This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Peroxisome Proliferator-Activated Receptor Pathway GenePolymorphism Associated With Extent of Coronary Artery
Disease in Patients With Type 2 Diabetes in the BypassAngioplasty Revascularization Investigation 2 Diabetes Trial
Sharon Cresci, MD; Jun Wu, MD, MS; Michael A. Province, PhD; John A. Spertus, MD, MPH;Michael Steffes, MD, PhD; Janet B. McGill, MD; Edwin L. Alderman, MD; Maria Mori Brooks, PhD;
Sheryl F. Kelsey, PhD; Robert L. Frye, MD; Richard G. Bach, MD; on behalf of the BARI 2D Study Group
Background—Coronary artery disease (CAD) is the major cause of death in patients with type 2 diabetes mellitus. Althoughdemographic and clinical factors associated with extent of CAD in patients with diabetes mellitus have been described, geneticfactors have not. We hypothesized that genetic variation in peroxisome proliferator-activated receptor (PPAR) pathway genes,important in diabetes mellitus and atherosclerosis, would be associated with extent of CAD in patients with diabetes mellitus.
Methods and Results—We genotyped 1043 patients (702 white, 175 blacks) from the Bypass Angioplasty Revascular-ization Investigation 2 Diabetes (BARI 2D) genetic cohort for 3351 variants in 223 PPAR pathway genes using a customtargeted-genotyping array. Angiographic end points were determined by a core laboratory. In whites, a single variant(rs1503298) in TLL1 was significantly (P�5.5�10�6) associated with extent of CAD, defined as number of lesions withpercent diameter stenosis �20%, after stringent Bonferroni correction for all 3351 single nucleotide polymorphisms.This association was validated in the diabetic subgroups of 2 independent cohorts, the Translational ResearchInvestigating Underlying Disparities in Acute Myocardial Infarction Patients’ Health Status (TRIUMPH) post–myocardial infarction registry and the prospective Family Heart Study (FHS) of individuals at risk for CAD.TLL1rs1503298 was also significantly associated with extent of severe CAD (�70% diameter stenosis; P�3.7�10�2)and myocardial jeopardy index (P�8.7�10�4). In general linear regression modeling, TLL1rs1503298 explained morevariance of extent of CAD than the previously determined clinical factors.
Conclusions—We identified a variant in a single PPAR pathway gene, TLL1, that is associated with the extent of CADindependently of clinical predictors, specifically in patients with type 2 diabetes mellitus and CAD.
Diabetes mellitus (DM) currently affects �17 millionpeople in the United States, and �1.6 million new cases
are diagnosed each year.1 Among persons with DM, coronaryatherosclerosis is highly prevalent and accounts for themajority of deaths.1 In patients with coronary artery disease(CAD), patients with DM have lower 10-year survival thanthose without DM,2 and the extent of CAD predicts mortal-ity.3 Atherosclerosis in patients with DM has an acceleratedphenotype, with more diffuse and extensive disease that
shows more rapid progression, suggesting a distinctive patho-genesis.4 The distinct pathogenesis of the accelerated athero-sclerosis observed among patients with DM is poorly under-stood,5 and the role of genetic factors is unknown.
Clinical Perspective on p 1434Bypass Angioplasty Revascularization Investigation 2 Di-
abetes (BARI 2D) was a multicenter, randomized clinical trialthat investigated the effect of different approaches to the
Received July 2, 2010; accepted July 26, 2011.From the Washington University School of Medicine, Department of Medicine, St. Louis, MO (S.C., J.W., M.A.P., J.B.M., R.G.B.); Saint Luke’s Mid
America Heart Institute and the University of Missouri–Kansas City (J.A.S.); University of Minnesota, Minneapolis (M.S.); Stanford University, Stanford,CA (E.L.A.); University of Pittsburgh, Pittsburgh, PA (M.M.B., S.F.K.); and Mayo Clinic, Rochester, MN (R.L.F.).
A complete list of the BARI 2D Study Group is provided in the Appendix in the online-only Data Supplement.Guest Editor for this article was Kari E. North, PhD.The online-only Data Supplement is available with this article at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIRCULATIONAHA.
111.029173/-/DC1.Correspondence to Sharon Cresci, MD, Cardiovascular Division, Washington University School of Medicine, 660 S Euclid Ave, Campus Box 8086,
treatment of both CAD and DM on long-term outcomes inpatients with type 2 DM.6,7 All patients in BARI 2D had DMand CAD suitable for, but not requiring, revascularization.Baseline extent of CAD in BARI 2D was defined angio-graphically and quantified by a core laboratory. Clinicalfactors contributing to extent of CAD among patients inBARI 2D, defined as the number of coronary lesions �20%diameter stenosis (DS), were recently reported;8 however, thetotal variance in extent of CAD explained by baseline clinicalfactors was �10%,8 providing an important rationale forexamining the putative role of genetic factors.
Peroxisome-proliferator activated receptor (PPAR) path-way genes are involved in cellular processes relevant to bothCAD and DM. Master regulators of lipid and glucose homeo-stasis, cardiac energy metabolism, vascular inflammation,and cell differentiation, PPARs have been implicated in thedevelopment and progression of both type 2 DM and athero-sclerosis in animal studies.9–12 We therefore hypothesizedthat investigating genetic variation within the PPAR genepathway would identify novel genes involved in diabeticatherosclerosis.
Using a custom PPAR pathway gene single nucleotidepolymorphism (SNP) microarray containing 3351 SNPs in223 PPAR pathway genes, we investigated genetic associa-tions of PPAR pathway genes with extent of CAD amongpatients with DM in the BARI 2D trial. We then sought tovalidate the discovered associations in 2 additional well-phe-notyped cohorts of patients with CAD and DM.
MethodsBypass Angioplasty RevascularizationInvestigation 2 Diabetes (BARI 2D) Cohort andBaseline Phenotypic DataEligibility criteria for BARI 2D have previously been described.6,7
Patients with type 2 DM having angiographically documented CADinvolving at least 1 major epicardial coronary artery (�50% DS witha positive stress test or �70% DS with classic angina) suitable fortreatment with either medical therapy or coronary revascularizationwere eligible for inclusion.6,7 Exclusion criteria included significantleft main disease (�50% DS), class III or IV congestive heart failure,and coronary revascularization within the preceding 12 months. Allpatients underwent coronary angiography, and the angiograms werereviewed by a central core laboratory (Stanford University) with astandard protocol used in previous trials and masked to patientrandomization or intended revascularization strategy.7,8 This assess-ment included quantifying the number and percent DS of each lesion,its location in the coronary tree using the BARI modification of theCoronary Artery Surgery Study map, and the lesion morphologyaccording to the American College of Cardiology/American HeartAssociation classification scheme.8 Percent DS was the ratio of theminimum diameter divided by the reference diameter (the mean ofthe proximal and distal diameter of the normal-appearing arteryadjacent to the lesion). All lesions �20% DS were counted. Inaddition to number of lesions, total myocardial jeopardy index wascalculated. The myocardial jeopardy index was calculated as thepercentage of the myocardium that was jeopardized by lesions�50% DS in any of the 3 main coronary arteries or theirbranches.8,13 In accordance with the published analysis of clinicalpredictors of extent of CAD in BARI 2D,8 number of coronarylesions �20% DS was selected to represent the overall extent ofatherosclerotic CAD; additional angiographic variables were in-cluded to represent the extent of severe atherosclerotic CAD (numberof lesions �70% DS and myocardial jeopardy index).
Bypass Angioplasty RevascularizationInvestigation 2 Diabetes (BARI 2D)Genetic CohortA total of 2368 patients were enrolled in BARI 2D from January2001 through March 2005. Of these, 1439 BARI 2D patientsconsented to genetic analysis; 1072 genomic samples (702 whites,175 blacks) were available for genotyping (353 samples from Brazilwere not available because of international restrictions on DNAshipment). To determine whether the genetic cohort was represen-tative of the entire BARI 2D cohort, clinical covariates and associ-ation with number of lesions �20% DS were determined for theentire genetic cohort (not stratified by race). Using the same clinicalcovariates and the same analytic strategy (variable entry and reten-tion criteria; entry P�0.15, retain P�0.05; age, and sex includedregardless of statistical significance) used in the entire cohort,8 wedetermined the magnitude of individual variables associated withextent of CAD (presented as semipartial correlation coefficients).
Validation Cohorts
Translational Research Investigating UnderlyingDisparities in Acute Myocardial Infarction Patients’Health Status StudyTranslational Research Investigating Underlying Disparities inAcute Myocardial Infarction Patients’ Health Status (TRIUMPH)was a large, prospective, observational cohort study of consecutivepatients with acute myocardial infarction presenting to 24 UShospitals from April 2005 to December 2008.14 Myocardial infarc-tion was diagnosed with contemporary definitions,15 and all patientshad an elevated troponin blood test. Thirty-one percent of TRIUMPHpatients had DM on enrollment. Extent of CAD was determined byangiography and defined as the number of vessels with �20% DS.We genotyped 386 whites using the PPAR pathway custom arrayconstituting the genetic cohort for the present analysis. The studywas approved by the Institutional Review boards of all participatinginstitutions, and written informed consent was provided by eachparticipant. A separate consent form for the acquisition of blood forgenetic analysis was signed by each participant.
Family Heart StudyThe Family Heart Study (FHS) has been described in detail.16,17
Briefly, adults (45 to 69 years of age) were recruited from 1 of 3studies: the Framingham study, the Utah study, and 2 AtherosclerosisRisk in Communities sites. Subjects and their family members wereeither randomly selected or selected because they had a high relativerisk of coronary heart disease.17 Of these, 14% had DM. The cardiaccomputed tomography subset of FHS (the FHS Subclinical Athero-sclerosis Network) recruited patients and their family memberspreviously enrolled in FHS who had been genotyped by genome-wide association study to undergo cardiac computed tomography todetermine the extent of coronary artery and aortic calcification.16,17
Coronary artery calcification was determined by a core laboratory(Wake Forest University Health Sciences, Winston Salem, NC) aspreviously described.16,17 In brief, 2 sequential scans were performed1 minute apart; the amount of calcium was quantified by theAgatston scoring method adjusted for slice thickness; and the totalcoronary artery calcification score was calculated as the sum of theAgatston scores from each scan and then averaged. We genotyped974 whites with the Illumina 550K SNP array.
Genotyping and Custom ArrayDNA was isolated and extracted with the Puregene genomic DNApurification kit (Gentra, Minneapolis, MN). DNA (4 �g) wasgenotyped using a custom-designed PPAR pathway–targeted geno-typing 3K array containing 3351 SNPs in 223 genes within the PPARpathway and 100 SNPs, allowing identification of ethnic origin bymapping by admixture linkage disequilibrium (Affymetrix, SantaClara, CA; Table I in the online-only Data Supplement).18
Cresci et al Polymorphism and Coronary Disease in BARI 2D 1427
Method of Selection of Genes on Custom ArrayThe PPAR pathway genes were selected if they met 1 of 2 criteria:were identified as a PPAR pathway gene in the published literatureor showed �2-fold upregulation or downregulation in in vitro geneexpression profiling experiments performed with mouse heartsoverexpressing PPAR� compared with nontransgenic littermates(data graciously provided by Dr Daniel P. Kelly, Burnham Institutefor Medical Research, Orlando, FL).
Method of Selection of Single NucleotidePolymorphisms on Custom ArrayThe SNPs were chosen from HapMap data. The projected popula-tions selected for tagging were Africans and whites, with a biastoward white coverage. A tagging SNP approach (from 5 kbupstream of the gene to 5 kb downstream of the gene; minor allelefrequency of �10% in white and black populations; the method ofCarlson et al19 was used to select tagging SNPs; r2 �0.8 required)was used. If a tagging SNP was in a large bin (containing �5 SNPswith r2 �0.8), we included 2 tagging SNPs on the array to ensuremaximal data acquisition (although genotyping only 1 tagging SNPis highly efficient, if it fails to genotype, a large amount of data couldpotentially be lost). In addition to the tagging SNPs, the array alsoincludes all nonsynonymous coding SNPs in each gene (regardless ofminor allele frequency). The SNP coverage provided by the finalpanel was 8500/�10 000 whites and 7000/�11 000 Yorubans.
Genotyping Quality Control DataData completeness was 99.64% and repeatability was 100%. Toensure high-quality data and to minimize false-positive results,filtering criteria were set as follows: genotype call rate �0.80 perSNP, minor allele frequency �0.01, and P value for Hardy-Weinberg equilibrium test �0.0001. Using the PLINK analysisprogram (version 1.06), we identified 5 white subjects whosereported sex did not match the sex estimated from the X chromosomedata. These 5 samples were excluded from further analysis. Clusteranalysis implemented in PLINK was used for screening any evidenceof subtle population stratifications based on pairwise identity-by-state sharing distance among all possible pairs of the SNPs. Acrossthe samples of the BARI 2D genetic cohort, 4 well-separatedsubgroups correlated with self-reported ethnic groups, and the 2major groups of blacks and whites were clearly distinguished. Inblacks, no further population stratification was detected according tomultidimensional scaling plots (Figure I in the online-only DataSupplement). In whites, subtle clustering was noted; we thereforeestimated 10 principal components using Eigenstrat.
Statistical Methods
Clinical CovariatesThe primary end point for our analysis was number of lesions �20%DS documented on the coronary angiogram. Myocardial jeopardyindex and number of lesions �70% DS were also determined forvariant(s) achieving significance in replication cohorts. Number oflesions and myocardial jeopardy index were evaluated as continuousvariables, with kernel density estimators used to plot their overalldistributions. Continuous variables were compared by Student t testsor Wilcoxon tests (depending on distributional properties); categor-ical variables were compared by �2 tests. Stepwise linear regressionwas used to identify baseline factors associated with number oflesions (log transformed) and myocardial jeopardy index.
Single Nucleotide Polymorphism AssociationTo prevent confounding from racial admixture, we performed ouranalyses separately in our 2 largest racial groups (whites, n�702;blacks, n�175). Single SNP associations for the number of lesions�20% DS were tested in general linear regression analyses using anadditive genetic model. Age, sex, body mass index, and priorrevascularization were included in the model as covariates. Signifi-cance level was 1.5�10�5 using stringent Bonferroni correction forall 3351 SNPs.
ValidationWe validated putative statistically significant associations in the 2separate cohorts of patients systematically assessed for extent ofCAD described above, TRIUMPH and FHS. Because the significantassociation was observed in white BARI 2D subjects, white TRIUMPHand FHS subjects were used for validation. None of the subjects wererelated in TRIUMPH. Because FHS had related subjects, we usedmixed models to take family structure into account. Values ofP�0.05 were considered statistically significant for validation anal-yses. Statistical analyses were performed with SAS 9.1 (SASInstitute Inc, Cary, NC), and plots were generated by R 2.3.1(CRAN, the comprehensive R archive network).
ResultsThe Bypass Angioplasty RevascularizationInvestigation 2 Diabetes (BARI 2D)Genetics CohortThe baseline characteristics of the patients making up thegenetic cohort of BARI 2D were similar to those of the entiregroup (Table). The primary end point representing the extentof CAD (number of coronary lesions �20% DS) was thesame in the BARI 2D genetic cohort (mean number ofcoronary lesions �20% DS � 4.6�2.2) and in the entireBARI 2D cohort (mean number of coronary lesions �20%DS � 4.6�2.3). To further confirm that the extent of CAD inthe genetics subgroup was representative of the entire cohort,the magnitude of association between clinical covariates andnumber of coronary lesions �20% DS was compared be-tween the 2 groups. As shown in Figure 1, individual clinicalcovariates in the BARI 2D genetic cohort showed correla-tions with the extent of CAD similar to those of the entireBARI 2D cohort. Similar to the entire BARI 2D cohort,female sex, black race, higher body mass index, and higher30-second pulse were associated with less extensive CAD,whereas older age, longer duration of DM, hypertension, andhigher tissue-type plasminogen activator antigen levels wereassociated with more extensive CAD.
Peroxisome Proliferator-Activated ReceptorPathway Single Nucleotide PolymorphismAssociations With Extent of CoronaryArtery DiseaseTo prevent confounding from racial admixture, we performedour analyses separately in our 2 largest racial groups (whites,n�702; blacks, n�175). In white BARI 2D subjects, singleSNP analysis identified 1 SNP (rs1503298) that was highlysignificantly associated (P�5.5�10�6) with number of cor-onary lesions �20% DS (Figure 2), even after stringentBonferroni correction for all 3351 SNPs. This SNP is locatedin intron 12 of the gene that encodes Tolloid-like 1 (TLL1), ametalloproteinase that regulates bone morphogeneticprotein-2 (BMP-2) and transforming growth factor-�. Thisassociation remained significant after inclusion of all baselineclinical covariates and principal components in the model.The mean�SE residual of number of lesions �20% by TLL1rs1503298 genotype after adjustment for age, sex, and bodymass index in white BARI 2D subjects is shown in Figure 3A.To put these data into a clinically relevant context, an averageBARI 2D subject (a 63-year-old man with a body mass indexof 30 kg/m2) with the TT genotype would have 4.43 coronarylesions �20% DS; with the CT genotype, 5.02 coronary lesions
Table. Baseline Characteristics of the Entire Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D) Cohort, theBARI 2D Genetic Cohort, and the White Subgroups of BARI 2D, Translational Research Investigating Underlying Disparities in AcuteMyocardial Infarction Patients’ Health Status (TRIUMPH), and Family Heart Study (FHS)
All Patients(n�2368)
Genetics Cohort(n�1072)
Genetics CohortWhites (n�702)
FHS DM Whites(n�139)
TRIUMPH DMWhites (n�100)
Age, mean (SD), y 62.4 (8.9) 63.1 (9.2) 64.3 (8.8) 63.6 (11.3) 58.8 (11.5)
Male, % 70.4 69.4 75.2 42.4 67.0
Race, %
White 70.4 66.6 100.0 100.0 100.0
Black 17 17.1 0 0 0
Asian 4.2 5.5 0 0 0
Indian/Native American 4.3 0.1 0 0 0
Other 4.1 5.9 0 0 0
Hispanic ethnicity 12.5 14.0 5.1
Region of world, %
United States 63.3 76.1 75.5 100.0 100.0
Canada 14.9 15.3 17.2 0 0
Mexico 3.6 4.0 0 0 0
Czechoslovakia 3.2 4.6 7.3 0 0
BMI categories, %
Normal or underweight, �25 kg/m2 9.7 7.9 5.1 23.0 11.1
Overweight, 25–�30 kg/m2 34 30.9 28.8 41.7 26.3
Class 1 obese, 30–�35 kg/m2 32.1 31.8 35.3 24.5 35.4
Class 2 obese, 35–�40 kg/m2 15.3 17.1 19.4 7.9 11.1
Class 3 obese, �40 kg/m2 9 12.2 11.6 2.9 16.2
Blood pressure �130/80 mm Hg,% 52.4 48.8 45.1 47.5 79.0
History of hypertension, % 82.5 81.9 81.6 53.2 77.0
History of myocardial infarction, % 32 29.5 30.0 12.6 24.0
History of congestive heart failure, % 6.6 8.4 8.9 0.7 13.0
Anginal status, %
None 17.9 21.1 22.7 87.8 27.0
Anginal equivalents only 21.4 25.2 26.4
Stable CCS1 14.3 14.4 14.7
Stable CCS2 28.8 24.5 23.5
Stable CCS3 7.5 5.2 5.3
Stable CCS4 1.2 0.7 0.7
Unstable angina 9.5 8.9 6.8
Duration of DM, %
�5 y 33.3 32.1 32.1
5–�10 y 23.5 23.1 22.9
10–�20 y 29.2 30.4 30.1
�20 y 14.1 14.3 14.6
History of insulin use, % 29.3 29.0 27.4
Glycemia measurements, %
HbA1c �7.0% 41.7 45.5 48.0 47.4
7.0%�HbA1c�8.0% 25.3 27.9 23.6 12.6
HbA1c �8.0% 33 26.6 28.5 40.0
Total cholesterol �200 mg/dL, % 19.0 18.5 16.6 29.6 34.0
�20% DS; and with the CC genotype, 5.46 coronary lesions�20% DS (Figure 3B). Of note, in general linear regressionmodeling, this SNP explained more variance of the phenotype(number of coronary lesions �20% DS) than the previouslydetermined clinical factors. TLL1 rs1503298 explained 2.75%of the variance compared with sex (1.61%), age (0.12%), andbody mass index (0.01%). Subjects with the TLL1 rs1503298CT genotype had 22% more coronary lesions �20% DScompared with those with the TT genotype, and those withthe CC genotype had 37% more coronary lesions �20% DS.
In the smaller subgroup of black BARI 2D subjects(n�175), no SNP was found to be significantly associatedwith number of coronary lesions �20% DS after stringentBonferroni correction for all 3351 SNPs. This included TLL1rs1503298 (P�0.25).
ValidationFor validation, the genetic association observed in BARI 2Dwas tested in both the overall group and the diabetic sub-groups of 2 independent cohorts of patients with CAD,TRIUMPH and FHS. These validation cohorts were chosenbecause both TRIUMPH and FHS were independent popula-tions with detailed information on extent of CAD by quanti-
tative imaging studies, both had detailed information onsubjects with and without carefully defined DM, both hadextensive phenotypic data to allow consistent adjustment forcovariates, and both were genotyped for relevant poly-morphisms. Validation analyses were performed on (white)subjects with and without DM within these same cohorts toobtain insight into whether the association was specific toindividuals with DM.
When the same model was applied in TRIUMPH with theidentical SNPs used in BARI 2D, a significant associationwith extent of CAD was selectively demonstrated for TLL1rs1503298 among patients with DM (n�98; P�3.3�10�2)but not in the overall cohort (n�386; P�1.8�10�1). Acongruent trend was found in FHS with the use of a lesssensitive measure of CAD extent. The association betweenrs1503298 and extent of CAD defined by coronary calciumscore was demonstrated selectively in FHS individuals withDM (n�139; P�8.5�10�2) but did not replicate in the largerFHS cohort (n�974; P�5.7�10�1).
Associations With Severe Coronary Artery DiseaseWe also investigated whether the TLL1rs1503298 variant wasassociated with the extent of severe CAD and myocardial
Figure 1. Comparison of clinical covariates associated with number of lesions �20% in the Bypass Angioplasty RevascularizationInvestigation 2 Diabetes (BARI 2D) genetic cohort (right) compared with the entire BARI 2D cohort (left). DM indicates diabetes mellitus;ABI, ankle-brachial index; and t-PA, tissue-type plasminogen activator.
Figure 2. Results of tests for a trend inthe association between number oflesions �20% and each single nucleo-tide polymorphism on a custom chip inwhites in Bypass Angioplasty Revascu-larization Investigation 2 Diabetes (BARI2D). Red line depicts the threshold forstatistical significance after Bonferronicorrection for multiple comparisons.
jeopardy index in BARI 2D, variables previously associatedwith mortality in other studies.3,13 TLL1 rs1503298 wassignificantly associated with both number of coronary lesions�70% DS (P�3.7�10�2) and myocardial jeopardy index(P�8.7�10�4; Figure 4).
DiscussionUsing a gene pathway microarray approach, we have identi-fied a novel PPAR pathway genetic polymorphism that issignificantly associated with extent of CAD in patients withDM. The TLL1 rs1503298 variant, located in intron 12 ofTLL1, showed associations with rigorous and objectivelyderived measures of atherosclerotic coronary disease in BARI2D and explained more variance of the phenotype than thepreviously determined clinical factors. The association ofextent of CAD observed with TLL1 rs1503298 in BARI 2Dwas validated by replication in the diabetic patients in 2independent populations, TRIUMPH and FHS.
The association of TLL1 rs1503298 with extent of CADwas observed among white patients with type 2 DM in theBARI 2D trial, in whom clinically stable CAD was definedand quantified by core laboratory interpretation of coronaryangiograms. This association was validated by replication inthe white diabetic subgroups of 2 independent populationswith CAD: the TRIUMPH cohort, in whom CAD in patientswith acute myocardial infarction was defined angiographi-cally, and the FHS cohort, in whom individuals at high risk ofCAD had their extent of coronary artery calcification deter-
mined by computed tomography scanning. The observationthat this variant was reproducibly associated with extent ofdisease quantified by disparate methods among individualswith disparate clinical manifestations supports a potential rolefor this genetic polymorphism in the pathogenesis of diabeticatherosclerosis. Further reinforcing the clinical relevance ofthe association between this SNP and CAD in patients withDM, the TLL1 rs1503298 polymorphism showed strongassociations not only with the primary end point of number oflesions �20% DS but also with number of angiographicallysevere (�70% DS) lesions and with myocardial jeopardyindex, a semiquantitative method used to estimate the amountof potential myocardial ischemia attributable to the locationand severity of coronary lesions in an individual; thesemeasures of extent of severe CAD have previously beenshown to correlate with prognosis in patients with CAD.3,13
To the best of our knowledge, this is the first demonstrationof a significant association of genetic variation specificallywith extent of the atherosclerotic disease that develops amongpatients with DM, in which atherosclerosis has been known todisplay a distinctively aggressive phenotype.20–22
The observation of an association between extent of CADin patients with DM and this particular gene in the PPARpathway–regulated processes is noteworthy. TLL1 encodes aprotein, Tolloid-like 1 (TLL1), that has been identified withinthe cascade of cellular processes related to vascular inflam-mation and calcification. Because vascular calcification is aprominent feature of the phenotype of diabetic atherosclerosis,23
genetic variability in TLL1 may have strong biologicalplausibility as a contributor to its pathogenesis.23 TLLI is ametalloproteinase that cleaves the BMP-2 antagonists Chor-
Figure 3. A, Mean (�SE) residual of number of lesions �20%by TLL1 rs1503298 genotype after adjustment for age, sex, andbody mass index (BMI) in white Bypass Angioplasty Revascular-ization Investigation 2 Diabetes (BARI 2D) subjects. B, Numberof lesions �20% by genotype for the average BARI 2D subject(a 63-year-old man with a BMI of 30 kg/m2).
Figure 4. Mean (�SE) residual of number of lesions �70% (A)and myocardial jeopardy index (MJI; B) by TLL1 rs1503298 ge-notype after adjustment for age, sex, and body mass index inwhite Bypass Angioplasty Revascularization Investigation 2 Dia-betes (BARI 2D) subjects.
Cresci et al Polymorphism and Coronary Disease in BARI 2D 1431
din and gremlin, thereby upregulating BMP-2/4 pathwaygenes that promote inflammation and extracellular matrixneogenesis.24,25 BMP-2 itself is an important mediator oftissue calcification, and endothelial BMP-2 expression can beupregulated by vascular oxidative stress, intravascular pres-sure, and elevated glucose.26–29 BMP-2 appears to play acritical role in programming the phenotype of the calcifyingvascular cell, the mural cell type that directs osteogenicmineralization in the vessel wall.24,25,27,30–32 Moreover, TLL1is also a member of the BMP-1–related activators of theprocollagen C-proteinase enhancer.33,34 Because TLL1 isat the intersection of the BMP-2/4 and frizzled-relatedprotein-2 signals that drive collagen deposition, TLL1 isalso likely to be a critical regulator of collagenous matrixdeposition, a hallmark of robust calcification processes incardiovascular tissues and in bone.33,34 BMP-2 is upregu-lated in human atherosclerotic plaques26,35 and has beenidentified as a key multifunctional regulator of vascularcalcium accrual in atherosclerosis and diabetic arterioscle-rosis.36,37 Thus, TLL1 has a direct link to fundamentalcellular processes involved in vascular calcification, awidely recognized feature of atherosclerosis that is char-acteristically increased in patients with DM compared withthose without DM.4,20 –22
Previous genetic association studies have observed associ-ations between polymorphisms of PPAR pathway genes andthe risk of developing type 2 DM, obesity, insulin resistance,abnormal lipid profiles, and cardiovascular disease in Euro-pean and American populations (reviewed elsewhere9,38). Inseparate candidate gene studies, an SNP in intron 7 of thePPAR� gene (PPARA IVS7 2498)39 and 2 SNPs in PPARG(the gene encoding PPAR�) were associated with increasedrisk of myocardial infarction in patients with type 2 DM.40
Also using a candidate gene approach, Regieli et al41 reportedan association between the Pro12Ala variant (rs1801282) ofPPARG and both angiographic extent of CAD and 10-yearrisk of ischemic events, including death, among 679 partici-pants in the Regression Growth Evaluation Statin Study(REGRESS) without differentiation of result by diabeticstatus. This association was not confirmed in other studies.42
Of note, in the present analysis, the IVS7 2498 SNP inPPARA and the Pro12Ala polymorphism in PPARG (bothpresent on the PPAR pathway array used) were not identifiedas variants with a significant association with extent of CADin the BARI 2D cohort.
Our findings should be considered in the context ofseveral potential limitations. First, to prevent confounding fromracial admixture, we performed analyses in the largest racialgroups (whites and black) separately. The association betweenTLL1rs1503298 and extent of CAD was observed among whitepatients with DM. Similar analyses did not confirm an associa-tion between TLL1 and extent of CAD among black participantsin BARI 2D, although the smaller sample size may have limitedthe power to detect a significant relationship if present, or theremay be different patterns of linkage disequilibrium betweenTLL1 rs1503298 and the causal variant in different ethnicgroups. Our findings therefore cannot be extrapolated to otherracial groups and should be examined independently in ade-quately powered, racially characterized cohorts, particularly
given the burden of DM and CAD in black and Hispanicpatients. In addition, this study used the genetic cohort of BARI2D, and although the genetic cohort had baseline characteristicssimilar to those of the overall BARI 2D cohort, selection bias ispossible. However, because patients (and their treating physi-cians) were unaware of their genetic code, it is unlikely that therewas a direct association between genotype and participation inthe genetic cohort of BARI 2D to make selection bias signifi-cantly problematic. Furthermore, it should be noted that diabeticpatients with the most extensive, or unstable, CAD for whomimmediate revascularization was considered clinically indicatedwere excluded from BARI 2D by design. Therefore, the ob-served association has not been tested in this potentially impor-tant subgroup of patients with DM and CAD. Finally, it shouldbe noted that the mechanisms of action of TLL1 could poten-tially be via a pathway other than PPAR. Without furtherconfirmation, caution is warranted in attributing the observedassociation to a direct or indirect effect of PPAR activation.
ConclusionsUsing a gene pathway microarray approach, we identified andvalidated a novel PPAR pathway gene variant, TLL1 rs1503298,that is significantly associated with extent of CAD specifically inwhite patients with DM. The reproducibility of the observedassociation across populations, despite varied clinical manifes-tations of CAD and varied methods of quantifying the extent ofatherosclerosis, underscores its potential importance and rele-vance to the fundamental pathophysiology of diabetic athero-sclerosis. Moreover, the physiological pathways affected by thisgenetic variant may provide novel targets for further investiga-tion and therapeutic intervention to address the accelerated rateof progression and the high risk of adverse events associatedwith CAD in patients with DM.
AcknowledgmentsWe would like to thank Dr Dwight A. Towler for critical review ofthis manuscript.
Sources of FundingBARI 2D is funded by the National Heart, Lung, and Blood Instituteand the National Institute of Diabetes and Digestive and KidneyDiseases (U01 HL061744, U01 HL061746, U01 HL061748, andU01 HL063804). Significant supplemental funding is provided byGlaxoSmithKline, Collegeville, PA; Bristol-Myers Squibb MedicalImaging, Inc, North Billerica, MA; Astellas Pharma US, Inc,Deerfield, IL; Merck & Co, Inc, Whitehouse Station, NJ; AbbottLaboratories, Inc, Abbott Park, IL; and Pfizer, Inc, New York, NY.Generous support is given by Abbott Laboratories Ltd, MediSenseProducts, Mississauga, ON, Canada; Bayer Diagnostics, Tarrytown,NY; Becton, Dickinson and Co, Franklin Lakes, NJ; J.R. CarlsonLabs, Arlington Heights, IL; Centocor, Inc, Malvern, PA; Eli Lillyand Co, Indianapolis, IN; LipoScience, Inc, Raleigh, NC; MerckSante, Lyon, France; Novartis Pharmaceuticals Corp, East Hanover,NJ; and Novo Nordisk, Inc, Princeton, NJ. Dr Cresci is funded byNational Institutes of Health (NIH) R21HL089681; NIH SpecializedCenter for Clinically-Oriented Research in Cardiac Dysfunction andDisease P50 HL077113, P60 DK20579; and The Longer LifeFoundation. The content is solely the responsibility of the authorsand does not necessarily represent the official views of the NationalHeart, Lung, and Blood Institute, the National Institute of Diabetesand Digestive and Kidney Diseases, or the NIH.
References1. Lloyd-Jones D, Adams RJ, Brown TM, Carnethon M, Dai S, De SG,
Ferguson TB, Ford E, Furie K, Gillespie C, Go A, Greenlund K, Haase N,Hailpern S, Ho PM, Howard V, Kissela B, Kittner S, Lackland D,Lisabeth L, Marelli A, McDermott MM, Meigs J, Mozaffarian D, Mus-solino M, Nichol G, Roger VL, Rosamond W, Sacco R, Sorlie P, RogerVL, Thom T, Wasserthiel-Smoller S, Wong ND, Wylie-Rosett J. Heartdisease and stroke statistics—2010 update: a report from the AmericanHeart Association. Circulation. 2010;121:e46–e215.
2. The final 10-year follow-up results from the BARI randomized trial. J AmColl Cardiol. 2007;49:1600–1606.
3. Natali A, Vichi S, Landi P, Severi S, L’Abbate A, Ferrannini E. Coronaryatherosclerosis in type II diabetes: angiographic findings and clinicaloutcome. Diabetologia. 2000;43:632–641.
4. Hayden MR, Tyagi SC, Kolb L, Sowers JR, Khanna R. Vascularossification-calcification in metabolic syndrome, type 2 diabetes mellitus,chronic kidney disease, and calciphylaxis-calcific uremic arteriolopathy:the emerging role of sodium thiosulfate. Cardiovasc Diabetol. 2005;4:4.
5. Jacoby RM, Nesto RW. Acute myocardial infarction in the diabeticpatient: pathophysiology, clinical course and prognosis. J Am CollCardiol. 1992;20:736–744.
6. Brooks MM, Frye RL, Genuth S, Detre KM, Nesto R, Sobel BE, KelseySF, Orchard TJ. Hypotheses, design, and methods for the Bypass Angio-plasty Revascularization Investigation 2 Diabetes (BARI 2D) Trial. Am JCardiol. 2006;97:9G–19G.
7. Frye RL, August P, Brooks MM, Hardison RM, Kelsey SF, MacGregorJM, Orchard TJ, Chaitman BR, Genuth SM, Goldberg SH, Hlatky MA,Jones TL, Molitch ME, Nesto RW, Sako EY, Sobel BE. A randomizedtrial of therapies for type 2 diabetes and coronary artery disease. N EnglJ Med. 2009;360:2503–2515.
8. Schwartz L, Kip KE, Alderman E, Lu J, Bates ER, Srinivas V, Bach RG,Mighton LD, Feit F, King S 3rd, Frye RL. Baseline coronary angio-graphic findings in the Bypass Angioplasty Revascularization Investi-gation 2 Diabetes trial (BARI 2D). Am J Cardiol. 2009;103:632–638.
9. Cresci S. The PPAR genes, cardiovascular disease and the emergence ofPPAR pharmacogenetics. Expert Opin Pharmacother. 2005;6:2577–2591.
10. Libby P, Plutzky J. Inflammation in diabetes mellitus: role of peroxisomeproliferator-activated receptor-alpha and peroxisome proliferator-acti-vated receptor-gamma agonists. Am J Cardiol. 2007;99:27B–40B.
11. Madrazo JA, Kelly DP. The PPAR trio: regulators of myocardial energymetabolism in health and disease. J Mol Cell Cardiol. 2008;44:968–975.
12. Bouhlel MA, Staels B, Chinetti-Gbaguidi G. Peroxisome proliferator-ac-tivated receptors–from active regulators of macrophage biology to phar-macological targets in the treatment of cardiovascular disease. J InternMed. 2008;263:28–42.
13. Graham MM, Faris PD, Ghali WA, Galbraith PD, Norris CM, Badry JT,Mitchell LB, Curtis MJ, Knudtson ML. Validation of three myocardialjeopardy scores in a population-based cardiac catheterization cohort. AmHeart J. 2001;142:254–261.
14. Smolderen KG, Spertus JA, Reid KJ, Buchanan DM, Krumholz HM,Denollet J, Vaccarino V, Chan PS. The association of cognitive andsomatic depressive symptoms with depression recognition and outcomesafter myocardial infarction. Circ Cardiovasc Qual Outcomes. 2009;2:328–337.
15. Alpert JS, Thygesen K, Antman E, Bassand JP. Myocardial infarctionredefined: a consensus document of the Joint European Society of Car-diology/American College of Cardiology Committee for the Redefinitionof Myocardial Infarction. J Am Coll Cardiol. 2000;36:959–969.
16. Djousse L, Arnett DK, Carr JJ, Eckfeldt JH, Hopkins PN, Province MA,Ellison RC. Dietary linolenic acid is inversely associated with calcifiedatherosclerotic plaque in the coronary arteries: the National Heart, Lung,and Blood Institute Family Heart Study. Circulation. 2005;111:2921–2926.
17. Higgins M, Province M, Heiss G, Eckfeldt J, Ellison RC, Folsom AR,Rao DC, Sprafka JM, Williams R. NHLBI Family Heart Study: objectivesand design. Am J Epidemiol. 1996;143:1219–1228.
18. Collins-Schramm HE, Phillips CM, Operario DJ, Lee JS, Weber JL,Hanson RL, Knowler WC, Cooper R, Li H, Seldin MF. Ethnic-differencemarkers for use in mapping by admixture linkage disequilibrium. Am JHum Genet. 2002;70:737–750.
19. Carlson CS, Eberle MA, Rieder MJ, Yi Q, Kruglyak L, Nickerson DA.Selecting a maximally informative set of single-nucleotide poly-morphisms for association analyses using linkage disequilibrium. Am JHum Genet. 2004;74:106–120.
20. Carter A, Murphy MO, Turner NJ, Halka AT, Ghosh J, Serracino-InglottF, Walker MG, Syed F. Intimal neovascularisation is a prominent featureof atherosclerotic plaques in diabetic patients with critical limb ischaemia.Eur J Vasc Endovasc Surg. 2007;33:319–324.
21. Hayden MR, Tyagi SC. Vasa vasorum in plaque angiogenesis, metabolicsyndrome, type 2 diabetes mellitus, and atheroscleropathy: a malignanttransformation. Cardiovasc Diabetol. 2004;3:1.
22. Hayden MR, Tyagi SC. Intimal redox stress: accelerated atherosclerosisin metabolic syndrome and type 2 diabetes mellitus: atheroscleropathy.Cardiovasc Diabetol. 2002;1:3.
23. Virmani R, Burke AP, Kolodgie F. Morphological characteristics ofcoronary atherosclerosis in diabetes mellitus. Can J Cardiol. 2006;22(suppl B):81B–84B.
24. Demer LL, Tintut Y. Vascular calcification: pathobiology of a multi-faceted disease. Circulation. 2008;117:2938–2948.
25. Shao JS, Cheng SL, Sadhu J, Towler DA. Inflammation and theosteogenic regulation of vascular calcification: a review and perspective.Hypertension. 2010;55:579–592.
26. Bostrom K, Watson KE, Horn S, Wortham C, Herman IM, Demer LL.Bone morphogenetic protein expression in human atherosclerotic lesions.J Clin Invest. 1993;91:1800–1809.
27. Csiszar A, Smith KE, Koller A, Kaley G, Edwards JG, Ungvari Z.Regulation of bone morphogenetic protein-2 expression in endothelialcells: role of nuclear factor-kappaB activation by tumor necrosis factor-alpha, H2O2, and high intravascular pressure. Circulation. 2005;111:2364–2372.
28. Csiszar A, Lehoux S, Ungvari Z. Hemodynamic forces, vascular oxi-dative stress, and regulation of BMP-2/4 expression. Antioxid RedoxSignal. 2009;11:1683–1697.
29. Zhang M, Zhou SH, Zhao S, Li XP, Liu LP, Shen XQ. Pioglitazone candownregulate bone morphogenetic protein-2 expression induced by highglucose in human umbilical vein endothelial cells. Pharmacology. 2008;81:312–316.
30. Csiszar A, Ahmad M, Smith KE, Labinskyy N, Gao Q, Kaley G, EdwardsJG, Wolin MS, Ungvari Z. Bone morphogenetic protein-2 induces pro-inflammatory endothelial phenotype. Am J Pathol. 2006;168:629–638.
31. Zebboudj AF, Shin V, Bostrom K. Matrix GLA protein and BMP-2regulate osteoinduction in calcifying vascular cells. J Cell Biochem.2003;90:756–765.
32. Li X, Yang HY, Giachelli CM. BMP-2 promotes phosphate uptake,phenotypic modulation, and calcification of human vascular smoothmuscle cells. Atherosclerosis. 2008;199:271–277.
33. Ge G, Zhang Y, Steiglitz BM, Greenspan DS. Mammalian tolloid-like 1binds procollagen C-proteinase enhancer protein 1 and differs from bonemorphogenetic protein 1 in the functional roles of homologous proteindomains. J Biol Chem. 2006;281:10786–10798.
34. Kobayashi K, Luo M, Zhang Y, Wilkes DC, Ge G, Grieskamp T, YamadaC, Liu TC, Huang G, Basson CT, Kispert A, Greenspan DS, Sato TN.Secreted Frizzled-related protein 2 is a procollagen C proteinase enhancerwith a role in fibrosis associated with myocardial infarction. Nat CellBiol. 2009;11:46–55.
35. Kwon HM, Hong BK, Kang TS, Kwon K, Kim HK, Jang Y, Choi D, ParkHY, Kang SM, Cho SY, Kim HS. Expression of osteopontin in calcifiedcoronary atherosclerotic plaques. J Korean Med Sci. 2000;15:485–493.
36. Scatena M, Liaw L, Giachelli CM. Osteopontin: a multifunctionalmolecule regulating chronic inflammation and vascular disease. Arte-rioscler Thromb Vasc Biol. 2007;27:2302–2309.
37. Dhore CR, Cleutjens JP, Lutgens E, Cleutjens KB, Geusens PP, KitslaarPJ, Tordoir JH, Spronk HM, Vermeer C, Daemen MJ. Differentialexpression of bone matrix regulatory proteins in human atheroscleroticplaques. Arterioscler Thromb Vasc Biol. 2001;21:1998–2003.
38. Cresci S. Pharmacogenetics of the PPAR genes and cardiovasculardisease. Pharmacogenomics. 2007;8:1581–1595.
39. Doney AS, Fischer B, Lee SP, Morris AD, Leese G, Palmer CN. Asso-ciation of common variation in the PPARA gene with incident myocardialinfarction in individuals with type 2 diabetes: a Go-DARTS study. NuclRecept. 2005;3:4.
40. Doney AS, Fischer B, Leese G, Morris AD, Palmer CN. Cardiovascularrisk in type 2 diabetes is associated with variation at the PPARG locus: aGo-DARTS study. Arterioscler Thromb Vasc Biol. 2004;24:2403–2407.
Cresci et al Polymorphism and Coronary Disease in BARI 2D 1433
41. Regieli JJ, Jukema JW, Doevendans PA, Zwinderman AH, van der GY,Kastelein JJ, Grobbee DE. PPAR gamma variant influences angiographicoutcome and 10-year cardiovascular risk in male symptomatic coronaryartery disease patients. Diabetes Care. 2009;32:839–844.
42. Pischon T, Pai JK, Manson JE, Hu FB, Rexrode KM, Hunter D, RimmEB. Peroxisome proliferator-activated receptor-gamma2 P12A poly-morphism and risk of coronary heart disease in US men and women.Arterioscler Thromb Vasc Biol. 2005;25:1654–1658.
CLINICAL PERSPECTIVEIn patients with diabetes mellitus, atherosclerosis is more extensive and rapidly progressive. The reasons for this are notknown but may relate in part to genetic variation that influences vascular disease severity. Using microarray technologyto interrogate �3000 tagged single nucleotide polymorphism markers of peroxisome proliferator-activated receptorpathway genes, we have discovered a polymorphism in the TLL1 gene that is significantly associated with angiographicextent of coronary artery disease among patients with type 2 diabetes mellitus. This genetic polymorphism explains morevariance of the phenotype than previously determined clinical factors. This association was initially observed in whitepatients with type 2 diabetes mellitus in the Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D)trial and then validated specifically in the diabetic subgroups of 2 other independent white patient cohorts withquantitatively coronary imaging. TLL1 encodes a metalloproteinase that regulates procollagen processing and themetabolism of inhibitors of bone morphogenetic proteins involved with atherosclerotic calcification. These observationsidentify a novel putative genetic contributor to the biology of diabetic coronary atherosclerosis and suggest that targetingthe TLL1/bone morphogenetic proteins pathway may afford specific therapeutic intervention for diabetic vascular disease.
G. BachMcGill, Edwin L. Alderman, Maria Mori Brooks, Sheryl F. Kelsey, Robert L. Frye and Richard
Sharon Cresci, Jun Wu, Michael A. Province, John A. Spertus, Michael Steffes, Janet B.Angioplasty Revascularization Investigation 2 Diabetes Trial
With Extent of Coronary Artery Disease in Patients With Type 2 Diabetes in the Bypass Peroxisome Proliferator-Activated Receptor Pathway Gene Polymorphism Associated
is online at: Circulation Information about subscribing to Subscriptions:
http://www.lww.com/reprints Information about reprints can be found online at: Reprints:
document. Permissions and Rights Question and Answer this process is available in the
click Request Permissions in the middle column of the Web page under Services. Further information aboutOffice. Once the online version of the published article for which permission is being requested is located,
can be obtained via RightsLink, a service of the Copyright Clearance Center, not the EditorialCirculationin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions: