Methods

Subjects
Since 1998, probands were recruited from bipolar outpatients in the Chung Shan Medical University Hospital and the Taichung Rehabilitation Hospital in Taiwan. Controls were subjects from local volunteer blood donors who have no family or personal history of major affective disorder and other psychiatric disorders were matched to cases on the basis of ethnic or geographic origin, sex, and age. The gender ratios are identical in both case and control groups (male to female is 53% to 47 %) and the age of populations are 38 ± 13.7 and 26.4 ± 7.9 in case and control groups, respectively. The age of onset is 28.4 ± 12.7 in all patients and is 30.3 ± 14.5 and 26.4 ± 10.2 in male and female patients, respectively. Clinical interviews were conducted by experienced psychiatrists (leading by Dr. Lai) for all subjects after the study procedure had been fully explained and information on general demographic data, such as age, sex, and ethnicity, was obtained. Patients assessment was purely by direct clinical interview by the treating clinician according the procedure described in the DSM-IV diagnoses of lifetime Major depressive disorder and bipolar I. Additional information required to reach diagnosis was also collected from all clinical and hospital records where available but the comorbidity with other psychiatric/neurological disorders or medical problems were not considered as an inclusion/exclusion criteria. This study was approved by the University Ethics Committee and written informed consent was obtained from all participants. In total, 108 BPD and 103 controls (all Taiwanese Han) were recruited for this study.

DNA preparation
Approximately 10 ml of peripheral blood was collected from the recruits using EDTA anticoagulant venous blood tubes, and DNA prepared using a BIO-RAD InstaGene™ Whole Blood Kit (Bio-Rad Laboratory, Hercules, CA) according to the manufacturer's protocol.

Primer design for searching for, and typing of, TPH1 gene polymorphisms
All primers used were designed using the Primer3 program [36] or OLIGO 5 Primer Analysis Software (Molecular Biology Insights, Inc., cascade. CO., USA). The primers used to screen the TPH1 gene (Table 1) were designed using the published sequence (GenBank accession number: AC005728). The primers used in exon-wide SNP scanning were designed from the intronic sequence roughly 50 bp upstream and downstream of each exon to amplify the entire exon sequences. In addition, overlapping fragments covering from 1,151 bp upstream of the 5' promoter region to 1,572 bp downstream of the 3'UTR were also amplified to screen for polymorphisms within these regions. The forward and reverse primers for base excision sequence scanning (BESS) product amplification were labeled, respectively, with fluorescent 6-FAM and HEX to facilitate variant detection.

SNP identification in the coding and regulatory regions of the TPH1 gene
In order to identify all SNPs in the coding and regulatory regions of the TPH1 gene, we designed primers to generate PCR products for use in SNP identification using the BESS Base Reader Kit (Epicentre Technologies, Madison, WI) according to the manufacturer's protocol. Each identified SNP was sequenced to confirm the sequence variant. The screening panel included 50 unrelated subjects with or without BPD.
The BESS T & G Base Reader Kit, which identifies all types of point mutation, deletion, insertion, repeat expansion, and frameshift mutation at sites involving thymine or guanine, was used to systematically search for sequence variants in pooled samples. DNA sequence variants are detected by cleavage of the amplification products at modified nucleotides, generating a defined series of fragments which can be easily separated on a standard sequencing gel and detected using a fluorescent dye detection system. Briefly, PCR products were generated using FAM-labeled forward primers and HEX-labeled reverse primers. PCR amplification was performed in a 25 μl volume containing 1 unit of Taq polymerase, 1× PCR buffer, 0.2 μM of each labeled primer, an appropriate concentration of MgCl2, 200 μM BESS T/G Scan dNTP Mix, and 50 ng of genomic DNA. Thermal cycling conditions were a pre-denaturation of 3 min at 94°C; 35 cycles of 30 sec at 94°C, 30 sec at the appropriate annealing temperature indicated in Table 1, and 30 sec of extension at 72°C; and a final extension at 72°C for 5 min.
For the excision reaction, 5 μl of the amplification reaction was mixed with 1 μl of the BESS T/G-Scan Excision Enzyme Mix and 1 μl of 10X BESS T/G-Scan Excision Enzyme Buffer, the mixture incubated for 30 minutes at 37°C, and the reaction stopped by adding 5 μl of Stop/Loading Buffer. One microliter of the excision reaction products was mixed with gel loading solution containing 12 μl of formamide and 8 μl of GENESCAN™-500 size standards (Applied Biosystems, Forster City, CA, USA), then the mixture was denatured for 5 min at 95°C, loaded onto a capillary polymer of the ABI 310 Genetic Analyzer, and run for 30 min for size separation. Analysis was performed using the GENESCAN 672 program (Applied Biosystems).
To reduce the cost and speed up the process, a pooling methodology was used. A preliminary test indicated that the sensitivity of BESS T/G-Scan analysis allowed the detection of alleles with a frequency in the population greater than 4.5 % (data not shown), i.e., The assay can recognize 1 heterozygous individual in a DNA pool from this individual and 9 homozygotes (equivalent to a minor allele with a frequency of 5%). DNA from 10 individuals was therefore pooled (10 ng/μl of DNA from each individual) and five such pools, representing 100 chromosomes, were prepared from 50 randomly selected subjects for polymorphism identification.

Selection of SNP markers for the TPH1 gene
In addition to the SNPs identified in the present study, five SNP markers within and flanking the human TPH1 gene were also selected to test the association with BPD. Based on the literature [34] and a public SNP database [35], SNP markers roughly 10 kb apart and with a relatively high minor allele frequency were selected. The TPH1 intron 7 A218C polymorphism, which has been suggested to be associated with Taiwanese BPD in our previously work, also has been included in the present study (i.e. in7SNP1; A20004C). The genomic localizations of the 10 SNPs examined relative to the transcription start site are given in Table 1.

SNP genotyping
Except for the intron 7 A218C polymorphism, which was genotyped using a modified Amplification Refractory Mutation System (ARMS), all other SNPs of the TPH1 gene were genotyped using a multiple SNP genotyping system which involves multiplex PCR and multiple single base extensions (MSBE). Briefly, two multiplex PCR reactions were performed. One was in a 20 μl volume containing 70 ng of genomic DNA, 2.4 μl of primer mix (0.1 μM of the primer pairs for 5'flankingSNP3, in1SNP1, in2SNP1, and in6SNP1, and 0.2 μM of the primer-pair for in3SNP1), 400 μM dNTPs, 1× PCR buffer, 1.5 mM MgCl2, and 1 U of Taq polymerase. The conditions used were an initial denaturation step of 5 min at 95°C, 30 cycles of amplification (30 s at 95°C, 60 s at 51°C, 90 s at 72°C), and a final extension step of 10 min at 72°C. The other multiplex PCR was performed under the same conditions, but using 0.8 μl of primer mix containing 0.1 μM of the primer pairs for 5'flankingSNP1, 5'flankingSNP2, 3'UTRSNP1, and 3'UTRSNP2. A 4 μl aliquot of the PCR products was treated with 5 U of shrimp alkaline phosphatase (SAP) and 0.1 U of exonuclease I in a total volume of 10 μl to remove primers and unincorporated dNTPs; the reaction was carried out at 37°C for 1 hour, then terminated by incubation at 72°C for 15 minutes to inactivate the enzymes. The multiplex SBE reaction was carried out using SNP-specific primers and fluorescent-labeled terminators (the ABI PRISM SNaPshot Multiplex Kit). The short-extension reaction was performed on a thermal cycling machine for 25 cycles of 10 s at 96°C, 5 s at 50°C, and 30 s at 60°C. After short-extension, excess ddNTPs were removed from the SBE products by addition of 0.5 U of SAP to the reaction mixture and incubation at 37°C for 1 hour. The purified SBE products were electrophoresed on a ABI PRISM 310 Genetic Analyzer and analyzed using GeneScan software (ABI PRISM).

Statistical analysis
The chi-squared test for allelic and genotypic distributions between patients and controls was performed using the CROSSTAB program implemented by SPSS. The Hardy-Weinberg equilibrium was analyzed using the HWE program, version 2.33 [37]. Pairwise LD coefficients D' [38] among the 10 SNPs were estimated and statistical significances were determined by using the SNP Alyze® program (SNP and Disease Association Analysis software; Dynacom Co., Ltd. Kanagawa, Japan). In addition, the PHASE 2.0 program [39,40] was used to construct haplotypes and perform a case-control permutation test, then the Fisher's exact test was applied to test differences in haplotype frequencies between cases and controls. All Fisher's exact tests (two tails) and estimation of the odds ratio of BPD associated with a particular haplotype were performed using the PROC FREQ program implemented by SAS package (SAS Institute Inc., Cary, NC, USA). Haplotype-tag SNPs were selected using SNPtagger software [41]. To consider the multiple comparisons, a Bonferroni correction was applied in this study thus the p value for reaching significance is 0.005 for 10 SNPs. In addition, the potential confounders such as personality disorders, substance abuses or organic disorders were not considered in the statistic analyses.