- Email: [email protected]
APOE dependent-association of PPAR-γ genetic variants with Alzheimer's disease risk
Neurobiology of Aging 32 (2011) 547.e1–547.e6
APOE dependent-association of PPAR-␥ genetic variants with Alzheimer’s disease risk Onofre Combarros ∗ , Eloy Rodríguez-Rodríguez, Ignacio Mateo, José Luis Vázquez-Higuera, Jon Infante, José Berciano, Pascual Sánchez-Juan Neurology Service and Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), “Marqués de Valdecilla” University Hospital (University of Cantabria), Santander, Spain Received 8 December 2008; received in revised form 11 May 2009; accepted 12 July 2009 Available online 6 August 2009
Abstract The peroxisome proliferator-activated receptor-␥ (PPAR-␥) is a ligand-inducible transcription factor that suppresses microglial inflammatory responses and inhibits amyloid beta (A) production through promoting cholesterol efflux from glial cells. PPAR-␥ agonists have been advanced as a new disease altering approach to Alzheimer’s disease (AD), with rosiglitazone therapy having improved cognition in those AD patients that did not possess an Apolipoprotein E (APOE) 4 allele. The current study was designed to explore the effect of interactions between PPAR-␥ and APOE gene polymorphisms on the AD risk. We examined genetic variations of PPAR-␥ by genotyping 7 haplotype tagging SNPs (htSNPs) (rs10510412, rs17793951, rs1801282, rs4135263, rs1151999, rs709149, and rs709154) in a group of 352 Spanish late-onset AD cases and 438 controls. The PPAR-␥ TCCA haplotype derived from SNPs in introns 4 (rs4135263), 5 (rs1151999), and 6 (rs709149 and rs709154) showed a strong protective effect against AD in APOE 4 allele noncarriers (p = 0.001, permutation p = 0.006, Bonferroni corrected p = 0.021), with a frequency of 39% in cases and 50% in controls. Our data suggest that PPAR-␥ genetic variants may modify the risk of AD in an APOE 4 allele-dependent fashion. © 2009 Elsevier Inc. All rights reserved. Keywords: Alzheimer’s disease; Peroxisome proliferator-activated receptor-␥; APOE; Polymorphism
1. Introduction A chronic inflammatory process might contribute to the neurodegeneration associated with Alzheimer’s disease (AD) (Wyss-Coray and Mucke, 2002), and NF-B, a major transcription factor controlling inflammation, is activated in microglia surrounding senile plaques in AD brains (Kaltschmidt et al., 1997). In addition, a number of studies suggest that increased cellular cholesterol levels induce high amyloid beta (A) production, which is central to the pathogenesis of AD (Puglielli et al., 2003; Shobab et al., 2005; Xiong et al., 2008). The peroxisome proliferator-activated receptor-␥ (PPAR-␥) is a ligand-inducible transcription factor which regulates genes involved in inflammation control and lipid and carbohydrate metabolism: PPAR-␥ inhibits proin∗
Corresponding author. Tel.: +34 942 202507; fax: +34 942 202655. E-mail address: [email protected] (O. Combarros).
0197-4580/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.neurobiolaging.2009.07.004
flammatory gene expression in microglia by suppressing the action of NF-B, and PPAR-␥ increases the expression of a network of genes (such as ABCA1) that promote cholesterol efflux from cells (Heneka and Landreth, 2007; van Neerven et al., 2008). Therefore, PPAR-␥ agonists have been advanced as a new disease altering approach to AD therapy (Landreth et al., 2008). A connection between Apolipoprotein E (APOE) and PPAR-␥ agonist treatment with rosiglitazone became apparent when a statistically significant improvement in cognition occurred in those AD patients that did not possess an APOE 4 allele (Risner et al., 2006). Studies of the association between PPAR␥ gene polymorphisms and AD have produced discordant results (Sauder et al., 2005; Koivisto et al., 2006; Hamilton et al., 2007; Scacchi et al., 2007; Helisalmi et al., 2008). To further elucidate the relationship between PPAR-␥ genetic variation and AD, independently or in concert with the APOE 4 allele, we analyzed seven htSNPs in the PPAR-␥
547.e2
O. Combarros et al. / Neurobiology of Aging 32 (2011) 547.e1–547.e6
gene in a large group of Spanish AD patients and controls.
2. Methods
from a limited geographical area in northern Spain. All patients and controls were ascertained to have parents and grandparents born in Northern Spain to ensure ethnicity. Consequently, possible confounding effects of the inclusion in the study of members of different ethnic groups have been minimized.
2.1. Subjects 2.2. Genotyping The study included 352 late-onset (≥65 years) AD patients (68% women; mean age at study 77.8 years; SD 6.0; range 66–97 years; mean age at onset 74.6 years; SD 5.9; range 65–93 years) who met NINCDS/ADRDA criteria for probable AD (McKhann et al., 1984). All AD cases were defined as sporadic because their family history did not mention any first-degree relative with dementia. AD patients were consecutively admitted to the Department of Neurology, University Hospital “Marqués de Valdecilla”, Santander, Spain, from January, 1997, to December 2001. The large majority of patients were living in the community and had been referred by their general practitioner; few had been admitted from hospital wards or nursing home facilities. Control subjects were 438 unrelated individuals (68% women; mean age 80.9 years; SD 7.4; range 65–100 years) randomly selected from a nursing home. These subjects had complete neurologic and medical examinations that showed that they were free of significant illness and had Mini Mental State Examination scores of 28 or more, which were verified by at least one subsequent annual following-up assessment. The controls arose from the same base population as the cases. The AD and control samples were Caucasians originating
Blood samples were taken after written informed consent had been obtained from the subjects or their representatives. The study was approved by the ethical committee of the University Hospital “Marqués de Valdecilla”. Genotyping of PPAR-␥ (rs10510412, rs17793951, rs1801282, rs4135263, rs1151999, rs709149, and rs709154) polymorphism was performed by a Taq-Man single-nucleotide-polymorphism assay (Applied Biosystems, Warrington, Cheshire, UK) and an ABI PRISM 7000 sequence detection system (Applied Biosystems). We used data from the HapMap project (http://www.hapmap.org) to select the 7 htSNPs capturing 84% of PPAR-␥ genetic variability in Caucasians. SNPs were chosen among those with minor allele frequencies ≥5% using Haploview v3.2 software (http://www.broad.mit.edu/mpg/haploview) with an r2 threshold of 0.8. The location of SNPs in PPAR-␥ gene used in the present study is described in Fig. 1. APOE genotyping was performed by amplification of the 4th exon of the APOE gene by PCR with biotinylated primers, followed by reverse hybridization on nitrocellulose strips, using the INNO-LIPA ApoE assay (Innogenetics NV, Ghent, Belgium).
Fig. 1. Genomic structure and relative location of studied haplotype tagging SNPs (indicated by grey boxes) in the PPAR-␥ gene. Lines represent the introns between exons (black boxes). Haplotype blocks were delineated using the solid spine of LD method implemented in Haploview 3.32. Pairwise linkage disequilibrium patterns between the 7 SNPs formed two different haplotype blocks (numbers in box represent r2 values, and the intensity of the color is proportional to the strength of the LD).
O. Combarros et al. / Neurobiology of Aging 32 (2011) 547.e1–547.e6
2.3. Statistical analysis Hardy–Weinberg equilibrium (HWE) was calculated for the 7 htSNPs genotypes in the control population using Pearson’s χ2 statistics. We assessed pairwise linkage disequilibrium (LD) between the 7 htSNPs by D and r2 statistics. Haplotype reconstruction and their frequencies in cases and controls were estimated by an expectation-maximization algorithm; blocks were delineated using the solid spine of LD method implemented in Haploview 3.32 (Fig. 1). Pearson’s χ2 statistics were performed to compare allele distribution of the patients and control for each htSNP. Haplotype frequencies were also assessed using Pearson’s χ2 . Rare haplotypes (total frequency <0.05) were excluded from the analysis. In order to obtain a measure of significance corrected for multiple testing, we ran 10,000 permutations to compute empiric p-values using Haploview. In addition, we used Bonferroni’s
547.e3
method to correct our nominal p-values for all 21 tests performed, consisting of 7 SNPs with 3 subgroups (total sample, APOE 4 allele carriers and APOE 4 allele noncarriers). Genotypic distributions of those SNPs in which we found an allelic association with AD were assessed by logistic regression using SPSS software. Odds ratios (ORs) and p-values adjusted by age and gender were also calculated by logistic regression.
3. Results In control groups, no significant deviations from Hardy–Weinberg equilibrium were found for any of the 7 SNPs. As shown in Table 1, the distribution of the allele and genotype frequencies in the total sample did not differ significantly between AD and control groups. After stratifying
Table 1 Distribution of PPAR-␥ polymorphisms in patients and controls stratified by APOE 4 allele. PPAR-␥ polymorphism
APOE 4 allele noncarriers
APOE 4 allele carriers
Total sample
Patients
Controls
Patients
Controls
Patients
Controls
rs10510412
GG GA AA Total Allele frequency G/A
102 (0.63) 52 (0.32) 7 (0.05) 161 0.80/0.20
245 (0.68) 107 (0.30) 9 (0.02) 361 0.83/0.17
122 (0.65) 52 (0.28) 13 (0.07) 187 0.79/0.21
54 (0.70) 20 (0.26) 3 (0.04) 77 0.83/0.17
224 (0.64) 104 (0.30) 20 (0.06) 348 0.79/0.21
299 (0.68) 127 (0.29) 12 (0.03) 438 0.83/0.17
rs17793951
AA AG GG Total Allele frequency A/G
75 (0.46) 64 (0.40) 23 (0.14) 162 0.66/0.34
158 (0.44) 166 (0.46) 37 (0.10) 361 0.67/0.33
84 (0.45) 85 (0.45) 18 (0.10) 187 0.68/0.32
25 (0.32) 41 (0.53) 11 (0.14) 77 0.59/0.41
159 (0.45) 149 (0.43) 41 (0.12) 349 0.67/0.33
183 (0.42) 207 (0.47) 48 (0.11) 438 0.65/0.35
rs1801282
CC CG GG Total Allele frequency C/G
139 (0.85) 21 (0.13) 4 (0.02) 164 0.91/0.09
310 (0.86) 48 (0.13) 3 (0.01) 361 0.93/0.07
153 (0.82) 28 (0.15) 6 (0.03) 187 0.89/0.11
66 (0.86) 8 (0.10) 3 (0.04) 77 0.91/0.09
292 (0.83) 49 (0.14) 10 (0.03) 351 0.90/0.10
376 (0.86) 56 (0.13) 6 (0.01) 438 0.92/0.08
rs4135623
TT TC CC Total Allele frequency T/C
116 (0.72) 41 (0.25) 5 (0.03) 162 0.84/0.16
262 (0.73) 95 (0.26) 4 (0.01) 361 0.86/0.14
131 (0.70) 48 (0.26) 7 (0.04) 186 0.83/0.17
58 (0.75) 18 (0.24) 1 (0.01) 77 0.87/0.13
247 (0.71) 89 (0.26) 12 (0.03) 348 0.84/0.16
320 (0.73) 113 (0.26) 5 (0.01) 438 0.86/0.14
rs1151999
CC CA AA Total Allele frequency C/A
33 (0.21) 78 (0.49) 48 (0.30) 159 0.45/0.55
91 (0.25) 190 (0.53) 80 (0.22) 361 0.52/0.48
37 (0.20) 106 (0.58) 41 (0.22) 184 0.49/0.51
14 (0.18) 41 (0.54) 21 (0.28) 76 0.45/0.55
70 (0.20) 184 (0.54) 89 (0.26) 343 0.47/0.53
105 (0.24) 231 (0.53) 101 (0.23) 437 0.50/0.50
rs709149
CC CT TT Total Allele frequency C/T
61 (0.37) 61 (0.37) 43 (0.26) 165 0.55/0.45
148 (0.41) 169 (0.47) 44 (0.12) 361 0.64/0.36
69 (0.37) 99 (0.53) 19 (0.10) 187 0.63/0.37
22 (0.29) 45 (0.58) 10 (0.13) 77 0.58/0.42
130 (0.37) 160 (0.45) 62 (0.18) 352 0.60/0.40
170 (0.39) 214 (0.49) 54 (0.12) 438 0.63/0.37
rs709154
AA AT TT Total Allele frequency A/T
75 (0.47) 59 (0.37) 26 (0.16) 160 0.65/0.35
166 (0.46) 159 (0.44) 36 (0.10) 361 0.68/0.32
83 (0.45) 85 (0.46) 17 (0.09) 185 0.68/0.32
29 (0.38) 41 (0.53) 7 (0.09) 77 0.64/0.36
158 (0.46) 144 (0.42) 43 (0.12) 345 0.67/0.33
195 (0.44) 200 (0.46) 43 (0.10) 438 0.67/0.33
Figures in parentheses indicate frequencies; in bold, permutation p-values <0.05 for allelic comparisons.
547.e4
O. Combarros et al. / Neurobiology of Aging 32 (2011) 547.e1–547.e6
Table 2 Haplotype association analysis between PPAR-␥ gene and AD stratified by APOE 4 allele. Haplotype
APOE e4 allele noncarriers
APOE e4 allele carriers
Total sample
AD, control frequency
p-value
Permutation p-value
AD, control frequency
p-value
Permutation p-value
AD, control frequency
p-value
Permutation p-value
Block 1 GAC GGC AAC AAG
0.46, 0.50 0.33, 0.32 0.12, 0.10 0.08, 0.07
0.16 0.72 0.23 0.54
0.73 0.99 0.90 0.99
0.46, 0.42 0.32, 0.40 0.11, 0.09 0.10, 0.08
0.40 0.09 0.47 0.50
0.99 0.61 0.99 0.99
0.46, 0.49 0.33, 0.34 0.12, 0.10 0.09, 0.07
0.27 0.66 0.19 0.18
0.93 0.99 0.81 0.80
Block 2 TCCA TATT CACA
0.39, 0.50 0.31, 0.30 0.16, 0.12
0.001 0.77 0.58
0.006 0.99 0.99
0.45, 0.44 0.30, 0.30 0.15, 0.12
0.89 0.27 0.40
0.99 0.96 0.99
0.42, 0.49 0.31, 0.31 0.14, 0.13
0.009 0.85 0.90
0.059 0.99 0.99
Haplotype Block 1 consists of SNPs rs10510412, rs17793951, and rs1801282. Haplotype Block 2 consists of SNPs rs4135263, rs1151999, rs709149, and rs709154. Rare haplotypes (total frequency < 0.05) were excluded from the analysis.
by APOE (Table 1), in APOE 4 allele noncarriers there was a decreased frequency of the PPAR-␥ (rs709149) C allele in patients (55%) versus controls (64%) (OR = 0.69; 95% CI 0.53–0.90; p = 0.0056; permutation p = 0.03; p = 0.11 Bonferroni corrected for the total 21 tests performed), and the CC genotype of PPAR-␥ (rs709149) decreased significantly the risk of AD (OR = 0.37; 95% CI 0.22–0.62; p = 10−5 ). Logistic regression interaction term between APOE non-4 allele and PPAR-␥ (rs709149) C allele was statistically significant (OR = 0.32; 95% CI 0.12–0.86; p = 0.02). There were no major differences in allele or genotype frequencies of PPAR␥ polymorphisms in our total sample associated to gender subgroups. Based on pairwise linkage disequilibrium data of the 7 PPAR-␥ htSNPs, two main haplotype blocks were constructed (Fig. 1), block 1 consisting of SNPs rs10510412, rs17793951, and rs1801282 located within introns 1 and 2, and block 2 consisting of SNPs rs4135263, rs1151999, rs709149, and rs709154 located in introns 4, 5, and 6. Seven different haplotypes with frequencies >5% were estimated within the two blocks of the PPAR-␥ gene (Table 2). The most common block 2 haplotype derived from SNPs in introns 4, 5 and 6 (TCCA) showed a strong and significant protective effect against AD only in APOE 4 allele noncarriers (p = 0.001; permutation p = 0.006; p = 0.021, Bonferroni corrected with 21 tests), with a frequency of 39% in cases and 50% in controls.
4. Discussion A polymorphic site in exon 2 (mRNA transcripts ␥-1 and ␥-3) or intron 2 (mRNA transcript ␥-2) of the PPAR␥ gene, the Pro12Ala (C/G, rs1801282) polymorphism, has been widely studied in AD and has produced discordant results as to the effect of the G (Ala) allele was associated with higher risk for AD in patients who were 80 years or older (Scacchi et al., 2007), was related to lower onset age of AD (Koivisto et al., 2006), was protective against AD in
females who were APOE 4 allele carriers (Hamilton et al., 2007), or showed no association with AD risk (Sauder et al., 2005). Moreover, elderly individuals with normal cognition who were carriers of the PPAR-␥ Ala12 allele showed a greater risk of cognitive impairment and dementia overtime compared to noncarriers (West et al., 2008), whereas in other study, PPAR-␥ Ala12 allele carriers were associated with less risk of cognitive decline (Yaffe et al., 2008). Differently to all these previous published genetic association studies between PPAR-␥ gene and AD exclusively based on single SNP (rs1801282) analyses, Helisalmi et al. (2008) examined multiple SNPs that might have much greater potential to prove the hypothesis that there is a genetic link between AD and PPAR-␥ gene. Helisalmi et al. (2008) examined eight polymorphic sites (rs2920505, rs17036314, and rs2028759 in intron 1; rs1801282 in intron 2; rs2938395 in intron 3; and rs709149 and rs1797912 in intron 6) in PPAR-␥ gene and failed to observe any allele, genotype or haplotype association with AD either in the whole study or in APOE 4-stratified subgroups in the Finnish population. We found a common protective PPAR-␥ haplotype (TCCA) composed of intron 4 (rs4135263), intron 5 (rs1151999), and intron 6 (rs709149 and rs709154) polymorphisms which accounted for an estimated 39% of cases and 50% of controls among APOE 4 allele noncarriers. The previously reported Pro12Ala (C/G, rs1801282) polymorphism, in exon/intron 2, is at 57.2 kb distance from our most significantly associated htSNP (rs709149) in intron 6. Based on Hap Map data for Caucasians, these two SNPs are not in strong LD (D = 0.42; r2 = 0.009), and consequently, they are located in different haplotype blocks: rs1801282 is included in block 1 which we found not to be associated to AD, and rs709149 is in block 2. Characterized by our markers in introns 4, 5 and 6, block 2 includes most PPAR-␥ gene exons (Fig. 1), and therefore, we hypothesize that some yet unidentified functional variant, perhaps affecting post-transcriptional processing and the expression level of splice variants, might be tagged by the TCCA block 2 haplotype. Helisalmi et al. (2008) had a greater sample size that ours one and they also
O. Combarros et al. / Neurobiology of Aging 32 (2011) 547.e1–547.e6
stratified their population according to APOE genotype without finding an association of the rs709149 polymorphism; since the APOE 4, 3 and 2 allele frequencies differ in Finland and Spain, it would be interesting to understand if these variations contribute to the obtention of these different results. Brain extracts from AD patients contain less PPAR-␥ protein than controls (Sastre et al., 2006; Xiong et al., 2008); therefore, investigation of whether PPAR-␥ expression levels vary with different PPAR-␥ haplotypes would help to clarify their potential relevance to AD. It is tempting to speculate that subjects carrying the PPAR-␥ haplotype (TCCA) characterized by intron 4 (rs4135263), intron 5 (rs1151999), and intron 6 (rs709149 and rs709154) polymorphisms might show a lower AD risk due to an increased expression of PPAR-␥ in the brain. Indeed, PPAR␥ suppresses microglial inflammatory responses (Combs et al., 2000), and also stimulates cellular cholesterol efflux (Chinetti et al., 2001), which in turn could decrease AD risk. Our data suggest that PPAR-␥ genetic variants may modify the risk of AD in an APOE 4 allele-dependent fashion: the protective effect against AD of the PPAR-␥ TCCA haplotype was only observed in the absence of the APOE 4 allele. We acknowledge the possibility of false positive observation mainly due to stratification with APOE, so we have used an over-conservative Bonferroni correction for all 21 tests performed consisting of 7 SNPs with 3 subgroups (total sample, APOE 4 allele carriers and APOE 4 allele noncarriers). APOE 4 allele, the major genetic risk factor for AD, is associated with enhanced brain inflammation: following intracerebroventricular injection of LPS, microglial and NF-B activation are more pronounced in the hippocampi of transgenic APOE4 than in APOE3 mice (Ophir et al., 2005). Additionally, using cultured astrocytes prepared from human APOE3 and APOE4 knock-in mice, the amount of lipids released from APOE3- and APOE4-expressing astrocytes is APOE-isoform-dependent, with the order of potency being APOE3 > APOE4 (Gong et al., 2002); so, it is possible to speculate that cholesterol levels in astrocytes of APOE4 carriers are higher than in those of APOE3 carriers, leading to increased generation of A in the brain of APOE4 carriers. As the protective effect against AD of the PPAR-␥ introns 4, 5 and 6 haplotype (TCCA) might be counteracted by the increased inflammation and cholesterol retention in the brain associated to the APOE 4 allele, we hypothesize that the association between PPAR-␥ and AD should be only evident in the absence of the APOE 4 allele. In fact, more than 500 patients with mild to moderate AD were treated for 6 months with the PPAR-␥ agonist rosiglitazone, resulting in a statistically significant improvement in cognition in those patients that did not possess an APOE 4 allele, whereas patients with APOE 4 did not respond to the drug (Risner et al., 2006). Additional studies using different sets of patients and control subjects are required to verify the validity of our hypotheses (genetic and biologic).
547.e5
5. Conflict of interest The authors hereby declare that they have no conflicts of interest, and that all human studies have been approved by the appropriate ethics committee and have, therefore, performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki
Acknowledgement The authors thank Weyma Notel’s help in Fig. 1 edition.
References Chinetti, G., Lestavel, S., Bocher, V., Remaley, A.T., Neve, B., Torra, I.P., Teissier, E., Minnich, A., Jaye, M., Duverger, N., Brewer, H.B., Fruchart, J.C., Clavey, V., Staels, B., 2001. PPAR-␣ and PPAR-␥ activators induce cholesterol removal from human macrophage foam cells through stimulation of the ABCA1 pathway. Nat. Med. 7, 53–58. Combs, C.K., Johnson, D.E., Karlo, J.C., Cannady, S.B., Landreth, G.E., 2000. Inflammatory mechanisms in Alzheimer’s disease: inhibition of -amyloid-stimulated proinflammatory responses and neurotoxicity by PPAR␥ agonists. J. Neurosci. 20, 558–567. Gong, J.S., Kobayashi, M., Hayashi, H., Zou, K., Sawamura, N., Jujita, S.C., Yanagisawa, K., Michikawa, M., 2002. Apolipoprotein E (ApoE) isoform-dependent lipid release from astrocytes prepared from human ApoE3 and ApoE4 knock-in mice. J. Biol. Chem. 277, 29919–29926. Hamilton, G., Proitsi, P., Jehu, L., Morgan, A., Williams, J., OˇıDonovan, M.C., Owen, M.J., Powell, J.F., Lovestone, S., 2007. Candidate gene association study of insulin signaling genes and Alzheimer’s disease: evidence for SOS2, PCK1, and PPARgamma as susceptibility loci. Am. J. Med. Genet. B: Neuropsychiatr. Genet. 144B, 508–516. Helisalmi, S., Tarvainen, T., Vepsäläinen, S., Koivisto, A.M., Hiltunen, M., Soininen, H., 2008. Lack of genetic association between PPARG gene polymorphisms and Finnish late-onset Alzheimer’s disease. Neurosci. Lett. 441, 233–236. Heneka, M.T., Landreth, G.E., 2007. PPARs in the brain. Biochim. Biophys. Acta 1771, 1031–1045. Kaltschmidt, B., Uherek, M., Volk, B., Baeuerle, P.A., Kaltschmidt, C., 1997. Transcription factor NF-B is activated in primary neurons by amyloid  peptides and in neurons surrounding early plaques from patients with Alzheimer disease. Proc. Natl. Acad. Sci. U.S.A. 94, 2642–2647. Koivisto, A.M., Helisalmi, S., Pihlajamäki, J., Hiltunen, M., Koivisto, K., Moilanen, L., Kuusisto, J., Helkala, E.L., Hänninen, T., Kervinen, K., Kesäniemi, Y.A., Laakso, M., Soininen, H., 2006. Association analysis of peroxisome proliferators-activated receptor gamma polymorphisms and late onset Alzheimer’s disease in the Finnish population. Dement. Geriatr. Cogn. Disord. 22, 449–453. Landreth, G., Jiang, Q., Mandrekar, S., Heneka, M., 2008. PPAR␥ agonists as therapeutics for the treatment of Alzheimer’s disease. Neurotherapeutics 5, 481–489. McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., Stadlan, E.M., 1984. Clinical diagnosis of Alzheimer disease: report of the NINCDS-ADRDA work group under the auspices of the Department of Health and Human Services Task Force on Alzheimer Disease. Neurology 24, 939–944. Ophir, G., Amariglio, N., Jacob-Hirsch, J., Elkon, R., Rechavi, G., Michaelson, D.M., 2005. Apolipoprotein E4 enhances brain inflammation by modulation of the NF-B signaling cascade. Neurobiol. Dis. 20, 709–718. Puglielli, L., Tanzi, R.E., Kovacs, D.M., 2003. Alzheimer’s disease: the cholesterol connection. Nat. Neurosci. 6, 345–351.
547.e6
O. Combarros et al. / Neurobiology of Aging 32 (2011) 547.e1–547.e6
Risner, M.E., Saunders, A.M., Altman, J.F., Ormandy, G.C., Craft, S., Foley, I.M., Zvartau-Hind, M.E., Hosford, D.A., Roses, A.D., Rosiglitazone in Alzheimer’s disease study group, 2006. Efficacy of rosiglitazone in a genetically defined population with mild-to-moderate Alzheimer’s disease. Pharmacogenomics J. 6, 246–254. Sastre, M., Dewachter, I., Rossner, S., Bogdanovic, N., Rosen, E., Borghgraef, P., Evert, B.O., Dumitrescu-Ozimek, L., Thal, D.R., Landreth, G., Walter, J., Klockgether, T., van Leuven, F., Heneka, M.T., 2006. Nonsteroidal anti-inflammatory drugs repress -secretase gene promoter activity by the activation of PPAR␥. Proc. Natl. Acad. Sci. U.S.A. 103, 443–448. Sauder, S., Kölsch, H., Lütjohann, D., Schulz, A., von Bergmann, M., Maier, W., Heun, R., 2005. Influence of peroxisome proliferators-activated receptor gamma gene polymorphism on 24S-hydroxycholesterol levels in Alzheimer’s patients. J. Neural Transm. 112, 1381–1389. Scacchi, R., Pinto, A., Gambina, G., Rosano, A., Corbo, R.M., 2007. The peroxisome proliferator-activated receptor gamma (PPAR-␥2) Pro12Ala polymorphism is associated with higher risk for Alzheimer’s disease in octogenarians. Brain Res. 1139, 1–5.
Shobab, L.A., Hsiung, G.Y., Feldman, H.H., 2005. Cholesterol in Alzheimer’s disease. Lancet Neurol. 4, 841–852. van Neerven, S., Kampmann, E., Mey, J., 2008. RAR/RXR and PPAR/RXR signalling in neurological and psychiatric diseases. Prog. Neurobiol. 85, 433–451. West, N.A., Haan, M.N., Morgenstern, H., 2008. The PPAR-gamma Pro12Ala polymorphism and risk of cognitive impairment in a longitudinal study. Neurobiol. Aging, doi:10.1016/j.neurobiolaging.2008.06.005. Wyss-Coray, T., Mucke, L., 2002. Inflammation in neurodegenerative disease: a double-edged sword. Neuron 35, 419–432. Xiong, H., Callaghan, D., Jones, A., Walker, D.G., Lue, L.F., Beach, T.G., Sue, L.I., Woulfe, J., Xu, H., Stanimirovic, D.B., Zhang, W., 2008. Cholesterol retention in Alzheimer’s brain is responsible for high and ␥-secretase activities and A production. Neurobiol. Dis. 29, 422– 437. Yaffe, K., Kanaya, A.M., Lindquist, K., Hsueh, W.C., Cummings, S.R., Beamer, B., Newman, A., Rosano, C., Li, R., Harris, T., Health ABC Study, 2008. PPAR-␥ Pro12Ala genotype and risk of cognitive decline in elders. Neurobiol. Aging 29, 78–83.
APOE dependent-association of PPAR-␥ genetic variants with Alzheimer’s disease risk Onofre Combarros ∗ , Eloy Rodríguez-Rodríguez, Ignacio Mateo, José Luis Vázquez-Higuera, Jon Infante, José Berciano, Pascual Sánchez-Juan Neurology Service and Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), “Marqués de Valdecilla” University Hospital (University of Cantabria), Santander, Spain Received 8 December 2008; received in revised form 11 May 2009; accepted 12 July 2009 Available online 6 August 2009
Abstract The peroxisome proliferator-activated receptor-␥ (PPAR-␥) is a ligand-inducible transcription factor that suppresses microglial inflammatory responses and inhibits amyloid beta (A) production through promoting cholesterol efflux from glial cells. PPAR-␥ agonists have been advanced as a new disease altering approach to Alzheimer’s disease (AD), with rosiglitazone therapy having improved cognition in those AD patients that did not possess an Apolipoprotein E (APOE) 4 allele. The current study was designed to explore the effect of interactions between PPAR-␥ and APOE gene polymorphisms on the AD risk. We examined genetic variations of PPAR-␥ by genotyping 7 haplotype tagging SNPs (htSNPs) (rs10510412, rs17793951, rs1801282, rs4135263, rs1151999, rs709149, and rs709154) in a group of 352 Spanish late-onset AD cases and 438 controls. The PPAR-␥ TCCA haplotype derived from SNPs in introns 4 (rs4135263), 5 (rs1151999), and 6 (rs709149 and rs709154) showed a strong protective effect against AD in APOE 4 allele noncarriers (p = 0.001, permutation p = 0.006, Bonferroni corrected p = 0.021), with a frequency of 39% in cases and 50% in controls. Our data suggest that PPAR-␥ genetic variants may modify the risk of AD in an APOE 4 allele-dependent fashion. © 2009 Elsevier Inc. All rights reserved. Keywords: Alzheimer’s disease; Peroxisome proliferator-activated receptor-␥; APOE; Polymorphism
1. Introduction A chronic inflammatory process might contribute to the neurodegeneration associated with Alzheimer’s disease (AD) (Wyss-Coray and Mucke, 2002), and NF-B, a major transcription factor controlling inflammation, is activated in microglia surrounding senile plaques in AD brains (Kaltschmidt et al., 1997). In addition, a number of studies suggest that increased cellular cholesterol levels induce high amyloid beta (A) production, which is central to the pathogenesis of AD (Puglielli et al., 2003; Shobab et al., 2005; Xiong et al., 2008). The peroxisome proliferator-activated receptor-␥ (PPAR-␥) is a ligand-inducible transcription factor which regulates genes involved in inflammation control and lipid and carbohydrate metabolism: PPAR-␥ inhibits proin∗
Corresponding author. Tel.: +34 942 202507; fax: +34 942 202655. E-mail address: [email protected] (O. Combarros).
0197-4580/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.neurobiolaging.2009.07.004
flammatory gene expression in microglia by suppressing the action of NF-B, and PPAR-␥ increases the expression of a network of genes (such as ABCA1) that promote cholesterol efflux from cells (Heneka and Landreth, 2007; van Neerven et al., 2008). Therefore, PPAR-␥ agonists have been advanced as a new disease altering approach to AD therapy (Landreth et al., 2008). A connection between Apolipoprotein E (APOE) and PPAR-␥ agonist treatment with rosiglitazone became apparent when a statistically significant improvement in cognition occurred in those AD patients that did not possess an APOE 4 allele (Risner et al., 2006). Studies of the association between PPAR␥ gene polymorphisms and AD have produced discordant results (Sauder et al., 2005; Koivisto et al., 2006; Hamilton et al., 2007; Scacchi et al., 2007; Helisalmi et al., 2008). To further elucidate the relationship between PPAR-␥ genetic variation and AD, independently or in concert with the APOE 4 allele, we analyzed seven htSNPs in the PPAR-␥
547.e2
O. Combarros et al. / Neurobiology of Aging 32 (2011) 547.e1–547.e6
gene in a large group of Spanish AD patients and controls.
2. Methods
from a limited geographical area in northern Spain. All patients and controls were ascertained to have parents and grandparents born in Northern Spain to ensure ethnicity. Consequently, possible confounding effects of the inclusion in the study of members of different ethnic groups have been minimized.
2.1. Subjects 2.2. Genotyping The study included 352 late-onset (≥65 years) AD patients (68% women; mean age at study 77.8 years; SD 6.0; range 66–97 years; mean age at onset 74.6 years; SD 5.9; range 65–93 years) who met NINCDS/ADRDA criteria for probable AD (McKhann et al., 1984). All AD cases were defined as sporadic because their family history did not mention any first-degree relative with dementia. AD patients were consecutively admitted to the Department of Neurology, University Hospital “Marqués de Valdecilla”, Santander, Spain, from January, 1997, to December 2001. The large majority of patients were living in the community and had been referred by their general practitioner; few had been admitted from hospital wards or nursing home facilities. Control subjects were 438 unrelated individuals (68% women; mean age 80.9 years; SD 7.4; range 65–100 years) randomly selected from a nursing home. These subjects had complete neurologic and medical examinations that showed that they were free of significant illness and had Mini Mental State Examination scores of 28 or more, which were verified by at least one subsequent annual following-up assessment. The controls arose from the same base population as the cases. The AD and control samples were Caucasians originating
Blood samples were taken after written informed consent had been obtained from the subjects or their representatives. The study was approved by the ethical committee of the University Hospital “Marqués de Valdecilla”. Genotyping of PPAR-␥ (rs10510412, rs17793951, rs1801282, rs4135263, rs1151999, rs709149, and rs709154) polymorphism was performed by a Taq-Man single-nucleotide-polymorphism assay (Applied Biosystems, Warrington, Cheshire, UK) and an ABI PRISM 7000 sequence detection system (Applied Biosystems). We used data from the HapMap project (http://www.hapmap.org) to select the 7 htSNPs capturing 84% of PPAR-␥ genetic variability in Caucasians. SNPs were chosen among those with minor allele frequencies ≥5% using Haploview v3.2 software (http://www.broad.mit.edu/mpg/haploview) with an r2 threshold of 0.8. The location of SNPs in PPAR-␥ gene used in the present study is described in Fig. 1. APOE genotyping was performed by amplification of the 4th exon of the APOE gene by PCR with biotinylated primers, followed by reverse hybridization on nitrocellulose strips, using the INNO-LIPA ApoE assay (Innogenetics NV, Ghent, Belgium).
Fig. 1. Genomic structure and relative location of studied haplotype tagging SNPs (indicated by grey boxes) in the PPAR-␥ gene. Lines represent the introns between exons (black boxes). Haplotype blocks were delineated using the solid spine of LD method implemented in Haploview 3.32. Pairwise linkage disequilibrium patterns between the 7 SNPs formed two different haplotype blocks (numbers in box represent r2 values, and the intensity of the color is proportional to the strength of the LD).
O. Combarros et al. / Neurobiology of Aging 32 (2011) 547.e1–547.e6
2.3. Statistical analysis Hardy–Weinberg equilibrium (HWE) was calculated for the 7 htSNPs genotypes in the control population using Pearson’s χ2 statistics. We assessed pairwise linkage disequilibrium (LD) between the 7 htSNPs by D and r2 statistics. Haplotype reconstruction and their frequencies in cases and controls were estimated by an expectation-maximization algorithm; blocks were delineated using the solid spine of LD method implemented in Haploview 3.32 (Fig. 1). Pearson’s χ2 statistics were performed to compare allele distribution of the patients and control for each htSNP. Haplotype frequencies were also assessed using Pearson’s χ2 . Rare haplotypes (total frequency <0.05) were excluded from the analysis. In order to obtain a measure of significance corrected for multiple testing, we ran 10,000 permutations to compute empiric p-values using Haploview. In addition, we used Bonferroni’s
547.e3
method to correct our nominal p-values for all 21 tests performed, consisting of 7 SNPs with 3 subgroups (total sample, APOE 4 allele carriers and APOE 4 allele noncarriers). Genotypic distributions of those SNPs in which we found an allelic association with AD were assessed by logistic regression using SPSS software. Odds ratios (ORs) and p-values adjusted by age and gender were also calculated by logistic regression.
3. Results In control groups, no significant deviations from Hardy–Weinberg equilibrium were found for any of the 7 SNPs. As shown in Table 1, the distribution of the allele and genotype frequencies in the total sample did not differ significantly between AD and control groups. After stratifying
Table 1 Distribution of PPAR-␥ polymorphisms in patients and controls stratified by APOE 4 allele. PPAR-␥ polymorphism
APOE 4 allele noncarriers
APOE 4 allele carriers
Total sample
Patients
Controls
Patients
Controls
Patients
Controls
rs10510412
GG GA AA Total Allele frequency G/A
102 (0.63) 52 (0.32) 7 (0.05) 161 0.80/0.20
245 (0.68) 107 (0.30) 9 (0.02) 361 0.83/0.17
122 (0.65) 52 (0.28) 13 (0.07) 187 0.79/0.21
54 (0.70) 20 (0.26) 3 (0.04) 77 0.83/0.17
224 (0.64) 104 (0.30) 20 (0.06) 348 0.79/0.21
299 (0.68) 127 (0.29) 12 (0.03) 438 0.83/0.17
rs17793951
AA AG GG Total Allele frequency A/G
75 (0.46) 64 (0.40) 23 (0.14) 162 0.66/0.34
158 (0.44) 166 (0.46) 37 (0.10) 361 0.67/0.33
84 (0.45) 85 (0.45) 18 (0.10) 187 0.68/0.32
25 (0.32) 41 (0.53) 11 (0.14) 77 0.59/0.41
159 (0.45) 149 (0.43) 41 (0.12) 349 0.67/0.33
183 (0.42) 207 (0.47) 48 (0.11) 438 0.65/0.35
rs1801282
CC CG GG Total Allele frequency C/G
139 (0.85) 21 (0.13) 4 (0.02) 164 0.91/0.09
310 (0.86) 48 (0.13) 3 (0.01) 361 0.93/0.07
153 (0.82) 28 (0.15) 6 (0.03) 187 0.89/0.11
66 (0.86) 8 (0.10) 3 (0.04) 77 0.91/0.09
292 (0.83) 49 (0.14) 10 (0.03) 351 0.90/0.10
376 (0.86) 56 (0.13) 6 (0.01) 438 0.92/0.08
rs4135623
TT TC CC Total Allele frequency T/C
116 (0.72) 41 (0.25) 5 (0.03) 162 0.84/0.16
262 (0.73) 95 (0.26) 4 (0.01) 361 0.86/0.14
131 (0.70) 48 (0.26) 7 (0.04) 186 0.83/0.17
58 (0.75) 18 (0.24) 1 (0.01) 77 0.87/0.13
247 (0.71) 89 (0.26) 12 (0.03) 348 0.84/0.16
320 (0.73) 113 (0.26) 5 (0.01) 438 0.86/0.14
rs1151999
CC CA AA Total Allele frequency C/A
33 (0.21) 78 (0.49) 48 (0.30) 159 0.45/0.55
91 (0.25) 190 (0.53) 80 (0.22) 361 0.52/0.48
37 (0.20) 106 (0.58) 41 (0.22) 184 0.49/0.51
14 (0.18) 41 (0.54) 21 (0.28) 76 0.45/0.55
70 (0.20) 184 (0.54) 89 (0.26) 343 0.47/0.53
105 (0.24) 231 (0.53) 101 (0.23) 437 0.50/0.50
rs709149
CC CT TT Total Allele frequency C/T
61 (0.37) 61 (0.37) 43 (0.26) 165 0.55/0.45
148 (0.41) 169 (0.47) 44 (0.12) 361 0.64/0.36
69 (0.37) 99 (0.53) 19 (0.10) 187 0.63/0.37
22 (0.29) 45 (0.58) 10 (0.13) 77 0.58/0.42
130 (0.37) 160 (0.45) 62 (0.18) 352 0.60/0.40
170 (0.39) 214 (0.49) 54 (0.12) 438 0.63/0.37
rs709154
AA AT TT Total Allele frequency A/T
75 (0.47) 59 (0.37) 26 (0.16) 160 0.65/0.35
166 (0.46) 159 (0.44) 36 (0.10) 361 0.68/0.32
83 (0.45) 85 (0.46) 17 (0.09) 185 0.68/0.32
29 (0.38) 41 (0.53) 7 (0.09) 77 0.64/0.36
158 (0.46) 144 (0.42) 43 (0.12) 345 0.67/0.33
195 (0.44) 200 (0.46) 43 (0.10) 438 0.67/0.33
Figures in parentheses indicate frequencies; in bold, permutation p-values <0.05 for allelic comparisons.
547.e4
O. Combarros et al. / Neurobiology of Aging 32 (2011) 547.e1–547.e6
Table 2 Haplotype association analysis between PPAR-␥ gene and AD stratified by APOE 4 allele. Haplotype
APOE e4 allele noncarriers
APOE e4 allele carriers
Total sample
AD, control frequency
p-value
Permutation p-value
AD, control frequency
p-value
Permutation p-value
AD, control frequency
p-value
Permutation p-value
Block 1 GAC GGC AAC AAG
0.46, 0.50 0.33, 0.32 0.12, 0.10 0.08, 0.07
0.16 0.72 0.23 0.54
0.73 0.99 0.90 0.99
0.46, 0.42 0.32, 0.40 0.11, 0.09 0.10, 0.08
0.40 0.09 0.47 0.50
0.99 0.61 0.99 0.99
0.46, 0.49 0.33, 0.34 0.12, 0.10 0.09, 0.07
0.27 0.66 0.19 0.18
0.93 0.99 0.81 0.80
Block 2 TCCA TATT CACA
0.39, 0.50 0.31, 0.30 0.16, 0.12
0.001 0.77 0.58
0.006 0.99 0.99
0.45, 0.44 0.30, 0.30 0.15, 0.12
0.89 0.27 0.40
0.99 0.96 0.99
0.42, 0.49 0.31, 0.31 0.14, 0.13
0.009 0.85 0.90
0.059 0.99 0.99
Haplotype Block 1 consists of SNPs rs10510412, rs17793951, and rs1801282. Haplotype Block 2 consists of SNPs rs4135263, rs1151999, rs709149, and rs709154. Rare haplotypes (total frequency < 0.05) were excluded from the analysis.
by APOE (Table 1), in APOE 4 allele noncarriers there was a decreased frequency of the PPAR-␥ (rs709149) C allele in patients (55%) versus controls (64%) (OR = 0.69; 95% CI 0.53–0.90; p = 0.0056; permutation p = 0.03; p = 0.11 Bonferroni corrected for the total 21 tests performed), and the CC genotype of PPAR-␥ (rs709149) decreased significantly the risk of AD (OR = 0.37; 95% CI 0.22–0.62; p = 10−5 ). Logistic regression interaction term between APOE non-4 allele and PPAR-␥ (rs709149) C allele was statistically significant (OR = 0.32; 95% CI 0.12–0.86; p = 0.02). There were no major differences in allele or genotype frequencies of PPAR␥ polymorphisms in our total sample associated to gender subgroups. Based on pairwise linkage disequilibrium data of the 7 PPAR-␥ htSNPs, two main haplotype blocks were constructed (Fig. 1), block 1 consisting of SNPs rs10510412, rs17793951, and rs1801282 located within introns 1 and 2, and block 2 consisting of SNPs rs4135263, rs1151999, rs709149, and rs709154 located in introns 4, 5, and 6. Seven different haplotypes with frequencies >5% were estimated within the two blocks of the PPAR-␥ gene (Table 2). The most common block 2 haplotype derived from SNPs in introns 4, 5 and 6 (TCCA) showed a strong and significant protective effect against AD only in APOE 4 allele noncarriers (p = 0.001; permutation p = 0.006; p = 0.021, Bonferroni corrected with 21 tests), with a frequency of 39% in cases and 50% in controls.
4. Discussion A polymorphic site in exon 2 (mRNA transcripts ␥-1 and ␥-3) or intron 2 (mRNA transcript ␥-2) of the PPAR␥ gene, the Pro12Ala (C/G, rs1801282) polymorphism, has been widely studied in AD and has produced discordant results as to the effect of the G (Ala) allele was associated with higher risk for AD in patients who were 80 years or older (Scacchi et al., 2007), was related to lower onset age of AD (Koivisto et al., 2006), was protective against AD in
females who were APOE 4 allele carriers (Hamilton et al., 2007), or showed no association with AD risk (Sauder et al., 2005). Moreover, elderly individuals with normal cognition who were carriers of the PPAR-␥ Ala12 allele showed a greater risk of cognitive impairment and dementia overtime compared to noncarriers (West et al., 2008), whereas in other study, PPAR-␥ Ala12 allele carriers were associated with less risk of cognitive decline (Yaffe et al., 2008). Differently to all these previous published genetic association studies between PPAR-␥ gene and AD exclusively based on single SNP (rs1801282) analyses, Helisalmi et al. (2008) examined multiple SNPs that might have much greater potential to prove the hypothesis that there is a genetic link between AD and PPAR-␥ gene. Helisalmi et al. (2008) examined eight polymorphic sites (rs2920505, rs17036314, and rs2028759 in intron 1; rs1801282 in intron 2; rs2938395 in intron 3; and rs709149 and rs1797912 in intron 6) in PPAR-␥ gene and failed to observe any allele, genotype or haplotype association with AD either in the whole study or in APOE 4-stratified subgroups in the Finnish population. We found a common protective PPAR-␥ haplotype (TCCA) composed of intron 4 (rs4135263), intron 5 (rs1151999), and intron 6 (rs709149 and rs709154) polymorphisms which accounted for an estimated 39% of cases and 50% of controls among APOE 4 allele noncarriers. The previously reported Pro12Ala (C/G, rs1801282) polymorphism, in exon/intron 2, is at 57.2 kb distance from our most significantly associated htSNP (rs709149) in intron 6. Based on Hap Map data for Caucasians, these two SNPs are not in strong LD (D = 0.42; r2 = 0.009), and consequently, they are located in different haplotype blocks: rs1801282 is included in block 1 which we found not to be associated to AD, and rs709149 is in block 2. Characterized by our markers in introns 4, 5 and 6, block 2 includes most PPAR-␥ gene exons (Fig. 1), and therefore, we hypothesize that some yet unidentified functional variant, perhaps affecting post-transcriptional processing and the expression level of splice variants, might be tagged by the TCCA block 2 haplotype. Helisalmi et al. (2008) had a greater sample size that ours one and they also
O. Combarros et al. / Neurobiology of Aging 32 (2011) 547.e1–547.e6
stratified their population according to APOE genotype without finding an association of the rs709149 polymorphism; since the APOE 4, 3 and 2 allele frequencies differ in Finland and Spain, it would be interesting to understand if these variations contribute to the obtention of these different results. Brain extracts from AD patients contain less PPAR-␥ protein than controls (Sastre et al., 2006; Xiong et al., 2008); therefore, investigation of whether PPAR-␥ expression levels vary with different PPAR-␥ haplotypes would help to clarify their potential relevance to AD. It is tempting to speculate that subjects carrying the PPAR-␥ haplotype (TCCA) characterized by intron 4 (rs4135263), intron 5 (rs1151999), and intron 6 (rs709149 and rs709154) polymorphisms might show a lower AD risk due to an increased expression of PPAR-␥ in the brain. Indeed, PPAR␥ suppresses microglial inflammatory responses (Combs et al., 2000), and also stimulates cellular cholesterol efflux (Chinetti et al., 2001), which in turn could decrease AD risk. Our data suggest that PPAR-␥ genetic variants may modify the risk of AD in an APOE 4 allele-dependent fashion: the protective effect against AD of the PPAR-␥ TCCA haplotype was only observed in the absence of the APOE 4 allele. We acknowledge the possibility of false positive observation mainly due to stratification with APOE, so we have used an over-conservative Bonferroni correction for all 21 tests performed consisting of 7 SNPs with 3 subgroups (total sample, APOE 4 allele carriers and APOE 4 allele noncarriers). APOE 4 allele, the major genetic risk factor for AD, is associated with enhanced brain inflammation: following intracerebroventricular injection of LPS, microglial and NF-B activation are more pronounced in the hippocampi of transgenic APOE4 than in APOE3 mice (Ophir et al., 2005). Additionally, using cultured astrocytes prepared from human APOE3 and APOE4 knock-in mice, the amount of lipids released from APOE3- and APOE4-expressing astrocytes is APOE-isoform-dependent, with the order of potency being APOE3 > APOE4 (Gong et al., 2002); so, it is possible to speculate that cholesterol levels in astrocytes of APOE4 carriers are higher than in those of APOE3 carriers, leading to increased generation of A in the brain of APOE4 carriers. As the protective effect against AD of the PPAR-␥ introns 4, 5 and 6 haplotype (TCCA) might be counteracted by the increased inflammation and cholesterol retention in the brain associated to the APOE 4 allele, we hypothesize that the association between PPAR-␥ and AD should be only evident in the absence of the APOE 4 allele. In fact, more than 500 patients with mild to moderate AD were treated for 6 months with the PPAR-␥ agonist rosiglitazone, resulting in a statistically significant improvement in cognition in those patients that did not possess an APOE 4 allele, whereas patients with APOE 4 did not respond to the drug (Risner et al., 2006). Additional studies using different sets of patients and control subjects are required to verify the validity of our hypotheses (genetic and biologic).
547.e5
5. Conflict of interest The authors hereby declare that they have no conflicts of interest, and that all human studies have been approved by the appropriate ethics committee and have, therefore, performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki
Acknowledgement The authors thank Weyma Notel’s help in Fig. 1 edition.
References Chinetti, G., Lestavel, S., Bocher, V., Remaley, A.T., Neve, B., Torra, I.P., Teissier, E., Minnich, A., Jaye, M., Duverger, N., Brewer, H.B., Fruchart, J.C., Clavey, V., Staels, B., 2001. PPAR-␣ and PPAR-␥ activators induce cholesterol removal from human macrophage foam cells through stimulation of the ABCA1 pathway. Nat. Med. 7, 53–58. Combs, C.K., Johnson, D.E., Karlo, J.C., Cannady, S.B., Landreth, G.E., 2000. Inflammatory mechanisms in Alzheimer’s disease: inhibition of -amyloid-stimulated proinflammatory responses and neurotoxicity by PPAR␥ agonists. J. Neurosci. 20, 558–567. Gong, J.S., Kobayashi, M., Hayashi, H., Zou, K., Sawamura, N., Jujita, S.C., Yanagisawa, K., Michikawa, M., 2002. Apolipoprotein E (ApoE) isoform-dependent lipid release from astrocytes prepared from human ApoE3 and ApoE4 knock-in mice. J. Biol. Chem. 277, 29919–29926. Hamilton, G., Proitsi, P., Jehu, L., Morgan, A., Williams, J., OˇıDonovan, M.C., Owen, M.J., Powell, J.F., Lovestone, S., 2007. Candidate gene association study of insulin signaling genes and Alzheimer’s disease: evidence for SOS2, PCK1, and PPARgamma as susceptibility loci. Am. J. Med. Genet. B: Neuropsychiatr. Genet. 144B, 508–516. Helisalmi, S., Tarvainen, T., Vepsäläinen, S., Koivisto, A.M., Hiltunen, M., Soininen, H., 2008. Lack of genetic association between PPARG gene polymorphisms and Finnish late-onset Alzheimer’s disease. Neurosci. Lett. 441, 233–236. Heneka, M.T., Landreth, G.E., 2007. PPARs in the brain. Biochim. Biophys. Acta 1771, 1031–1045. Kaltschmidt, B., Uherek, M., Volk, B., Baeuerle, P.A., Kaltschmidt, C., 1997. Transcription factor NF-B is activated in primary neurons by amyloid  peptides and in neurons surrounding early plaques from patients with Alzheimer disease. Proc. Natl. Acad. Sci. U.S.A. 94, 2642–2647. Koivisto, A.M., Helisalmi, S., Pihlajamäki, J., Hiltunen, M., Koivisto, K., Moilanen, L., Kuusisto, J., Helkala, E.L., Hänninen, T., Kervinen, K., Kesäniemi, Y.A., Laakso, M., Soininen, H., 2006. Association analysis of peroxisome proliferators-activated receptor gamma polymorphisms and late onset Alzheimer’s disease in the Finnish population. Dement. Geriatr. Cogn. Disord. 22, 449–453. Landreth, G., Jiang, Q., Mandrekar, S., Heneka, M., 2008. PPAR␥ agonists as therapeutics for the treatment of Alzheimer’s disease. Neurotherapeutics 5, 481–489. McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., Stadlan, E.M., 1984. Clinical diagnosis of Alzheimer disease: report of the NINCDS-ADRDA work group under the auspices of the Department of Health and Human Services Task Force on Alzheimer Disease. Neurology 24, 939–944. Ophir, G., Amariglio, N., Jacob-Hirsch, J., Elkon, R., Rechavi, G., Michaelson, D.M., 2005. Apolipoprotein E4 enhances brain inflammation by modulation of the NF-B signaling cascade. Neurobiol. Dis. 20, 709–718. Puglielli, L., Tanzi, R.E., Kovacs, D.M., 2003. Alzheimer’s disease: the cholesterol connection. Nat. Neurosci. 6, 345–351.
547.e6
O. Combarros et al. / Neurobiology of Aging 32 (2011) 547.e1–547.e6
Risner, M.E., Saunders, A.M., Altman, J.F., Ormandy, G.C., Craft, S., Foley, I.M., Zvartau-Hind, M.E., Hosford, D.A., Roses, A.D., Rosiglitazone in Alzheimer’s disease study group, 2006. Efficacy of rosiglitazone in a genetically defined population with mild-to-moderate Alzheimer’s disease. Pharmacogenomics J. 6, 246–254. Sastre, M., Dewachter, I., Rossner, S., Bogdanovic, N., Rosen, E., Borghgraef, P., Evert, B.O., Dumitrescu-Ozimek, L., Thal, D.R., Landreth, G., Walter, J., Klockgether, T., van Leuven, F., Heneka, M.T., 2006. Nonsteroidal anti-inflammatory drugs repress -secretase gene promoter activity by the activation of PPAR␥. Proc. Natl. Acad. Sci. U.S.A. 103, 443–448. Sauder, S., Kölsch, H., Lütjohann, D., Schulz, A., von Bergmann, M., Maier, W., Heun, R., 2005. Influence of peroxisome proliferators-activated receptor gamma gene polymorphism on 24S-hydroxycholesterol levels in Alzheimer’s patients. J. Neural Transm. 112, 1381–1389. Scacchi, R., Pinto, A., Gambina, G., Rosano, A., Corbo, R.M., 2007. The peroxisome proliferator-activated receptor gamma (PPAR-␥2) Pro12Ala polymorphism is associated with higher risk for Alzheimer’s disease in octogenarians. Brain Res. 1139, 1–5.
Shobab, L.A., Hsiung, G.Y., Feldman, H.H., 2005. Cholesterol in Alzheimer’s disease. Lancet Neurol. 4, 841–852. van Neerven, S., Kampmann, E., Mey, J., 2008. RAR/RXR and PPAR/RXR signalling in neurological and psychiatric diseases. Prog. Neurobiol. 85, 433–451. West, N.A., Haan, M.N., Morgenstern, H., 2008. The PPAR-gamma Pro12Ala polymorphism and risk of cognitive impairment in a longitudinal study. Neurobiol. Aging, doi:10.1016/j.neurobiolaging.2008.06.005. Wyss-Coray, T., Mucke, L., 2002. Inflammation in neurodegenerative disease: a double-edged sword. Neuron 35, 419–432. Xiong, H., Callaghan, D., Jones, A., Walker, D.G., Lue, L.F., Beach, T.G., Sue, L.I., Woulfe, J., Xu, H., Stanimirovic, D.B., Zhang, W., 2008. Cholesterol retention in Alzheimer’s brain is responsible for high and ␥-secretase activities and A production. Neurobiol. Dis. 29, 422– 437. Yaffe, K., Kanaya, A.M., Lindquist, K., Hsueh, W.C., Cummings, S.R., Beamer, B., Newman, A., Rosano, C., Li, R., Harris, T., Health ABC Study, 2008. PPAR-␥ Pro12Ala genotype and risk of cognitive decline in elders. Neurobiol. Aging 29, 78–83.