Donate Help Contact The AHA Sign In Home
American Heart Association
Arteriosclerosis, Thrombosis, and Vascular Biology
Search: search_blue_button Advanced Search
« Previous Article | Table of Contents | Next Article »
Arteriosclerosis, Thrombosis, and Vascular Biology. 2007;27:2250-2257
Published online before print July 26, 2007, doi: 10.1161/ATVBAHA.107.149468
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Data Supplement
Right arrow All Versions of this Article:
27/10/2250    most recent
ATVBAHA.107.149468v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Morange, P.E.
Right arrow Articles by Juhan-Vague, I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Morange, P.E.
Right arrow Articles by Juhan-Vague, I.
(Arteriosclerosis, Thrombosis, and Vascular Biology. 2007;27:2250.)
© 2007 American Heart Association, Inc.

Thrombosis

Association of Plasminogen Activator Inhibitor (PAI)-1 (SERPINE1) SNPs With Myocardial Infarction, Plasma PAI-1, and Metabolic Parameters

The HIFMECH Study

P.E. Morange; N. Saut; M.C. Alessi; J.S. Yudkin; M. Margaglione; G. Di Minno; A. Hamsten; S.E. Humphries; D.A. Tregouet; I. Juhan-Vague

From INSERM, U626 (P.E.M., N.S., M.C.A., I.J.V.), Université de la Méditerranée, Marseille, France; the Diabetes and Cardiovascular Disease Academic Unit (J.S.Y.), Archway Campus, Royal Free and University College Medical School, London, UK; Instituto di Ricovero e Cura a Carattere Scientifico (M.M., G.D.M.), Ospedale Casa Sollievo della Sofferenza, San Giovani Rotondo, Italy; King Gustaf V Research Institute and Departments of Medicine and Cardiology (A.H.), Karolinska Hospital, Karolinska Institute, Stockholm, Sweden; the Center for Cardiovascular Genetics (S.E.H.), British Heart Foundation Laboratories, Rayne Building, Royal Free and University College Medical School, London, UK; and INSERM, UMR S 525 (D.A.T.), Université Pierre et Marie Curie-Paris6, UMR S 525, Paris, F-75634 France.

Correspondence to P.E. Morange, Lab. Hematology, CHU Timone. 264, Rue Saint-Pierre 13385 Marseille Cedex 05, France. E-mail pmorange{at}ap-hm.fr


*    Abstract
up arrowTop
 *Abstract
down arrowIntroduction
 down arrowMaterials and Methods
 down arrowResults
 down arrowDiscussion
 down arrowAppendix
 down arrowReferences
 
Objective— The purpose of this study was to investigate the effects of plasminogen activator inhibitor-1 (PAI-1) gene (SERPINE1) single nucleotide polymorphisms (SNPs) on the risk of myocardial infarction (MI), on PAI-1 levels, and factors related to the metabolic syndrome.

Methods and Results— Eleven SNPs capturing the common genetic variation of the SERPINE1 gene were genotyped in the HIFMECH study. In the 510 male cases and their 543 age-matched controls, a significant gene-smoking interaction was observed. In nonsmokers, the rs7242-G allele was more frequent in cases than in controls (0.486 versus 0.382, P=0.013) whereas the haplotype derived from the rs2227631 (–844A>G)-G and rs2227683-A alleles was {approx}3-fold lower in cases than in controls (0.042 versus 0.115, P=0.006). SERPINE1 haplotypes explained 3.5% (P=0.007) of the variability of PAI-1 levels, which was attributable to the combined effects of 3 SNPs, –844A>G, rs2227666, and rs2227694. The rs6092 (Ala15Thr) and rs7242 SNPs acted additively to explain 4.4% of the variability of plasma insulin levels and 1.6% of the variability of BMI (P<10–3 and P=0.023, respectively).

Conclusions— SERPINE1 haplotypes are mildly associated with plasma levels of PAI-1 and with the risk of MI in nonsmokers. They are also associated with insulin levels and BMI.

HIFMECH is a European case-control study for myocardial infarction (MI). In the 510 male cases and their 543 age-matched controls, SERPINE1 haplotypes were mildly associated with plasma levels of PAI-1 and with the risk of MI in nonsmokers. They were also associated with insulin levels and BMI.


Key Words: metabolic syndrome • myocardial infarction • PAI-1 • SERPINE1


*    Introduction
up arrowTop
 up arrowAbstract
 *Introduction
down arrowMaterials and Methods
 down arrowResults
 down arrowDiscussion
 down arrowAppendix
 down arrowReferences
 
In blood, fibrinolysis breaks down fibrin and maintains vessel patency, and in tissues it breaks down the extracellular matrix and controls cell adhesion and migration and thus participates in tissue remodeling. Fibrinolysis is primarily regulated by plasminogen activator inhibitor type-1 (PAI-1), which controls the extent of this potentially destructive protease system.1–3

Increased PAI-1 levels may predispose patients to the formation of atherosclerosis plaque prone to rupture, with a high lipid-to-vascular smooth muscle cells ratio as a result of decreased cell migration.4 In humans, there is clinical evidence that increased PAI-1 levels are associated with atherothrombosis.5,6 In large epidemiological studies, elevated plasma PAI-1 levels have been identified as a predictor of myocardial infarction (MI).7–11 Remarkably, the predictive ability of PAI-1 disappears after adjustment for markers of the metabolic syndrome (MetS),8,12–16 suggesting that the MetS is a prerequisite to high plasma PAI-1 levels in patients prone to atherothrombosis. Moreover, it has been hypothesized that PAI-1 participates in the development of key features of the MetS. Indeed, several studies17–20 showed that high plasma PAI-1 levels independently predict the development of type II diabetes. Whether PAI-1 plays a direct role in MI, MetS, or diabetes, or is only a bystander, is difficult to assess in humans. One way to verify this hypothesis is to look for a relation between single nucleotide polymorphisms (SNPs) influencing PAI-1 expression or activity, MI, and parameters belonging to the MetS. Several SERPINE1 (formerly PAI-1) gene SNPs have been identified,21–23 among which the polymorphism 4G/5G (rs1799889) located in position –675 of the promoter region has been quite extensively studied. The 4G allele has been shown to be associated with increased SERPINE1 transcription compared with the 5G allele in in vitro studies21,24 and with increased plasma PAI-1 levels in vivo.22 A large systematic review found that the 4G/4G genotype was associated with a modest 1.2-fold increased risk of MI.25 We have shown, in a previous report of the HIFMECH study, that the –675 4G/5G polymorphism is associated with the risk of MI but that this effect is considerably influenced by the presence of underlying MetS.26 Results about the relation between this polymorphism and variables related to the MetS were however contradictory, carriers of the 4G allele being more prone to obesity and MetS in some studies22,23,26–28 but not in others.29–32 One possible explanation for these observed discrepancies could be that other SERPINE1 SNPs in linkage disequilibrium (LD) with the 4G/5G polymorphism are associated with MetS, suggesting that haplotype analysis of SERPINE1 SNPs could be of great interest. For example, another potentially functional polymorphism of the promoter,33 –844A>G (rs2227631), in strong LD with the 4G/5G polymorphism, could be responsible, instead of the 4G/5G, for the association between PAI-1, MetS, and MI. Three haplotype association studies have recently been performed to address the influence of SERPINE1 SNPs on PAI-1 plasma levels and on the risk of cardiovascular disease.34–36 Kathiresan et al34 have shown that SERPINE1 SNPs explained about 5% of the variability of PAI-1 plasma levels and this association could be attributable to 3 SNPs, rs6465787, rs2227674, and –844A>G. This haplotype analysis could not completely exclude the possibility that the effect of the –844A>G SNP was the consequence of its strong LD with the –675 4G/5G. On the contrary, Ding et al,35 using a similar approach, showed that the SERPINE1 effect on PAI-1 plasma levels seems to be restricted to the –675 4G/5G polymorphism, however, they did not study the effect of the –844A>G. Neither of these studies found any relation between common haplotypes of the SERPINE1 and the overall risk of cardiovascular disease. However, Su et al,36 in a group of Chinese subjects, detected a SERPINE1-smoking interaction on CHD risk, such as the main haplotype carrying the –844A and –6754G allele significantly increased the risk of CHD in nonsmokers only.

The aim of our study was to simultaneously evaluate, in a case-control study of White individuals, the association of SERPINE1 SNPs with MI, plasma PAI-1 levels, and metabolic parameters using a haplotype-based approach. We aimed to test whether the association already described between the SERPINE1 variants and myocardial infarction could be first modulated by smoking and secondly be partly the result of the relationship between these variants and some features of the metabolic syndrome.


*    Materials and Methods
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 *Materials and Methods
down arrowResults
 down arrowDiscussion
 down arrowAppendix
 down arrowReferences
 
Study Subjects
Full details of the study design and recruitment criteria are presented elsewhere.26 Male survivors of a first MI aged <60 years (excluding patients with familial hypercholesterolemia and insulin dependent diabetes mellitus) and population-based individuals of the same age were recruited from the 4 centers as part of the HIFMECH study: Stockholm (Sweden), London (UK), Marseille (France), and San Giovanni Retondo (Italy). Consecutive patients were invited to participate, along with randomly selected healthy individuals from the same catchment areas. In all, a total of 510 postinfarction patients and 543 controls were included in the present study. Postinfarction patients were investigated 3 to 6 months after the acute event. Patients and control subjects were examined in parallel in the early morning after an overnight fast. Height and weight were recorded and the body mass index (BMI) was calculated as kg m–2. Smokers were considered as current or ex-smokers at the time of the MI onset.

Determination of PAI-1 antigen was centrally performed with a commercially available kit (Asserachrom PAI-1; Stago). Each plasma sample was run in duplicate. Interassay variation coefficient of pooled plasma from 30 healthy volunteers was 8%. Assay method for insulin has been described.37

Choice of PAI-1 Tag Polymorphisms
SERPINE1 has been sequenced by the Seattle SNPs program for Genomics Application project in 23 individuals of European ancestry (http://pga.gs.washington.edu/). From the identified SNPs spanning 13 kb of the SERPINE1 gene, the minimum number of SNPs (tag SNP) required to characterize 100% of the haplotypic diversity of the SERPINE1 gene was determined. 10 tag SNP with minor allele frequency >0.04 were found to be enough to fully characterize this gene and were further genotyped in the HIFMECH study. These are rs2227631 (–844A>G), rs6092 (Ala15Thr), rs7242, rs2227708, rs2227662, rs2227666, rs2227667, rs2227672, rs2227683, rs2227694. The rs1799889 (–675 4G/5G) was also genotyped as it is widely used in SERPINE1 genetic studies, has been shown to be functional,24 and is associated with PAI-1 plasma levels.

Genotyping were performed under contract by Kbioscience, Cambridge, UK (http://www.kbioscience.co.uk), except for 4G-675 5G, which was genotyped by allele specific polymerase chain reaction (PCR), using the following primers (5'>3'): forward: TCAGCCAGACAAGGTTGTTG, reverse: TTTTCCCCCAGGGCTGTCCA, 4G: GTCTGGACACGTGGGGA, 5G: GTCTGGACACGTGGGGG. PCR conditions were an initial denaturation step of 1’30 at 95°C, followed by 35 cycles of these 3 steps: 95°C: 30", 62°C: 45", 72°C: 1’ and then a final extension step of 5’ at 72°C. PCR were then kept at 15°C for immediate use or frozen for later use.

Statistical Analysis
Allele frequencies were estimated by gene counting, and departure from Hardy-Weinberg (HW) equilibrium was testing using a {chi}2 with 1 degree of freedom. Allele frequencies were compared between cases and controls by use of a {chi}2 with 1 degree of freedom. Conditional logistic regression analysis for matched case-control study was used to investigate the association between MI and explanatory variables. Genotypic association of SERPINE1 polymorphisms with plasma PAI-1 and insulin levels was first investigated by use of a classical linear model. Plasma PAI-1 and insulin were square-root and log-transformed to remove positive skewness, respectively.

LD analysis was carried by the THESIAS software38 (www. genecanvas.org) based on the SEM algorithm.39 The extent of LD was expressed in terms of D'.40 THESIAS was also used for haplotype analyses. For the haplotype analyses, systematic analyses of all possible combinations of 1 to 9 polymorphisms were carried out to reduce the haplotype dimension and to search for the most informative and parsimonious haplotype configuration in terms of prediction of the phenotypes variability using the previously described Akaike’s Information Criterion-based strategy.41,42 The homogeneity of allelic and haplotypic effects across North and South or across cases and controls was assessed by the Mantel-Haenszel statistics.43 All analyses were adjusted for age, gender, smoking, center, and case-control status when appropriate. A probability value of <0.05 was taken as statistically significant.


*    Results
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMaterials and Methods
 *Results
down arrowDiscussion
 down arrowAppendix
 down arrowReferences
 
Baseline Characteristics of Cases and Controls
Cases were more likely to be smokers and suffer from diabetes. Parameters of the MetS such as BMI, plasma insulin, and triglycerides (TG) levels were significantly higher in cases than in controls (supplemental Table I, available online at http://atvb.ahajournals.org). PAI-1 levels were significantly higher in cases than in controls (40.10±28.46 versus 29.48±22.50 ng/mL, P<10–4), an effect seen in both the North (42.64 versus 31.85 ng/mL) and South (38.28 versus 27.82 ng/mL). As expected, plasma PAI-1 levels were highly correlated with insulin levels, {triangleup}=0.44 and {triangleup}=0.37 in controls and cases, respectively (both P<10–4), but also to BMI ({triangleup}=0.43 and {triangleup}=0.24, P<10–4, respectively) and with TG ({triangleup}=0.38 and {triangleup}=0.30, P<10–4, respectively).

Description of Studied SERPINE1 SNPs
Among the 11 genotyped polymorphisms, 1 was found to be nonpolymorphic (rs2227662), whereas the rs2227708 was relatively rare (frequency of 0.012 in the whole HIFMECH study). Therefore, as shown in Figure 1, the present analysis focused on 9 polymorphisms, rs2227631 (–844A>G), rs1799889 (–675 4G/5G), rs6092 (Ala15Thr), rs2227666, rs2227667, rs2227672, rs2227683, rs2227694, and rs7242 (+11061 T>G). The genotype distribution of all the SNPs were in HW equilibrium and their allele frequencies were very similar in North and South (supplemental Table II). Pairwise LD was relatively strong between all polymorphisms, except for the rs2227683 and rs2227694 that showed moderate LD with the –844A>G and –675 4G/5G SNPs. As a consequence, 9 haplotypes with frequency greater than 3% were inferred and accounted for about 93% of the whole chromosomes (see below). As the pattern of LD and the resulting haplotypic structure were very similar in North and South (supplemental Table III), the following analyses were performed on the whole HIFMECH sample, while checking for this homogeneity of the associations across regions.


Figure 1
View larger version (12K):
[in this window]
[in a new window]

  		Figure 1. Location of the 11 polymorphisms analyzed in the SERPINE1 gene. Usual names, when existing, are given with the rs numbers. The exons are denoted as squares (white: untranslated, black: coding region). The 9th exon contains the end of the coding sequence and the 3'UTR.

Association of SERPINE1 SNPs With MI
In the whole sample, none of the polymorphisms was significantly associated with MI either using single-locus (supplemental Table IV) or haplotype (Table 1) analyses. The rs6092-Thr allele was carried by only 1 haplotype and tended to be less frequent in cases than in controls (0.099 versus 0.125), but this difference failed to reach significance (P=0.07). However, the association of SERPINE1 SNPs with MI was found to be modulated by smoking. Two SNPs were associated with MI only in nonsmokers (supplemental Table V). Although SERPINE1 haplotypes were not associated with MI in smokers (P=0.42), they were highly associated with MI in nonsmokers (P=0.004). The best model in terms of predicting MI status in nonsmokers included 3 polymorphisms, –844A>G, rs2227683, and rs7242 (Table 2). Table 2 also includes the information on the –675 4G/5G, to examine in more detail its contribution on the risk of MI. Consistent with univariate analysis, the rs7242-G allele carried by 1 frequent haplotype (H2) was more frequent in cases than in controls (0.486 versus 0.383, P=0.013). Because the effect of this haplotype that also carries the –844A allele was not significantly different from the 3 other haplotypes carrying the –844A allele (H1, H3, H4; P=0.27), it cannot be completely excluded that the effect of the rs7242 SNP was attributable to its LD with the –844A>G. In addition, the frequency of the haplotype carrying the –844G and rs2227683-A alleles (H6) was {approx}3-fold lower in nonsmoker cases than in nonsmoker controls (0.042 versus 0.115, P=0.006). It is important to note that the –6754G/5G does not participate in this gene x smoking interaction.


View this table:
[in this window]
[in a new window]

  		Table 1. Association of Main SERPINE1 Haplotypes With MI in the HIFMECH Study


View this table:
[in this window]
[in a new window]

  		Table 2. Association Between MI and Main SERPINE1 Haplotypes Derived From the rs2227631, rs1799889, rs2227683, and rs7242 Polymorphisms According to Smoking

Association of SERPINE1 SNPs With PAI-1 Levels
Although effects were generally larger in cases than control (supplemental Table VI), there was no significant evidence for genetic effect heterogeneity (all P>0.15) nor across smokers and non-smokers (data not shown). Therefore, Table 3 provides a full description of the single-SNP association analyses in the whole HIFMECH study. Four SNPs, –844A>G, –675 4G/5G, rs2227667, and rs2227672, were significantly associated with PAI-1 levels. Overall, the percentage of variance explained by these SNPs were 1.25% (P=0.002), 1.12% (P=0.004), 0.77% (P=0.02), and 0.72% (P=0.03), respectively. SERPINE1 haplotypes were significantly associated with PAI-1 levels (P=0.007) and explained 3.8% (P=0.05) and 3.1% (P=0.11) of the variability of PAI-1 levels in controls and cases, respectively (supplemental Table VII). The best model found in the systematic exploration of haplotype effects in relation to PAI-1 levels included the –844A>G (already found to be associated with the risk of MI in nonsmokers), rs2227666, and rs2227694 polymorphisms. Detailed haplotype analysis of these 3 polymorphisms is summarized in Figure 2. The information on the –6754G/5G SNP is also provided to get better insight into its contribution on PAI-1 levels variability. Firstly, the –844G allele was carried by 2 haplotypes that differed only at position rs2227694 and were both associated with similar PAI-1 levels (2.33 versus 2.14, P=0.51). By comparison to the most frequent A[4G]GG haplotype, these 2 haplotypes were associated with lower PAI-1 levels (2.55 versus 2.30,P=0.02), a result compatible with an increasing effect on PAI-1 levels of the –844A allele. The observation that the unique haplotype pair A[4G]GG and A[5G]GG that differed only at the –6754G/5G locus did not show difference in PAI-1 levels (2.55 versus 2.59, P=0.87) would additionally suggest that the effect of the –675 to 4G/5G polymorphism observed in univariate analysis was the consequence of its LD with other SERPINE1 SNPs. In addition, 2 haplotypes, A[4G]GA and A[4G]AG, were associated with higher mean PAI-1 levels than the most frequent haplotype (3.02 versus 2.55 and 3.01 versus 2.55, respectively; P=0.03 for both). Because the A[4G]AG haplotype is the only one carrying the rs2227666 A allele, these results would suggest an increasing effect on PA-I levels of the rs2227666A allele in addition to an increasing effect of the rs222794 A allele when associated with the –844A allele on the same haplotype. These effects were similar in controls and cases, in North and South, and in smokers and nonsmokers (data not shown).


View this table:
[in this window]
[in a new window]

  		Table 3. Association Between PAI-1 Gene Polymorphisms and PAI-1 Levels in the HIFMECH Study (n=948)


Figure 2
View larger version (14K):
[in this window]
[in a new window]

  		Figure 2. Association between plasma PAI-1 levels and SERPINE1 haplotypes derived from the rs2227631 (A/G), rs1799889 (4G/5G), rs2227666 (G/A), and rs2227694 (G/A) polymorphisms. Polymorphisms are ordered according to their position on the genomic sequence. Each bar and its 95% CI brackets corresponds to the expected mean of PAI-1 levels (square rooted) associated with 1 dose of haplotype under the assumption of additive haplotype effects.

Association of SERPINE1 SNPs With Metabolic Parameters
In univariate analysis, the rs7242 was found to be significantly associated with insulin levels in cases only (supplemental Table VIII). No single locus nor haplotype effects were observed in controls (test for homogeneity between cases and controls for the rs7242 P=0.039, supplemental data). Conversely, haplotype analysis revealed that the Ala15Thr and the rs7242 SNPs were strongly associated with insulin levels in cases. These 2 SNPs defined 3 haplotypes (Table 4) that were highly associated with insulin (R2=4.4%, {chi}2=15.89 with 2 df, P<10–3). By comparison to the most frequent Ala-T haplotype, both Ala-G and Thr-T haplotypes were associated with higher insulin levels (Table 4). These results were compatible with independent and increasing effects of the Thr15 (+0.24 [0.05 to 0.43], P=0.015) and rs7242 G (+0.17 [0.08 to 0.27], P<10–3) alleles. These effects were similar in smokers and nonsmokers (data not shown), and were independent of other SERPINE1 SNPs such as the –844A>G, –675 4G/5G, and rs227683 (supplemental Table IX). The same pattern of association was observed with BMI, the rs6092 and rs7242 SNPs explaining 1.6% of the variability of BMI (P=0.023) in cases only (supplemental Table X). No SERPINE1 genotype or haplotype was associated with a significant effect on TG levels.


View this table:
[in this window]
[in a new window]

  		Table 4. SERPINE1 Haplotype Analysis of the rs6092 and rs7242 Polymorphisms in Relation to Insulin Levels According to Case–Control Status


*    Discussion
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMaterials and Methods
 up arrowResults
 *Discussion
down arrowAppendix
 down arrowReferences
 
The main findings of the study are that different SERPINE1 haplotypes are associated with the risk of MI in nonsmokers and with plasma levels of PAI-1 in both cases and controls and with insulin and BMI in cases. In all, 6 SNPs were associated with these different clinical and biological phenotypes (Figure 1). The –844A>G is particularly of interest as it is both related with MI in nonsmokers and with plasma PAI-1 levels. As previously reported in Chinese Han subjects,36 a highly significant gene x smoking interaction was detected, characterized by a strong association of SERPINE1 haplotypes with MI in nonsmokers only. In nonsmokers, the rs7242 SNP was associated with MI, as the rs7242-G allele was more frequent in cases than in controls (0.51 versus 0.41). This allele was carried out by only 1 haplotype, a haplotype that also carries the –844A and –675 to 4G alleles, and that was found to be associated with higher risk of MI in nonsmoking Chinese.36 Because of the strong LD, it is not possible to determine whether rs7242 or the –844A>G is responsible for the observed association. However, haplotype analysis revealed that the –675 4G/5G is unlikely to explain this association. In addition, we observed that the frequency of the haplotype defined by the –844-G and rs2227683-A alleles was {approx}3-fold lower in cases than in controls. Thus, the –844A>G polymorphism is more relevant than the –675 4G/5G for the association with MI in nonsmokers.

The fact that the impact of the SERPINE1 polymorphisms on MI was only observed in nonsmokers remains puzzling. We can hypothesize that SERPINE1 SNPs effect is relatively modest so that it is overwhelmed by the strong effect of smoking on risk. However, only 84 cases and 201 controls were nonsmokers, and this result must be replicated in a larger study conducted in nonsmokers. In HIFMECH, besides its effect on the risk of MI in nonsmokers, the –844A>G SNP was also associated with PAI-1 plasma levels. It must be underlined that this SNP explained only a small amount of PAI-1 levels variability (1.25%) and that this association may not be of clinical relevance. As it has been also implicated in the regulation of the SERPINE1 gene, as a part of an Ets nuclear protein consensus sequence binding site,33 the effect of the –844A>G could be attributable to modifications of SERPINE1 expression. In the present study, the observation that adjustment for PAI-1 plasma levels did not modify the relation between SERPINE1 haplotypes and the MI risk in nonsmokers suggests that PAI-1 plasma levels might not reflect the local expression of PAI-1. This latter hypothesis is supported by the fact that PAI-1 mRNA expression is increased in the endothelial cells located in the vicinity of thrombi, and that this overexpression is not correlated with plasma PAI-1 antigen levels.44

Besides a direct effect via modification of PAI-1 expression in the arterial wall, SERPINE1 polymorphisms could influence the risk of MI by influencing well-known cardiovascular risk factors such as the MetS. Indeed, several studies conducted in vitro and in vivo support a role of SERPINE1 in the development MetS (reviewed in45). This led us to study the association between SERPINE1 SNPs and several variables of the MetS such as BMI, insulin, and TG. Haplotype analysis revealed that 2 polymorphisms, rs6092 and rs7242, independently affect BMI and plasma insulin levels in cases. The rs6092 could be of functional importance as it is an Ala15Thr located in the central hydrophobic core of the PAI-1 signal peptide.23 The Ala15 allele increases hydrophobicity and {alpha}-helix propensity, indicating that it could stabilize the {alpha}-helix confirmation of the signal peptide. These 2 properties are known to be important in signal peptide function and, therefore, the mutations might modulate the secreted PAI-1 level. Lopes et al23 have shown that the Ala15 allele tended to be associated with a higher risk of CHD in diabetic subjects. In the present study, in univariate analysis, the Ala15 allele tended to be more frequent in individuals with MI as compared with those without, however this difference did not reach significance (P=0.06). As regards the rs7242 polymorphism located in the 3' untranslated region of the SERPINE1 gene, no specific information on a functional role is available; however, studies to investigate its functional impact could be useful because this polymorphism was also found associated with MI in nonsmokers.

It is of note that these associations were not modified by adjustment for circulating PAI-1 levels. The reason for the restriction of the relation between insulin and SERPINE1 polymorphisms to cases with MI is not understood. It could be attributable to the effect of another factor, overrepresented in MI and more generally in a stressed inflammatory situation that reinforces the link between SERPINE1 and the MetS, but such a factor has yet to be identified.

In the first report of the HIFMECH study,26 only the –675 4G/5G polymorphism was studied in relation with the risk of MI and was found to interact with insulin to modulate the risk of MI, such that the risk mediated by higher insulin levels was only observed in 4G carriers. The haplotype analysis performed in this report suggested that this interaction was in fact the consequence of the joint effects of the rs6092 and rs7242 on insulin observed in cases only.

SERPINE1 haplotypes explained about 3.5% of the variability of PAI-1 levels, which is similar to that observed in the Framingham Heart Study.34 The present study confirmed results from this latter study by demonstrating that other polymorphisms located outside the SERPINE1 promoter influence PAI-1 levels, but this is in disagreement with results from the study of Ding et al35 which exclusively observed an effect of the –675 4G/5G polymorphism. The analysis here suggested that the observed haplotype effects were attributable to the combined effects of 3 SNPs, –844A>G, rs2227666, and rs2227694. In the Framingham Heart Study, it was impossible to distinguish which of the 2 polymorphisms (A–844G and –675 4G/5G) was responsible for the association with PAI-1 levels. In the present study, the data suggest that the effect of the –675 4G/5G polymorphism observed in univariate analysis was the consequence of its LD with the –844A>G. In view of the results of the Framingham Heart Study,34 2 others polymorphisms, rs6465787 and rs2227692, were also found associated with PAI-1 plasma levels. The rs6465787 SNP was not genotyped in our study because of its minor allele frequency (2%, 34) and its complete LD with the –844A>G, which meant it would be impossible to distinguish its potential effect on PAI-1 levels from that of the –844A>G SNP. The rs2227692 was not genotyped in our study as it is completely tagged by the rs2227667 and rs2227672 SNPs, the rs2227692-T allele being equivalent to the haplotype defined by the rs2227667-G and rs2227672-G alleles (see HapMap database at www. hapmap.org).

It is to note that, because of the LD among the 9 SNPs studied, a standard Bonferroni correction for multiple testing would not be appropriate because it would have been too conservative. Using the method proposed by Li and Ji,46 the number of independent components underlying the LD structure of the set of SNPs was estimated to be 7. Correcting for this number, which corresponds to consider significant any probability value ≤0.007 should be considered as significant, would not have altered the main conclusions of our analyses.

Conclusions
SERPINE1 haplotypes are not associated with the risk of MI in the overall cohort. However, they are mildly associated with plasma levels of PAI-1 and with the risk of MI in nonsmokers. They are also mildly associated with insulin levels and BMI. Six SNPs are associated with these different clinical and biological phenotypes, suggesting that modifications of SERPINE1 expression involved in these processes are regulated via different pathways. Besides, our haplotype analysis suggested that among the 2 promoter polymorphisms rs2227631 (–844A>G) and rs1799889 (–675 4G/5G) which are in tight LD, the former associated with MI, is more likely to be associated with PAI-1 levels.


*    Appendix
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMaterials and Methods
 up arrowResults
 up arrowDiscussion
 *Appendix
down arrowReferences
 
HIFMECH Investigators
Stockholm: A. Hamsten (coordinator), S. Boquist, C.G. Ericsson, P. Lundman, A. Samnegard, A. Silveira, P. Tornvall; London: J.S. Yudkin, V. Mohamed-Ali, A. Holmes; Marseille: I. Juhan-Vague, M.F. Aillaud, P.E. Morange, M.C. Alessi, P. Ambrosi, I. Canavy. F. Paganelli, R. Didelot, J. Ansaldi, M. Billerey; San Giovanni Rotondo: G. Di Minno, M. Margaglione, D. Cimino, N. Dello Iacono, A. Cimino, G. Gaeta, C. Blasich, G. Pucciarelli; London: S.E. Humphries; Leiden: V. van Hinsbergh, T. Kooistra; Milan: E. Tremoli, C. Banfi, L. Mussoni.


*    Acknowledgments
 
Sources of Funding

The HIFMECH Study was also supported by the European Commission (BMH4-CT96–0272), the Swedish Medical Research Council, the Swedish Heart-Lung Foundation, INSERM, and Université de la Méditerranée (INSERM U626), Fondation pour la Recherche Médicale (FRM) and Programme Hospitalier de Recherche Clinique (PHRC 1996). Steve Humphries is supported by the British Heart Foundation (RG2005/014).

Disclosures

None.


*    Footnotes
 
Original received April 25, 2007; final version accepted June 29, 2007.


*    References
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMaterials and Methods
 up arrowResults
 up arrowDiscussion
 up arrowAppendix
 *References
 

  1. Lijnen HR, Collen D. Mechanisms of physiological fibrinolysis. Baillieres Clin Haematol. 1995; 8: 277–290.[Medline] [Order article via Infotrieve]
  2. Dellas C, Loskutoff DJ. Historical analysis of PAI-1 from its discovery to its potential role in cell motility and disease. Thromb Haemost. 2005; 93: 631–640.[Medline] [Order article via Infotrieve]
  3. Lijnen HR. Pleiotropic functions of plasminogen activator inhibitor-1. J Thromb Haemost. 2005; 3: 35–45.[CrossRef][Medline] [Order article via Infotrieve]
  4. Sobel BE. Increased plasminogen activator inhibitor-1 and vasculopathy. A reconcilable paradox. Circulation. 1999; 99: 2496–2498.[Free Full Text]
  5. Kohler HP, Grant PJ. Plasminogen-activator inhibitor type 1 and coronary artery disease. N Engl J Med. 2000; 342: 1792–1801.[Free Full Text]
  6. Sobel BE, Taatjes DJ, Schneider DJ. Intramural plasminogen activator inhibitor type-1 and coronary atherosclerosis. Arterioscler Thromb Vasc Biol. 2003; 23: 1979–1989.[Abstract/Free Full Text]
  7. Hamsten A, de Faire U, Walldius G, Dahlen G, Szamosi A, Landou C, Blomback M, Wiman B. Plasminogen activator inhibitor in plasma: Risk factor for recurrent myocardial infarction. Lancet. 1987; 2: 3–9.[CrossRef][Medline] [Order article via Infotrieve]
  8. Juhan-Vague I, Pyke SDM, Alessi MC, Jespersen J, Haverkate F, Thompson SG. Fibrinolytic factors and the risk of myocardial infarction or sudden death in patients with angina pectoris. Circulation. 1996; 94: 2057–2063.[Abstract/Free Full Text]
  9. Thogersen AM, Jansson JH, Boman K, Nilsson TK, Weinehall L, Huhtasaari F, Hallmans G. High plasminogen activator inhibitor and tissue plasminogen activator levels in plasma precede a first acute myocardial infarction in both men and women: evidence for the fibrinolytic system as an independent primary risk factor. Circulation. 1998; 98: 2241–2247.[Abstract/Free Full Text]
  10. Collet JP, Montalescot G, Vicaut E, Ankri A, Walylo F, Lesty C, Choussat R, Beygui F, Borentain M, Vignolles N, Thomas D. Acute release of plasminogen activator inhibitor-1 in ST-segment elevation myocardial infarction predicts mortality. Circulation. 2003; 108: 391–394.[Abstract/Free Full Text]
  11. Smith A, Patterson C, Yarnell J, Rumley A, Ben-Shlomo Y, Lowe G. Which hemostatic markers add to the predictive value of conventional riskfactors for coronary heart disease and ischemic stroke? The Caerphilly Study. Circulation. 2005; 112: 3080–3087.[Abstract/Free Full Text]
  12. Salomaa V, Riley W, Kark JD, Nardo C, Folsom AR. Non-insulin-dependent diabetes mellitus and fasting glucose and insulin concentrations are associated with arterial stiffness indexes. The ARIC Study. Circulation. 1995; 91: 1432–1443.[Abstract/Free Full Text]
  13. Bavenholm P, de Faire U, Landou C, Efendic S, Nilsson J, Wiman B, Hamsten A. Progression of coronary artery disease in young male post-infarction patients is linked to disturbances of carbohydrate and lipoprotein metabolism and to impaired fibrinolytic function. Eur Heart J. 1998; 19: 402–410.[Abstract/Free Full Text]
  14. Folsom AR, Pankow JS, Williams RR, Evans GW, Province MA, Eckfeldt JH. Fibrinogen, plasminogen activator inhibitor-1, and carotid intima-media wall thickness in the NHLBI Family Heart Study. Thromb Haemost. 1998; 79: 400–404.[Medline] [Order article via Infotrieve]
  15. de Maat MP, Bladbjerg EM, Drivsholm T, Borch-Johnsen K, Moller L, Jespersen J. Inflammation, thrombosis and atherosclerosis: results of the Glostrup study. J Thromb Haemost. 2003; 1: 950–957.[CrossRef][Medline] [Order article via Infotrieve]
  16. Anand SS, Yi Q, Gerstein H, Lonn E, Jacobs R, Vuksan V, Teo K, Davis B, Montague P, Yusuf S. Study of Health Assessment and Risk in Ethnic Groups; Study of Health Assessment and Risk Evaluation in Aboriginal Peoples Investigators. Relationship of metabolic syndrome and fibrinolytic dysfunction to cardiovascular disease. Circulation. 2003; 108: 420–425.[Abstract/Free Full Text]
  17. Festa A, D’Agostino R Jr, Tracy RP, Haffner SM. Insulin Resistance Atherosclerosis Study. Elevated levels of acute-phase proteins and plasminogen activator inhibitor-1 predict the development of type II diabetes: the insulin resistance atherosclerosis study. Diabetes. 2002; 51: 1131–1137.[Abstract/Free Full Text]
  18. Festa A, Williams K, Tracy RP, Wagenknecht LE, Haffner SM. Progression of plasminogen activator inhibitor-1 and fibrinogen levels in relation to incident type II diabetes. Circulation. 2006; 113: 1753–1759.[Abstract/Free Full Text]
  19. Meigs JB, O’donnell CJ, Tofler GH, Benjamin EJ, Fox CS, Lipinska I, Nathan DM, Sullivan LM, D’Agostino RB, Wilson PW. Hemostatic markers of endothelial dysfunction and risk of incident type 2 diabetes: the Framingham Offspring Study. Diabetes. 2006; 55: 530–537.[Abstract/Free Full Text]
  20. Kanaya AM, Wassel Fyr C, Vittinghoff E, Harris TB, Park SW, Goodpaster BH, Tylavsky F, Cummings SR. Adipocytokines and incident diabetes mellitus in older adults: the independent effect of plasminogen activator inhibitor 1. Arch Intern Med. 2006; 166: 350–356.[Abstract/Free Full Text]
  21. Dawson SJ, Wiman B, Hamsten A, Green F, Humphries S, Henney AM. The two allele sequences of a common polymorphism in the promoter of theplasminogen activator inhibitor-1 (PAI-1) gene respond differently to interleukin-1 in HepG2 cells. J Biol Chem. 1993; 268: 10739–10745.[Abstract/Free Full Text]
  22. Henry M, Chomiki N, Scarabin PY, Alessi MC, Peiretti F, Arveiler D, Ferrieres J, Evans A, Amouyel P, Poirier O, Cambien F, Juhan-Vague I. Five frequent polymorphisms of the PAI-1 gene: lack of association between genotypes, PAI activity, and triglyceride levels in a healthy population. Arterioscler Thromb Vasc Biol. 1997; 17: 851–858.[Abstract/Free Full Text]
  23. Lopes C, Dina C, Durand E, Froguel P. PAI-1 polymorphisms modulate phenotypes associated with the metabolic syndrome in obese and diabetic Caucasian population. Diabetologia. 2003; 46: 1284–1290.[CrossRef][Medline] [Order article via Infotrieve]
  24. Eriksson P, Kallin B, van ’t Hooft FM, Bavenholm P, Hamsten A. Allele-specific increase in basal transcription of the plasminogen-activatorinhibitor 1 gene is associated with myocardial infarction. Proc Natl Acad Sci U S A. 1995; 92: 1851–1855.[Abstract/Free Full Text]
  25. Boekholdt SM, Bijsterveld NR, Moons AH, Levi M, Buller HR, Peters RJ. Genetic variation in coagulation and fibrinolytic proteins and their relation with acute myocardial infarction: a systematic review. Circulation. 2001; 104: 3063–3068.[Abstract/Free Full Text]
  26. Juhan-Vague I, Morange PE, Aubert H, Henry M, Aillaud MF, Alessi MC, Samnegard A, Hawe E, Yudkin J, Margaglione M, Di Minno G, Hamsten A, Humphries SE; HIFMECH Study Group. Plasma thrombin-activatable fibrinolysis inhibitor antigen concentration and genotype in relation to myocardial infarction in the north and south of Europe. Arterioscler Thromb Vasc Biol. 2002; 22: 867–873.[Abstract/Free Full Text]
  27. Panahloo A, Mohamed-Ali V, Lane A, Green F, Humphries SE, Yudkin JS. Determinants of plasminogen activator inhibitor 1 activity in treated NIDDM and its relation to a polymorphism in the plasminogen activator inhibitor 1 gene. Diabetes. 1995; 44: 37–42[Abstract]
  28. Mansfield MW, Stickland MH, Grant PJ. Environmental and genetic factors in relation to elevated circulating levels of plasminogen activator inhibitor-1 in Caucasian patients with non-insulin-dependent diabetes mellitus. Thromb Haemost. 1995; 74: 842–847.[Medline] [Order article via Infotrieve]
  29. Hoffstedt J, Andersson IL, Persson L, Isaksson B, Arner P. The common –675 4G/5G polymorphism in the plasminogen activator inhibitor–1gene is strongly associated with obesity. Diabetologia. 2002; 45: 584–587.[CrossRef][Medline] [Order article via Infotrieve]
  30. McCormack LJ, Nagi DK, Stickland MH, Mansfield MW, Mohamed-Ali V, Yudkin JS, Knowler WC, Grant PJ. Promoter (4G/5G) plasminogen activator inhibitor-1 genotype in Pima Indians: relationship to plasminogen activator inhibitor-1 levels and features of the insulin resistance syndrome. Diabetologia. 1996; 39: 1512–1518.[CrossRef][Medline] [Order article via Infotrieve]
  31. Viitanen L, Pihlajamaki J, Halonen P, Lehtonen M, Kareinen A, Lehto S, Laakso M. Association of angiotensin converting enzyme and plasminogen activatorinhibitor-1 promoter gene polymorphisms with features of the insulin resistance syndrome in patients with premature coronary heart disease. Atherosclerosis. 2001; 157: 57–64.[CrossRef][Medline] [Order article via Infotrieve]
  32. Freeman MS, Mansfield MW. The common –675 4G/5G polymorphism in the plasminogen activator inhibitor-1 gene is strongly associated with obesity. Diabetologia. 2002; 45: 1602–1603.[CrossRef][Medline] [Order article via Infotrieve]
  33. Grubric N, Stegnar M, Peternel P, Kaider A, Binder BR. A novel G/A and the 4G/5G polymorphism within the promoter the plasminogen activator inhibitor-1 gene in patients with deep vein thrombosis. Thromb Res. 1996; 84: 431–443.[CrossRef][Medline] [Order article via Infotrieve]
  34. Kathiresan S, Gabriel SB, Yang Q, Lochner AL, Larson MG, Levy D, Tofler GH, Hirschhorn JN, O’Donnell CJ. Comprehensive survey of common genetic variation at the plasminogen activator inhibitor-1 locus and relations to circulating plasminogen activator inhibitor-1 levels. Circulation. 2005; 112: 1728–1735.[Abstract/Free Full Text]
  35. Ding J, Nicklas BJ, Fallin MD, de Rekeneire N, Kritchevsky SB, Pahor M, Rodondi N, Li R, Zmuda JM, Harris TB. Plasminogen activator inhibitor type 1 gene polymorphisms and haplotypes are associated with plasma plasminogen activator inhibitor type 1 levels but not with myocardial infarction or stroke. Am Heart J. 2006; 152: 1109–1115.[CrossRef][Medline] [Order article via Infotrieve]
  36. Su S, Chen S, Zhao J, Huang J, Wang X, Chen R, Gu D. Plasminogen activator inhibitor-1 gene: selection of tagging single nucleotide polymorphisms and association with coronary heart disease. Arterioscler Thromb Vasc Biol. 2006; 26: 948–954.[Abstract/Free Full Text]
  37. Mohamed-Ali V, Gould MM, Gillies S, Goubet S, Yudkin JS, Haines AP. Association of proinsulin-like molecules with lipids and fibrinogen in non-diabetic subjects evidence against a modulating role for insulin. Diabetologia. 1995; 38: 1110–1116.[Medline] [Order article via Infotrieve]
  38. Tregouet D, Garelle V. A new JAVA interface implementation of THESIAS: Testing haplotype effects in association studies. Bioinformatics. In press.
  39. Tregouet DA, Escolano S, Tiret L, Mallet A, Golmard JL. A new algorithm for haplotype-based association analysis: the SEM algorithm. Ann Hum Genet. 2004; 68: 165–177.[CrossRef][Medline] [Order article via Infotrieve]
  40. Nei M. Molecular Evolutionary Genetics. New York, NY: Columbia University Press; 1987.
  41. Tregouet DA, Ricard S, Nicaud V, Arnould I, Soubigou S, Rosier M, Duverger N, Poirier O, Macé S, Kee F, Morrison C, Denèfle P, Tiret L, Evans A, Deleuze JF, Cambien F. In depth haplotype analysis of ABCA1 gene polymorphisms in relation to plasma ApoA1 levels and myocardial infarction. Arterioscler Thromb Vasc Biol. 2004; 24: 775–781.[Abstract/Free Full Text]
  42. Pearce E, Tregouet DA, Samnegard A, Morgan AR, Cox C, Hamsten A, Eriksson P, Ye S. Haplotype effect of the matrix metalloproteinase-1 gene on risk of myocardial infarction. Circ Res. 2005; 97: 1070–1076.[Abstract/Free Full Text]
  43. Paul SR, Donner A. A comparison of tests of homogeneity of odds ration in k 2x2 tables. Stat Med. 1989; 8: 1455–1468.[Medline] [Order article via Infotrieve]
  44. Lang IM, Marsh JJ, Olman MA, Moser KM, Loskutoff DJ, Schleef RR. Expression of type 1 plasminogen activator inhibitor in chronic pulmonary thromboemboli. Circulation. 1994; 89: 2715–2721.[Abstract/Free Full Text]
  45. Alessi MC, Juhan-Vague I. PAI-1 and the metabolic syndrome: links, causes, and consequences. Arterioscler Thromb Vasc Biol. 2006; 26: 2200–2207.[Abstract/Free Full Text]
  46. Li J, Ji L. Adjusting multiple testing in multilocus analysis using the eigenvalues of a correlation matrix. Heredity. 2005; 95: 221–227.[CrossRef][Medline] [Order article via Infotrieve]



This article has been cited by other articles:


Home page
 CirculationHome page
K. Aziz, K. Berger, K. Claycombe, R. Huang, R. Patel, and G. S. Abela
Noninvasive Detection and Localization of Vulnerable Plaque and Arterial Thrombosis With Computed Tomography Angiography/Positron Emission Tomography
Circulation, April 22, 2008; 117(16): 2061 - 2070.
[Abstract] [Full Text] [PDF]

Home page
 Arterioscler. Thromb. Vasc. Bio.Home page
L. Brazionis, K. Rowley, A. Jenkins, C. Itsiopoulos, and K. O'Dea
Plasminogen Activator Inhibitor-1 Activity in Type 2 Diabetes: A Different Relationship With Coronary Heart Disease and Diabetic Retinopathy
Arterioscler. Thromb. Vasc. Biol., April 1, 2008; 28(4): 786 - 791.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Data Supplement
Right arrow All Versions of this Article:
27/10/2250    most recent
ATVBAHA.107.149468v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Morange, P.E.
Right arrow Articles by Juhan-Vague, I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Morange, P.E.
Right arrow Articles by Juhan-Vague, I.

ATVB Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search
Copyright © 2007 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited.