American Journal of Medical Genetics (Neuropsychiatric Genetics) 105:761±764 (2001)

Alzheimer Disease Is Not Associated WithPolymorphisms in the Angiotensinogenand Renin Genes

Alison Taylor,1 Mario Ezquerra,2 Gurjog Bagri,1 Agustin Yip,1,3 Louisa Goumidi,2 Dominique Cottel,2

Douglas Easton,4 John Grimley Evans,5 John Xuereb,6 Nigel J. Cairns,7 Philippe Amouyel,2

Marie-Christine Chartier-Harlin,2 Carol Brayne,3 and David C. Rubinsztein1*1Department of Medical Genetics, Wellcome Trust Centre for Molecular Mechanisms in Disease, Cambridge Institute forMedical Research, Addenbrooke's Hospital, Cambridge, U.K.2INSERM 508, Institut Pasteur de Lille, Lille Cedex, France3Department of Public Health and Primary Care, University Forvie Site, Cambridge, U.K.4CRC Genetic Epidemiology Unit, Cambridge University Department of Public Health and Primary Care, StrangewaysResearch Laboratory, Cambridge, U.K.5Department of Clinical Geratology, Radcliffe In®rmary, Oxford, U.K.6Department of Pathology, Cambridge University, Addenbrooke's Hospital, Cambridge, U.K.7Brain Bank, Department of Neuropathology, Institute of Psychiatry, King's College, London, U.K.

Hypertension has been implicated as a riskfactor for Alzheimer disease (AD) anddementia in epidemiological studies ofhumans. It is thus possible that there arecommon genetic determinants for hyperten-sion and AD. Epidemiological, clinical, andexperimental data suggest that the renin-angiotensin-aldosterone system is a criticalregulator of blood pressure. The presence ofan MboI site in an RFLP in the renin geneand the Thr at the Met/Thr polymorphism atcodon 235 (M235T) of the angiotensinogengene have been reported to be associatedwith hypertension. These variants werestudied in autopsy-con®rmed AD cases andmatched controls from the U.K. While noassociation was detected with the reninpolymorphism, a weak deleterious effectwas observed in cases homozygous for theangiotensinogen Thr allele. However, thisassociation was not observed in a Frenchcohort of clinically diagnosed AD cases andcontrols, suggesting that the initial observa-tion was a type I error. Thus, these poly-morphisms are unlikely to be associatedwith AD risk. ß 2001 Wiley-Liss, Inc.

KEY WORDS: Alzheimer disease; demen-tia; renin; angiotensinogen;MRC CFAS study

INTRODUCTION

Alzheimer disease (AD) is characterized clinically bydementia and neuropathologically by the presence of b-amyloid plaques and neuro®brillary tangles. Althoughthe pathological changes are similar at all ages of onset,AD is often divided into early- and late-onset diseaseusing an arbitrary age-at-onset cutoff of 65 years. MostAD cases are late onset.

Mutations in the amyloid precursor protein, preseni-lin 1, and presenilin 2 can cause autosomal dominantforms of early-onset AD. The only locus unequivocallyassociated with late-onset AD risk is APOE. Howeverthis locus accounts for less than half the geneticvariance [Farrer et al., 1997; Rubinsztein and Easton,1999].

Hypertension has been implicated as a risk factor forAD and dementia in epidemiological studies of humans[Lis and Gaviria, 1997; Skoog, 1997]. Furthermore,spontaneously hypertensive rats are prone to memoryimpairment compared to nonhypertensive rats [Gattuet al., 1997]. It is thus possible that there are commongenetic determinants for hypertension and AD. Epide-miological, clinical, and experimental data suggest thatthe renin-angiotensin-aldosterone system is a criticalregulator of blood pressure [Ganong, 1981]. Hence,genes involved in this pathway are good candidates forquantitative trait loci in human hypertension [Rigatet al., 1990]. A link between the renin-angiotensin-aldosterone system and AD is possibly providedby studies of the DCP1/ACE gene, which encodes

Grant sponsor: The Medical Research Council; Grant sponsor:INSERM, Institut Pasteur de Lille.

*Correspondence to: David C. Rubinsztein, Department ofMedical Genetics, Wellcome Trust Centre for Molecular Mechan-isms in Disease, Cambridge Institute for Medical Research,Wellcome/MRC Building, Addenbrooke's Hospital, Hills Road,Cambridge, CB2 2XY, U.K. E-mail: dcr1000@cus.cam.ac.uk

Received 23 April 2001; Accepted 1 August 2001

DOI 10.1002/ajmg.10044

ß 2001 Wiley-Liss, Inc.

angiotensin-converting enzyme (ACE). Kehoe et al.[1999] recently reported an association between AD andthe insertion allele of an insertion/deletion (I/D) poly-morphism in intron 16 in ACE. Although this ®ndinghas not been consistently replicated, the effect of the I-allele on AD risk is supported by our subsequent meta-analysis [Narain et al., 2000]. Two other genes codingfor participants in the renin-angiotensin-aldosteronepathway that have been implicated in hypertension arerenin and angiotensinogen. The presence of an MboIsite in an RFLP in the renin gene was identi®ed as arisk factor for hypertension in the United ArabEmirates [Frossard et al., 1998] and with a familyhistory of hypertension in Japan [Okura et al., 1993]. Anumber of different studies have implicated thepresence of Thr at the Met/Thr polymorphism at codon235 (M235T) of the angiotensinogen gene with hyper-tension and preeclampsia and this polymorphism hasbeen associated with variation in plasma angiotensino-gen levels [Berg and Berg, 1998]. Thus, we studiedthese two variants in our English autopsy-con®rmedAD cases and controls and then tried to replicate our®ndings in a French population composed of clinicallydiagnosed AD cases and controls.

MATERIALS AND METHODS

English Samples

Anonymized cases (n� 239) with autopsy-con®rmedAD with an onset after 65 years (using CERAD criteria)were obtained from Brain Banks in Cambridge, Oxford,and London. Cases were of English Caucasian originand comprised 86 males and 153 females; mean age atdeath was 81.2�7.8. Our English Caucasian nonde-mented controls aged 73 years and older with MiniMental State Examination (MMSE) scores of 24 ormore were collected around Oxford and Cambridge(n�342) as part of ongoing community-based studies:the MRC Multicentre Study of Cognitive Function andAgeing and Cambridge City Study [MRC CFAS, 1999;Paykel et al., 1994]. Controls comprised 140 males and202 females and had a mean age of 82.1�3.8 years.There was no signi®cant difference in the sex distribu-tions in the two groups, or the age at death of the casescompared to the age of examination of the controls.These studies have been approved by the Adden-brooke's Hospital local ethics committee. The genotypesat the apoE locus were reported previously [Narainet al., 2000] and the renin and angiotensinogenpolymorphisms were typed essentially as described[Berg and Berg, 1998; Frossard et al., 1998]. Thenumbers in the tables re¯ect the samples that weresuccessfully genotyped for the locus of interest andAPOE.

French Samples

The M235T angiotensinogen polymorphism wasanalyzed in 239 French AD patients composed of 86males and 149 females and 244 French controlscomposed of 80 males and 164 females. Genderinformation was missing from four cases. A diagnosis

of probable AD was made according to DSM-III-R andNINCDS-ADRDA criteria. The age at onset of the caseswas 73� 8.6 years. All controls did not meet DSM-III-Rcriteria for dementia and had normal cognitive function(MMSE >27; age, 73�8.5 years). Informed consentwas obtained from patients and relatives. A fragmentof the angiotensinogen gene containing the M235Tpolymorphism was ampli®ed using the primers 50-GATGCGCACAAGGTCCTGTC-30 and 50-GCCAGCA-GAGAGGTTTGCCT-30. PCR was performed in a ®nalvolume of 50 ml containing 40 ng genomic DNA, 0.75 mMof each primer, 1 mM MgCl2, 0.2 mM of each dNTP, and2 U Taq polymerase. The cycling conditions were 948Cfor 5 min, 35 cycles of 948C for 30 sec, 548C for 30 sec,728C for 30 sec, and a ®nal extension of 5 min. AfterPCR ampli®cation, allele-speci®c hybridization (ASO)was performed to identify the genotypes using theprimers 50-TCCCTGATGGGAGCC-30 (T-allele corre-sponding to a Thr) and 50-TCCCTGACGGGAGCC (C-allele corresponding to a Met), which were 50-labeledwith DIG (Boehringer Mannheim) and hybridized at47.58C (C-allele) and 458C (T-allele). The genotypes atthe apoE locus were reported previously [Lambert et al.,1998].

RESULTS

The genotype frequencies of the M235T angiotensi-nogen polymorphism appeared to be different in theEnglish AD cases and controls (chi-square2�7.66,P� 0.02; Table I). The TT genotype appeared to beassociated with increased AD risk compared to the MMhomozygotes (OR�1.72; 95% CI 1.01±2.95). However,the allele frequencies at this locus did not differ in ADcases and controls. No signi®cant effects emerged whenthe data were adjusted for age, sex, or the presence ofthe apoE E4 allele (P> 0.05 Mantel-Haenszel test).

The effect we had observed was of marginal sig-ni®cance and the genotype frequencies of the controlgroup were marginally out of Hardy-Weinberg equili-brium (chi-square1�5.3; P<0.05). Accordingly, wesought to replicate this ®nding in an independentseries of 239 French AD patients and 244 Frenchcontrols. In this sample, no signi®cant differences wereobserved in genotype or allele frequencies at this locusin the cases and controls (Table II). No signi®canteffects emerged when the data were adjusted for age,sex, or the presence of the apoE E4 allele (P> 0.05

TABLE I. Allele and Genotype Frequencies of AngiotensinogenM235T Polymorphism in English AD Cases and Controls*

AD Controls Odds ratio 95% CI

AllelesM 229 (0.55) 348 (0.60) 1.00T 185 (0.45) 232 (0.40) 1.21 0.94±1.56

GenotypesMM 67 (0.32) 95 (0.33) 1.00MT 95 (0.46) 158 (0.54) 0.85 0.57±1.28TT 45 (0.22) 37 (0.13) 1.72 1.01±2.95

*Odds ratios and 95% con®dence intervals are shown. Alleles: chi-square1�2.18, P� 0.14; genotypes: chi-square2�7.66, P� 0.02.

762 Taylor et al.

Mantel-Haenszel test). The genotype frequencies of theFrench control group did not deviate from Hardy-Weinberg expectations (chi-square1�0.8; P>0.05).

The MboI renin polymorphism was typed in ourEnglish AD cases and controls (Table III). We detectedno signi®cant differences in allele or genotype frequen-cies between the two groups. Furthermore, no signi®-cant effects emerged when the data were adjusted forage, sex, or the presence of the apoE E4 allele (P> 0.05Mantel-Haenszel test). The genotype frequencies of thepooled cases and controls did not deviate from Hardy-Weinberg expectations (chi-square1�3.83; P>0.05).

DISCUSSION

Our initial observation in a U.K. sample suggested amoderate effect of the M235T polymorphism on the riskof developing AD. However, this ®nding of an associa-tion of the angiotensin TT genotype with AD risk wasnot replicated in a French sample. Even if the twopopulations studied are different regarding diagnosticcriteria [autopsy con®rmed (U.K.) vs. probable AD(French)], it is likely that our initial observation was atype I error. Since many loci have been and will betyped for AD associations, P values of much less than0.05 are required in order to be con®dent of a signi®cant®nding. Because of this multiple testing issue, it may beadvisable to con®rm initial observations of signi®cantassociations in independent samples prior to publica-tion. This may limit the numbers of false positiveassociations reported. No association was observed atthe renin locus.

Our sample has 95% power to detect loci with oddsratios of 2 or more (5% signi®cance) assuming the allele

frequencies we have observed in our control popula-tions for these genes [Breslow and Day, 1980]. Hence, itis unlikely that we have missed major effects at thepolymorphic loci we examined. While our data suggestthat the genes studied have no major effect on AD risk,it is possible that there are other polymorphisms inthese genes that do in¯uence AD risk. However, thepolymorphisms we studied were selected on the basis ofreported associations with hypertension risk.

The association between hypertension and AD risk isintriguing and may be due to common underlyinggenetic effects. In this study, we have tested candidatesthat may be functionally involved in hypertension riskin AD. However, the reverse situation may also needconsideration; it is possible that genes and pathwaysinvolved in AD risk may be relevant to hypertensionrisk.

ACKNOWLEDGMENTS

We are very grateful to Professor M. Esiri forproviding cases with autopsy-con®rmed AD. A.Y. isgrateful to the Cambridge Overseas Trust and theEuropean Dana Alliance for the Brain and J.G.E. isgrateful to the Clothworkers' Foundation for funding.M.E. and L.G. are recipients of grants from INSERM(Poste Vert) and Institut Pasteur-Conseil ReÂgionalNord Pas de Calais, respectively. D.C.R. is a GlaxoWellcome Research Fellow.

REFERENCES

Berge KE, Berg K. 1998. Polymorphisms at the angiotensinogen (AGT) andangiotensin II type 1 receptor (AT1R) loci and normal blood pressure.Clin Genet 53:214±219.

Breslow NE, Day NE. 1980. Statistical methods in core research, vol. 1, thedesign and analysis of case-control studies. Lyon: IARC Scienti®cPublications.

Farrer LA, Cupples LA, Haines JL, Hyman B, Kukull WA, Mayeux R,Myers RH, Pericak-Vance MA, Risch N, van Duijn CM. 1997. Effects ofage, sex, and ethnicity on the association between apolipoprotein Egenotype and Alzheimer disease: a meta-analysisÐAPOE and Alzhei-mer Disease Meta Analysis Consortium. JAMA 278:1349±1356.

Frossard PM, Lestringant G, Elsahat YI, John A, Obineche E. 1998. AnMboI two-allele polymorphism may implicate the human renin gene inprimary hypertension. Hypertens Res 21:221±225.

Ganong WF. 1981. Review of medical physiology, 10th ed. Los Alton: LangeMedical Publications. p 364±365.

Gattu M, Terry AV Jr, Pauly JR, Buccafusco JJ. 1997. Cognitiveimpairment in spontaneously hypertensive rats: role of central nicotinicreceptorsÐpart II. Brain Res 771:104±114.

Kehoe PG, Russ C, McIlory S, Williams H, Holmans P, Holmes C, LiolitsaD, Vahidassr D, Powell J, McGleenon B, Liddell M, Plomin R, Dynan K,Williams N, Neal J, Cairns NJ, Wilcock G, Passmore P, Lovestone S,Williams J, Owen MJ. 1999. Variation in DCP1, encoding ACE, isassociated with susceptibility to Alzheimer disease. Nat Genet 21:71±72.

Lambert JC, Pasquier F, Cottel D, Frigard B, Amouyel P, Chartier-HarlinMC. 1998. A new polymorphism in the APOE promoter associatedwith risk of developing Alzheimer's disease. Hum Mol Genet 7:533±540.

Lis CG, Gaviria M. 1997. Vascular dementia, hypertension, and the brain.Neurol Res 19:471±480.

MRC CFAS. 1999. Cognitive function and dementia in six areas of Englandand Wales: the distribution of MMSE and prevalence of GMS organicitylevel in the MRC CFA study. Psychol Med 28:319±335.

Narain Y, Yip A, Murphy T, Brayne C, Easton D, Grimley Evans J,Xuereb J, Cairns N, Esiri MM, Furlong RA, Rubinsztein DC. 2000. The

TABLE III. Allele and Genotype Frequencies of Renin MboIPolymorphism in English AD Cases and Controls*

AD Controls Odds ratio 95% CI

Alleles1 315 (0.72) 468 (0.74) 1.002 125 (0.28) 166 (0.26) 1.12 0.85±1.47

Genotypes11 108 (0.49) 165 (0.52) 1.0012 99 (0.45) 138 (0.44) 1.10 0.77±1.5622 13 (0.06) 14 (0.04) 1.42 0.64±3.14

*Odds ratios and 95% con®dence intervals are shown. Alleles: chi-square1� 0.65, P� 0.42; genotypes: chi-square2� 0.86, P�0.65.

TABLE II. Allele and Genotype Frequencies of AngiotensinogenM235T Polymorphism in French AD Cases and Controls*

AD Controls Odds ratio 95% CI

AllelesM 278 (0.58) 289 (0.59) 1.00T 200 (0.42) 199 (0.41) 1.04 0.81±1.35

GenotypesMM 81 (0.34) 89 (0.36) 1.00MT 116 (0.49) 111 (0.45) 1.15 0.77±1.71TT 42 (0.18) 44 (0.18) 1.05 0.62±1.76

*Odds ratios and 95% con®dence intervals are shown. Alleles: chi-square1� 0.11, P� 0.74; genotypes: chi-square2� 0.48, P�0.79.

Alzheimer Disease 763

ACE gene and Alzheimer disease susceptibility. J Med Genet 37:695±669.

Okura T, Kitami Y, Hiwada K. 1993. Restriction fragment lengthpolymorphisms of the human renin gene: association study with afamily history of essential hypertension. J Hum Hypertens 7:457±461.

Paykel ES, Brayne C, Huppert FA, Gill C, Barkley C, Gelhaar E, BeardsallL, Girling DM, Pollitt P, O'Connor D. 1994. Incidence of dementia in apopulation older than 75 years in the United Kingdom. Arch GenPsychiatry 51:325±332.

Rigat B, Hubert C, Alhenc-Gelas F, Cambien F, Corvol P, Soubrier F. 1990.An insertion-deletion polymorphism in the angiotensin I convertingenzyme gene accounting for half the variance in serum enzyme levels. JClin Invest 86:1343±1346.

Rubinsztein DC, Easton DF. 1999. Apolipoprotein E genetic variation andAlzheimer's disease: a meta-analysis. Dement Geriatr Cogn Disord10:199±209.

Skoog I. 1997. The relationship between blood pressure and dementia: areview. Biomed Pharmacother 51:367±375.

764 Taylor et al.