Association Analysis of Common Variants of ELN, NOS2A, APOE and ACE2 to Intracranial Aneurysm
Background and Purpose— Previous studies have shown positive evidence of linkage of the intracranial aneurysm (IA) at chromosome 7q11, 17cen, 19q13, and Xp22. These regions contain elastin (ELN), nitric oxide synthetase 2A (NOS2A), apolipoprotein E (APOE), and angiotensin-I converting enzyme 2 (ACE2), which are considered to be promising candidate genes for IA. We aimed to examine the association of single-nucleotide polymorphisms (SNPs) with IA in these candidate genes.
Methods— To identify polymorphisms in NOS2A and ACE2, all exons and exon-intron boundaries were screened by direct sequencing in 30 randomly selected controls. The program tagSNPs was used to select an optimal set of haplotype-tagging SNPs. For ELN and APOE, SNPs were selected from previous reports. These selected SNPs were then genotyped in 362 cases with IA and 332 residential area matched controls. THESIAS software was used to investigate the association of alleles and haplotypes with IA by adjusting with covariates.
Results— We genotyped 8 SNPs in ELN, 8 SNPs in NOS2A, 3 ε alleles in APOE and 1 SNP in ACE2. No alleles or haplotypes of 4 candidate genes revealed any significant association with IA.
Conclusions— Investigated polymorphisms in this study were not associated with IA.
It has recently been recognized that genetic factors have an impact on the pathogenesis of intracranial aneurysm (IA). Genome-wide linkage analyses have revealed linkages to several chromosomal regions.1–7 Among them, 7q11,1,4 17cen,1,6 19q132,3,6 and Xp222,6 are potentially interesting because they have been replicated in several studies. Elastin (ELN), nitric oxide synthetase 2A (NOS2A), apolipoprotein E (APOE) and angiotensin-I converting enzyme 2 (ACE2) are located on 7q11, 17cen, 19q13 and Xp22, respectively, and they are considered promising candidate genes for IA.
Human ELN consists of 34 exons and spans 45 kb of genomic DNA. The association of ELN haplotypes with IA or subarachnoid hemorrhage (SAH) was reported in Japanese and Dutch studies, albeit with genetic heterogeneity between the studies.1,8 However, other studies have failed to show an association.9,10 Besides, a Finnish group and we have demonstrated the absence of a linkage to 7q11.2,11
Human NOS2A consists of 26 exons and 25 introns spanning 37 kb of genomic DNA.12,13 Sadamasa et al reported that knocking out the iNOS (NOS2A) gene reduced the size of cerebral aneurysms in mice, suggesting its potential role in the progression of IA.14,15
The most common genetic alleles of APOE are ε2, ε3 and ε4. In a prospective case-control study by Kokubo et al,16 the ε4 allele was reported to be a risk factor for SAH in eastern Japan.
Human ACE2 contains 18 exons spanning ≈40 kb of genomic DNA,17 and resides in chromosome Xp22 where many genes escape inactivation.18 The I/I genotype of ACE was reported as a risk factor for SAH in Poland.19 ACE2 is a homolog of ACE, and they negatively regulate each other,20 suggesting that ACE2 could also be a risk factor for IA.
To validate these findings, we studied the association of single-nucleotide polymorphisms (SNPs) and haplotypes in these candidate genes with IA in a western Japan–based population.
Materials and Methods
The study population consisted of 362 unrelated case subjects with IA, who were diagnosed by digital subtraction angiography or by operations in collaborating hospitals in western Japan. The residential areas of cases and controls were matched to eliminate the effect of population stratification by heterogeneity. Control subjects met the following criteria: (1) confirmation that they did not harbor IA by digital subtraction angiography, 3-dimensional computed tomography, or by magnetic resonance angiography, (2) age at diagnosis of ≥40 years old, (3) no medical history of any stroke including IA and SAH, and (4) no family history of IA and SAH in first-degree relatives. The study was approved by the Ethics Committee of Kyoto University. For all subjects, we interviewed for their risk factors profile, including past medical history, family history, smoking habit, and alcohol consumption. Smoking habit was defined as current smokers of ≥1 cigarette per day, former smokers, and nonsmokers. For statistical analysis, current smokers and former smokers were dealt as smokers. Drinkers were defined as regular drinkers who drink >150 g of alcohol per week.
SNP Screening in NOS2A and ACE2
To identify polymorphisms in NOS2A (GeneBank accession No. NT_010799) and ACE2 (NT_011757), all exons, intron-exon boundaries, putative promoter sequence and the 3′UTR were analyzed by direct sequencing in 30 randomly selected controls. Primers for coding exons were designed from an intronic sequence >50 bp away from the intron-exon boundaries and commercially synthesized by PROLIGO (PROLIGO Primers & Probes; http://www.proligo.com). After polymerase chain reaction (PCR) amplification, products were electrophoresed and purified using a QIAquick Gel Extraction Kit (Qiagen Inc), followed by sequencing on an ABI Prism 3100 Avant DNA sequencer (Applied Biosystems). We checked the SNP database (dbSNP; http://www.ncbi.nlm.nih.gov/SNP/) as reference. Primers and PCR conditions of each gene are available from the author on request.
In NOS2A and ACE2, among all the SNPs identified by direct sequencing, we selected a minimized number of haplotype-tagging (ht) SNPs to be genotyped using the program tagSNPs (tagSNPs version 1; http://www-rcf.usc.edu/≈stram/tagSNPs.html).21 We ran the program with the following criteria: common haplotypes were defined as the minimal set of haplotypes that covers 80% of existing haplotypes, sets of htSNPs resolving the common haplotypes were selected at an Rh2 (the squared correlation between estimated and true haplotype dosage) threshold of 0.8.22 Exceptionally, all nonsynonymous SNPs (ie, SNPs located in coding regions and results in amino acid variation in the protein products of the gene) were forced in as a set of htSNPs. Selected htSNPs were genotyped in 362 cases and 332 controls. Nonsynonymous SNPs were analyzed by bioinfomatics using PolyPhen software (http://tux.embl-heidelberg.de/ramensky/) to predict whether they were damaging to the structure or function of the protein products.
In ELN (NT_007758), we selected 8 of 18 SNPs identified in previous reports1,8 in the following process: because the significant association of ELN haplotypes with IA or SAH was found in intron 20 (INT20)/INT22 in a Japanese study and INT4/INT5/INT21 in a Dutch study,1,8 these 5 intronic SNPs were selected to be genotyped. In addition, all 3 exonic SNPs (exon 5 [EX5], EX20 and EX22) were selected, and a total of 8 SNPs were genotyped in ELN. In APOE (NT_011109), we genotyped ε alleles in EX4.16
Genotyping of ELN, NOS2A, APOE and ACE2 was performed by PCR-restriction fragment length polymorphism (PCR-RFLP) protocol. APOE was genotyped as previously reported.23 Because of the lack of a proper restriction enzyme for INT4 in ELN, real-time PCR TaqMan analysis was conducted on the 7300/7500 Real Time PCR System (Applied Biosystems).
Haplotype and Linkage Disequilibrium Analysis
We investigated haplotypes with a frequency of >5% for each gene.24 ELN, INT20/INT22 and INT4/INT5/INT21 were also investigated. Considering the possibility of interchromosomal interaction, we examined pair-wise haplotypes that consisted of all the SNPs genotyped in ELN, NOS2A and APOE. Linkage disequilibrium (LD) calculations were conducted by means of r2 and D′ using Genotype2LDBlockVO.2 (http://cgi.uc.edu/cgi-bin/kzhang/genotype2LDBlock.cgi). Association Analysis The association of alleles and haplotypes with IA was analyzed using THESIAS software (http://genecanvas.ecgene.net/)25 by adjusting with covariates including age, sex, hypertension, smoking habit and heavy alcohol consumption. We also analyzed association of polymorphisms with SAH. Allele frequencies of control subjects in 2 major residential areas (Osaka and Kyoto) were compared by χ2 test using SAS software (Version 8.2. SAS Institute, Inc). For ACE2, data for each sex were analyzed separately because it is on the X chromosome. Bonferroni correction was done as needed (probability value after correction [Pcorr]).
Assuming an autosomal disease allele with population frequency of 0.20 that contributes to IA with a relative risk of ≥1.25, sampling would require an equal number of 314 cases and controls to provide 80% power for a significant threshold of P=0.05. (Genetic Power Calculator, http://statgen.iop.kcl.ac.uk/gpc/cc2.html).
As shown in Table 1, the percentage of females and hypertension was higher among cases than controls. No significant difference was found in either smoking habit or alcohol consumption. Identification and Selection of SNPs In NOS2A, we identified 12 SNPs (supplemental Table I, available online at http://stroke.ahajournals.org), of which 2 (INT16: IVS16+88 G>T, and EX19: Ex19 2503 A>G) were novel and 2 were nonsynonymous, EX16 (S608L) and EX19 (T747A). S608L was predicted to have a possible damaging structure or function of NOS2A by PolyPhen. Serine608 is conserved among 5 species including rat and mouse (Homolo Gene: 55473; http:// www.ncbi.nlm.nih.gov), whereas Threonin747 was conserved in only 2 species, human and dog. Of 12 SNPs identified in NOS2A, 8 SNPs (INT7, INT7′, INT8, INT12, EX16, INT16, EX19, and EX22) were selected according to the tagSNP program. In ACE2, only 1 registered SNP (rs2285666) in INT3 was identified. In ELN and APOE, 8 SNPs and 3 ε alleles were selected as already stated.
In ELN, 8 SNPs and 8 haplotypes including INT20/INT22 and INT4/INT5/INT21 were analyzed. We observed no significant association of polymorphisms with either IA (Table 2) or SAH (data not shown). All haplotypes also failed to show an association (Table 3). LD analysis revealed a weak LD pattern unlike that of NOS2A (data not shown).
In NOS2A, a total of 8 htSNPs and 4 haplotypes were analyzed, and all these SNPs were in Hardy-Weinberg equilibrium after Bonferroni correction. No SNPs or haplotypes were associated with either IA (Table 2, Table 3) or SAH (data not shown).
In APOE, no association was observed between ε alleles and the occurrence of either IA (Table 3) or SAH (data not shown). In ACE2, analysis of the SNP demonstrated a lack of association either in males or females (Table 2). Besides this, none of the pair-wise haplotypes consisting of all SNPs in ELN, NOS2A and APOE could have shown the association (data not shown).
In the analysis of regional differences of allele frequency, the frequency of EX5, INT20 and INT21 in ELN was significantly different between Osaka and Kyoto (P=0.0042, P=0.0385 and P=0.0113, respectively; Table 4), whereas no difference was observed in either NOS2A or APOE (supplemental Table II). Even after applying Bonferroni correction, the probability value of EX5 was statistically significant (Pcorr=0.034). Characteristics of control subjects in Osaka and Kyoto were listed in Table 1.
In the present study, we examined the association of polymorphisms of ELN, NOS2A, APOE and ACE2 with IA. For ELN and APOE, we selected the SNPs to be analyzed based on previous association studies.1,8,16 For NOS2A and ACE2, because there were no previously published association studies, we sequenced all exons and exon-intron boundaries to search for SNPs. Considering various modes of associations, we also test the associations of polymorphisms with a related phenotype of IA, SAH by using a large number of cases and controls that promised us sufficient statistical power. Furthermore, we investigated interchromosomal interactions among these genes. Thus, within the present experimental settings, design and quality enabled us to detect signals as weak as a relative risk of 1.25.
We tested the association of ELN SNPs reported by Onda et al1 and Ruigrok et al8 but failed to show an association with either IA or SAH. One explanation for the disagreement could be haplotype heterogeneity among study populations. In fact, LD analysis of ELN showed very weak LD even in the same ethnic group, being consistent with other reports1,8 and HapMap LD data (http://www.hapmap.org/cgi-perl/gbrowse/gbrowse/hapmap). Considering that LD is negatively correlated with recombination rates,26 ELN is likely to have a recombination hotspot; therefore, it is easy to have haplotype heterogeneity even among adjacent populations. So, there is a possibility that untested SNPs in this study were associated with IA or SAH. However, the most likely explanation for the disagreement would be that a significant association of ELN haplotype with IA may represent LD with an unknown gene.
For APOE, Kokubo et al reported the positive association of ε4 allele with SAH in eastern Japan.16 Our study, however, could not confirm their findings, suggesting that polymorphisms of APOE may not be a major genetic risk factor for either SAH or unruptured IA in western Japan.
For NOS2A, knockout mice were proven to have reduced sizes of aneurysms.14,15 The present study, however, could not show any association with either IA or SAH. The apparent discrepancy may be attributable to differences in species or in study protocols. Although knockout mice model a loss of function of NOS2A, our study investigated qualitative functional changes. In addition, minor allele frequencies of 2 nonsynonymous SNPs (S608L and T747A) were below 7%, which made it difficult to detect positive signals attributable to the limitation of statistical power. Indeed, our study indicates that NOS2A is not likely to take a major role in the pathogenesis of IA or SAH. However, the effect of a rare polymorphism, such as S608L, needs more cautious interpretation because Serine608 is conserved in various species and S608L is predicted to be a deleterious mutation. Although haploinsufficiency is not likely to be associated with IA, S608L cannot be discarded as a risk factor for IA in its homozygous state. Further study will be needed for this rare polymorphism.
ACE2 is a homolog of ACE, the I/I genotype of which has been proven to be associated with SAH in Polish population.19 In the present study, however, no association was observed.
We examined the association of SNPs and haplotypes of 4 promising candidate genes with IA. However, investigated polymorphisms in this study were not associated with either IA or SAH.
This work was supported by a grant from the Ministry of Education, Science, Sports, and Culture of Japan (Kiban Kenkyuu S: 14207016) and a grant from the Japan Society for the Promotion of Science (15012231). We are grateful to Benjamin Seery for critical reading of the manuscript. We thank Miho Yoshida, Norio Matsuura, and Maki Utsunomiya for technical assistance, and the following doctors for patient recruitment and help in ascertaining magnetic resonance angiography examinations: Shinsuke Tominaga, Hiroshi Hasegawa, and Toshihiko Inui (Tominaga Hospital), Shyunichi Yoneda and Yoshito Naruo (Nihonbashi Hospital), Yoo Kang, Shoichi Tani, and Atsuhito Matsumoto (Osaka Saiseikai Izuo Hospital), Hiroyasu Yamakawa (Gero-spring Hospital), Atsushi Kawarazaki (Kawarazaki Hospital), Masayuki Matsuda (Shiga University of Medical Science), Michiyasu Suzuki and Sadahiro Nomura (Yamaguchi University School of Medicine), Takaaki Kaneko, Nozomu Murai, and Susumu Kanemoto (Hikone Municipal Hospital), Tatsuhito Yamagami and Motoharu Fujii (Kyoto Kizugawa Hospital), Hikaru Ohishi and Kiminari Ohtaka (Senboku Kumiai Sougou Hospital), Kenji Kikuchi and Yutaka Yamazaki (Yuri Kumiai Sougou Hospital), Jun Takahashi, Nobuhiro Mikuni, Ken-ichiro Kikuta, and Yasushi Takagi (Kyoto University Graduate School of Medicine).
Drs Y. Mineharu, K. Inoue, and S. Inoue contributed equally to this work.
- Received March 18, 2005.
- Revision received April 29, 2005.
- Accepted June 2, 2005.
van der Voet M, Olson JM, Kuivaniemi H, Dudek DM, Skunca M, Ronkainen A, Niemela M, Jaaskelainen J, Hernesniemi J, Helin K, Leinonen E, Biswas M, Tromp G. Intracranial aneurysms in Finnish families: confirmation of linkage and refinement of the interval to chromosome 19q13.3. Am J Hum Genet. 2004; 74: 564–571.
Roos YB, Pals G, Struycken PM, Rinkel GJ, Limburg M, Pronk JC, van den Berg JS, Luijten JA, Pearson PL, Vermeulen M, Westerveld A. Genome- wide linkage in a large Dutch consanguineous family maps a locus for intracranial aneurysms to chromosome 2p13. Stroke. 2004; 35: 2276–2281.
Yamada S, Utsunomiya M, Inoue K, Nozaki K, Inoue S, Takenaka K, Hashimoto N, Koizumi A. Genome-wide scan for Japanese familial intracranial aneurysms: linkage to several chromosomal regions. Circulation. 2004; 110: 3727–3733.
Ruigrok YM, Seitz U, Wolterink S, Rinkel GJ, Wijmenga C, Urban Z. Association of polymorphisms and haplotypes in the elastin gene in Dutch patients with sporadic aneurysmal subarachnoid hemorrhage. Stroke. 2004; 35: 2064–2068.
Hofer A, Hermans M, Kubassek N, Sitzer M, Funke H, Stogbauer F, Ivaskevicius V, Oldenburg J, Burtscher J, Knopp U, Schoch B, Wanke I, Hubner F, Deinsberger W, Meyer B, Boecher-Schwarz H, Poewe W, Raabe A, Steinmetz H, Auburger G. Elastin polymorphism haplotype and intracranial aneurysms are not associated in central Europe. Stroke. 2003; 34: 1207–1211.
Yamada S, Utsunomiya M, Inoue K, Nozaki K, Miyamoto S, Hashimoto N, Takenaka K, Yoshinaga T, Koizumi A. Absence of linkage of familial intracranial aneurysms to 7q11 in highly aggregated Japanese families. Stroke. 2003; 34: 892–900.
Chartrain NA, Geller DA, Koty PP, Sitrin NF, Nussler AK, Hoffman EP, Billiar TR, Hutchinson NI, Mudgett JS. Molecular cloning, structure, and chromosomal localization of the human inducible nitric oxide synthase gene. J Biol Chem. 1994; 269: 6765–6772.
Fukuda S, Hashimoto N, Naritomi H, Nagata I, Nozaki K, Kondo S, Kurino M, Kikuchi H. Prevention of rat cerebral aneurysm formation by inhibition of nitric oxide synthase. Circulation. 2000; 101: 2532–2538.
Sadamasa N, Nozaki K, Hashimoto N. Disruption of gene for inducible nitric oxide synthase reduces progression of cerebral aneurysms. Stroke. 2003; 34: 2980–2984.
Kokubo Y, Chowdhury AH, Date C, Yokoyama T, Sobue H, Tanaka H. Age-dependent association of apolipoprotein E genotypes with stroke subtypes in a Japanese rural population. Stroke. 2000; 31: 1299–1306.
Tipnis SR, Hooper NM, Hyde R, Karran E, Christie G, Turner AJ. A human homolog of angiotensin-converting enzyme. Cloning and functional expression as a captopril-insensitive carboxypeptidase. J Biol Chem. 2000; 275: 33238–33243.
Carrel L, Cottle AA, Goglin KC, Willard HF. A first-generation X-inactivation profile of the human X chromosome. Proc Natl Acad Sci U S A. 1999; 96: 14440–14444.
Slowik A, Borratynska A, Pera J, Betlej M, Dziedzic T, Krzyszkowski T, Czepko R, Figlewicz DA, Szczudlik A. II genotype of the angiotensin-converting enzyme gene increases the risk for subarachnoid hemorrhage from ruptured aneurysm. Stroke. 2004; 35: 1594–1597.
Hixson JE, Vernier DT. Restriction isotyping of human apolipoprotein E by gene amplification and cleavage with HHAI. J Lipid Res. 1990; 31: 545–548.
Johnson GC, Esposito L, Barratt BJ, Smith AN, Heward J, Di Genova G, Ueda H, Cordell HJ, Eaves IA, Dudbridge F, Twells RC, Payne F, Hughes W, Nutland S, Stevens H, Carr P, Tuomilehto-Wolf E, Tuomilehto J, Gough SC, Clayton DG, Todd JA. Haplotype tagging for the identification of common disease genes. Nat Genet. 2001; 29: 233–237.
Greenwood TA, Rana BK, Schork NJ. Human haplotype block sizes are negatively correlated with recombination rates. Genome Res. 2004; 14: 1358–1361.