452 Gene expression, PCR, gene array, in situ hybridization histochemistry, mRNA
Interaction between ALOX5AP and CYP3A5 gene variants significantly increases the risk for cerebral infarctions in Chinese Li-Fen Chi, Xing-Yang Yi, Min-Jie Shao, Jing Lin and Qiang Zhou In this study, we investigated associations between susceptibility genes and cerebral infarctions in a Chinese population, and whether gene–gene interactions increase the risk of cerebral infarctions. Overall, 292 patients with cerebral infarctions and 259 healthy control individuals were included. Eight variants in five candidate genes were examined for the risk of stroke, including the SG13S32 (rs9551963), SG13S42 (rs4769060), SG13S89 (rs4769874), and SG13S114 (rs10507391) variants of the 5-lipoxygenase activating protein (ALOX5AP) gene, the G860A (rs751141) variant of the soluble epoxide hydrolase (EPHX2) gene, the A1075C (rs1057910) variant of the CYP2C9*2 gene, the C430T (rs1799853) variant of the CYP2C9*3 gene, and the A6986G (rs776746) variant of the CYP3A5 gene. Gene–gene interactions were explored using generalized multifactor dimensionality reduction methods. There were no statistically significant differences in the frequencies of the genotypes of the eight candidate genes. The generalized multifactor dimensionality reduction analysis showed a significant gene–gene interaction between SG13S114 and A6986G, with scores of 10 for cross-validation consistency and 9 for
the sign test (P = 0.0107). These gene–gene interactions predicted a significantly higher risk of cerebral infarction (adjusted for age, hypertension, and diabetes mellitus; odds ratio = 1.80495%, confidence interval: 1.180–2.759, P = 0.006). A two-loci gene interaction confers a significantly higher risk for cerebral infarction. The combinational analysis used in this study may be helpful in the elucidation of genetic risk factors for common and c 2014 complex diseases. NeuroReport 25:452–457 Wolters Kluwer Health | Lippincott Williams & Wilkins.
Introduction
membrane glycerophospholipids, can be released by phospholipases in response to several stimuli, such as ischemia [5]. Free AA is then available for metabolism by the cyclooxygenase, lipoxygenase (LOX), and cytochrome P450 (CYP) pathways, which generate prostanoids, leukotrienes, and epoxyeicosatrienoic acids (EETs), respectively [6,7]. These products play different roles during inflammation.
Cerebrovascular disease is the main cause of death and disability worldwide [1]. In China, where 2.6 million individuals each year are estimated to experience a stroke for the first time, ischemic stroke was recently estimated to account for 43.7 – 78.9% of all strokes [2]. The development of stroke is influenced by a variety of cardiovascular risk factors including hypertension, smoking, diabetes mellitus (DM), and genetic predispositions. Stroke is a common and complex trait related to gene– gene and gene–environment interactions. Single-locus methods may not be appropriate for the study of complex cardiovascular disorders, which do not follow a simple Mendelian pattern of inheritance. It is possible that particular loci contribute toward certain complex diseases only by their interactions with other genes, and the main effects of individual loci may be too small to be observed [3]. Strokes are considered to be largely caused by an inflammation-mediated destabilization and rupture of atherosclerotic lesions [4]. The main sources of inflammation mediators are the products of arachidonic acid (AA) metabolism. AA, normally found esterified to cell c 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins 0959-4965
NeuroReport 2014, 25:452–457 Keywords: generalized multifactor dimensionality reduction, genetics, interaction, stroke, variants Department of Neurology, 3rd Affiliated Hospital of Wenzhou Medical College, Zhejiang, China Correspondence to Xing-Yang Yi, MSc, Department of Neurology, 3rd Affiliated Hospital of Wenzhou Medical College, 108 Wansong Road, Rui’an, Wenzhou 325200, Zhejiang, China Tel: + 86 577 658 66222; fax: + 86 577 658 66013; e-mail: [email protected] Received 9 October 2013 accepted 28 November 2013
A genetic variation in the components of the inflammatory response has been implicated as a risk factor, particularly through interactions between the cyclooxygenase, LOX, and CYP pathways [8,9]. ALOX5AP encodes the 5-LOX-activating protein, which is a key enzyme involved in leukotriene biosynthesis within the LOX pathway. A linkage study identified variations in ALOX5AP to be factors in susceptibility to myocardial infarction [10]. A subsequent study of an Icelandic population found that a four-SNP ALOX5AP haplotype (termed HapA, including SG13S25, SG13S32, SG13S89, and SG13S114) doubled the risk of coronary heart disease and incidence of stroke [11]. However, to date, studies relating ALOX5AP gene polymorphisms and the risk of ischemic stroke in various populations have yielded widely variable conclusions. DOI: 10.1097/WNR.0000000000000114
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Gene variants increase stroke risk Chi et al. 453
AA is metabolized by CYP epoxygenases into four EETs. Soluble epoxide hydrolase, encoded by the EPHX2 gene, metabolizes EETs into less biologically active dihydroxyeicosatrienoic acids [12]. Accumulating evidence indicates that EETs have diverse cardiovascular protective effects [13]. CYP2C9 enzymes are considered the major epoxygenase enzymes [14], whereas CYP3A5 enzymes have minimal epoxygenase activity [15]. Therefore, in this study, we evaluated the ALOX5AP gene in the LOX pathway and the soluble epoxide hydrolase (EPHX2), CYP2C9*2, CYP2C9*3, and CYP3A5 genes in the CYP pathway as susceptibility genes for stroke in a small population of stroke patients and matched control individuals from southeastern China using a case– controlled design. We hypothesized that interactions between variants of multiple genes would confer a higher risk of stroke than a single variant in one gene. Our hypothesis was tested using generalized multifactor dimensionality reduction (GMDR) with eight different variants in five candidate genes.
All participants were Chinese Han in origin and unrelated. Participants were excluded from both the experimental and the control groups if they had arteritis, infections, tumors, blood diseases, serious cardiopulmonary, liver, or kidney problems, thyroid disease, or autoimmune diseases. The diagnostic criteria and the collection of the clinical data
The following vascular risk factors for each individual were determined: age, sex, hypertension, DM, cigarette smoking, alcohol intake, total plasma cholesterol, triglycerides, and low-density lipoprotein cholesterol. Hypertension was defined as the mean of three independent measures of blood pressure of at least 140/ 90 mmHg or the use of antihypertensive drugs. DM was diagnosed as when a patient had a fasting glucose level greater than 7.8 mM, greater than 11.1 mM 2 h after oral glucose challenge, or both. A detailed medical assessment of history was performed and information on stroke risk factors was obtained from each participant.
Materials and methods Study populations
Ethical approval for this study was provided by the Ethical Committee of the Third Affiliated Hospital of Wenzhou Medical College. Informed consent was obtained from each patient before study enrollment. The study population included 292 ischemic stroke patients and 259 control individuals. Patients were consecutively recruited between March 2010 and March 2012 from the Third Affiliated Hospital of Wenzhou Medical College. Each diagnosis of a cerebral infarction was made on the basis of the results of a neurological examination using computerized tomography or MRI according to the ninth revision of the International Classification of Diseases. The infarctions were classified according to a modified version of the Trial of Org 10172 in Acute Stroke Treatment classification system [16]. Patients with cerebral infarctions related to atherothrombotic and small artery disease were included. The exclusion criteria were as follows: (a) cardiogenic cerebral embolisms, (b) unclear etiologies, (c) a history of strokes, and (d) a lack of willingness to participate in this study. On the basis of sample-size calculations, we included 292 inpatients with cerebral infarctions from the Department of Neurology, among whom 173 were men and 119 were women, average age 68.56±11.09 years. The control participants were selected from among outpatients with no history of cerebral infarctions as confirmed by assessment of medical history, physical, and laboratory examinations at our center. They had no family history of cerebral infarctions and were not genetically related to the cerebral-infarction patients. A total of 259 volunteers participated in this study, of whom 148 were men and 111 were women, average age 63.92±9.29 years.
Marker selection and genotyping
Ten milliliters of blood was drawn from an arm vein into a sterile tube containing ethylenediaminetetraacetic acid, and stored at – 801C until genotype analysis was carried out. Genomic DNA was extracted from peripheral blood using a modified phenol/chloroform method and purified using the UNIQ-10 kit (Sangon Biotech Co. Ltd, Shanghai, China). Amplification of the target sequences was performed in a multiplex reaction containing 5 ng of DNA, 0.95 ml of water, 0.625 ml of PCR buffer containing 15 mM MgCl2, 1 ml of 2.5 mM dNTP, 0.325 ml of 25 mM MgCl2, 1 ml of PCR primers (Table 1), and 0.5 U HotStar Taq (Qiagen Co. Ltd, Shanghai, China). The reaction was incubated at 941C for 15 min, followed by 45 cycles at 941C for 20 s, 561C for 30 s, and 721C for 1 min, and a final incubation at 721C for 3 min. After PCR amplification, the remaining dNTPs were dephosphorylated by adding 1.53 ml water, 0.17 ml of SAP buffer, and 0.3 U of shrimp alkaline phosphatase (Sequenom Inc., San Diego, California, USA). After incubation at 371C for 40 min, the enzyme was deactivated by incubating at 851C for 5 min. The extension primers were then added to the reaction in a buffer containing 0.755 ml water, 0.2 ml of 10 iPLEX buffer, 0.2 ml termination mix, 0.041 ml of iPLEX enzyme (Sequenom Inc.), and 0.804 ml of 10 mM extension primer. The single base extension reaction consisted of an initial denaturation at 941C for 30 s and then 941C for 5 s, followed by five cycles of 521C for 5 s and 801C for 5 s for a total of 40 cycles. A final extension at 721C was performed for 3 min. The reaction mix was desalted by adding 6 mg of cation exchange resin (Sequenom Inc.), mixed, and resuspended
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
454
NeuroReport 2014, Vol 25 No 7
in 25 ml of water. Once the primer extension reaction was completed, the samples were spotted onto a 384-well spectroCHIP (Sequenom Inc.) using MassARRAY Nanodispenser (Sequenom Inc.) and genotyped using the matrix-assisted laser desorption ionization time-of-flight mass spectrometer. Genotype calling was performed in real time using MassARRAY RT software, version 3.0.0.4 and analyzed using the MassARRAY Typer software, version 3.4 (Sequenom Inc.). Eight variants in five candidate genes were examined for both the patients and the control participants: SG13S32 (rs9551963), SG13S42 (rs4769060), SG13S89 (rs4769874), SG13S114 (rs10507391) variants of the 5-LOX-activating protein (ALOX5AP) gene, the G860A (rs751141) variant of the soluble epoxide hydrolase (EPHX2) gene, the A1075C (rs1057910) variant of the CYP2C9*2 gene, the C430T (rs1799853) variant of the CYP2C9*3 gene, and the A6986G (rs776746) variant of the CYP3A5 gene (Table 1). Statistical analysis
Summary statistics are expressed as the mean±SEM or as percentages. The w2-test was used to analyze the deviation of Hardy–Weinberg equilibrium for genotype frequencies and to compare genotype frequencies. Continuous variables were compared between patients with ischemic stroke and control participants using Student’s t-test. Discrete variables were compared using w2 tests or Fisher’s exact test. For gene–gene interaction analyses, the GMDR method was applied. As the participants were of the same ethnicity, sex and age were modeled as covariates. The GMDR computes the maximum-likelihood estimates and the scores of all individuals under the null hypothesis. Then, the cumulative score is calculated within each multifactor cell, which is labeled either as high risk if the average score meets or exceeds a preassigned threshold of 0 or as low risk if the score is less than 0. An exhaustive search of all possible one-locus to seven-locus models was performed for all variants. The model with the minimum prediction error, the maximum cross-validation consistency score, and a P value of 0.05 or lower automatically derived from the sign test in GMDR software was Table 1
considered to be the best model. This was confirmed by a permutation test implemented in the GMDR software as well [17]. Furthermore, we carried out multivariate logistic regression analyses adjusted for some risk factors to assess the independent contribution of gene–gene interactions to a risk of stroke. Adjusted odds ratios with 95% confidence intervals were computed. A P value of less than 0.05 was considered statistically significant. The GMDR v0.7 program was used in this study (http:// www.healthsystem.virginia.edu/internet/addiction-genomics/Software). All other analyses were carried out using SPSS 16.0 (SPSS Inc., Chicago, Illinois, USA).
Results General clinical characteristics of the participants
The demographic characteristics of the control participants and ischemic stroke patients are presented in Table 2. Briefly, the mean age of the stroke patients was 68.56±11.09 years and that of the control participants was 63.92±9.29 years, and 59.2% of the stroke patients and 57.5% of the control participants were men. There was a significant difference in the age of the participants in the two groups (P < 0.001), but not in their sex. As expected, stroke patients had a higher prevalence of risk factors for vascular diseases, including a history of hypertension (P < 0.01) and diabetes (P = 0.032). However, differences in conventional risk factors including smoking and alcohol intake and levels of LDL-C, total plasma cholesterol, and triglycerides did not reach statistical significance between the two groups (P > 0.05) (Table 2).
Association between specific single-nucleotide polymorphism genotypes and risk of stroke
The single-nucleotide polymorphisms and their genotype frequencies are shown in Table 3. The genotype distributions of the eight candidate genes examined in this study were consistent with the Hardy–Weinberg equilibrium model (P > 0.05). However, there were no statistically significant differences in the frequencies of the genotypes of the eight candidate genes between the stroke patients and the control participants (P > 0.05).
Amplification of alleles using the sense and antisense primers
SNPs SG13S32 (rs9551963) SG13S42 (rs4769060) SG13S89 (rs4769874) SG13S114 (rs10507391) C430T (rs1799853) A1075C (rs1057910) A6986G (rs776746) G860A (rs751141)
Upstream primer
Downstream primer
ACGTTGGATGGGGTTCAAGAGAGAAATTCC ACGTTGGATGGAAGGGTAGAAGTGTCTCAG ACGTTGGATGCACCAGGGAGCAAGCATTAG ACGTTGGATGTCCAGATGTATGTCCAAGCC ACGTTGGATGCTGCGGAATTTTGGGATGGG ACGTTGGATGCTACACAGATGCTGTGGTGC ACGTTGGATGGATGAAGGGTAATGTGGTCC ACGTTGGATGTTTTCTAGATCCCTGCTCTG
ACGTTGGATGAGTTCTTGACCTCACCAACC ACGTTGGATGCGTGGTAATGGGTTTTGAGG ACGTTGGATGTTTCAGGCATGCTCTGCACC ACGTTGGATGCTCTTAAGGTAGGTCTATGG ACGTTGGATGACCCACCCTTGGTTTTTCTC ACGTTGGATGTGTCACAGGTCACTGCATGG ACGTTGGATGCCATAATCTCTTTAAAGAGC ACGTTGGATGAGCAGATGACTCTCCATAGC
SNP, single-nucleotide polymorphism.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Gene variants increase stroke risk Chi et al. 455
Table 2
Clinical characteristics of the participants
Characteristics Age (years) Men (%) Hypertension (%) Diabetes mellitus (%) Cigarette smoking (%) Alcohol intake (%) TC (mM) LDL-C (mM) TG (mM)
Stroke patients (n = 292)
Control participants (n = 259)
P value
68.56±11.09* 59.2 80.1* 33.6* 44.3 46.6 5.25±1.11 3.14±1.25 1.85±1.04
63.92±9.29 57.5 23.2 25.9 41.7 45.2 5.07±1.12 2.96±1.16 1.73±0.99
< 0.001 0.729 < 0.001 0.032 0.547 0.797 0.089 0.098 0.078
Age, TC, TG, and LDL-C PLT values are given as the mean±SD. The other values are given as the percentage of the group. LDL-C, low-density lipoprotein cholesterol; PLT, platelets; TC, total cholesterol; TG, triglycerides. *P < 0.05 versus control.
Table 3
Association between specific SNP genotypes and risk of stroke
SNPs
Genotype
SG13S32
AA AC CC AA AG GG AA AG GG AA AT TT CC AA AC CC AA AG GG GG AG AA
SG13S42
SG13S89
SG13S114
C430T A1075C
A6986G
G860A
Control participants [n (%)] 101 122 36 103 122 34 0 4 255 36 134 89 259 140 19 0 28 92 139 165 81 13
Stroke patients [n (%)]
(39.0) (47.1) (13.9) (39.8) (47.1) (13.1) (0) (1.5) (98.5) (13.9) (51.7) (34.4) (100) (92.7) (7.3) (0) (10.8) (35.5) (53.7) (63.7) (31.3) (5.0)
103 148 41 111 142 39 0 12 280 50 149 93 292 266 25 1 31 111 150 184 97 11
P value
(35.3) (50.7) (14.0) (38.0) (48.6) (13.4) (0) (4.1) (95.9) (17.1) (51.0) (31.9) (100) (91.1) (8.6) (0.3) (10.6) (38.0) (51.4) (63.0) (33.2) (3.8)
0.646
0.914
0.081
0.55
– 0.554
0.829
0.717
SNP, single-nucleotide polymorphism.
Table 4 Comparison of the best models, prediction accuracies, cross-validation consistencies, and P values identified by GMDR for ischemic stroke Best model
Training balanced accuracy
Testing balanced accuracy
1a 1,2 1,2,3 1,2,4,5 1,2,3,4,5 1,2,3,4,5,6 1,2,3,4,5,6,7
0.5399 0.5820 0.6165 0.6627 0.7210 0.7500 0.7566
0.4710 0.5632 0.4532 0.4331 0.4867 0.4925 0.4865
Sign test (P) 3 9 2 2 5 5 4
(0.9453) (0.0107) (0.9893) (0.9893) (0.9893) (0.6230) (0.8281)
Cross-validation consistency 7/10 10/10 5/10 3/10 10/10 10/10 10/10
GMDR, generalized multifactor dimensionality reduction. a SG13S114, A6986G, EPHX2, SG13S32, SG13S42, A1075C, and SG13S89 are symbolized as 1–7, respectively.
Gene–gene interactions
High-order interactions were explored for ischemic stroke using the GMDR method. With covariate adjustments, the best model for ischemic stroke included SG13S114 and A6986G was scored 10/10 for cross-validation
consistency and 9/10 by the sign test (P = 0.0107), as shown in Table 4. Furthermore, the significant interactions among the above two-locus models were confirmed by a permutation test (P = 0.027). A one-locus model was also computed for each variant and the minimum P value
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
456
NeuroReport 2014, Vol 25 No 7
Fig. 1
SG13S114 AT
AA
TT
AA
A6986GG
3.8
−1.3
2.6
20.0 GA 2.4
−4.4
3.9
−16.6
12.8
−6.7
−2.3
−9.3
−29.4 21.0 GG
15.2 9.3
−15.7
−5.4
In each cell, the left bar represents a positive score and the right bar represents a negative score. High-risk cells are indicated by dark shading, low-risk cells by light shading.
Table 5 Multiple regression analysis of the major risk factors for cerebral infarction Risk factors Age Hypertension Diabetes mellitus Interactive variable
Wald
OR
95% CI
P value
12.017 136.521 0.408 7.420
1.660 13.037 0.859 1.804
1.246–2.210 8.474–20.055 0.540–1.369 1.180–2.759
0.001 0.000 0.523 0.006
CI, confidence interval; OR, odds ratio.
was 0.9453, suggesting that their contribution to stroke risk was because of the joint action of the two genes, but not a single locus. Figure 1 shows the best model for ischemic stroke identified by GMDR. The three interactions making larger contributions to this model were between SG13S114 AA and A6986G AA, SG13S114 AA and A6986G GG, and SG13S114 TT and A6986G AA. Logistic regression analysis
As shown in Fig. 1, the relative risk of the nine combinations of the two genes was considered as an interactive variable, with high risk assigned a 1 and low risk assigned a 0. Table 5 shows that certain gene–gene interactions predict a significantly higher risk of cerebral infarction (adjusted for age, hypertension, DM; odds ratio = 1.804, 95% confidence interval 1.180–2.759, P = 0.006).
Discussion The results that we presented in this study do not support a single locus within the eight candidate genes as
a risk gene for ischemic stroke. Rather, using the GMDR method to search gene–gene interactions that predict a high risk of stroke, we principally found that interactions between ALOX5AP SG13S114 and CYP3A5 A6986G were significant combinational predictors of ischemic stroke. There was a 1.804-fold increased risk for ischemic stroke in individuals with these combined genetic factors. Our findings add to the evidence that interactions between genes with marginal effects in their own rights can increase the risk of common and complex diseases, such as stroke. Despite our experiments, the nature of the two-factor interactions between ALOX5AP and CYP3A5 remains unclear. As stroke does not have a typical pattern of Mendelian inheritance, our observations could be considered as quite expected. One possible explanation for the two-factor interaction is that these two genes are all involved in AA metabolism and its impaired equilibrium. In addition to cyclooxygenase, the LOX and CYP pathways have been identified as a major pathway for AA metabolism. ALOX5AP encodes 5-LOX-activating protein, which activates arachidonate 5-lypoxygenase and leads to leukotriene synthesis. Leukotrienes are proinflammatory mediators, secreted by various types of inflammatory cells that cluster at injured sites in blood vessels, which change vascular permeability and promote atherosclerosis formation [18]. After ALOX5AP gene mutations, coding proteins can increase 5-LOX activity, promote leukotriene-mediated damage to vascular walls, and form vulnerable plaques, which lead to ischemic strokes. A meta-analysis has shown that ALOX5AP and SG13S114 mutations are independent risk factors in Iberian populations for ischemic stroke [19]. In contrast, a study in Chinese patients reported that the SG13S114A allele is not an independent risk gene for ischemic infarction [20]. The results are consistent with our single-locus findings. AA is metabolized in endothelial cells by CYP epoxygenases into four EETs, which have diverse cardiovascular protective effects, including antiapoptotic effects in endothelial cells and anti-inflammatory and antiangiogenic effects, suggesting antiartheroscelerotic effects of EETs that may be beneficial for stroke treatment [21–23]. The CYP3A5 enzyme is a type of epoxygenase enzyme. CYP3A5 gene mutations can reduce EET generation [24]. A6986G is the most common mutation of CYP3A5. As both the ALOX5AP SG13S114 and the CYP3A5 A6986G genes are involved in AA metabolism, their separate effects on leukotrienes and EETs can eliminate the dynamic equilibrium of AA metabolism. Furthermore, a CYP3A5 genetic polymorphism has been investigated with respect to the risk for high blood pressure [25]. The mechanism underlying this
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Gene variants increase stroke risk Chi et al. 457
association may be related to sodium reabsorption. As hypertension is the most important risk factor of cerebral infarction, this CYP3A5 gene polymorphism may also be associated with atherosclerosis and cerebral infarction. Therefore, the interaction between these two genes may lead to the progression of atherosclerosis and increase susceptibility to cerebral infarction through a common mechanism of action.
8
9
10 11
Conclusion In the present study, we show that interactions between the ALOX5AP SG13S114 and CYP3A5 A6986G genes can engender a significantly higher risk of stroke than a single risk factor. This gene–gene interaction may serve as a novel area for stroke research. The two-locus combination may provide a target for prevention of this disease.
12 13
14
15
Acknowledgements This work was supported by the Zhejiang Provincial Natural Science Foundation (No. Y2110379) and the Zhejiang Province Science and Technology Research Foundation (No. 2012c33106) of China. The authors thank all the nurses in the Department of Neurology of the Third Affiliated Hospital of Wenzhou Medical College.
16
17
18
Conflicts of interest
There are no conflicts of interest.
19
References
20
1
2
3 4 5
6 7
Rothwell PM, Coull AJ, Giles MF, Howard SC, Silver LE, Bull LM, et al. Change in stroke incidence, mortality, case-fatality, severity, and risk factors in Oxfordshire, UK from 1981 to 2004 (Oxford Vascular Study). Lancet 2004; 363:1925–1933. Jiang B, Wang WZ, Chen H, Hong Z, Yang QD, Wu SP, et al. Incidence and trends of stroke and its subtypes in China: results from three large cities. Stroke 2006; 37:63–68. Culverhouse R, Suarez BK, Lin J, Reich T. A perspective on epistasis: limits of models displaying no main effect. Am J Hum Genet 2002; 70:461–471. Stoll G, Bendszus M. Inflammation and atherosclerosis: novel insights into plaque formation and destabilization. Stroke 2006; 37:1923–1932. Morisseau C, Hammock BD. Epoxide hydrolases: mechanisms, inhibitor designs, and biological roles. Annu Rev Pharmacol Toxicol 2005; 45:311–333. Calder PC. Polyunsaturated fatty acids and inflammatory processes: new twists in an old tale. Biochimie 2009; 91:791–795. Funk CD. Prostaglandins and leukotrienes: advances in eicosanoid biology. Science 2001; 294:1871–1875.
21
22
23
24 25
Jerrard-Dunne P, Sitzer M, Risley P, Buehler A, von Kegler S, Markus HS. Inflammatory gene load is associated with enhanced inflammation and early carotid atherosclerosis in smokers. Stroke 2004; 35:2438–2443. Jerrard-Dunne P, Sitzer M, Risley P, Steckel DA, Buehler A, von Kegler S, et al. Interleukin-6 promoter polymorphism modulates the effects of heavy alcohol consumption on early carotid artery atherosclerosis: the Carotid Atherosclerosis Progression Study (CAPS). Stroke 2003; 34:402–407. Topol EJ, Smith J, Plow EF, Wang QK. Genetic susceptibility to myocardial infarction and coronary artery disease. Hum Mol Genet 2006; 15:117–123. 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. Zeldin DC. Epoxygenase pathways of arachidonic acid metabolism. J Biol Chem 2001; 276:36059–36062. Larsen BT, Gutterman DD, Hatoum OA. Emerging role of epoxyeicosatrienoic acids in coronary vascular function. Eur J Clin Invest 2006; 36:293–300. Daikh BE, Lasker JM, Raucy JL, Koop DR. Regio- and stereoselective epoxidation of arachidonic acid by human cytochromes P450 2C8 and 2C9. J Pharmacol Exp Ther 1994; 271:1427–1433. Rifkind AB, Lee C, Chang TK, Waxman DJ. Arachidonic acid metabolism by human cytochrome P450s 2C8, 2C9, 2E1, and 1A2: regioselective oxygenation and evidence for a role for CYP2C enzymes in arachidonic acid epoxygenation in human liver microsomes. Arch Biochem Biophys 1995; 320:380–389. Han SW, Kim SH, Lee JY, Chu CK, Yang JH, Shin HY, et al. A new subtype classification of ischemic stroke based on treatment and etiologic mechanism. Eur Neurol 2007; 57:96–102. Ritchie MD, Hahn LW, Roodi N, Bailey LR, Dupont WD, Parl FF, et al. Multifactor-dimensionality reduction reveals high-order interactions among estrogen-metabolism genes in sporadic breast cancer. Am J Hum Genet 2001; 69:138–147. Dichgans M. Genetics of ischaemic stroke. Lancet Neurol 2007; 6: 149–161. Domingues-Montanari S, Fernandez-Cadenas I, del Rio-Espinola A, Corbeto N, Krug T, Manso H, et al. Association of a genetic variant in the ALOX5AP with higher risk of ischemic stroke: a case–control, meta-analysis and functional study. Cerebrovasc Dis 2010; 29:528–537. He YL, Zhu M, Jin XP, Zhou YL. Relationship of polymorphism of SG13S114A/T in ALOX5AP gene with atherosclerotic cerebral infarction. Zhejiang Da Xue Xue Bao Yi Xue Ban 2009; 38:630–633. Heizer ML, McKinney JS, Ellis EF. 14,15-Epoxyeicosatrienoic acid inhibits platelet aggregation in mouse cerebral arterioles. Stroke 1991; 22: 1389–1393. Node K, Huo Y, Ruan X, Yang B, Spiecker M, Ley K, et al. Anti-inflammatory properties of cytochrome P450 epoxygenase-derived eicosanoids. Science 1999; 285:1276–1279. Wang Y, Wei X, Xiao X, Hui R, Card JW, Carey MA, et al. Arachidonic acid epoxygenase metabolites stimulate endothelial cell growth and angiogenesis via mitogen-activated protein kinase and phosphatidylinositol 3-kinase/Akt signaling pathways. J Pharmacol Exp Ther 2005; 314: 522–532. Burk O, Wojnowski L. Cytochrome P450 3A and their regulation. Naunyn Schmiedebergs Arch Pharmacol 2004; 369:105–124. Bochud M, Eap CB, Elston RC, Bovet P, Maillard M, Schild L, et al. Association of CYP3A5 genotypes with blood pressure and renal function in African families. J Hypertens 2006; 24:923–929.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Interaction between ALOX5AP and CYP3A5 gene variants significantly increases the risk for cerebral infarctions in Chinese Li-Fen Chi, Xing-Yang Yi, Min-Jie Shao, Jing Lin and Qiang Zhou In this study, we investigated associations between susceptibility genes and cerebral infarctions in a Chinese population, and whether gene–gene interactions increase the risk of cerebral infarctions. Overall, 292 patients with cerebral infarctions and 259 healthy control individuals were included. Eight variants in five candidate genes were examined for the risk of stroke, including the SG13S32 (rs9551963), SG13S42 (rs4769060), SG13S89 (rs4769874), and SG13S114 (rs10507391) variants of the 5-lipoxygenase activating protein (ALOX5AP) gene, the G860A (rs751141) variant of the soluble epoxide hydrolase (EPHX2) gene, the A1075C (rs1057910) variant of the CYP2C9*2 gene, the C430T (rs1799853) variant of the CYP2C9*3 gene, and the A6986G (rs776746) variant of the CYP3A5 gene. Gene–gene interactions were explored using generalized multifactor dimensionality reduction methods. There were no statistically significant differences in the frequencies of the genotypes of the eight candidate genes. The generalized multifactor dimensionality reduction analysis showed a significant gene–gene interaction between SG13S114 and A6986G, with scores of 10 for cross-validation consistency and 9 for
the sign test (P = 0.0107). These gene–gene interactions predicted a significantly higher risk of cerebral infarction (adjusted for age, hypertension, and diabetes mellitus; odds ratio = 1.80495%, confidence interval: 1.180–2.759, P = 0.006). A two-loci gene interaction confers a significantly higher risk for cerebral infarction. The combinational analysis used in this study may be helpful in the elucidation of genetic risk factors for common and c 2014 complex diseases. NeuroReport 25:452–457 Wolters Kluwer Health | Lippincott Williams & Wilkins.
Introduction
membrane glycerophospholipids, can be released by phospholipases in response to several stimuli, such as ischemia [5]. Free AA is then available for metabolism by the cyclooxygenase, lipoxygenase (LOX), and cytochrome P450 (CYP) pathways, which generate prostanoids, leukotrienes, and epoxyeicosatrienoic acids (EETs), respectively [6,7]. These products play different roles during inflammation.
Cerebrovascular disease is the main cause of death and disability worldwide [1]. In China, where 2.6 million individuals each year are estimated to experience a stroke for the first time, ischemic stroke was recently estimated to account for 43.7 – 78.9% of all strokes [2]. The development of stroke is influenced by a variety of cardiovascular risk factors including hypertension, smoking, diabetes mellitus (DM), and genetic predispositions. Stroke is a common and complex trait related to gene– gene and gene–environment interactions. Single-locus methods may not be appropriate for the study of complex cardiovascular disorders, which do not follow a simple Mendelian pattern of inheritance. It is possible that particular loci contribute toward certain complex diseases only by their interactions with other genes, and the main effects of individual loci may be too small to be observed [3]. Strokes are considered to be largely caused by an inflammation-mediated destabilization and rupture of atherosclerotic lesions [4]. The main sources of inflammation mediators are the products of arachidonic acid (AA) metabolism. AA, normally found esterified to cell c 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins 0959-4965
NeuroReport 2014, 25:452–457 Keywords: generalized multifactor dimensionality reduction, genetics, interaction, stroke, variants Department of Neurology, 3rd Affiliated Hospital of Wenzhou Medical College, Zhejiang, China Correspondence to Xing-Yang Yi, MSc, Department of Neurology, 3rd Affiliated Hospital of Wenzhou Medical College, 108 Wansong Road, Rui’an, Wenzhou 325200, Zhejiang, China Tel: + 86 577 658 66222; fax: + 86 577 658 66013; e-mail: [email protected] Received 9 October 2013 accepted 28 November 2013
A genetic variation in the components of the inflammatory response has been implicated as a risk factor, particularly through interactions between the cyclooxygenase, LOX, and CYP pathways [8,9]. ALOX5AP encodes the 5-LOX-activating protein, which is a key enzyme involved in leukotriene biosynthesis within the LOX pathway. A linkage study identified variations in ALOX5AP to be factors in susceptibility to myocardial infarction [10]. A subsequent study of an Icelandic population found that a four-SNP ALOX5AP haplotype (termed HapA, including SG13S25, SG13S32, SG13S89, and SG13S114) doubled the risk of coronary heart disease and incidence of stroke [11]. However, to date, studies relating ALOX5AP gene polymorphisms and the risk of ischemic stroke in various populations have yielded widely variable conclusions. DOI: 10.1097/WNR.0000000000000114
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Gene variants increase stroke risk Chi et al. 453
AA is metabolized by CYP epoxygenases into four EETs. Soluble epoxide hydrolase, encoded by the EPHX2 gene, metabolizes EETs into less biologically active dihydroxyeicosatrienoic acids [12]. Accumulating evidence indicates that EETs have diverse cardiovascular protective effects [13]. CYP2C9 enzymes are considered the major epoxygenase enzymes [14], whereas CYP3A5 enzymes have minimal epoxygenase activity [15]. Therefore, in this study, we evaluated the ALOX5AP gene in the LOX pathway and the soluble epoxide hydrolase (EPHX2), CYP2C9*2, CYP2C9*3, and CYP3A5 genes in the CYP pathway as susceptibility genes for stroke in a small population of stroke patients and matched control individuals from southeastern China using a case– controlled design. We hypothesized that interactions between variants of multiple genes would confer a higher risk of stroke than a single variant in one gene. Our hypothesis was tested using generalized multifactor dimensionality reduction (GMDR) with eight different variants in five candidate genes.
All participants were Chinese Han in origin and unrelated. Participants were excluded from both the experimental and the control groups if they had arteritis, infections, tumors, blood diseases, serious cardiopulmonary, liver, or kidney problems, thyroid disease, or autoimmune diseases. The diagnostic criteria and the collection of the clinical data
The following vascular risk factors for each individual were determined: age, sex, hypertension, DM, cigarette smoking, alcohol intake, total plasma cholesterol, triglycerides, and low-density lipoprotein cholesterol. Hypertension was defined as the mean of three independent measures of blood pressure of at least 140/ 90 mmHg or the use of antihypertensive drugs. DM was diagnosed as when a patient had a fasting glucose level greater than 7.8 mM, greater than 11.1 mM 2 h after oral glucose challenge, or both. A detailed medical assessment of history was performed and information on stroke risk factors was obtained from each participant.
Materials and methods Study populations
Ethical approval for this study was provided by the Ethical Committee of the Third Affiliated Hospital of Wenzhou Medical College. Informed consent was obtained from each patient before study enrollment. The study population included 292 ischemic stroke patients and 259 control individuals. Patients were consecutively recruited between March 2010 and March 2012 from the Third Affiliated Hospital of Wenzhou Medical College. Each diagnosis of a cerebral infarction was made on the basis of the results of a neurological examination using computerized tomography or MRI according to the ninth revision of the International Classification of Diseases. The infarctions were classified according to a modified version of the Trial of Org 10172 in Acute Stroke Treatment classification system [16]. Patients with cerebral infarctions related to atherothrombotic and small artery disease were included. The exclusion criteria were as follows: (a) cardiogenic cerebral embolisms, (b) unclear etiologies, (c) a history of strokes, and (d) a lack of willingness to participate in this study. On the basis of sample-size calculations, we included 292 inpatients with cerebral infarctions from the Department of Neurology, among whom 173 were men and 119 were women, average age 68.56±11.09 years. The control participants were selected from among outpatients with no history of cerebral infarctions as confirmed by assessment of medical history, physical, and laboratory examinations at our center. They had no family history of cerebral infarctions and were not genetically related to the cerebral-infarction patients. A total of 259 volunteers participated in this study, of whom 148 were men and 111 were women, average age 63.92±9.29 years.
Marker selection and genotyping
Ten milliliters of blood was drawn from an arm vein into a sterile tube containing ethylenediaminetetraacetic acid, and stored at – 801C until genotype analysis was carried out. Genomic DNA was extracted from peripheral blood using a modified phenol/chloroform method and purified using the UNIQ-10 kit (Sangon Biotech Co. Ltd, Shanghai, China). Amplification of the target sequences was performed in a multiplex reaction containing 5 ng of DNA, 0.95 ml of water, 0.625 ml of PCR buffer containing 15 mM MgCl2, 1 ml of 2.5 mM dNTP, 0.325 ml of 25 mM MgCl2, 1 ml of PCR primers (Table 1), and 0.5 U HotStar Taq (Qiagen Co. Ltd, Shanghai, China). The reaction was incubated at 941C for 15 min, followed by 45 cycles at 941C for 20 s, 561C for 30 s, and 721C for 1 min, and a final incubation at 721C for 3 min. After PCR amplification, the remaining dNTPs were dephosphorylated by adding 1.53 ml water, 0.17 ml of SAP buffer, and 0.3 U of shrimp alkaline phosphatase (Sequenom Inc., San Diego, California, USA). After incubation at 371C for 40 min, the enzyme was deactivated by incubating at 851C for 5 min. The extension primers were then added to the reaction in a buffer containing 0.755 ml water, 0.2 ml of 10 iPLEX buffer, 0.2 ml termination mix, 0.041 ml of iPLEX enzyme (Sequenom Inc.), and 0.804 ml of 10 mM extension primer. The single base extension reaction consisted of an initial denaturation at 941C for 30 s and then 941C for 5 s, followed by five cycles of 521C for 5 s and 801C for 5 s for a total of 40 cycles. A final extension at 721C was performed for 3 min. The reaction mix was desalted by adding 6 mg of cation exchange resin (Sequenom Inc.), mixed, and resuspended
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
454
NeuroReport 2014, Vol 25 No 7
in 25 ml of water. Once the primer extension reaction was completed, the samples were spotted onto a 384-well spectroCHIP (Sequenom Inc.) using MassARRAY Nanodispenser (Sequenom Inc.) and genotyped using the matrix-assisted laser desorption ionization time-of-flight mass spectrometer. Genotype calling was performed in real time using MassARRAY RT software, version 3.0.0.4 and analyzed using the MassARRAY Typer software, version 3.4 (Sequenom Inc.). Eight variants in five candidate genes were examined for both the patients and the control participants: SG13S32 (rs9551963), SG13S42 (rs4769060), SG13S89 (rs4769874), SG13S114 (rs10507391) variants of the 5-LOX-activating protein (ALOX5AP) gene, the G860A (rs751141) variant of the soluble epoxide hydrolase (EPHX2) gene, the A1075C (rs1057910) variant of the CYP2C9*2 gene, the C430T (rs1799853) variant of the CYP2C9*3 gene, and the A6986G (rs776746) variant of the CYP3A5 gene (Table 1). Statistical analysis
Summary statistics are expressed as the mean±SEM or as percentages. The w2-test was used to analyze the deviation of Hardy–Weinberg equilibrium for genotype frequencies and to compare genotype frequencies. Continuous variables were compared between patients with ischemic stroke and control participants using Student’s t-test. Discrete variables were compared using w2 tests or Fisher’s exact test. For gene–gene interaction analyses, the GMDR method was applied. As the participants were of the same ethnicity, sex and age were modeled as covariates. The GMDR computes the maximum-likelihood estimates and the scores of all individuals under the null hypothesis. Then, the cumulative score is calculated within each multifactor cell, which is labeled either as high risk if the average score meets or exceeds a preassigned threshold of 0 or as low risk if the score is less than 0. An exhaustive search of all possible one-locus to seven-locus models was performed for all variants. The model with the minimum prediction error, the maximum cross-validation consistency score, and a P value of 0.05 or lower automatically derived from the sign test in GMDR software was Table 1
considered to be the best model. This was confirmed by a permutation test implemented in the GMDR software as well [17]. Furthermore, we carried out multivariate logistic regression analyses adjusted for some risk factors to assess the independent contribution of gene–gene interactions to a risk of stroke. Adjusted odds ratios with 95% confidence intervals were computed. A P value of less than 0.05 was considered statistically significant. The GMDR v0.7 program was used in this study (http:// www.healthsystem.virginia.edu/internet/addiction-genomics/Software). All other analyses were carried out using SPSS 16.0 (SPSS Inc., Chicago, Illinois, USA).
Results General clinical characteristics of the participants
The demographic characteristics of the control participants and ischemic stroke patients are presented in Table 2. Briefly, the mean age of the stroke patients was 68.56±11.09 years and that of the control participants was 63.92±9.29 years, and 59.2% of the stroke patients and 57.5% of the control participants were men. There was a significant difference in the age of the participants in the two groups (P < 0.001), but not in their sex. As expected, stroke patients had a higher prevalence of risk factors for vascular diseases, including a history of hypertension (P < 0.01) and diabetes (P = 0.032). However, differences in conventional risk factors including smoking and alcohol intake and levels of LDL-C, total plasma cholesterol, and triglycerides did not reach statistical significance between the two groups (P > 0.05) (Table 2).
Association between specific single-nucleotide polymorphism genotypes and risk of stroke
The single-nucleotide polymorphisms and their genotype frequencies are shown in Table 3. The genotype distributions of the eight candidate genes examined in this study were consistent with the Hardy–Weinberg equilibrium model (P > 0.05). However, there were no statistically significant differences in the frequencies of the genotypes of the eight candidate genes between the stroke patients and the control participants (P > 0.05).
Amplification of alleles using the sense and antisense primers
SNPs SG13S32 (rs9551963) SG13S42 (rs4769060) SG13S89 (rs4769874) SG13S114 (rs10507391) C430T (rs1799853) A1075C (rs1057910) A6986G (rs776746) G860A (rs751141)
Upstream primer
Downstream primer
ACGTTGGATGGGGTTCAAGAGAGAAATTCC ACGTTGGATGGAAGGGTAGAAGTGTCTCAG ACGTTGGATGCACCAGGGAGCAAGCATTAG ACGTTGGATGTCCAGATGTATGTCCAAGCC ACGTTGGATGCTGCGGAATTTTGGGATGGG ACGTTGGATGCTACACAGATGCTGTGGTGC ACGTTGGATGGATGAAGGGTAATGTGGTCC ACGTTGGATGTTTTCTAGATCCCTGCTCTG
ACGTTGGATGAGTTCTTGACCTCACCAACC ACGTTGGATGCGTGGTAATGGGTTTTGAGG ACGTTGGATGTTTCAGGCATGCTCTGCACC ACGTTGGATGCTCTTAAGGTAGGTCTATGG ACGTTGGATGACCCACCCTTGGTTTTTCTC ACGTTGGATGTGTCACAGGTCACTGCATGG ACGTTGGATGCCATAATCTCTTTAAAGAGC ACGTTGGATGAGCAGATGACTCTCCATAGC
SNP, single-nucleotide polymorphism.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Gene variants increase stroke risk Chi et al. 455
Table 2
Clinical characteristics of the participants
Characteristics Age (years) Men (%) Hypertension (%) Diabetes mellitus (%) Cigarette smoking (%) Alcohol intake (%) TC (mM) LDL-C (mM) TG (mM)
Stroke patients (n = 292)
Control participants (n = 259)
P value
68.56±11.09* 59.2 80.1* 33.6* 44.3 46.6 5.25±1.11 3.14±1.25 1.85±1.04
63.92±9.29 57.5 23.2 25.9 41.7 45.2 5.07±1.12 2.96±1.16 1.73±0.99
< 0.001 0.729 < 0.001 0.032 0.547 0.797 0.089 0.098 0.078
Age, TC, TG, and LDL-C PLT values are given as the mean±SD. The other values are given as the percentage of the group. LDL-C, low-density lipoprotein cholesterol; PLT, platelets; TC, total cholesterol; TG, triglycerides. *P < 0.05 versus control.
Table 3
Association between specific SNP genotypes and risk of stroke
SNPs
Genotype
SG13S32
AA AC CC AA AG GG AA AG GG AA AT TT CC AA AC CC AA AG GG GG AG AA
SG13S42
SG13S89
SG13S114
C430T A1075C
A6986G
G860A
Control participants [n (%)] 101 122 36 103 122 34 0 4 255 36 134 89 259 140 19 0 28 92 139 165 81 13
Stroke patients [n (%)]
(39.0) (47.1) (13.9) (39.8) (47.1) (13.1) (0) (1.5) (98.5) (13.9) (51.7) (34.4) (100) (92.7) (7.3) (0) (10.8) (35.5) (53.7) (63.7) (31.3) (5.0)
103 148 41 111 142 39 0 12 280 50 149 93 292 266 25 1 31 111 150 184 97 11
P value
(35.3) (50.7) (14.0) (38.0) (48.6) (13.4) (0) (4.1) (95.9) (17.1) (51.0) (31.9) (100) (91.1) (8.6) (0.3) (10.6) (38.0) (51.4) (63.0) (33.2) (3.8)
0.646
0.914
0.081
0.55
– 0.554
0.829
0.717
SNP, single-nucleotide polymorphism.
Table 4 Comparison of the best models, prediction accuracies, cross-validation consistencies, and P values identified by GMDR for ischemic stroke Best model
Training balanced accuracy
Testing balanced accuracy
1a 1,2 1,2,3 1,2,4,5 1,2,3,4,5 1,2,3,4,5,6 1,2,3,4,5,6,7
0.5399 0.5820 0.6165 0.6627 0.7210 0.7500 0.7566
0.4710 0.5632 0.4532 0.4331 0.4867 0.4925 0.4865
Sign test (P) 3 9 2 2 5 5 4
(0.9453) (0.0107) (0.9893) (0.9893) (0.9893) (0.6230) (0.8281)
Cross-validation consistency 7/10 10/10 5/10 3/10 10/10 10/10 10/10
GMDR, generalized multifactor dimensionality reduction. a SG13S114, A6986G, EPHX2, SG13S32, SG13S42, A1075C, and SG13S89 are symbolized as 1–7, respectively.
Gene–gene interactions
High-order interactions were explored for ischemic stroke using the GMDR method. With covariate adjustments, the best model for ischemic stroke included SG13S114 and A6986G was scored 10/10 for cross-validation
consistency and 9/10 by the sign test (P = 0.0107), as shown in Table 4. Furthermore, the significant interactions among the above two-locus models were confirmed by a permutation test (P = 0.027). A one-locus model was also computed for each variant and the minimum P value
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
456
NeuroReport 2014, Vol 25 No 7
Fig. 1
SG13S114 AT
AA
TT
AA
A6986GG
3.8
−1.3
2.6
20.0 GA 2.4
−4.4
3.9
−16.6
12.8
−6.7
−2.3
−9.3
−29.4 21.0 GG
15.2 9.3
−15.7
−5.4
In each cell, the left bar represents a positive score and the right bar represents a negative score. High-risk cells are indicated by dark shading, low-risk cells by light shading.
Table 5 Multiple regression analysis of the major risk factors for cerebral infarction Risk factors Age Hypertension Diabetes mellitus Interactive variable
Wald
OR
95% CI
P value
12.017 136.521 0.408 7.420
1.660 13.037 0.859 1.804
1.246–2.210 8.474–20.055 0.540–1.369 1.180–2.759
0.001 0.000 0.523 0.006
CI, confidence interval; OR, odds ratio.
was 0.9453, suggesting that their contribution to stroke risk was because of the joint action of the two genes, but not a single locus. Figure 1 shows the best model for ischemic stroke identified by GMDR. The three interactions making larger contributions to this model were between SG13S114 AA and A6986G AA, SG13S114 AA and A6986G GG, and SG13S114 TT and A6986G AA. Logistic regression analysis
As shown in Fig. 1, the relative risk of the nine combinations of the two genes was considered as an interactive variable, with high risk assigned a 1 and low risk assigned a 0. Table 5 shows that certain gene–gene interactions predict a significantly higher risk of cerebral infarction (adjusted for age, hypertension, DM; odds ratio = 1.804, 95% confidence interval 1.180–2.759, P = 0.006).
Discussion The results that we presented in this study do not support a single locus within the eight candidate genes as
a risk gene for ischemic stroke. Rather, using the GMDR method to search gene–gene interactions that predict a high risk of stroke, we principally found that interactions between ALOX5AP SG13S114 and CYP3A5 A6986G were significant combinational predictors of ischemic stroke. There was a 1.804-fold increased risk for ischemic stroke in individuals with these combined genetic factors. Our findings add to the evidence that interactions between genes with marginal effects in their own rights can increase the risk of common and complex diseases, such as stroke. Despite our experiments, the nature of the two-factor interactions between ALOX5AP and CYP3A5 remains unclear. As stroke does not have a typical pattern of Mendelian inheritance, our observations could be considered as quite expected. One possible explanation for the two-factor interaction is that these two genes are all involved in AA metabolism and its impaired equilibrium. In addition to cyclooxygenase, the LOX and CYP pathways have been identified as a major pathway for AA metabolism. ALOX5AP encodes 5-LOX-activating protein, which activates arachidonate 5-lypoxygenase and leads to leukotriene synthesis. Leukotrienes are proinflammatory mediators, secreted by various types of inflammatory cells that cluster at injured sites in blood vessels, which change vascular permeability and promote atherosclerosis formation [18]. After ALOX5AP gene mutations, coding proteins can increase 5-LOX activity, promote leukotriene-mediated damage to vascular walls, and form vulnerable plaques, which lead to ischemic strokes. A meta-analysis has shown that ALOX5AP and SG13S114 mutations are independent risk factors in Iberian populations for ischemic stroke [19]. In contrast, a study in Chinese patients reported that the SG13S114A allele is not an independent risk gene for ischemic infarction [20]. The results are consistent with our single-locus findings. AA is metabolized in endothelial cells by CYP epoxygenases into four EETs, which have diverse cardiovascular protective effects, including antiapoptotic effects in endothelial cells and anti-inflammatory and antiangiogenic effects, suggesting antiartheroscelerotic effects of EETs that may be beneficial for stroke treatment [21–23]. The CYP3A5 enzyme is a type of epoxygenase enzyme. CYP3A5 gene mutations can reduce EET generation [24]. A6986G is the most common mutation of CYP3A5. As both the ALOX5AP SG13S114 and the CYP3A5 A6986G genes are involved in AA metabolism, their separate effects on leukotrienes and EETs can eliminate the dynamic equilibrium of AA metabolism. Furthermore, a CYP3A5 genetic polymorphism has been investigated with respect to the risk for high blood pressure [25]. The mechanism underlying this
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Gene variants increase stroke risk Chi et al. 457
association may be related to sodium reabsorption. As hypertension is the most important risk factor of cerebral infarction, this CYP3A5 gene polymorphism may also be associated with atherosclerosis and cerebral infarction. Therefore, the interaction between these two genes may lead to the progression of atherosclerosis and increase susceptibility to cerebral infarction through a common mechanism of action.
8
9
10 11
Conclusion In the present study, we show that interactions between the ALOX5AP SG13S114 and CYP3A5 A6986G genes can engender a significantly higher risk of stroke than a single risk factor. This gene–gene interaction may serve as a novel area for stroke research. The two-locus combination may provide a target for prevention of this disease.
12 13
14
15
Acknowledgements This work was supported by the Zhejiang Provincial Natural Science Foundation (No. Y2110379) and the Zhejiang Province Science and Technology Research Foundation (No. 2012c33106) of China. The authors thank all the nurses in the Department of Neurology of the Third Affiliated Hospital of Wenzhou Medical College.
16
17
18
Conflicts of interest
There are no conflicts of interest.
19
References
20
1
2
3 4 5
6 7
Rothwell PM, Coull AJ, Giles MF, Howard SC, Silver LE, Bull LM, et al. Change in stroke incidence, mortality, case-fatality, severity, and risk factors in Oxfordshire, UK from 1981 to 2004 (Oxford Vascular Study). Lancet 2004; 363:1925–1933. Jiang B, Wang WZ, Chen H, Hong Z, Yang QD, Wu SP, et al. Incidence and trends of stroke and its subtypes in China: results from three large cities. Stroke 2006; 37:63–68. Culverhouse R, Suarez BK, Lin J, Reich T. A perspective on epistasis: limits of models displaying no main effect. Am J Hum Genet 2002; 70:461–471. Stoll G, Bendszus M. Inflammation and atherosclerosis: novel insights into plaque formation and destabilization. Stroke 2006; 37:1923–1932. Morisseau C, Hammock BD. Epoxide hydrolases: mechanisms, inhibitor designs, and biological roles. Annu Rev Pharmacol Toxicol 2005; 45:311–333. Calder PC. Polyunsaturated fatty acids and inflammatory processes: new twists in an old tale. Biochimie 2009; 91:791–795. Funk CD. Prostaglandins and leukotrienes: advances in eicosanoid biology. Science 2001; 294:1871–1875.
21
22
23
24 25
Jerrard-Dunne P, Sitzer M, Risley P, Buehler A, von Kegler S, Markus HS. Inflammatory gene load is associated with enhanced inflammation and early carotid atherosclerosis in smokers. Stroke 2004; 35:2438–2443. Jerrard-Dunne P, Sitzer M, Risley P, Steckel DA, Buehler A, von Kegler S, et al. Interleukin-6 promoter polymorphism modulates the effects of heavy alcohol consumption on early carotid artery atherosclerosis: the Carotid Atherosclerosis Progression Study (CAPS). Stroke 2003; 34:402–407. Topol EJ, Smith J, Plow EF, Wang QK. Genetic susceptibility to myocardial infarction and coronary artery disease. Hum Mol Genet 2006; 15:117–123. 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. Zeldin DC. Epoxygenase pathways of arachidonic acid metabolism. J Biol Chem 2001; 276:36059–36062. Larsen BT, Gutterman DD, Hatoum OA. Emerging role of epoxyeicosatrienoic acids in coronary vascular function. Eur J Clin Invest 2006; 36:293–300. Daikh BE, Lasker JM, Raucy JL, Koop DR. Regio- and stereoselective epoxidation of arachidonic acid by human cytochromes P450 2C8 and 2C9. J Pharmacol Exp Ther 1994; 271:1427–1433. Rifkind AB, Lee C, Chang TK, Waxman DJ. Arachidonic acid metabolism by human cytochrome P450s 2C8, 2C9, 2E1, and 1A2: regioselective oxygenation and evidence for a role for CYP2C enzymes in arachidonic acid epoxygenation in human liver microsomes. Arch Biochem Biophys 1995; 320:380–389. Han SW, Kim SH, Lee JY, Chu CK, Yang JH, Shin HY, et al. A new subtype classification of ischemic stroke based on treatment and etiologic mechanism. Eur Neurol 2007; 57:96–102. Ritchie MD, Hahn LW, Roodi N, Bailey LR, Dupont WD, Parl FF, et al. Multifactor-dimensionality reduction reveals high-order interactions among estrogen-metabolism genes in sporadic breast cancer. Am J Hum Genet 2001; 69:138–147. Dichgans M. Genetics of ischaemic stroke. Lancet Neurol 2007; 6: 149–161. Domingues-Montanari S, Fernandez-Cadenas I, del Rio-Espinola A, Corbeto N, Krug T, Manso H, et al. Association of a genetic variant in the ALOX5AP with higher risk of ischemic stroke: a case–control, meta-analysis and functional study. Cerebrovasc Dis 2010; 29:528–537. He YL, Zhu M, Jin XP, Zhou YL. Relationship of polymorphism of SG13S114A/T in ALOX5AP gene with atherosclerotic cerebral infarction. Zhejiang Da Xue Xue Bao Yi Xue Ban 2009; 38:630–633. Heizer ML, McKinney JS, Ellis EF. 14,15-Epoxyeicosatrienoic acid inhibits platelet aggregation in mouse cerebral arterioles. Stroke 1991; 22: 1389–1393. Node K, Huo Y, Ruan X, Yang B, Spiecker M, Ley K, et al. Anti-inflammatory properties of cytochrome P450 epoxygenase-derived eicosanoids. Science 1999; 285:1276–1279. Wang Y, Wei X, Xiao X, Hui R, Card JW, Carey MA, et al. Arachidonic acid epoxygenase metabolites stimulate endothelial cell growth and angiogenesis via mitogen-activated protein kinase and phosphatidylinositol 3-kinase/Akt signaling pathways. J Pharmacol Exp Ther 2005; 314: 522–532. Burk O, Wojnowski L. Cytochrome P450 3A and their regulation. Naunyn Schmiedebergs Arch Pharmacol 2004; 369:105–124. Bochud M, Eap CB, Elston RC, Bovet P, Maillard M, Schild L, et al. Association of CYP3A5 genotypes with blood pressure and renal function in African families. J Hypertens 2006; 24:923–929.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
