Coronary artery disease (CAD) is the leading cause of morbidity and mortality worldwide [1]. In Egypt it is responsible for 21% of deaths [2] and reported a prevalence of 8.3% [3]. It is one of the complex diseases that are multifactorial or polygenic disorders, caused by multiple genetic variants with low penetrance in combination with various environmental and lifestyle factors [4]. CAD is characterized by long-term atheromatous plaque formation which culminates into atherothrombotic obstructive lesions leading to tissue damage [5]. Atherosclerosis and thrombosis are the two manifestations underlying CAD. It is known that events such as endothelial dysfunction, inflammation, abnormal lipoprotein and homocysteine metabolism, as well as dysfunctional coagulation and fibrinolysis play a key role in the pathogenesis of CAD [6]. In recent years, several genetic variants robustly associated with CAD have been detected by genome-wide association studies (GWAS), mainly in people of European descent [7]. In order to unravel the genetic profile of a sample of the Egyptian CAD patients with respect to these susceptibility loci, twelve variants were analyzed; Additional file 1: Table S1.

Seven polymorphisms associated with thrombotic risk; (i) Factor V (FV): FV Leiden G1691A (R534Q), leads to activated protein C (APC) resistance on the cleavage of factor V and occurs in 20–50% of patients with venous thromboembolism (VTE). It represents one of the most important genetic risk factors for inherited thrombophilia that increases in homozygous state. The contribution of this polymorphism to the pathogenesis of CAD remains controversial [8, 9] (ii) FVR2 haplotype A4070G (H1299R) was found to be in complete linkage disequilibrium with Factor V Leiden and may confer mild APC resistance. When interacting with the Factor V Leiden mutation could produce a more severe APC resistance phenotype and increased the risk for VTE associated with Factor V Leiden by approximately 3-fold [10]. (iii) Prothrombin, (PTH; Factor II) G20210A; the A allele was previously shown to be associated with increased prothrombin levels due to an alteration in the mRNA processing. Its presence with a hypercoagulable state, imply a moderate association with a higher CAD risk [11, 12]. In addition, A allele carriers had about 3-fold elevated risk for cerebral and deep vein thrombosis. This risk increased significantly when combined with FV Leiden [13]. (iv) Factor XIII (FXIII) 163G > T (V34L); most of clinical studies have demonstrated a lower incidence of both venous and arterial thrombosis in carriers of the mutant allele (34Leu) compared to Val/Val carriers [14]. This could be explained by lower plasma level of FXIII in carriers of L34 with increased its activation that result in less stabilized and thinner fibrin clot due to easy dissociation of its 2A and 2B subunits [15] (v) 5,10-Methylenetetrahydrofolate Reductase (MTHFR) C677T and (vi) MTHFR A1298C. The 677TT homozygosity and 677CT/1298 AC compound heterozygote genotypes have been shown to have decreased MTHFR activity and higher predisposition to arterial and venous thrombosis in the presence of additional risk factors [16, 17]. (vii) Plasminogen Activator Inhibitor 1 (PAI-1, Serpin E1) 4G/5G; the 4G allele was previously found to be associated with higher PAI-1 transcription rates and considered to be a mild risk factor for VTE and myocardial infarction (MI) [18].

Other five variants associated with atherosclerosis; (viii) Beta-Fibrinogen (FGB) -455G > A; confers elevated β-fibrinogen plasma levels, which are directly related to increase the risk of acute coronary syndromes (ACS) [19], premature MI [20] and ischemic stroke [21]. However controversial findings of some researchers that show a protective role of this polymorphism against non-fatal acute MI [22] still present. (ix) Human Platelet Antigen 1 (HPA-1; Gp IIIa; integrin b3) L33P (1a/b); studies have shown that, in the normal population, the HPA-1b phenotype results in increased platelet aggregation and fibrinogen binding, thus enhancing the risk of MI [23, 24]. (x) Angiotensin-Converting Enzyme (ACE) 287 bp insertion/deletion (I/D): D allele has been associated with the elevated ACE activity and plasma levels; thereby represents a risk factor for MI in older patients and in smokers [25].)xi) Apolipoprotein B (Apo B) R3500Q, that is considered one of the most common single site mutations in the human ApoB gene; results in mild to severe hypercholesterolaemia and an increased risk for early onset atherosclerosis [26]. (xii) Apolipoprotein E (Apo E) E2/E3/E4; an important predictor of the plasma lipid profile with E2 shows the lowest and E4 shows the highest LDL (low density lipoprotein) and total cholesterol levels; E4 allele was associated with increased susceptibility to early-onset MI, particularly in smokers [27].

Taken together, this study aimed to assess the presence of previously mentioned mutations and polymorphisms in a sample of Egyptian CAD patients and to correlate the co-existence and combination of these variants with the clinical and laboratory risk factors in attempt to be helpful in refining a risk profile for our CAD patients, in order to determine the need for and the intensity of follow-up or to influence decisions about staged implementation of preventive interventions with behavioral or drug therapy.

To our knowledge, this is the first study conducted to investigate the interaction between twelve atherosclerosis and thrombosis related gene variants and the environmental risk factors that could increase the susceptibility to coronary artery disease in Egyptian CAD patients. The previous studies have confirmed that understanding the gene–environment interactions is very important for the development of novel preventive and therapeutic approaches to reduce the risk of CAD [45‐47].

In the current study FVR was the most variant associated with CAD patients, combined with the FVL variant to increase CAD risk in one of the patient clusters and with both ACE and MTHFR 667C > T in the second cluster. Because CAD is a multifactorial disorder, its genetic components may be a combined effect of a number of genes with each playing only a small role. Butt et al. [48] have reported that the predisposition imparted by individual genes may act independently or interact with other genes to result in an additive effect or a synergistic coeffect. The FV:R2 A4070G variant causes amino acid replacement of histidine for arginine at the codon 1299 in the FV B domain. This polymorphism, which marked by the HR2 haplotype was reported to affect the plasma level of FV and contribute to the APC resistance [49]. Although the homo mutant genotype was absent in CAD patients and controls in the current work, this variant confers high susceptibility to CAD risk under the heterozygous and allelic models. Segers et al. [50] suggested that the reduced level of expression of FV: R2 is the most likely explanation for the enhanced thrombophilic phenotype in R2 carriers. Although there are no conclusive data available that associate the R2 haplotype per se with a relevant increased risk of thrombosis (i.e. taken in consideration that FV: Leiden frequency was comparable between CAD patient and control groups in the current study), it was observed previously that male carriers of the FV: R2 polymorphism, have increased circulating levels of FVIII [51]. Although there was no biochemical explanation has been given for this observation, it has been speculated that altered FV anticoagulant activity of FV: R2 results in diminished the proteolytic control of FVIII levels by the protein C pathway [50]. Since increased levels of FVIII are an independent risk factor for the occurrence of venous thrombosis, this may explain the potential association between FV: R2 and thrombosis. Nevertheless, we found that patients carrying the G allele unexpectedly displayed a mild phenotype (i.e. lower Genseni score and lower number of MVD) than wild-type genotype carriers. It seems therefore that the HR2 haplotype in our population preferentially increases the CAD susceptibility but correlates with mild disease severity. It should be kept in mind, however, that thrombosis is a multifactorial disease in which genetic and acquired risk factors interact dynamically. Therefore, the role that FV plays in the etiology of thrombosis is largely dependent on other factors that influence the impact of any altered FV phenotype [50] and the disease outcome depends on the sum of other multiple interacting factors determine the severity of the disease. This finding should be validated in future larger scale studies to be conclusive.

The mutant homozygote TT genotype of MTHFR 667C > T (rs1801133) was found to be more frequent among the studied CAD Egyptian patients (21.7%) and was not found in the controls. This variant showed significantly higher susceptibility for CAD under homozygote and allelic models and it was one of the most effective associated gene variants in combination with other variants for clustering our CAD patients by multivariate analysis (i.e. the combined occurrence of the MTHFR 677TT, ACE DD and FVR 4070AG genotypes seems to magnify the CAD risk in our population). MTHFR gene is located on chromosome 1p36.3 and contains more than hundred SNPs, the most studied variants are rs1801133 and rs1801131 that included in the current study. The 667C > T transition in exon 5 of the first SNP is resulting in a change of alanine to valine (A222V) near the FAD (flavin-adenine-dinucleotide) cofactor binding region, and the other variant caused by the point mutation of A to C at nucleotide 1298 of coding sequence leads to substitution of glutamine by alanine at the amino acid number 470 in the protein structure [52]. In silico tools predicted the former variant (i.e. 667C > T) in exon 5 to be deleterious in contrast to the later neutral one. This could explain why MTHFR 667C > T; was associated with CAD risk in our population as the enzyme with valine was shown previously to be associated with increased thermoliability and reduced enzymatic activity to about 30% if mutant homozygous and 65% if heterozygous [53]. This reduced activity resulting in hyperhomocysteinemia when folate status is low that has been linked to thrombosis and considered as a major independent risk factor for cardiovascular diseases [54]. Similar to our results, when Almawi et al. [55] studied FV 1691 G > A, PT 20210 G > A, and MTHFR 677 C > T gene variants in angiographically documented CAD, they found a strong association of hyperhomocysteinemia and homozygosity of the MTHFR 677 C > T, but not FV: Leiden or PT 20210 G > A mutations with CAD. The results of the present study are also in agreement with Neto et al. [56] who found the MTHFR 677 T/T genotype to predispose to premature atherosclerotic coronary artery disease, and with Yamada et al. [57] and Schürks et al. [58] who reported increased risk for MI and ischemic stroke, respectively. Otherwise, our findings were in contrast to Hsu et al. [59], Yilmaz et al. [60] and Caner et al. [61] who reported that the MTHFR C677T mutation was not associated with the risk of CAD or venous thrombosis among Chinese in Taiwan nor among Turkish patients, respectively. These contradictory results between different studies could be attributed to relatively small sample size, low frequency of gene variants and ethnic heterogeneity [50].

The plasminogen activator inhibitor-1 (PAI-1) is a 52 kDa glycoprotein encoded by the serine protease inhibitor (SERPINE1) gene [62]. It is secreted by endothelial cells and is stored and released from the platelets during activation to downregulate the fibrinolysis process by inhibiting plasminogen to plasmin conversion induced by either tissue-plasminogen activator or urokinase [63]. The current study functional insertion/deletion promoter variant at -675 position (rs1799889), has been associated with increased risk of atherosclerosis and thrombosis previously with inconsistent results [20, 64‐67]. As both the -675 4G and 5G alleles have a binding site for a common transcription activator and the 5G allele has an additional binding site for a repressor, this leads to 5G allele is slightly less transcriptionally active than the 4G and subsequently is associated with lower PAI-1 plasma levels and lower risk of thrombus stabilization [68]. Our results showed no association between PA1 4G/5G and the studied sample of CAD patients. This is in line with a previous study on Egyptian population [69] and others [20, 65, 70, 71] that also showed lack of association of this risk variant with MI or CAD. A number of evidences could elucidate our findings; first, the pathogenesis of CAD is complex and multifactorial with multiple interacting environmental and genetic determinants; hence PAI-1 is influenced not only by the gene SNP, but also by environmental factors, such as blood sugar, insulin and triglyceride concentrations among others [66]. Considering this fact, in subgroup stratified analysis, the 4G/4G genotype has been shown to be associated with obesity under the dominant genetic model to be in concordance with Mertens et al. [72], who found increased PAI-1 levels are associated with increased visceral obesity due to its production by ectopic fat depots [73]. In addition, Fernandes and Sandrim [74], found that obese women with 4G/4G genotypes were at increased risk of thrombotic diseases. The 5G/5G genotype in our cases has shown to be protective in patients with positive family history for CAD, to be in line with Margaglione et al. [75]. Second, variability in plasma concentrations of PAI-1 has been reported in different ethnic groups worldwide. The 4G/5G variant has been shown to be the predominant effector in some cases [76‐78], while in others environmental factors such as smoking are involved [79].

Human platelet antigens (HPA), platelet surface receptors, play a key role in the adhesion, activation, and aggregation of platelets. They are formed of two glycoprotein (GP) subunits (GPIIb/IIIa) [80]. The GPIIIa subunit is polymorphic [81]. One of the most studied GPIIIa variant, HPA-1 a/b (c.176 T > C; Lys59Pro), also known as PIA1/A2, has been suggested as an important genetic key player in the CAD pathogenesis [82]. In the current study, our data demonstrated an association between HPA-1 a/b polymorphism and the occurrence of CAD. We found a significant higher prevalence of HPA-1 a/b and HPA b/b genotypes among CAD patients compared to controls. This was consistent with the studies of Weiss et al. [83] and Abboud et al. [84] which reported a strong association between the HPA-1 b allele and genotypes and the risk of CAD. The mutant allele encodes proline instead of lysine at an amino acid position located near the ligand binding site [81]. Bennet et al. [85] showed no impact on ligand binding using static model system. Similarly, our in silico analysis demonstrated L59P mutation to be benign and tolerated. However, in vitro cell culture showed enhanced binding of both fibrinogen and von Willebrand factor (vWF) to the receptor. In the normal population, the HPA-1b phenotype results in increased platelet aggregation and increased fibrinogen binding, thus conferred an increased risk of MI [23, 86], especially in early onset heart disease [83], ischemic stroke [87] and resistance to aspirin [88]. In contrast, other studies reported the absence of any association between the HPA-1a/b missense mutation and the occurrence of coronary diseases [89‐92]. This discrepancy could be explained by the finding of Zotz et al. [93] who reported that HPA-1 a/b is a risk determinant in patients with already existing atheromatous burden, but not for the development of atheromata.

One of the most popular genetic risk factors in CAD is the ACE insertion/deletion (rs1799752) polymorphism [94]. The ACE gene is located on chromosome 17q23.3 with 26 exons and 25 introns spreading over about 2 Kb [95]. The presence (insertion) or absence (deletion) of an Alu repetitive element (287-bp repeat sequence) in intron 16 results in I and D alleles, respectively [92]. This common polymorphism accounts for 47% of variations within ACE gene [94]. The results of the current study showed that DD genotype could be a considerable risk factor for CAD susceptibility, particularly in females. There was an increased frequency of D allele in CAD patients (0.72) compared to controls (0.44). No association of the gene was observed with other cardiovascular risk factors highlighting more strong effect of ACE with CAD. In 1992, Cambien and his colleagues reported the earliest significant relationship between the DD homozygosity and the risk of CAD [96], and that was followed by the study of Morris et al. [97] who suggested that the DD genotype represents a robust risk factor for MI and sudden death compared to II genotype. In addition, DD genotype was associated with coronary artery spasm that is considered one of the mechanisms of MI [98] and increased risk of restenosis with atherosclerotic lesions after percutaneous transluminal angioplasty [99]. Similar findings were further reported in two meta-analysis studies conducted by Samani et al. [100] and Zhou et al. [101]. Furthermore, numerous studies implicated the DD homozygosity as a key risk factor in other cardiovascular diseases, including hypertension [102], coronary and carotid atherosclerosis [103], cardiomyopathy [104], peripheral arterial disease [105], ischemic stroke [106], and recurrent venous thromboembolism [107]. In contrast, Lindpaintner et al. [108] did not confirm the association between D variant and increased risk of CAD in a large population of American males, and this finding was consistent with that of ours as stratification analysis by gender revealed significant association among females only. ACE, the key regulator in the renin angiotensin system (RAS), has two main physiological functions; converts angiotensin I into a physiologically active peptide angiotensin II (Ang II) and degrades bradykinin, a potent vasodilator [105]. Ang II is the atherogenic component of RAS that increases vascular permeability, stimulates proliferation and migration of vascular smooth muscle cells, induces the expression of inflammatory mediators, and enhances the deposition of extracellular matrix [109]. Reduced bradykinin by ACE decreased endothelium-dependent vasodilators nitric oxide and prostacyclin thus controls blood pressure [103]. The circulatory ACE level is genetically determined by the I/D variant [105]. Individuals with DD genotype have a twofold increase in the serum ACE concentration in the normal population, with ID subjects having intermediate levels [110], leading to higher levels of circulating Ang II and vasoconstriction which are common in MI and hypertensive patients [95]. The results of these studies correlate with those of Suehiro et al. [111] which showed an association between D allele and higher expression of the ACE mRNA that may affect the renin-angiotensin system in local regions. The mechanisms underlying this variation in levels of circulating enzymes by the absence of the Alu repeat sequence within the intron in DD genotype still need to be elucidated.

Two apolipoprotein variants were examined in our study; ApoB-100 R3500Q mutation and ApoE (rs7412 and rs429358). ApoB is the primary apolipoprotein of chyomicrons, VLDL and LDL, which is responsible for carrying fat molecules throughout the body to all tissue cells. The ApoB on the LDL particle acts as a ligand for LDL receptors. High levels of ApoB are the primary drivers of atheroma formation [112]. It is well established that ApoB100 levels are associated with coronary heart disease [113, 114]. Mutations in ApoB gene can cause familial hypercholesterolemia [115]. The R3500Q (rs5742904) mutation in ApoB was found to be a major determinant of LDL levels and coronary artery calcification in the Amish population [112]. Substitution of Arginine with Glutamine is believed to prevent proper folding of ApoB and thereby reducing the ability of LDL particle to bind to the LDL receptor [116]. R3500Q mutation was absent in our study population, possible due to low sample size or ethnic variations. This was consistent with population-based surveys which reported low frequency of R3500Q mutation in non-Amish population [117]. The ApoB gene mutations were not detected in Lebanon [118], Russia [119], Iran [26], and Turkey [120].

The other studied apolipoprotein ApoE is found in chylomicrons and intermediate density lipoprotein (IDL) that is essential for the normal catabolism of triglyceride-rich lipoprotein constituents and cholesterol metabolism [121]. The ApoE gene, mapped on chromosome 19q13.32 in a cluster with ApoC1 and ApoC2, is highly polymorphic [122], with 3 major alleles according to the amino acids at positions 130 (rs429358; c.388 T > C) and 176 (rs7412; c.526C > T) in exon 4: ApoE2 (cys130/cys176), ApoE3 (cys130/arg176), and ApoE4 (arg130/arg176) (ensemble.org). The ApoE polymorphisms were found to affect mRNA transcription and result in increased plasma cholesterol and triglyceride with impaired clearance of lipids from the bloodstream [121]. Moreover, knockout mice lacking ApoE developed extreme hypercholesterolemia after a high-fat diet [123]. In the present study, E3 (rs429358*T, rs7412*C) was the predominant form of gene in both patients and controls followed by E2 variant (rs429358*T, rs7412*T) among the study population, E2 carriers showed protection against developing CAD compared to E3 and E4. Both E3/E3 and E3/E4 genotypes had higher disease risk more than three times when compared to E2 genotypes. ApoE variants were also significantly associated with abnormal triglyceride and LDL-c among our CAD patients. Similar to our findings, E3 isoform was the most frequent ApoE variant in several populations, including Americans [124], Indians [125], Turkish [126], and Saudians [127], thus is considered the wild type ApoE genotype. The presence of E4 isoform (rs429358*C, rs7412*C) was previously associated with increased risk of CAD and mortality [128], atherosclerosis [129], and ischemic cerebrovascular disease [130]. Further meta-analyses studies showed ApoE2 allele to be protective while E4 allele rendered higher susceptibility to heart diseases [119, 131‐133]. Although these allelic forms differ from each other by only one or two amino acids at 2 sites, these differences alter the ApoE protein structure and function, and thereby have significant physiological consequences [134]. E2 variant with 2 cystienes binds poorly to LDL cell surface receptor compared to E3 and E4, thus can increase the LDLR number, and lowering cholesterol level [122]. On the other hand, E4 isoform with two arginine amino acids exhibits enhanced transfer from HDL to TG-rich lipoproteins, promoting hepatic remnant clearance by ApoE receptors and decreasing LDLR, thereby increasing circulating cholesterol levels [122]. The same variant has also been associated with increased calcium ion levels and apoptosis following mechanical injury [135].

With the caveat that our sample size for this analysis was low, and the current analyses are likely to have been underpowered, our study confirms that CAD development requires more complex interactions among several risk factors [89], including genetic and environmental one. More specifically, it has been found that five variants; rs1800595 (FVR), rs1801133 (MTHFR), rs5918 (HPA-1), rs1799752 (ACE), and rs7412 and rs429358 (ApoE) were significantly associated with CAD susceptibility under different genetic models in the current sample of Egyptian CAD patients. These variants interacted with different traditional risk factors to cluster the current study population into the patient and control subgroups by multivariate analysis application. However, there are some limitations in our study should be considered. First, the study is cross-sectional for the point of the coronary disease investigation and all the study participants were stable CAD patients. Thus, the clinical phenotype was determined up to the time that the study was performed thereby allowing us to detect an association, but not to predict the outcome. Hence, follow-up studies are highly recommended to estimate the effect of the study risk factors on the disease outcome. In addition, including CAD patients with different clinical presentations are recommended to be useful to differentiate the distribution of the genetic pattern analyzed. Second, it is possible that some of our control subjects could have subclinical atherosclerosis that might result in an underestimation of the true risk associated with certain alleles. Additional studies examining coronary angiography free controls, will be of even greater value to more clearly establish the magnitude of risk imparted by the presence of these genetic risk factors. Third, the lack of an association between some variants and CAD may be due to the relatively small sample size that warrant a larger cohort in the same population and in different ethnicity populations to validate the current results.