Background Single nucleotide polymorphisms (SNPs) on chromosome 9p21 have been recently associated with intracranial aneurysms and stroke. In this study, we tested the association between the rs1333040C>T polymorphism on the 9p21 locus and sporadic brain arteriovenous malformations (BAVMs).
Methods We studied 78 patients with sporadic BAVMs and 103 unaffected controls. Genomic DNA was isolated from peripheral blood and the rs1333040C>T polymorphism was assessed by PCR–restriction fragment length polymorphism using the BsmI restriction endonuclease.
Results We found that the distribution of the three genotypes (TT/TC/CC) of the rs1333040 polymorphism was significantly different between cases and controls (p=0.02). Using dominant, recessive and additive genetic models, we found that the TT genotype and the T allele were significantly more common in the BAVM group than in controls. We also evaluated whether the rs1333040 polymorphism was associated with prototypical angio-architectural features of BAVMs (such as nidus size, venous drainage pattern and Spetzler–Martin grading) and with the occurrence of seizures and bleeding. We detected a significant association between the homozygous T allele in the recessive model and BAVMs with a nidus >4 cm in diameter. Deep venous drainage was significantly more frequent among subjects carrying at least one T allele in the dominant model. Patients with seizures showed a significant association with the TT genotype and the T allele in all genetic models examined whereas those who experienced intracranial bleeding showed a significant association with the T allele in the trend model.
Conclusions This is the first study demonstrating an association between an SNP of the 9p21 region and sporadic BAVMs. Our results emphasise the relevance of this chromosomal locus as a common risk factor for various forms of cerebrovascular diseases.
- CEREBROVASCULAR DISEASE
Statistics from Altmetric.com
Single nucleotide polymorphisms (SNPs) on chromosome 9p21 have been recently associated with stroke and intracranial aneurysms in many genome wide association studies (GWAS).1–7 Interestingly, the 9p21 locus has also been associated with several other vascular disorders, such as coronary artery disease (CAD), abdominal aortic aneurysm (AAA) and peripheral artery occlusive disease (PAOD).2 Among these SNPs, the rs1333040 C>T polymorphism is one of those more strongly associated with both intracranial1 ,4 and extracranial vascular diseases.2 In the study by Nakaoka et al,8 stepwise logistic regression analysis identified rs1333040 as an independent predictor of intracranial aneurysms among other SNPs analysed in the 9p21 region. The association between this SNP and intracranial aneurysms was recently replicated by Deka et al in a white population different from those analysed in the original GWAS.9
In this study, we tested the possible association between the rs1333040 C>T polymorphism on 9p21 and sporadic brain arteriovenous malformations (BAVMs).
Materials and methods
The BAVM group consisted of subjects undergoing neurosurgical treatment or follow-up for BAVMs (n=78) at the Department of Neurosurgery of the A Gemelli University Hospital of Rome, Italy. In all cases, the diagnosis of BAVM was based on the results of brain MRI and digital subtraction angiography. The control group consisted of individuals consecutively admitted to the same Department of Neurosurgery, from January to December 2011, because of a traumatic brain injury. To be included as controls, subjects needed a contrast enhanced brain CT scan negative for any type of vascular lesion. In addition, a negative history of stroke and other cardiovascular diseases was required. This was assessed through collection of a complete medical history, analysis of medical records and physical examination. At the end of the recruitment period, 103 individuals fulfilled these inclusion criteria and were enrolled as controls. Informed consent was signed by all participating individuals. The study was approved by the ethics committee of our university hospital. All patients and controls were Italians from the administrative region of Lazio.
Venous blood (2 ml) was collected and stored at −80°C before use. Genomic DNA was isolated from whole blood using a commercially available DNA extraction kit (Illustra Blood GenomicPrep Mini Spin Kit, GE Healthcare). Polymorphism spanning fragments were amplified by a PCR–restriction fragment length polymorphism assay. PCR reaction was carried out in a total volume of 25 ml using 5 ml of template DNA, 1 × PCR buffer (10 mM Tris-HCl pH 8.5, 50 mM KCl), 1.5 mM MgCl2, 0.0625 mM of each deoxynucleotide triphosphate (dATP, dTTP, dGTP, dCTP), 0.5 mM of each primers (FW: TCT GGA AGC ACT GGG AAG GAT G and REV: TTG ATT TGG GAG CCA CTG TTG) and 1.6 U of Taq DNA polymerase (Roche). Thermal cycling was performed in a PE Applied Biosystems 9700 Sequence Detector with the following programme: 4 min at 95°C, 40 cycle of 1 min at 94°C, 1 min a 56°C and 1 min at 72°C, with a final extension step at 72°C for 7 min. Genotyping of the rs1333040C>T polymorphism was identified using restriction endonuclease BsmI (Biolabs). The amplified DNA (5 ml), 10 U of BsmI and 1 ml of 10× recommended restriction buffer in a final volume of 10 ml were incubated at 65°C for 1 h. Digested products were separated on 2% agarose gel by electrophoresis.
Statistical analysis was performed with SPSS Statistical Package V.12.0 (Chicago, Illinois, USA). The association between rs1333040 genotypes and BAVMs was assessed using dominant, recessive and additive (trend) genetic models. Hardy–Weinberg equilibrium and categorical variables were compared using Pearson's χ2 test. A value of p<0.05 was considered statistically significant.
Demographic characteristics of the studied population are shown in table 1. Mean age in the BAVM group was 37.71±19.2 years (range 5–76 years) while in the control group mean age was 50.22±20.61 years (range 18–96 years). There was no significant difference in age between patients and controls (p=0.11). The cohort of BAVM individuals included 46 men (58.9%) while the number of men in the control group was 72 (69.9%). The male/female ratio was not statistically different between the two groups (p=0.12).
Genotype distribution is shown in table 2. Genotypes were in Hardy–Weinberg equilibrium. The distribution of the three possible genotypes of rs1333040 (TT, TC and CC) was statistically different between cases and controls (p=0.02). The TT genotype was significantly associated with BAVMs both in the dominant (p=0.009) and recessive (p=0.006) models (table 2). The T allele was significantly associated with BAVMs in the additive (trend) model (p=0.01) (table 2).
We also took into consideration a number of important clinical and angio-architectural features of BAVMs and evaluated their possible association with the genotypes of rs1333040 (table 3). We detected a significant association between the TT genotype in the recessive model and the presence of BAVMs with a nidus >4 cm in diameter (p=0.02). Deep venous drainage was also significantly more frequent among subjects carrying at least one T allele in the dominant model (p=0.02). A Spetzler–Martin score ≥3 was associated with the presence of the T allele in the additive (trend) model (p=0.01). Finally, patients who presented with seizures during the natural history of the disease showed a significant association with the TT genotype and the T allele in the dominant, recessive and additive models (p=0.02, p=0.02 and p<0.001, respectively) while those who experienced spontaneous rupture of the BAVM with subsequent intracranial bleeding showed a significant association with the T allele in the trend model (p=0.02).
BAVMs are tangles of abnormal vessels directly shunting blood from the arterial to the venous circulation without an interposed capillary bed. They may occur either sporadically or as part of hereditary genetic disorders involving other organs in addition to the brain, such as in the case of Osler–Weber–Rendu syndrome.10 BAVMs are also characteristic of certain congenital non-hereditary conditions, such as Wyburn–Mason syndrome, in which they are associated with retinal angiomas,11 and Sturge–Weber syndrome, a neurocutaneous disorder in which BAVMs are associated with a large port wine stain birthmark on the forehead.11 ,12 Their detection rate is approximately 1.1 per 100 000 adults per year.13 The literature does not recognise a specific genetic cause for sporadic BAVMs, which are therefore considered a multigenic/multifactorial disease.
Identification of chromosome 9p21 as a marker of cardiovascular diseases has been a landmark discovery in cardiovascular genetics. It was discovered in the first surge of GWAS for CAD in 2007 and published simultaneously by four independent groups only a few weeks apart.14–17 Today, 9p21 represents the most replicated marker of CAD and myocardial infarction in the medical literature.18 Interestingly, the same locus is also strongly associated with carotid atherosclerosis,19 ischaemic stroke,20 peripheral artery occlusive disease,21 heart failure,17 ,19 AAA22 and intracranial aneurysms.3 ,4
This is the first study reporting an association between an SNP of chromosome 9p21 and BAVMs.23 Inclusion of BAVMs among the vascular diseases that are associated with the 9p21 locus suggests that BAVMs may share common pathogenetic mechanisms with a number of other cardiovascular and cerebrovascular diseases, including intracranial aneurysms and stroke. Importantly, the 9p21 region lacks associations with common cardiovascular risk factors, indicating that the locus exerts its effect through completely novel mechanisms which need further investigation for their potential amenability to therapeutic intervention. Indeed, the pathways through which the 9p21 locus influences the risk of vascular diseases are still unknown. The majority of the risk variants identified by GWAS in the 9p21 region are in non-coding regions, suggesting that the observed effects may be mediated by the modulation of gene expression. Genes mapping in the 9p21 region include the cyclin dependent kinase inhibitors CDKN2A (p16INK4a), including its alternative reading frame (ARF) transcript variant (p19ARF), and CDKN2B (p15INK4b), along with a recently discovered non-coding RNA sequence named ANRIL that undergoes splicing and is transcribed from the opposite strand to CDKN2A/B. These genes are involved in the regulation of cell cycle progression and influence key physiological processes, such as replicative senescence, apoptosis, stem cell self-renewal and repair of aged tissue.24–26 Sequence variation can influence expression by cis or trans mechanisms. Trans acting elements influence transcript levels of both alleles via diffusible factors and are usually located distant to the genes they regulate whereas cis acting elements act on genes on the same chromosome in an allele specific manner and are usually located close to the genes they regulate. As most reported risk variants in the 9p21 region do not appear in mature transcripts and there are no known or predicted microRNAs mapping to this region, these variants are unlikely to produce diffusible trans acting factors and are therefore likely to influence expression of nearby genes in cis.24
The rs1333040 that we evaluated in this study is in close proximity to ANRIL. ANRIL is expressed in atherosclerotic tissues27 and its transcript DQ485453 has been detected in primary coronary smooth muscle cells, macrophages, and carotid endarterectomy and AAA tissue samples.28 Folkersen et al identified eight ANRIL transcripts also in lymphoblastic cells and carotid artery, medial aorta and mammary artery plaque tissue,29 while Holdt et al found differential expression of ANRIL in peripheral blood mononuclear cells of patients with CAD and carotid, aortic and femoral plaques.30 In addition, GWAS have identiﬁed ANRIL as a susceptibility gene for type 2 diabetes mellitus,31 glioma32 and basal cell carcinoma.33
Another novel finding of our study is the association between rs1333040 and BAVM nidus dimension. This is the first SNP to be associated with this parameter in the medical literature. As BAVMs are evolving lesions which can spontaneously increase in size with time, or recur after incomplete surgery or embolisation through the recruitment of new arteriovenous shunts,34 ,35 identification of a genetic marker that is associated with nidus dimension might be useful to screen patients harbouring malformations more prone to increase in size over time. This might theoretically have an influence on treatment decision. Our findings also show an association between the occurrence of seizures and rs1333040. Rs1333040 also appears significantly associated with the occurrence of seizures and bleeding in subjects with BAVMs. The pathophysiological mechanisms underlying these associations remain to be understood although they might simply depend on the presence of larger nidi.36
This study has some potential limitations. It was performed on a relatively small number of patients from the Lazio administrative region of Italy. Therefore, our findings should be considered as hypothesis generating and need confirmation in larger prospective studies and other ethnic groups. In addition, we cannot exclude the fact that the reported associations are caused by a gene in linkage disequilibrium with the rs1333040C>T polymorphism. This possibility needs to be further investigated through transmission disequilibrium tests and/or linkage disequilibrium mapping.
This is the first study demonstrating an association between an SNP in the chromosomal 9p21 locus and BAVMs. This SNP also appears to be associated with the incidence of larger dysplastic nidi and could be a candidate marker to screen patients harbouring lesions prone to increase in dimension by time or recur after treatment. Our results extend and strengthen the role of the 9p21 chromosomal region as a common risk factor for cerebrovascular diseases.
Contributors CLS designed the study, collected patients and controls, performed the statistical analyses and wrote the manuscript. IG performed the genetic analyses. AP designed the study and collected patients and controls. SD collected patients and controls, and helped with the design of the study. IGi performed the statistical analyses. AA collected patients and controls. CDR collected patients and controls, helped with the design of the study and supervised the project. GM supervised the project and helped with the design of the study. RP designed the study, helped with the genetic and statistical analyses, supervised the project, and wrote and edited the manuscript.
Funding The study was supported by a D1 intramural grant from the Catholic University School of Medicine, Rome, Italy.
Competing interests None.
Ethics approval The study was approved by the ethics committee of the Catholic University.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement The authors are happy to share their data with other investigators.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.