Abstract
Coronary artery disease (CAD) is a complex disease with various components, genetic as well as environmental. Previous reports correlating ALOX5AP gene variants and CAD showed conflicting results depending on the population studied. In this study, we examined the contribution of ALOX5AP genetic predisposition to CAD in a group of CAD patients and controls carefully selected from the Lebanese population. We genotyped SNPs for ALOX5AP variants in 289 catheterized patients aged ≤52 years with >50% stenosis in at least one main coronary artery and 227 catheterized control subjects aged 60 years and above with 0% stenosis. Chi-square (χ2) tests and logistic regression showed no significant difference in the allele and genotype frequencies between the CAD or myocardial infarction (MI) cases and the healthy controls. Haplotype analysis using PHASE showed that the distribution of the risk haplotypes among cases and controls were not significantly different and had no attributable risk to CAD (P = 1.00 and P = 0.5, respectively) or MI (P = 0.2 and P = 0.5, respectively). Our data revealed that ALOX5AP gene variants are not predictors of CAD risk or MI risk among Lebanese patients.
Keywords: CAD, Stenosis, ALOX5AP, SNP, Inflammation
Introduction
Coronary artery disease (CAD) is a condition characterized by a defect in blood supply to the heart. It includes myocardial infarction (MI) and remains the main cause of fatality in many countries (Mackay et al. 2004). Several clinical and environmental factors have been well established as risk factors for CAD. Clinical risk factors include diabetes mellitus, hypercholesterolemia, and hypertension, whereas environmental risk factors comprise smoking, high-fat diet, and obesity (Mackay et al. 2004; Assmann et al. 1999).
Genome-wide association and linkage studies have identified a large number of CAD susceptibility genes, most of which encode factors involved in the inflammatory pathways associated with CAD (Lusis et al. 2004; Robin et al. 2007; Abchee et al. 2006; Yusuf et al. 2004). These include thelymphotoxin-α (LTA), galectin-2 (LGALS2), proteasome subunit α-type 6 gene (PSMA6), and leukotriene A4 hydrolase (LTA4H) (Ozaki et al. 2002, 2004, 2006; Helgadottir et al. 2004, 2006). The ALOX5AP gene encodes the 5-lipoxygenase-activating protein (FLAP) and was found to be significantly associated with CAD and MI (Helgadottir et al. 2004, 2006).
In a study involving 713 MI cases, Helgadottir et al. (2004)identified a haplotype A (HapA), consisting of four SNPs within the ALOX5AP gene region, that was associated with twofold increased risk of MI. In the same study, they identified another ALOX5AP gene haplotype B (HapB) to be significantly associated with MI in a British population (P < 0.001). A replication study on an Italian population revealed an association between HapB and increased risk for CAD (Odds ratio (OR) = 1.67; P = 0.032). Haplotype analyses in this population additionally revealed significant association between CAD and an ALOX5AP haplotype named HapC (OR = 2.41; P = 0.03), present within the same region than that of HapB. HapC was not considered in previous studies; the frequency of HapC population was low (Girelli et al. 2007). Finally, in a recent study, HapB was found to increase the risk for MI in a German population (P < 0.0001) (Linsel-Nitschke et al. 2008).
Other studies, however, showed no correlation between ALOX5AP gene variants and CAD. Two US-based studies revealed lack of association of HapB with acute coronary syndrome (OR = 1.12; P = 0.31) or MI (P = 0.08) (Zee et al. 2006; Morgan et al. 2007). A study in a Japanese cohort provided no evidence for an association of ALOX5AP gene variants and MI. (Kajimoto et al. 2005) Similar results were seen in a North American cohort indicating a lack of association of HapB with stroke (Lohmussaar et al. 2005; Meschia et al. 2005).
This population-specific association of ALOX5AP variants and CAD initiated an investigation of the correlation of these variants with CAD and MI patients in the Lebanese population. Results from this study may help in early identification of subjects at high risk, thus providing greater opportunity for therapeutic applications and prophylaxis in the population.
Materials and methods
Subjects
From a total of 5,000 individuals recruited between 2007 and 2009, 516 Lebanese subjects were selected. These subjects were referred to the cardiovascular care units to undergo cardiac catheterization due to clinical indications and not for the purposes of the study itself at either the American University of Beirut Medical Center or the Rafic Hariri University Hospital in Beirut. Angiography was performed to visualize the left main coronary artery, the left anterior descending artery, the left circumflex artery, and the right coronary artery. All coronary arteries were imaged from different angles. The extent of coronary artery stenosis and the number of diseased arteries were documented for each case. Angiogram results were utilized as the main outcome variable.
Trained research assistants discussed informed consent with patients and obtained permission prior to the angiographic procedure. Data concerning life style factors, demographic variables, medical history, family history, therapeutic applications, and clinical information about subjects were documented. The Institutional Review Board at the Lebanese American University approved the study protocol.
Cases and controls selection
Under the assumption that young patients with CAD will likely carry a lower contribution from lifestyle and environmental factors and reflect a higher genetic contribution, 289 young cases with positive family history of CAD were selected (≤52 years old) . Of these, 288 had all the variables required for adjusted logistic regression estimates of ORs. All 289 were used in unadjusted regressions against the controls.
They have been angiographically confirmed to have a significant coronary artery narrowing (>50% stenosis) in at least one main coronary artery. Among these cases, a subgroup of 55 patients was diagnosed with MI based on their documented medical records. For the statistical analysis, the 289 patients were grouped according to the number of diseased vessels into single vessel disease (SVD) patients and multiple vessel disease (MVD) patients. Only the 288 fully annotated samples were used in MVD vs. SVD analysis.
The control group included subjects that have undergone angiography and were confirmed to be CAD-free (0% stenosis). This is a more stringent selection compared to other studies where controls were either population-based (not necessarily confirmed to be CAD-free) or shown by angiography to have non-significant artery stenosis (<50%) (Ozaki et al. 2002, 2004, 2006; Helgadottir et al. 2004, 2006); however, healthy subjects with minor stenosis may have sub-clinical artery narrowing and may become cases later in life with fully manifested clinical signs (Luo et al. 2007). Taking that into consideration, we have excluded the controls with non-significant coronary stenosis (<50% stenosis) from our study. In addition, we chose control subjects with a much higher mean age than that of the cases (≥60 years old; the difference between the mean ages was 22 years) (Table 1), proposing that such group has a relatively low genetic load.
Table 1.
Characteristics of the study subjects
| Cases (n = 289) | Controls (n = 227) | P values | |
|---|---|---|---|
| Age (years) | 46.7 | 68.65 | <0.0001a |
| Sex (male, %) | 84.8 | 46.7 | <0.0001b |
| BMI (kg/m2) | 29.43 | 29.2 | 0.62a |
| C-smoking (%) | 58.9 | 22 | <0.0001b |
| Hypertension (%) | 45.5 | 64.3 | <0.0001b |
| Diabetes (%) | 24 | 19.4 | 0.213b |
| Hyperlipidemia (%) | 56.6 | 43.2 | 0.02b |
| t-Chol (mg/dl) | 209 | 185 | <0.0001a |
| HDL (mg/dl) | 38.67 | 47.5 | <0.0001a |
| LDL (mg/dl) | 133.3 | 109.4 | <0.0001a |
| TG (mg/dl) | 218.4 | 151.8 | <0.0001a |
| FBS (mg/dl) | 120.4 | 109.3 | 0.025a |
| Uric acid (mg/dl) | 6 | 6 | 0.86a |
In hypertension, blood pressure > 140/90 mm Hg. In diabetes, two fasting glucose consecutive readings > 125 mg/dl. In hyperlipidemia, LDL > 130 mg/dl
aMeans compared by Student’s t test
bPercentages compared by χ2 test
Blood collection and DNA extraction
Venous blood specimens were collected by cardiologists performing coronary angiography; all patients underwent coronary catheterization by Judkins’ technique. A blood sample of 20 ml was collected in a sterile syringe from the femoral artery catheterization site of every patient consenting to enroll in the database. EDTA-treated and plain tubes were placed on ice to be delivered to the laboratory. DNA was extracted from a 5-ml peripheral blood samples using phenol and stored at −70°C until genotyping. Concentrations of DNA samples were determined by ultraviolet spectrophotometry.
Genotyping
Using the NCBI PubMed Entrez database, we selected the four ALOX5AP SNPs (SG13S377, SG13S114, SG13S41, and SG13S35) from which HapB and HapC are derived.
Genotyping was carried out using TaqMan real-time PCR assay. For each reaction, 30 ng of genomic DNA was used, and the end-point PCR was carried out on the ABI primer 7900 HT (Applied Biosystems, Foster City, CA).
Statistical analysis
Statistical analyses reported in Table 1 were performed using SPSS version 16.0 for Windows (Statistical Package for the Social Science, SPSS Inc., Chicago, IL). Discrete variables were compared by the χ2 test. Continuous variables were expressed as means, and significance of quantitative data was assessed using Student’s t test. A P value of a two-tailed P < 0.05 was considered significant and P < 0.01 highly significant.
OR tests were used to compare the Allele frequencies between cases vs. controls. Logistic regressions were performed for all genotype data, treating homozygous and heterozygous states as distinct categories in order to identify heterozygous advantage, if present. They also estimated ORs adjusted for the effects of gender, smoking, hyperlipidemia, and hypertension. Logistic regression tests were also used to compare the genotype frequency between patient groups with SVD and MVD. Haplotype analysis was performed using PHASE v2.1 to estimate haplotype frequencies among cases vs. controls, with OR tests performed for each of those configurations.
Tests were performed using singular value decomposition to implement linear regression using a logit linkage function with standard variance estimators. OR values with 95% confidence intervals (CIs), P values, and statistical test powers were calculated assuming 95% two-tailed CIs to estimate the association between the haplotypes and the clinical outcome (CAD and MI). These computations were performed for unadjusted homozygous and heterozygous states, as well as adjusted to control for significant risk of smoking, gender, hypertension, and hyperlipidemia. ORs with 95% (CIs), P values, and powers were also computed based on allele frequencies for each of the CAD and MI categories. Similar analyses were performed for SVD vs. MVD; however, specific assignment of allele frequency to subjects was not performed, excluding the computation of risk factor adjustments. Results for these computations are displayed in Table 2 for CAD, Table 3 for MI, and Table 4 for SVD vs. MVD.
Table 2.
Genotype and allele frequencies of ALOX5AP SNPs in the study subjects grouped to CAD cases and controls
| ALOX5AP genotype (%) | CAD cases (%) | CAD cases counts (n = 289) | Controls (%) | Controls counts (n = 227) | OR, 95% CI | P value power | Adj. OR, 95% CI | P value power |
|---|---|---|---|---|---|---|---|---|
| SG13S377 | ||||||||
| GG | 73.4 | 212 | 69.6 | 158 | 0.293 | 0.059 | 1.131 | 0.164 |
| 0.06–1.38 | 0.343 | 0.88–1.45 | 0.163 | |||||
| GA | 23.5 | 68 | 29.5 | 67 | 0.219 | 0.028 | 0.861 | 0.178 |
| 0.05–1.05 | 0.475 | 0.63–1.18 | 0.149 | |||||
| AA | 3.1 | 9 | 0.9 | 2 | – | – | – | – |
| – | – | – | – | |||||
| A allele | 14.9 | 15.7 | 0.940 | 0.401 | – | – | ||
| G allele | 85.1 | 84.3 | 0.63–1.41 | 0.044 | – | – | ||
| SG13S114 | ||||||||
| TT | 20 | 58 | 17.6 | 40 | 1.151 | 0.298 | 1.153 | 0.324 |
| 0.68–1.94 | 0.076 | 0.62–2.12 | 0.066 | |||||
| TA | 50.2 | 145 | 52 | 118 | 1.021 | 0.460 | 1.125 | 0.324 |
| 0.68–1.53 | 0.031 | 0.67–1.87 | 0.366 | |||||
| AA | 29.8 | 86 | 30.4 | 69 | – | – | – | – |
| – | – | – | – | |||||
| T allele | 45.1 | 43.6 | 1.063 | 0.367 | – | – | ||
| A allele | 54.9 | 56.4 | 0.75–1.50 | 0.053 | – | – | ||
| SG13S41 | ||||||||
| AA | 74.4 | 215 | 78 | 177 | 0.326 | 0.045 | 1.484 | 0.350 |
| 0.09–1.19 | 0.397 | 0.20–11.0 | 0.058 | |||||
| AG | 21.4 | 62 | 20.7 | 47 | 0.352 | 0.062 | 1.563 | 0.335 |
| 0.09–1.34 | 0.336 | 0.02–12.1 | 0.0.063 | |||||
| GG | 4.2 | 12 | 1.3 | 3 | – | – | – | – |
| – | – | – | – | |||||
| A allele | 85.1 | 88.3 | 0.757 | 0.146 | – | – | ||
| G allele | 14.9 | 11.7 | 0.45–1.27 | 0.182 | – | – | ||
| SG13S35 | ||||||||
| GG | 82.3 | 238 | 83.3 | 189 | 0.245 | 0.101 | 1.132 | 0.190 |
| 0.03–2.11 | 0.247 | 0.86–1.49 | 0.140 | |||||
| GA | 15.9 | 46 | 16.3 | 37 | 0.250 | 0.107 | 0.887 | 0.283 |
| 0.03–2.24 | 0.235 | 0.59–1.34 | 0.083 | |||||
| AA | 1.7 | 5 | 0.5 | 1 | – | – | – | – |
| – | – | – | – | |||||
| G allele | 90.3 | 91.4 | 0.875 | 0.334 | – | – | ||
| A allele | 9.7 | 8.6 | 0.48–1.60 | 0.063 | – | – | ||
Table 3.
Genotype and allele frequencies of ALOX5AP SNPs in the study subjects grouped to MI cases and controls
| ALOX5AP genotype (%) | MI cases (%) | MI cases counts (n = 55) | Controls(%) | Controls counts (n = 222) | OR, 95% CI | P value power | Adj. OR, 95% CI | P value power |
|---|---|---|---|---|---|---|---|---|
| SG13S377 | ||||||||
| GG | 69.1 | 38 | 69.4 | 154 | 0.247 | 0.084 | 0.650 | 0.032 |
| 0.03–1.81 | 0.280 | 0.41–1.02 | 0.459 | |||||
| GA | 27.3 | 15 | 29.7 | 66 | 0.227 | 0.077 | 0.598 | 0.046 |
| 0.03–1.74 | 0.296 | 0.33–1.09 | 0.391 | |||||
| AA | 3.6 | 2 | 0.9 | 2 | – | – | – | – |
| – | – | – | – | |||||
| G allele | 82.75 | 45.5 | 84.25 | 187 | 0.896 | 0.393 | – | – |
| A allele | 17.25 | 9.5 | 30.15 | 35 | 0.41–1.97 | 0.046 | – | – |
| SG13S114 | ||||||||
| TT | 21.8 | 12 | 17.6 | 39 | 1.249 | 0.301 | 0.908 | 0.428 |
| 0.54–2.88 | 0.075 | 0.32–2.55 | 0.038 | |||||
| TA | 47.3 | 26 | 51.4 | 114 | 0.926 | 0.412 | 0.731 | 0.261 |
| 0.47–1.83 | 0.041 | 0.28–1.90 | 0.094 | |||||
| AA | 30.9 | 17 | 31.1 | 69 | – | – | – | – |
| – | – | – | – | |||||
| T allele | 45.4 | 25 | 43.3 | 96 | 1.094 | 0.384 | – | – |
| A allele | 54.6 | 30 | 56.8 | 126 | 0.60–1.98 | 0.048 | – | – |
| SG13S41 | ||||||||
| AA | 65.5 | 36 | 78 | 174 | 0.207 | 0.030 | 0.527 | 0.002 |
| 0.04–1.07 | 0.469 | 0.34–0.82 | 0.812 | |||||
| AG | 29.1 | 16 | 20.3 | 45 | 0.356 | 0.116 | 0.724 | 0.130 |
| 0.07–1.94 | 0.221 | 0.41–1.27 | 0.202 | |||||
| GG | 5.5 | 3 | 1.4 | 3 | – | – | – | – |
| – | – | – | – | |||||
| A allele | 80 | 44 | 88.5 | 196.5 | 0.519 | 0.049 | – | – |
| G allele | 20 | 11 | 11.5 | 25.5 | 0.24–1.13 | 0.378 | – | – |
| SG13S35 | ||||||||
| GG | 89.1 | 49 | 83.3 | 185 | 0.749 | 0.056 | 0.582 | 0.052 |
| 0.52, 0.107 | 0.354 | 0.30–1.12 | 0.370 | |||||
| GA | 10.9 | 6 | 16.2 | 36 | 0.471 | 0.006 | 0.659 | 0.195 |
| 0.26–0.48 | 0.710 | 0.26–1.70 | 0.136 | |||||
| AA | 0 | 0 | 0.5 | 1 | – | – | – | – |
| – | – | – | – | |||||
| G allele | 94.5 | 52 | 91.4 | 203 | 1.622 | 0.225 | – | – |
| A allele | 5.5 | 3 | 8.6 | 19 | 0.46–5.69 | 0.114 | – | – |
Table 4.
Genotype frequencies of ALOX5AP SNPs in CAD cases grouped to SVD cases and MVD cases
| ALOX5AP genotype | SVD (%) | SVD counts (n = 40) | MVD (%) | MVD counts (n = 248) | OR 95% CI | P value power | Adj. OR, 95% CI | P value power |
|---|---|---|---|---|---|---|---|---|
| SG13S377 | ||||||||
| GG | 80 | 32 | 72.2 | 179 | 0.699 | 0.370 | 1.514 | 0.034 |
| 0.08–5.78 | 0.052 | 0.97–3.26 | 0.446 | |||||
| GA | 17.5 | 7 | 24.6 | 61 | 1.089 | 0.470 | 1.715 | 0.050 |
| 0.12–10.0 | 0.030 | 0.90–3.26 | 0.376 | |||||
| AA | 2.5c | 1 | 3.2c | 8 | – | – | – | – |
| – | – | – | – | |||||
| G allele | 88.75 | 35.5 | 84.5 | 209.5 | 0.690 | 0.242 | ||
| A Allele | 11.25 | 4.5 | 15.5 | 38.5 | 0.24–1.95 | 0.104 | ||
| SG13S114 | ||||||||
| TT | 12.5 | 5 | 21.0 | 52 | 2.562 | 0.041 | 1.197 | 0.386 |
| 0.89–7.40 | 0.412 | 0.35–4.04 | 0.047 | |||||
| TA | 45.0 | 18 | 51.2 | 127 | 1.738 | 0.067 | 1.869 | 0.065 |
| 0.84–3.59 | 0.321 | 0.83–4.21 | 0.326 | |||||
| AA | 42.5 | 17 | 27.8 | 69 | – | – | – | – |
| – | – | – | – | |||||
| T allele | 35 | 14 | 46.6 | 115.5 | 1.619 | 0.087 | ||
| A Allele | 65 | 26 | 53.4 | 132.5 | 0.81–3.25 | 0.273 | ||
| SG13S41 | ||||||||
| AA | 72.5 | 29 | 75.0 | 186 | 1.425 | 0.330 | 4.663 | 0.052 |
| 0.29–6.93 | 0.064 | 0.72–30.0 | 0.367 | |||||
| AG | 22.5 | 9 | 21.4 | 53 | 1.309 | 0.377 | 3.256 | 0.122 |
| 0.24–7.07 | 0.050 | 0.44–23.8 | 0.213 | |||||
| GG | 5c | 2 | 3.6c | 9 | – | – | – | – |
| – | – | – | – | |||||
| A allele | 83.75 | 33.5 | 85.7 | 212.5 | 1.161 | 0.374 | ||
| G allele | 16.25 | 6.5 | 14.3 | 35.5 | 0.47–2.89 | 0.051 | ||
| SG13S35 | ||||||||
| GG | 95 | 38 | 80.2 | 199 | 1.076 | 0.392 | 2.076 | 0.022 |
| 0.64–1.82 | 0.046 | 1.02–4.23 | 0.520 | |||||
| GA | 5 | 2 | 17.7 | 44 | 4.521 | 0.001 | 1.059 | 0.461 |
| 1.75–11.7 | 0.874 | 0.32–3.37 | 0.031 | |||||
| AA | 0c | 0 | 2.1c | 5 | – | – | – | – |
| – | – | – | – | |||||
| G allele | 97.5 | 39 | 89.1 | 221 | 0.210 | 0.065 | ||
| A allele | 2.5 | 1 | 10.9 | 27 | 0.03–1.59 | 0.327 | ||
For multiple SNP haplotypes determined by PHASE, we computed ORs with 95% CIs, P values, and powers based on the two-tailed 95% CI for CAD, displayed in Table 5, and for MI, displayed in Table 6.
Table 5.
Comparison of haplotypes percentages among control group and cases
| Haplotype | Population (% ± SD) | Control (% ± SD) | CAD cases (% ± SD) | Odds ratio, 95% CI | P value power |
|---|---|---|---|---|---|
| G-T-A-G | 18.23 ± 0.32 | 19.15 ± 0.45 | 17.84 ± 0.38 | 0.917 | 0.648 |
| 0.59–1.43 | 0.057 | ||||
| G-T-A-A (HapC) | 0.84 ± 0.19 | 0.81 ± 0.25 | 0.85 ± 0.21 | 1.050 | 0.480a |
| 0.15–7.16a | 0.028 | ||||
| G-T-G-G | 10.101 ± 0.32 | 6.81 ± 0.45 | 11.48 ± 0.36 | 1.775 | 0.0372 |
| 0.94–3.33 | 0.430 | ||||
| G-T-G-A | 0.11 ± 0.14 | 0.11 ± 0.21 | 0.12 ± 0.14 | – | – |
| – | – | ||||
| G-A-A-G | 49.55 ± 0.41 | 51.14 ± 0.60 | 48.88 ± 0.47 | 0.914 | 0.694 |
| 0.65–1.29 | 0.073 | ||||
| G-A-A-A | 2.98 ± 0.29 | 1.44 ± 0.49 | 3.62 ± 0.33 | 2.571 | 0.070a |
| 0.73–9.01a | 0.314 | ||||
| G-A-G-G | 2.25 ± 0.35 | 2.35 ± 0.47 | 2.21 ± 0.39 | 0.939 | 0.294 |
| 0.29–3.00a | 0.032 | ||||
| G-A-G-A | 0.10 ± 0.096 | 0.02 ± 0.09 | 0.13 ± 0.13 | – | – |
| – | – | ||||
| A-T-A-G | 9.761 ± 0.36 | 11.67 ± 0.63 | 8.96 ± 0.41 | 0.745 | 0.843 |
| 0.42–1.32 | 0.171 | ||||
| A-T-A-A | 5.10 ± 0.28 | 5.45 ± 0.54 | 4.95 ± 0.29 | 0.903 | 0.600 |
| 0.41–1.97 | 0.044 | ||||
| A-T-G-G | 0.96 ± 0.25 | 1.03 ± 0.41 | 0.93 ± 0.29 | 0.902 | 0.546a |
| 0.15–5.25a | 0.033 | ||||
| A-T-G-A | 0.02 ± 0.09 | 0.02 ± 0.08 | 0.02 ± 0.10 | – | – |
| – | – | ||||
| A-A-A-G (HapB) | 0.01 ± 0.037 | 0.0001 ± 0.01 | 0.012 ± 0.06 | – | – |
| – | – | ||||
| A-A-A-A | 0.003 ± 0.03 | 0.01 ± 0.07 | 0.0017 ± 0.03 | – | – |
| – | – | ||||
| A-A-G-G | 0.001 ± 0.01 | 0.0041 ± 0.04 | 0 ± 0.00 | – | – |
| – | – |
aThese values are inaccurate for small sample sizes
Table 6.
Comparison of haplotypes percentages among control group and the MI cases
| Haplotype | Population (% ± SD) | Control (% ± SD) | MI cases (% ± SD) | OR, 95% CI | P value power |
|---|---|---|---|---|---|
| G-T-A-G | 16.90 ± 0.52 | 17.73 ± 0.59 | 13.47 ± 0.88 | 0.722 | 0.225 |
| 0.32–1.68 | 0.114 | ||||
| G-T-A-A (HapC) | 0.84 ± 0.32 | 1.00 ± 0.38 | 0.15 ± 0.33 | – | – |
| – | – | ||||
| G-T-G-G | 10.21 ± 0.45 | 9.22 ± 0.47 | 14.30 ± 0.96 | 1.643 | 0.134 |
| 0.68–3.96 | 0.197 | ||||
| G-T-G-A | 0.10 ± 0.15 | 0.05 ± 0.11 | 0.33 ± 0.44 | – | – |
| – | – | ||||
| G-A-A-G | 50.80 ± 0.58 | 51.24 ± 0.61 | 48.98 ± 1.06 | 0.914 | 0.381 |
| 0.51–1.65 | 0.049 | ||||
| G-A-A-A | 3.56 ± 0.42 | 3.76 ± 0.52 | 2.73 ± 0.00 | 0.718 | 0.356a |
| 0.12–4.18a | 0.056 | ||||
| G-A-G-G | 1.50 ± 0.43 | 1.33 ± 0.39 | 2.20 ± 1.09 | 1.669 | 0.319a |
| 0.20–14.04a | 0.068 | ||||
| G-A-G-A | 0.14 ± 0.11 | 0.03 ± 0.07 | 0.58 ± 0.44 | – | – |
| – | – | ||||
| A-T-A-G | 11.26 ± 0.51 | 10.83 ± 0.55 | 12.99 ± 0.85 | 1.229 | 0.325 |
| 0.50–3.00 | 0.066 | ||||
| A-T-A-A | 3.32 ± 0.41 | 3.72 ± 0.48 | 1.63 ± 0.41 | – | – |
| – | – | ||||
| A-T-G-G | 1.32 ± 0.37 | 1.02 ± 0.37 | 2.55 ± 0.79 | 2.539 | 0.194a |
| 0.31–21.12a | 0.136 | ||||
| A-T-G-A | 0.03 ± 0.08 | 0.02 ± 0.07 | 0.05 ± 0.20 | – | – |
| – | – | ||||
| A-A-A-G (HapB) | 0.03 ± 0.10 | 0.03 ± 0.10 | 0.05 ± 0.22 | – | – |
| – | – | ||||
| A-A-A-A | 0.01 ± 0.04 | 0.01 ± 0.05 | 0.00 ± 0.00 | – | – |
| – | – | ||||
| A-A-G-G | 0.00 ± 0.02 | 0.00 ± 0.02 | 0.00 ± 0.01 | – | – |
| – | – |
aThese values are inaccurate for small sample sizes
Results
Clinical characteristics and demographic data of the study population are shown in Table 1. Males constitute a significantly higher percentage among cases. Using χ2 test for categorical variable and student t test for continuous variables, the cases and controls groups were comparable for diabetes (OR, 1.31), BMI, and uric acid.
Cases showed significantly higher levels of total cholesterol, LDL cholesterol, triglyceride (all highly significant), and fasting blood sugar (FBS; significant). In contrast, lower levels of HDL cholesterol (highly significant) were detected, hence reflecting a role of dyslipidemia as a risk factor for CAD in our population.
Both χ2 and logistic regression tests are constructed assuming sufficiently large samples so that asymptotic approximations hold. In Tables 1–3, for SG13S377, SG13S41, and SG13S35, the hypothesized non-pathogenic state was very rare in our populations. In these cases, error bars on the hypothesized base condition for logistic regression are large. In several such cases, the inversion of the poorly conditioned matrix resulted in the loss of one component, indicating that the weight of the logistic regression came entirely from the heterozygous state vs. the most common homozygous state. In the most extreme regressions, the allele frequency ORs deviated dramatically from the logistic regression estimates of the heterozygous and homozygous ORs. Also, it is clear that sample sizes impacted test resolution when rather dramatic ORs resulted in non-significant results. In Table 4, the estimates of haplotype frequency were often so low that the error bars exceeded the estimates of the population fractions. Also, category sizes were often too small for probabilities to be meaningfully obtained for those ORs. ORs were not computed in those conditions. For Table 4, the population fractions were applied to the number of samples in disease vs. control groups to get numbers for the OR calculations.
Logistic regression analyses resolved no significant difference in the allele and genotype frequencies of the four ALOX5AP SNPs among CAD cases and controls (Table 2). When considering a subgroup of cases with MI and comparing the allele and genotype frequencies between the MI patients and controls, no significant difference was observed (Table 3). There was no significant association between allele frequencies and the various subgroups. The GG and GA genotypes of the SG13S35 SNP and the GG genotype of the SG13S377 SNP appear to be nearly significantly protective against both MI in general, and among MI subjects, in SVD vs. MVD; however, the uncertainty due to small samples among MI AA genotypes for both SNPs may not be ruled out as a possible source of artifact. Subgrouping the CAD cases into SVD cases and MVD cases did not resolve a significant variation in the individual polymorphisms comprising HapB between the two groups prior to adjustment (Table 4), though a trend at the 90% two-tailed level was notable for SG13S377, SG13S114, and SG13S41 after adjustment; however, this result is also vulnerable to uncertainty due to small sample sizes at this level of stratification.
The distribution of HapB between CAD cases and CAD-free controls showed too few samples in this category for standard OR normal-variance approximations to apply. The most frequent haplotype was G-A-A-G (Table 5). The same results appeared in estimation to the association of HapB with MI compared to the control group (Table 6). When CAD cases were compared to controls with no family history of CAD (n = 121), the SG13S377 genotype showed nearly significant association (P = 0.032), where HapB frequency was significantly higher in the controls with no FxCAD (P = 0.021) (Data not shown).
The association of HapC with CAD was also under-represented in our population, but risk was also seen to be insignificant, and therefore population-specific (Girelli et al. 2007), and its correlation with CAD in this study population was examined. Table 5 shows that HapC SNPs are not significantly correlated with CAD (OR = 1.050 and P value = 0.480), but we lacked sufficient samples to draw conclusions about MI.
Discussion
Inflammation plays an essential role in the pathogenesis of CAD. Several studies reported the contribution of the leukotriene pathway to atherosclerosis and CAD (Lotzer et al. 2005). Consequently, we have chosen ALOX5AP, which has been previously implicated in the inflammatory pathway, to examine in the Lebanese population.
The difference in allele, genotype, and HapB frequencies for the variants under study between the various populations studied may be attributed to the differences in ethnic background leading to different SNP linkage disequilibrium and haplotype structures. This difference was clearly observed in the Japanese population where a number of HapB frequencies were too low for performing an association analysis, and the authors relied on other haplotypes to detect the association of ALOX5AP gene with MI (Kajimoto et al. 2005). The same situation exists in our population as well, as was noted extensively in the “Results” section.
Evidence to the effect of ethnic differences on the genetic susceptibility to CAD is also found in the LTA4H gene (encoding the LTA4 hydrolase), which increases the risk for MI depending on the ethnic background of the subjects under study (Helgadottir et al. 2006). Moreover, the previously published reports included different study subjects with different categorization criteria and different phenotypes (MI, CAD, or stroke), which complicated the comparison between different studies.
In our study population, we resolved no difference in the genotype frequency of the polymorphisms comprising HapB when we compared SVD cases and MVD cases. This result again provides no evidence for the role of HapB in the disease severity. While the accumulating evidence on the role of inflammatory mediators in the initiation, progression, and severity of the disease builds(Laxton et al. 2005; Lusis et al. 2004), the impact of the inflammatory mediator FLAP is difficult to assess due to the strong nearly monoallelic structure of our population.
Our selection for the subjects was based on stringent inclusion criteria based on measured clinical outcomes (angiography results) for the purpose of genetic enrichment. Therefore, the results variability between our study and others may have been due either to less stringent selection of cases and/or improper categorization of the controls.
The primary indicator of disease was catheterization. Subject recruitment was therefore among those for whom the procedure had been ordered for other indications, rather than by random selection of volunteers since the procedure is invasive. Therefore, proportions among interactive variables may be distorted by selection bias, which constitutes a limitation to this study. Specifically, the selection criteria for ordering the procedure includes variables that may themselves be indicators of inflammatory response, which would render the control and disease groups to be artificially homogeneous and enriched for inflammatory alleles. Furthermore, history of conditions identified as risk factors included neither age of onset and detailed clinical history nor details of medication, treatment, and other factors; however, blood panels measuring FBS and lipids, as well as BMI and other measures, were obtained. We sought to limit the impact of the above-sited factors by selecting patients with earlier onset (younger than 52 years of age); however, this also limits the probe to genetic factors that promotes disease process at earlier ages, rather than those that may inhibit CAD among some older patients. This also implies that the set of patients selected would be from among those whose risk factors may have identified the subject as being at sufficient risk to justify catheterization at a younger age (e.g., family history) that may also impact sampling bias by genetic association.
Finally, the susceptibility HapB is constructed from polymorphisms of no clear effect on the function or the synthesis of the FLAP. Moreover, these variants might be associated with disease-causal SNPs in ALOX5AP gene, and this association might vary between populations according to the genetic structure of each race. Thus, other haplotypes in ALOX5AP gene may be associated with the disease.
Our study is the first genetic study on the association of the inflammatory gene ALOX5AP with CAD that includes cases and controls with detailed clinical and epidemiological criteria of categorization. This is also the first study on the ALOX5AP gene in Middle Eastern populations.
Acknowledgment
We thank the patients for agreeing to participate in this study. We thank Hana Sbeite and Nour Moukalled for their help with the subjects’ recruitment and data collection. We thank May Sanyoura for her help with sample management and Dora Mouallem for her help with manuscript preparation. We thank the American University of Beirut Medical Center and the Rafic Hariri University Hospital for their collaboration and support. This work is supported by the European Commission (FGENTCARD, LSHG-CT-2006-037683).
References
- Abchee A, Puzantian H, Azar ST, Shbaklo H, Nasrallah A, Sawaya FJ, et al. Predictors of coronary artery disease in the Lebanese population. Thromb Res. 2006;117:631–637. doi: 10.1016/j.thromres.2005.05.022. [DOI] [PubMed] [Google Scholar]
- Assmann G, Carman R, Cullen P, Fruchart JC, Jossa F, Lewis B, et al. Coronary heart disease: reducing the risk: a worldwide view. Int Task Force Prev Coron Heart Dis Circ. 1999;100:1930–1938. doi: 10.1161/01.cir.100.18.1930. [DOI] [PubMed] [Google Scholar]
- Girelli D, Martinelli N, Trabetti E, Olivieri O, Cavallari U, Malerba G, et al. ALOX5AP gene variants and risk of coronary artery disease: an angiography-based study. Eur J Hum Genet. 2007;15:959–966. doi: 10.1038/sj.ejhg.5201854. [DOI] [PubMed] [Google Scholar]
- Helgadottir A, Manolescu A, Thorleifsson G, Gretarsdottir S, Jonsdottir H, Thorsteinsdottir U, et al. The gene encoding 5-lipoxygenase activating protein confers risk of myocardial infarction and stroke. Nat Genet. 2004;36:233–239. doi: 10.1038/ng1311. [DOI] [PubMed] [Google Scholar]
- Helgadottir A, Manolescu A, Helgason A, Thorleifsson G, Thorsteinsdottir U, Gudbjartsson DF, et al. A variant of the gene encoding leukotriene A4 hydrolase confers ethnicity-specific risk of myocardial infarction. Nat Genet. 2006;38:68–74. doi: 10.1038/ng1692. [DOI] [PubMed] [Google Scholar]
- Kajimoto K, Shioji K, Ishida C, Iwanaga Y, Kokubo Y, Tomoike H, et al. Validation of the association between the gene encoding 5-lipoxygenase-activating protein and myocardial infarction in a Japanese population. Circ J. 2005;69:1029–1034. doi: 10.1253/circj.69.1029. [DOI] [PubMed] [Google Scholar]
- Laxton R, Pearce E, Kyriakou T, Ye S. Association of the lymphotoxin-alpha gene Thr26Asn polymorphism with severity of coronary atherosclerosis. Genes Immun. 2005;6:539–541. doi: 10.1038/sj.gene.6364236. [DOI] [PubMed] [Google Scholar]
- Linsel-Nitschke P, Gotz A, Medack A, Konig IR, Bruse P, Lieb W, et al. Genetic variation in the arachidonate 5-lipoxygenase-activating protein (ALOX5AP) is associated with myocardial infarction in the German population. Clin Sci (Lond) 2008;115:309–315. doi: 10.1042/CS20070468. [DOI] [PubMed] [Google Scholar]
- Lohmussaar E, Gschwendtner A, Mueller JC, Org T, Wichmann E, Hamann G, et al. ALOX5AP gene and the PDE4D gene in a central European population of stroke patients. Stroke. 2005;36:731–736. doi: 10.1161/01.STR.0000157587.59821.87. [DOI] [PubMed] [Google Scholar]
- Lotzer K, Funk CD, Habenicht AJ. The 5-lipoxygenase pathway in arterial wall biology and atherosclerosis. Biochim Biophys Acta. 2005;1736:30–37. doi: 10.1016/j.bbalip.2005.07.001. [DOI] [PubMed] [Google Scholar]
- Luo AK, Jefferson BK, Garcia MJ, Ginsburg GS, Topol EJ. Challenges in the phenotypic characterisation of patients in genetic studies of coronary artery disease. J Med Genet. 2007;44:161–165. doi: 10.1136/jmg.2006.045732. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lusis AJ, Mar R, Pajukanta P. Genetics of atherosclerosis. Annu Rev Genomics Hum Genet. 2004;5:189–218. doi: 10.1146/annurev.genom.5.061903.175930. [DOI] [PubMed] [Google Scholar]
- Mackay JH, Powell SJ, Charman SC, Rozario C. Resuscitation after cardiac surgery: are we ageist? Eur J Anaesthesiol. 2004;21:66–71. doi: 10.1017/s0265021504001115. [DOI] [PubMed] [Google Scholar]
- Meschia JF, Brott TG, Brown RD, Jr, Crook R, Worrall BB, Kissela B, et al. Phosphodiesterase 4D and 5-lipoxygenase activating protein in ischemic stroke. Ann Neurol. 2005;58:351–361. doi: 10.1002/ana.20585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Morgan TM, Krumholz H, Lifton RP, Spertus JA. Nonvalidation of reported genetic risk factors for acute coronary syndrome in a large-scale replication study. Jama. 2007;297:1551–1561. doi: 10.1001/jama.297.14.1551. [DOI] [PubMed] [Google Scholar]
- Ozaki K, Ohnishi Y, Iida A, Sekine A, Yamada R, Tsunoda T, et al. Functional SNPs in the lymphotoxin-alpha gene that are associated with susceptibility to myocardial infarction. Nat Genet. 2002;32:650–654. doi: 10.1038/ng1047. [DOI] [PubMed] [Google Scholar]
- Ozaki K, Inoue K, Sato H, Iida A, Ohnishi Y, Sekine A, et al. Functional variation in LGALS2 confers risk of myocardial infarction and regulates lymphotoxin-alpha secretion in vitro. Nature. 2004;429:72–75. doi: 10.1038/nature02502. [DOI] [PubMed] [Google Scholar]
- Ozaki K, Sato H, Iida A, Mizuno H, Nakamura T, Miyamoto Y, et al. A functional SNP in PSMA6 confers risk of myocardial infarction in the Japanese population. Nat Genet. 2006;38:921–925. doi: 10.1038/ng1846. [DOI] [PubMed] [Google Scholar]
- Robin NH, Tabereaux PB, Benza R, Korf BR. Genetic testing in cardiovascular disease. J Am Coll Cardiol. 2007;50:727–737. doi: 10.1016/j.jacc.2007.05.015. [DOI] [PubMed] [Google Scholar]
- Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, Lanas F, et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study. Lancet. 2004;364:937–952. doi: 10.1016/S0140-6736(04)17018-9. [DOI] [PubMed] [Google Scholar]
- Zee RY, Cheng S, Hegener HH, Erlich HA, Ridker PM. Genetic variants of arachidonate 5-lipoxygenase-activating protein, and risk of incident myocardial infarction and ischemic stroke: a nested case-control approach. Stroke. 2006;37:2007–2011. doi: 10.1161/01.STR.0000229905.25080.01. [DOI] [PubMed] [Google Scholar]