Angiopoietin-2 is an important mediator of angiogenesis, and we hypothesized that genetic variants of ANGPT2 (the gene encoding angiopoietin-2) would result in abnormal angiogenesis and contribute to stroke susceptibility. To test our hypothesis, we investigated the association of variants in the promoter of ANGPT2 with stroke in a multi-centre case-control study. We found that the C allele of rs3739390 conferred a 1.42-fold risk of lacunar infarction {adjusted OR (odds ratio), 1.42 [95% CI (confidence interval), 1.08–1.87]; P=0.012} and a 2.10-fold higher transcriptional activity than did the corresponding G allele rs3739390G. The haplotype G-G-T conferred a 1.54-fold risk of atherothrombotic stroke and a 1.64-fold risk for haemorrhagic stroke, whereas the haplotype G-C-C conferred approx. a 2.0-fold risk of each subtype of stroke. In conclusion, our results indicate that haplotypes in the promoter of ANGPT2 confer a high risk of stroke in a Chinese population.

INTRODUCTION

Stroke is the second most common cause of death worldwide and a major cause of long-term disability, with a great impact on public health. Each year, 2.5 million people have strokes, and more than 1 million die from stroke-related causes in China. Furthermore, more than 7 million patients have survived a stroke but are disabled [1]. Increased age, hypertension, cigarette smoking, alcohol abuse, DM (diabetes mellitus), hypercholesterolaemia and a history of coronary heart disease contribute to the risk of stroke [2].

Stroke is not one disease but rather a heterogeneous group of disorders reflecting different pathological processes, such as thrombotic vascular occlusion, embolism of carotid and cardiac sources, and intracerebrovascular rupture [3]. Evidence supports the involvement of angiogenesis in stroke. Acceleration of angiogenesis might be expected to enhance the outcome of stroke. Angiogenesis within the penumbra of a damaged brain region has been found after the onset of ischaemic stroke in humans, and the extent of angiogenesis has been correlated with the patient's survival time [4]. Atherosclerosis, which has been accepted as the common basis of most strokes, is also influenced by angiogenesis.

Angiogenesis is a highly regulated process that requires the orchestrated interaction of endothelial cells, extracellular matrix and surrounding cells. These cellular activities are mediated by a cascade of growth factors [5]. VEGF (vascular endothelial growth factor), its receptors and the angiopoietin/Tie2 pathway are believed to play fundamental roles in the process of angiogenesis.

Angiopoietin-2 is a context-dependent agonist/antagonist for the vascular-specific Tie2 receptor and is highly expressed in endothelial cells at sites of normal and pathological angiogenesis [6]. Exogenous angiopoietin-2 does not usually activate Tie2 in cultured endothelial cells [711], although some exceptions have been reported [1214]. In addition, angiopoietin-2 is selectively expressed in the endothelial cells of actively remodelling vessels, e.g. in the female reproductive system and in tumours [1518]. Moreover, transgenic expression of angiopoietin-2 results in embryonic lethality from vascular defects, similar to those observed in Angpt1 (angiopoietin-1)-null mice [8]. Together, these findings support the prevailing view that angiopoietin-2 is induced in remodelling endothelia as an autocrine Tie2 blocker and vessel-destabilizing agent [8,16,1820]. Subsequent analyses of Angpt2 (angiopoietin-2)-knockout mice show surprisingly that angiopoietin-2 acts as a Tie2 activator during lymphatic vessel development and is required for post-natal vascular remodelling in the eye [21]. Interestingly, the defects in vessel remodelling in Angpt2-null mice could not be rescued by angiopoietin-1, suggesting that angiopoietin-2 functions as a Tie2 antagonist [21]. More recently, Daly et al. [22] have reported a model in which angiopoietin-2 expression is induced in stressed endothelial cells, where it acts as an autocrine Tie2 agonist and protective factor. As with angiopoietin-1, angiopoietin-2 activates Tie2/Akt signalling in vivo and inhibits vascular leak [22], consistent with a role as a Tie2 activator.

Clinical studies have shown that angiopoietin-2 is up-regulated in patients with ACS (acute coronary syndrome) [23] and hypertension [24], both of which are risk factors of stroke. Although we still do not know whether elevated angiopoietin-2 levels are a cause or an effect of ACS and hypertension, increased angiopoietin-2 levels have been reported as a biomarker of cardiovascular risk (myocardial infarction, stroke and death) in patients with hypertension [25]. We, therefore, hypothesized that genetic variants of ANGPT2 (the gene encoding angiopoietin-2) would contribute to the susceptibility of stroke. To test the hypothesis, we investigated the association of variants in the promoter of ANGPT2 with stroke in a multi-centre case-control stroke study.

MATERIALS AND METHODS

Subjects

The Multi-Centre Chinese Stroke Study is a large prospective case-control study initiated in 2000, which was designed to assess genetic and environmental risk factors for stroke. The recruitment of stroke patients and controls has been described previously [26,27]. Briefly, both cohorts were recruited from the same demographic area and at the same time (from November 2000 to November 2001) from seven clinical centres (Yanzhou, Xi'an, Chongqing, Wuhan Xiehe, Wuhan Tongji, Beijing and Tianjin). Only patients with one of the following three subtypes of stroke were included: cerebral thrombosis, lacunar infarction and intracerebral haemorrhage. Patients were excluded when they were diagnosed as having transient ischaemic attack, subarachnoid haemorrhage, embolic brain infarction, brain tumours, cerebrovascular malformation or autoimmune diseases. Confirmation of stroke was based on the results of a strict neurological examination, CT (computer tomography) or MRI (magnetic resonance imaging), according to the International Classification of Diseases (9th revision) [27a]. Controls were frequency matched, selected simultaneously from the same demographic area as the patients and were matched for age (±5 years) and gender. Controls included 21.5% in-patients with minor illness from the Departments of Ophthalmology, Gastroenterology or Orthopaedics, and 78.5% community-based inhabitants free of neurological disease. The controls followed the same exclusion criteria as the patients.

Initially, 2000 patients and 2000 controls were recruited. Before data assessment, we excluded 905 subjects at different experimental stages because of a lack of definite diagnosis (24 cases), absence of plasma sample (76 cases and 93 controls), insufficient DNA sample (127 cases and 202 controls) and failure to detect ANGPT2 genotypes (260 cases and 123 controls). No significant differences were found in clinical characteristics between the included and the excluded subjects. Finally, 1513 patients (mean±S.D. age, 60.3±9.4 years; 62.7% male) and 1582 controls (mean±S.D. age, 59.7±8.3 years; 57.5% male) were eligible for genetic association analysis.

The study was approved by the Ministry of Science and Technology of China and the local ethics committees of the collaborative hospitals. All subjects were of Han nationality, and provided written informed consent.

Determination of biochemical variables and clinical data collection

Blood samples were collected after a 12 h overnight fast. In subjects with acute medical events, the drawing of blood was delayed for at least 6 weeks. The plasma was separated by centrifugation, and plasma and cell buffy coat were kept at −70 °C. Genomic DNA was extracted, and biochemical variables, including blood glucose, TC (total cholesterol), TAGs [triacylglycerols (triglycerides)] and HDL-C (high-density lipoprotein cholesterol), were determined within 3 months using an automatic analyser (Hitachi 7060). Non-HDL-C was calculated using the Friedewald formula. These variables were determined in a CDC (Center for Disease Control and Prevention)-qualified laboratory at FuWai hospital.

Genotyping of ANGPT2 variants

Genomic DNA was isolated from the white blood cell buffy coat as described previously [28]. Three SNPs (single nucleotide polymorphisms) were genotyped, including rs3739390 (+442G/C), rs2525507 (−220G/A) and rs3739391 (+398C/T), located in the ANGPT2 promoter region (5.0 kb). SNPs were selected based on the HapMap CHB (Chinese Han in Beijing) sample using the pairwise option of the Haploview version of the Tagger program. An r2 of 0.8 was selected as a threshold for all analyses. Among the three SNPs, rs3739390 is in strong LD (linkage disequilibrium) with five other SNPs according to the HapMap CHB sample: rs11996740, rs3739392, rs12546205, rs2922826 and rs17624399. These six variants are in complete LD (D′=1, r2=1) with one another. The SNP rs2515507 is in strong LD with six other variants, including rs1562715, rs2959769, rs2922816, rs2442509, rs2515508 and rs2515509. These seven variants are also in complete LD (D′=1, r2=1) with one another. The SNP rs3739391 does not capture other variants (r2=0.8).

These three variants were genotyped by using PCR RFLP (restriction-fragment-length polymorphism). Variant rs3739390 was analysed by amplification of a 753 bp sequence with the following primers: forward, 5′-GAACAAAGGACCGTGAAAGC-3′; and reverse, 5′-ACACCCTGGGATAGAATAAAAGA-3′. The resultant PCR products were digested with SmlI (New England Biolabs) and separated on 2% (w/v) agarose gels, which yielded two DNA fragments of 504 and 259 bp for the G allele, whereas the C allele was undigested.

Variant rs2515507 was analysed by amplification of a 1175 bp sequence with the following primers: forward, 5′-ACTTGCCCGAGTCTAATACCA-3′; and reverse, 5′-TTCACGGTCCTTTGTTCCTC-3′. The resultant PCR products were digested with SspI (New England Biolabs) and separated on 1% (w/v) agarose gels, which yielded two DNA fragments of 746 and 429 bp for the A allele, whereas the G allele was undigested.

Variant rs3739391 was analysed by amplification of a 228 bp sequence with the following primers: forward, 5′-TGCCACATTCTTTCTTCAGTAATA-3′; and reverse, 5′-GTTCTTCCCACTGCAATCTG-3′. The resultant PCR products were digested with Hpy188 (New England Biolabs) and separated on 4% (w/v) agarose gels, which yielded two DNA fragments of 141 and 87 bp for the C allele, whereas the T allele was undigested.

Reactions were performed with an ABI 9700 (Applied Biosystems) in a 96-well format. Oligo6.0 was used to design the primers.

A total of 3095 subjects were genotyped for these variants. Some genotyping results were confirmed by direct sequencing of the products with a DNA sequencer (ABI prism 377; PerkinElmer).

in vitro transcriptional activity of the ANGPT2 promoter variants

To study whether the variant +442C/G could influence ANGPT2 expression, ANGPT2 gene promoter transcriptional activity was compared among the different promoter constructs. The promoter regions (from −818 to +476) from homozygous individuals carrying the variant +442C or +442G were generated using PCR. The fragment was then subcloned into the pGL3-basic vector (Promega) using SacI and XhoI restriction enzymes. The reporter construct containing +442C was termed pGL3/+442C, and the one containing +442G was pGL3/+442G. The ANGPT2 promoter sequences in the constructs were confirmed by sequencing.

HUVECs (human umbilical vein endothelial cells) were purchased from Cascade Biologics. The reporter constructs (1.5 μg) were co-transfected with 50 ng of Renilla luciferase vector using Lipofectin® (Invitrogen). After 48 h, cells were collected and luciferase activity was determined using the dual-luciferase reporter assay system (Promega) with a luminometer (SIRIUS). Firefly luciferase expression levels were adjusted based on Renilla luciferase activity in controls. Three independent experiments were performed for each reporter construct.

Statistical analysis

Values are means±S.D. or medians (range). A χ2 test was used to test for qualitative variables, genotype/allele frequencies and for the HW (Hardy–Weinberg) equilibrium of the variants. Differences of the quantitative variables between groups were analysed using a Student's t test. Because the levels of plasma TAGs were highly skewed, the Mann–Whitney U test was used to examine differences in TAG levels between the groups. The associations between each variant and stroke and its subtypes were detected using multiple regression analyses. The covariates selected for the logistic regression models included conventional risk factors, as follows: age, gender, BMI (body mass index), cigarette smoking status and status of alcohol abuse. One-way ANOVA was used to compare the relative luciferase activities of the ANGPT2 promoter containing either +442C or +442G. All statistics were performed with the SPSS 13.0 package. A value of P<0.05 was taken as significant (two-tailed). A PHASE 2.0 software platform was used to analyse the haplotype association of the variants in the case-control study. The lowest frequency threshold for haplotype analysis was 0.05.

RESULTS

Clinical characteristics

The clinical characteristics of patients and controls are shown in Table 1. As expected, cases had a higher prevalence of conventional cardiovascular risk factors, including aging, cigarette smoking, alcohol intake, history of hypertension and DM, higher BP (blood pressure), higher levels of glucose and TAG, and lower levels of HDL-C, whereas TC was lower in cases, especially in those with haemorrhagic stroke (Table 1).

Table 1
Baseline characteristics of the subjects

Values are means±S.D., medians (range), or the number of individuals (%). P values are compared with the controls. DBP, diastolic BP; SBP, systolic BP.

    Stroke subtype 
Characteristic Controls (n=1582) Stroke patients (n=1513) P value Atherothrombosis (n=674) P value Lacunar (n=419) P value Haemorrhage (n=420) P value 
Age (years) 59.7±8.3 60.3±9.4 0.088 61.5±9.6 0.000 60.7±8.5 0.059 58.0±9.5 0.000 
Male gender (%) 57.5 62.7 0.003 63.2 0.012 63.2 0.034 61.4 0.149 
BMI (kg/m224.2±3.3 24.3±3.5 0.269 24.3±3.6 0.433 24.6±3.1 0.029 24.0±3.5 0.519 
SBP (mmHg) 129±18 147±23 0.000 147±23 0.000 142±21 0.000 152±24 0.000 
DBP (mmHg) 80±10 88±13 0.000 87±13 0.000 86±12 0.000 92±13 0.000 
TC (mmol/l) 5.01±1.00 4.73±1.02 0.000 4.86±1.03 0.001 4.74±0.97 0.000 4.53±0.99 0.000 
TAG (mmol/l) 1.48 (15.10) 1.65 (13.87) 0.000 1.72 (8.41) 0.000 1.71 (12.79) 0.000 1.46 (13.67) 0.667 
HDL-C (mmol/l) 1.06±0.30 0.90±0.28 0.000 0.86±0.26 0.000 0.92±0.26 0.000 0.89±0.32 0.000 
Non-HDL-C (mmol/l) 3.14±0.94 2.94±0.95 0.000 3.04±0.97 0.017 2.89±0.92 0.000 2.84±0.96 0.000 
Glucose (mmol/l) 5.88±1.74 6.68±2.73 0.000 6.87±2.86 0.000 6.47±2.71 0.000 6.58±2.51 0.000 
Cigarette smoking (%)   0.000  0.000  0.016  0.000 
 Never 62.0 52.1  48.5  56.8  53.2  
 Former 12.5 20.8  22.1  18.4  21.0  
 Current 25.5 25.5  29.4  24.8  25.8  
Alcohol intake (%)   0.001  0.007  0.427  0.002 
 Non-drinker 67.7 62.2  61.7  65.6  59.4  
 Drinker 32.3 37.8  38.3  34.4  40.6  
Hypertension history (%) 28.7 61.9 0.000 60.8 0.000 58.5 0.000 66.8 0.000 
DM history (%) 5.7 12.3 0.000 16.7 0.000 11.9 0.000 5.7 0.952 
    Stroke subtype 
Characteristic Controls (n=1582) Stroke patients (n=1513) P value Atherothrombosis (n=674) P value Lacunar (n=419) P value Haemorrhage (n=420) P value 
Age (years) 59.7±8.3 60.3±9.4 0.088 61.5±9.6 0.000 60.7±8.5 0.059 58.0±9.5 0.000 
Male gender (%) 57.5 62.7 0.003 63.2 0.012 63.2 0.034 61.4 0.149 
BMI (kg/m224.2±3.3 24.3±3.5 0.269 24.3±3.6 0.433 24.6±3.1 0.029 24.0±3.5 0.519 
SBP (mmHg) 129±18 147±23 0.000 147±23 0.000 142±21 0.000 152±24 0.000 
DBP (mmHg) 80±10 88±13 0.000 87±13 0.000 86±12 0.000 92±13 0.000 
TC (mmol/l) 5.01±1.00 4.73±1.02 0.000 4.86±1.03 0.001 4.74±0.97 0.000 4.53±0.99 0.000 
TAG (mmol/l) 1.48 (15.10) 1.65 (13.87) 0.000 1.72 (8.41) 0.000 1.71 (12.79) 0.000 1.46 (13.67) 0.667 
HDL-C (mmol/l) 1.06±0.30 0.90±0.28 0.000 0.86±0.26 0.000 0.92±0.26 0.000 0.89±0.32 0.000 
Non-HDL-C (mmol/l) 3.14±0.94 2.94±0.95 0.000 3.04±0.97 0.017 2.89±0.92 0.000 2.84±0.96 0.000 
Glucose (mmol/l) 5.88±1.74 6.68±2.73 0.000 6.87±2.86 0.000 6.47±2.71 0.000 6.58±2.51 0.000 
Cigarette smoking (%)   0.000  0.000  0.016  0.000 
 Never 62.0 52.1  48.5  56.8  53.2  
 Former 12.5 20.8  22.1  18.4  21.0  
 Current 25.5 25.5  29.4  24.8  25.8  
Alcohol intake (%)   0.001  0.007  0.427  0.002 
 Non-drinker 67.7 62.2  61.7  65.6  59.4  
 Drinker 32.3 37.8  38.3  34.4  40.6  
Hypertension history (%) 28.7 61.9 0.000 60.8 0.000 58.5 0.000 66.8 0.000 
DM history (%) 5.7 12.3 0.000 16.7 0.000 11.9 0.000 5.7 0.952 

Association of ANGPT2 variants with a risk of stroke

The genotype frequencies for the three variants were in accordance with HW equilibrium (Table 2). The estimated risk of subjects carrying +442G/C was significantly higher than those carrying +442C/C, but comparable with +442G/G carriers, indicating a dominant effect of the risk allele C. Similarly, the risk alleles +398T and −220A were also consistent with the dominant model. Thus the dominant model was applied to analyse the association of variants with stroke.

Table 2
ANGPT2 variants and the association with the risk of stroke

Controls, n=1582; atherothrombosis, n=674; lacunar, n=419; haemorrhage, n=420. *Determined using a χ2 test between patients and controls; †calculated by multivariate logistic regression analysis (adjusted for age, gender, BMI, and cigarette and alcohol consumption).

Variant Risk allele (%) P value for HW P value* Genotype (n  Crude OR (95% CI)* Adjusted OR (95% CI)† Adjusted P value† 
rs3739390   GG GC CC GC+CC   
 Controls 11.0 0.306  1258 (79.5%) 301 (19.0%) 23 (1.5%)  
 Stroke subtype          
  Atherothrombosis 12.0  0.344 524 (77.7%) 138 (20.5%) 12 (1.8%) 1.111 (0.893–1.383) 1.177 (0.912–1.517) 0.210 
  Lacunar 13.8  0.008 308 (73.5%) 107 (25.5%) 4 (1.0%) 1.399 (1.091–1.795) 1.4209 (1.080–1.867) 0.012 
  Haemorrhage 12.3  0.203 322 (76.7%) 92 (21.9%) 6 (1.4%) 1.182 (0.914–1.528) 1.194 (0.865–1.647) 0.281 
rs3739391   CC CT TT CT+TT   
 Controls 20.2 0.671  1004 (63.5%) 516 (32.6%) 62 (3.9%)  
 Stroke subtype          
  Atherothrombosis 22.2  0.188 408 (60.5%) 233 (34.6%) 33 (4.9%) 1.132 (0.941–1.363) 1.114 (0.897–1.382) 0.328 
  Lacunar 20.3  0.582 272 (64.9%) 124 (29.6%) 23 (5.5%) 0.939 (0.749–1.176) 0.961 (0.752–1.228) 0.750 
  Haemorrhage 21.9  0.682 262 (62.4%) 133 (31.7%) 25 (6.0%) 1.048 (0.839–1.308) 0.951 (0.721–1.255) 0.723 
rs2515507   GG GA AA GA+AA   
 Controls 5.8 0.052  1402 (88.6%) 179 (11.3%) 1 (0.1%)   
 Stroke subtype          
  Atherothrombosis 5.8  0.944 598 (88.7%) 73 (10.8%) 3 (0.4%) 0.990 (0.745–1.316) 1.069 (0.768–1.488) 0.694 
  Lacunar 5.2  0.355 378 (90.2%) 39 (9.3%) 2 (0.5%) 0.845 (0.591–1.208) 0.891 (0.602–1.318) 0.563 
  Haemorrhage 5.0  0.281 380 (90.5%) 38 (9.0%) 2 (0.5%) 0.820 (0.572–1.176) 0.963 (0.620–1.494) 0.865 
Variant Risk allele (%) P value for HW P value* Genotype (n  Crude OR (95% CI)* Adjusted OR (95% CI)† Adjusted P value† 
rs3739390   GG GC CC GC+CC   
 Controls 11.0 0.306  1258 (79.5%) 301 (19.0%) 23 (1.5%)  
 Stroke subtype          
  Atherothrombosis 12.0  0.344 524 (77.7%) 138 (20.5%) 12 (1.8%) 1.111 (0.893–1.383) 1.177 (0.912–1.517) 0.210 
  Lacunar 13.8  0.008 308 (73.5%) 107 (25.5%) 4 (1.0%) 1.399 (1.091–1.795) 1.4209 (1.080–1.867) 0.012 
  Haemorrhage 12.3  0.203 322 (76.7%) 92 (21.9%) 6 (1.4%) 1.182 (0.914–1.528) 1.194 (0.865–1.647) 0.281 
rs3739391   CC CT TT CT+TT   
 Controls 20.2 0.671  1004 (63.5%) 516 (32.6%) 62 (3.9%)  
 Stroke subtype          
  Atherothrombosis 22.2  0.188 408 (60.5%) 233 (34.6%) 33 (4.9%) 1.132 (0.941–1.363) 1.114 (0.897–1.382) 0.328 
  Lacunar 20.3  0.582 272 (64.9%) 124 (29.6%) 23 (5.5%) 0.939 (0.749–1.176) 0.961 (0.752–1.228) 0.750 
  Haemorrhage 21.9  0.682 262 (62.4%) 133 (31.7%) 25 (6.0%) 1.048 (0.839–1.308) 0.951 (0.721–1.255) 0.723 
rs2515507   GG GA AA GA+AA   
 Controls 5.8 0.052  1402 (88.6%) 179 (11.3%) 1 (0.1%)   
 Stroke subtype          
  Atherothrombosis 5.8  0.944 598 (88.7%) 73 (10.8%) 3 (0.4%) 0.990 (0.745–1.316) 1.069 (0.768–1.488) 0.694 
  Lacunar 5.2  0.355 378 (90.2%) 39 (9.3%) 2 (0.5%) 0.845 (0.591–1.208) 0.891 (0.602–1.318) 0.563 
  Haemorrhage 5.0  0.281 380 (90.5%) 38 (9.0%) 2 (0.5%) 0.820 (0.572–1.176) 0.963 (0.620–1.494) 0.865 

The frequency of homozygous or heterozygous +442C was significantly more common in patients with lacunar infarction than in the controls (26.5 compared with 20.5%; P=0.008). Multivariate analysis showed that +442C conferred a risk of lacunar infarction {adjusted OR (odds ratio), 1.42 [95% CI (confidence interval), 1.08–1.87]; P=0.012}, independent of conventional factors, including age, gender, BMI and cigarette or alcohol consumption. No association of variants −220G/A and +398C/T with any subtype of stroke was found. After Bonferroni correction, the association of variants +442 with lacunar infarction was still significant, where a P value of 0.017 (0.05/3) was considered significant.

We also tested haplotypes based on these markers for association. Haplotype analysis was performed by using PHASE 2.0 software, in which the common haplotype G-G-C (−220/+442/+398) was defined as the reference. As shown in Table 3, after adjustment for conventional risk factors using a multivariate regression model, the haplotype G-G-T (−220/+442/+398) was associated with an increased risk for atherothrombotic stroke [OR, 1.54 (95% CI, 1.15–2.08)] and haemorrhagic stroke [OR, 1.64 (95% CI, 1.17–2.32)], whereas the haplotype with the risk allele rs3739390 (G-C-C; −220/+442/+398) conferred a risk of atherothrombotic stroke [OR, 1.84 (95% CI, 1.18–2.87)], lacunar infarction [OR, 2.16 (95% CI, 1.30–3.57)] and haemorrhagic stroke [OR, 2.10 (95% CI, 1.27–3.46)].

Table 3
Impact of the ANGPT2 haplotypes (−220/+442/+398) on the risk of stroke

*Calculated using a χ2 test; †calculated by multivariate logistic regression analysis after adjustment for age, gender, BMI, and cigarette and alcohol consumption; ‡Bonferroni's multiple correction. NS, not significant.

  Stroke subtype Atherothrombosis compared with control Lacunar compared with control Haemorrhage compared with control 
Haplotype Control Atherothrombosis Lacunar Haemorrhage OR (95% CI)† P value Pcorr value‡ OR (95% CI)† P value Pcorr value‡ OR (95% CI)† P value Pcorr value‡ 
G-G-C 0.751 0.703 0.717 0.702 1.00   1.00   1.00   
G-G-T 0.082 0.118 0.101 0.125 1.54 (1.15–2.08) 0.004 0.016 1.30 (0.90–1.89) 0.161 NS 1.64 (1.17–2.32) 0.004 0.016 
G-C-C 0.031 0.053 0.059 0.060 1.84 (1.18–2.87) 0.006 0.024 2.16 (1.30–3.57) 0.002 0.008 2.10 (1.27–3.46) 0.003 0.012 
G-C-T 0.079 0.067 0.073 0.063 0.90 (0.63–1.29) 0.572 NS 0.98 (0.65–1.48) 0.911 NS 0.84 (0.54–1.30) 0.433 NS 
A-G-C 0.016 0.022 0.019 0.020 1.50 (0.79–2.88) 0.215 NS 1.32 (0.59–2.98) 0.497 NS 1.29 (0.58–2.89) 0.535 NS 
A-G-T 0.041 0.037 0.026 0.029 0.96 (0.60–1.59) 0.879 NS 0.67 (0.35–1.29) 0.229 NS 0.74 (0.40–1.40) 0.355 NS 
A-C-C 0.000 0.000 0.003 0.001 − − NS − − NS − − NS 
A-C-T 0.000 0.000 0.003 0.000 − − NS − − NS − − NS 
Global P value*  0.005 0.033 0.002          
  Stroke subtype Atherothrombosis compared with control Lacunar compared with control Haemorrhage compared with control 
Haplotype Control Atherothrombosis Lacunar Haemorrhage OR (95% CI)† P value Pcorr value‡ OR (95% CI)† P value Pcorr value‡ OR (95% CI)† P value Pcorr value‡ 
G-G-C 0.751 0.703 0.717 0.702 1.00   1.00   1.00   
G-G-T 0.082 0.118 0.101 0.125 1.54 (1.15–2.08) 0.004 0.016 1.30 (0.90–1.89) 0.161 NS 1.64 (1.17–2.32) 0.004 0.016 
G-C-C 0.031 0.053 0.059 0.060 1.84 (1.18–2.87) 0.006 0.024 2.16 (1.30–3.57) 0.002 0.008 2.10 (1.27–3.46) 0.003 0.012 
G-C-T 0.079 0.067 0.073 0.063 0.90 (0.63–1.29) 0.572 NS 0.98 (0.65–1.48) 0.911 NS 0.84 (0.54–1.30) 0.433 NS 
A-G-C 0.016 0.022 0.019 0.020 1.50 (0.79–2.88) 0.215 NS 1.32 (0.59–2.98) 0.497 NS 1.29 (0.58–2.89) 0.535 NS 
A-G-T 0.041 0.037 0.026 0.029 0.96 (0.60–1.59) 0.879 NS 0.67 (0.35–1.29) 0.229 NS 0.74 (0.40–1.40) 0.355 NS 
A-C-C 0.000 0.000 0.003 0.001 − − NS − − NS − − NS 
A-C-T 0.000 0.000 0.003 0.000 − − NS − − NS − − NS 
Global P value*  0.005 0.033 0.002          

Variant +442C increased ANGPT2 transcriptional activity

Dual luciferase assays were performed to determine the effect of the variant +442G/C on ANGPT2 promoter transcriptional activity. We generated two reporter constructs for the G or C alleles (pGL3/+442G and pGL3/+442C respectively), and transfected the constructs into HUVECs. +442C led to 2.10-fold higher transcriptional activity compared with +442G (P=0.014; Figure 1).

Impact of the −442G/C variant on ANGPT2 transcriptional activity

Figure 1
Impact of the −442G/C variant on ANGPT2 transcriptional activity

Dual-luciferase assays were performed to determine the effect of the −442G/C variant on ANGPT2 promoter activity. We generated two reporter constructs for either the G or C allele [pGL3/−442G (pGL3+442G) and pGL3/−442C (pGL3+442C) respectively], and transfected the constructs into HUVECs. *P<0.05 compared with pGL3/−442G. The pGL3-basic construct was used as a negative control without any promoter sequence.

Figure 1
Impact of the −442G/C variant on ANGPT2 transcriptional activity

Dual-luciferase assays were performed to determine the effect of the −442G/C variant on ANGPT2 promoter activity. We generated two reporter constructs for either the G or C allele [pGL3/−442G (pGL3+442G) and pGL3/−442C (pGL3+442C) respectively], and transfected the constructs into HUVECs. *P<0.05 compared with pGL3/−442G. The pGL3-basic construct was used as a negative control without any promoter sequence.

DISCUSSION

In the present study, we have found that the ANGPT2 promoter variant +442C increased the transcriptional activity of the gene and that this change was translated into a higher risk of stroke. The variant +442C conferred an increased risk of lacunar infarction. Haplotype G-G-T (−220/+442/+398) was associated with an increased risk of atherothrombotic stroke and haemorrhagic stroke, whereas the haplotype with the risk allele rs3739390 (G-C-C; −220/+442/+398) was associated with an almost doubling of the risk of stroke. Next, we examined whether the disease-associated allele was related to specific vascular risk factors, such as hypertension, diabetes, age and gender. No significant association was observed between the +442 allele and these risk factors, suggesting that the contribution of variant +442 to the risk of vascular disease is independent of those conventional vascular risk factors.

Population stratification may lead to a spurious association; however, in the present study, adjustment for clinical sites did not affect the association of the genotype with stroke development. The distribution of the genotypes was in agreement with HW equilibrium in patients and controls. Importantly, all subjects were of a Han nationality. Furthermore, we examined the population stratification effect by genotyping microsatellite markers randomly selected in the whole genome region among stroke patients and control subjects, and no association was found between these markers and stroke [27,29]. Although these strategies were taken to exclude any false associations, our results need replication in similar and divergent populations.

Stroke is not one disease, but rather a heterogeneous group of disorders reflecting different pathological processes. Atherothrombotic stroke mainly results from large-artery atherosclerosis, and intracerebral haemorrhages may be due to microaneurysms in intracerebral arteries (whether this is the main cause is still under debate), whereas lacunar infarction is usually caused by lipohyalinosis or microatheromata with thrombosis of the vascular lumen (approx. 30–100 μm). Although stroke results from a number of different pathological processes, predisposing factors for each stroke subtype are surprisingly similar, such as increased age, hypertension, smoking and DM.

The role of ischaemia in regulating intra-plaque neovessel formation has been postulated, and neoangiogenesis may represent a prerequisite for plaque growth [30,31]. Evidence supports the notion that angiopoietin-2 provides a key role in destabilizing the vasculature in a manner that is necessary for its subsequent remodelling [5]. Angiopoietin-2 was initially thought to be a natural antagonist of angiopoietin-1, blocking the stabilizing effects of angiopoietin-1 and facilitating endothelial cell activation in response to VEGF [8]. In the presence of endogenous VEGF, angiopoietin-2 promotes a rapid increase in capillary diameter, remodelling of the basal lamina and new vessel growth. On the other hand, studies have demonstrated that angiopoietin-2 has angiogenic activities in adult tissues and in cultured endothelial cells [13,14,22,32]. High concentrations of angiopoietin-2 in HUVECs [13] or angiopoietin-2 treatment in a particular condition [14] stimulates autophosphorylation of Tie2, which was associated with increased survival and tube formation in fibrin matrix respectively. Furthermore, angiopoietin-2 stimulates migration and tube-like structure formation of murine brain capillary endothelial cells through c-Fes and c-Fyn [32]. The autocrine angiopoietin-2 induced by the transcription factor FOXO-1 (forkhead box O-1) is an unexpected activator of the Tie2/Akt pathway and inhibits vascular leak [22]. In addition, angiopoietin-2 expression has been shown to correlate with MMP-2 (matrix metalloproteinase-2) activity in the atherosclerotic plaque and might play a role in the development of (unstable) plaque microvessels [33]. These findings indicate that high levels of angiopoietin-2 may promote plaque progression by increasing neoangiogenesis [13,14,22,32] and MMP-2 activity [33]. Consistent with this hypothesis, we found that the ANGPT2 promoter variant +442C increased ANGPT2 promoter transcriptional activity (2.10-fold) in vitro compared with the wild-type promoter and contributed to stroke risk.

SHRSP (spontaneously hypertensive stroke-prone rats), a model for genetic stroke susceptibility, suffer stroke spontaneously and injury becomes worse after experimental stroke, in part due to abnormal cerebro-vascular development. Wang et al. [34] found that angiopoietin-2 expression is significantly up-regulated in SHRSP compared with controls, whereas angiopoietin-1 expression fell markedly after stroke in SHRSP. They, therefore, proposed that angiopoietin-1 deficiency and angiopoietin-2 excess may lead to increased vascular permeability and a combination of increased cerebral oedema or inflammatory cell translocation, both of which are known to exacerbate stroke injury in humans [34]. Those findings are consistent with our present results.

In 2003, Wardlaw et al. [35] hypothesized that cerebral small-vessel endothelial (i.e. blood–brain barrier) dysfunction with leakage of plasma components into the vessel wall and surrounding brain tissue, which leads to neuronal damage, may contribute to the development of lacunar infarction stroke [35]. Indeed, angiopoietin-2 alone can cause inflammation in vivo by promoting vascular leakage [36], and the combination of VEGF and angiopoietin-2 may lead to blood–brain barrier disruption because it increases MMP-9 activity and inhibits ZO-1 (zonula occludens 1 protein) expression [37]. Thus +442C may contribute to a risk of lacunar infarction by increasing the transcriptional activity of angiopoietin-2.

In conclusion, the results of the present study provide evidence, for the first time to our knowledge, that the haplotypes in the promoter of ANGPT2 confer a high risk of stroke in a Chinese population.

FUNDING

This work was supported by the National High Technology Research and Development Programme of China (863 Programme) [grant number 2006AA02Z477 (to R.H.)], and the National Natural Science Foundation of China [grant number 30500199 (to J.C.)].

Abbreviations

     
  • ACS

    acute coronary syndrome

  •  
  • BMI

    body mass index

  •  
  • BP

    blood pressure

  •  
  • CI

    confidence interval

  •  
  • DM

    diabetes mellitus

  •  
  • HDL-C

    high-density lipoprotein cholesterol

  •  
  • HUVEC

    human umbilical vein endothelial cell

  •  
  • HW

    Hardy–Weinberg

  •  
  • LD

    linkage disequilibrium

  •  
  • MMP

    matrix metalloproteinase

  •  
  • OR

    odds ratio

  •  
  • SHRSP

    spontaneously hypertensive stroke-prone rats

  •  
  • SNP

    single nucleotide polymorphism

  •  
  • TAG

    triacylglycerol

  •  
  • TC

    total cholesterol

  •  
  • VEGF

    vascular endothelial growth factor

References

References
1
Tunstall-Pedoe
H.
Kuulasmaa
K.
Amouyel
P.
Arveiler
D.
Rajakangas
A. M.
Pajak
A.
Myocardial infarction and coronary deaths in the World Health Organization MONICA Project: registration procedures, event rates, and case-fatality rates in 38 populations from 21 countries in four continents
Circulation
1994
, vol. 
90
 (pg. 
583
-
612
)
2
Mitsios
N.
Gaffney
J.
Kumar
P.
Krupinski
J.
Kumar
S.
Slevin
M.
Pathophysiology of acute ischaemic stroke: an analysis of common signalling mechanisms and identification of new molecular targets
Pathobiology
2006
, vol. 
73
 (pg. 
159
-
175
)
3
Leys
D.
Deplanque
D.
Mounier-Vehier
C.
Mackowiak-Cordoliani
M. A.
Lucas
C.
Bordet
R.
Stroke prevention: management of modifiable vascular risk factors
J. Neurol.
2002
, vol. 
249
 (pg. 
507
-
517
)
4
Krupinski
J.
Kaluza
J.
Kumar
P.
Kumar
S.
Wang
J. M.
Role of angiogenesis in patients with cerebral ischaemic stroke
Stroke
1994
, vol. 
25
 (pg. 
1794
-
1798
)
5
Yancopoulos
G. D.
Davis
S.
Gale
N. W.
Rudge
J. S.
Wiegand
S. J.
Holash
J.
Vascular-specific growth factors and blood vessel formation
Nature
2000
, vol. 
407
 (pg. 
242
-
248
)
6
Yuan
H. T.
Khankin
E. V.
Karumanchi
S. A.
Parikh
S. M.
Angiopoietin 2 is a partial agonist/antagonist of tie2 signaling in endothelium
Mol. Cell. Biol.
2009
, vol. 
29
 (pg. 
2011
-
2022
)
7
Davis
S.
Papadopoulos
N.
Aldrich
T. H.
Maisonpierre
P. C.
Huang
T.
Kovac
L.
Xu
A.
Leidich
R.
Radziejewska
E.
Rafique
A.
, et al. 
Angiopoietins have distinct modular domains essential for receptor binding, dimerization and superclustering
Nat. Struct. Biol.
2003
, vol. 
10
 (pg. 
38
-
44
)
8
Maisonpierre
P. C.
Suri
C.
Jones
P. F.
Bartunkova
S.
Wiegand
S. J.
Radziejewski
C.
Compton
D.
McClain
J.
Aldrich
T. H.
Papadopoulos
N.
, et al. 
Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis
Science
1997
, vol. 
277
 (pg. 
55
-
60
)
9
Parikh
S. M.
Mammoto
T.
Schultz
A.
Yuan
H. T.
Christiani
D.
Karumanchi
S. A.
Sukhatme
V. P.
Excess circulating angiopoietin-2 may contribute to pulmonary vascular leak in sepsis in humans
PLoS Med.
2006
, vol. 
3
 pg. 
e46
 
10
Saharinen
P.
Kerkela
K.
Ekman
N.
Marron
M.
Brindle
N.
Lee
G. M.
Augustin
H.
Koh
G. Y.
Alitalo
K.
Multiple angiopoietin recombinant proteins activate the Tie1 receptor tyrosine kinase and promote its interaction with Tie2
J. Cell Biol.
2005
, vol. 
169
 (pg. 
239
-
243
)
11
Scharpfenecker
M.
Fiedler
U.
Reiss
Y.
Augustin
H. G.
The Tie-2 ligand angiopoietin-2 destabilizes quiescent endothelium through an internal autocrine loop mechanism
J. Cell Sci.
2005
, vol. 
118
 (pg. 
771
-
780
)
12
Harfouche
R.
Hussain
S. N.
Signaling and regulation of endothelial cell survival by angiopoietin-2
Am. J. Physiol. Heart Circ. Physiol.
2006
, vol. 
291
 (pg. 
H1635
-
H1645
)
13
Kim
I. J.
Kim
H.
Moon
S. O.
Kwak
H. J.
Kim
N. G.
Koh
G. Y.
Angiopoietin-2 at high concentration can enhance endothelial cell survival through the phosphatidylinositol 3′-kinase/Akt signal transduction pathway
Oncogene
2000
, vol. 
19
 (pg. 
4549
-
4552
)
14
Teichert-Kuliszewska
K.
Maisonpierre
P. C.
Jones
N.
Campbell
A. I.
Master
Z.
Bendeck
M. P.
Alitalo
K.
Dumont
D. J.
Yancopoulos
G. D.
Stewart
D. J.
Biological action of angiopoietin-2 in a fibrin matrix model of angiogenesis is associated with activation of Tie2
Cardiovasc. Res.
2001
, vol. 
49
 (pg. 
659
-
670
)
15
Goede
V.
Schmidt
T.
Kimmina
S.
Kozian
D.
Augustin
H. G.
Analysis of blood vessel maturation processes during cyclic ovarian angiogenesis
Lab. Invest.
1998
, vol. 
78
 (pg. 
1385
-
1394
)
16
Holash
J.
Maisonpierre
P. C.
Compton
D.
Boland
P.
Alexander
C. R.
Zagzag
D.
Yancopoulos
G. D.
Wiegand
S. J.
Vessel cooption, regression, and growth in tumors mediated by angiopoietins and VEGF
Science
1999
, vol. 
284
 (pg. 
1994
-
1998
)
17
Stratmann
A.
Risau
W.
Plate
K. H.
Cell type-specific expression of angiopoietin-1 and angiopoietin-2 suggests a role in glioblastoma angiogenesis
Am. J. Pathol.
1998
, vol. 
153
 (pg. 
1459
-
1466
)
18
Zagzag
D.
Hooper
A.
Friedlander
D. R.
Chan
W.
Holash
J.
Wiegand
S. J.
Yancopoulos
G. D.
Grumet
M.
In situ expression of angiopoietins in astrocytomas identifies angiopoietin-2 as an early marker of tumor angiogenesis
Exp. Neurol.
1999
, vol. 
159
 (pg. 
391
-
400
)
19
Hanahan
D.
Signaling vascular morphogenesis and maintenance
Science
1997
, vol. 
277
 (pg. 
48
-
50
)
20
Holash
J.
Wiegand
S. J.
Yancopoulos
G. D.
New model of tumor angiogenesis: dynamic balance between vessel regression and growth mediated by angiopoietins and VEGF
Oncogene
1999
, vol. 
18
 (pg. 
5356
-
5362
)
21
Gale
N. W.
Thurston
G.
Hackett
S. F.
Renard
R.
Wang
Q.
McClain
J.
Martin
C.
Witte
C.
Witte
M. H.
Jackson
H.
, et al. 
Angiopoietin-2 is required for postnatal angiogenesis and lymphatic patterning, and only the latter role is rescued by Angiopoietin-1
Dev. Cell
2002
, vol. 
3
 (pg. 
411
-
423
)
22
Daly
C.
Pasnikowski
E.
Burova
E.
Wong
V.
Aldrich
T. H.
Griffiths
J.
Ioffe
E.
Daly
T. J.
Fandl
J. P.
Papadopoulos
N.
McDonald
D. M.
, et al. 
Angiopoietin-2 functions as an autocrine protective factor in stressed endothelial cells
Proc. Natl. Acad. Sci. U.S.A.
2006
, vol. 
103
 (pg. 
15491
-
15496
)
23
Lee
K. W.
Lip
G. Y.
Blann
A. D.
Plasma angiopoietin-1. angiopoietin-2, angiopoietin receptor tie-2, and vascular endothelial growth factor levels in acute coronary syndromes
Circulation
2004
, vol. 
110
 (pg. 
2355
-
2360
)
24
Nadar
S. K.
Blann
A.
Beevers
D. G.
Lip
G. Y.
Abnormal angiopoietins 1&2, angiopoietin receptor Tie-2 and vascular endothelial growth factor levels in hypertension: relationship to target organ damage [a sub-study of the Anglo-Scandinavian Cardiac Outcomes Trial (ASCOT)]
J. Intern. Med.
2005
, vol. 
258
 (pg. 
336
-
343
)
25
Patel
J. V.
Lim
H. S.
Varughese
G. I.
Hughes
E. A.
Lip
G. Y.
Angiopoietin-2 levels as a biomarker of cardiovascular risk in patients with hypertension
Ann. Med.
2008
, vol. 
40
 (pg. 
215
-
222
)
26
Li
Z.
Sun
L.
Zhang
H.
Liao
Y.
Wang
D.
Zhao
B.
Zhu
Z.
Zhao
J.
Ma
A.
Han
Y.
, et al. 
Elevated plasma homocysteine was associated with hemorrhagic and ischemic stroke, but methylenetetrahydrofolate reductase gene C677T polymorphism was a risk factor for thrombotic stroke: a Multicenter Case-Control Study in China
Stroke
2003
, vol. 
34
 (pg. 
2085
-
2090
)
27
Wang
Y.
Zhang
W.
Zhang
Y.
Yang
Y.
Sun
L.
Hu
S.
Chen
J.
Zhang
C.
Zheng
Y.
Zhen
Y.
, et al. 
VKORC1 haplotypes are associated with arterial vascular diseases (stroke, coronary heart disease, and aortic dissection)
Circulation
2006
, vol. 
113
 (pg. 
1615
-
1623
)
27a
Practice Management Information Corporation
PMIC
International Classification of Diseases, 9th revision
Clinical Modification
1995
4th edn
Los Angeles
Practice Management Information Corporation
28
Miller
S. A.
Dykes
D. D.
Polesky
H. F.
A simple salting out procedure for extracting DNA from human nucleated cells
Nucleic Acids Res.
1988
, vol. 
16
 pg. 
1215
 
29
Pritchard
J. K.
Rosenberg
N. A.
Use of unlinked genetic markers to detect population stratification in association studies
Am. J. Hum. Genet.
1999
, vol. 
65
 (pg. 
220
-
228
)
30
Williams
J. K.
Armstrong
M. L.
Heistad
D. D.
Vasa vasorum in atherosclerotic coronary arteries: response to vasoactive stimuli and regression of atherosclerosis
Circ. Res.
1988
, vol. 
62
 (pg. 
515
-
523
)
31
Patterson
J. C.
Capillary rupture with intimal hemorrhage as a causative factor in coronary thrombosis
Arch. Pathol.
1938
, vol. 
25
 (pg. 
474
-
487
)
32
Mochizuki
Y.
Nakamura
T.
Kanetake
H.
Kanda
S.
Angiopoietin 2 stimulates migration and tube-like structure formation of murine brain capillary endothelial cells through c-Fes and c-Fyn
J. Cell Sci.
2002
, vol. 
115
 (pg. 
175
-
183
)
33
Post
S.
Peeters
W.
Busser
E.
Lamers
D.
Sluijter
J. P.
Goumans
M. J.
de Weger
R. A.
Moll
F. L.
Doevendans
P. A.
Pasterkamp
G.
Vink
A.
Balance between angiopoietin-1 and angiopoietin-2 is in favor of angiopoietin-2 in atherosclerotic plaques with high microvessel density
J. Vasc. Res.
2008
, vol. 
45
 (pg. 
244
-
250
)
34
Wang
M. M.
Klaus
J. A.
Joh
H. D.
Traystman
R. J.
Hurn
P. D.
Postischemic angiogenic factor expression in stroke-prone rats
Exp. Neurol.
2002
, vol. 
173
 (pg. 
283
-
288
)
35
Wardlaw
J. M.
Sandercock
P. A.
Dennis
M. S.
Starr
J.
Is breakdown of the blood-brain barrier responsible for lacunar stroke, leukoaraiosis, and dementia?
Stroke
2003
, vol. 
34
 (pg. 
806
-
812
)
36
Roviezzo
F.
Tsigkos
S.
Kotanidou
A.
Bucci
M.
Brancaleone
V.
Cirino
G.
Papapetropoulos
A.
Angiopoietin-2 causes inflammation in vivo by promoting vascular leakage
J. Pharmacol. Exp. Ther.
2005
, vol. 
314
 (pg. 
738
-
744
)
37
Zhu
Y.
Lee
C.
Shen
F.
Du
R.
Young
W. L.
Yang
G. Y.
Angiopoietin-2 facilitates vascular endothelial growth factor-induced angiogenesis in the mature mouse brain
Stroke
2005
, vol. 
36
 (pg. 
1533
-
1537
)