lable at ScienceDirect

Neurobiology of Aging 35 (2014) 727.e1e727.e3

Contents lists avai

Neurobiology of Aging

journal homepage: www.elsevier .com/locate/neuaging

Brief communication

Rare and common variants in the Apolipoprotein E gene in healthy oldest old

Lauren C. Tindale a,b, Stephen Leach a, Kevin Ushey c, Denise Daley c, Angela R. Brooks-Wilson a,b,*

aCanada’s Michael Smith Genome Sciences Centre, BC Cancer Agency, Vancouver, British Columbia, CanadabDepartment of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, British Columbia, CanadacUniversity of British Columbia James Hogg Research Centre, Institute for Heart and Lung Health, Vancouver, British Columbia, Canada

a r t i c l e i n f o

Article history:Received 11 July 2013Received in revised form 4 September 2013Accepted 6 September 2013Available online 11 October 2013

Keywords:APOEApolipoprotein EHealthy agingOldest oldLongevity

* Corresponding author at: Genome Sciences Center10th Avenue, Vancouver, BC V5Z 1L3, Canada. Tel.: þ1675 8178.

E-mail address: abrooks-wilson@bcgsc.ca (A.R. Bro

0197-4580/$ e see front matter � 2014 Elsevier Inc. Ahttp://dx.doi.org/10.1016/j.neurobiolaging.2013.09.010

a b s t r a c t

Apolipoprotein E (APOE) alleles are associated with longevity in genome-wide scans, with ε4 correlatedwith shorter life, and ε2 with longer life, than ε3. We hypothesized that rare APOE variants with largeindividual effects might also contribute to long-term good health. The APOE exons and promoter wereresequenced in DNA samples from 376 healthy oldest old aged �85yrs with no self-reported history ofcancer, cardiovascular disease, diabetes, major pulmonary disease or Alzheimer disease (“Super-Seniors”)and 376 population-based controls aged 41e54. Forty variants were observed: 32 were rare (minor allelefrequency <2%); 9 were nonsynonymous. Controls were more likely to have an ε4 allele (Pearson c2 ¼6.61, p ¼ 0.04). Among the Super-Seniors, APOE allele status was not associated with body mass index orMini Mental State Examination score. There was no excess of rare APOE variants in healthy oldest oldcompared with midlife controls, or vice-versa; however, this does not rule out an effect of some vari-ants on ApoE function. Our findings were consistent with ε4 being a risk factor for early mortality.

� 2014 Elsevier Inc. All rights reserved.

1. Introduction

Apolipoprotein E (APOE) is a significant association in genome-wide association scans of longevity (Deelen et al., 2011; Nebelet al., 2011; Sebastiani et al., 2012). The ApoE isoforms are risk fac-tors for cardiovascular disease and Alzheimer disease likely becauseof their involvement in elevated lipid levels, oxidative stress, andinflammation (Huebbe et al., 2011). Two single nucleotide poly-morphisms (SNPs), rs429358 and rs7412, and their respective dif-ferences at amino acids 112 and 158, define the 3 major isoforms ofApoE: ε2 (cys112, cys158), ε3 (cys112, arg158), and ε4 (arg112,arg158). Genotypes at rs429358 and rs7412 also define the dip-lotypes ε2/ε2, ε2/ε3, ε2/ε4, ε3/ε3, ε3/ε4, and ε4/ε4. Although notconsistently, the APOE ε4 allele has been associated with earlymortality, whereas the ε2 allele has been associated with longevity(Frisoni et al., 2001; Schupf et al., 2012; Schächter et al., 1994). The 3major ApoE isoforms, ApoE2, ApoE3, and ApoE4, are determined bythe genotype at rs429358 and rs7412. The literature is divided onwhether variants in the 50 upstream region and first intron have aneffect on disease risk after considering linkage disequilibrium withrs429358 and rs7412 (Belbin et al., 2007;Maloneyet al., 2009; Parra-Bonilla et al., 2003; Yu et al., 2007). In addition to common variants,

, BC Cancer Agency, 675 West604 675 8153; fax: þ1 604

oks-Wilson).

ll rights reserved.

rare variants may also play an important role in susceptibility todisease (Cohen et al., 2004). For instance, rare alleles have beenfound to contribute to low plasma high-density lipoproteincholesterol levels (Cohen et al., 2004), and a significantly highernumber of rare missense or nonsense variants were observed in 4genes associated with hypertriglyceridemia (Johansen et al., 2010).

In our study of healthy aging, we compared “Super-Seniors”whowere older than 85 years and self-reported never having had can-cer, cardiovascular disease, diabetes, major pulmonary disease orAlzheimer disease, with population-based midlife controls(Halaschek-Wiener et al., 2008, 2009). We hypothesized that rarevariants in APOE, in addition to the more common known variants,may contribute to longevity and healthy aging. We resequenced thepromoter region and exons of APOE to: (1) test if there was anexcess of rare variants in either Super-Seniors or controls; (2)confirm an inverse association of healthy aging with ε4; and (3) testif bodymass index (BMI) or Mini Mental State Examination (MMSE)score was correlated with APOE status in Super-Seniors.

2. Methods

2.1. Study population

Research ethics board approval was received from the jointClinical Research Ethics Board of the BC Cancer Agency and theUniversity of British Columbia. All subjects gave written informedconsent. Recruitment and characterization of subjects were

L.C. Tindale et al. / Neurobiology of Aging 35 (2014) 727.e1e727.e3727.e2

described previously (Halaschek-Wiener et al., 2008, 2009). For thecurrent analysis, 376 Super-Senior cases, 32%males and 68% females,aged 85e108 years (89.1 years [SD 3.1]) and 376 controls, 39% malesand 61% females, aged 41e54 years (47.4 years [SD 2.9]) of Europeanancestry were selected randomly to fit (along with technical con-trols) on two 384-well sequencing plates. Weight was not reportedfor 3 Super-Seniors and 8 controls. Twenty-five incomplete MMSEscores (due to vision or language deficits, or unwillingness to com-plete the examination) were excluded.

2.2. Resequencing and functional predictions

All 4 exons and 622 base pairs (bp) upstream of the APOE genewere amplified in 6 amplicons using PCR, and the primer pairs inSupplementary Table S1, and sequenced using the Sanger methodas previously described (Brooks-Wilson et al., 2004) (also seeSupplementaryMethods). Sequence readswere aligned to GenBankreference files and genetic variants scored using Mutation Surveyor(DNA Variant Analysis Software v3.0). Non-synonymous variantswere input into SIFT (Ng and Henikoff, 2001) and PolyPhen-2(Adzhubei et al., 2010) to predict their functional effects.

2.3. Statistical analysis

Statistics was done using JMP 10 and R 2.15.3. c2 and Fisher exacttests were done on common variants for genotype and ε2, ε3, ε4allele frequency, and on rare variant counts between Super-Seniorsand controls. Analyses of variance were run for BMI and MMSEaccording to ε2, ε3, ε4 allele frequency.

3. Results

Forty variants were observed (39 SNPs and one 12-bp deletion;Fig.1), 23 of whichwere not listed in dbSNP 137. Flanking sequencesare shown in Supplementary Table S2. There were 17 rare variants(minor allele frequency [MAF] <2%) found exclusively in Super-Seniors and 11 rare variants exclusively in controls, plus 4 rare var-iants that were seen in both groups (Supplementary Table S3).

Fig. 1. Distribution of variants observed across the apolipoprotein E gene. Variants in contro(MAF <2%) are in red; common variants are in black. Brackets above the gene show the locatiDNA sequence), with (þ1) as the start of translation, and major allele greater than minorspecified according to the processed protein sequence. Diagram not to scale.

No subjects had more than 1 rare variant. Among the 32 rare vari-ants, 7 were nonsynonymous and 1 was synonymous. Of the 7nonsynonymous rare variants, 4 were predicted to be damaging byPolyPhen-2, a tool used to predict the possible impact of an aminoacid substitution (Adzhubei et al., 2010). Possibly damaging variantsincluded g.137 T>C (L28P), found in 2 Super-Seniors and 2 controls,andg.805C>G (R251G), found in 1 Super-Senior. Probablydamagingvariants included g.460 C>T (R136C), found in 1 control, andg.886G>T (G278W), found in 1 Super-Senior. Therewas no evidenceof a difference in proportion of rare variants in Super-Seniorscompared with controls using Fisher exact test (p ¼ 0.39).

Of the 8 common variants, 2 were nonsynonymous: rs429358(g.388T>C; C112R) and rs7412 (g.526C>T; R158C); these SNPsdefine the APOE ε2, ε3, and ε4 alleles. The results of Fisher exacttests for all common variants are shown in Supplementary Table S4.Because 8 common variants were tested, we used the Bonferronicorrection to adjust the alpha cutoff to 0.006. rs429358 was theonly common variant that showed marginal evidence of a differ-ence in genotype frequency proportions (p ¼ 0.03) and allele fre-quency proportions (p ¼ 0.01) between Super-Seniors and controls.rs7412 did not reach statistical significance. The set of diplotypes atrs429358 and rs7412 was not significantly different between Super-Seniors and controls (p¼ 0.06; Supplementary Table S5). Therewas,however, evidence of an association in the ε2, ε3, ε4 frequencybetween Super-Seniors and controls (Pearson c2 ¼ 6.61, p ¼ 0.04).Controls were more likely to have an ε4 allele. The ε2 allele wasmore common in Super-Seniors but was not a major contributor tothe overall c2 test statistic score. There were 4 centenarians amongthe Super-Seniors, all of whom had an ε3/ε3 diplotype.

Descriptive statistics for BMI, MMSE, and age are shown inSupplementary Table S6. The mean BMI of controls was 1.6 kg/m2

(SE 0.32) greater than the mean BMI of Super-Seniors (t ratio ¼5.22; p < 0.0001) using a t test (Supplementary Fig. 1S). UsingANOVA, no differences in mean BMI score according to diplotypewere detected in Super-Seniors (F ¼ 0.59, p ¼ 0.71). There was,however, evidence of a difference inmean BMI score by diplotype incontrols (F ¼ 2.58, p ¼ 0.03). Using Tukey’s multiple comparison, adifference in mean BMI of 6.29 kg/m2 (SE 2.03), p ¼ 0.03 could be

ls are listed above the gene and those in cases (Super-Seniors) are below. Rare variantsons of the amplicons sequenced. Variants are named according to location (g ¼ genomicallele as observed in study data. Amino acid numbers of nonsynonymous variants are

L.C. Tindale et al. / Neurobiology of Aging 35 (2014) 727.e1e727.e3 727.e3

detected between only 1 pair of groups, ε2/ε4 and ε4/ε4 controls(Supplementary Table S7). In Super-Seniors, therewas evidence of adifference in mean MMSE score by diplotype (F ¼ 2.45, p ¼ 0.03).Tukey’s multiple comparison showed that a difference of 3.83 (SE1.23), p ¼ 0.03 could be detected between ε2/ε4 and ε4/ε4(Supplementary Table S8). For this analysis, 1 outlier with a lowMMSE score of 16 was excluded because it was an influential point.There were no apparent relationships between MMSE and BMI, orMMSE and age (Supplementary Fig. S2).

4. Discussion

We did not find an excess of rare variants in Super-Seniors orcontrols; however, it is possible that some variants observed mayaffect ApoE function. Four of 7 non-synonymous rare variants werepredicted by PolyPhen-2 (Adzhubei et al., 2010) to be possibly (2) orprobably damaging (2). Because these variants are individually veryrare, they should not be over-interpreted. No common promoterregion variants showed evidence of a difference in distributionbetween Super-Seniors and controls.

There was marginal evidence of a difference in rs429358 pro-portion in genotype and allele frequency between Super-Seniorsand controls. ε2, ε3, ε4 allele frequency showed evidence againstindependence between Super-Seniors and controls, with the majordifference in ε4, but not ε2. Note that rs429358, which determinesthe cys/arg difference at amino acid 112 that differentiates ε3 fromε4, has a much higher MAF (0.156) than rs7412 (MAF, 0.080), whichdetermines the arg/cys difference at residue 158 that differentiatesε2 from ε3. In addition, rs429358 and rs7412 are not in high linkagedisequilibrium with each other (Yu et al., 2007) and would notnecessarily be expected to behave uniformly with regard to theseassociations.

In controls there was evidence of a difference in mean BMI scorebetween ε2/ε4 and ε4/ε4 diplotypes. In Super-Seniors there wasevidence of a difference inmeanMMSE score between ε2/ε4 and ε4/ε4 diplotypes. This could be an artifact of small counts for thesediplotypes, or it could indicate that these 2 “extreme” diplotypesmay have an impact on phenotype. No correlation was found be-tween MMSE and BMI, or MMSE and age in Super-Seniors. Ifinability to finish the MMSE test was related to aging then thiscould have introduced a bias against low scores. Also, if controlswith a higher BMI were more likely to not report their weight, thenthis may have introduced bias against high BMI.

Our findings for ε4 allele frequency support it being a risk factorfor early mortality. ε2 allele frequency was higher in Super-Seniorsbut was not a major contributor to the overall c2 test statistic score.It is possible that significance for some variants was not achieveddue to the control group including individuals who would go on tobecome Super-Seniors. Further study with an increased sample sizewould be needed in order to conclusively test for rare variant as-sociations in APOE. Alternatively, increasing the number of genessequenced to include additional longevity, Alzheimer disease-, orcardiovascular disease-associated genes would provide a morecomprehensive view of the role of rare variants in genes likely toaffect healthy aging.

Disclosure statement

The authors have no actual or potential conflicts of interest.

Acknowledgements

Collection of the Super-Seniors and controls was supported by aNew Emerging Team grant from the Canadian Institutes of HealthResearch.

Appendix A. Supplementary material

Supplementary data associated with this article can be found,in the online version, at http://dx.doi.org/10.1016/j.neurobiolaging.2013.09.010.

References

Adzhubei, I.A., Schmidt, S., Peshkin, L., Ramensky, V.E., Gerasimova, A., Bork, P.,Kondrashov, A.S., Sunyaev, S.R., 2010. PolyPhen-2. Nat. methods 7, 248e249.

Belbin, O., Dunn, J.L., Ling, Y., Morgan, L., Chappell, S., Beaumont, H., Warden, D.,Smith, D.A., Kalsheker, N., Morgan, K., 2007. Regulatory region single nucleotidepolymorphisms of the apolipoprotein E gene and the rate of cognitive decline inAlzheimer’s disease. Hum. mol. genet. 16, 2199e2208.

Brooks-Wilson, A.R., Kaurah, P., Suriano, G., Leach, S., Senz, J., Grehan, N., Butterfield, Y.S.,Jeyes, J., Schinas, J., Bacani, J., Kelsey, M., Ferreira, P., MacGillivray, B., MacLeod, P.,Micek, M., Ford, J., Foulkes, W., Australie, K., Greenberg, C., LaPointe, M., Gilpin, C.,Nikkel, S., Gilchrist, D., Hughes, R., Jackson, C.E., Monaghan, K.G., Oliveira, M.J.,Seruca, R., Gallinger, S., Caldas, C., Huntsman, D., 2004. Germline E-cadherin mu-tations in hereditary diffuse gastric cancer: assessment of 42 new families and re-view of genetic screening criteria. J. Med. Genet. 41, 508e517.

Cohen, J.C., Kiss, R.S., Pertsemlidis, A., Marcel, Y.L., McPherson, R., Hobbs, H.H., 2004.Multiple rare alleles contribute to low plasma levels of HDL cholesterol. Science305, 869e872.

Deelen, J., Beekman, M., Uh, H., Helmer, Q., Kuningas, M., Christiansen, L., Kremer, D.,van der Breggen, R., Suchiman, H.E., Lakenberg, N., van den Akker, E.B.,Passtoors, W.M., Tiemeier, H., van Heemst, D., de Craen, A.J., Rivadeneira, F., deGeus, E.J., Perola, M., van der Ouderaa, F.J., Gunn, D.A., Boomsma, D.I.,Uitterlinden, A.G., Christensen, K., van Duijn, C.M., Heijmans, B.T., Houwing-Duistermaat, J.J., Westendorp, R.G., Slagboom, P.E., 2011. Genome-wide associ-ation study identifies a single major locus contributing to survival into old age;the APOE locus revisited. Aging Cell 10, 686e698.

Frisoni, G.B., Louhija, J., Geroldi, C., Trabucchi, M., 2001. Longevity and the epsilon2allele of apolipoprotein E: the Finnish Centenarians Study. J. Gerontol. A Biol.Sci. Med. Sci. 56A, M75eM78.

Halaschek-Wiener, J., Vulto, I., Fornika, D., Collins, J., Connors, J.M., Le, N.D.,Lansdorp, P.M., Brooks-Wilson, A., 2008. Reduced telomere length variation inhealthy oldest old. Mech. Ageing. Dev. 129, 638e641.

Halaschek-Wiener, J., Amirabbasi-Beik, M., Monfared, N., Pieczyk, M., Sailer, C.,Kollar, A., Thomas, R., Agalaridis, G., Yamada, S., Oliveira, L., Collins, J.,Meneilly, G., Marra, M., Madden, K., Le, N., Connors, J., Brooks-Wilson, A., 2009.Genetic variation in healthy oldest-old. PLoS One 4, e6641.

Huebbe, P., Nebel, A., Siegert, S., Moehring, J., Boesch-Saadatmandi, C., Most, E.,Pallauf, J., Egert, S., Muller, M.J., Schreiber, S., Nothlings, U., Rimbach, G., 2011.APOE epsilon4 is associated with higher vitamin D levels in targeted replace-ment mice and humans. FASEB J. 25, 3262e3270.

Johansen, C.T., Wang, J., Lanktree, M.B., Cao, H., McIntyre, A.D., Ban, M.R.,Martins, R.A., Kennedy, B.A., Hassell, R.G., Visser, M.E., Schwartz, S.M.,Voight, B.F., Elosua, R., Salomaa, V., O’Donnell, C.J., Dallinga-Thie, G.M.,Anand, S.S., Yusuf, S., Huff, M.W., Kathiresan, S., Hegele, R.A., 2010. Excess of rarevariants in genes identified by genome-wide association study of hyper-triglyceridemia. Nat. Genet. 42, 684e687.

Maloney, B., Ge, Y.-W., Petersen, R.C., Hardy, J., Rogers, J.T., Pérez-Tur, J., Lahiri, D.K., 2009.Functional characterization of three single-nucleotide polymorphisms present inthe humanAPOEpromoter sequence: differential effects in neuronal cells and onDNA-protein interactions. Am. J. Med. Genet. B Neuropsychiatr. Genet. 1, 185e201.

Nebel, A., Kleindorp, R., Caliebe, A., Nothnagel, M., Blanche, H., Junge, O., Wittig, M.,Ellinghaus, D., Flachsbart, F., Wichmann, H.E., Meitinger, T., Nikolaus, S.,Franke, A., Krawczak, M., Lathrop, M., Schreiber, S., 2011. A genome-wide as-sociation study confirms APOE as the major gene influencing survival in long-lived individuals. Mech. Ageing. Dev. 132, 324e330.

Ng, P., Henikoff, S., 2001. Predicting Deleterious amino acid Substitutions. GenomeRes. 11, 863e874.

Parra-Bonilla, G., Arboleda, G., Yunis, J., Solano, E., Pardo, R., Arango, G., Hedmont, D.,Arboleda, H., 2003. Haplogroup analysis of the risk associated with APOE pro-moter polymorphisms (�219T/G, �491A/T and �427T/C) in Colombian Alz-heimer’s disease patients. Neurosci. Lett. 349, 159e162.

Schächter, F., Faure-Delanef, L., Guénot, F., Rouger, H., Froguel, P., Lesueur-Ginot, L.,Cohen, D., 1994. Genetic associations with human longevity at the APOE andACE loci. Nat. Genet. 6, 29e32.

Schupf, N., Barral, S., Perls, T., Newman, A., Christensen, K., Thyagarajan, B.,Province, M., Rossi, W.K., Mayeux, R., 2012. Apolipoprotein E and familiallongevity. Neurobiol. Aging 34, 1287e1291.

Sebastiani, P., Solovieff, N., Dewan, A.T., Walsh, K.M., Puca, A., Hartley, S.W.,Melista, E., Andersen, S., Dworkis, D.A., Wilk, J.B., Myers, R.H., Steinberg, M.H.,Montano, M., Baldwin, C.T., Hoh, J., Perls, T.T., 2012. Genetic signatures ofexceptional longevity in humans. PLoS One 7, e29848.

Yu, C.-E., Seltman, H., Peskind, E.R., Galloway, N., Zhou, P.X., Rosenthal, E.,Wijsman, E.M., Tsuang, D.W., Devlin, B., Schellenberg, G.D., 2007. Comprehen-sive analysis of APOE and selected proximate markers for late-onset Alzheimer’sdisease: patterns of linkage disequilibrium and disease/marker association.Genomics 89, 655e665.