Association study of ABCA1 polymorphisms in singapore populations 4

Chapter Materials and Methods Materials and Methods 4.1 General Methods 4.1.1 Nomenclature for ABCA1 SNPs Locations of sequence variations in the ABCA1 proximal promoter were described relative to the transcriptional start site (TSS) mapped by Santamarina-Fojo et al. (2000). Sequence variations in the UTR and coding regions were numbered with reference to the ABCA1 cDNA Genbank accession NM_005502; in addition, coding SNPs (cSNPs) were also identified by their amino acid positions. Intronic SNPs were described in relation to their distances from the exon-intron splice junctions, with ‘+’ and ‘-‘ designating positions relative to the 5’ or 3’ splice junctions respectively, whichever was nearest to the SNP (den Dunnen and Antonarakis, 2001). To illustrate, IVS1+92T>C implies a T>C variant within intron 1, 92 bp downstream from the end of exon 1. Insertion/deletion polymorphisms were abbreviated as indels. 4.1.2 Genomic DNA Extraction Genomic DNA was isolated from whole blood using a slightly modified method of Parzer and Mannhalter (1991). Five ml of blood were transferred to a sterile 15 ml centrifuge tube, incubated in 10 ml of lysis buffer (0.32 M sucrose, mM MgCl2, 0.01 M Tris-HCl, 1% Triton X-100, pH 8.0) for 15 on ice, and then centrifuged at 2000 xg for 10 at room temperature. The supernatant was discarded. The pellet was resuspended in washing buffer (10 mM NaCl, 10 mM EDTA, pH 8.0) and spun as before. The pellet, which should appear almost white (if still reddish, the previous washing step was repeated), was mixed with 350 ul of 20% sodium N-laurylsacrosine, 250 ul of 7.5 M ammonium acetate, 3.5 ml of M guanidine hydrochloride and 125 ul of 10 mg/ml proteinase K, and incubated at 60oC with intermittent shaking. When digestion was 38 Chapter Materials and Methods completed in approximately 10-15 min, the tubes were placed on ice. To precipitate DNA, ice-cold absolute ethanol was added and the tube gently inverted a few times. The DNA, which might be visible as strands depending on yield, was transferred to 1.5 ml microfuge tubes and dried overnight on the bench. The DNA pellet was resuspended in TE buffer (10 mM Tris-HCl, mM EDTA, pH 8.0) and stored at 4oC. 4.1.3 Polymerase Chain Reaction (PCR) Generally, PCR was carried out in 20 ul reaction volumes comprising of 50-200 ng genomic DNA, 0.2 mM of each dNTP (MBI Fermentas), 0.4% dimethylsulfoxide (SigmaAldrich), 20 pmol of each primer (Operon), 1.5 mM MgCl2, 0.2 U Taq polymerase (DynaZymeII, Finnzymes Oy) and 1x manufacturer’s buffer. To accommodate for the high GC content in the ABCA1 gene promoter, the PCR recipe was modified: 2x PCRx enhancer (Invitrogen) in place of dimethylsulfoxide, and 2.5 U native Taq polymerase from Invitrogen in place of DynaZymeII. Thermal cycling was performed on a GeneAmp 9700 machine (Applied Biosystems, ABI) using a Touchdown protocol: initial denaturation at 95oC for 3-5 followed by 34 cycles of denaturation at 95oC for 30 sec, annealing at 63oC for 30 sec with temperature decrement of 0.5oC per cycle until 56oC and 56oC for the remaining 20 cycles, and extension at 72oC for 30 sec; and a final extension at 72oC for min. Touchdown PCR was chosen because it eliminates the need for optimization of annealing temperature as well as reduces spurious priming of templates (Don et al., 1991). PCR primers were obtained from literature or designed using Oligo (Molecular Biology Insights Inc). 4.1.4 Preparation of Native Polyacrylamide Gels The mini polyacrylamide vertical gel systems, miniVE or SE240 from Amersham Biosciences were used. Each pair of glass plates (glass and notched silica plates for SE240) was rinsed in tap water and dried. Prior to use, the plates were wiped with 70% 39 Chapter Materials and Methods ethanol, spacers placed between the plates and clamped in the gel casting stand. Sixteen ml of gel mixture was sufficient for casting two mini gels. The gel mixture consisted of an appropriate volume of acrylamide stock solution (prepared from 30% or 50% stocks for regular and single-strand conformation polymorphism (SSCP) gels respectively), glycerol (SSCP gels only) and stock 5x TBE buffer added to a final 1x. Prior to casting of gels, 100 ul of 10% ammonium persulphate solution and 20 ul of TEMED were added to initiate polymerization. The gel mixture was poured between the clamped plates using a serological pipette, combs put in place and gels left to set at room temperature. After the gels had set, the combs were removed and the wells were rinsed thrice with distilled water before use. 4.1.5 Gel Visualization by Ethidium Bromide Staining Ethidium bromide was added to molten agarose at a final concentration of 0.2 ug/ml before pouring into the gel tray. Polyacrylamide gels were stained post-electrophoresis in water containing 10 ug/ml of ethidium bromide for min. Stained gels were viewed under ultraviolet (UV) illumination and images captured on Polaroid T57 black and white film. 4.1.6 Gel Visualization by Silver Staining Polyacrylamide gels were removed carefully from the glass plates and soaked in 10% ethanol for (fixation) in plastic containers, followed by freshly prepared 1% nitric acid (oxidization of non-DNA containing areas) for min, and then transferred to a transparent glass dish containing 0.1% silver nitrate for 20 with shaking. Excess silver ions were rinsed away in deionized water, followed by a final rinse in the developing solution (3% sodium carbonate with 0.05% v/v formaldehyde added just prior to use). The remaining developing solution was added to the gels and incubated with shaking. To aid in visualization of bands, a sheet of white paper was placed beneath the glass tray. If the brownish DNA bands had not appeared and the developing solution had already turned 40 Chapter Materials and Methods dark, the solution was replaced; this was necessary to reduce background staining. When sufficient intensity of the DNA bands had been achieved (often within ~10 min), the developing solution was discarded and 10% acetic acid added to quench the reaction. To maintain a permanent record of the stained gels, they were shrunken first in 80% methanol to ~70% of original size, placed between cellophane sheets, clamped in place with a Perspex drying frame and binder clips, and air-dried for two days. Although the gels were shrunken to ~70% of original size, it was necessary since high percentage polyacrylamide gels tended to crack upon drying. Despite the smaller gel size upon preservation, there was no loss in resolution of bands. Indeed, shrunken gels gave sharper bands (Mohamed et al., 1989). 4.1.7 Big Dye Terminator Sequencing PCR products were purified directly or following agarose gel electrophoresis using the QIAQuick kit (Qiagen). Purified PCR templates were quantitated using UV absorbance at 260 nm before they were sequenced using fluorescent dye terminator chemistry. Each 20 ul sequencing reaction comprised of ul BigDye Ready Terminator mix (version or 3, ABI), ul sequencing buffer (400 mM Tris-HCl, pH 8.0, 10 mM MgCl2), 3.2 pmol primer and 3-10 ng purified PCR template. Thermal cycling was performed on a GeneAmp9700 machine (ABI) with 25 cycles of 96oC for 10 sec, 50oC for sec and a final extension of 60oC for in 9600 emulation mode. Sequencing products were precipitated using isopropanol (BigDye Terminator Ready Mix version 2) or sodium acetate-ethanol (BigDye Terminator Ready Mix version 3). Precipitated sequencing products were then resuspended in 10 ul of HiDi formamide (ABI), transferred into 96-well optical microplates and ran on the ABI3100 Genetic Analyzer with the appropriate instrument settings. The DNA traces were inspected visually for SNPs. 41 Chapter Materials and Methods 4.2 SNP Discovery 4.2.1. DNA Samples For SNP discovery using resequencing and DHPLC, DNA samples were extracted from discarded cord blood samples of 16 unrelated, anonymous neonates representing each of the three main ethnic groups (Chinese, Malays and Indians) of Singapore. These samples were obtained from the National University Hospital. Ethical approval was obtained for the usage of the DNA samples for the purpose. 4.2.2 Resequencing of ABCA1 Proximal Promoter A segment of the ABCA1 proximal promoter spanning 594 bp downstream and 297 bp upstream of the transcriptional start site was amplified using Touchdown PCR, purified and sequenced using the amplification as well as internal primers (Figure 4.1). 4.2.3 DHPLC Screening of ABCA1 Exons An outline of the DHPLC procedure is provided in Figure 4.2. For DHPLC assay development, primers were designed to amplify most of the 50 ABCA1 exons individually using Oligo (Table 4.1). Amplicon sizes ranged between 200 and 800 bp. Oven temperature for the optimal separation of heteroduplex molecules in each amplicon was selected using DHPLCMelt (http://insertion.stanford.edu/melt.html) or WAVEMAKER™ (Transgenomic). Ideally, amplicons with one predominant melting domain are preferred since it means that heteroduplex detection could be achieved at a single optimal melting temperature (Topt). In some cases, this criterion could be simply fulfilled by redesigning the primers. In other situations, the sequences themselves possessed intrinsic properties that would entail analysis at two different oven temperatures. Following Touchdown PCR, PCR products were denatured at 95oC for and allowed to reanneal for 30 at room temperature before DHPLC analysis. The 42 Chapter Materials and Methods Transgenomic WAVE DHPLC system comprised of a 96-well autosampler mounted on a Peltier cooling rack, pumps, buffer degasser, 20 ul sample loop, DNASep (Transgenomic) separation column housed in a temperature-controlled oven, online variable wavelength UV detector and a Windows NT computer running the WAVEMAKER™ software for analysis protocols and data collection. Buffers were: buffer A, 0.1 M triethylammonium acetate (TEAA; Transgenomic) pH 7.0; and buffer B, 0.1 M TEAA and 25% HPLC-grade acetonitrile (Fisher Scientific). Eluted products were monitored online by UV absorbance at 260 nm. Sample injection volumes were 3-5 ul. Starting and ending gradient conditions were determined using WAVEMAKER™ with necessary adjustment when the peaks eluted too close to the injection or wash peaks. A standard linear gradient of 2% buffer B and a flow rate of 0.9% buffer B per ml were used throughout. Prior to heteroduplex detection, a subset of PCR samples was analyzed by DHPLC in non-denaturing mode at 50oC or on agarose gels to ensure specificity of PCR. The actual optimal temperature (Topt) for heteroduplex analysis was determined semi-empirically by serial injections at temperatures starting 2oC below the predicted Topt (given by DHPLCMelt or WAVEMAKER™) to 2oC above in 1oC increments. The temperature that resulted in first major time shift of DNA peak was selected for variant detection. The actual DHPLC screening temperatures used are provided in Table 4.1. Representative samples showing heteroduplex peaks were re-amplified and sequenced with the amplification primers. Samples that yielded homoduplex peak profiles served as reference. 4.2.4 In silico SNP Discovery from ABCA1 ESTs The SNPFINDER (Buetow et al., 1999; http://lpgws.nci.nih.gov:82/perl/snp/snp_cgi.pl) program was used to mine candidate SNPs from overlapping regions of ABCA1-specific, non-redundant ESTs in the UniGene cluster Hs.211562. To maximize the number of sequences for multiple alignment, no reverse strand confirmation was specified. High 43 Chapter Materials and Methods confidence putative SNPs in which mismatches were seen across multiple, independent sequences and were flanked by high quality bases, were validated by sequencing. 4.2.5 Inferring Functional Significance of ABCA1 SNPs The functional importance of ABCA1 variants was inferred by various methods. Putative transcription factor binding sites in the proximal were identified by searching against the TRANSFAC database (Kel et al., 2003) using MATCH (http://www.gene- regulation.com/cgi-bin/pub/programs/match/bin/match.cgi). Coding SNPs causing amino acid replacements were classified according to BLOSUM62 (Henikoff and Henikoff, 1992) and Grantham’s scores (Grantham, 1974). The BLOSUM62 scoring matrix is derived from the observed amino acid substitutions in a large set of approximately 2000 conserved amino acid blocks from 500 protein families. Positive and negative BLOSUM62 scores denote frequent and rare amino acid substitutions respectively. Grantham’s D values are judged on three physiochemical properties of amino acids (polarity, volume and composition) and are correlated to the evolutionary exchangeability of amino acid residues. Sequences from ABCA1 orthologues from diverse species were aligned using ClustalW. 44 Chapter Materials and Methods -515-494U PrU -564T>C -417-400U -463C>T TSS +1 -407G>C -302C>T -278C>G -99G>C -14C>T +280+297L PrL Primers PrU -515--494U -417--400U PrL +280-+297L Location* -594 to -574 -515 to -494 -417 to -400 -52 to -69 +297 to +280 Sequence (5’Æ3’) CAAAAGCAGCCCATTACCCAG ACAAAATGATTGGCGTCCTGA CGGAAAGCACGATTTAG AGCCGCCCACGACACAT CCGAGCACTTCCCGAAGC Figure 4.1 Resequencing of the ABCA1 gene proximal promoter. A 891 bp fragment was amplified with the primers PrU and +280-+297L and sequenced using nested as well as the PCR primers. Positions of SNPs identified in this study are indicated and are numbered with respect to the transcriptional start site (TSS) identified by Santamina-Fojo et al. (2001). 45 Chapter Materials and Methods ABCA1 genomic sequence AF275948 Exon and flanking intron sequence Primer design Melting curve simulation WAVEMAKER or DHPLCMelt (i) predicted melting domain (ii) >1 predicted melting domain Primer synthesis PCR Denaturation and reannealling Experimental Topt determination using 1oC increment starting 2oC below predicted melting temperature(s) to 2oC above. Figure 4.2 Outline of exon-centric SNP discovery in the ABCA1 gene using DHPLC. 46 Chapter Materials and Methods Table 4.1 PCR primers and DHPLC analysis temperatures for SNP screening in ABCA1 exons. Region Forward primer Reverse primer Amplicon size (bp) Exon size (bp) Topt Exon 5’-CCGGCTCCACGTGCTTTC-3’ 5’-CCTCCGGTCCCATCTGAGTTGC-3’ 5’-GGGTTTGGCTCTGGGTACTGC-3’ 5’-GCTGCTCTCCAGGAATCAGTTCG-3’ 5’-GGCTTTGCAGCAATAACTGATG-3’ 5’-AGCCACATGCCATATCCAGAC-3’ 5’-GGGAAGACACAGAGGTATATGG-3’ 5’-AGTCTCCTGAGTTCATGCCTTA-3’ 5’-GAGCCTCAAAATCGCTTCA-3’ 5’-TACTCGAGGTGATGGATACCC-3’ 5’-GGAGCCAGGCTGAGTTCTATG-3’ 5’-GCAACCAGCTCCCCTAGTTTC-3’ 5’-CCTGATATGGCGATGCTC-3’ 5’-CTGGATGGTTTGGCAATTCCT-3’ 5’-TGCCTTGTGCTTTGATACATTC-3’ 5’-CCCAAAAGTCTGAAAGAACACTA-3’ 5’-AATGAACAAAATGCAAACTTAC-3’ 5’-ACCAATTAGCTCTGAGACATCTA-3’ 5’-GTAATGCTGATGCTGCTCGTC-3’ 5’-CCAAGGCCAGAACTAAGGGA-3’ 5’-GCTGGGGTTTCAACTAAGAAC-3’ 5’-TTCCAGAAGAGCCGTGAGTA-3’ 5’-TTTCAGGGAGGAGCCAAACG-3’ 5’-GTGGTGCCCTGTGACTTTAGC-3’ 5’-GTGCTCCTCAAGATTTAGTTGG-3’ 5’-CCTGCCTGAACCTTATTGTAAC-3’ 5’-TGGCTGCCAACGTTCTCATGAA-3’ 5’-CCCAAGACCAGGCTGGTGTG-3’ 5’-GGCTTGGTAAGGGTAGTAGGGT-3’ 5’-CCCCATCTGGCACAGTATAAACT-3’ 5’-TCTACATTGCATTGCTCGTGAA-3’ 5’-TTCTCCTCCCTTAGCCCGTGTT-3’ 5’-CGCCTCTGCATTGTCTCTAACA-3’ 5’-CCAGAGGTACTCACAGCGAAGA-3’ 5’-GGTAGCCCACCACTCCCCTAAAG-3’ 5’-ATCAGCTGCCTGTCCTTGGACTA-3’ 5’-ATGATGCTGAGCTTGGCTCATAC-3’ 5’-CAGCAGCACTAGGTTAAAGAAAG-3’ 5’-TGCTCCTTTATCATCGTATTCCC-3’ 5’-ACATGTTCACATTCGGCAACTCC-3’ 5’-CCCATAGGCCCACTTGTAGTTA-3’ 5’-GCTGAGGCAGGAGAATCGCTTGA-3’ 5’-TGGGCCAGGGGACACTGTATTCT-3’ 5’-GCCAGGGACAAGTTTCTGTTACC-3’ 5’-GGTGCGGTTGGTAACAGA-3’ 5’-AGCCTAGCCATGAGATACAGC-3’ 5’-CCAGTGCTTACCCCTGCTAA-3’ 5’-AACAGAGCAGGGAGATGGTG-3’ 5’-GAGGCTGTTCACACGGCACAC-3 5’-AGGCACCCCAGCAAGCATTAG-3’ 5’-CCAGCTGTCCACCCATTCTACAC-3’ 5’-TTCCTGCTATCTCCCACTACCA-3’ 514 221 o 60 C 510 136 56oC 382 66 60oC 470 94 406 142 54, 59oC o 58 C 406 119 o 61 C 463 122 538 177 54, 59oC o 57 C 485 93 526 241 53, 58oC o 59 C 354 140 60oC 347 117 60oC 333 198 59oC 281 206 o 60 C 330 176 474 223 54, 57oC o 61 C 354 222 o 61 C 525 205 60oC 200 114 o 58 C 517 172 61oC 560 132 339 143 56, 61oC o 62 C 333 138 61oC 446 221 59oC 457 73 o 62 C 636 203 59oC Exon 1A Exon Exon Exon Exon Exon Exon Exon Exon Exon 10 Exon 11 Exon 12 Exon 13 Exon 14 Exon 15 Exon 16 Exon 17 Exon 18 Exon 19 Exon 20 Exon 21 Exon 22 Exon 23 Exon 24 Exon 25 47 Chapter Materials and Methods Table 4.1 Continued from previous page. PCR primers and DHPLC analysis temperatures for SNP screening in ABCA1 exons. Region Forward primer Reverse primer Exons 26-27 5’-TGGCAGTGGTGACTACTCTCGTA-3’ 5’-GGATCTATCACCTTGGCTAAAG-3’ 5’-GCAATGCCGAAAACACCTAAG-3’ 5’-CTCCGGCATCCATATCTTGAC-3’ 5’-TTGGTGCAGCACAGTCATAGG-3’ 5’-TGGTCTGCTCGAATCTTACCC-3’ 5’-CCCTGATGCCGAATACAG-3’ 5’-TGCACGGTGTTAGAAGACTCTCC-3’ 5’-CCATCTCTGGTAATCCTACTCTTG-3’ 5’-GATATCTCACTCATTCCTGCTTCC-3’ 5’-TCAGGTTTCCGGTCACACTG-3’ 5’-CCCACTGTTTCAGTCTGTTATTTG-3’ 5’-TTGCATCATTAGGAATAGGCTC-3’ 5’-CCAAGGCTTTCTTCAATCCAAG-3’ 5’-ACCTTGTACACACTCGCACTG-3’ 5’-CGTTTAACCTGCCAACTACTC-3’ 5’-ACTTGGTTCTCAGGAGCC-3’ 5’-CCCCTGCCAACTTTACC-3’ 5’-GTGTTTTGCCTTGGTATGTGAC-3’ 5’-GCTGTTCCCCTACAATGAG-3’ 5’-GTCCATGTCCTCACTGATTG-3’ 5’-TTCCTGATGATAGCCAGAGC-3’ 5’-GCATTGTATATTTGATTTAGGGGTG-3’ 5’-AATTAAACACTGTCCTCTGGCT-3’ 5’-CTTCACTCCCATATTTCAGAACTTG-3’ 5’-ATCTCCATTAAAGCATCCTACAGC-3’ 5’-TCTGACCGGATCTCTGCATTGTG-3’ 5’-CAGGTGCTCCACGGGTTCTAAG-3’ 5’-ATCCTGGAGACTGTGGCAAGTAG-3’ 5’-GCCCTTTTATTAAGCAAGTCAGC-3’ 5’-TGGCTTAGGGTTATTGACAAGT-3’ 5’-TCTGAAATCATTCTCTTGACATACA-3’ 5’-AGAATTGAAGGTTTGAGGTAGTTAC-3’ 5’-TTTATGTGCTTCTTTACCCTTTACT-3’ 5’-TTTCAAGTAGCAGCAGATGTATTGG-3’ 5’- CAGCCTGAAGTCAATGCGTGTG-3’ 5’-TGTATGTGTAGGACAGCATGATAA-3’ 5’-GGGAAGACAAGCCAATCATACAAC-3’ 5’-AAATGAATTTGAAAGTTACTCTAAT-3’ 5’-GATCGCATATTCTACTTGGA-3’ 5’-ACTGAACAGCATCATCCCTATATCC-3’ 5’-GTTAAGTAACTTGCCCAAGAGTCAC-3’ 5’-GGGTTCCCAGGGTTCAGTAT-3’ 5’-GATCAGGAATTCAAGCACCAA-3’ 5’-TAATTCTGTATGCTCCTACTTGACC-3’ 5’-CATTGCATTGCATTGAATAGTATCAG-3’ Exon 28 Exon 29 Exon 30 Exon 31 Exon 32 Exon 33 Exon 34 Exon 35 Exon 36 Exon 37 Exons 38-39 Exon 40 Exon 41 Exon 42 Exon 43 Exon 44 Exon 45 Exon 46 Exon 47 Exon 48 Exon 49 Exon 50 Amplicon Size (bp) 528 Exon size (bp) 163 Topt 310 149 55, o 56 C o 60 C 327 125 o 61 C 496 99 318 190 55, o 60 C o 61 C 289 95 o 58 C 177 33 o 58 C 270 105 o 58 C 337 75 o 55 C 390 170 o 59 C 344 178 o 60 C 494 361 276 124 194 130 55, o 60 C 54, o 59 C o 60 C 321 121 o 57 C 369 63 o 56 C 342 107 o 55 C 364 142 o 55 C 311 135 o 59 C 333 104 o 57 C 448 93 o 56 C 436 245 o 59 C 413 141 52, o 57 C 48 Chapter Materials and Methods 4.3 Genetic Association Study 4.3.1 Cases and Controls Cases were unrelated patients who had been admitted to the Cardio-Thoracic Surgery Unit of the Singapore General Hospital for coronary bypass graft surgery between 1989 and 2000. Eligibility criterion included at least 50% stenosis in one or more major coronary arteries as confirmed by coronary angiography. A detailed family history for CAD, hypertension, diabetes and smoking, along with clinical data such as height, weight, blood pressure and smoking habits were documented. Further inclusion criteria for patients in the association study were absence of valve disease, cardiomyopathy, hypertension or diabetes. The mean age of the male cases analyzed in this study was 58.17 ± 9.14 years and ranged between 32 and 97 years. Unrelated controls were obtained from two sources. The first set of controls was randomly selected when they were undergoing mandatory pre-employment and annual medical checkups at the Singapore Anti-Tuberculosis Association Chest and Heart Clinic between 1989 and 1995. Physical examination and laboratory tests such as blood hemoglobin estimation, urine analysis for albumin and sugar, chest X-ray and resting electrocardiogram were carried out. The second set of controls were randomly called during a national rhinitis survey study in 2000 to attend a rhinologic examination in the Ear, Neck and Throat outpatient clinic of the National University Hospital in Singapore. Medical histories were documented by questionnaire. Inclusion criteria for controls were an absence of personal or familial history of CAD, hypertension and diabetes. The mean age of the first collection of male controls analyzed in the study was 45.71 ± 12.98 years and ranged between 10 and 84 years while the mean age of the male controls from the second collection was 34.36 ± 14.25 years and ranged between and 74 years. Controls from the two collections were combined and analyzed together. 49 Chapter Materials and Methods All study participants gave verbal or written informed consent with additional ethics approval when required. 4.3.2 Blood Collection Blood samples were drawn from subjects after a 12-14 hour fast. For cases, it was further required that blood samples were drawn during pre-operation review or at least three months post-myocardial infarction (MI). Plasma was separated from blood cells by centrifugation and aliquots were stored at -20oC until lipid analysis. The remaining blood cell fraction was retained for genomic DNA isolation. 4.3.3 Plasma Lipid and Lipoprotein Measurements Total cholesterol (TC) and triglyceride (TG) levels were estimated using standard colourimetric enzymatic assay kits from Roche Diagnostics on a Cobas Mira autoanalyzer. The cholesterol content of HDL-C was measured after solubilizing ApoBcontaining lipoproteins (LDL-C, VLDL-C, chylomicrons). LDL-C was estimated using Friedewald’s equation (1972): LDL-C = TC – HDL-C –TG/5 where TG < 400 mg/dL. ApoAI and ApoB levels were determined using commercial immunoturbidometric assays (Roche Diagnostics). Lp(a) levels were measured using an enzyme immunoassay kit (Biopool). 4.3.4 SNP Genotyping Table 4.2 summarizes the PCR primer sequences, genotyping methods, electrophoresis conditions for analyzing the SNPs used in the association study. Detailed methods are given in the following sections. 50 Chapter Materials and Methods 4.3.4.1 Restriction Fragment Length Polymorphism (RFLP) Assays The SNPs -14C>T, 237indelG, R219K, V825I, M883I and IVS44+18T>C were genotyped using restriction enzymes. PCR products were incubated overnight in 2-5 U of the appropriate restriction enzyme (New England Biolabs) and 1x manufacturer’s buffer in a final volume of 20 ul at the recommended temperatures. Fragments were resolved using either agarose or native polyacrylacrylamide gels in 1x TBE buffer (89 mM Tris-borate, 20 mM EDTA, pH 8.0) and visualized by ethidium bromide. Representative RFLP gels are provided in Figures 4.3-4.8. 4.3.4.2 Single-Strand Conformation Polymorphism (SSCP) Genotyping Assay for 8895A>G Denaturing loading dye (95% formamide, 0.05% bromophenol blue, 0.05% xylene cyanol blue, mM EDTA) was added directly to PCR products in 2:1 ratio, and the mixture was denatured at 99oC for 3-5 min, snap-cooled on ice and electrophoresed using 20% mini native polyacrylamide gels. The native polyacrylamide gel was prepared from a 50% stock solution composed of 49 parts acrylamide and part bis-acrylamide, and included glycerol at a final 5% (v/v). Prior to sample loading, the gels (dimensions: cm x 10 cm x 0.1 cm) and upper tank buffer were pre-chilled to approximately 4oC. A running temperature of 15oC was maintained by circulating water from a regulated water bath through the electrophoresis unit (Hoefer 240 Mighty Small, Amersham Biosciences). A constant voltage of 200-240 V was used. Using these conditions, the single-stranded DNA bands were sufficiently resolved in about 7-8 hours. Two runs were performed per day on one electrophoresis unit, one in the morning and another in the late afternoon, permitting 80 samples to be genotyped per day per electrophoresis unit. DNA bands were visualized by silver staining. 51 Chapter Materials and Methods Table 4.2 Genotyping of the seven ABCA1 SNPs used in the case-control association study. Forward primer (5’ Æ 3’) Reverse primer (5’ Æ 3’) PCR Size (bp) Genotyping - CGGCTCCACGTGCTTTC CCACTCACTCTCGTCCGCAATTAC 177 BsmA1 2.5% agarose gel C: 177 bp T: 25, 144 bp - GCTGGATTAGCAGTCCTCATTG CCCCAACTCAAAACCACAAAG 301, 302 R219K TGCCTTGTGCTTTGATACATTC CCCAAAAGTCTGAAAGAACACTA 538 2786G>A V825I GGTAGCCCACCACTCCCCTAAAG ATCAGCTGCCTGTCCTTGGACTA 525 Exon 18 2963G>A M883I ATGATGCTGAGCTTGGCTCATAC AGGTCAACAGCACTTACTTTCTGG 171 Intron 44 IVS44+18T>C - AGAATTGAAGGTTTGAGGTAGTTAC TTTATGTGCTTCTTTACCCTTTACT 342 Exon 50 3’UTR 8995A>G - ATGAGAGAACTATTGTTTGGG CTGAAGTCTTACACCTTTAGCG 109 Bsl1 10% polyacrylamide gel 2g: 148, 153 bp 3g: 93, 56, 153 bp EcoN1 2% agarose G: 538 bp A: 242, 296 bp DpnII 2.5% agarose gel G: 62, 423, 40 bp A: 62, 241, 182, 40 bp BsmI 3% agarose gel A: 171 bp G: 150, 21 bp HaeIII 2% agarose gel C: 129, 213 bp T: 342 bp SSCP 20% polyacrylamide + 5% glycerol native gel. 15oC, hours, 200V Location SNP Proximal Promoter -14C>T Exon 2a 5’UTR 237indelG (2g/3g) Exon 969G>A Exon 17 b Amino acid change a Primers from Pullinger et al. (2000). Primers from Wang et al. (2000). A mismatch of G (underlined) is introduced into the lower primer. All PCR products were amplified using Touchdown PCR. b 52 Chapter Materials and Methods C 177 bp A 177 bp 141 bp T 36 bp 141 bp B Figure 4.3 Genotyping of -14C>T using BsmA1 restriction enzyme. (A) Restriction sites and predicted fragment sizes. (B) Following an overnight incubation of PCR products with BsmA1, the fragments were separated on a 2.5% agarose gel. Lanes 1,2,4,6,8,11,12=genotype CC; lanes 3,7,9,10,14=genotype CT; lanes 5,13= genotype TT. 53 Chapter Materials and Methods A 148 bp 2G 301 bp 148 bp 93 bp 3G 153 bp 56 bp 153 bp 302 bp 93 bp 149 bp B Figure 4.4 Genotyping of 237indelG using Bsl1 restriction enzyme. (A) Restriction sites and predicted fragment sizes. (B) Following an overnight incubation of PCR products with Bsl1, the fragments were separated on 10% native polyacrylamide gel. Lanes 1,3=genotype 2g2g; lanes 4-6=genotype 2g3g; lanes 2,7=genotype 3g3g. 54 Chapter Materials and Methods A G 538 bp 538 bp 296 bp A 242 bp 296 bp B Figure 4.5 Genotyping of R219K (G>A) using EcoN1 restriction enzyme. (A) Restriction sites and predicted fragment sizes. (B) Following an overnight incubation of PCR products with EcoN1, the fragments were separated on 2.5% agarose gel. Lanes 1,11=genotype GG; lanes 2-8,10,14=genotype GA; lanes 9,12,13=genotype AA. 55 Chapter Materials and Methods A 62 bp 485 bp G 525 bp 62 bp 423 bp 62 bp 241 bp 62 bp A 303 bp 40 bp 182 bp 40 bp 485 bp B Figure 4.6 Genotyping of V825I (G>A) using DpnII restriction enzyme. (A) Restriction sites and predicted fragment sizes. (B) Following an overnight incubation of PCR products with DpnII, the fragments were separated on 2% agarose gel. Lanes 1-3,5,7=genotype GG; lanes 4,6,8-10=genotype GA; lane 11=genotype AA. 56 Chapter Materials and Methods A A 171 bp 171 bp 148 bp G 23 bp 148 bp B Figure 4.7 Genotyping of M883I (G>A) using Bsm1 restriction enzyme. (A) Restriction sites and predicted fragment sizes. (B) Following an overnight incubation of PCR products with Bsm1, the fragments were separated on 3% agarose gel. Lanes 12,13,16-18=genotype GG; lanes 2,5,7,9,14=genotype GA; lanes 1,3,4,6,8,10,11,15=genotype AA. 57 Chapter Materials and Methods A T 213 bp C 129 bp 342 bp 213 bp B Figure 4.8 Genotyping of IVS44+18T>C using HaeIII restriction enzyme. (A) Restriction sites and predicted fragment sizes. (B) Following an overnight incubation of PCR products with HaeIII, the fragments were resolved on 2% agarose gel. Lanes 1,2,10,12=genotype TT; lanes 3,4,6-9,11=CT; lane 5=genotype CC. 58 Chapter Materials and Methods 4.4 Data Analysis Distributions of quantitative variables traits were explored using summary statistics and also graphically (histograms and boxplots) to identify the shape of the distributions as well as extreme outliers prior to statistical analysis. Prior to formal statistical analysis, quantitative variables displaying skewed distributions (TG and Lp(a)) were transformed by taking the natural logarithm of the observed values so as to induce symmetry and reduce skewness. Extreme outliers were excluded from analysis. Differences in means of quantitative traits between two groups were performed using independent t- tests, and when comparing more than two groups with or without adjustment for covariates, linear regression analysis was applied. Relationships between pairs of quantitative traits were examined visually using matrix scatterplots and also summarized numerically using Pearson coefficient of correlation, r. Body-mass index (BMI) was computed as the weight in kilograms divided by the square of height in metres. Allele frequencies were estimated directly from the observed genotype counts. Exact tests for Hardy-Weinberg equilibrium (HWE) were determined using the method of Guo and Thompson (1992). The pairwise LD metric, D’, was calculated using the LDSHELL program (http://www.iop.kcl.ac.uk/IoP/Departments/PsychMed/GEpiBSt/software.shtml). Significance testing for D’ was also carried out by the program. The second LD metric, r2, was calculated using the D’ values and two-locus haplotype frequencies provided by LDSHELL. Significance testing for r2 was by approximation to the χ2 distribution with one degree of freedom using the formula, χ2~2Nr2, where N represents the number of individuals. Power calculation assuming effect sizes of allelic OR 1.5 or 2.0 was implemented in WinPepi (Abramson and Gahlinger, 2001; http://www.myatt.demon.co.uk/index.htm). 59 Chapter Materials and Methods In general, comparisons of qualitative variables between groups were performed using logistic regression analysis instead of classical contingency table analysis. Logistic regression not only allows significance testing of the association but also estimates the strength of the effect expressed as an odds ratio (OR) as well as has the flexibility to adjust for covariates. In examining single-locus effects on the binary trait CAD, various genotype-based models were implemented. Under the assumption of HWE, the additive model is equivalent to testing for allelic effect using a 2x2 contingency table of allele counts, but unlike the latter, it is robust to departure from HWE (Sasieni, 1997; Schaid and Jacobsen, 1999). In the general genotype model, two ORs are computed, one for the heterozygote and another for one of the two homozygotes, all comparisons relative to a reference homozygous category. The significance of the two ORs in a general genotype model indicates a dominant or recessive allele effect. For a dominant allele effect, both homozygous and heterozygous ORs would vary significantly different from unity, whereas for a recessive allele effect, only the homozygous OR would be significantly different from unity. Alternatively, genotypes could also be recoded accordingly to test for a specific dominant or recessive risk effect. Assuming a locus with C and T alleles and genotypes CC, CT, TT, the covariate coding for the various models are: additive, 0, or reflecting dosage of the T allele; general genotype, 0, or 2, reference category can be either CC or TT; dominant, collapsing CC and CT into single category as and TT as 1; and recessive, collapsing CT and TT into single category coded as 0, and CC as 1. In examining multi-locus effects on association with CAD using unphased genotypes, a preliminary multivariable logistic regression model composed of main effects was first built. The predictor variables considered the seven ABCA1 genetic risk factors as well as the non-genetic risk factors of age, BMI and smoking. CAD was the dependent variable. A backward stepwise strategy was used to select variables. Cutoffs for entry or removal of a variable were set at P values of 0.20 and 0.25 respectively. Note 60 Chapter Materials and Methods that the high P value threshold was used as a screening criterion for variable selection because studies have shown a more traditional level such as 0.05 often fails to identify variables known to be important (Bendel and Afifi, 1977, Mikey and Greenland, 1989). Next, first-order interactions between the main effects were added one at a time to the preliminary main effects model. Interaction terms that caused a significant likelihood ratio test (P[...]... 32 Exon 33 Exon 34 Exon 35 Exon 36 Exon 37 Exons 38-39 Exon 40 Exon 41 Exon 42 Exon 43 Exon 44 Exon 45 Exon 46 Exon 47 Exon 48 Exon 49 Exon 50 Amplicon Size (bp) 528 Exon size (bp) 163 Topt 310 149 55, o 56 C o 60 C 327 125 o 61 C 49 6 99 318 190 55, o 60 C o 61 C 289 95 o 58 C 177 33 o 58 C 270 105 o 58 C 337 75 o 55 C 390 170 o 59 C 344 178 o 60 C 49 4 361 276 1 24 1 94 130 55, o 60 C 54, o 59 C o 60 C... C 342 107 o 55 C 3 64 142 o 55 C 311 135 o 59 C 333 1 04 o 57 C 44 8 93 o 56 C 43 6 245 o 59 C 41 3 141 52, o 57 C 48 Chapter 4 Materials and Methods 4. 3 Genetic Association Study 4. 3.1 Cases and Controls Cases were unrelated patients who had been admitted to the Cardio-Thoracic Surgery Unit of the Singapore General Hospital for coronary bypass graft surgery between 1989 and 2000 Eligibility criterion included... hemoglobin estimation, urine analysis for albumin and sugar, chest X-ray and resting electrocardiogram were carried out The second set of controls were randomly called during a national rhinitis survey study in 2000 to attend a rhinologic examination in the Ear, Neck and Throat outpatient clinic of the National University Hospital in Singapore Medical histories were documented by questionnaire Inclusion... Pullinger et al (2000) Primers from Wang et al (2000) A mismatch of G (underlined) is introduced into the lower primer All PCR products were amplified using Touchdown PCR b 52 Chapter 4 Materials and Methods C 177 bp A 177 bp 141 bp T 36 bp 141 bp B Figure 4. 3 Genotyping of -14C>T using BsmA1 restriction enzyme (A) Restriction sites and predicted fragment sizes (B) Following an overnight incubation of. .. conditions for analyzing the SNPs used in the association study Detailed methods are given in the following sections 50 Chapter 4 Materials and Methods 4. 3 .4. 1 Restriction Fragment Length Polymorphism (RFLP) Assays The SNPs -14C>T, 237indelG, R219K, V825I, M883I and IVS 44+ 18T>C were genotyped using restriction enzymes PCR products were incubated overnight in 2-5 U of the appropriate restriction enzyme (New... 2.5% agarose gel Lanes 1,2 ,4, 6,8,11,12=genotype CC; lanes 3,7,9,10, 14= genotype CT; lanes 5,13= genotype TT 53 Chapter 4 Materials and Methods A 148 bp 2G 301 bp 148 bp 93 bp 3G 153 bp 56 bp 153 bp 302 bp 93 bp 149 bp B Figure 4. 4 Genotyping of 237indelG using Bsl1 restriction enzyme (A) Restriction sites and predicted fragment sizes (B) Following an overnight incubation of PCR products with Bsl1, the... 55 Chapter 4 Materials and Methods A 62 bp 48 5 bp G 525 bp 62 bp 42 3 bp 62 bp 241 bp 62 bp A 303 bp 40 bp 182 bp 40 bp 48 5 bp B Figure 4. 6 Genotyping of V825I (G>A) using DpnII restriction enzyme (A) Restriction sites and predicted fragment sizes (B) Following an overnight incubation of PCR products with DpnII, the fragments were separated on 2% agarose gel Lanes 1-3,5,7=genotype GG; lanes 4, 6,8-10=genotype... Chapter 4 Materials and Methods A T 213 bp C 129 bp 342 bp 213 bp B Figure 4. 8 Genotyping of IVS 44+ 18T>C using HaeIII restriction enzyme (A) Restriction sites and predicted fragment sizes (B) Following an overnight incubation of PCR products with HaeIII, the fragments were resolved on 2% agarose gel Lanes 1,2,10,12=genotype TT; lanes 3 ,4, 6-9,11=CT; lane 5=genotype CC 58 Chapter 4 Materials and Methods 4. 4... effect sizes of allelic OR 1.5 or 2.0 was implemented in WinPepi (Abramson and Gahlinger, 2001; http://www.myatt.demon.co.uk/index.htm) 59 Chapter 4 Materials and Methods In general, comparisons of qualitative variables between groups were performed using logistic regression analysis instead of classical contingency table analysis Logistic regression not only allows significance testing of the association. .. buffer in a final volume of 20 ul at the recommended temperatures Fragments were resolved using either agarose or native polyacrylacrylamide gels in 1x TBE buffer (89 mM Tris-borate, 20 mM EDTA, pH 8.0) and visualized by ethidium bromide Representative RFLP gels are provided in Figures 4. 3 -4. 8 4. 3 .4. 2 Single-Strand Conformation Polymorphism (SSCP) Genotyping Assay for 8895A>G Denaturing loading dye . sequencing. 4. 2.5 Inferring Functional Significance of ABCA1 SNPs The functional importance of ABCA1 variants was inferred by various methods. Putative transcription factor binding sites in the. -515 49 4U -515 to -49 4 ACAAAATGATTGGCGTCCTGA -41 7 40 0U -41 7 to -40 0 CGGAAAGCACGATTTAG PrL -52 to -69 AGCCGCCCACGACACAT +280-+297L +297 to +280 CCGAGCACTTCCCGAAGC PrL PrU -515 49 4U -41 7. Outline of exon-centric SNP discovery in the ABCA1 gene using DHPLC. Chapter 4 Materials and Methods 47 Table 4. 1 PCR primers and DHPLC analysis temperatures for SNP screening in ABCA1