Introduction

Stroke is the second leading cause of death throughout the world, causing more serious chronic disabilities than any other diseases [1]. Atherosclerosis in carotid arteries represents a risk for ischemic stroke. The currently accepted theory for the development of atherosclerosis postulates that atherosclerosis begins with endothelial injury, making the endothelium susceptible to the accumulation of lipids and the deposition of thrombus [2]. Thus, lipids play a major role in the pathology of atherosclerosis. Elevated triglyceride (TG) levels lead to formation of more dense and, therefore, more atherogenic low density lipoprotein (LDL). In addition, increased serum TG causes production of atherogenic chylomicron and very low density lipoprotein (VLDL) remnants [3]. Transport and redistribution of lipids among various cells and tissues, an important step in lipoprotein metabolism, is carried out by specific apolipoproteins (apo). The major apo include apoE, apoB, apoA-I, apoA-II, apoA-IV, apoC-I, apoC-II, and apoC-III [4].

A new member of the apolipoprotein gene family, apolipoprotein A5 (APOA5), located approximately 30 kb downstream from the apoAI-apoCIII-apoAIV gene cluster on chromosome 11q23, was identified by Pennacchio et al. in 2001 [5]. The protein product of APOA5 gene, APOAV, is a component of several lipoprotein fractions, including VLDL, high density lipoprotein (HDL) and chylomicrons [6]. Although the exact function of APOAV is not known, it was demonstrated that APOA5 gene may play an important role in TG metabolism and is involved in the activation of lipoprotein lipase (LPL) [5, 7]. Functional studies in mice have shown that altering the level of APOAV substantially affects plasma TG levels. Transgenic mice overexpressing the human APOA5 gene exhibit one-third lower plasma TG levels than controls, whereas APOA5 knockout mice have four times as much plasma TG levels, with no effect on plasma cholesterol levels [5]. Recombinant adenoviral vector-mediated transfer of APOA5 into mice was associated with markedly decreased (−70 %) serum TG levels caused primarily by the reduced TG content of the VLDL fraction [7].

There are several single nucleotide polymorphisms (SNPs) on human APOA5 gene (S19W, −1131T>C and G185C), which were shown to be associated with TG levels [5, 8–23]. Increased TG levels in serum may lead to atherosclerosis in the carotid arteries, which is a major cause of ischemic stroke. A possible relation between APOA5 polymorphisms and risk of ischemic stroke was tested before with −1131C allele of APOA5 in Hungarians [18], and in Chinese Han population [23] but the other SNPs of APOA5 were not investigated in this respect.

Since three important APOA5 SNPs namely S19W, −1131T>C and G185C have not been analyzed in relation to ischemic stroke in Turkish population up to now, the present study was aimed to explore the association of these polymorphisms with serum lipid parameters and the risk of ischemic stroke. To our knowledge, this is the first study investigating the relation between APOA5 G185C SNP and ischemic stroke risk.

Materials and methods

Study population

The present sample of 272 unrelated adult Turkish patients with acute hemispheric ischemic stroke and 123 symptom-free Turkish controls was expanded from the population described before by Can Demirdöğen et al. [24]. Blood samples of the participants were obtained from Gülhane Military Medical Academy Hospital Neurology Department, Ankara. Informed consent was obtained from all participants before study entry. The study was approved by the Ethical Committee of the Medical Faculty and was carried out according to the principles of the Declaration of Helsinki.

Cases were selected among patients suffering from atherothrombotic ischemic stroke admitted to the neurology services of the Medical Faculty, Ankara, within 24 h after onset. Recruitment of the patients was performed consecutively. Stroke was defined as clinical designation for a rapidly developing loss of brain functions that lasted at least 24 h and had no apparent cause other than that of vascular origin. The cerebral infarction was initially diagnosed on the basis of neurological examination and brain computer tomography (CT) scan and then transthoracic echocardiographic examination, Holter study and Transcranial Doppler emboli detection procedure to rule out emboli source. In order to be considered eligible, the patients should meet following criteria: having anterior circulation stroke, no other major illnesses, including autoimmune diseases, neoplasms, coagulopathies, hepatic or renal failure, no known embolic source (aortic arch, cardiac or carotid), no family history of hematological, autoimmune or chronic inflammatory diseases, and no history of myocardial infarction within 3 weeks. Our classification system is similar to Trial of ORG 10172 in Acute Stroke Treatment (TOAST). We included TOAST “Large-vessel disease” group and Oxfordshire Community Stroke Project (OCSP) “total anterior” and “partial anterior circulation infarcts” groups into the study.

Control subjects were selected randomly from the neurology outpatient clinics who did not have stroke or transient ischemic attack at any time. All exclusion criteria were applied to the controls exactly plus not having ischemic heart disease, carotid stenosis (lumen narrowing) >70 % or ulcerated carotid plaque. All subjects underwent bilateral carotid Doppler ultrasound (CUSG) and transthoracic echocardiographic studies.

Laboratory analysis

A detailed history of conventional vascular risk factors and conditions from each participant was obtained. Hypertension was defined as systolic blood pressure >140 mm Hg and/or diastolic blood pressure >90 mm Hg and/or use of antihypertensive drugs. Diabetes was defined as fasting glucose ≥6.99 mmol/L and/or use of pharmacological treatment. Obesity was assigned when body mass index was 30 or higher. Smoking status of an individual was assigned “yes” if the individual is currently smoking or have quitted less than 3 months ago. Routine laboratory tests, including electrocardiogram, chest X-ray, complete blood count, leukocyte differential, erythrocyte sedimentation rate, routine biochemistry tests including fasting glucose, lipid profile (TG, total cholesterol, LDL-cholesterol, HDL-cholesterol, VLDL-cholesterol), creatinine, sodium, potassium, bilirubin, and liver function tests, routine urine tests and rheumatologic screening tests were performed for all participants. Standard enzymatic colorimetric methods and commercially available kits were used for determination of serum lipid parameters on auto-analyzer (Olympus AU 2700, Mishima, Japan). Blood samples for lipid tests were taken in the morning after an overnight fast of at least 12 h. When it was not possible to obtain fasting blood due to death of the patient before morning, that subject is not accepted to the study group. Ten milliliters of blood were taken from patients and controls into tubes without anticoagulant and the tubes were centrifuged at 3,000 g for 15 min at room temperature to obtain serum. Blood samples for DNA isolation were taken into Na-EDTA containing tubes and stored at −20 °C until use. All laboratory measurements were done blinded to clinical characteristics.

Genotype determination

DNA isolation was carried out using salting-out method [25]. S19W (rs3135506) SNP of APOA5 gene was genotyped by a TaqMan SNP Genotyping Assay (Applied Biosystems) using Rotor Gene Q real-time polymerase chain reaction (PCR) machine (Qiagen, Hilden, Germany) in allelic discrimination mode. PCR was carried out using the TaqMan Universal PCR Master Mix (Applied Biosystems), TaqMan SNP Genotyping Assay and approximately 20 ng genomic DNA in a 25 μL volume. Thermal cycling program was as follows: an initial melting temperature of 95 °C for 10 min followed by 40 cycles of 92 °C 15 s and 60 °C 60 s. The quality of genotyping was controlled using DNA samples of known genotypes, kindly provided by Dr. Maria-Jose Ariza from Spain.

APOA5 −1131T>C (rs662799) and G185C (rs2075291; c.553G>T) SNPs were determined using standard PCR protocols, followed by restriction enzyme digestions, as previously described [5, 21]. The primer pairs used for the determination of −1131T>C SNP were forward 5′-GAT TGA TTC AAG ATG CAT TTA GGA C-3′ and reverse 5′-CCC CAG GAA CTG GAG CGA AAT T-3′ [5]. The following set of primers was used for G185C SNP: forward 5′-AGA CAC CAA GGC CCA GTT GCT GGG-3′ and reverse 5′-ATG CCG CTC ACC AGG CTC TCG GCG-3′ [21]. PCR mixture in a total volume of 50 μL contained 400 ng genomic DNA, 200 μM dNTPs, 20 pmol of each primer, 1.25 mM MgCl2 and 2.5 U of Taq polymerase for −1131T>C SNP, and 400 ng genomic DNA, 200 μM dNTPs, 20 pmol of each primer, 2.0 mM MgCl2 and 2.5 U of Taq polymerase for G185C SNP. Eppendorf Mastercycler (Hamburg, Germany) was used for PCR. The PCR conditions for the amplification of −1131T>C SNP region consisted of an initial melting temperature of 96 °C for 2 min followed by 32 cycles of melting (94 °C, 15 s), annealing (55 °C, 30 s) and extension (72 °C, 30 s). A final extension step (72 °C, 3 min) terminates the reaction. 187 bp PCR products were digested with 5U MseI at 65 °C for 20 h, and resulted in 167 and 20 bp fragments for the T allele and a non-digested 187 bp fragment for the C allele. Digested products were resolved by gel electrophoresis (2.5 % agarose gel) and visualized by ethidium bromide staining. The PCR conditions for the amplification of G185C SNP region of APOA5 consisted of an initial melting temperature of 96 °C for 5 min followed by 35 cycles of melting (96 °C, 30 s), annealing (58 °C, 30 s) and extension (72 °C, 30 s). A final extension step (72 °C, 7 min) terminates the reaction. 138 bp PCR products were digested with 5U HaeIII at 37 °C for 20 h, and resulted in 87 and 51 bp fragments for the G allele and a non-digested 138 bp fragment for the T allele. Allele frequencies were obtained by direct gene counting.

Statistical analyses

Continuous variables were expressed as either mean ± SD or median and interquartile range. Normality of the sample distribution of each continuous variable was tested with the Kolmogorov–Smirnov test. Differences of normally distributed continuous variables were evaluated by the independent samples t test and data which were not normally distributed were compared using Mann–Whitney U test. Categorical variables were expressed as proportions and compared using χ2 test or Fisher’s exact test. Allele frequencies were determined by the gene counting method and departure from the Hardy–Weinberg equilibrium was evaluated by the χ2 test. In order to determine the effects of vascular risk factors, lipid parameters and APOA5 genotypes in the prediction of ischemic stroke, binary logistic regression analyses with backward selection method was used. Age and sex were also included as covariates. 2-tailed probability values with 95 % confidence intervals were estimated for each odds ratio (OR). The Hosmer–Lemeshow goodness-of-fit test was used for calibration. The statistical significance level was considered as P < 0.05. These statistical tests were conducted by using SPSS 16.0 statistical software package (SPSS Inc., Chicago, IL, USA). Power calculation to test the sufficiency of the sample size was carried out using G Power 3.1 software. A power value >50 % was considered adequate.

Results

Study population

Demographic features, results of the clinical laboratory tests and the prevalence of conventional risk factors of patients and control subjects are given in Table 1. The age of study population varied from 20 to 84 years in stroke patients and from 37 to 90 years in controls. The prevalence of conventional risk factors, hypertension, diabetes, smoking and obesity were found to be higher in patients compared to the controls.

Table 1 Clinical characteristics, lipid parameters and prevalence of conventional risk factors of ischemic stroke patients and controls
Full size table

According to clinical laboratory test results given in Table 1, the level of HDL-cholesterol was significantly lower and the levels of total cholesterol, LDL-cholesterol and VLDL-cholesterol were significantly higher in ischemic stroke patients when compared to controls; whereas TG levels did not differ significantly between the two groups.

APOA5 genotypes and allele frequencies

The distribution of genotypes and alleles resulting from S19W (c.56C>G), and −1131T>C and G185C (c.553G>T) SNPs of APOA5 in stroke patients and controls are presented in Table 2. There was no deviation of genotype and allele frequencies from Hardy–Weinberg equilibrium.

Table 2 Distribution of genotypes and allele frequencies for APOA5 19S/W, −1131T>C and G185C (c.553G>T) SNPs in stroke patients and controls
Full size table

S19W SNP of APOA5 exhibited similar distribution in stroke patients and controls. The homozygote polymorphic genotype (19WW) was seen in 0.8 % of patients and 1.7 % of controls. The frequency of the minor allele (19W) was found as 0.090 in the patient group and 0.062 in the controls (P = 0.191).

Genotype frequencies for the APOA5 −1131T>C SNP were not significantly different between patients and controls. The homozygote mutant genotype (CC) was found in 1.5 % of patients and 0.8 % of controls. No significant difference was found between patients and controls in terms of −1131C allele frequencies, which were 0.106 in patients and 0.102 in controls.

Homozygote mutant genotype for the G185C SNP (185CC) was not found in the study population of this work. There was no difference between the patient and control groups in terms of G185C genotype distributions. The rare allele frequency for the G185C SNP (185C) was calculated to be exactly the same (0.004) in cases and controls.

APOA5 S19W and −1131T>C SNPs and serum lipid parameters

In order to estimate the impact of the APOA5 S19W and −1131T>C SNPs on lipid metabolism, the mean serum lipid levels were compared with respect to the S19W and −1131T>C genotypes, separately in stroke patients and controls (Table 3).

Table 3 Lipid parameters of stroke patients and controls in different APOA5 S19W and −1131T>C genotype groups
Full size table

APOA5 S19W SNP affected lipid parameters in both patients and controls. Total cholesterol and LDL-cholesterol levels were significantly higher for stroke patients having at least one minor allele (19W) compared to those having the wild type genotype (19SS). There was a trend for higher TG levels in 19W allele carrying stroke patients compared to non-carriers (P = 0.051). In the control group, TG and VLDL-cholesterol levels were significantly higher in minor allele carriers (19SW+WW) compared to wild type individuals.

In stroke patients, LDL-cholesterol levels were significantly higher in −1131C allele carriers (P = 0.041). There was a trend for higher total cholesterol levels in −1131C allele carrier stroke patients than in non-carrier (−1131TT) stroke patients (P = 0.060). By contrast, in the control group, significant differences were found for TG and VLDL-cholesterol; −1131C allele carriers had significantly higher TG and VLDL-cholesterol levels compared to non-carriers.

Effects of −1131T>C genotypes on ischemic stroke in hypertensive, diabetic and obese subjects

Frequencies of ischemic stroke patients within different vascular risk groups (hypertensive, diabetic and obese subjects) are compared among −1131T>C genotypes (Table 4). APOA5 −1131TC and −1131CC genotypes were ascribed to a single group because only five individuals with CC homozygote genotype were observed.

Table 4 Effects of −1131T>C genotypes on ischemic stroke in hypertensive, diabetic and obese subjects
Full size table

The frequency of ischemic stroke among hypertensive subjects was 77.8 % (row “all”). When subjects were divided according to −1131T>C genotypes, ischemic stroke was observed in 76.9 % of hypertensive wild type −1131TT subjects (non-carriers). On the other hand, in hypertensive carriers (−1131TC+CC) the prevalence of ischemic stroke was higher (82.5 %). Similar results were obtained for diabetes and obesity; prevalence of ischemic stroke was higher in −1131C allele carrier diabetic or obese subjects compared to non-carriers. Differences were, however, not statistically significant.

Logistic regression analysis

Logistic regression analysis with backward likelihood method was used to determine the effects of vascular risk factors, lipid parameters and APOA5 genotypes in the prediction of ischemic stroke. When age, sex, hypertension, smoking status, diabetes, obesity, lipid parameters (total cholesterol, TG, LDL-cholesterol, HDL-cholesterol, VLDL-cholesterol), APOA5 S19W, −1131T>C and G185C genotypes were added as covariates, logistic regression revealed hypertension (OR = 3.413), smoking (OR = 3.208), VLDL-cholesterol (OR = 13.262) and HDL-cholesterol (OR = 0.285) to be the strongest determinants of ischemic stroke (Table 5). APOA5 genotypes were not associated with the risk of ischemic stroke. Calibration of the logistic regression model was tested by Hosmer–Lemeshow goodness-of-fit test and turned out to be satisfactory (Chi-square = 6.892; 8 degrees of freedom; P = 0.548). The model predicted 72.5 % of cases correctly.

Table 5 Logistic regression analysis of vascular risk factors (age, sex, hypertension, diabetes, smoking and obesity), lipid parameters (total cholesterol, TG, HDL-cholesterol, LDL-cholesterol and VLDL-cholesterol), APOA5 S19W, −1131T>C, and G185C genotypes in stroke patients and controls
Full size table

Discussion

Atherothrombotic stroke, which is the most common type of stroke, occurs when a blood clot forms on an atherosclerotic plaque within a blood vessel in the brain and blocks blood flow to that part of the brain. Elevated TG levels may lead to atherosclerotic plaque formation and APOA5 SNPs are related with serum TG levels [3]. Therefore, this study was designed to investigate the relationship between APOA5 SNPs and serum lipid parameters and search for the significance of these SNPs as stroke risk factors. Three major APOA5 SNPs, S19W, −1131T>C and G185C were tested as risk factors for ischemic stroke in Turkish population. We did not detect a significant difference in frequencies of rare alleles (19W, −1131C and 185C) between stroke patients and controls. Neither these polymorphic alleles were identified to be significantly associated with ischemic stroke after adjustment for possible confounders. Therefore, these SNPs of APOA5 cannot be considered as independent ischemic stroke risk factors in the studied Turkish subjects. On the contrary, in a study performed with Hungarian stroke patients and controls, the frequency of −1131C allele was found to be approximately two times higher in patients when compared to controls [18]. In Chinese Han population, −1131T>C SNP of APOA5 gene was independently associated with the development of ischemic stroke; it was suggested that CC homozygote genotype may have a promoting effect on ischemic stroke [23]. Since APOA5 −1131T>C SNP shows great variations between populations (Table 6), the disagreement between our results and the studies performed on Hungarians and Chinese Han population may be due to the differences in allele frequencies among populations.

Table 6 Comparison of APOA5 19W, −1131C and 185C allele frequencies in different populations
Full size table

Table 6 summarizes the APOA5 19W, −1131C and 185C allele frequencies in different populations and compares them with the frequencies obtained in control subjects of the present study. The minor allele frequency for the APOA5 S19W SNP was 0.062 in 123 control subjects of this study and was very close to this value in the study of Hodoglugil et al. [20], who also studied Turkish population (Table 6). The frequency of this allele was less than 10 % in most of the European populations, such as 0.060 [26] and 0.081 [27] in British population, 0.068 in Spanish [28], 0.064 (in men) and 0.078 (in women) in Czech [29] and 0.092 in Romanian [30] populations. 19W allele frequency was higher in Brazilians and reported to be 0.137 [31].

The rare allele frequency (−1131C) of APOA5 −1131T>C SNP in the control subjects was found to be 0.102 in the present study. In general, the frequency of −1131C allele was less than 10 % in Caucasians [5, 8, 26, 32–35], whereas it was found as 12.8 % in another Turkish study [20]. In contrast, the frequency of this allele was over 30 % in the Japanese [9, 10, 14, 36] and Chinese populations [12, 15, 37]. Lai et al. [38] reported the −1131C allele frequency of Asian Indians, Malay and Chinese in Singapore as 0.232, 0.269 and 0.294 respectively. C allele frequency was 0.285 in Koreans [39]. In the present study, the genotype frequencies of APOA5 −1131T>C polymorphism for −1131TT, −1131TC and −1131CC genotypes for control subjects were 80.5, 18.7 and 0.8 %, respectively which were consistent with the genotype frequencies (88.3, 11.1 and 0.6 %) found in a study with Caucasians [34]. The genotype frequencies of −1131T>C SNP in Chinese [15, 19] and Japanese populations [36] were considerably different.

In this work, the frequency of 185C allele in control subjects was found as 0.004 and this result is consistent with the studies performed on Caucasians [40], including Turks [20]. Hoduglugil et al. [20] found the frequency of this allele to be very low (0.006) in Turkish population and Hubacek et al. [40] supposed that this variant is probably absent in Caucasian population since they did not detect 185C allele in 420 healthy subjects. On the contrary, the frequency of this allele was higher in Chinese [13, 19, 21, 41, 42] and Japanese populations [36].

Studies in animals and humans have demonstrated that APOA5 gene may play an important role in TG metabolism, although the exact function of APOAV in TG metabolism is not known yet [5, 7]. Presence of a C allele at −1131 position of APOA5 gene affects serum lipid parameters [5, 8–12, 14–20, 22]. APOA5 −1131T>C polymorphism was found to be associated with significantly elevated TG levels in Caucasians [5, 8] including Turks [20, 22], Hungarians [17, 18], African Americans and Hispanics [8], Japanese [9, 10, 14], as well as Chinese [12, 15, 16, 19]. Likewise, we found that TG as well as total cholesterol, LDL-cholesterol and VLDL-cholesterol levels were higher in the −1131C allele carriers compared with the wild type individuals, both in stroke patients and controls. In addition to elevated TG levels, this SNP was also associated with reduced HDL-cholesterol levels in both Asian and Caucasian populations [5, 8, 11]. Similarly, HDL-cholesterol levels of −1131C allele carriers was slightly lower in controls of this study. The mechanism behind the association of the APOA5 −1131T>C SNP with the elevated plasma TG level is not evident yet. It is possible that the promoter region polymorphism of APOA5 gene affects the transcriptional activity of the gene, leading to low levels of circulating APOAV and thus, modulates plasma TG levels [8].

APOA5 S19W polymorphism was also correlated with lipid parameters and both stroke patients and controls with 19SW+WW genotype had higher TG levels in the present study. Similarly, rare allele of the S19W SNP was significantly associated with increased TG levels in a previous study on Turks [20]. TG levels were also significantly higher in Spanish subjects carrying the 19WW genotype in a large, well-characterized Spanish Mediterranean population [28]. De Andrade et al. [31] recently published results of a study on the influence of APOA5 S19W polymorphism on the TG levels in Brazilians. They found the 19W allele to be associated with increased TG levels [31].

APOA5 G185C polymorphism has also been found to correlate strongly with increased TG levels in Chinese population [13, 21]. Tang et al. [21] reported that the minor 185C allele carriers had significantly higher TG levels (2.31 mmol/L) when compared to the wild type 185GG genotype (1.68 mmol/L, P = 0.002) in a study with Chinese. The increase in TG level may be caused by the modulation of the function of APOAV due to replacement of glycine with cysteine at amino acid 185 caused by the c553G>T polymorphism [13]. However, in Caucasians [25] including Turks [20] a correlation between G185C polymorphism and TG level was not found. In the present study, serum lipid parameters could not be analyzed with respect to G185C polymorphism, since there are only five heterozygote individuals (four patients and one control) and no homozygote mutant in our population.

When we analyzed the effects of conventional vascular risk factors (hypertension, diabetes and obesity) on ischemic stroke risk with respect to APOA5 −1131T>C genotypes, we observed that among hypertensive, diabetic and obese subjects, −1131C allele carriers had an increased proportion of ischemic stroke compared to non-carriers. This result might be due to the increased TG levels observed in −1131C subjects. Hypertension may change the flow properties of the blood and this situation was independently associated with higher blood viscosity [43]. Blood stickiness is linked to the early development of atherosclerosis, which results from the build-up of cholesterol, fats and biological debris in the tissue lining the inside of blood vessels. This build-up can obstruct blood flow to the heart and brain and thereby cause a heart attack or stroke [44]. It was suggested that oxidative stress may constitute a major pathogenic factor in the development of hypertension and type 2 diabetes [45], as in the case of atherosclerosis and ischemic stroke [46, 47]. Fat accumulation is also correlated with systemic oxidative stress in humans and mice [48]. Therefore, individuals carrying at least one −1131C allele had an increased risk of hypertension-, diabetes- and obesity-associated ischemic stroke risk.

The study of APOA5 polymorphisms came into prominence in Turkish population due to the variations of the allele frequencies in different populations and limited number of studies related with APOA5 polymorphisms and ischemic stroke risk. To the best of our knowledge, this is the first study addressing the relation between APOA5 G185C polymorphism and stroke risk. Additionally, these three polymorphisms of APOA5 were analyzed for the first time in terms of their relation to ischemic stroke in Turkish subjects. Although we observed higher frequency of hypertension-, diabetes- and obesity-related ischemic stroke in individuals carrying −1131CC or −1131TC genotypes, the minor alleles of APOA5 did not represent an independent risk for ischemic stroke. Limitations of the present study include number of control subjects. However, it should be noted that the sample was clinically well defined and strictly selected so as to keep mean age of controls and stroke patients similar, as age is the strongest determinant of stroke. It was highly difficult to find elderly subjects who meet the required criteria as controls, i.e. no history of ischemic stroke, transient ischemic attack and ischemic heart disease at any time or myocardial infarction within 3 weeks and no more than 70 % carotid stenosis.

In conclusion, the present study determined distribution of the three genetic variants of APOA5 in ischemic stroke patients and respective control group. In addition, influence of these polymorphisms on serum lipid profile was analyzed. Although two of the studied polymorphisms were correlated with TG levels, they were not independently related with ischemic stroke risk in the studied Turkish subjects.