Qingli Sun, Yu Fu, Aping Sun, Yanhong Shou, Mei Zheng, Xiaogang Li, Dongsheng Fan

Department of Neurology, Third Hospital of Peking University, Beijing 100191, China

INTRODUCTION

Epidemiological studies have found that genetic factors play an important role in stroke onset[1]. Genes associated with susceptibility to stroke are considered an important topic in current studies regarding the mechanisms underlying stroke.

E-selectin, an important inflammatory cytokine, is closely associated with hypertension, the mechanism of abnormal blood clotting, abnormal lipid metabolism,the formation of atherosclerosis and other risk factors for ischemic stroke[2]. E-selectin genetic polymorphisms affect ligand recognition and binding, significantly increasing the affinity for the ligand, thus influencing disease development[3].

Worldwide studies have found that E-selectin gene polymorphisms contribute to the early onset of atherosclerosis[4], and are a potential risk factor for ischemic stroke.The existing studies[5-8]are mostly single-center with a small sample size and low reliability. Meta-analysis methodology is available for the combination and comprehensive evaluation of several independent studies with the same objective,and compensates for the inadequacy of a single study. This study sought to summarize and analyze the correlation between E-selectin and ischemic stroke through a meta-analysis, in a broader attempt to effectively evaluate their relationship and provide evidence-based medical evidence.

DATA AND METHODS

Retrieval strategy

Using key words of ＂E-selectin, stroke,ischemic stroke, cerebral infarction, gene,polymorphism, genetic polymorphism＂ in English, a computer-based online retrieval of PubMed, Elsevier, Ovid database, the China National Knowledge Infrastructure and the Wanfang database was conducted.Papers were also retrieved personally from reference lists to ensure all relevant papers were identified. The literature types included journal papers, conference papers and postgraduate dissertations. Case-controlled studies regarding E-selectin gene polymorphism and ischemic stroke were extracted through the above means. The search was conducted between January 1998 and December 2010.

Inclusion and exclusion criteria

Inclusion criteria

(1) Case-controlled studies published in or before December 2010 pertaining to E-selectin gene polymorphism and ischemic stroke. (2) The number of cases with different genotypes in the experimental and control groups should be directly or indirectly provided.(3) Genotype distribution in the involved subjects should comply with the Hardy-Weinberg genetic equilibrium. (4) For multiple repeated studies, those most recently published or with a large sample size were preferred.

Exclusion criteria

(1) Duplicated research. (2) The control group was not in Hardy-Weinberg genetic equilibrium. (3) Data description or sample information was unclear, and statistical methods were inappropriate. (4) No control group was included. (5) Animal trials. (6) Abstracts, comments,reviews and other non-original research.

Literature evaluation and extraction

The first and third authors independently extracted relevant data from publications, including the name of the first author, the publication year, the nationality of the included subjects, the genetic test methods, and the number of cases with different genotypes. The third party was available to negotiate a solution if a dispute occurred in the data extraction process. The quality of the included literature was evaluated based on the genetic association and the guidelines for meta-analysis proposed by Little et al[9], as follows: (1) efficiency of the genotyping identification method; (2) data of the included subjects; (3) a mixed and stratified population; (4)statistical analysis. The maximum score of the evaluation scale was 26 points, a score of 1 to 8 was poor, 9 to 18 was moderate, and 19 to 26 was good.

Outcome measures

Genotype and allele distributions in the E-selectin gene S128R locus.

Statistical analysis

The distribution of genotypes in the included studies was tested for Hardy-Weinberg equilibrium. The publication bias of the included studies was determined with funnel plot analysis. RVEMAN 5.1 statistical software(International Society of Evidence-Based Medicine) was utilized to calculate OR values of (AC+CC)/AA in all of the S128R sites and OR values of the C/A allele, and their 95% CI. The heterogeneity was tested with q test,and a level of P ＞ 0.10 was considered as no heterogeneity. If heterogeneity was absent, the Peto fixed-effect model was adopted for data consolidation and for calculation of the total OR value; if heterogeneity was significant, the random-effect model was adopted for data consolidation.

RESULTS

Literature retrieval

A total of 655 documents were screened out, leaving nine publications which emphasized E-selectin S128R polymorphisms and ischemic stroke. Two reviews were removed after full texts were read, while the remaining seven studies included two duplicated studies, one of which was also excluded. Finally, six eligible papers were included in the final analysis[5-8,10-11], according to the flowchart in Figure 1. Among the involved six studies,there were 1 220 patients in the ischemic stroke group and 1 283 patients in the control group, and their genotype distribution was consistent with the Hardy-Weinberg genetic equilibrium.

Figure 1 Flowchart of the included studies.

Baseline analysis and quality evaluation

Among the six included studies, four studies were conducted in a Chinese population, while the other two focused on Japanese and Iranian populations. General information of the subjects in the studies is shown in Table 1, and the genotype and allele frequency distributions are presented in Table 2.

The quality of the included literature was evaluated in accordance with the genetic association study and meta-analysis guidelines proposed by Little et al[9]. The publication of Mehran et al[10]was found to be of good quality (scoring 19 points), while the remaining five studies were assessed as medium quality, scoring 11-17 points. The publication bias of the involved literatures was detected with RVEMAN 5.1 statistical software. Funnel plot analysis showed no obvious bias(supplementary Figures 1-2 online).

Heterogeneity of the involved studies

Heterogeneity tests showed that the (AC+CC)/AA genotype in the E-selectin S128R gene A/C locus had χ2= 5.90, I2= 15%, P = 0.32; C/A allele had χ2= 6.32,I2= 21%, P = 0.28. The heterogeneity of four studies in a Chinese population showed that the (AC+CC)/AA genotype in the S128R A/C locus had χ2= 1.57, I2= 0%,P = 0.67; C/A allele had χ2= 1.93, I2= 0%, P = 0.59. The genotype and allele frequency showed no heterogeneity based on the P = 0.10 level.

Table 1 General information of the studies included in the meta-analysis

Table 2 Genotype and allele frequency distributions from the included studies

Meta-analysis on E-selectin genotype and allele distribution with ischemic stroke

According to the heterogeneity test, the Peto fixed-effect model and data combination, OR values were calculated and a forest map was prepared. The meta-analysis showed that the combined OR value of the (AC+CC)/AA genotype in the E-selectin S128R gene A/C locus was 1.93 (95% CI: 1.55-2.41, Z value of OR significance test was 5.80, P ＜ 0.000 01; Figure 2). The OR value of the C/A allele was 1.80 (95% CI: 1.47-2.22, Z = 5.59,P ＜ 0.000 01; Figure 3).

Figure 2 Forest map of the (AC+CC)/AA genotype in the E-selectin S128R gene A/C locus and ischemic stroke.

Figure 3 Forest map of the C/A allele in the E-selectin S128R gene A/C locus and ischemic stroke.

Meta-analysis of E-selectin genotype and allele distribution with ischemic stroke in Chinese subjects

Four studies focusing on Chinese populations were merged and the meta-analysis showed that the combined OR value of the (AC+CC)/AA genotype in the E-selectin S128R gene A/C locus was 2.36 (95% CI:1.68-3.31, Z = 4.99, P ＜ 0.000 01; Figure 4). The OR value of the C/A allele was 2.25 (95% CI: 1.63-3.12, Z =4.89, P ＜ 0.000 01; Figure 5).

Figure 4 Forest map of the (AC+CC)/AA genotype in the E-selectin S128R gene A/C locus and ischemic stroke in a Chinese population.

Figure 5 Forest map of the C/A allele in the E-selectin S128R gene A/C locus and ischemic stroke in a Chinese population.

DISCUSSION

E-selectin, also known as human leukocyte differentiation antigen CD62E, mediates adherence of activated endothelial cells to leukocytes, subsequently inducing leukocyte activation and producing a series of inflammatory reactions, leading to the formation of atherosclerosis[12]. After cerebral ischemia occurs,leukocytes migrate across endothelial cells into the brain parenchyma, secrete various inflammatory mediators to damage brain tissue, and enlarge the ischemic regions[13-16]. Animal experiments showed that E-selectin levels increased after focal cerebral ischemia[17-18].E-selectin levels in ischemic stroke patients in the acute phase were significantly higher than those in a control group, and the alteration contributed to the development of stroke[19].

The human E-selectin gene is about 13 kb long, is located on the long arm of chromosome number 1, and consists of 14 exons and 13 introns. In 1994, Wenzel et al[4]first confirmed the conversion of the E-selectin gene at the 561 position, adenine A →C (A561C),resulting in amino acids at the 128 position changing from serine (Ser) into arginine (Arg) (S128R). The polymorphism was associated with early onset of atherosclerosis in a Caucasian German population.

An increasing number of studies have demonstrated that E-selectin S128R polymorphism correlates with atherosclerosis[2]and coronary heart disease[20-22]. In addition, a correlation between this polymorphism and myasthenia gravis, prostate cancer and other diseases,was also found[23-24].

In this study, all publications regarding E-selectin gene polymorphism and ischemic stroke were fully searched in accordance with the requirements of the meta-analysis.

We combined and summarized several similar studies to increase the sample size, improve test performance, and conduct a more comprehensive analysis. A metaanalysis has several kinds of bias, and publication bias is the most apparent. The funnel plot is an easy way to determine publication bias. In this study, no obvious bias was found through the visual funnel plot.

The current study included 1 220 ischemic stroke patients and 1 283 controls, who were all Asian. Among them, there were 626 Chinese patients with ischemic stroke and 629 Chinese subjects in the control group.

There was no heterogeneity in the genotype and allele frequency among the study results, so the Peto fixed-effect model was adopted for data combination.

The meta-analysis showed that the E-selectin S128R genotype distribution was significantly different in the stroke and control groups. The AC+CC genotype and C allele frequencies in patients with ischemic stroke were significantly higher than those in the control group.

Relative risk analysis of genotype frequency found that the risk of ischemic stroke in the AC+CC genotype was 1.93 times as high as that of the AA genotype. There were significant differences in the allele frequency between the two groups, and the risk of ischemic stroke with the C allele was 1.80 times as high as that of the A allele. Four Chinese studies suggested that the risk of ischemic stroke in subjects with the AC+CC genotype and the C allele was significantly increased. The CC genotype was very rare in both the ischemic stroke group and control group, and was reported in only two studies[5,10].

Previous studies have shown that the AC/CC genotype and C allele may be associated with ischemic stroke, and gene polymorphisms may directly or indirectly affect the production or performance of their product. E-selectin plasma levels in patients carrying the AC+CC genotype were higher than in those carrying the AA genotype[25],and the C allele significantly enhanced thromboxane and cellulose production[26]. E-selectin S128R polymorphism was also found to contribute to the increase in plasma E-selectin levels in patients with acute ischemic stroke[7].

The present study indirectly revealed the mechanism underlying genetic factors in the pathogenesis of ischemic stroke at the gene level.

In summary, the present meta-analysis found that E-selectin gene S128R polymorphism is associated with risk of ischemic stroke, with the AC/CC genotype and C allele conferring susceptibility for ischemic stroke.

However, the studied subjects were Asian, and the quality of the included literature was not very high, so the results have certain limitations. Further studies are needed to conclusively determine the correlation between E-selectin gene polymorphism and ischemic stroke through multicenter, large-scale, multi-ethnic clinical studies.

Author contributions:Qingli Sun, Yu Fu and Aping Sun screened the literature. Yanhong Shou and Mei Zheng evaluated study quality. Qingli Sun was responsible for statistical analysis and wrote the manuscript. Xiaogang Li and Dongsheng Fan were responsible for the study conception and design, and gave guidance in the study.

Conflicts of interest:None declared.

Supplementary information:Supplementary data associated with this article can be found, in the online version, by visiting www.nrronline.org, and entering Vol. 6, No. 22, 2011 item after selecting the “NRR Current Issue” button on the page.

[1]Catto AJ. Genetic aspects of the hemostatic system in cerebrovascular disease. Neurology. 2001;57(5 Suppl 2):S24-30.

[2]Ghilardi G, Biondi ML, Turri O, et al. Ser128Arg gene polymorphism for E-selectin and severity of atherosclerotic arterial disease. J Cardiovasc Surg (Torino). 2004;45(2):143-147.

[3]Revelle BM, Scott D, Beck PJ. Single amino acid residues in the E- and P-selectin epidermal growth factor domains can determine carbohydrate binding specificity. J Biol Chem. 1996;271(27):16160-16170.

[4]Wenzel K, Felix S, Kleber FX, et al. E-selectin polymorphism and atherosclerosis: an association study. Hum Mol Genet. 1994;3(11):1935-1937.

[5]Wang F, Liu RZ, Liu ZW. Association of plasma E-selectin, S128R gene polymorphisms and cerebral infarction in Shanxi Han population. Zhongxiyi Jiehe Xinnao Xueguan Bing Zazhi. 2010;8(5):549-551.

[6]Chen JN, Wei GY, Fu XL, et al. Relationship of E-selectin A561C gene polymorphisms in patients with cerebral infarction and its effect on plasma lipid level. Zhongguo Xiandai Yixue Zazhi. 2010;20(16):2468-2470.

[7]Zhang YL, Tang TT, Fang BY. Polymorphisms of E-selectin S128R gene and genetic susceptibility of cerbral thrombosis. Zhongguo Xiandai Yixue Zazhi. 2009;19(13):1989-1991.

[8]Wei YS, Xu QQ, Meng LQ, et al. Polymorphisms of E-selectin gene and genetic susceptibility of ischemic stroke in Zhuang nationality of Guangxi province. Disan Junyi Daxue Xuebao. 2005;27(21):2160-2163.

[9]Little J, Bradley L, Bray MS, et al. Reporting, appraising, and integrating data on genotype prevalence and gene-disease associations. Am J Epidemiol. 2002;156(4):300-310.

[10]Mehran H, Hajilooi M, Rafiei AR, et al. E-selectin genetic variation as a susceptibility factor for ischemic stroke. Cerebrovasc Dis.2009;28(1):26-32.

[11]Hattori H, Sato H, Ito D, et al. A561C polymorphism of E-selectin is associated with ischemic cerebrovascular disease in the Japanese population without diabetes mellitus and hypercholesterolemia. Brain Res. 2006;1108(1):221-223.

[12]Yoshida M, Takano Y, Sasaoka T, et al. E-selectin polymorphism associated with myocardial infarction causes enhanced leukocyte-endothelial interactions under flow conditions.Arterioscler Thromb Vasc Biol. 2003;23(5):783-788.

[13]Kozuka K, Kohriyama T, Nomura E, et al. Endothelial markers and adhesion molecules in acute ischemic stroke-sequential change and differences in stroke subtype. Atherosclerosis. 2002;161(1):161-168.

[14]Mussack T, Biberthaler P, Gippner-Steppert C, et al. Early cellular brain damage and systemic inflammatory response after cardiopulmonary resuscitation or isolated severe head trauma: a comparative pilot study on common pathomechanisms.Resuscitation. 2001;49(2):193-199.

[15]Omi H, Okayama N, Shimizu M, et al. Participation of high glucose concentrations in neutrophil adhesion and surface expression of adhesion molecules on cultured human endothelial cells: effect of antidiabetic medicines. J Diabetes Complications.2002;16(3):201-208.

[16]Rao RM, Haskard DO, Landis RC. Enhanced recruitment of Th2 and CLA-negative lymphocytes by the S128R polymorphism of E-selectin. J Immunol. 2002;169(10):5860-5865.

[17]Zhang R, Chopp M, Zhang Z, et al. The expression of P- and E-selectins in three models of middle cerebral artery occlusion.Brain Res. 1998;785(2):207-214.

[18]Haring HP, Berg EL, Tsurushita N, et al. E-selectin appears in nonischemic tissue during experimental focal cerebral ischemia.Stroke. 1996;27(8):1386-1392.

[19]Jilma B, Marsik C, Kovar F, et al. The single nucleotide polymorphism Ser128Arg in the E-selectin gene is associated with enhanced coagulation during human endotoxemia. Blood.2005;105(6):2380-2383.

[20]Tripathi R, Singh PK, Tewari S, et al. Genetic predisposition of E-selectin gene (S128R) polymorphism in patients with coronary artery disease (CAD). Indian J Med Res. 2009;130(4):423-427.

[21]Zak I, Sarecka B, Krauze J. Synergistic effects between 561A ＞ C and 98G ＞ T polymorphisms of E-selectin gene and hypercholesterolemia in determining the susceptibility to coronary artery disease. Heart Vessels. 2008;23(4):257-263.

[22]Li Y, Wei YS, Wang M, et al. Association between the Ser128Arg variant of the E-selectin and risk of coronary artery disease in the central China. Int J Cardiol. 2005;103(1):33-36.

[23]Chen H, Cui B, Wang S, et al. The common variants of E-selectin gene in Graves' disease. Genes Immun. 2008;9(2):182-186.

[24]Panoussopoulos GS, Theodoropoulos G, Michalopoulos NV.Analysis of E-Selectin S128R gene polymorphism in pancreatic cancer. J Surg Oncol. 2010;102(6):604-607.

[25]Cherian P, Hankey GJ, Eikelboom JW, et al. Endothelial and platelet activation in acute ischemic stroke and its etiological subtypes. Stroke. 2003;34(9):2132-2137.

[26]Mlekusch W, Exner M, Schillinger M, et al. E-Selectin and restenosis after femoropopliteal angioplasty: prognostic impact of the Ser128Arg genotype and plasma levels. Thromb Haemost.2004;91(1):171-179.