Government & Nonprofit

ATP-binding cassette transporter A1 locus is not a major determinant of HDL-C levels in a population at high risk for coronary heart disease

ATP-binding cassette transporter A1 (ABCA1) transports cellular cholesterol to lipid-poor apolipoproteins. Mutations in the ABCA1 gene are linked to rare phenotypes, familial hypoalphalipoproteinemia (FHA) and Tangier disease (TD), characterized by
of 6
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  ATP-binding cassette transporter A1 locus is not a major determinantof HDL-C le v els in a population at high risk for coronary heartdisease Sakari Kakko a, * ,1 , Jani Kelloniemi a , Peter  v on Rohr b , Ina Hoeschele b ,Minna Tamminen a , Margaret E. Brousseau c , Y. Antero Kesa¨niemi a , MarkkuJ. Sa v olainen a a Department of Internal Medicine and Biocenter Oulu, Uni  v ersity of Oulu, P.O. BOX 5000, 90014 Oulu, Finland  b Departments of Dairy Science and Statistics, Virginia Polytechnic Institute and State Uni  v ersity, Blacksburg, VA, USA c The Lipid Metabolism Laboratory, JM-USDA Human Nutrition Research Center on Aging at Tufts Uni  v ersity and Department of Medicine, NewEngland Medical Center, Boston, MA, USA Recei v ed 27 February 2002; recei v ed in re v ised form 23 May 2002; accepted 31 May 2002 Abstract ATP-binding cassette transporter A1 (ABCA1) transports cellular cholesterol to lipid-poor apolipoproteins. Mutations in theABCA1 gene are linked to rare phenotypes, familial hypoalphalipoproteinemia (FHA) and Tangier disease (TD), characterized bymarkedly decreased plasma high-density lipoprotein cholesterol (HDL-C) le v els. The aim was to test if the ABCA1 locus is a majorlocus regulating HDL-C le v els in the homogenous Finnish population with a high pre v alence of coronary heart disease (CHD).Firstly, the ABCA1 locus was tested for linkage to HDL-C le v els in 35 families with premature CHD and low HDL-C le v els.Secondly, 62 men with low HDL-C le v els and CHD were screened for the fi v e mutations known to cause FHA. Thirdly,polymorphisms of the ABCA1 gene were tested for an association with HDL-C le v els in a population sample of 515 subjects. TheABCA1 locus was not linked to HDL-C le v els in the CHD families, and no carriers of the FHA mutations were found. The AA596genotype was associated with higher HDL-C le v els compared with the GG and GA genotypes in the women, but not in the men. TheG596A genotypes explained 4% and the A2589G genotypes 3% of the  v ariation in plasma HDL-C le v els in women. The data suggestthat the ABCA1 locus is of minor importance in the regulation of HDL-C in Finns. #  2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords:  Genetic; Genetic polymorphisms; High-density lipoproteins; Coronary heart disease; Quantitati v e trait locus 1. Introduction CHD is a major cause of morbidity and mortality inaffluent societies. Plasma high-density lipoprotein cho-lesterol (HDL-C) le v els are in v ersely associated with theCHD risk [1] and low HDL-C is the main lipidabnormality associated with early and familial CHD[2]. The genetic basis of HDL-C regulation seems to belargely heterogeneous, since se v eral loci on differentchromosomes ha v e been found to be linked significantlyor suggesti v ely to plasma HDL-C le v els in pre v iouslypublished whole-genome scans [3    / 9], and in additionse v eral candidate genes of lipoprotein metabolism [10    / 14] ha v e been found to be associated with or linked toplasma HDL-C le v els. Howe v er, the major loci affectingplasma HDL-C le v els, i.e. loci explaining most of theHDL-C  v ariation at the population le v el, are stillunknown.ABCA1 has been shown to be in v ol v ed in apolipo-protein-mediated lipid efflux from peripheral cells [15]and the ABCA1 gene on the chromosomal location9q31.1 has been found to be the locus for the extremelyrare Tangier disease (TD) characterized by a  v irtual * Corresponding author. Tel.:  / 358-8-315-4570; fax:  / 358-8-315-4543 E-mail address: (S. Kakko). 1 v olainen.htmAtherosclerosis 166 (2003) 285    / 290www.else v$ - see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved.PII: S0021-9150(02)00232-0  absence of plasma HDL-C and also, in some families,for familial hypoalphalipoproteinemia (FHA) [16    / 20].In addition to the rare mutations leading to a se v erephenotype, either TD or FHA, the ABCA1 genecontains polymorphisms that ha v e been found to beassociated with HDL-C le v els in some [21], but not all,studies [22    / 24] and with the CHD risk [22    / 24].Therefore, ABCA1 seems to be a promising candidategene for plasma HDL-C regulation in the generalpopulation. The aim of the current study was todetermine, using three different approaches, whetherthe ABCA1 gene is a major locus regulating plasmaHDL-C le v els at the population le v el. Firstly, theABCA1 locus was tested for linkage with HDL-C le v elsin families with premature CHD and low HDL-C le v els.Secondly, men with low HDL-C le v els and CHD werescreened for fi v e mutations known to cause FHA[16,19]. Thirdly, the ABCA1 gene with three singlenucleotide polymorphisms (SNPs) was tested for asso-ciation with plasma HDL-C le v els in a populationsample. 2. Materials and methods  2.1. Subjects Families with early CHD and low HDL-C le v els werecollected for the linkage sub-study based on probandswith premature CHD (i.e. acute myocardial infarction,coronary artery bypass operation or coronary angio-plasty before the age of 55 years) and low HDL-C le v els( B / 1.1 mmol/l). The probands were selected from thefiles of Oulu Uni v ersity Hospital. In addition, theprobands were to ha v e normal to moderately ele v atedle v els of triglycerides and total cholesterol ( B / 3.5 and B / 7 mmol/l, respecti v ely) and no diabetes. Complexfamilies were recruited to ha v e as many CHD patients aspossible. The total number of families was 35 (with 375subjects genotyped). In CHD patients, lipid measure-ments were taken at least 3 months after myocardialinfarction or coronary bypass operation. The CHD andlow HDL-C group for FHA mutation screening con-sisted of 62 Caucasian men. They had all undergone acoronary bypass graft operation and had had plasmaHDL-C le v els below 0.9 mmol/l, 3    / 48 months after thebypass operation. The population sample (253 men and262 women, age 40    / 61 years) for the association sub-study consisted of the control cohort of our the OuluProject Elucidating Risk of Atherosclerosis (OPERA)study and the subjects were selected from among theinhabitants of the town of Oulu as described earlier [25].To keep the population sample representati v e of thegeneral population, the subjects (32 men and threewomen) with CHD were not excluded. Ele v en personson lipid-lowering medication (statins, fibrates or resins)were excluded from the sample. The study was appro v edby the Ethical Committee of the Uni v ersity of Oulu, andall the subjects ga v e written informed consent.  2.2. Lipid and lipoprotein measurements Blood samples were obtained after an o v ernight fast.Plasma lipoprotein fraction were separated by sequen-tial ultracentrifugation and concentrations of choles-terol and triglycerides in the plasma and lipoproteinfractions were determined by enzymatic colorimetricmethods, as described in detail earlier [26].  2.3. DNA analysis Genomic DNA was extracted from peripheral bloodnuclear cells by the salting-out method [27]. For linkagesub-study, 14 commercially a v ailable microsatellitemarkers (ABI Prism Linkage Mapping Set 1, AppliedBiosystems, Foster City, CA, USA) in chromosome ninewith two additional markers, namely D9S300 andD9S311 (with distances of 1 and 4 cM, respecti v ely, tothe ABCA1 locus) were genotyped with an ABI 377automatic sequencer (Applied Biosystems). Thescreened FHA mutations were C6370T,  D (E1833,D1834), C2665T, T3212C and  D L633 [16,19]. The genotyped SNPs of the ABCA1 locus were the G596A(changing arginine, amino acid 159 of the ABCA1protein, to lysine, R159K), A2589G (changing isoleu-cine 823 to methionine, I823M), and G3456C poly-morphisms (changing glutamic acid 1112 to asparticacid, E1112D). The genotyping processes could befound in detail in the web-site.  2.4. Statistical analyses All the statistical analyses for the association sub-study were carried out using  SSPS  v ersion 10.0 (SPSSInc., Chicago, IL, USA). Lipid and lipoprotein  v alueswere adjusted for age, body mass index (BMI), smokingand alcohol consumption in men and women as well asfor hormone replacement therapy in women. Thedifferences in lipid and lipoprotein  v alues between thedifferent genotypes were e v aluated by one-way analysisof   v ariance (ANOVA). Because of multiple groupcomparisons, the Bonferroni correction was used todetermine the statistically significant differences. Theresults are gi v en as means (standard de v iations, S.D.s)or percentages. The general factorial procedure of thegeneral linear model (GLM) with Type III sums of squares was used to construct the model to explain the v ariation in HDL-C le v els. The genotypes of e v erypolymorphism indicated were added as an independentfactor into the model without any assumption of themode of inheritance (i.e. without any combination of genotypes). The  h 2 -statistics was used to estimate how S. Kakko et al. / Atherosclerosis 166 (2003) 285    /   290 286  much of the  v ariation was explained by each factor orco v ariate.Haplotype frequencies were estimated with the  ARLE-QUIN  ( v ersion 1.1) program (Genetics and BiometryLaboratory, Uni v ersity of Gene v a, Switzerland). Thestatistical significance of the linkage disequilibria be-tween each pair of polymorphisms was calculated bycomparing the frequencies of the obser v ed haplotypes(defined by two polymorphisms) with their expectedfrequencies under the assumption of total linkageequilibrium by chi-square test. The most likely haplo-types for e v ery subject were reconstructed from thegenotypes of the three ABCA1 gene polymorphismsbased on the estimated haplotype frequencies.Quantitati v e multipoint linkage analysis was per-formed with the  SOLAR  program [28]. This programuses a  v ariance components approach to map thepositions of putati v e quantitati v e trait loci (QTLs).The total obser v ed phenotypic  v ariance is split intocomponents attributed to QTL, to residual polygeniceffects and to non-genetic residual effects. Plasma HDL-C le v els were used as a continuous  v ariable in theanalysis. The sex and age of the study subjects were usedas co v ariates in the analysis. The allele frequencies of e v ery microsatellite marker were estimated from thestudy material using a maximum likelihood procedureand used in the calculation of identity-by-descentcoefficients for the QTL co v ariance matrices in the v ariance components analysis.In the power calculation of the association sub-study[29] the S.D. was assumed to be 0.35 (the S.D. of adjusted HDL-C in women), the power of the study 0.80and  P  - v alue for the significance less than 0.05. For theG596A polymorphisms (with the genotype frequenciestaken from the sample), the difference in plasma HDL-Cle v els between the 596GG and 596GA genotypes shouldbe about 0.14 and 0.28 mmol/l between the homozygousgenotypes to be significant. For the A2589G poly-morphisms the differences (the genotype frequenciestaken from the sample) should be about 0.18 and 0.42mmol/l, respecti v ely. 3. Results 3.1. Characteristics of study subjects The characteristics of the study subjects are shown inTable 1. 3.2. Linkage sub-study Genetic factors explained 49% of the  v ariability in theplasma HDL-C le v els in families with early CHD andlow HDL-C le v els and a maximum lod score of 0.45 wasyielded for chromosome 9, which is far from e v ensuggesti v e linkage [30]. The lod score cur v e could befound in the web-site. 3.3. FHA mutation screening  A total number of 62 men with CHD and low HDL-Cle v els were screened for fi v e FHA mutations [16,19], butnone of them were found. 3.4. Association sub-study The frequencies of the rare allele of the SNPs were22.0 (A596), 11.2 (G2589) and 2.6%, (C3456). Thegenotypes obtained were in Hardy    / Weinberg equili-brium.There were no significant differences in the lipid andlipoprotein  v alues between the genotypes of any of thepolymorphisms in men. No statistically significantdifferences were detected between the genotypes of theG3456C polymorphism in women (data not shown). Inwomen, the 596AA genotype was associated withsignificantly higher HDL-C le v els compared with the596GG genotype (a mean of 1.80  v s. 1.55 mmol/l,  P   / 0.04). Alcohol consumption  v aried considerably be-tween the A2589G genotypes, although significantlyonly in women ( P  B / 0.001). Therefore, the subjectsbelonging to the highest quartiles of alcohol consump-tion were excluded from the analysis (o v er 138 g of alcohol for a week in men and o v er 30 g in women), andafter the exclusion, the women with the 2589AAgenotype ( n  / 153) had lower HDL-C le v els than thewomen with the 2589AG genotype ( n  / 30) (a mean of 1.52  v s. 1.67 mmol/l,  P  B / 0.05). There was only onewoman with the rare 2589GG genotype with alcoholconsumption below the highest quartile.Linkage disequilibrium was detected between thepolymorphisms at the nucleotides 596 and 2589 ( P  B / 0.001) and 596 and 3456 ( P  B / 0.05), but not between2589 and 3456. The estimated frequencies of haplotypesare shown in Table 2. The most likely haplotypes werereconstructed for e v ery study subject ha v ing all threeABCA1 gene polymorphisms genotyped ( n  / 498) andthe number of the reconstructed haplotypes in thetertiles of plasma HDL-C le v els could be found in Table2. In men, no significant differences in the haplotypefrequencies were detected. In women, the haplotype 1(defined in Table 2) was associated with low plasmaHDL-C le v els (80% of haplotypes in lowest HDL-Ctertile  v s. 63% in the highest,  P  B / 0.01), whereashaplotypes 5 and 6 were associated with high plasmaHDL-C le v els (12  v s. 20%,  P   / 0.08 and 3  v s. 10%,  P  B / 0.05, respecti v ely). The haplotypes 5 and 6 were the onlyones with the allele A596 and when combined, theirfrequency was significantly higher in the highest HDL-Ctertile than in the lowest HDL-C tertile (29  v s. 14%,  P  B / 0.01). S. Kakko et al. / Atherosclerosis 166 (2003) 285    /   290  287  GLM models suggested that BMI, alcohol consump-tion and the genotypes of the G596A ( P   / 0.02) andA2589G polymorphisms ( P   / 0.03) were associated withthe HDL-C le v els, explaining 9, 4, 4 and 3%, respec-ti v ely, of the  v ariation in the plasma HDL-C le v els inwomen. If the genotypes of A2589G polymorphismswere excluded from the model, the G596A genotypesexplained about 3% of the  v ariation. In men, BMI andalcohol consumption were associated with the plasmaHDL-C le v els, explaining 6 and 3% of the  v ariation,respecti v ely, whereas none of the genotypes weresignificantly associated with HDL-C le v els. 4. Discussion Gi v en the strong in v erse relationship between theplasma HDL-C le v els and the CHD risk, the establish-ment of major genetic loci regulating plasma HDL-Cle v els at the population le v el could gi v e us importantinsight into atherogenesis. Although a strong geneticcomponent regulating plasma HDL-C le v els has beenclearly demonstrated [31], no major locus has been found. Numerous association studies of   v arious candi-date genes and whole-genome linkage scans ha v e yieldedcontradictory results indicating probably great geneticheterogeneity as described in the Introduction. TheABCA1 gene on chromosomal location 9q31.1, whichis a locus for rare TD and FHA [16    / 20], seemed to be apromising candidate locus for HDL-C regulation e v enat the population le v el.In the study of Wang et al. [21], the G2589 allele of the ABCA1 gene was associated with higher plasmaHDL-C le v els. The association was studied in sixdifferent populations, but only detected in Inuits. Noassociation between this  v ariant and the plasma HDL-Cle v els was detected in the other association studies [22    / 24], whereas an association between the CHD risk(determined either angiographically or with clinicalendpoints) and the polymorphisms of the ABCA1 genewas found in these studies. In our study material, wewere able to find an association between the polymorph-isms of the ABCA1 gene and plasma HDL-C le v els onlyin women. This association explained a moderateproportion (about 7%) of the  v ariation in plasmaHDL-C le v els, meaning that the ABCA1 locus is only Table 1Characteristics of the study subjectsPopulation sample Families with early CHD and low HDL-C Men with CHD and low HDL-CMen Women Men WomenNumber of subjects 253 262 212 163 62Age (years) 50.8 (6.0) 51.7 (6.0) 48.6 (13.1) 49.3 (16.5) 58.1 (8.1)CHD patients 32 (13%) 3 (1%) 90 (42%) 19 (12%) 62 (100%)BMI (kg/m 2 ) 26.5 (3.5) 26.3 (4.5) 26.9 (3.9) 25.5 (4.7) 26.7 (2.5)Current smokers 102 (40%) 72 (27%) 49 (23%) 23 (14%) 6 (10%)Alcohol consumption (g/week) 91 (100) 23 (36) 70 (109) 18 (33) n.d.Total cholesterol (mmol/l) 5.77 (1.08) 5.51 (1.03) 5.37 (1.06) 5.49 (1.15) 5.65 (0.72)HDL cholesterol (mmol/l) 1.23 (0.30) 1.56 (0.38) 1.07 (0.29) 1.39 (0.33) 0.77 (0.09)LDL cholesterol (mmol/l) 3.71 (0.94) 3.30 (0.91) 3.33 (0.87) 3.30 (0.97) 3.96 (0.73)Triglycerides (mmol/l) 1.56 (0.81) 1.19 (0.68) 1.78 (1.02) 1.49 (1.54) 2.01 (1.08)Values are expressed as means (S.D.s) or as number of subjects (%). BMI, body mass index. HDL, high-density lipoprotein. LDL, low-densitylipoprotein. n.d., not determined. CHD, coronary heart disease.Table 2The haplotypes of three polymorphisms of the ABCA 1 gene and their estimated numbers in the tertiles of plasma HDL-C le v els, separately for menand women of the population sampleHaplotype Percentage Allele Men Women596 2589 3456  B 1.08 1.08    / 1.32   1.32  B 1.37 1.37    / 1.68   1.681 70.3 G A G 104 105 105 121 104 98**2 2.4 G A C 5 1 4 4 4 43 5.2 G G G 5 5 9 5 8 84 0.2 G G C 1 0 0 0 0 05 16.1 A A G 18 29 19 18 24 316 5.8 A G G 7 14 7 4 10 15*Total 100 140 154 144 152 150 156*,  P  B 0.05 and **,  P  B 0.01 by  x 2 -test for the difference between the lowest and highest tertiles. The common allele shown in bold. S. Kakko et al. / Atherosclerosis 166 (2003) 285    /   290 288  of minor importance in HDL-C regulation. The findingwas supported by our linkage sub-study, in which noe v idence for linkage between the ABCA1 locus and theHDL-C le v els was found. No known FHA mutationswere found in men with CHD and low HDL-C le v els.After negati v e linkage results and only weak associa-tions between ABCA1 gene polymorphisms and plasmaHDL-C le v els, no further sequencing for finding no v elFHA mutations in the ABCA1 gene was done.An association between the HDL-C le v els and thepolymorphisms of the ABCA1 gene was detected only inwomen. The reason for this is unknown. It might bepossible that this sex difference is due to differenten v ironmental factors between the sexes. The men inthe population sample smoked more frequently andconsumed more alcohol than the women (Table 1),which probably masked the associations between theplasma HDL-C le v els and the polymorphisms of theABCA1 gene. On the other hand, no studies describingthe effect of sex hormones on ABCA1 function ha v ebeen published.The lack of e v idence of linkage between the ABCA1locus and HDL-C le v els could also be due to the specificselection of the study material based on low le v els of HDL-C. This selection could ha v e reduced the o v erall v ariation in HDL-C and, consequently, also the  v aria-tion attributable to putati v e QTL. Howe v er, this seemsan unlikely explanation, since the  v ariances of plasmaHDL-C le v els are similar in the population sample andin the families with low HDL-C le v els and prematureCHD (Table 1). The findings of the linkage sub-studyare supported by the association sub-study, in which theABCA1 locus seemed to ha v e only a minor effect onHDL-C le v els. It would require hundreds of families todetect this minor effect in a linkage study.Based on pre v ious studies in TD and FHA families[16    / 20], the ABCA1 gene seemed to be a promisinglocus regulating plasma HDL-C le v els also in thegeneral population. In our series by using three differentmethods, we were able to exclude it as a major locusdetermining plasma HDL-C le v els. This finding issupported by most of the association studies publishedso far [22    / 24]. Acknowledgements The authors gratefully acknowledge the expert tech-nical assistance of Saara Korhonen, Saija Kortetja¨r v i,Marja-Leena Kyto¨kangas, Sari Pyrho¨nen, Sirpa Ran-nikko and Eila Saarikoski. This study was supported bygrants from the Research Council for Health of theAcademy of Finland, the Finnish Foundation forCardio v ascular Research, the Sigrid Juselius Founda-tion, the Finnish Cultural Foundation, the Paa v o NurmiFoundation, the Finnish Medical Foundation and theFinnish Medical Society Duodecim. References [1] The Framingham Study, Gordon T, Castelli WP, Hjortland MC,Kannel WB, Dawber TR. High density lipoprotein as a protecti v efactor against coronary heart disease. Am J Med 1977;62:707    / 14.[2] Genest JJ, Jr, Martin-Munley SS, McNamara JR, et al. Familiallipoprotein disorders in patients with premature coronary arterydisease. Circulation 1992;85:2025    / 33.[3] Almasy L, Hixson JE, Rainwater DL, et al. Human pedigree-based quantitati v e-trait-locus mapping: localization of two genesinfluencing HDL-cholesterol metabolism. Am J Hum Genet1999;64:1686    / 93.[4] Peacock JM, Arnett DK, Atwood LD, et al. Genome scan forquantitati v e trait loci linked to high-density lipoprotein choles-terol: The NHLBI Family Heart Study. Arterioscler Thromb VascBiol 2001;21:1823    / 8.[5] Arya R, Duggirala R, Almasy L, et al. Linkage of high-densitylipoprotein-cholesterol concentrations to a locus on chromosome9p in Mexican Americans. Nat Genet 2001;30:102    / 5.[6] Kort EN, Ballinger DG, Ding W, et al. E v idence of linkage of familial hypoalphalipoproteinemia to a no v el locus on chromo-some 11q23. Am J Hum Genet 2000;66:1845    / 56.[7] Coon H, Leppert MF, Eckfeldt JH, et al. Genome-wide linkageanalysis of lipids in the Hypertension Genetic EpidemiologyNetwork (HyperGEN) Blood Pressure Study. ArteriosclerThromb Vasc Biol 2001;21:1969    / 76.[8] Imperatore G, Knowler WC, Pettitt DJ, Kobes S, Fuller JH,Bennett PH, Hanson RL. A locus influencing total serumcholesterol on chromosome 19p: results from an autosomalgenomic scan of serum lipid concentrations in Pima Indians.Arterioscler Thromb Vasc Biol 2000;20:2651    / 6.[9] Soro A, Pajukanta P, Lilja HE, et al. Genome scans pro v idee v idence for low-HDL-C loci on chromosomes 8q23, 16q24.1-24.2, and 20q13.11 in Finnish families. Am J Hum Genet2002;70:1333    / 40.[10] Inazu A, Jiang XC, Haraki T, et al. Genetic cholesteryl estertransfer protein deficiency caused by two pre v alent mutations as amajor determinant of increased le v els of high density lipoproteincholesterol. J Clin In v est 1994;94:1872    / 82.[11] Franceschini G. Apolipoprotein function in health and disease:insights from natural mutations. Eur J Clin In v est 1996;26:733    / 46.[12] Kui v enho v en JA, Pritchard H, Hill J, Frohlich J, Assmann G,Kastelein J. The molecular pathology of lecithin:cholesterolacyltransferase (LCAT) deficiency syndromes. J Lipid Res1997;38:191    / 205.[13] Fisher RM, Humphries SE, Talmud PJ. Common  v ariation in thelipoprotein lipase gene: effects on plasma lipids and risk of atherosclerosis. Atherosclerosis 1997;135:145    / 59.[14] Guerra R, Wang J, Grundy SM, Cohen JC. A hepatic lipase(LIPC) allele associated with high plasma concentrations of highdensity lipoprotein cholesterol. Proc Natl Acad Sci USA1997;94:4532    / 7.[15] Oram JF, Lawn RM. ABCA1. The gatekeeper for eliminatingexcess tissue cholesterol. J Lipid Res 2001;42:1173    / 9.[16] Brooks-Wilson A, Marcil M, Clee SM, et al. Mutations in ABC1in Tangier disease and familial high-density lipoprotein deficiency.Nat Genet 1999;22:336    / 45.[17] Bodzioch M, Orso E, Klucken J, et al. The gene encoding ATP-binding cassette transporter 1 is mutated in Tangier disease. NatGenet 1999;22:347    / 51. S. Kakko et al. / Atherosclerosis 166 (2003) 285    /   290  289
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!