A Novel Risk Haplotype of ALOX5AP Gene is Associatedwith Ischemic Stroke in Chinese Han Population

Dongzhi Yang & Ying He & Manyu Li & Congcong Shi &Guoying Song & Qing Wang & Yujia Fan &

Qingchuan Feng & Hong Zheng

Received: 26 August 2013 /Accepted: 9 October 2013# Springer Science+Business Media New York 2013

Abstract Previous studies have implicated that two at-risk haplotypes (HapA and HapB) of gene-encoding5-lipoxygenase-activating protein (ALOX5AP) weresignificantly associated with stroke. The aim of this studywas to explore the association between haplotypes ofALOX5AP gene and risk for ischemic stroke (IS) in ChineseHan population. A total of 492 patients with IS and 490matched control subjects were recruited. Six ALOX5APSNPs (SG13S377, SG13S114, SG13S41, SG13S89,SG13S32 and SG13S35) were genotyped by SNaPshotminisequence technique. A common genetic variantSG13S114/AA in the ALOX5AP gene was associated withIS in this Chinese cohort (OR=2.514, 95 % CI=1.667~3.790). HapA (TGA) and HapB (AAAG) had no significantdifference in the patients (36.3 and 18.5 %, respectively)and controls (37.6 and 16.3 %, respectively) (P=0.631 andP=0.375, respectively). But, the frequency of Hap (GAAG)was significantly higher in the patients than that in the controlsafter Bonferroni's adjustment (P =0.006). To conclude,SG13S114/AA of the ALOX5AP gene was associated withan increased risk for IS. A novel risk haplotype, Hap (GAAG)was a genetic risk factor for IS in this Chinese population.

Keywords ALOX5AP . Haplotype . Ischemic stroke .

Polymorphism

Introduction

Ischemic stroke (IS) is a complex multifactorial disorder, andthe process of developing IS is generally attributed toatherosclerosis with arterial wall inflammation that ultimatelyleads to plaque rupture, fissure or erosion (Zintzaras et al.2009). It has been reported that increased 5-lipoxygenase-activating protein (ALOX5AP) activity could lead to theaccumulation of leukotrienes in fatty deposits on the arterialwall. The subsequent breakdown of these deposits by theimmune system may then lead to the development ofatherosclerosis and an increased risk for IS (Quarta et al. 2009).

In the initial study, Helgadottir et al. showed that a four-single-nucleotide polymorphism (SNP) haplotype, HapA(SG13S25, SG13S114, SG13S89 and SG13S32), wasstrongly associated with a nearly twofold greater risk formyocardial infarction (MI) and stroke in an Icelandicpopulation. In addition, another association of a differentfour-SNP haplotype, HapB (SG13S114, SG13S377,SG13S41 and SG13S35), with risk for MI was reported in aBritish population (Helgadottir et al. 2004). Several studieshave tried to replicate this association in different populations.A study in an Italian population revealed an associationbetween HapB and an increased risk for coronary arterydisease (CAD). Haplotype analyses in this populationadditionally revealed a significant association between CADand a new haplotype named HapC (Girelli et al. 2007).However, Zee et al. (2006) found no association of HapA orHapB with stroke in a US population. Lõhmussaar et al.(2005) indicated a lack of association of HapA with IS in acentral European population. To date, a few studies on theassociation between the ALOX5AP gene polymorphism andthe risk for stroke in Chinese Han population were available(Sun et al. 2011; Wang et al. 2011). However, the associationresults were still inconsistent and controversial. Therefore, the

D. YangDepartment of Bioengineering, Zhengzhou University, Zhengzhou,450052, China

Y. He (*) :M. Li :C. Shi :G. Song :Q. Wang :Y. Fan :Q. Feng :H. Zheng (*)Department of Medical Genetics, Basic Medical College,Zhengzhou University, Zhengzhou, 450052, Chinae-mail: heying39@zzu.edu.cne-mail: hzheng@zzu.edu.cn

J Mol NeurosciDOI 10.1007/s12031-013-0147-9

aim of the present study is to determine the associationbetween the haplotypes of ALOX5AP gene and IS in centralChinese Han population.

Materials and Methods

Study Population

A total of 492 patients with ischemic stroke (males/females=290:202, mean age 56.7±8.3 years) were enrolled fromHenanProvincial Hospital in central China, which is a high-incidence region of IS. The IS was defined as a loss of globalor focal cerebral function persisting for >24 h withcorresponding infarction on brain imaging with a probablevascular cause (Saleheen et al. 2005). IS subtypes wereclassified based on the Trial of Org10172 in Acute StrokeTreatment (TOAST) classification (Adams et al. 1993). Brainimaging with computer tomography (CT) and/or magneticresonance imaging (MRI) as well as ancillary diagnosticinvestigations and standardized blood tests were completed.Patients with arterial fibrillation, cerebral hemorrhage,peripheral vascular diseases or kidney diseases were excludedin this study.

The control group consisted of 490 individuals (males/females=268:222, mean age 56.2±8.9 years) who wereselected from the same demographic area and were matchedto cases on the basis of age (±5 years), gender and residency.All the controls were unrelated native Henan Han and withoutcerebrovascular and cardiovascular diseases, cancer, hepaticor renal diseases. Signed consent formwas obtained from eachparticipant. The study protocols were approved by the EthicsCommittee on Human Research of Zhengzhou University.

Genotyping of ALOX5AP Genomic Variants

Genomic DNA was extracted from peripheral white bloodcells using the phenol/chloroform method (Sangon, Shanghai,China). Six ALOX5AP SNPs (SG13S114, SG13S377,SG13S41, SG13S35, SG13S89 and SG13S32) were selectedfrom which HapA and HapB are derived. Of the four HapASNPs (SG13S25, SG13S114, SG13S89 and SG13S32),SG13S25 was not selected because this SNP was shown tobe monomorphic in a large Chinese cohort. Thus, we selectedthree SNPs SG13S114, SG13S89 and SG13S32 from HapAfor this study. The four HapB SNPs (SG13S114, SG13S377,SG13S41 and SG13S35) were all included in this study.Touchdown PCR amplifications were performed using thefollowing six pairs of primers (Sangon, Shanghai, China).All primers, as listed in Table 1, were designed using thePrimer3 program (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi). All the samples were genotyped bySNaPshot minisequence technique.

A 20 μl mixture was prepared for each reaction, including1× HotStarTaq buffer, 2.8 mMMg2+, 0.3 mM dNTP, 0.1 μMof SG13S144F/R, SG13S89F/R, SG13S32F/R andSG13S41F/R, 0.2 μM of SG13S35F/R, 0.3 μM ofSG13S377F/R, 1 U HotStarTaq polymerase (Qiagen) and1 μl template DNA. The cycling program was 95 °C for15 min, 9 cycles of 94 °C for 20 s, 62–0.5 °C/cycle for 40 s,72 °C for 1.5 min, 28 cycles of 94 °C for 20 s, 57 °C for 30 s,72 °C for 1.5 min and 72 °C for 5 min. In order to eliminatethe excess of primers and dNTPs, 2 U SAP (Promega, USA)and 1 U Exo I (EpiCentre, Palmerston North, New Zealand)were added into 8 μl PCR products. The mixture wasincubated at 37 °C for 80 min, followed by incubation at75 °C for 15 min, and then multiplex PCR SNaPshot reactionwas performed. The extension primers were as follows:

SG13S114SF: ccccccccccccccccccccCAAGCCTCTCTTTGCAATTCTASG13S89SR: ccccccccccAGCATTAGCAATGCATTATCACASG13S32SF: cccccccccccccccccccCAACCGAGGAGGAATTGCTSG13S41SF: cccccccAGTCCCATTCTGAGGAACTGAGSG13S35SF: CCTGGGATGTGGTCCTTTCSG13S377SR: ccccATCACAAAACTGTGGGAGGC

The reaction mixture included 3 μl SNaPshot mix (AppliedBiosystems, USA), 2 μl extension primer mix (0.1 μM forSG13S41SF and 0.4 μM for others) and 1 μl purified PCRproduct. The cycling program was 96 °C for 1 min, 28 cyclesof 96 °C for 10 s, 52 °C for 5 s and 60 °C for 30 s. In order topurify the extension products, 1 U SAP (Promega) was addedto the extension product and incubated at 37 °C for 60 min,followed by incubation at 75 °C for 15 min. Finally, a mixtureof 9 μl HiDi formamide, 0.25μl Liz120 size standard and 1 μlpurified extension product was denatured at 95 ° for 5 min,

Table 1 Primer sequences (5′–3′) used for PCR

SNP Primer sequence Segment (bp)

SG13S377F TGCACCACTATGCCCATCTAA 223

SG13S377R AATGAAGCAAATGACCCATGC

SG13S114F CCTCTGTCCCTCCATTGTCAC 373

SG13S114R GAATGGCATTTTGGGGTATGA

SG13S41F GTGTCCAAATCTCCCCTTCTT 178

SG13S41R GGCTGAATTAGGTCCCTTCCA

SG13S35F GTCTGATGGTCCAGGCTGAAG 627

SG13S35R CCCAGGATCATCCCAGTTGTA

SG13S89F AGATGCGTACCCCACTTTCCT 473

SG13S89R GTGGAAACATGCCTGAAGGAG

SG13S32F AGATTTTCAACCCTGCCGTCT 305

SG13S32R TCTTCCCTACCCACTGGATCA

J Mol Neurosci

quenched on ice for 2 min at minimum and then loaded onABI3730xl (ABI, USA). The GeneMapper4.0 (AppliedBiosystems) was ran to analyse the results.

Statistical Analysis

The statistical tests were performed with SPSS 16.0 softwarepackage (SPSS Inc., USA). All the continuous variables wereexpressed as mean value±SD. The differences in the allele andgenotype frequencies were calculated using the chi-square test.Testing for deviation of genotype distribution from Hardy–Weinberg equilibrium and haplotype-based case–control studywas performed using the SHEsis software (http://analysis.bio-x.cn) (Shi and He 2005). In addition, logistic regression wascarried out to adjust for confounding factors. Bonferroni'sadjustment was used for multiple comparisons. The adjustedP value for significance was set at 0.05/6=0.008. Odds ratiowith 95 % confidence intervals (95 % CI) was calculated to testthe association between risk factors and IS.

Results

Characteristics of the Subjects

The characteristics of the study population were described inTable 2. The cases and controls appear well matched in ageand sex. Compared with those of the control group, thepercentages of hypertension, diabetes mellitus and smokerswere higher in the IS group. In addition, the IS patients hadsignificantly higher total cholesterol and total triglyceridesthan the control subjects.

Association Analysis Between the ALOX5APPolymorphisms and IS

No significant deviation from the Hardy–Weinbergequilibrium test was found at all six SNPs in the IS group

and controls (P >0.05). After multivariate logistic regressionanalysis and Bonferroni's adjustment were used to adjust theeffects of the conventional risk factors, both the SG13S114homozygous A/A genotype frequency (18.7 %) and A allelefrequency (41.3 %) in the IS group were significantly higherthan those (9.0 %, 32.7 %) in the control group (P=0.000)(Table 3). A genotype association test with dominant,recessive and additive models were performed, and SG13S114was associated with IS in the recessive and additive geneticmodels (P=0.000, Table 4). Furthermore, the SG13S377,SG13S89, SG13S32 and SG13S41 polymorphisms showedno significant difference (P>0.008, Table 3).

Linkage Disequilibrium Test and Haplotype Analysis

Linkage disequilibrium test was performed, and the resultswere shown in Fig. 1. The SG13S114 and SG13S377 SNPs,and the SG13S377 and SG13S32 SNPs were in strong linkagedisequilibrium (D ′=0.968 and D ′=0.906, respectively). Thedistribution of HapA and HapB has no significant differencebetween the IS group and control (P=0.631 and P=0.375,respectively) (Tables 5 and 6). However, the frequencies ofHap (GAAG) were still significantly higher in the IS patientsthan those in the controls after Bonferroni's adjustment(P=0.006); see Table 6.

Discussion

ALOX5AP gene is located in the chromosome 13q12-13,containing five exons and four introns with a size of 31 kb.ALOX5AP encodes arachidonate 5-lipoxygenase activatingprotein, which plays a key role in the biosynthesis of pro-inflammatory leukotriene B mediators, providing a potentiallink between inflammation and cardiovascular disease(Linsel-Nitschke et al. 2008). ALOX5AP is a good candidategene underlying the association with IS, although bothpositive and negative associations have been described invarious ethnic populations (Ström et al. 2012; Barral et al.2012). The complex entity and heterogeneity of IS might beresponsible for the conflicting findings among studies.

The present study examined the association between thevariants of ALOX5AP and the risk for ischemic stroke incentral Chinese Han population. For the SG13S114polymorphism, SG13S114 A/A genotype frequencies(18.7 %) and A allele frequencies (41.3 %) of the IS groupwere significantly higher than those (9.0 and 32.7 %,respectively) of the control group (P <0.008). Multivariatelogistic regression analysis result showed that theSG13S114AA genotype was still associated with a 2.514-foldrisk for ischemic stroke by adjusting the conventional riskfactors (95 % CI, 1.667 to 3.790; P=0.000). Next, genotypeassociation test with dominant, recessive and additive models

Table 2 Characteristics of IS patients and control subjects

Characteristics IS patients(n =492)

Controls(n =490)

P value

Sex (females/males) 202/290 222/268 >0.05

Age (years) 56.7±8.3 56.2±8.9 >0.05

Smokers, n (%) 212 (43.09) 140 (28.57) <0.05

Hypertension, n (%) 266 (54.06) 208 (42.45) <0.05

Diabetes, n (%) 108 (21.95) 56 (11.43) <0.05

Total triglyceride (mmol/l) 2.06±1.41 1.83±1.36 <0.05

Total cholesterol (mmol/l) 4.89±1.23 4.48±0.95 <0.05

The IS group was compared with the control group by the χ2 andU tests

J Mol Neurosci

was performed, and the SG13S114AA genotype wasassociated with IS in the recessive and additive genetic models(P=0.000). This association was consistent with the resultsreported previously by Zhang et al. (2006) who showed thatthe SG13S114AA genotype was strongly associated withatherosclerotic stroke in Chinese males (95 % CI, 1.11 to2.35; P =0.012). Domingues-Montanari et al. (2010)presented a large meta-analysis and showed that theSG13S114 SNP is a risk factor for IS in white populations,and they found that T allele carriers of the SG13S114 presentlower levels of ALOX5AP gene expression, which mightindicate that low RNA levels of ALOX5AP is a risk factor

for IS. The role of this gene polymorphism in IS merits furtherinvestigation.

The SG13S377, SG13S89, SG13S32 and SG13S41polymorphisms showed no significant difference, and theywere not associated with IS in this Chinese Han cohort. Asfor the SG13S35 polymorphism, the G allele frequency was100 % and the T allele frequency was 0 % in our study.

Next, we tested the association of the haplotypes with IS.Of the four HapA SNPs (SG13S25, SG13S114, SG13S89 andSG13S32), SG13S25 was not selected because this SNP wasshown to be almost monomorphic in the Chinese population(Zhao et al. 2012; Zhang et al. 2006). So, in our study, three

Table 3 Genotype and allelicdistribution of ALOX5APpolymorphisms in the ISand control subjects

P value and OR (95 % CI) wereadjusted for confounding factorsa Express the adjusted P valuefor significance P <0.008

SNP IS subjects (n =492, %) Control subjects (n=490, %) P value Adjusted OR (95 % CI)

SG13S114

TT 178 (36.2) 214 (43.7) 1.00

AT 222 (45.1) 232 (47.3) 0.311 1.150 (0.877~1.508)

AA 92 (18.7) 44 (9.0) 0.000a 2.514 (1.667~3.790)

T allele 578 (58.7) 660 (67.3) 1.00

A allele 406 (41.3) 320 (32.7) 0.000a 1.449 (1.205~1.742)

SG13S377

GG 316 (64.2) 330 (67.3) 1.00

AG 160 (32.5) 144 (29.4) 0.285 1.160 (0.883~1.524)

AA 16 (3.3) 16 (3.3) 0.905 1.044 (0.513~2.124)

G allele 792 (80.5) 804 (82.0) 1.00

A allele 192 (19.5) 176 (18.0) 0.378 1.107 (0.883~1.389)

SG13S35

GG 492 (100) 490 (100) – –

AG 0 (0) 0 (0) – –

AA 0 (0) 0 (0) – –

G allele 984 (100) 980 (100) –

A allele 0 (0) 0 (0) – –

SG13S41

AA 470 (95.5) 452 (92.2) 1.00

AG 22 (4.5) 38 (7.8) 0.032 0.557 (0.324~0.956)

GG 0 (0) 0 (0) – –

A allele 962 (97.8) 942 (96.1) 1.00

G allele 22 (2.2) 38 (3.9) 0.035 0.567 (0.333~0.966)

SG13S89

GG 446 (90.6) 448 (91.4) 1.00

AG 44 (9.0) 40 (8.2) 0.662 1.105 (0.706~1.729)

AA 2 (0.4) 2 (0.4) 0.996 1.004 (0.141~7.162)

G allele 936 (95.1) 936 (95.5) 1.00

A allele 48 (4.9) 44 (4.5) 0.684 1.091 (0.718~1.659)

SG13S32

AA 226 (45.9) 190 (38.8) 1.00

AC 208 (42.3) 228 (46.5) 0.053 0.767 (0.586~1.004)

CC 58 (11.8) 72 (14.7) 0.053 0.677 (0.456~1.006)

A allele 660 (67.1) 608 (62.0) 1.00

C allele 324 (32.9) 372 (38.0) 0.020 0.802 (0.667~0.966)

J Mol Neurosci

SNP haplotype combinations were made up of eight commonhaplotypes, and the most frequent haplotype was HapA(TGA). But, the distribution of HapA has no significantdifference between the IS group (36.2 %) and the control(37.6 %) (P=0.631). While Zhang et al. (2006) found thatSG13S25 and SG13S32 were not polymorphic among stroke

patients or normal controls, they just analysed two SNPhaplotype combinations and showed no significantdifferences between the controls and the overall stroke orstroke subtypes.

The four HapB SNPs (SG13S114, SG13S377, SG13S41and SG13S35) constituted a total of seven common potential

Table 4 Detailed associationresults of the SNPs betweenIS and control in differentgenetic models

P value and OR (95 % CI) wereadjusted for confounding factorsa Express the adjusted P valuefor significance P <0.008

SNPs Model Genotype Case (n , %) Control (n , %) χ2 P OR (95 % CI)

SG13S114 Dominant TT 178 (36.2) 214 (43.7) 1.000

AT+AA 314 (63.8) 276 (56.3) 5.750 0.016 1.368 (1.059~1.767)

Recessive TT+AT 400 (81.3) 446 (91) 1.000

AA 92 (18.7) 44 (9.0) 19.438 0.000a 2.331 (1.588~3.422)

Additive TT 178 (36.2) 214 (43.7) 1.000

AT 222 (45.1) 232 (47.3) 1.028 0.311 1.150 (0.877~1.508)

AA 92 (18.7) 44 (9.0) 19.985 0.000a 2.514 (1.667~3.790)

SG13S377 Dominant GG 316 (64.2) 330 (67.3) 1.000

AG+AA 176 (35.8) 160 (32.7) 1.061 0.303 1.149 (0.882~1.496)

Recessive GG+AG 476 (96.7) 474 (96.7) 1.000

AA 16 (3.3) 16 (3.3) 0.000 0.991 0.996 (0.492~2.014)

Additive GG 316 (64.2) 330 (67.3) 1.000

AG 160 (32.5) 144 (29.4) 1.141 0.285 1.160 (0.883~1.524)

AA 16 (3.3) 16 (3.3) 0.014 0.905 1.044 (0.513~2.124)

SG13S35 Dominant GG 492 (100) 490 (100) 1.000

AG+AA 0 (0) 0 (0) – – –

Recessive GG+AG 490 (100) 490 (100) 1.000

AA 0 (0) 0 (0) – – –

Additive GG 492 (100) 490 (100) 1.000

AG 0 (0) 0 (0) – – –

AA 0 (0) 0 (0) – – –

SG13S41 Dominant AA 470 (95.5) 452 (92.2) 1.000

AG+GG 22 (4.5) 38 (7.8) 4.614 0.032 0.557 (0.324~0.956)

Recessive AA+AG 492 (100) 490 (100) 1.000

GG 0 (0) 0 (0) – – –

Additive AA 470 (95.5) 452 (92.2) 1.000

AG 22 (4.5) 38 (7.8) 4.614 0.032 0.557 (0.324~0.956)

GG 0 (0) 0 (0) – – –

SG13S89 Dominant GG 446 (90.6) 448 (91.4) 1.000

AG+AA 46 (9.4) 42 (8.6) 0.182 0.669 1.100 (0.710~1.705)

Recessive GG+AG 490 (99.6) 488 (99.6) 1.000

AA 2 (0.4) 2 (0.4) 0.000 0.997 0.996 (0.140~7.098)

Additive GG 446 (90.6) 448 (91.4) 1.000

AG 44 (9.0) 40 (8.2) 0.191 0.662 1.105 (0.706~1.729)

AA 2 (0.4) 2 (0.4) 0.000 0.996 1.004 (0.141~7.162)

SG13S32 Dominant AA 226 (45.9) 190 (38.8) 1.000

AC+CC 266 (54.1) 300 (61.2) 5.154 0.023 0.745 (0.578~0.961)

Recessive AA+AC 434 (88.2) 418 (85.3) 1.000

CC 58 (11.8) 72 (14.7) 1.804 0.179 0.776 (0.535~1.124)

Additive AA 226 (45.9) 190 (38.8) 1.000

AC 208 (42.3) 228 (46.5) 3.734 0.053 0.767 (0.586~1.004)

CC 58 (11.8) 72 (14.7) 3.743 0.053 0.677 (0.456~1.006)

J Mol Neurosci

haplotypes in the patients and controls, and the most frequenthaplotype was Hap (GTAG). But, HapB (AAAG) had nosignificant difference in the patients (18.5 %) and in thecontrols (16.3 %) (P=0.375). Our results were so differentfrom those of the reports of Kajimoto et al. (2005). They saidthat HapB frequencies were too low to perform an associationanalysis, and the authors relied on other haplotypes to detectthe association of ALOX5AP gene with MI in the Japanesepopulation. But, our study results were consistent withZintzaras et al. (2009) who performed a meta-analysis andfound significant heterogeneity among studies (PQ=0.03, I2=

63 %), a non-significant association between the HapA andstroke risk (OR=1.13, 95 % CI 0.88–1.45) and no associationof HapB with stroke risk (OR=1.03, 95 % CI 0.77–1.37).HapC was not considered in previous studies by reason thatthe frequency of HapC population was low (Girelli et al.2007). In our Chinese Han cohort, the HapC (GTAA) cannotbe discovered.

However, of the four HapB SNP-constituted haplotypes(SG13S114, SG13S377, SG13S41 and SG13S35), we foundthat the frequencies of Hap (GAAG) were significantly higherin the patients (20.5 %) than those in the controls (13.9 %)after Bonferroni's adjustment (critical P value=0.05 / sixtests=0.008). Hap (GAAG) allele was associated with anincreased risk for IS (OR=1.599, 95 % CI=1.143~2.239).Our results indicated that Hap (GAAG) was a genetic riskfactor for IS in central Chinese Han population. The effect ongene expression can be dependent on the haplotypes, whichhave a cooperative influence on gene transcriptionalregulation. Thus, the data suggest a contribution ofALOX5AP gene variation to interindividual variability in riskfor IS, which is based on multifactorial etiology. Besides, in aLebanese population, the most frequent haplotype was Hap(GAAG) which has a frequency of 51.1 %, but there is noassociation of Hap (GAAG) and CAD risk (Alwan et al.2010). Bisoendial et al. (2012) reported the association ofALOX5AP haplotypes with abdominal aortic aneurysms(AAA) in Western Australian men population. They foundthat the most frequent haplotypes were Hap (GTGC) and Hap(GTAG) for HapA SNPs and HapB SNPs, respectively. Thefrequency of Hap (GAAG) was 12.4 %, and there is noassociation with AAA risk. In a word, the difference in allele,genotype, and HapA, HapB, HapC and Hap (GAAG)frequencies for the variants under study between the variouspopulations studied may be attributed to the differences inethnic background, leading to different SNP linkagedisequilibriums and haplotype structures.

Table 6 Comparison of haplotype frequencies of ALOX5AP between IScases and controls

Haplotype IS group(2n=984, %)

Control group(2n=980, %)

P value OR (95 % CI)

AAAG(HapB)

182 (18.5) 160 (16.3) 0.375 1.161 (0.834~1.616)

AAGG 8 (0.9) 12 (1.2) 0.746 0.661 (0.185~2.358)

ATAG 2 (0.2) 4 (0.4) 0.997 0.497 (0.045~5.498)

GAAG 202 (20.5) 136 (13.9) 0.006a 1.599 (1.143~2.239)

GAGG 14 (1.4) 12 (1.2) 0.794 1.156 (0.390~3.425)

GTAG 576 (58.5) 640 (65.3) 0.026 0.746 (0.576~0.965)

GTGG 0 (0.0) 16 (1.6) – –

The allele order is SG13S377, SG13S114, SG13S41 and SG13S35 fromleft to righta Expresses the adjusted P value for significance P <0.008

Table 5 Comparison of haplotype frequencies of ALOX5AP between IScases and controls

Haplotype IS group(2n=984, %)

Controlgroup(2n=980,%)

P value OR (95 % CI)

TGA (HapA) 356 (36.2) 368 (37.6) 0.631 0.938 (0.724~1.216)

TGC 220 (22.4) 280 (28.6) 0.025 0.721 (0.540~0.926)

TAA 2 (0.2) 6 (0.6) 0.616 0.332 (0.034~3.203)

TAC 0 (0) 6 (0.6) 0.104 –

AAA 10 (1.0) 2 (0.2) 0.221 5.201 (0.584~43.131)

AAC 18 (1.8) 20 (2.0) 0.812 0.894 (0.354~2.255)

AGA 292 (29.7) 230 (23.5) 0.028 1.375 (1.034~1.829)

AGC 86 (8.7) 68 (6.9) 0.269 1.301 (0.815~2.075)

The allele order is SG13S114, SG13S89 and SG13S32 from left to right;the adjusted P value for significance P<0.008

Fig. 1 The results of linkage disequilibrium tests. The order isSG13S114, SG13S32, SG13S35, SG13S377, SG13S41 and SG13S89from left to right

J Mol Neurosci

The limitation of the present study is the relatively smallsample size. Further study with a larger population is requiredto confirm the findings. In addition, we did not test the plasmalevel of LTB4 (a key product of the FL AP/5-LO pathway) inthe different genotypes of SG13S114, both in the IS group andcontrols.

In summary, the present study suggested that a commongenetic variant SG13S114/AA in the ALOX5AP gene wasassociated with an increased risk for ischemic stroke in thecentral Chinese population. Hap (GAAG) of ALOX5AP wasconfirmed to be associated with increased risk for IS in thecentral Chinese Han population.

Acknowledgments We acknowledge the technical assistance of thestaff members of the Neurology Department of the First AffiliatedHospital of Zhengzhou University. This study was supported by grantsfrom the Provincial Health Department of Henan Province (projectnumber 2011020078). We also thank all the patients and healthy controlsfor providing blood samples.

Conflict of interest None to declare.

References

AdamsHP Jr, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL,Marsh EE 3rd (1993) Classification of subtype of acute ischemicstroke. Definitions for use in a multicenter clinical trial. TOAST.Trial of Org 10172 in Acute Stroke Treatment. Stroke 24(1):35–41

Alwan A, Youhanna SC, Platt DE, El-Sibai M, Yerezian JS, Deeb ME,Khazen G, Saadé S, Zreik TG, El Bayeh H, Maalouf A, Abchee A,Zalloua PA (2010) ALOX5AP gene variants show differentialassociation with coronary artery disease in different populations. JCommun Genet 1(3):107–115

Barral S, Fernández-Cadenas I, Bis JC, Montaner J, Ikram AM, LaunerLJ, Fornage M, Schmidt H, Brickman AM, Seshadri S, Mayeux R(2012) No association of ALOX5AP polymorphisms with risk ofMRI-defined brain infarcts. Neurobiol Aging 33(3):629.e1–629.e3

Bisoendial RJ, Tanck MW, Golledge J, Broekhuizen LN, Legemate DA,Stroes ES, Norman PE (2012) The association between the geneencoding 5-lipoxygenase activating protein and abdominal aorticaneurysms. Atherosclerosis 220(2):425–428

Domingues-Montanari S, Fernández-Cadenas I, del Rio-Espinola A,Corbeto N, Krug T, Manso H, Gouveia L, Sobral J, Mendioroz M,Fernández-Morales J, Alvarez-Sabin J, Ribó M, Rubiera M, ObachV, Martí-Fàbregas J, Freijo M, Serena J, Ferro JM, Vicente AM,Oliveira SA, Montaner J (2010) Association of a genetic variant inthe ALOX5AP with higher risk of ischemic stroke: a case-control,meta-analysis and functional study. Cerebrovasc Dis 29(6):528–537

Girelli D, Martinelli N, Trabetti E, Olivieri O, Cavallari U, Malerba G,Busti F, Friso S, Pizzolo F, Pignatti PF, Corrocher R (2007)ALOX5AP gene variants and risk of coronary artery disease: anangiography-based study. Eur J Hum Genet 15(9):959–966

Helgadottir A, Manolescu A, Thorleifsson G, Gretarsdottir S, JonsdottirH, Thorsteinsdottir U, Samani NJ, Gudmundsson G, Grant SF,

Thorgeirsson G, Sveinbjornsdottir S, Valdimarsson EM,Matthiasson SE, Johannsson H, Gudmundsdottir O, GurneyME, Sainz J, Thorhallsdottir M, Andresdottir M, Frigge ML,Topol EJ, Kong A, Gudnason V, Hakonarson H, Gulcher JR,Stefansson K (2004) The gene encoding 5-lipoxygenaseactivating protein confers risk of myocardial infarction andstroke. Nat Genet 36:233–239

Kajimoto K, Shioji K, Ishida C, Iwanaga Y, Kokubo Y, Tomoike H,Miyazaki S, Nonogi H, Goto Y, Iwai N (2005) Validation of theassociation between the gene encoding 5-lipoxygenase-activatingprotein and myocardial infarction in a Japanese population. Circ J69(9):1029–1034

Linsel-Nitschke P, Götz A,Medack A, König IR, Bruse P, LiebW,MayerB, Stark K, Hengstenberg C, Fischer M, Baessler A, Ziegler A,Schunkert H, Erdmann J (2008) Genetic variation in thearachidonate 5-lipoxygenase-activating protein (ALOX5AP) isassociated with myocardial infarction in the German population.Clin Sci 115:309–315

Lõhmussaar E, Gschwendtner A, Mueller JC, Org T, Wichmann E,Hamann G, Meitinger T, Dichgans M (2005) AlOX5AP gene andthe PDE4D gene in a central European population of stroke patients.Stroke 36:731–736

Quarta G, Stanzione R, Evangelista A, Zanda B, Di Angelantonio E,Marchitti S, Di Castro S, Di Vavo M, Volpe M, Rubattu S (2009)Phosphodiesterase 4D and 5-lipoxygenase activating protein genesand risk of ischemic stroke in Sardinians. Eur J Hum Genet 17(11):1448–1453

Saleheen D, Bukhari S, Haider SR, Nazir A, Khanum S, Shafqat S, AnisMK, Frossard P (2005) Association of phosphodiesterase 4D genewith ischemic stroke in a Pakistani population. Stroke 36(10):2275–2277

Shi YY, He L (2005) SHEsis, a powerful software platform for analysesof linkage disequilibrium, haplotype construction, and geneticassociation at polymorphism loci. Cell Res 15:97–98

Ström JO, Strid T, Hammarström S (2012) Disruption of the alox5ap geneameliorates focal ischemic stroke: possible consequence of impairedleukotriene biosynthesis. BMC Neurosci 13(1):146

Sun H,Wu H, Zhang J,Wang J, Lu Y, Ding H, Xiao H, Zhang J (2011) Atagging SNP in ALOX5AP and risk of stroke: a haplotype-basedanalysis among eastern Chinese Han population. Mol Biol Rep38(7):4731–4738

Wang G, Wang Y, Sun H, Cao W, Zhang J, Xiao H, Zhang J (2011)Variants of the arachidonate 5-lipoxygenase-activating protein(ALOX5AP) gene and risk of ischemic stroke in Han Chinese ofeastern China. J Biomed Res 25(5):319–327

Zee RY, Cheng S, Hegener HH, Erlich HA, Ridker PM (2006) Geneticvariants of arachidonate 5-lipoxygenase-activating protein, and riskof incidentmyocardial infarction and ischemic stroke: a nested case–control approach. Stroke 37(8):2007–2011

Zhang WL, Yang XM, Shi J, Sun K, Hui RT (2006) Polymorphism ofSG13S114T/A in the ALOX5AP gene and the risk for stroke in alarge Chinese cohort. Yi Chuan Xue Bao 33(8):678–684

Zhao J, Wang X, Xu J et al (2012) Association of inflammatoryresponse gene polymorphism with atherothrombotic stroke inNorthern Han Chinese. Acta Biochim Biophys Sin 44(12):1023–1030

Zintzaras E, Rodopoulou P, Sakellaridis N (2009) Variants of thearachidonate 5-lipoxygenase-activating protein (ALOX5AP) geneand risk of stroke: a HuGE gene-disease association review andmeta-analysis. Am J Epidemiol 169(5):523–532

J Mol Neurosci