Effect of six candidate genes on early aging in a French population

(Aging 15: 111-116, 2003),©2003, Editrice Kurtis

Aging Clin Exp Res, Vol. 15, No. 2 111

ABSTRACT. Background and aims: The objective ofthis study was to examine the association between an ag-ing indicator previously defined from a nationwide pop-ulation study, and lipids and apolipoproteins, angiotensinconverting enzyme, paraoxonase activities, and six can-didate genes related to the aging process. Methods:Two hundred and fifty-six healthy Caucasian men(69.8±4.0 years) were included in the study. Totalcholesterol, triglycerides, HDL-cholesterol, lipoprotein(a),apolipoprotein A1, B and E concentrations, and theactivities of paraoxonase, arylesterase, and angiotensin-converting enzymes were determined by standardized lab-oratory methods. A multiplex assay was used to genotypethe studied polymorphisms: apolipoprotein E, lipoproteinlipase, paraoxonase, methylenetetrahydrofolate reduc-tase, cystathionine β-synthase and angiotensin-convert-ing enzymes. Results: Paraoxonase polymorphism atcodon 192 (Gln/Arg) was the only one significantly as-sociated with the aging indicator, Gln homozygotes be-ing more advanced in aging compared with Arg allele car-riers. It was also observed that the aging indicator waspositively correlated with serum concentrations of totalcholesterol, triglycerides and apolipoprotein B, and neg-atively with the activities of basal and stimulated paraox-onase and arylesterase. Multiple regression analysisshowed that triglycerides and basal paraoxonase activi-ty explain 13.6% of the variance of the aging indicator.Conclusions: Triglyceride concentration and paraox-onase gene and activities may contribute to the aging pro-cess. Taking into account the smallness of the samplesize, and the poor level of significance due to the im-plication of paraoxonase polymorphism at codon 192,these results need to be verified in further studies on agreater number of subjects.(Aging Clin Exp Res 2003; 15: 111-116)©2003, Editrice Kurtis

INTRODUCTIONCardiovascular disease (CVD) is age-related and a

leading cause of death worldwide. Elderly individualswho have avoided CVD should possess a protective andbiological background in both genetic profiles and plasmaconstituents against CVD. Consequently, many studieshave investigated genetic polymorphisms and serum vari-ables associated with CVD in a general healthy elderlypopulation to search for the biological factors contribut-ing to successful aging (1, 2). Therefore, we chose to ex-amine the associations between lipids, apolipoproteins,ACE and PON activities, six common bi-allele polymor-phisms, and the aging indicator (AI), representing the dif-ference of biological age (BA) from chronological age(CA). The studied candidate genes are apolipoprotein(apo) E, lipoprotein lipase (LPL), paraoxonase (PON),methylenetetrahydrofolate reductase genes (MTHFR),cystathionine β-synthase (CBS) and angiotensin-convert-ing enzymes (ACE). The polymorphisms of these genesare: apoE at codons 112/158 (ε2, ε3, ε4 alleles), ACE(Alu insertion/deletion alleles, I/D), MTHFR (Ala222Valor 677C/T), PON (Gln192Arg), LPL (Ser447Ter) andCBS (Ile278Thr).

ApoE is one of the major proteins of several lipoproteinparticles and plays a central role in lipoprotein metabolism.Human apoE gene is located on chromosome 19 and hasthree common alleles, designated as ε2, ε3, ε4, encodingthree major apoE isoforms, E2, E3 and E4. The differencein these three isoforms lies in the substitution of amino acidat codons 112 and 158 (3). In comparison with the mostcommon isoform E3, E2 is associated with higher con-centrations of apoE and triglycerides (TG) and lower totalcholesterol (TC), whereas E4 is related to lower concen-trations of apoE and TG, and higher TC in plasma in Eu-ropean populations (3). Many studies have confirmedthat the ε4 allele is a genetic risk factor for both CVD (4)

Aging Clinical and Experimental Research

Effect of six candidate genes on early aging in aFrench populationYongjing Xia1, René Gueguen2, Monique Vincent-Viry1, Gérard Siest1, and Sophie Visvikis1

1Laboratoire du Centre de Médecine Préventive and Unité INSERM 525, 2Département Statistique duCentre de Médecine Préventive, Vandoeuvre-lès-Nancy, France

Key words: Aging, candidate genes, genetic polymorphism, paraoxonase.Correspondence: G. Siest, Ph.D., Centre de Médecine Préventive, 2 Rue du Doyen Jacques Parisot, 54500 Vandoeuvre-lès-Nancy, France. E-mail: Gé[email protected] October 30, 2001; accepted in revised form October 2, 2002.

ORIGINAL ARTICLES

and Alzheimer’s disease (AD) (5) with unfavored longevity(6), whereas ε2 is regarded as a protective factor resultingin longevity, according to a French centenarian populationstudy (7), although 1% of apoE ε2/ε2 carriers suffer fromhypertriglyceridemia (3). ACE is an enzyme which catalyzesthe generation of angiotensin II and also inactivatesbradykinin, the two peptides playing a key role in modu-lating vascular tone. Serum ACE levels are under geneticcontrol, and up to 50% of the variation is due to the I/Dpolymorphism of the ACE gene, highest concentrations be-ing found in DD homozygous (8). Although ACE D allelehas been proposed as a genetic risk marker for CVD (9),this allele was once found to be frequent in centenariansbut later ruled out (7). Mild hyperhomocysteinemia is a riskfactor for CVD. Genetic mutation in MTHFR and CBSgenes may account for reduced enzyme activities and el-evated plasma homocysteine levels (10). LPL is an enzymemediating the TG hydrolysis of chylomicrons and VLDL inthe circulation. Several polymorphisms have been identi-fied at its gene locus, and the polymorphism at position447 (Ser→Ter) has been found to be associated withserum TG concentration and/or CVD (11). The PONgene is polymorphic, and the common mutation is re-sponsible for (Gln→Arg) amino acid substitution in position192 of the protein, the Gln isoform being associatedwith low serum paraoxonase activity, whereas the Arg iso-form is associated with high activity and quicker peroxidehydrolysis. Earlier studies have demonstrated an associa-tion between this polymorphism and CVD. This relation-ship may be partly related to the association with lipid andlipoprotein in HDL (12), which suggests that serumparaoxonase contributes to the ability of HDL to attenuateoxidative changes (13).

An association study is an acceptable method of in-vestigating the biological determinants of human longevi-ty (7). However, new methods still need to be sought,since there are not many healthy centenarians available tocompose a large sample. Thus, we first developed a sta-tistical model estimating an individual’s theoretical agebased upon biological parameters, as an indicator of bi-ological (rather than chronological) aging (14). This indi-cator was applied to evaluate whether the specific allelesof these six candidate genes and the studied serum con-stituents were associated with early aging in a healthyFrench sample population.

METHODSSubjectsThe sample population came from the Centre de

Médecine Préventive (CMP) in Nancy (France), wheresubjects were consecutive attenders for health screening ev-ery five years. They were recruited between September1996 and October 1998 according to the possibility of da-ta collection, and all subjects gave their informed consentfor inclusion in this study. The sample population was com-

posed of 256 men aged 69.8±4.0 (range 53-86) years.Subjects were unrelated and free of chronic diseases andknown pathological states. Clinical measurements such asweight, height, systolic blood pressure (SBP), diastolicblood pressure (DBP) and waist-to-hip ratio (WHR) weresystematic measurements of the general health exami-nation provided by CMP; body mass index (BMI) wascalculated by Quetelet’s formula: weight/height2 (kg/m2).

Laboratory proceduresFor biological measurements of serum and genomic

DNA extraction, venous blood was collected from supinesubjects by venipuncture after overnight fasting. Sera werefrozen in liquid nitrogen (-196°C) until analysis. SerumTC, TG and glucose were measured using standard enzy-matic methods (Merck, Darmstadt, Germany) automated onan AU 5021 apparatus (Olympus-Merck, Nogent-sur-Marne, France). Determination of serum apoE concentra-tion was performed by immunoturbidimetry on a Cobas-Mi-ra (Roche-Diagnostics, Meylan, France) using reagentssupplied by Daiichi Pure Chemicals Co. Ltd. (Tokyo,Japan). HDL-cholesterol was measured by the phospho-tungstate method with reagents supplied by Roche-Boehringer (Meylan, France), followed by measurement ofTC on the same Cobas-Mira as above. Serum lipoprotein(a)[Lp(a)], apoA1 and apoB were determined by im-munonephelometry on a Behring nephelometer analyzer,with Dade-Behring reagents (Paris la Défense, France).Serum PON activities and serum ACE activity were mea-sured on a Cobas-Fara analyzer (Roche-Boehringer, Mey-lan, France). Of the two, the former used 1 mM paraoxonand the latter 0.76 mM synthetic substrate furylacryloyl-phenylalanylglycyl-glycine as substrates (15, 16). The activityof arylesterase was assessed using 1 mM phenylacetate assubstrate on a Uvikon 943 double beam UV/VIS spec-trophotometer (Kontron, France) according to the methodof Eckerson et al. (15). The basal activity of PON wascalled PON(1), and the NaCl-stimulated activity of PON,PON(2). Hematological data, such as mean globular volumeof red blood cells (MGV) and hemoglobin, were measuredon a Coulter® MAXM (Beckman Coulter, Gagny, France).

Determination of genetic polymorphismsNuclear blood cells of EDTA blood samples were stored

at -80°C until DNA extraction, according to the method ofMiller et al. (17). The common apoE polymorphism was de-termined by HhaI digestion following polymerase chain re-action (PCR), with the method of Hixson and Vernier(18). ACE I/D polymorphism genotyping was performedaccording to Evans et al. (19). Genotyping MTHFR C677Twas done according to Frosst et al. (20). Then these threeDNA polymorphism determinations were verified by themultiplex assay of Cheng et al. (21). In addition, threeother genotypes were determined by this assay: C→Gmutation (Ser447→Stop codon) in exon 9 of the LPL

Y. Xia, R. Gueguen, M. Vincent-Viry, et al.

112 Aging Clin Exp Res, Vol. 15, No. 2


gene; Gln→Arg mutation in the PON gene at codon 192;and Ile→Thr mutation in the CBS gene at position 278.

Statistical analysisStatistical analyses were performed by BMDP® statis-

tical software (University of California, Los Angeles).One-way analysis of variance (ANOVA) was used to testthe associations between AI and genetic polymorphisms,concentration of lipids and apolipoproteins, and activitiesof ACE and PON. Linear regression analysis was used toestimate the relationships between AI and serum con-stituents. The multiple comparison test of Tukey andScheffé was used to assess differences in mean values ofAI for each candidate gene. A p-value of 0.05 was re-tained as significant.

Eight subjects with the apoE ε2/ε4 genotype were ex-cluded from this study because of the dominant effect ofapoE ε2 on apoE ε4 allele (22). Since very few subjectshad apoE ε2/ε2 or apoE ε4/ε4 genotypes, subjectswere classified into three groups: subjects carrying apoEε2/ε2 and apoE ε2/ε3 genotypes; subjects with apoEε3/ε3 genotype; and subjects carrying apoE ε4/ε3 andapoE ε4/ε4 genotypes. We also grouped for LPL 447polymorphism, subjects having G/G genotype with sub-jects having C/G genotype (since G allele has the same ef-fects on triglyceride concentration), and for PON poly-morphism (codominance of the two alleles), subjects hav-ing Arg/Arg and Gln/Arg genotypes. Serum HDL-C, TG,Lp(a), apoE concentrations, PON activities [PON(1) andPON(2)] and arylesterase activity were log-transformedto improve normality for statistical testing.

The statistical approach used to define the BA wasbased on a multiple regression of age by several quanti-tative variables showing a significant evolution with age(14). The BA of one individual was then calculated fromthis regression equation: BA= a+Σβi (xi–gi(a)), where xi in-dicates the observed value of variable xi for each variable

included in the model, gi (a) is the regression curve of xi onage, a is the CA of the individual, and βi is the coefficientof regression found for each variable xi. AI is the differencebetween BA and CA (AI= BA – CA). By construction, AIis independent of age: its mean is zero, whatever the age,in the reference group used to estimate the regression co-efficients. The following variables were used to define BA:hearing loss; WHR; SBP; number of healthy teeth; serumglucose concentration; serum total cholesterol concen-tration (TC); mean globular volume (MGV) of red bloodcells; and hemoglobin value.

RESULTSThe clinical and biological characteristics of the sam-

ple population are listed in Table 1. The mean age ofthe studied sample population (69.8±4.0 years old)was older than 65, the low-limitation age for the elder-ly. All biological and morphological parameters hadvalues consistent with the generally good health ofthese subjects. The allele frequencies of the six genepolymorphisms were very close to those observed in oth-er French populations and were in Hardy-Weinbergequilibrium (Table 2) (23, 24).

Table 3 shows the mean value of the AI according togenotype. The genetic polymorphism of PON at codon192 appeared to be significantly associated with the AI(p=0.0486) according to the three genotypes. It wasfound that the Arg allele is the least frequent and that thevalue of AI for those heterozygous subjects was outside thevalue of the AI interval defined for all Gln homozygotes,thus grouping Arg heterozygous and homozygous carri-ers. Comparing the mean value of AI according to the Glncarrier and Arg carrier groups, we found that the p-valuewas 0.0237. Since a positive value for this indicatormeant that aging is advanced and a negative value re-flected that the aging process was retarded, the Arg car-riers (related to high paraoxonase activity) seemed to be

Paraoxonase and early aging in a French population



Table 1 - Biological variables in the sample population (men, 69.8±4.0 years old).

Variables Mean±SD Variables Mean±SD

Age (years) 69.8±4.0 Triglycerides (mmol/L)a 1.38±1.46BMI (kg/m2) 27.1±3.3 Apolipoprotein A1 (g/L) 1.66±0.24WHR 0.97±0.06 Apolipoprotein B (g/L) 1.10±0.21SBP (mmHg) 140.6±14.0 Lipoprotein (a) (g/L)a 0.18±0.24DBP (mmHg) 79.1±8.0 Apolipoprotein E (mg/L)a 41.5±10.4Hemoglobin (g/L) 148.5±12.5 ACE (U/L) 50.4±19.5Glucose (mmol/L) 5.41±1.42 PON(1) (nmol/mn/mL)a 191.0±97.5Total cholesterol (mmol/L) 5.68±0.83 PON(2) (nmol/mn/mL)a 402.4±304.9HDL-Cholesterol (mmol/L)a 1.30±0.35 Arylesterase (μmol/mn/mL)a 103.6±30.2

Values are means±SD; a log-transformed for statistical testing.BMI: body mass index; WHR: waist-to-hip ratio; SBP: systolic blood pressure; DBP: diastolic blood pressure; ACE: angiotensin-converting enzyme; PON: paraox-onase; PON(1): basal paraoxonase activity; PON(2): stimulated paraoxonase activity.

more retarded in the aging process than Gln homozygouscarriers (related to low paraoxonase activity). RegardingapoE, LPL and MTHFR polymorphisms, results obtainedafter pooling heterozygous and homozygous subjects,as described in Methods, were not different from those ob-tained in keeping the three most frequent apoE, LPL orMTHFR genotypes. The only observed difference con-cerns PON polymorphism, presenting a co-dominance ofthe two alleles (Table 3).

TC, TG and apoB concentrations were significantly andpositively correlated with the AI, whereas PON [PON(1)and PON(2)] and arylesterase activities were significantlyand negatively associated with the AI (p<0.05). Regression

coefficients were negative for PON and arylesterase ac-tivities, suggesting that elevated PON activity leads to a de-lay in the aging process, whereas TC, TG and apoBconcentrations lead to acceleration (Table 4).

To further quantify the association between the AIand the studied variables, multiple regression analysiswas applied to estimate the percentage of variation of theAI explained by serum variables significantly found asso-ciated with the AI value. Only PON basal activity and TGconcentration were related to AI variance (Table 5).PON(1) activity alone explained 7.12% of the AI withTG concentration 13.6%.




Table 2 - Allele frequencies of candidate genes in the sample pop-ulation.

Gene/allele Allele Gene/allele Allele frequency frequency

APOE ACE (I/D)ε2 0.114 I 0.438ε3 0.800 D 0.562ε4 0.086 CBS (Ile278Thr)LPL (C447G) Ile 0.924C 0.870 Thr 0.076G 0.130 MTHFR (C677T)PON (Gln192Arg) C 0.654Gln 0.684 T 0.346Arg 0.316

APOE: apolipoprotein E; LPL: lipoprotein lipase; PON: paraoxonase;ACE: angiotensin-converting enzyme; CBS: cystathionine β-synthase; MTHFR: methylenetetrahydrofolate reductase.

Table 4 - Regression analyses of serum variables with value of theaging indicator.

Serum constituents Correlation coefficients

(N=159)Total cholesterol (mmol/L) 0.181*HDL-cholesterol (mmol/L) -0.126 (NS)Triglycerides (mmol/L) 0.234**Apolipoprotein A1 (g/L) -0.123 (NS)Apopolipoprotein B (g/L) 0.169*Lipoprotein (a) (g/L) -0.103 (NS)Apolipoprotein E (mg/L) 0.113 (NS)ACE (U/L) -0.083 (NS)PON(1) (nmol/mn/mL) -0.268***PON(2) (nmol/mn/mL) -0.257***Arylesterase (μmol/mn/mL) -0.165*

NS: not significant; *p<0.05, **p<0.01, ***p<0.001.ACE: angiotensin-converting enzyme; PON: paraoxonase; PON(1): basalparaoxonase activity; PON(2): stimulated paraoxonase activity.

Table 3 - Mean values of the aging indicator (AI) according to genotypes.

Gene candidates Genotypes p-value

APOE ε2/ε3 ε3/ε3 ε3/ε4AI -2.46 -0.45 -0.29 0.34 (NS)LPL(447) C/C C/G G/GAI -0.73 -0.93 -0.26 0.96 (NS)PON(192) Gln/Gln Gln/Arg Arg/ArgAI 0.26 -1.83 -0.53 0.0486*PON(192) Gln/Gln Gln/Arg+ Arg/ArgAI 0.26 -1.57 0.0237*ACE(I/D) I/I I/D D/DAI -0.56 -0.74 -0.89 0.96 (NS)CBS(278) Ile/Ile Ile/ThrAI -0.67 -1.25 0.61 (NS)MTHFR(677) C/C C/T T/TAI -0.16 -1.00 -1.82 0.38 (NS)

NS: not significant; *p<0.05. APOE: apolipoprotein E; LPL: lipoprotein lipase; PON: paraoxonase; ACE: angiotensin-converting enzyme; CBS: cystathionineβ-synthase; MTHFR: methylenetetrahydrofolate reductase.

DISCUSSIONIn this study, we used a statistical indicator, AI, to

study the association of serum lipid parameters and six ge-netic polymorphisms with aging, in a group of healthy el-derly men around 70 years old. The construction of thisaging indicator has been fully described elsewhere (14).The sample population used here was based on a largerstudy named Stanislas Cohort (25), in which some grand-parents were included. Since not enough women qualifiedas subjects, we could only use the male group.

Surprisingly, the significant association of PON ge-netic polymorphism at codon 192 with aging was in-teresting, as subjects carrying the Arg allele had a bio-logical age 1.5 years below their chronological age.These results suggest that higher PON activity is betterfor “successful” aging. This seems clear, since PON is anenzyme bound to HDL sub-fractions and may provideprotection essentially through hydroperoxide hydrolysisin the antiatherogenic properties of HDL by avoiding ox-idative lipoprotein modifications (26). However, some re-ports found a relationship between the Arg allele (highPON activity) and higher serum TG and apoB, but low-er TC and LDL-C concentrations. Thus, the respectiveroles of PON Arg allele and PON activity are complex inassociation with CVD. Epidemiological studies carriedout in healthy subjects, patients with angiographicallydocumented CVD, and non-insulin dependent diabetesmellitus (NIDDM) patients having or not a history ofCVD led to over-representation of the Arg allele inCVD patients and in NIDDM patients with CVD (12,13). Other studies suggest that the Arg allele is associ-ated with increased risk for CVD, whereas others did notshow evidence of associations between the Arg 192 Glnpolymorphism of PON and either myocardial or CVDrisk. However, in this study, we do not show statisticaland significant differences in lipid profiles between var-ious PON genotypes (data not shown). The observed as-sociation between PON polymorphism and the mean AIvalues obtained in each PON genotype may partly be ex-plained by the multiple implications of PON polymor-phism in lipid metabolism and by the new physiologicalrole of PON activity in the hydrolysis of lipid hy-droperoxides generated in lipoproteins upon oxidativestress. This was also suggested by Heijmans et al. (27),

who did not find relations between common PON genevariants (Met55Leu, Arg192Gln) and the risk of car-diovascular mortality in a cohort of elderly people aged85 years and over. In addition, we did not find over-rep-resentation of the Arg allele in this sample population,compared with the frequencies found in Caucasianpopulations.

The association of TG concentration with the AIfound in this study is also a new result. Besides, we re-ported a decrease in the TG concentration in a popu-lation of subjects with AD, when values were com-pared with those of control subjects of the same age (un-published data). This result is independent of the com-mon genetic polymorphism of apoE, and may be linkedto the more general nutritional habits and energymetabolism of middle-aged subjects. In contrast, wecould not find associations between apoE and ACEpolymorphisms with aging in this French sample pop-ulation, as observed in other studies in French cente-narians (1, 7). These authors demonstrated that apoE ε4is associated with a shorter life expectancy, as this iso-form correlates with increased susceptibility for bothCVD and AD, and that apoE ε2 is more frequent inlong-lived individuals (1, 7). They also found that theACE (DD) genotype is associated with increased lifeexpectancy (1), but this finding was refuted by a recentstudy by Blanché et al. (7).

The lack of association between apoE polymorphismand the AI may be explained by the age difference(about 30 years) between our study and the two abovestudies on French centenarians. Many studies fail toshow an association between MTHFR polymorphismand longevity in European populations (28). However,Heijmans et al. (29) found a decrease in the allele fre-quency of MTHFR 677T in an elderly European originalpopulation compared with young controls (4.4 vs12.2%), and concluded that an association of this genevariant (677T) leads to a risk of mortality 3.7-fold high-er in elderly subjects. Perhaps our sample size pre-vented us from obtaining the same results. Meanwhile,to our knowledge, no study has investigated the asso-ciation of CBS and LPL polymorphisms with aging,and we did not demonstrate any significant associa-tion in this study.

Besides the fact that TC is part of the estimate of BA,significant correlations of TC, TG and apoB concen-trations with the AI may be due to their implication inCVD, which reduces successful aging. However, Kervi-nen et al. (6) found concentrations of TC and LDL-C tobe lower in the elderly population. The work of Strand-berg et al. (30) also showed a decreased concentra-tion of TC with age in an elderly population, perhaps ex-plained by less physical activity, or by changes in nutri-tional status, and the lower need for lipids in themetabolism in the elderly population.

Paraoxonase and early aging in a French population



Table 5 - Multiple regression analysis on the aging indicator inmen.

Independent variables β s R2×100

PON(1) (nmol/mn/mL) -0.0207 0.0054 7.12Triglycerides (mmol/L) 2.718 0.801 13.6

β: regression coefficient; s: standard deviation of β; R2: variance explained.PON(1): basal paraoxonase activity.

CONCLUSIONSWe report here that PON polymorphism at codon

192, basal PON activity, and serum concentration ofTG are biological determinants of the aging indicator, andsuggest that they should be taken into account whenstudying aging in an elderly male population. It seems thatcarriers of the Arg allele are more protected against agingthan Gln/Gln homozygous subjects. Because of the small-ness of our sample size and the poor level of signifi-cance, these results need to be validated in further studiescomposed of a larger number of subjects.

REFERENCES1. Schächter F, Faure-Delanef L, Guenot F, et al. Genetic asso-

ciations with human longevity at the APOE and ACE loci.Nature Genetics 1994; 6: 29-32.

2. Luoma PV. Gene activation, apolipoprotein AI-/high densitylipoprotein, atherosclerosis prevention and longevity. PharmacolToxicol 1997; 81: 57-64.

3. Siest G, Pillot T, Regis-Bailly A, et al. Apolipoprotein E: An im-portant gene and protein to follow in laboratory medicine. ClinChem 1995; 41: 1068-86.

4. Stengard JH, Weiss KM, Sing CF. An ecological study of as-sociation between coronary heart disease mortality rates in menand the relative frequencies of common allelic variations in thegene coding for apolipoprotein E. Hum Genet 1998; 103:234-41.

5. Strittmatter WJ, Roses AD. Apolipoprotein E and Alzheimer’sdisease. Proc Natl Acad Sci USA 1995; 92: 4725-7.

6. Kervinen K, Savolainen MJ, Salokannel J, et al. Apolipopro-tein E and B polymorphisms-longevity factors assessed innonagenarians. Atherosclerosis 1994; 105: 89-95.

7. Blanche H, Cabanne L, Sahbatou M, Thomas G. A study ofFrench centenarians: are ACE and APOE associated withlongevity? Life Science 2001; 324: 129-35.

8. Tiret L, Rigat B, Visvikis S, et al. Evidence, from combined seg-regation and linkage analysis, that a variant of the angiotensinI-converting enzyme (ACE) gene controls plasma ACE levels.Am J Hum Genet 1992; 51: 197-205.

9. Krege JH, Moyer JS, Langenbach LL, et al. Angiotensin-converting enzyme gene and atherosclerosis. ArteriosclerThromb Vasc Biol 1997; 17:1245-50.

10. Kang SS, Wong PW. Genetic and nongenetic factors for mod-erate hyperhomocyst(e)inemia. Atherosclerosis 1996; 119:135-8.

11. Jemaa R, Fumeron F, Poirier O, et al. Lipoprotein lipasegene polymorphisms: associations with myocardial infarctionand lipoprotein levels, the ECTIM study. J Lipid Res 1995; 36:2141-6.

12. Mackness MI, Mackness B, Durrington PN, Connelly PW, HegeleRA. Paraoxonase: biochemistry, genetics and relationship toplasma lipoproteins. Curr Opin Lipidol 1996; 7: 69-76.

13. Laplaud PM, Dantoine T, Chapman MJ. Paraoxonase as a riskmarker for cardiovascular disease: facts and hypotheses. ClinChem Lab Med 1998; 36: 431-41.

14. Gueguen R. Proposition of an aging indicator from generalhealth examination in France. Clin Chem Lab Med 2002;40: 235-9.

15. Eckerson HW, Wyte CM, La Du BN. The human serumparaoxonase/arylesterase polymorphism. Am J Hum Genet1983; 35: 1126-38.

16. Dingeon B, Thome H. Enzyme de conversion de l’angiotensine:dosage automatique sur analyseur centrifuge COBAS-BIO etCOBAS-FARA, ou pourquoi faire cher et compliqué quand onpeut faire simple et économique. ISB 1990; 16: 201-6 (inFrench).

17. Miller SA, Dykes DD, Polesky HF. A simple salting out pro-cedure for extracting DNA from human nucleated cells. NucleicAcids Res 1988; 16: 1215.

18. Hixson JE, Vernier DT. Restriction isotyping of humanapolipoprotein E by gene amplification and cleavage withHhaI. J Lipid Res 1990; 31: 545-8.

19. Evans AE, Poirier O, Kee F, et al. Polymorphisms of the an-giotensin-converting-enzyme in subjects who die from coronaryheart disease. Q J Med 1994; 87: 211-4.

20. Frosst P, Blom HJ, Milos R, et al. A candidate genetic risk fac-tor for vascular disease: a common mutation in methylenete-trahydrofolate reductase. Nat Genet 1995; 10:111-3.

21. Cheng S, Pallaud C, Grow MA, et al. A multilocus genotypingassay for cardiovascular disease. Clin Chem Lab Med 1998; 36:561-6.

22. Bohnet K, Régis-Bailly A, Vincent-Viry M, Gueguen R, Siest G,Visvikis S. Apolipoprotein E genotype ε4/ε2 in the STANIS-LAS cohort study. Dominance of the epsilon2 allele? AnnHum Genet 1996; 60: 509-16.

23. Pallaud C, Maurice M, Cheng S, et al. Multilocus approach tocardiovascular risk. Scand J Clin Lab Invest 1999; 59: 168-76.

24. Sass C, Zannad F, Herbeth B, et al. Apolipoprotein E4,lipoprotein lipase C447 and angiotensin-I converting enzymedeletion alleles were not associated with increased wall thick-ness of carotid and femoral arteries in healthy subjects from theStanislas cohort. Atherosclerosis 1998; 140: 89-95.

25. Siest G, Visvikis S, Herbeth B, et al. Objectives, design and re-cruitment of a familial and longitudinal cohort for studying gene-environment interactions in the field of cardiovascular risk: theStanislas cohort. Clin Chem Lab Med 1998; 36: 35-42.

26. Mackness B, Durrington PN, Mackness MI. Human serumparaoxonase. Gen Pharmacol 1998; 31: 329-36.

27. Heijmans BT, Westendorp RGJ, Lagaay AM, Knook DL,Kluft C, Slagboom PE. Common paraoxonase gene variants,mortality risk and fatal cardiovascular events in elderly subjects.Atherosclerosis 2000; 149: 91-7.

28. Brattström L, Zhang Y, Hurtig M, et al. A commonmethylenetetrahydrofolate reductase gene mutation and longevi-ty. Atherosclerosis 1998; 141: 315-19.

29. Heijmans BT, Westendorp RGJ, Slagboom PE. Commongene variants, mortality and extreme longevity in humans.Exp Gerontol 2000; 35: 865-77.

30. Strandberg TE, Valvanne J, Erkinjuntti T, Sorva A, Tilvis RS.Serum lipids, health, and one-year mortality in randomized agecohorts of 75, 80, and 85 years: the Helsinki Ageing Study.Nutr Metab Cardiovasc Dis 1992; 2: 101-5.




Documents

Effect of six candidate genes on early aging in a French population