Alzheimers disease is not associated with the hypertension genetic risk factors PLA2 or G protein 3 either independently or interactively with apolipoprotein eкод для вставкиСкачать
American Journal of Medical Genetics (Neuropsychiatric Genetics) 88:465–468 (1999) Alzheimers Disease is not Associated With the Hypertension Genetic Risk Factors PLA2 or G Protein ␤3, Either Independently or Interactively With Apolipoprotein E Terrence Town,1* Daniel Paris,1 Timothy A. Parker,1 Amy Kundtz,1 Jun Tan,1 Ranjan Duara,2 Michael Gold,3 Fiona Crawford,1 and Michael Mullan1 1 The Roskamp Institute, Department of Psychiatry, University of South Florida, Tampa, Florida Mount Sinai Memory Disorder Clinic, University of Miami Medical School, Miami, Florida 3 Department of Neurology, University of South Florida, Tampa, Florida 2 Growing evidence suggests that hypertension and Alzheimers disease (AD) may share a common etiology. To evaluate the contribution to AD of genetic factors associated with hypertension, we genotyped clinic and community-based AD cases and controls for polymorphisms within the pancreatic PLA2 gene and the G protein ␤3 subunit gene, both of which are located on chromosome 12. Our results do not support an independent association between either of these genes and AD. We further assessed the possibility that either of these genes may interact with the apolipoprotein E gene, a known risk factor for hypertension and AD, on predicting AD. We were unable to find statistical interaction between either the pancreatic PLA2 or G␤3 genes and the apolipoprotein E gene on risk for AD. These results do not support a shared genetic etiology between hypertension and AD. Possibly, a clinical association between these diseases could be due to pathophysiologic interactions. Am. J. Med. Genet. (Neuropsychiatr. Genet.) 88:465–468, 1999. © 1999 Wiley-Liss, Inc. KEY WORDS: Alzheimers disease; hypertension; phospholipase A2; G protein; chromosome 12 Contract grant sponsor: National Institutes of Health; Contract grant number: AG12412-02. *Correspondence to: Terrence Town, Roskamp Institute, University of South Florida, 3515 E. Fletcher Ave., Tampa, FL 33613. E-mail: firstname.lastname@example.org Received 19 June 1998; Accepted 30 December 1998 © 1999 Wiley-Liss, Inc. INTRODUCTION Occurrence of hypertension (HTN) has been suggested to be directly associated with risk for vascular dementia and Alzheimers disease (AD), possibly through induction of cerebral infarcts, ischaemic subcortical white matter lesions, and blood-brain barrier dysfunction [Lis and Gaviria, 1997; Skoog, 1997]. Furthermore, in an animal model of HTN, spontaneously hypertensive rats were shown to develop cognitive and learning abnormalities when compared with normotensive rats, suggesting that HTN may contribute to the cognitive impairment associated with AD [Gattu et al., 1997]. The ⑀4 allele of the apolipoprotein E gene (APOE) has been consistently implicated as a risk factor for AD and as a predictor of HTN [Pasquier and Leys, 1997; Sparks, 1997]. Based on such evidence for a shared etiology between AD and HTN, we investigated a possible genetic association between AD and two genes on chromosome 12 previously found to be associated with HTN: pancreatic PLA2 and G protein ␤3 subunit (G␤3). We further assessed a possible interaction between polymorphisms within either of these genes and APOE on risk for AD. MATERIALS AND METHODS Study Subjects We genotyped community-based control subjects, community-based AD patients, and clinic-based AD patients. Community-based AD subjects and controls were recruited as part of a community-based screen for dementia in the elderly (age 60 or older) in the Miami area of Florida. Subjects who met criteria for healthy elderly (must score >26 on the Mini Mental State Examination [MMSE], achieve >10 accurate recalls of three words in four separate trials, generate >30 category-specific words in 3 min, and achieve a normal score on the Hamilton depression scale) were included in the control group. Subjects who did not meet one or 466 Town et al. more of these criteria were asked to undergo further examination at the Mount Sinai Memory Disorder Clinic. Individuals who returned for a complete examination and were diagnosed normal were also included in the control set; those diagnosed with AD were included in the community-based AD sample; and those with other dementia diagnoses or pending further investigation were not included in this study. Clinicbased AD cases were drawn from a sample in which patients were recruited from over 40 clinical centers throughout the United States. All AD subjects met criteria for probable or possible AD (according to National Institute of Neurological and Communicative Disorders and Stroke-Alzheimers Disease and Related Disorders Association criteria). Peripheral blood samples were collected from study subjects, and DNA was extracted from whole blood using the PureGene娂 kit (Gentra Systems) or by standard phenol-chloroform methods. We tested the hypothesis that the clinic-based sample and the community-based sample would differ in their genetic profiles for hypertension. This expectation arose because the clinic-based AD sample had been selected for a therapeutic trial in which HTN/ vascular disease was a strict exclusion criterion whereas in the community-based sample hypertensives were not excluded if this systemic disorder could not explain the dementia. We did not find a significant difference (by genotype, p ⳱ .781; by allele, p ⳱ .453) in the G␤3 genetic profiles for hypertension between the clinic- and community-based samples. Therefore, as there are no known phenotypic differences between G␤3 and PLA2-associated HTN, we focused on the less biased community-based AD sample in our PLA2 study. Control subjects (n ⳱ 167, mean age ⳱ 71.81 ± 7.37, 59.0% female) and community-based AD cases (n ⳱ 77, mean age ⳱ 75.72 ± 7.58, 53.7% female) included in the PLA2 genotyping strategy did not differ by gender (p ⳱ .468), but control subjects were significantly younger than cases (p < .001). For G␤3 genotyping, control subjects (n ⳱ 181, mean age ⳱ 71.02 ± 8.93, 71.7% female), community-based AD cases (n ⳱ 79, mean age ⳱ 74.96 ± 8.43, 50.7% female), and clinic-based AD cases (n ⳱ 70, mean age ⳱ 74.54 ± 8.33, 76.2% female) were included. Clinic-based AD cases and controls had similar gender frequencies (p ⳱ .503), while community-based AD samples and controls did not (p < .01), yet the gender comparison of the combined AD group to controls was not significantly different (p ⳱ .057). Both AD groups were significantly older than controls (p < .01). All samples (cases and controls) were drawn from primarily Caucasian and Hispanic populations and no significant differences (p > .05) were noted between Caucasians and Hispanics within each sample on G␤3 or PLA2 status. APOE and G␤3 Polymorphism Genotyping APOE genotyping was performed according to the methods of Wenham and colleagues . G␤3 genotyping was performed as follows: approximately 200 ng of genomic DNA was used in a total volume of 25 L for a polymerase chain reaction (PCR) designed to amplify a 268 bp fragment containing a coding polymorphism [C825T, Siffert et al., 1998] within exon 10 of the heterotrimeric G protein ␤3 subunit gene. Forward (5⬘TGACCCACTTGCCACCCGTGC-3⬘) and reverse (5⬘GCAGCAGCCAGGGCTGGC-3⬘) oligonucleotides were used as reported by Siffert and colleagues . Thermocycling was performed for 35 cycles at 94°C for 1 min, 66°C for 40 sec, and 72°C for 1 min, and a final extension step at 72°C for 10 min. PCR products were digested with 1.25 U of BsaJ I (NEB, Beverly, MA), separated on 2.0% agarose gels, and viewed by UV transillumination. The uncut PCR product corresponds to the T/T genotype while complete digestion results in fragments of 116 bp and 152 bp (C/C genotype). Pancreatic PLA2 Polymorphism Genotyping PCR was designed to amplify a 430-bp fragment containing a Taq I dimorphic site within the first intron of the pancreatic PLA2 gene. This Taq I dimorphism was previously identified by Frossard and Lestringant using a southern blot technique . Because of the difficulty in amplifying this fragment directly from some samples, a nested PCR strategy was used. Forward (5⬘-GAGTACAGTGGTGCGATCTC-3⬘) and reverse (5⬘-CCATTCTCCTGCCTCAGCCT-3⬘) oligonucleotides were used with ∼200 ng of genomic DNA in a TABLE I. G␤3 Polymorphism Frequencies* Clinicbased AD C/C 29 (41.4%) T/C 33 (47.1%) T/T 8 (11.4%) n ⳱ 70 2 ⳱ .385,** p ⳱ .825 91 (65.0%) C T 49 (35.0%) n ⳱ 140 2 ⳱ .230,** p ⳱ .632 ORs with 95% CIs Communitybased AD ORs with 95% CIs Combined AD ORs with 95% CIs 1.52 [.51–4.57] 1.17 [.39–3.54] 1 35 (44.3%) 1.11 [.46–2.69] 1.07 [.44–2.55] 1 64 (43.0%) 1.24 [.58–2.67] 1.09 [.51–2.33] 1 72 (39.8%) 1.12 [.78–1.61] 1 227 (62.7%) 1.26 [.76–2.09] 1 33 (41.8%) 11 (13.9%) n ⳱ 79 2 ⳱ .486,** p ⳱ .784 103 (65.2%) 55 (34.8%) n ⳱ 158 2 ⳱ .294,** p ⳱ .588 1.05 [.69–1.60] 1 *Control ages and AD ages of onset ⱖ 60 years; OR ⳱ odds ratio; CI ⳱ confidence interval. **Comparison made to control. 66 (44.3%) 19 (12.8%) n ⳱ 149 2 ⳱ .400,** p ⳱ .819 194 (65.1%) 104 (34.9%) n ⳱ 298 2 ⳱ .406,** p ⳱ .524 Controls 83 (45.9%) 26 (14.4%) n ⳱ 181 135 (37.3%) n ⳱ 362 PLA2, G␤3 Genes, and Alzheimer’s Disease TABLE II. PLA2 Polymorphism Frequencies* Communitybased AD A1/A1 A1/A2 A2/A2 13 (16.9%) 36 (46.8%) 28 (36.4%) n ⳱ 77 A1 A2 62 (40.3%) 92 (59.7%) n ⳱ 154 Controls 36 (21.6%) 87 (52.1%) 44 (26.3%) n ⳱ 167 2 ⳱ 2.622, p ⳱ .270 159 (47.6%) 175 (52.4%) n ⳱ 334 2 ⳱ 2.306, p ⳱ .129 ORs with 95% CIs 1 1.15 [.52–2.55] 1.46 [.62–3.44] 1 1.21 [.79–1.85] *Control ages and AD ages of onset ⱖ60 years; OR ⳱ odds ratio; CI ⳱ confidence interval. total volume of 25 L. Thermocycler conditions were as follows: 94°C for 10 min followed by 35 cycles of 94°C for 30 sec, 60°C for 30 sec, and 72°C for 20 sec, and a final extension step at 72°C for 10 min. One microliter of this product was then used in a 25 L PCR with forward (5⬘-AGTGGTGCGATCTCGGCTCA-3⬘) and reverse (same as above) oligonucleotides to yield the 430 bp fragment. Thermocycler conditions were as follows: 94°C for 10 min followed by 30 cycles of 94°C for 30 sec, 63°C for 30 sec, and 72°C for 30 sec, and a final extension step at 72°C for 10 min. PCR products were digested with 1.5 U of Taq I (Promega), separated on 2.0% agarose gels, and viewed by UV transillumination. The uncut PCR product corresponds to the A1/A1 genotype while complete digestion results in fragments of 263 bp and 167 bp (A2/A2 gentotype). Statistical Analysis The likelihood ratio 2 statistic was used to compare genotype and allelic distributions, as well as gender distributions, between control subjects and AD cases. A t-test for independent samples was used to compare control ages with AD ages of onset. Multiple logistic regression models were used to assess odds ratios (ORs) for each genotype or allele, as well as to assess possible interaction between either the PLA2 polymorphism or the G␤3 polymorphism and APOE genotype. Age and gender were included in each logistic regression model as controlling factors, and while age was a significant (p < .05) term across all analyses, gender was not (p > .05). Odds ratios and corresponding 95% confidence intervals (CIs) were calculated according to standard methods. Alpha levels were set at .05 for each analysis. Each of these analyses was performed using SPSS for Windows, release 7.5.1. Power analyses were carried out using Power and Precision, release 1.0. RESULTS As shown in Table I, we did not detect an association between the G␤3 polymorphism and AD, either by genotype or by allele. We did find the expected APOE ⑀4 elevation when comparing each of the AD groups shown in Table I with controls, both by allele (p < .001) and by genotype (p < .01, data not shown). Possible interaction between APOE and the G␤3 polymorphism on prediction of AD was assessed, and no significant 467 interactive term was found when including clinic-based AD patients (p ⳱ .974), community-based AD cases (p ⳱ .904), or the combined group (p ⳱ .948) in the model. We went on to assess a possible association between the PLA2 polymorphism and community-based AD cases, and were unable to find such an association (see Table II). We did find the expected APOE ⑀4 increase, again, when comparing the groups shown in Table II (by allele, p < .001; by genotype, p ⳱ .001, data not shown). We were unable to find interaction between APOE and the PLA2 polymorphism on risk for AD (p ⳱ .833). DISCUSSION Occurrence of HTN has been suggested to be directly associated with risk for vascular dementia and AD, though the mechanism underlying such an association remains unclear [Skoog, 1996; Lis and Gaviria, 1997]. We investigated a possible genetic etiologic link between HTN and AD, and genotyped AD cases and controls for polymorphisms within the pancreatic PLA2 and G␤3 genes. Both of these genes had previously been found to be strongly associated with HTN, and the G␤3 polymorphism was further shown to differentially effect G protein activity [Frossard and Lestringant, 1995; Siffert et al., 1998]. Additionally, both of these polymorphisms map to chromosome 12, where other groups have found linkage to AD [Pericak-Vance et al., 1997; Rogaeva et al., 1998]. We did not detect an association between either of these polymorphisms and AD, either independently or synergistically with APOE. Thus, our results do not support a genetic etiologic link between HTN and AD. Statistical power to detect a clinically relevant difference (a disease allele difference of 12% between AD cases and controls, on the order of that observed with APOE ⑀4) between either the G␤3 or PLA2 polymorphisms and AD was calculated; our G␤3 and PLA2 polymorphism data sets reveal 91 and 71% power, respectively. Further, power to detect a significant interaction (where each term explains 5% of disease variance in a multiple regression model) between either the G␤3 or PLA2 polymorphisms and APOE on prediction of AD was calculated, and these analyses reveal 99 and 96% power, respectively. Thus, overall likelihood of a type II error in detecting clinically relevant differences in genetic profiles between cases and controls in our data set is minimal. Yet, our apparent lack of association between HTN genetic risk factors and AD may be due to a relatively low information content of the polymorphisms examined in this study. Additionally, it is possible that, by examining an AD sample, we have selected a specific subset of hypertensive individuals who are etiologically distinct from hypertensive cases at large. To examine this hypothesis, the study of further candidate genes for HTN in AD is warranted. If the results with these HTN genetic risk factors are indicative of no genetic etiologic link between HTN and AD, it is possible that the physiologic consequences of HTN themselves contribute to AD. Alternatively, aspects of the AD process itself (such as the possible consequences of ␤-amyloidenhanced vasoconstriction) could lead to HTN. 468 Town et al. ACKNOWLEDGMENTS We are grateful to Mr. and Mrs. Robert Roskamp for their generous support which helped to make this work possible. We would like to thank Danielle Fallin for her assistance in G␤3 genotyping. REFERENCES Frossard P, Lestringant G. 1995. Association between a dimorphic site on chromosome 12 and clinical diagnosis of hypertension in three independent populations. Clin Genet 48:284–287. Lis CG, Gaviria M. 1997. Vascular dementia, hypertension, and the brain. Neurol Res 19:471–480. Gattu M, Terry AV Jr, Pauly JR, Buccafusco JJ. 1997. Cognitive impairment in spontaneously hypertensive rats: role of central nicotinic receptors. Part II. Brain Res 771:104–114. Pasquier F, Leys D. 1997. Why are stroke patients prone to develop dementia? J Neurol 244:135–142. Pericak-Vance MA, Bass MP, Yamaoka LH, Gaskell PC, Scott WK, Terwedow HA, Menold MM, Conneally PM, Small GW, Vance JM, Saunders AM, Roses AD, Haines JL. 1997. Complete genomic screen in late-onset familial Alzheimer’s disease. Evidence for a new locus on chromosome 12. J Am Med Assoc 278:1237–1241. Rogaeva E, Premkumar S, Song Y, Sorbi S, Brindle N, Paterson A, Duara R, Levesque G, Yu G, Nishimura M, Ikeda M, O’Toole C, Kawarai T, Jorge R, Vilarino D, Bruni AC, Farrer LA, St George-Hyslop PH. 1998. Evidence for an Alzheimer disease susceptibility locus on chromosome 12 and for further locus heterogeneity. J Am Med Assoc 280:614–618. Siffert W, Rosskoph D, Siffert G, Busch S, Moritz A, Erbel R, Sharma A, Ritz E, Wichmann H, Jakobs K, Horsthemke B. 1998. Association of a human G-protein ␤3 subunit variant with hypertension. Nat Genet 18:45–48. Skoog I. 1997. The relationship between blood pressure and dementia: a review. Biomed Pharmacother 51:367–375. Sparks DL. 1997. Coronary artery disease, hypertension, ApoE, and cholesterol: a link to Alzheimer’s disease? Ann NY Acad Sci 826:128–146. Whenham PR, Price WH, Blundell G. 1991. Apolipoprotein E genotyping by one-stage PCR. Lancet 337:1158–1159.