth JE, Shao W, Sharpe L, et al. 1994. A possible vulnerability locus for bipolar affective disorder on chromosome 21q22.3. Nat Genet 8:291–296. Lahiri DK, Nurnberger JI Jr. 1991. A rapid non-enzymatic method for the preparation of HMW DNA from blood for RFLP studies. Nucleic Acids Res 19:5444. Leckman JF, Zhang H, Alsobrook JP, Pauls DL. 2001. Symptom dimensions in obsessive-compulsive disorder: Toward quantitative phenotypes. Am J Med Genet 105:28–30. Liu J, Juo SH, Terwilliger JD, Grunn A, Tong X, Brito M, et al. 2001. A follow-up linkage study supports evidence for a bipolar affective disorder locus on chromosome 21q22. Am J Med Genet 105:189–194. Livak KJ. 1999. Allelic discrimination using fluorogenic probes and the 50 nuclease assay. Genet Anal 14:143–149. Mattick JS. 2004. RNA regulation: A new genetics? Nat Rev Genet 5: 316–323. Tishkoff SA, Verrelli BC. 2003. Role of evolutionary history on haplotype block structure in the human genome: Implications for disease mapping. Curr Opin Genet Dev 13:569–575. Warsh JJ, Andreopoulos S, Li PP. 2004. Role of intracellular calcium signaling in the pathophysiology and pharmacotherapy of bipolar disorder: current status. Clin Neurosci Res 4:201–213. Yoon I-S. 2002. Molecular analysis of altered calcium homeostasis in bipolar affective disorder using differential display and candidate gene approaches. PhD Thesis. University of Toronto. Yoon IS, Li PP, Siu KP, Kennedy JL, Macciardi F, Cooke R, et al. 2001. Altered TRPC7 gene expression in bipolar disorder. Biol Psychiatry 50:620–626. Mountain JL, Risch N. 2004. Assessing genetic contributions to phenotypic differences among ‘racial’ and ‘ethnic’ groups. Nat Genet 36:S48–S53. Zaykin DV, Westfall PH, Young SS, Karnoub MA, Wagner MJ, Ehm MG. 2002. Testing association of statistically inferred haplotypes with discrete and continuous traits in samples of unrelated individuals. Hum Hered 53:79–91. Nagamine K, Kudoh J, Minoshima S, Kawasaki K, Asakawa S, Ito F, et al. 1998. Molecular cloning of a novel putative Ca2þ channel protein (TRPC7) highly expressed in brain. Genomics 54:124–131. Zhang W, Chu X, Tong Q, Cheung JY, Conrad K, Masker K, et al. 2003. A novel TRPM2 isoform inhibits calcium influx and susceptibility to cell death. J Biol Chem 278:16222–16229.