ASSOCIATION OF CETP AND ADTRP GENES WITH CORONARY ARTERY DISEASE IN MULTI ETHNIC SINGAPOREAN POPULATION

... consisting of Telugas and Tamils from southeastern India and a minority of Sikhs and Pathans from north India and accounts for 7.9% of the Singapore population The immigrant Chinese and Indians... ethnic Chinese were not different from that in the ethnic Indians, in spite of a much lower CAD incidence in the ethnic Chinese (57) Because of Singapore's multiracial population structure and differential... 76.7% of the resident Singaporean population The Malay population account for 13.9% of the Singapore population (52) Moreover, the Indian migrants in Singapore were from the Indian subcontinent, with

ASSOCIATION OF CETP AND ADTRP WITH CORONARY ARTERY

DISEASE IN MULTI-ETHNIC SINGAPOREAN POPULATION

Acknowledgements

I would like to thank:

A/Prof.Heng and A/Prof.Liu, who have been my great supervisor and supervisor respectively, their love of learning and insistence on perfection has been great sources of inspiration I am thankful to them for lending their trust, attention, patience and support during this research and preparation of this thesis; this would not have been possible without their help

co-I wish to thank from Lye Hui Jen, Huang Ning and June-Mui Goh for teaching the techniques in molecular biology to me, and Ke Tingjing, for helping me with the experiments

I wish to thank from my GIS friends: Foo Jia Nee, Low Hui Qi, Astrid Irwanto, Herty Liany, Zhao Wanting, Rajkumar Dorajoo, and Li Jingmei, who taught me statistical analysis

I wish to thank from National University of Singapore for scholarship and all

of my good friends in Singapore, for their support and encouragement

Naeimeh Tayebi

December 2012

1.5.1 Various genetic epidemiological approaches 15

Chapter 2: Methods 22

3.4.3 Linkage disequilibrium between rs708272 and rs1800775

sites in the three ethnic groups of Singapore

42

3.4.4 Impact of CETP polymorphisms on plasma lipid profile 42

Summary

Singapore’s multi-ethnic population structure and different prevalence and mortality due to CAD in the ethnic groups provide an excellent opportunity to discover further ethological factors CAD In the study leading to this thesis,

two polymorphisms in the Cholesteryl Ester Transfer Protein (CETP), which

were results of a candidate gene approach and the polymorphism of

androgen-dependent tissue factor pathway inhibitor (TFPI) regulating protein (ADTRP),

that was a result of validation of GWAS, were investigated in relation to CAD and plasma lipid levels among the Chinese, Malay and Indian ethnic groups in Singapore The genotypic distribution of all the three polymorphisms was in line with a population at Hardy Weinberg equilibrium (HWE)

The study of two polymorphisms (rs708272 and rs1800775) within the CETP

genes included 662 CAD cases with angiographically confirmed CAD (CAD+) and 927 consecutive individuals from the general population This group is labeled as CAD- since those with a history of heart disorders or with ECG abnormalities were excluded For evaluation of allelic and genotypic

association of rs708272 and rs1800775 with CAD, subjects with dyslipidemia,

hypertension and diabetes were excluded and this subset of CAD- group was

labeled as healthy control group For ADTRP gene (rs6903956), 645 CAD+

and 755 CAD- have been selected We then tested the association of these Single Nucleotide Polymorphisms (SNPs) with CAD and lipid profiles The

B2 frequency of rs708272 was significantly lower in the Malays than the

Indians and Chinese in the healthy control A similar significant trend was

observed for the A allele of rs1800775 polymorphism of CETP Significantly lower B2 and A allele frequencies were observed in the Chinese cases compared to healthy controls The absence of the B2 allele was associated

with CAD with an odds ratio (OR) of 2.0 (95% CI 1.2 to 3.4) after adjustment for the confounding effects of age, cigarette smoking, BMI and gender The

B2 allele of rs708272 and A allele of rs1800775 were significantly associated

with higher plasma HDL-C levels in the male Chinese CAD- after adjusting

for the significant confounding effects of age, BMI and smoking The B2

allele had similar HDL-C elevating effects in the Indian women However, in the combined population, rs708272 showed only highly significant association with plasma HDL-C

The association of rs6903956 within ADTRP gene was also investigated with CAD in the multi-ethnic Singaporean population The risk allele A of

rs6903956 was associated significantly only in the Chinese cases with an OR

of 2.03 (95% CI 1.04-3.96, p-value = 0.037) when analyzed by each ethnic

group separately In a meta-analysis, rs6903956 showed highly significant association with CAD both before (observed p =1.39×10-4; OR= 1.66; 95% CI

1.28-2.15) and after adjustment (p-value = 4.63×10-3; OR =1.86; 95% CI 2.87) for conventional risk factors of age, gender, BMI, smoking status and ethnicity No significant association was observed between rs6903956 genotypes and lipid profiles in the Chinese, Malays and Indians, suggesting

1.21-that the association of this SNP with CAD is not mediated through plasma lipids The limitations of two studies are small sample groups especially for the Malay and Indian populations Therefore, further analysis in larger numbers of cases and controls will help to confirm these results

List of Tables

1.5.1 The 36 loci associated with CAD/MI at Genome-Wide Levels

of statistical significance

18

3.1 Previous studies of polymorphisms in CETP gene 34

3.4.2 Genotype and allele frequencies of the rs708272 and rs1800775

promoter polymorphisms in the three ethnic groups of Singapore

48

3.4.2a The P-values between different Genotypes of the rs708272 and

rs1800775 promoter polymorphisms in the three ethnic groups of

Singapore before and after adjustment for age

49

3.4.3 Linkage disequilibrium coefficients between the rs708272 and

rs1800775 sites in the three ethnic groups of Singapore for CAD

cases and CAD-

50

3.4.4a Genotypic lipid levels (Mean  SD) of the rs708272

polymorphism in the Chinese, Malays and Indian CAD- men

51

3.4.4b Genotypic lipid levels (Mean  SD) of the rs708272

polymorphism in the Chinese, Malays and Indian CAD- women

52

3.4.4c Genotypic lipid levels (Mean  SD) of the rs1800775 promoter

polymorphism in the Chinese, Malays and Indian CAD- men

53

3.4.4d Genotypic lipid levels (Mean  SD) of the rs1800775 promoter

polymorphism in the Chinese, Malays and Indian CAD- women

54

4.4.1 Clinical and biochemical characteristics of the Singaporean study

population

72

4.4.2a Association of rs6903956 genotypes with CAD 73

4.4.3 Genotypic lipid levels (mean ± SD) of rs6903956 in the Chinese,

Malays and Indians CAD- controls

75

3.4.4 Mean plasma HDL-C levels of the rs708272 and rs1800775

composite genotypes in male Chinese subjects

55

3.4.5 Mean genotypic plasma HDL-C levels in male Chinese

according to A) BMI quartiles and rs708272 genotypes; B) BMI quartiles and rs1800775 genotypes; C) Cigarette smoking and rs708272 genotypes; D) Cigarette smoking and rs1800775 genotypes

56

List of Abbreviations

CAD: coronary artery disease

SNP: Single Nucleotide polymorphism

MI: myocardial infarction

SMC: smooth muscle cell

ApoA: apolipoprotein A

ApoB: apolipoprotein B

CETP: Cholesteryl Ester Transfer Protein

ADTRP: Androgen-Dependent TFPI regulating protein

HDL-C: high density lipoprotein cholesterol

DBP: diastolic blood pressure

SBP: systolic blood pressure

CI: 95% confidence interval

OR: odds ratio

C: degrees in centigrade

min: minute

ml: milliliter

DNA: deoxyribonucleic acid

dNTP: deoxy nucleotide triphosphate

EDTA: ethylene diamine tetra-acetic acid

LDL-C: low density lipoprotein cholesterol

Log: logarithm

Lp(a): lipoprotein(a)

mg/dl: milligram per deciliter

PCR: polymerase chain reaction

RFLP: restriction fragment length polymorphism

Chapter1: Introduction to Coronary Artery Disease

Introduction to Coronary Artery Disease

1.1 Definition of coronary artery disease(CAD)

Coronary artery disease (CAD) is one of the most common diseases related to the vasculature CAD is the predominant cause of disability and death in all industrialized nations (1) It is the narrowing or blockage of the coronary arteries, usually caused by atherosclerosis In more than 90% of cases, the cause for myocardial ischemia is reduction in coronary blood flow because of atherosclerotic coronary lesions Atheromatous plaque consists of cholesterol and fatty deposits on the inner walls of the arteries These plaques can diminish blood flow to the heart muscle by physically clogging the artery or

by causing abnormal artery tone and function Consequently, there is an imbalance between oxygen demand and blood supply

1.2 Prevalence of CAD

About 80% of cardiac mortality is due to CAD Annually, about 5 million individuals are diagnosed as having CAD, resulting in approximately 600,000 deaths every year (1) The frequency of CAD has progressively increased in most industrialized nations to almost epidemic proportions CAD has become more and more prevalent and is one of the principal causes of death in Eastern and developing countries

According to the World Health Organization (WHO), 23.6 million deaths each year will be recorded by 2030 due to CAD (2) In the United States, the total number of individuals affected by CAD is 15.8 million and the annual

Chapter1: Introduction to Coronary Artery Disease

incidence of myocardial infarction (MI) is 565,000 new attacks and 300,000 recurrent attacks, with an annual mortality of 157,000 (3) In United Kingdom, according to British Heart Foundation, 101,000 deaths each year is attributed

to MI and according to Ministry of Health, Labor and Welfare of Japan, the total number of Japanese affect by CAD is 910,000 and nearly 50,000 people die annually from MI However, there are vast differences of CAD morbidity and mortality in different age groups, genders and races Men uniformly have higher incidence than women during their life The CAD mortality in men is also higher (4) The incidence and mortality of CAD also increases with age The data from National Center for Health Statistics showed that the incidence

of CAD dramatically increases in men over 45 years and in women over 55 years The proportion of CAD deaths increased from 12% in men aged 35-44 years to 27% in men aged 65-74 years and from almost 1% to 23% in women between the corresponding age groups (5) Various races often demonstrate varying degree of morbidity and mortality from CAD In the United States, a study revealed that male blacks had less hazard of CAD mortality than male whites, whereas female blacks showed an equal or higher rate than that of their female whites’ counterparts (6, 7) In Trinidad (West Indies), a survey showed that the incidence of the first CAD was 16.4 per 1000 person-years in men of Indian origin, 6.8 per 1000 in men of African origin, 6-5 per 1000 in those with European origin and 2.4 per 1000 in men with mixed origin (8) In Asia, Japanese have been found to have low incidence and mortality for CAD (9)

Moreover, the incidence of CAD in Korea was as low as 8.5 per 100,000 in

1989 (10)

The 1998 National Health Survey showed the level of knowledge among Singaporeans has increased on the benefits of leading a healthy lifestyle including more exercise and less smoking Nonetheless, the levels of high blood cholesterol and hypertension (HTN) among them continue to increase

In addition, the prevalence of obesity and diabetes remained unchanged (11) The proportion of death due to CAD was 31.9%, 22.4% and 17.2% for the Indians, Malays and Chinese respectively (12) Some cross-sectional studies had indicated that some of the established risk factors for CAD, such as total cholesterol (TC), low density lipoprotein cholesterol (LDL-C), HTN and smoking, failed to explain the high excess of CAD in Singaporean Indians (13) However, Indians in Singapore were found to have lower high density lipoprotein cholesterol (HDL-C), Apo protein A1 (apoA1) and raised lipoprotein (Lp (a)) levels compared to the ethnic Chinese (14)

1.3 Pathogenesis of CAD

The most important cause of CAD is diminished coronary perfusion relative to myocardial demand There is presumed to be a complex dynamic interaction among fixed atherosclerotic narrowing of the epicardial coronary arteries, intraluminal thrombosis overlying a disrupted atherosclerotic plaque, platelet aggregation and vasospasm Over 90% of the patients with one of the ischemic

Chapter1: Introduction to Coronary Artery Disease

heart diseases have the advanced coronary atherosclerosis with one or more stenotic lesions causing at least 75% reduction of the cross sectional area of at least one of the major sub pericardial arteries The reduction will block the augmented coronary flow needed to meet even the moderate increases in myocardial demand About one third of the patients have single-vessel, another third have two-vessel, and slightly more than a third have three-major-vessel disease (15) The atheromatous plaque includes a focal plaque within the intima of the vessel, having a core of lipid and a covering fibrous cap The lipid-core includes numerous lipid-loaded foam cells originating from macrophages and smooth muscle cells (SMC), and an extracellular lipid-pool which contains cholesterol, cholesterol ester, Lp(a) (16) and oxidized LDL (17) Other components found inside the plaque include collagen, proteoglycan (18), fibrin (19), T-lymphocytes (20) and complement components (21) Focal rupture or gross ulceration or both of the luminal surfaces of atheromatous plaques may result in exposure of highly thrombogenic substances that induce thrombus formation Hemorrhage into a plaque may also occur from rupture of either the overlying fibrous cap or the thin-walled capillaries that vascularize the plaque (22-24)

1.4 Conventional risk factors of CAD

The risk factors that predispose to the atherosclerosis and resultant CAD have been identified in several prospective studies in well-defined population groups (25) The method of establishing risk factors is one way to explore the

etiology of multifactorial diseases such as CAD Over past several decades, epidemiological studies have provided numerous evidences to establish the following risk factors of CAD

1.4.1 Hyperlipidemia

Hyperlipidemia is acknowledged to be a major risk factor for atherosclerosis Chronic hyperlipidemia may itself impair endothelial function With chronic hyperlipidemia, lipoproteins accumulate within the intima at sites of increased endothelial permeability The oxidative modification of lipid by free radicals generated in the macrophages or endothelial cells in the arterial wall yields oxidized LDL (26) Oxidized LDL is cytotoxic to endothelial cells and SMCs and is also immunogenic, including the production of antibodies to oxidized lipoproteins (27)

The major evidence implicating hypercholesterolemia in the genesis of atherosclerosis includes the following:

- Genetic polymorphism defects in abundant genes involved in cholesterol metabolism contribute to atherosclerosis susceptibility such as

Cholesterylester transfer protein (CETP) gene variations

- There is a major decline in the rate of progression of the disease when the levels of serum cholesterol are lowered by diet or drugs, resulting in the lowering of the risk for CAD (28).Cholesterol lowering increases overall

Chapter1: Introduction to Coronary Artery Disease

survival risk and reduces risk of atherosclerosis related events in patients with established CAD with elevated (29) or average (30) cholesterol levels as well

as in patients with hypercholesterolemia but without overt related disease (31)

1.4.3 Cigarette smoking

Cigarette smoking is a well-established risk factor in the development of CAD A number of epidemiological studies have identified smoking as a

major risk factor for CAD (33) Animal and in vitro experiments showed that

the endothelium was injured due to smoking (34) Furthermore, cessation of smoking was found to be correlated with decreased CAD risk (35)

1.4.4 Diabetes Mellitus

It has long been noticed that diabetic patients have much higher risk for CAD Some studies showed that the CAD mortality accounted for more than one third of total deaths in diabetic patients (36) Although diabetic patients often have other CAD risk factors such as obesity, HTN, hyperlipidemia and elevated fibrinogen, diabetes mellitus still remained an independent risk factor for CAD after these risk factors were taken into account (37)

1.4.5 Coagulation factors

The markers for thrombotic function include fibrinolysis and inflammation such as elevated levels of plasminogen activator inhibitor-1, plasma fibrinogen and C reactive protein (CRP) Studies have shown that CRP levels predict the risk of MI (38) Other factors associated with CAD risk include lack of exercise, lifestyle, weight gain etc Epidemiologic studies also indicate a protective role for moderate intake of alcohol

1.5 Genetic factors

It is well known that genetic factors play an important role in the etiology of CAD and contribute to the individual's susceptibility or resistances to the disease The biological complexity of atherosclerosis implies that approximately 40% - 60% of unexplained variation in cardiovascular disease risk can be attributed to genetic factors; hence it is necessary to identify these specific molecular and genetic determinants (39-43)

Chapter1: Introduction to Coronary Artery Disease

1.5.1 Various genetic epidemiological approaches

One of the genetic epidemiological approaches to determine genetic variants is based on candidate genes studies This technique has reported multiple positive findings for cardiovascular phenotypes With this technique, we can find association, provided information about the functionality in the pathway

of genes, which have been tested and are believed to have a role in the biological pathway of the trait Furthermore, this method takes advantage of the increased statistical efficiency of association analysis compared to a genome wide association study However, candidate gene studies only tested one to a few variants for association with CAD and these approaches cannot discover unknown novel variants and also cannot evaluate how strong each variant contribute to the susceptibility to CAD (44)

In parallel with candidate gene studies, other strategies such as genome wide linkage analysis were carried out to interrogate the entire human genome and

it is based on the Mendelian cosegregation of a genetic marker within a family However, great efforts had to be made in order to collect sufficient numbers of affected sibling pairs The application of this strategy to multifactorial disease is relatively limited (44)

Another approach is linkage disequilibrium mapping, which has proved a valuable tool for locating disease genes Although all existing linkage disequilibrium mapping methods implicitly presume that individual

and haplotypes cannot always be uniquely resolved based on genotype data In fact, this method relies on the availability of a set of individual high-risk chromosomes and a set of individual normal chromosomes The genetic compositions of high-risk chromosomes such as haplotypes across a set of genetic markers are inferred using the genotypes of the affected individuals and the genotypes of their relatives The haplotypes on the normal chromosomes can be constructed either using chromosomes that are not associated with the disease in the same set of pedigrees or unrelated normal chromosomes in the same population (45)

With improved genotyping technologies and the completion of the human HapMap project, genome wide association studies (GWASs) have gained wide acceptance due to its comprehensive coverage of much of the human genome (46) Over the last 6 years, GWAS has proved to be one of the most successful approaches to study coronary atherosclerosis disease and have led to discovery

of some novel cellular genetic loci Since the identification of the 9p21.3 locus

in 2007, which is so far the most robustly replicated genetic signal for CAD around the world, several other loci have been identified by the Welcome Trust Case Control Consortium (WTCCC) and the MI Genetics Consortium Recently, using GWAS, 36 loci (1p13, 1q41, 2q36, 3q22, 6p24, 6q25, 9p21, 10q11, 12q24, 15q22 and so on) of novel single nucleotide polymorphisms (SNPs) associated with CAD and MI have been identified ( Table 1.5.1) and many of them have been validated in other large Caucasian replication studies

Chapter1: Introduction to Coronary Artery Disease

(47) In fact, one major limitation of GWAS is a high rate of false positives; (48) thus, it is essential that rigorous replication studies in many independent populations from different research groups replicate the findings and conform

to certain standards to ensure proper validation of identified loci (49) Studies show that almost all the currently known CAD and MI loci were discovered in the European populations Although the 9p21.3 locus has been replicated in the Asian populations (47), it is important to validate other identified novel loci of European GWAS in the Asian cohort as well Variants identified by GWAS may not be causal but in strong linkage disequilibrium with the causal ones that are located in genes or functional regions nearby Fine mapping will

be needed for identifying causal variants and to derive more detailed information about the region identified by GWAS analysis In order to achieve fine mapping, deep sequencing of the surrounding region is often done In

addition, in vitro and in vivo studies using transgenic animal models have

identified or confirmed functional genes, creating potential new therapeutic targets for decreasing the prevalence of heart disease (50, 51)

Table 1.5.1 The 36 loci associated with CAD/MI at Genome-Wide Levels of Statistical Significance

PCSK9 MIA3 WDR12 TTC32-WDR35 MRAS GUCY1A3 ANKS1A C6orf10-BTNL2 PHACTR1 C6orf105 TCF21 LPA BCAP29 ZC3HC1 CDNK2A/B/CDK2B-AS1 DAB2IP

ABO KIAA1462 CXCL12/HNRNPA3P1

LIPA CYP17A1/CNNM2/NT5C2 PDGFD/DYNC2H1 ZNF259/APOA5/4/1/APOC3 ATP2B1

SH2B3 COL4A1/COL4A2 HHIPL1 MORF4L1/ADAMTS7/RPL21P116 ADAMTS7

RASD1/SMCR3/PEMT/RAL1 SMG6/SRR

UBE27/GIP/ATP5G1/SNF8 LDLR/SMARCA4 SLC5A3/MRPS6/KCNE2/C21orf82

0.78(A) 0.91(A) 0.82(T) 0.74(C) 0.15(C) 0.39(T) 0.18(C) 0.76(T) 0.75(G) 0.59(G) 0.67(C) 0.10(A) 0.62(C) 0.02(C) 0.75(C) 0.62(C) 0.46(G) 0.14(A) 0.21(C) 0.43(G) 0.87(C)

0.34(T) 0.89(G) 0.29(T) 0.13(G) 0.39(T) 0.44(T) 0.44(G) 0.43(C) 0.60(C) 0.57(A) 0.56(G) 0.37(C) 0.53(T) 0.77(G) 0.15(T)

1.11(1.08-1.15) 1.17(1.13-1.22) 1.08(1.05-1.11) 1.14(1.09-1.20) 1.14(1.09-1.19) 1.25(1.12-1.39) 1.12(1.07-1.16) 1.23(1.11-1.37) 1.07(1.05-1.10) 1.17(1.07-1.27) 1.10(1.06-1.13) 1.51(1.34-1.70) 1.08(1.06-1.10) 1.51(1.33-1.70) 1.10(1.05-1.15) 1.09(1.07-1.12) 1.29(1.23-1.36) 1.09(1.04-1.15) 1.10(1.07-1.13) 1.10(1.06-1.13) 1.09(1.07-1.13)

1.08(1.05-1.12) 1.12(1.08-1.16) 1.09(1.05-1.13) 1.13(1.10-1.16) 1.21(1.11-1.33) 1.07(1.04-1.10) 1.07(1.05-1.09) 1.07(1.05-1.10) 1.07(1.03-1.11) 1.08(1.06-1.10) 1.07(1.05-1.09) 1.07(1.05-1.09) 1.06(1.04-1.08) 1.14(1.09-1.18) 1.18(1.12-1.24)

Lipoprotein and cholesterol metabolisms Unknown

LDL metabolism Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Lipoprotein metabolism Unknown

Unknown Increased proliferation of smooth muscle cells Unknown

Unknown Unknown Neointima formation after arterial injury, platelet activation

in atherosclerotic lesions Lipoprotein metabolism Unknown

Unknown Unknown Unknown Unknown Unknown Unknown Neointima formation, vascular remodeling Unknown

Unknown Unknown Lipoprotein metabolism Unknown

Unknown

Chapter1: Introduction to Coronary Artery Disease

1.6 Background of present study

Singapore is a relatively young country with the Chinese, Malay and India immigrants (12) The Chinese community consists mainly of descendants of Han Chinese settlers from the southern provinces of China, such as Fujian and Guangdong, and currently represents the dominant racial population in Singapore, accounting for 76.7% of the resident Singaporean population The Malay population account for 13.9% of the Singapore population (52) Moreover, the Indian migrants in Singapore were from the Indian subcontinent, with the majority consisting of Telugas and Tamils from southeastern India and a minority of Sikhs and Pathans from north India and accounts for 7.9% of the Singapore population

The immigrant Chinese and Indians have lived together with native Malays for the past two to three generations, with shared socio-economic environment and to some extent integrated dietary habits However, their incidence and mortality of CAD, as mentioned above, are markedly different The Singaporean Indians has the highest mortality from CAD in both men and women (49) As mentioned earlier, overseas Indians around the world have been widely reported to have higher CAD death rate compared to their compatriots (53) It has been observed that Indians have low levels of HDL-C (54) The high prevalence of diabetes mellitus among Indians is well known (55) These two points seem to be consistent with the high incidence and mortality of CAD in the Indians However, smoking is not always more

prevalent among the Indians (56) Similarly, HTN is not always more common

in the Indians (57) Total cholesterol levels in the Indians were also not higher than those of the native people or migrants with other origins (54)

In Singapore, the ethnic Indians showed no difference in cholesterol level compared to the Chinese and Malays (58) In addition, studies in Singapore and Malaysia showed that the Indians had high prevalence of CAD and this pattern keeps uniform at any age group The ethnic differences in CAD risk could not be explained by conventional CAD risk factors like HTN, smoking, hypercholesterolemia and even diabetes mellitus (58)

In the study of Lau et al (1993), the lipid profile in Hong Kong Chinese

population, which had a lower prevalence of CAD compared with Caucasians, was studied to determine whether lipid-modifying disease prevention programs were necessary Indices of obesity, TC, TG, Apo A-I, and Apo B concentrations were measured The lipid profile was similar to that of the North American population, with an overall prevalence of LDL-C > or = 4.1 mmol/L of 21% and 11% in men and women respectively The high prevalence of subjects with undesirable lipid values in this Chinese population, which at present had a CAD prevalence of one-eighth to one-quarter compared with that in Caucasians, suggested that the incidence of the disease may rise in the future to pose a similar problem to that in Caucasian populations Cut off values of lipids conferring increased cardiovascular risks need to be determined in this population, so that public health strategies may

Chapter1: Introduction to Coronary Artery Disease

be formulated (59) In Singapore, the total cholesterol levels in the ethnic Chinese were not different from that in the ethnic Indians, in spite of a much lower CAD incidence in the ethnic Chinese (57)

Because of Singapore's multiracial population structure and differential incidence and mortality of CAD in those ethnic groups, an excellent opportunity is provided to explore further the etiological factors of CAD Therefore, our main aim in this thesis is to do research in order to understand the genetic basis of the susceptibility of cardiovascular disease in the multi-ethnic Singaporean population Hence, the specific aims are to study two genes in order to figure out whether there was any association between these genes with CAD and lipid levels, one gene was a result of candidate gene approach and the other was a validation of GWAS The first gene based on

candidate gene approach was Cholesteryl Ester Transfer Protein (CETP), in

which two polymorphisms (rs708272 and rs1800775) of this gene were studied The second gene was Androgen-dependent tissue factor pathway

inhibitor (TFPI) regulating protein (ADTRP) based on GWAS and the SNP

rs6903956 within this gene on chromosome 6p24.1 was validated in the ethnic Singaporean population

multi-Chapter 2

Methods

Chapter2: Methods & statistical Analysis

2.1 Record of demographic information

After obtaining opinion consent, all the subjects from 1995 and 2002 at Singapore’s National Heart Centre were asked to fill in a questionnaire form, which includes the following items: age, height, weight, race, smoking history, medical history, and family history The sample of the questionnaire form is presented in the Appendices

Weight was measured by an electronic weight and height machine Height was measured by the same machines with which weight was measured The reading was taken to the nearest centimeter (cm) with subjects standing straight without shoes Body Mass Index (BMI) was estimated by the formula: BMI = Wt (Kg)/Ht2 (m2) The systolic and diastolic blood pressures were recorded after the subjects were asked to rest for at least 30 min A 12-lead electrocardiogram (ECG) was performed Chest X-ray was also taken

2.2 Blood sampling

All the subjects were requested to fast overnight (at least 10 hr) Blood was collected between 8 - 10 am after subjects had been seated for at least 30 min Blood samples were collected by direct venipuncture in vacutainer tubes [5 ml

in plain, 3 ml in EDTA, 4 ml in 3.8% trisodium citrate (9 parts of blood with 1 part of coagulant)] using multisampling device with minimal stasis Citrated plasma was separated within one hour Separation was carried out at 2800 rpm (revolutions per minute) for 15 min at room temperature in a centrifuge and plasma was divided into three vials for the separate coagulation factor assays

The plasma was snap-frozen in liquid nitrogen and then stored at -80 0C until analyzed At the same time, blood was also separated by centrifuge and the serum was divided into three tubes The supernatant was obtained and two aliquots of serum were stored at -70 0C until analyzed Packed cells were stored at -20 0C for the extraction of genomic DNA

2.3 Measurement of blood lipids

Serum Precipitation for HDL-C Estimation:

Serum was first precipitated by a phosphotungstic acid/magnesium chloride kit Briefly, 500 µl of the reagent was added into 200 µl of serum The mixture was allowed to stand at room temperature for at least 10 minutes so that LDL and VLDL would have enough time to precipitate with polyanions and magnesium chloride After centrifuging at 1000 rpm for 5 minutes, the mixture was separated and the supernatant was used for the enzymatic estimation of HDLC

Enzymatic Estimation for TC, HDL-C and TG:

TC, HDLC and TG were detected by Cobas Mira autoanalyzer using corresponding reagent kits supplied by manufacturers The reaction principle

is based on peroxidase based enzymatic colorimetric assay

Chapter2: Methods & statistical Analysis

were calculated in milligrams per deciliter (mg/dl) Quality control for HDLC,

TC, and TG was verified routinely with the reference laboratory values (Wellcome Quality Assurance Program, UK)

2.4 Extraction of genomic DNA

The extraction of genomic DNA was performed by Phenol/Chloroform Method:

Frozen packed blood cells, from 5 ml of whole blood, were thawed at 37 0C After centrifuging at 35000 rpm, 4 0C for 10 min, the plasma (supernatant) was separated and transferred into 15 ml Falcon conical centrifuge tubes Hypotonic buffer (TE, pH 8.0) was added to the plasma in order to lyse the red blood cells After another cycle of centrifuging at 35000 rpm, 4 0C for 10 min, the supernatant was separated and transferred to a new falcon tube This step (Adding TE buffer and spinning down) was repeated 4 times until the supernatant becomes clear Next step was to lyse the white blood cells by adding 100 µl 10% SDS and protein digesting enzyme 30 µl 10 mg/ml proteinase K This mixture was incubated at 370C overnight

Next day, equal volume of phenol (about 2 ml) was added into the mixture and centrifuged at 35000 rpm for 10 minutes to separate the phenol layer from the aqueous layer The aqueous layer was removed with plastic Pasteur pipettes and transferred into new falcon tube The phenol step was repeated 1 time more

The equal volume of chloroform and isoamyl-alcohol mixture (24 parts chloroform: 1 part isoamyl-alcohol, about 2 ml) was added to the aqueous layer and centrifuged the mixture at 35000 rpm for 10 minutes to separate the organic layer from the aqueous layer The aqueous layer was separated with plastic Pasteur pipettes and transferred to a new tube The chloroform and isoamyl-alcohol step was repeated 1 time as well

Next step, 0.1 volume of 3M sodium acetate and 2.5 volume of cold 100% ethanol were added to DNA solution and slowly inverted the tube in order to mix The solution was then kept in -20 0C for short-term to form DNA Then,

it was centrifuged at 35000 rpm for 10 minutes and discarded the supernatant The DNA pellet was washed using 70% cold ethanol to remove excess salt centrifuged at 35000 rpm for 10 minutes and discarded the supernatant The DNA pellet was dissolved with 100-300 µl of TE buffer and then the concentration of DNA was measured with nanodrop

2.5 Optimization

In order to develop and verify our genotyping technique (primer concentration and temperature annealing) before proceeding to the adult population, cord blood samples were used from neonates born in the National University Hospital, Singapore All mothers of the neonates gave informed consent prior

to collection of their cord blood

Chapter2: Methods & statistical Analysis

Exo I (10U/ul) 0.5ul 1ul

Table 2.6 Components of cycle sequencing reaction

Bright PCR product band Faint PCR product

2) To sequence double-stranded DNA on the GeneAmp® PCR System 9700

3) Ethanol/sodium acetate precipitation of extension products after cycle

sequencing

4) Addition of Hi-Di Formamide for DNA sequence analysis

Finally, the mixture was transferred into 96 well sequencing plate and it was

loaded in the 16-capillary DNA sequencer (ABI PRISM®3100 Genetic

Analyzer) and proceeded to sequence analysis

2.7 Statistical analysis

Distribution of genotypes:

Whether or not the distributions of genotypes of a polymorphism were at

Hardy-Weinberg equilibrium was detected by comparing the frequencies of

genotypes observed with those expected from the Hardy-Weinberg law (p2

+2pq +q2) If the p value was bigger than 0.05, the null hypothesis that

distribution of genotypes for the investigated polymorphism was at

Hardy-Weinberg equilibrium was accepted

Allele Frequencies:

The allele frequencies were estimated by direct gene counting, which, in turn,

reflected the frequencies of specific genotypes Given two alleles A and B, the

frequency of allele A (denoted as P) was calculated as:

P= AA+AB/2

Chapter2: Methods & statistical Analysis

Where AA is the case number of genotype AA, AB is the case number of genotype AB and N is the total case number in the sample

Pearson Chi-square Test:

Pearson chi-square test is often used to test whether two discrete variables are independent or associated The chi-square statistic was calculated by summing the squared residual divided by the expected frequencies in all cells:

X 2 = ∑ (O-E) 2

E

Where O is the observed number and E is the expected number The calculated chi-square is then compared to the critical points of the theoretical chi-square distribution to check how likely or unlikely this calculated value is if the two variables are in fact independent (i.e., not associated) It must be indicated that the chi-square test provides little information about the extent, or form, or direction of the association between the two variables Chi-square value not only depends on the goodness of fit of the independence model but also on the sample size The conditions for chi-square test include (1) samples are obtained randomly; (2) the expected value must not be too small

Logistic regression model:

Association analysis of allele and genotype in each of the three independent ethnic samples was performed using logistic regression with or without adjustment for age, gender, smoking and BMI The association was expressed

as an odds ratio (OR) with 95% confidence interval (CI) In the combined samples, the association analysis was performed using the Cochran Armitage

trend test in PLINK The Breslow-Day test (P-BD) was performed to assess

the homogeneity of the ORs across three different ethnic samples. Power

analysis was carried out using the program implemented in the following web site: http://osse.bii.a-star.edu.sg/calculation2.php

Univariate model:

Before doing univariate analysis, the normality and equality of variance were examined by Kolmogorov-Smirnov test Log10-transformation was thus introduced into the necessary variables such as lipoprotein (a) and triglyceride The association of plasma lipids with genotypes was investigated only in control group using univariate analysis, as the lipid levels of the cases may be affected by treatment with lipid-lowering drugs Confounding factors such as age, gender, smoking and BMI were included as covariates in the univariate analysis In addition, linkage disequilibrium between the rs708272 and rs1800775 sites was indicated by r2

Most statistical analyses were carried out using SPSS version 20 and statistical significance was defined by p-value threshold of 0.05 (two-tailed) in all analyses Cochran Armitage trend and Breslow-Day analyses were performed

using PLINK

Chapter 3: The CETP rs708272 and rs1800775 Polymorphisms

The Cholesteryl Ester Transfer Protein

rs708272 and rs1800775 Polymorphisms

3.1 Introduction

Serum high-density lipoprotein (HDL) levels may be influenced by variety of

factors, which can be broadly categorized as either environmental or genetic

factors The former includes smoking, alcohol consumption and obese status

The latter is primarily composed of polymorphisms in various genes that alter

the metabolic and functional status of coding protein (60), such as ApoB,

ApoE-C1, ApoA5 genes and Cholesteryl Ester Transfer Protein (CETP), and

that will in turn exert influence on plasma HDL and related lipid levels (61)

The observed association between CETP polymorphisms and HDL levels is

primarily attributable to its intrinsic property of transferring triglyceride (TG)

from apolipoprotein B-containing lipid particles to HDL in exchange for

cholesterol ester during the reverse cholesterol transfer process (62) This

results in accumulation of TG on HDL particles and increases the catabolic

rate of apolipoprotein A1 (63), the major component of HDL particles

It is documented that CETP is highly polymorphic and several polymorphic

sites on this gene are associated with fluctuations of plasma CETP activity and

consequently with variations in HDL levels In addition, CETP is considered

to be a key protein in reverse cholesterol transport, a protective system against

atherosclerosis Therefore, CETP gene was selected for this study (64,65)

Taq1B allele is located in intron 1 of CETP gene (rs708272C>T; rs708272B,

minor allele frequency (MAF in Chinese) = 0.44) and -629C>A located in the

CETP promoter region (rs1800775, MAF in Chinese=0.48) are two

Chapter 3: The CETP rs708272 and rs1800775 Polymorphisms

linkage disequilibrium with rs1800775 and it was accepted as an indicator of

the concentration of plasma CETP (65) Clinical studies have demonstrated a

robust association of the rs708272 allele with low CETP activity, decreased

total cholesterol, and increased HDL-C (65)

As LDL supports reverse cholesterol transport to the liver, patients with rare

genetic defects in CETP present with numerous lipid abnormalities (66)

Studies showed that the low CETP levels can associate with increased CAD

risk (67), possibly because of functions other than cholesterol transport

Table 3.1 shows the previous studies regarding the association of the

polymorphisms in CETP gene with CAD and lipid profiles

In this thesis, we reported the allele frequencies and the association of the

CETP rs708272 and rs1800775 polymorphisms with lipid levels and CAD in

the Chinese, Malays and Asian Indians

Table 3.1: Previous studies of polymorphisms in CETP gene

Conclusion Polymorphism(s)

Phenotype Reference

Population

TaqIB &V405 that affect the activity of CETP and plasma levels of

HDL TaqIB & V405

TaqIB & -629C>A were associated with increased risk for CHD/

no link between I450V polymorphism &CHD TaqIB /-629C>A/ I450V

CAD/DM/Lipid profile

79

Iranian

451Q allele, associated with HDL-C/ higher TC and ApoB/

increased risk of significant stenosis B2 allele of Taq1B polymorphism had an increase in HDL-C & associated with a decreased risk of coronary stenosis No significant effect of different A373P and I405V alleles was found on the lipid profile and on coronary stenosis

Taq1B, I405V, R451Q , A373P

Lipid profile /CAD