Association Between ADAMTS1 Matrix Metalloproteinase Gene Variation, Coronary Heart Disease, and Benefit of Statin Therapy
Objective— The purpose of this study was to investigate the association between the Ala227Pro polymorphism in the ADAMTS1 metalloproteinase gene and coronary heart disease and benefit from statin therapy in 2 independent cohorts.
Methods and Results— The frequency of the ADAMTS1 227Pro minor allele was 0.24 in 2421 male subjects from CARE, a randomized trial of pravastatin versus placebo. In the placebo arm, homozygotes (6.3% of study population) had a significantly increased risk of fatal coronary disease or nonfatal myocardial infarction (D/MI) compared with noncarriers (OR 2.12, 95% CI 1.07 to 4.19, P=0.03), and in the entire study the benefit of pravastatin in reducing the risk of D/MI was greater in these subjects (OR 0.21, 95% CI 0.06 to 0.69) than in heterozygotes (OR 0.74, 95% CI 0.48 to 1.14) or noncarriers (OR 0.99, 95% CI 0.68 to 1.42; Pinteraction=0.044). Results were tested in 1565 male subjects from WOSCOPS, also a randomized trial of pravastatin versus placebo. Similar to the results in CARE, in the placebo arm subjects homozygous for the minor allele were at increased risk of D/MI (OR 1.72, P=0.052) and in the entire study the benefit of pravastatin in reducing D/MI was greater in these subjects (OR 0.24, 95% CI 0.09 to 0.68) than in heterozygotes (OR 0.73, 95% CI 0.48 to 1.11) or noncarriers (OR 0.65, 95% CI 0.20 to 2.09) (Pinteraction=0.029).
Conclusions— In men not on pravastatin, those homozygous for the 227Pro allele of ADAMTS1 have a nearly 2-fold increased risk of coronary heart disease events compared with noncarriers. In this high-risk group, treatment with pravastatin is highly efficacious, reducing the odds of fatal coronary disease or nonfatal MI by approximately 75%, as compared with 25% in noncarriers or heterozygotes.
Statins effectively lower cholesterol levels and consistently and significantly reduce mortality and coronary events in patients with and without a history of coronary artery disease.1 Yet despite producing profound clinical benefits, statins induce only modest changes in the angiographic severity of atherosclerotic lesions.2 These observations have led to the concept that statins confer clinical benefit in part by reducing inflammation and stabilizing atherosclerotic plaques.3
A critical destabilizing influence on plaques are the matrix metalloproteinases (MMPs) secreted by inflammatory cells.4 MMPs degrade the extracellular matrix and cause a loss of structural integrity that permits further inflammatory cell infiltration and predisposes the fibrous cap to rupture precipitating an acute coronary syndrome. ADAMTS-1 is a protein whose name is an acronym for “a disintegrin-like and metalloproteinase with thrombospondin motifs”.5 ADAMTS-1 has been shown to cleave versican,6 a key proteoglycan that regulates vascular smooth muscle cell (VSMC) migration and contributes to the structural integrity of the fibrous cap in atherosclerotic lesions.7
Recently, the Ala227Pro polymorphism of the ADAMTS1 gene has been reported to be associated with incident heart disease in a single cohort of healthy individuals.8 We investigated the association between the Ala227Pro polymorphism and incident and recurrent coronary events in 2 independent cohorts from the Cholesterol and Recurrent Events (CARE) and West of Scotland Coronary Prevention Study (WOSCOPS) trials,9,10 as well as whether the Ala227Pro polymorphism was associated with the magnitude of clinical benefit of statin therapy in these 2 randomized placebo-controlled trials of pravastatin.
The CARE trial randomized 4159 patients (86% men) residing in the United States or Canada with a prior myocardial infarction (MI) and a total cholesterol level <240 mg/dL to pravastatin 40 mg or placebo for a median of 5 years.9 WOSCOPS randomized 6595 men residing in Scotland without a history of MI and low-density lipoprotein levels >155 mg/dL to pravastatin 40 mg or placebo for a median of 5 years.10 Fasting lipid profiles were compared between genotypes at baseline and between genotypes stratified by treatment arm using the last measured lipid profile (at 5 years in 93% of subjects). The primary end point for both trials was fatal coronary disease or nonfatal MI, although defined slightly differently in each trial. To ensure consistency in disease phenotype in the present analysis, we used an end point of fatal coronary disease (defined as fatal MI or sudden death) or nonfatal MI for both trials. In the CARE trial, a total of 2998 subjects provided DNA. The baseline characteristics, treatment allocation, and rate of the primary end point were similar to the overall cohort. In WOSCOPS, genetic analyses were performed in a nested case-control study of 1565 subjects (350 cases with the end point and 1215 controls, matched for age and smoking status; of these subjects, 820 were treated with placebo and 745 with pravastatin).11 The parent trials and this genetic substudy were approved by the relevant institutional review boards, and informed consent was obtained from all subjects.
As part of a comprehensive genotyping project of candidate genes potentially related to cardiovascular disease and metabolism,12 we analyzed a missense SNP in ADAMTS1, rs428785. This SNP is characterized by a G→C substitution at basepair +1134 of NM_006988 that results in the nonconservative replacement of alanine by proline at amino acid 227. We also genotyped a second SNP in ADAMTS1 that is characterized by a G→C substitution at basepair 409 of NM_006988 (rs402007) in the 5′ untranslated region. However, because the strong linkage disequilibrium between rs428785 and rs402007 (Lewontin’s D′=0.998; P<0.00001) produced virtually identical results for the 2 SNPs, we opted to focus our analysis on the coding SNP.
Genomic DNA was extracted from samples of peripheral whole blood using PURIGENE DNA purification kits (Gentra Systems). The genotype of each sample was determined using kinetic, allele specific polymerase chain reaction (PCR) as previously described.13 Samples were genotyped in random order by individuals blinded with respect to clinical outcomes. The genotyping error rate calculated from genotyping of 180 duplicated DNA samples was below 0.06%.
To achieve comparability to the demographics of the WOSCOPS population (White males) that would be used in the validation phase, analyses in CARE were restricted to the 2432 White males among the cohort of 2998 genotyped subjects, 2421 (99.5%) of whom were successfully genotyped at the ADAMTS1 locus. The χ2 test was used to assess whether the frequencies of the ADAMTS1 genotypes in this cohort were consistent with those expected under Hardy-Weinberg equilibrium. Baseline characteristics of subjects stratified by genotype were compared using analysis of variance (ANOVA) and χ2 tests as appropriate. Given that genotype could modify the effect of statin therapy on the primary end point, it was decided a priori to analyze the risk associated with genotype separately in the placebo and pravastatin treatment arms. The unadjusted incidence of the primary end point in subjects within each treatment arm was compared across genotypes using a Cochran-Armitage test for trend. As follow-up was >99.9% complete and as the validation analyses in WOSCOPS would require conditional logistic regression, for the sake of consistency we used logistic regression in CARE to calculate odds ratios and 95% confidence intervals for the effect of genotype on the incidence of the primary end point and the effect of pravastatin therapy on the incidence of the primary end point, adjusting for age, history of diabetes, history of hypertension, body mass index, systolic blood pressure, diastolic blood pressure, and baseline lipids. Regression models contained all subjects and included terms for the interaction between genotype and pravastatin treatment to determine the effect of genotype in placebo subjects and to assess for modification of the effect of pravastatin treatment on the primary end point by genotype.
Analogous statistical analyses were then performed in a validation population consisting of the WOSCOPS nested case-control study using conditional logistic regression where subjects were stratified on age and smoking status. The covariates used were the same as in CARE, with the exception of age, which was a stratification variable. As we were validating the results from CARE (and would reject results that were nominally significant but opposite in direction from what was observed in CARE), we constructed directional statistical tests with 1-sided probability values for risk and statin benefit.
To estimate the effect sizes using all available data, we created a nested case-control study from the CARE cohort (using 241 cases and 2140 controls matched for age and smoking status) and combined this dataset with the WOSCOPS nested case-control dataset to create a pooled CARE-WOSCOPS dataset. This dataset was analyzed using conditional logistic regression in the same manner as the WOSCOPS data. We corrected for multiplicity using a Benjamini-Hochberg False Discovery Rate (FDR) adjustment,14 based on all the SNPs that were associated with an interaction with statin therapy in CARE that were then tested in WOSCOPS.
The frequency of the minor allele of the ADAMTS1 +1134 G→C polymorphism, which causes a nonconservative change from alanine to proline at amino acid 227 (Ala227Pro), was 0.24 in the CARE White male cohort, and the genotypes showed no departure from Hardy-Weinberg equilibrium (P=0.31). The baseline characteristics of subjects stratified by genotype are shown in Table 1. There were no significant differences except for a modestly different prevalence of hypertension across the three genotypes. Although the proportion of subjects with more than 1 prior MI did not differ statistically across the 3 groups, there was a trend toward a higher proportion in subjects homozygous for the 227Pro allele than in those who were not (21.6% versus 16.0%, P=0.07). There was no association between ADAMTS1 genotype and lipid levels at follow-up in either the placebo or pravastatin arms (P=NS for all lipid parameters; supplemental Figure I, available online at http://atvb.ahajournals.org).
Genotype and Risk of Coronary Events
The incidence of the primary end point of fatal coronary disease or nonfatal MI was significantly associated with ADAMTS1 genotype in the placebo arm, with rates of 9.5%, 12.3%, and 17.4% in subjects who were noncarriers, heterozygotes, and homozygotes for the 227Pro allele, respectively (P=0.026 for trend, Figure 1). After adjusting for age, history of diabetes, history of hypertension, body mass index, systolic blood pressure, diastolic blood pressure, and baseline lipids, the ORs for the risk of the end point in heterozygotes and homozygotes compared with noncarriers were 1.32 (95% CI 0.89 to 1.95, P=0.17) and 2.12 (95% CI 1.07 to 4.19, P=0.03), respectively. There was directional consistency for the individual components of the composite end point (supplemental Table I). In contrast, there was no increased risk of coronary events by genotype across pravastatin treated subjects (Figure 1; Table 2), suggesting an interaction with statin therapy.
Genotype and Efficacy of Pravastatin
In the overall cohort, treatment with pravastatin reduced the risk of fatal coronary disease or nonfatal MI (OR 0.81, 95% CI 0.62 to 1.06). However, there was significant heterogeneity in the benefit of pravastatin therapy in relation to ADAMTS1 genotype. When stratified by genotype, pravastatin significantly reduced the incidence of fatal coronary disease or nonfatal MI in subjects who were homozygotes for the 227Pro allele (OR 0.21, 95% CI 0.06 to 0.69), whereas more modest effects were seen in heterozygotes (OR 0.74, 95% CI 0.48 to 1.14) and noncarriers (0.99, 95% CI 0.68 to 1.42), with a significant test for interaction (P=0.044). Again, there was directional consistency for the individual components of the composite end point (supplemental Table II).
To see whether we could replicate the findings in CARE, we then examined the association of the ADAMTS1 Ala227Pro polymorphism with both clinical outcomes and pravastatin efficacy in a separate population based on a nested case-control study from WOSCOPS. In the control subjects, the minor allele frequency was 0.23 and the genotypes showed no departure from Hardy-Weinberg equilibrium (P=0.53).
Genotype and Risk of Coronary Events
In the placebo arm, compared with noncarriers, heterozygous carriers were not at an increased risk for fatal coronary disease or nonfatal MI (adjusted OR 0.85, 95% CI 0.60 to 1.21), but there was a trend for homozygous carriers of the 227Pro allele to be at greater risk (OR 1.64, 95% CI 0.85 to 3.18). The probability value for the directional test for increased risk in homozygous carriers versus heterozygotes or noncarriers was 0.052 with an OR of 1.72 and a 1-sided 95% CI of >0.99. As was seen in CARE, there was no increased risk of coronary events by genotype across subjects treated with pravastatin (Table 3).
Genotype and Efficacy of Pravastatin
Mirroring the pattern observed in CARE, in WOSCOPS pravastatin significantly reduced the incidence of fatal coronary disease or nonfatal MI in subjects who were homozygotes for the 227Pro allele (OR 0.24, 95% CI 0.09 to 0.68), whereas more modest effects were seen in heterozygotes (OR 0.73, 95% CI 0.48 to 1.11) and noncarriers (0.65, 95% CI 0.47 to 0.89). The interaction probability value for the directional test for greater risk reduction with pravastatin in homozygotes than in heterozygotes or noncarriers was 0.029.
CARE and WOSCOPS Pooled
Given the consistency of the CARE results in WOSCOPS, we pooled CARE and WOSCOPS data to obtain more reliable odds ratio estimates for each genotype.
In the pooled dataset that combined the CARE and WOSCOPS nested case-control studies, compared with noncarriers, subjects heterozygous for the ADAMTS1 227Pro allele were not at increased risk for fatal coronary disease or nonfatal MI (adjusted OR 1.04, 95% CI 0.80 to 1.34), but subjects homozygous for the ADAMTS1 227Pro allele were at a significantly increased risk (OR 1.78, 95% CI 1.11 to 2.87) compared with noncarriers (Figure 2).
Pravastatin Versus Placebo
In the pooled dataset, treatment with pravastatin significantly reduced the odds of fatal coronary disease or nonfatal MI by 77% in subjects homozygous for the ADAMTS1 227Pro allele (OR 0.23, 95% CI 0.10 to 0.49). In contrast, in subjects who were noncarriers or heterozygous carriers of the ADAMTS1 227Pro allele, treatment with pravastatin only reduced the odds of fatal coronary disease or nonfatal MI by 25% (Figure 3). The probability value for the interaction between ADAMTS1 genotype and pravastatin treatment was P=0.008, with a false discovery rate of 17%.
We found that the Ala227Pro polymorphism in ADAMTS1 is significantly associated with both clinical outcomes and the magnitude of benefit with pravastatin therapy. Specifically, males homozygous for the 227Pro allele were at nearly a 2-fold increased risk of fatal coronary disease or nonfatal MI compared with noncarriers. The magnitude of risk for coronary events in those homozygous for the 227Pro allele (OR 1.78) was greater than that of smoking (OR 1.49) and similar to that of diabetes (OR 1.78) in the CARE trial.9 In contrast to what was observed in male subjects treated with placebo, among male subjects treated with pravastatin the 227Pro allele was not associated with an increased risk of fatal coronary disease or nonfatal MI, suggesting an interaction with statin therapy. To that end, we report the novel observation that in male subjects homozygous for the 227Pro allele, treatment with pravastatin was highly efficacious, reducing the odds of fatal coronary disease or nonfatal MI by 77%, as compared with approximately 25% in subjects without this allele or in heterozygotes. Importantly, the findings in CARE were also seen in the independent WOSCOPS trial.
ADAMTS-1 is the founding member of a novel family of multidomain extracellular proteases that contain an amino-terminal metalloproteinase, a disintegrin-like domain, and carboxy-terminal thrombospondin type I motifs. ADAMTS-1 has been shown to cleave versican, a proteoglycan that is present in both early and advanced atherosclerotic lesions, and cleavage of which may permit greater invasion of the intima by VSMCs and undermine the structural integrity of the fibrous cap.6,7,15 ADAMTS-1 also cleaves and alters the extracellular location of tissue-factor pathway inhibitor-2,16 which is an inhibitor of plaque destabilizing MMPs.17 In a murine carotid artery ligation model, overexpression of ADAMTS-1 resulted in increased neointimal formation and favored positive remodeling over negative remodeling, supporting an active role of ADAMTS-1 in atherosclerosis.18 Administration of statins has been shown to decrease matrix metalloproteinase (MMP) content and activity in the heart, carotid arteries, and aorta,19 and decrease circulating MMP levels in humans, independent of changes in lipid levels.20 If statins have a similar effect on ADAMTS-1 activity, it could explain, in part, the observed interaction between genotype and therapy. Thus, one possible model is that the ADAMTS1 227Pro allele leads to increased cleavage of versican, promotion of atherosclerosis, and plaque instability, and hence greater risk of coronary heart disease. In individuals homozygous for the 227Pro allele, statins may be of greater benefit either because of enhanced inhibition of ADAMTS-1 via statin-mediated pathways, or more simply because in such individuals the effect of statins on plaque stabilization is particularly important.
Validation is critical for any genetic association study given the risk of false-positive associations because of multiple testing. However, mitigating against this possibility in our study are the consistent effects we observed in the 2 independent cohorts we evaluated, although the probability value for the association of genotype with cardiovascular disease was only 0.052 in WOSCOPS. Furthermore, our findings in these 2 cohorts are supported by data from a third independent study conducted in healthy individuals from 4 geographically diverse US communities (the ARIC cohort). In that study, the ADAMTS1 227Pro allele was also significantly associated with an increased risk of incident coronary heart disease in Whites.8 Of note, though, our observation of an interaction with the clinical efficacy of statin therapy has not been studied in cohorts other than the 2 we studied. Therefore, in addition to showing a significant interaction in 2 independent cohorts, to account for multiple testing, we also calculated a false discovery rate that was found to be 17%. This result compares favorably to false discovery rates reported in other genetic studies.21 It should also be noted that this test is relatively conservative as it does not take into account the replication of the direction of the association seen in the 2 separate cohorts. Nonetheless, additional replication of our observed interaction between the ADAMTS Ala227Pro polymorphism and the clinical benefit of statin therapy will be necessary.
An unresolved issue is whether heterozygous carriers of the 227Pro allele, who constitute 36% of these populations, are at increased risk of coronary events. An excess risk was seen in these subjects in CARE, but not in WOSCOPS. However, the 95% confidence interval for the risk estimate in heterozygous carriers in WOSCOPS was broad and could not exclude up to a 21% excess risk. Future studies in large cohorts should clarify the risk in heterozygous carriers. Additionally, studies in women and non-Whites will be important to define the effect of the 227Pro allele is those populations.
As with any genetic epidemiology study, associations can be demonstrated, but causality cannot be proven. The ADAMTS1 Ala227Pro polymorphism results in a nonsynonymous, nonconservative replacement of alanine with proline, whose cyclic side chain could distort normal protein structure. Nonetheless, it remains possible that the ADAMTS1 polymorphism we studied is in linkage disequilibrium with the true causative locus. More comprehensive genotyping of the ADAMTS1 gene using tagging SNPs will be important to potentially refine the locus of interest. Furthermore, molecular studies of the expression levels and function of the variant ADAMTS-1 protein and immunohistochemical studies of atheroma in animals engineered to express the 227Pro allele will be important future experiments as the mechanism of action appears to be distinct from modulation of measured lipid levels. As is the nature of genetic epidemiological studies, significant associations are only a first, albeit important step, in gaining a better understanding of the fundamental biology.
In summary, we have shown that a polymorphism in the gene encoding ADAMTS-1, a MMP involved in atherosclerosis, is significantly associated with both long-term cardiovascular events and the magnitude of benefit with pravastatin therapy in men. Our findings add to a growing body of literature documenting the importance of metalloproteinases in human atherosclerosis and cardiovascular disease and support the importance of genetic variability in determining clinical events and response to therapy. Although patients homozygous for the 227Pro allele demonstrated a greater clinical benefit with pravastatin therapy than did heterozygotes or noncarriers, both of the latter groups still enjoyed a significant risk reduction with pravastatin therapy. Thus, determination of ADAMTS1 genotype should not be used clinically to restrict statin therapy. Rather, we believe that additional molecular and clinical studies of ADAMTS-1 and the 227Pro variant could offer important further insights into the biology of acute coronary syndromes, generate novel targets for therapeutic intervention, and ultimately lead to better-tailored therapy for our patients.
Sources of Funding
Dr Sabatine was supported in part by grant R01 HL072879 from the National Heart, Lung, and Blood Institute. Genotyping was supported by Celera; statistical analyses were supported by Bristol-Myers Squibb and Celera. The parent clinical trials, CARE and WOSCOPS, were supported by research grants from Bristol-Myers Squibb.
M.S.S. has received a research grant from Schering-Plough and has received honoraria and served on scientific advisory boards for Bristol-Myers Squibb.
L.P., K.L.S., T.G.K., K.R., Z.T., and K.E.Z. are employees of and/or have ownership interest in Bristol-Myers Squibb.
O.A.I., D.U.L., C.H.T., and J.J.D. are or were employees of and have or had ownership interests in Celera.
C.J.P. has received a research grant from Wyeth; honoraria from AstraZeneca and GlaxoSmithKline; and has served on advisory boards for AstraZeneca, Merck & Co, GlaxoSmithKline, and Organon.
M.A.P. has received research grants from Amgen, Atherogenics, Baxter, Bristol-Myers Squibb, Celladon, Novartis, and sanofi-aventis; is a co-inventor on a patent awarded to Brigham and Women’s Hospital regarding the use of inhibitors of the rennin-angiotensin system in selected survivors of myocardial infarction; and has served on advisory boards for AstraZeneca, Bristol-Myers Squibb, Cytokinetics, Daiichi Sankyo, GlaxoSmithKline, Medtronic, Mitsubishi, Novartis, Pharmacopeia, and Via Pharmaceuticals.
J.S. has received honoraria from AstraZeneca, Pfizer, and Merck & Co., and served on advisory boards for Nicox, Merck & Co, and Pfizer.
H.C. reports no potential conflicts of interest.
F.M.S. has received honoraria from AstraZeneca; served on advisory boards for Aegerion and Genzyme; and served as an expert witness for Pfizer.
E.B. has received research grants from Bristol Myers Squibb, Merck & Co., and Schering-Plough, and has served as a consultant for Merck & Co. and Schering-Plough.
The authors thank Dr. Nicholas C. Dracopoli for his helpful comments.
Original received July 22, 2007; final version accepted December 11, 2007.
Baigent C, Keech A, Kearney PM, Blackwell L, Buck G, Pollicino C, Kirby A, Sourjina T, Peto R, Collins R, Simes R. Efficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90,056 participants in 14 randomised trials of statins. Lancet. 2005; 366: 1267–1278.
Nissen SE, Tuzcu EM, Schoenhagen P, Brown BG, Ganz P, Vogel RA, Crowe T, Howard G, Cooper CJ, Brodie B, Grines CL, DeMaria AN. Effect of intensive compared with moderate lipid-lowering therapy on progression of coronary atherosclerosis: a randomized controlled trial. JAMA. 2004; 291: 1071–1080.
Kuno K, Kanada N, Nakashima E, Fujiki F, Ichimura F, Matsushima K. Molecular cloning of a gene encoding a new type of metalloproteinase-disintegrin family protein with thrombospondin motifs as an inflammation associated gene. J Biol Chem. 1997; 272: 556–562.
Sandy JD, Westling J, Kenagy RD, Iruela-Arispe ML, Verscharen C, Rodriguez-Mazaneque JC, Zimmermann DR, Lemire JM, Fischer JW, Wight TN, Clowes AW. Versican V1 proteolysis in human aorta in vivo occurs at the Glu441-Ala442 bond, a site that is cleaved by recombinant ADAMTS-1 and ADAMTS-4. J Biol Chem. 2001; 276: 13372–13378.
Wight TN, Merrilees MJ. Proteoglycans in atherosclerosis and restenosis: key roles for versican. Circ Res. 2004; 94: 1158–1167.
Morrison AC, Bare LA, Chambless LE, Ellis SG, Malloy M, Kane JP, Pankow JS, Devlin JJ, Willerson JT, Boerwinkle E. Prediction of coronary heart disease risk using a genetic risk score: the Atherosclerosis Risk in Communities Study. Am J Epidemiol. 2007; 166: 28–35.
Sacks FM, Pfeffer MA, Moye LA, Rouleau JL, Rutherford JD, Cole TG, Brown L, Warnica JW, Arnold JMO, Wun C-C, Davis BR, Braunwald E, for the Cholesterol and Recurrent Events Trial Investigators. The effect of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels. N Engl J Med. 1996; 335: 1001–1009.
Packard CJ, O’Reilly DS, Caslake MJ, McMahon AD, Ford I, Cooney J, Macphee CH, Suckling KE, Krishna M, Wilkinson FE, Rumley A, Lowe GD. Lipoprotein-associated phospholipase A2 as an independent predictor of coronary heart disease. West of Scotland Coronary Prevention Study Group. N Engl J Med. 2000; 343: 1148–1155.
Iakoubova OA, Sacks FM, Chokkalingam AP, Sabatine MS, Ray KK, Campos H, Kirchgessner TG, Ranade K, Tong CH, Tsuchihashi Z, Packard CP, Pfeffer MA, Ploughman LM, Zerba KE, Shepherd J, Shaw PM, Devlin JJ, Dracopoli NC Association of polymorphisms in the IgA Fc receptor and ABCA1 genes with myocardial infarction and other coronary events in the CARE and WOSCOPS trials. Circulation. 2004; 110 (Suppl III): III–427.
Germer S, Holland MJ, Higuchi R. High-throughput SNP allele-frequency determination in pooled DNA samples by kinetic PCR. Genome Res. 2000; 10: 258–266.
Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Statist Soc B. 1995; 57: 289–300.
Torres-Collado AX, Kisiel W, Iruela-Arispe ML, Rodriguez-Manzaneque JC. ADAMTS1 interacts with, cleaves, and modifies the extracellular location of the matrix inhibitor tissue factor pathway inhibitor-2. J Biol Chem. 2006; 281: 17827–17837.
Jonsson-Rylander AC, Nilsson T, Fritsche-Danielson R, Hammarstrom A, Behrendt M, Andersson JO, Lindgren K, Andersson AK, Wallbrandt P, Rosengren B, Brodin P, Thelin A, Westin A, Hurt-Camejo E, Lee-Sogaard CH. Role of ADAMTS-1 in atherosclerosis: remodeling of carotid artery, immunohistochemistry, and proteolysis of versican. Arterioscler Thromb Vasc Biol. 2005; 25: 180–185.
Nagashima H, Aoka Y, Sakomura Y, Sakuta A, Aomi S, Ishizuka N, Hagiwara N, Kawana M, Kasanuki H. A 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor, cerivastatin, suppresses production of matrix metalloproteinase-9 in human abdominal aortic aneurysm wall. J Vasc Surg. 2002; 36: 158–163.
Koh KK, Son JW, Ahn JY, Jin DK, Kim HS, Choi YM, Kim DS, Jeong EM, Park GS, Choi IS, Shin EK. Comparative effects of diet and statin on NO bioactivity and matrix metalloproteinases in hypercholesterolemic patients with coronary artery disease. Arterioscler Thromb Vasc Biol. 2002; 22: e19–23.
Samani NJ, Erdmann J, Hall AS, Hengstenberg C, Mangino M, Mayer B, Dixon RJ, Meitinger T, Braund P, Wichmann HE, Barrett JH, Konig IR, Stevens SE, Szymczak S, Tregouet DA, Iles MM, Pahlke F, Pollard H, Lieb W, Cambien F, Fischer M, Ouwehand W, Blankenberg S, Balmforth AJ, Baessler A, Ball SG, Strom TM, Braenne I, Gieger C, Deloukas P, Tobin MD, Ziegler A, Thompson JR, Schunkert H. Genomewide association analysis of coronary artery disease. N Engl J Med. 2007; 357: 443–453.