B R A I N R E S E A R C H 1 3 7 1 ( 2 0 1 1 ) 1 6 1 – 1 7 0

ava i l ab l e a t www.sc i enced i r ec t . com

www.e l sev i e r . com/ loca te /b ra i n res

Research Report

Atorvastatin and pitavastatin improve cognitive function andreduce senile plaque and phosphorylated tau in aged APPmice

Tomoko Kurata, Kazunori Miyazaki, Miki Kozuki, Violeta-Lukic Panin,Nobutoshi Morimoto, Yasuyuki Ohta, Makiko Nagai, Yoshio Ikeda,Tohru Matsuura, Koji Abe⁎

Department of Neurology, Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama University, 2-5-1 Shikatacho,kitaku, Okayama 700-8558, Japan

A R T I C L E I N F O

⁎ Corresponding author. 2-5-1 Shikatacho, kitE-mail address: toko11@cc.okayama-u.acAbbreviations: Aβ, amyloid-β peptide; AD,

weight; BW, body weight; CNS, central nervglutaryl coenzyme A; LDL-C, low-density lipphosphorylated tau-positive dystrophic neur

0006-8993/$ – see front matter © 2010 Elsevidoi:10.1016/j.brainres.2010.11.067

A B S T R A C T

Article history:Accepted 18 November 2010Available online 25 November 2010

In addition to simply reducing the serum level of cholesterol, 3-hydroxy-3-methyl glutarylcoenzyme A (HMG-CoA) reductase inhibitors (statins) have various pleiotrophic effects suchas reducing oxidative stress, neuroinflammation, and neurotoxicity. However, such apleiotrophic effect has not been fully studied in a new statin (pitavastatin). We examinedand compared the effects of two strong statins (atorvastatin, 30 mg/kg/day, p.o.;pitavastatin, 3 mg/kg/day, p.o.) on the serum level of lipids, cognitive dysfunction, senileplaque (SP) and phosphorylated tau-positive dystrophic neuritis (pτDN) in amyloidprecursor protein (APP) transgenic (Tg) mice from 5months (M) of age to 20 M. These twostatins improved behavioral memory and reduced the numbers of SP and pτDN at 15 and20 M without affecting serum lipid levels, but preserved mice brain weight in pitavastatingroup at 20 M. These protective effects of statins took 10 M from the beginning of treatmentto show an improvement in the present model mice, and sensitivity to the statin treatmentwas linked to behavioral memory, SP and pτDN in this order. These findings suggest thatearly treatment with both atorvastatin and pitavastatin prevented subsequent worsening ofcognitive function and the amyloidogenic process, probably due to pleiotrophic effects,suggesting a therapeutic potential for Alzheimer's disease (AD).

© 2010 Elsevier B.V. All rights reserved.

Keywords:Alzheimer's diseaseTransgenic mouseAtorvastatinPitavastatin

1. Introduction

Alzheimer's disease (AD) is the most common neurodegener-ative dementia disease in the aged. AD is neuropathologicallycharacterized by abnormal accumulations of amyloid plaquesand neurofibrillary tangles (NFTs) throughout cortical and

aku, Okayama 700-8558, J.jp (K. Abe).Alzheimer's disease; APPous system; HDL-C, highoprotein cholesterol; M,ite; SP, senile plaque; T-c

er B.V. All rights reserved

limbic brain regions. Cognitive dysfunction in AD is widelybelieved to result from progressive synaptic dysfunction andneurodegenration initiated by soluble aggregated amyloid-βpeptide (Aβ) and further involves aggregates of phosphorylatedtau (pτ), a principal component of NFTs. Transgenic (Tg) micethat overexpress mutant amyloid β protein precursor (βAPP) in

apan. Tel.: +81 86 235 7365; fax: +81 86 235 7368.

, amyloid precursor protein; 8-ARM, radial 8 arm maze; BrW, brain-density lipoprotein cholesterol; HMG-CoA, 3-hydroxy-3-methyl

month; MC, methylcellulose; NFTs, neurofibrillary tangles; pτDN,ho, total cholesterol; Tg, transgenic; TG, triglyceride

.

162 B R A I N R E S E A R C H 1 3 7 1 ( 2 0 1 1 ) 1 6 1 – 1 7 0

thebrainhad substantialAβamyloiddeposits (Hsiao et al., 1996;Kawarabayashi et al., 2001) with memory disturbance andsubsequent typical pathological characteristics of AD.

Evidence from cell culture experiments and animal studiesrecently suggested a strong relationship between serumcholesterol level and AD. Diet-induced hypercholesterolemiaresulted in significantly increased levels of formic acid-extractable Aβ peptides in the central nervous system (CNS),whereas the levels of total Aβ were strongly correlated withthe levels of both plasma and CNS total cholesterol (Refoloet al., 2000). A cholesterol-lowering drug, BM15.766, reducedplasma cholesterol, brain Aβ peptides, and the brain Aβ-loadbymore than 2-fold (Refolo et al., 2001). However, treatment ofguinea pigs with high-dose simvastatin led to a strongreduction in Aβ1–40 and Aβ1–42 levels in brain homogenateswithout affecting total brain cholesterol levels (Fassbenderet al., 2001), and treatment of presenilin-1(PS-1)(M146V) × APP(K670N)/APP(M671L) human double knock-in (PSAPP) Tg micewith atorvastatin markedly reduced the brain Aβ content byabout 38–52% with the brain cholesterol level reduced only by10% (Petanceska et al., 2002). These effects of statins suggest apositive ‘pleiotrophic effect’ in the AD brain, in addition tosimply reducing the serum level of cholesterol.

Atorvastatin is a strong 3-hydroxy-3-methyl glutaryl coen-zyme A (HMG-CoA) reductase inhibitor currently in worldwideclinical use, which reduces brain Aβ. On the other hand,pitavastatin is a novel potent inhibitor of HMG-CoA reductaseand showed a strong effect of lowering plasma total choles-terol (T-cho) and triglycerides (TG); 10 times stronger thansimvastatin or pravastatin in reducing T-cho levels in animals(Kajinami et al., 2000; Suzuki et al., 1999). Because these twostatins have a very strong effect against hyperlipidemia inhumans and animals, both could reduce accumulation of Aβin the brain and improve clinical behavioral memory scores.However, there has been no study of pitavastatin in terms ofeffects on cognitive function and brain Aβ levels using an ADmodel animal.

We therefore examined and compared the efficacy ofatorvastatin and pitavastatin on serum levels of lipids,cognitive function, brain Aβ levels, and pτ.

2. Results

2.1. Physical parameters

At 5 M, BW was 20.9±2.3 g (n=19, mean±SD) in the non-Tgcontrol group, 16.5±2.6 g (n=18) in the APP vehicle group,17.0±2.3 g (n=16) in the APP atorvastatin group, and 17.2±1.7 g(n=16) in the APP pitavastatin group (Table 1). Body weightwas significantly lower in all three APP groups (**p<0.01) thanthe non-Tg control group at 5 M, but there were no differenceswithin three APP groups. At 10, 15 and 20 M, BW was slightlyincreased, but was not different among the four groups. BrWwas not different among the four groups at 10 and 15 M, butwas significantly decreased in the APP vehicle group at 20 Mcompared to the non-Tg control group (**p<0.01). BrW in theAPP pitavastatin and APP atorvastatin groups showed asignificant improvement (#p<0.05) compared to the APPvehicle group at 20 M (Table 1).

2.2. Behavioral memory

At 10 M, the mice in all four groups took less time to eat alleight pellets (Fig. 1A) and had a lower number of errorchoices (Fig. 1C) until 8th day of each trial, but there was nochange in the number of correct choices during the 8 days(Fig. 1B). A two-way repeated measure analysis of varianceshowed a significant time saving in the non-Tg control groupat blocks 1, 3, 6 and 7 (+p<0.05) compared to the APP vehiclegroup, but not to the two APP statin groups (Fig. 1A).Significant differences were not recognized in the numberof correct choices or the number of error choices among thefour groups (Fig. 1B, C).

At 15 M, the mice in the non-Tg control and two APP statingroups gradually took less the time, but the mice in the APPvehicle group did not (Fig. 1D). In detail, a two-way repeatedmeasure analysis of variance showed a significant time savingin the non-Tg control and APP pitavastatin groups at blocks 1,4 and 7 (#p<0.05), and in the non-Tg control and two APP statingroups at block 8 (*p<0.05) compared with the APP vehiclegroup (Fig. 1D). There was nomarked difference in the numberof correct choices between the non-Tg control and two APPstatin groups, while the APP vehicle group showed a decreaseat 8 blocks (*p<0.05) compared with the non-Tg control andtwo APP statin groups (Fig. 1E). All four groups graduallyreduced the number of errors, but the score of errors wassignificantly higher in the APP vehicle group in all 8 blocks(*p<0.05) compared with the non-Tg control and two APPstatin groups (Fig. 1F).

At 20 M, the mice of the non-Tg control and two APP statingroups gradually reduced the time during the 8 consecutivedays, but the APP vehicle group did not (Fig. 1G). The non-Tgcontrol and APP pitavastatin groups showed significant timesavings at blocks 3, 5, 6 and 7 (#p<0.05) compared with the APPvehicle group (Fig. 1G). No marked difference was found in thenumber of correct choices in the non-Tg control and two APPstatin groups, except for the APP vehicle group with a declineat blocks 3 and 4 (*p<0.05) compared with the other 3 groups(Fig. 1H). As for the number of errors, all four groups graduallyreduced them, and the number of errors was significantlyhigher in the APP vehicle group at blocks 3, 4, 5, 7 and8 (*p<0.05) compared with the other 3 groups and at block 6(#p<0.05) compared with the non-Tg control and APP pitavas-tatin groups (Fig. 1I).

2.3. Biochemical parameters

The serum TG, T-cho and LDL-C levels were not significantlydecreased in the non-Tg control and APP vehicle groupscompared with the three APP statin groups at 10, 15 and 20 M.On the other hand, the serum HDL levels and the ratio of LDLandHDL (L/H ratio) tended to be lower in the APP vehicle groupthan the non-Tg control and two APP statin groups, but thedifferences were not significant (Table 1).

2.4. Immunohistochemical analyses

SP consisted of a core and surrounding DNs. In non-Tg controlmouse brains, SP was not detected at any ages. In all three APPgroups, a number of SP cores were stained with the 4G8

Table 1 – Physical and biochemical parameters in non-Tg control and APP mice.

Non-Tg control group APP vehicle group APP atorvastatin group APP pitavastatin group

5 M(n=19)

10 M(n=8)

15 M(n=6)

20 M(n=5)

5 M(n=18)

10 M(n=7)

15 M(n=6)

20 M(n=5)

5 M(n=16)

10 M(n=7)

15 M(n=4)

20 M(n=5)

5 M(n=16)

10 M(n=5)

15 M(n=5)

20 M(n=6)

Body weight (g) 20.9±2.3 23.8±1.2 26.2±1.7 24.7±1.5 16.5±2.6 ⁎⁎ 24.3±1.5 24.6±1.3 20.6±0.7 17.0±2.3 ⁎⁎ 24.6±1.8 22.3±0.8 20.1±1.7 17.2±1.7 ⁎⁎ 21.8±1.2 23.8±0.5 23.3±1.3Brain weight (g) 0.43±0.01 0.42±0.02 0.47±0.01 0.43±0.01 0.44±0.01 0.40±0.01 ⁎⁎ 0.43±0.01 0.43±0.01 0.43±0.01 0.42±0.01 0.44±0.03 0.46±0.01 #

TG (mg/dL) 68.8±14.3 66.3±8.6 58.8±11.2 20.4±3.1 36.3±9.4 17.2±2.6 30.1±3.5 70.3±28.4 22.0±7.5 31.0±5.9 28.8±15.1 29.0±7.5T-cho (mg/dL) 61.1±5.4 59.8±6.4 85.0±8.3 52.0±5.2 67.8±4.4 58.4±7.4 57.6±1.6 65.5±6.0 63.4±5.9 57.2±8.0 62.4±4.3 66.7±5.4HDL-cho (mg/dL) 30.6±4.6 29.7±5.1 40.3±3.8 31.3±5.1 37.0±5.0 25.0±1.0 36.6±3.2 41.7±3.8 45.0±9.1 38.3±8.0 43.0±0.6 44.8±4.0LDL-cho (mg/dL) 16.4±2.1 10.5±2.1 18.3±2.8 11.4±0.8 18.0±3.1 12.5±2.5 10.9±0.9 14.0±0.6 12.0±2.1 11.5±1.2 12.7±1.3 12.8±1.8Ratio of LDLand HDL

0.63±0.43 0.38±0.17 0.47±0.21 0.43±0.18 0.56±0.40 0.50±0.17 0.32±0.13 0.34±0.07 0.27±0.05 0.35±0.16 0.30±0.06 0.29±0.11

⁎⁎ p<0.01 significant difference compared with the non-Tg control group at the same age.# p<0.05 significant difference compared with the APP vehicle group at the same age.

163BR

AIN

RESEA

RC

H1371

(2011)

161–170

Fig. 1 – Behavioral memory scores on the 8-ARM test at 10, 15 and 20 M. Running time and numbers of correct and error choicesare shown in mice in the non-Tg (○), APP vehicle group (×), APP atorvastatin group (▲), APP pitavastatin group (■). Noteprogressive worsening in the APP vehicle group and improvements in the atorvastatin and pitavastatin groups. +p<0.05;significant difference in the non-Tg group vs APP vehicle group. *p<0.05; significant improvement in the non-Tg and two APPstatin groups vs the APP vehicle group. #p<0.05; significant improvement in the non-Tg and APP pitavastatin groups vs APPvehicle group.

164 B R A I N R E S E A R C H 1 3 7 1 ( 2 0 1 1 ) 1 6 1 – 1 7 0

antibody, and these plaques spread throughout the cerebralcortices from 10 M with increasing numbers at 15 and 20 M(Fig. 2A). The number of SP was small at 10 M, and nodifference was detected among the three APP groups(Fig. 2A, a–c, arrowheads, and B). At 15 and 20 M, SPdramatically increased both in number and size with age inall three APP groups, but the number and size of SP in thecerebral corticeswas smaller in the twoAPP statin groups thanin the APP vehicle group (Fig. 2A, d-i).

Quantitative analysis showed that the sums of each SPsize in 1 mm2 of the cerebral cortex were 1.06±1.72×103 μm2

in the APP vehicle group, 1.46±1.09×103 μm2 in the APPatorvastatin group, and 1.47±1.33×103 μm2 in the APP pita-

vastatin group at 10 M, respectively, and significant differ-ences were not recognized among the three APP groups(Fig. 2B). At 15 M, the sum of each SP size was 21.41±11.31×103 μm2 in the APP vehicle group, 6.23±4.54×103 μm2 inthe APP atorvastatin group (*p<0.05 vs APP vehicle group),and 5.59±4.54×103 μm2 in the APP pitavastatin group(**p<0.01 vs APP vehicle group) (Fig. 2B). At 20 M, the sum ofeach SP size was 61.48±16.49×103 μm2 in the APP vehiclegroup, 30.85±7.79×103 μm2 in the APP atorvastatin group(**p<0.01 vs APP vehicle group), and 40.50±9.61×103 μm2 inthe APP pitavastatin group(*p<0.05 vs APP vehicle group)(Fig. 2B). APP statin groups showed no significant difference toeach other.

165B R A I N R E S E A R C H 1 3 7 1 ( 2 0 1 1 ) 1 6 1 – 1 7 0

2.5. Pτ analysis

Pτ was not detected at any age in non-Tg control mice brains,while pτ was clearly labeled in DNs around the SP core by theAT8 antibody in all three APP groups (Fig. 3A). Only a few pτDNwere detected per slice and their sizes were very small in allthreeAPP groupsat 10 M (Fig. 3A, a–c, arrowheads). Quantitativeanalysis showed that thenumbers ofpτDNwere1.28±1.49/SP in

Fig. 2 – Representative photomicrographs of Aβ stainings (A) andstronger and larger plaques in brains of APP vehicle mice and thbar=100 μm). Mice groups N, Ve, At and Pi represent the non-Tgrespectively. (*p<0.05, **p<0.01 vs APP vehicle group).

the APP vehicle group, 1.90±2.51/SP in the APP atorvastatingroup and 0.26±0.64/SP in the APP pitavastatin group at 10 M,and significant differences were not recognized among thethree APP groups (Fig. 2B). At 15 M, pτDNs were stained muchstronger than those at 10 M inall threeAPP groups. Quantitativeanalysis showed that thenumbersofpτDNwere3.27±3.35/SP inthe APP vehicle group, 1.51±2.95/SP in the APP atorvastatingroup and 3.11 ± 2.97/SP in the APP pitavastatin group. The

sums of senile plaque (SP) size (B) at 10, 15 and 20 M. Notee improvements in two APP statin mice brains (A, scale, APP vehicle, APP atorvastatin or APP pitavastatin groups,

166 B R A I N R E S E A R C H 1 3 7 1 ( 2 0 1 1 ) 1 6 1 – 1 7 0

number of pτDNswas smaller in the two statin group than thatin the APP vehicle group, but the difference did not reach thesignificant level at 15 M (Fig. 3A, d–f and B). At 20 M, the numberof pτDN increased in each SPmoremarkedly in the APP vehiclegroup than in the two APP statins groups (Fig. 3A g–i).Quantitative analysis showed that the numbers of pτDN were8.91±3.05/SP in the APP vehicle group, 5.58±1.30/SP in the APPatorvastatin group and 4.86±2.23/SP in the APP pitavastatingroup (* p<0.05 vs APP vehicle group) (Fig. 3B).

Fig. 3 – Representative photomicrographs of pτ stainings (A) andstaining of a small subset of dystrophic neurites stained with angroups N, Ve, At and Pi represent the non-Tg, APP vehicle, APP atAPP vehicle group).

3. Discussion

In this study, BW was significantly lower in all three APPgroups (**p<0.01) than the non-Tg control group at 5 M, butthere were no differences within three APP groups at 5 M. At10, 15 and 20 M, BW was not different among the four groups.BrWwas significantly smaller in the APP vehicle group than inthe non-Tg control group at 20 M (**p<0.01), but the APP

numbers of pτDN per SP (B) at 10, 15 and 20 M. Note strongerti-pτ (A, scale bar=100 μm, arrowheads and inserts). Miceorvastatin or APP pitavastatin group, respectively (*p<0.05 vs

167B R A I N R E S E A R C H 1 3 7 1 ( 2 0 1 1 ) 1 6 1 – 1 7 0

pitavastatin group showed a significant improvement com-pared to the APP vehicle group (Table 1, #p<0.05). In our study,neither atorvastatin nor pitavastatin treatment affectedserum triglycerides, total cholesterol, HDL or LDL levels at10 M, 15 M and 20 M (Table 1), while the L/H ratio showed adecreasing tendency in the APP statin groups compared to theAPP vehicle group (Table 1).

Analysis of behavioralmemory showed an improvement at15 and 20 M in the two APP statin groups (Fig. 1 D, G, F, I).Several clinical studies reported that treatment with statins(atorvastatin, simvastatin, cerivastatin, fluvastatin, lovastat-in, and pravastatin) decreased the prevalence of Alzheimer'sdisease (AD) (Jick et al., 2000; Wolozin et al., 2000; Carlssonet al., 2008) and improved cognitive function and neuropath-ological markers such as neuritic plaque and neurofibrillarytangles (NFT) in AD patients (Sparks et al., 2006; Li et al., 2007).Treatment of mice with high-dose simvastatin for 3 monthsenhanced memory in both APP and non-Tg control mice (Liet al., 2006), significantly decreased brain oxidative stress, glialactivation, cortical soluble Aβ levels and the number of Aβplaque-associated dystrophic neurites (Tong et al., 2009), andeffectively counteracted astroglial and microglial activation,in accord with the recognized role of statins in preventingexpression and secretion of proinflammatory cytokines fromactivated glial cells (Cordle and Landreth, 2005). As theseeffects on the patients and animals were independent of thecholesterol-lowering action of statins (Jick et al., 2000;Nachtigal et al., 2006), the benefits probably arise from theirpleotropic effects (Cordle and Landreth, 2005; Cimino et al.,2007). The present study also showed that both atorvastatinand pitavastatin improved cognitive function without affect-ing serum lipid levels at 15 and 20 M, again suggesting theirpleiotrophic effects.

As shown in Fig. 2, average areas of SP detected by Aβstaining in the cerebral cortexwere significantly smaller in thetwo APP statin groups than in the APP vehicle group at 15 and20 M (Fig. 2A, d-i, and B), which supports several previousstudies showing reduced Aβ levels by lovastatin, pravastatinand simvastatin in APP mice (Chauhan et al., 2004; Li et al.,2006). A recent report found that simvastatin reduced thebrain level of soluble Aβ at 12 M in Tg 2576 mice treated from9 M (before Aβ deposition), but not if initiated at 14 Mwhen Aβwas already deposited (Li et al., 2006). In the present study,treatment of mice with statins started from 5 M (again beforeAβ deposition), and the average areas of SP were significantlysmaller in the two APP statin groups than in the APP vehiclegroup at 15 and 20 M (Fig. 2), which is well correlated with thebehavior memory at these ages in the mice (Fig. 2).

In the present study, atorvastatin showed a reducingtendency in the number of pτDN per SP at 15 and 20 M(Fig. 3B), and pitavastatin showed a significant reduction inthe number of pτDN per SP in APP (*p<0.05) at 20 M comparedto the APP vehicle group (Fig. 3B). These observations areinteresting regarding the mechanism of statins to prevent ADbecause phosphorylated tau and its aggregates, NFTs, corre-late with dementia (Arriagada et al., 1992). A previous studyshowed a marked reduction in phosphorylation of tau withatorvastatin or simvastatin in both normocholesterolemic andhypercholesterolemic mice (Boimel et al., 2009), again sug-gesting a direct effect of statins on tau pathology regardless of

plasma cholesterol levels. Our present study suggests that theeffect of reducing the pτDN number by pitavastatin andatorvastatin could be related to preserving BrW and clinicalscore (Table 1 and Fig. 1G–I).

In this study, treatment with statins started from 5 M of age,but the effect of statins to improve functional memory, reduceSP and reduce pτDN became evident at 15 and 20 M (Figs. 2–4).Thus, the protective effect of statins took 10 M even in thepresentmodelmice, andbehavioralmemory, SPandpτDNweresensitive to statin treatment in this order. Growing evidenceindicates that although tau is an essentialmediator of cognitivedecline, Aβ accumulation is a critical upstream event. Apertinent study to support this shows that cognitive deficits inan AD mouse model are attenuated by crossing AD miceoverexpressing APP with tau−/− mice: AD/tau−/− mice exhibitunaltered Aβ levels compared with AD/tau+/+ mice (Robersonet al., 2007). This work demonstrates a critical role for tau incognitive decline, but also demonstrates that cognitive deficitsare only observed in ADmicewith elevated Aβ levels. In amorerecent study, it was shown that exposure to Aβ induces tauphosphorylation (Hu et al., 2008). These studies thereforedemonstrate an intimate relationship between Aβ accumula-tionand tauphosphorylation in thecognitivedeficits ofAD.Andso, the decrease of pτDN number was detected only in20months, though the improvement of clinical score and thereduction of the area of SP were apparent in 15month.

In this study, we examined and compared the efficacy ofatorvastatin and pitavastatin on BW, BrW, serum levels oflipids, cognitive function, brain Aβ levels, and pτ, and thesignificant difference between them was not detected. How-ever, the APP pitavastatin group showed a little better effectabout the time savings in 8-ARM test and the numbers of pτDNthan the APP atorvastatin group at 20 M. The two statins areboth lipophilic, but they differ in some points. Although moststatins such as atorvastatin are hepaticallymetabolized by thecytochrome P450 (CYP) 3A4 isoenzyme, pitavastatin is princi-cally metabolized by CYP 2 C9. Pitavastatin was found toenhance LDL receptor expression in vitro studies involvinghepatoblastoma cells, as well as the amount of LDL binding tothe LDL receptor. Pitavastatin also exhibited more potentinduction of LDL receptor mRNA expression compared withatorvastatin (Morikawa et al., 2000; Nomura et al., 2009). Andso, pitavastatin could have a little better potential forimproving cognitive function compared with atorvastatin,but this needs further study.

Several previous studies demonstrated that soluble oligo-mers of Aβ could be very important in synaptic and cognitivedysfunction in the early stages of AD (Klein et al., 2001; Selkoe,2002). In vitro studies demonstrated that an exogenously-applied Aβ oligomer induced tau phosphorylation (De Feliceet al., 2008) and caused neuronal death (Kayed et al., 2003;Lambert et al., 1998), and an in vivo study demonstrated thatan Aβ oligomer caused not only synaptic alteration, but alsoother features of typical AD pathology such as tau phosphor-ylation, microglial and astrocytic activations, and neuronalloss before amyloid deposition (Tomiyama et al., 2010). Thisstudy showed a good correlation between treatment withstatins and improvements in cognitive function and reductionin SP and pτDN (Figs. 2–4), but the correlation could be due tothe Aβ oligomer, and this needs further study.

168 B R A I N R E S E A R C H 1 3 7 1 ( 2 0 1 1 ) 1 6 1 – 1 7 0

The present findings suggest that treatment with 2 strongstatins was effective to achieve cognitive improvement andreduce the pivotal amyloidogenic process via their pleotrophiceffects, which took 10 M before showing a significant differ-ence in clinical (Table 1, Fig. 1) and pathological parameters(Figs. 2 and 3). Our results indicate that both atorvastatin andpitavastatin have high potential as a therapeutic approach toAD.

4. Experimental procedures

4.1. Animals and drug preparation

Fifty-four 12-week-old (3 months) female hemizygous Tgmiceexpressing a familial AD mutant, human APPK670N, M671L(line Tg2576) (B6.SJL) and 20 female non-Tg mice (B6.SJLHybrid) were provided from Taconic company (NY, USA) andplaced on a basal diet.

When the mice reached 5 M of age, the above non-Tg mice(n=20) were started on a daily dose of 0.5% methylcellulose(MC) in 0.1 ml water by oral gavage for the subsequent 5–15 M,and the above APPmice (n=54) were divided into the followingthree groups, i.e., APP vehicle group (n=17), APP atorvastatingroup (n=19) and APP pitavastatin group (n=18), receivingdaily oral doses of 0.5% MC only (APP vehicle group), 0.5% MCplus atorvastatin (30 mg/kg/day) or 0.5% MC plus pitavastatin(3 mg/kg/day) for the subsequent 5–15 months by oral gavage,respectively. Atorvastatin and pitavastatin were provided byPfizer Inc. (Grotan, CT) or Kowa Co. Ltd. (Tokyo, Japan),respectively, and were given to the mice as a suspensionwith 5% MC in 0.1 ml water everyday.

Every 5 M at 10, 15 or 20 M on age, data of the mice wereobtained for analyses of their physical parameters, behavioralmemory, and serum levels of TG, T-cho, high-density lipo-protein cholesterol (HDL-C), and low-density lipoproteincholesterol (LDL-C). The mice were then sacrificed underdeep anesthesia with pentobarbital (40 mg/kg, i.p.). Afterdecapitaion, their brains were removed and the brain weightswere measured. All experimental procedures were approvedby the Animal Committee of the Graduate School of Medicineand Dentistry, Okayama University.

4.2. Physical parameters, behavioral memory, and serumsubstances

As physical parameters, body weight (BW) and brain weight(BrW) were measured at 5, 10, 15 or 20 M and 10, 15 or 20 M,respectively, by a portable balance (Sartorius Mechatronics,Goettingen, Germany).

The behavioral memory was evaluated with a radial eight-arm maze (8-ARM) task, described in detail elsewhere (Okadaet al., 1995). In brief, food pellets were randomly scattered overthe entire maze surface. After adaptation, all mice weretrained once a day for 8 consecutive days, during which asingle food pellet was placed in each of the food cups of theeight arms. For each trial, a mouse was placed on the centerplatform facing a randomly selected arm and was allowed tomake arm choices until either all 8 pellets had been eaten or5 min had elapsed, whichever came first. The initial entry of

an arm was scored as a correct choice, whereas reentry to anarm previously visited was scored as an error. The runningtime elapsed before the mouse ate all 8 pellets, the number ofcorrect choices in the initial eight chosen arms and thenumber of errors (defined as choosing arms already visited)was assessed. At 10, 15 and 20 M, a block consisted of 3 trialsand each mouse was subjected to one trial daily for8 consecutive days. During the 3 weeks of the behavioralmemory examination, each mouse was placed on a partialfood-deprivation schedule designed to maintain the deficien-cy of BW within 10%, except for free access to water.

Serum levels of TG, T-cho, HDL-C, and LDL-C weremeasured by a standard biochemical method in SRL, Inc(Tokyo, Japan).

4.3. Immunohistochemistry

After measuring BrW, the brains were immersed and fixed in4% paraformaldehyde with 0.1 M phosphate buffer (PB, pH 7.6)for 8 h, embedded in paraffin, and 5-μm-thick sections wereprepared for subsequent immunostaining. For immunostain-ings, the brain sections were pretreated with formic acid for3 min for both Aβ and phosphorylated tau (pτ) stainings,followed by heating 3 times with a 500Wmicrowave for 5 minin 10 mM (pH 6.0) citric acid buffer for only pτ staining. Thesepretreated sections were then immersed in 0.5% periodic acidto block intrinsic peroxidase, and treated with 5% normalhorse serum in 50 mM phosphate-buffered saline (PBS, pH 7.4)containing 0.05% Tween 20 to block any non-specific antibodyresponse, and were finally incubated overnight with eachprimary antibody. The following primary antibodies wereused in this study: monoclonal antibody 4G8 (1:1000, Signet,Dedham, MA, USA) for Aβ immunostaining, and AT8 (1:400,Innogenetics, Gent, Belgium) for pτ at serine202/threonine205. A biotinylated anti-mouse antibody was used as thesecond antibody, and specific labeling was visualized by aVectastain Elite ABC kit (Vector, Burlingame, CA). To guaran-tee specific staining primary antibodies, brain sections werealso stained without primary antibodies.

4.4. Detection and analyses

The above stained sections were digitized with a digitalmicroscope camera (Olympus BX-51; Olympus Optical Co.,Japan). After acquisition with a digital camera, the file for theexperimental image is opened in Photoshop (Adobe; San Jose,CA). The region of interest is defined in Photoshop using theMarquee tool. This tool allows an area of any pixel dimensionto be determined by the operator by using the “optionswindow.” To do this, the appropriately sized region of interestis identified, and then the area of senile plaque (SP) detectedby 4G8 per each 1 mm2 cerebral cortex and number of pτpositive dystrophic neurites (pτDN) per SP in hemisphericcoronal sections were counted at three levels (frontal cortex,basal ganglia, and posterior hippocampus). Three serialsections were used for each level and all deposit counts inthe cerebral cortex per animal were determined and added.

Data are expressed as mean±SD. Statistical analyses wereperformed using ANOVA with repeated measures (multiplecomparisons). Planned comparisons were used for Tukey–

169B R A I N R E S E A R C H 1 3 7 1 ( 2 0 1 1 ) 1 6 1 – 1 7 0

Kramer post-hoc analysis. P<0.05 was considered significant.All statistical analyzes were performed with Statcel statisticalpackage (Statcel 2; OMS Inc., Tokorozawa, Japan).

Acknowledgments

Thisworkwas partly supported by a Grant-in-Aid for ScientificResearch (B) 21390267 and the Ministry of Education, Science,Culture and Sports of Japan, and by Grants-in-Aid from theResearch Committee of CNS Degenerative Diseases(Nakano I),and grants (Itoyama Y, Imai T, Sobue G) from the Ministry ofHealth, Labour and Welfare of Japan.

R E F E R E N C E S

Arriagada, P.V., Growdon, J.H., Hedley-Whyte, E.T., Hyman, B.T., 1992.Neurofibrillary tangles but not senile plaques parallel durationand severity of Alzheimer's disease. Neurology 42, 631–639.

Boimel, M., Grigoriadis, N., Lourbopoulos, A., Touloumi, O.,Rosenmann, D., Abramsky, O., Rosenmann, H., 2009. Statinsreduce the neurofibrillary tangle burden in a mouse model oftauopathy. J. Neuropathol. Exp. Neurol. 68, 314–325.

Carlsson, C.M., Gleason, C.E., Hess, T.M., Moreland, K.A., Blazel,H.M., Koscik, R.L., Schreiber, N.T., Johnson, S.C., Atwood, C.S.,Puglielli, L., Hermann, B.P., McBride, P.E., Stein, J.H., Sager, M.A.,Asthana, S., 2008. Effect s of simvastatin on cerebrospinalfluid biomarkers and cognition in middle-aged adults at riskfor Alzheimer's disease. J. Alzheimers Dis. 13, 187–197.

Chauhan, N.B., Siegel, G.J., Feinstein, D.L., 2004. Effects oflovastatin and pravastatin on amyloid processing andinflammatory response in TgCRND8 brain. Neurochem. Res. 29,1897–1911.

Cimino, M., Gelosa, P., Gianella, A., Nobili, E., Tremoli, E., Sironi, L.,2007. Statins: multiple mechanisms of action in the ischemicbrain. Neuroscientist 13, 208–213.

Cordle, A., Landreth, G., 2005. 3-Hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors attenuate beta-amyloid-induced microglial inflammatory responses. J. Neurosci. 25,299–307.

De Felice, F.G., Wu, D., Lambert, M.P., Fernandez, S.J., Velasco, P.T.,Lacor, P.N., Bigio, E.H., Jerecic, J., Acton, P.J., Shughrue, P.J.,Chen-Dodson, E., Kinney, G.G., Klein, W.L., 2008. Alzheimer'sdisease-type neuronal tau hyperphosphorylation induced byAβ oligomers. Neurobiol. Aging 29, 1334–1347.

Fassbender, K., Simons, M., Bergmann, C., Stroick, M., Lutjohann,D., Keller, P., Runz, H., Kuhl, S., Bertsch, T., von Bergmann, K.,Hennerici, M., Beyreuther, K., Hartmann, T., 2001. Simvastatinstrongly reduces levels of Alzheimer's disease beta-amyloidpeptides Abeta 42 and Abeta 40 in vitro and in vivo. Proc. NatlAcad. Sci. USA 98, 5856–5861.

Hsiao, K., Chapman, P., Nilsen, S., Eckman, C., Harigaya, Y.,Younkin, S., Yang, F., Cole, G., 1996. Correlative memorydeficits, Aβ elevation, and amyloid plaques in transgenic mice.Science 274, 99–102.

Hu, M., Waring, J.F., Gopalakrishnan, M., Li, J., 2008. Role ofGSK-3beta activation and alpha7 nAChRs inAbeta(1-42)-induced tau phosphorylation in PC12 cells.J. Neurochem. 106, 1371–1377.

Jick, H., Zornberg, G.L., Jick, S.S., Seshadri, S., Drachman, D.A., 2000.Statins and the risk of dementia. Lancet 356, 1627–1631.

Kajinami, K., Mabuchi, H., Saito, Y., 2000. NK-104: a novel syntheticHMG-CoA reductase inhibitor. Expert Opin. Investig. Drugs 9,2653–2661.

Kawarabayashi, T., Younkin, L.H., Saido, T.C., Shoji, M., Ashe, K.H.,Younkin, S.G., 2001. Age-dependent changes in brain, CSF, andplasma amyloid β protein in the Tg2576 transgenic mousemodel of Alzheimer's disease. J. Neurosci. 21, 372–381.

Kayed, R., Head, E., Thompson, J.L., McIntire, T.M., Milton, S.C.,Cotman, C.W., Glabe, C.G., 2003. Common structure of solubleamyloid oligomers implies common mechanism ofpathogenesis. Science 300, 486–489.

Klein, W.L., Krafft, G.A., Finch, C.E., 2001. Targeting small Aβoligomers: the solution to an Alzheimer's disease conundrum?Trends Neurosci. 24, 219–224.

Lambert, M.P., Barlow, A.K., Chromy, B.A., Edwards, C., Freed, R.,Liosatos, M., Morgan, T.E., Rozovsky, I., Trommer, B., Viola, K.L.,Wals, P., Zhang, C., Finch, C.E., Krafft, G.A., Klein, W.L., 1998.Diffusible, nonfibrillar ligands derived from Aβ1-42 are potentcentral nervous system neurotoxins. Proc. Natl Acad. Sci. USA95, 6448–6453.

Li, L., Cao, D., Kim, H., Lester, R., Fukuchi, K., 2006. Simvastatinenhances learning and memory independent of amyloid loadin mice. Ann. Neurol. 60, 729–739.

Li, G., Larson, E.B., Sonnen, J.A., Shofer, J.B., Petrie, E.C.,Schantz, A., Peskind, E.R., Raskind, M.A., Breitner, J.C.,Montine, T.J., 2007. Statin therapy is associated withreduced neuropathologic changes of Alzheimer disease.Neurology 28, 878–885.

Morikawa, S., Umetani, M., Nakagawa, S., Yamazaki, H., Suganami,H., Inoue, K., Kitahara, M., Hamakubo, T., Kodama, T., Saito, Y.,2000. Relative induction of mRNA for HMG CoA reductase andLDL receptor by five different HMG-CoA reductase inhibitors incultured human cells. J. Atheroscler. Thromb. 7, 138–144.

Nachtigal, P., Jamborova, G., Pospisilova, N., Pospechova, K.,Solichova, D., Zdansky, P., Semecky, V., 2006. Atorvastatin hasdistinct effects on endothelial markers in different mousemodels of atherosclerosis. J. Pharm. Pharm. Sci. 9,222–230.

Nomura, S., Inami, N., Shouzu, A., Omoto, S., Kimura, Y.,Takahashi, N., Tanaka, A., Urase, F., Maeda, Y., Ohtani, H.,Iwasaka, T., 2009. The effects of pitavastatin, eicosapentaenoicacid and combined therapy on platelet-derived microparticlesand adiponectin in hyperlipidemic, diabetic patients. Platelets20, 16–22.

Okada, M., Tamura, A., Urae, A., Nakagomi, T., Kirino, T., Mine, K.,Fujiwara, M., 1995. Long-term spatial cognitive impairmentfollowing middle cerebral artery occlusion in rats. A behavioralstudy. J. Cereb. Blood Flow. Metab. 15, 505–512.

Petanceska, S.S., DeRosa, S., Olm, V., Diaz, N., Sharma, A.,Thomas-Bryant, T., Duff, K., Pappolla, M., Refolo, L.M., 2002.Statin therapy for Alzheimer's disease: will it work? J. Mol.Neurosci. 19, 155–161.

Refolo, L.M., Pappolla, M.A., Malester, B., LaFrancois, J.,Bryant-Thomas, T., Wang, R., Tint, G.S., Sambamurti, K.,Duff, K., 2000. Hypercholesterolemia accelerates theAlzheimer's amyloid pathology in a transgenic mouse model.Neurobiol. Dis. 7, 321–331.

Refolo, L.M., Pappolla, M.A., LaFrancois, J., Malester, B., Schmidt,S.D., Thomas-Bryant, T., Tint, G.S., Wang, R., Mercken, M.,Petanceska, S.S., Duff, K.E., 2001. A cholesterol-lowering drugreduces beta-amyloid pathology in a transgenicmouse model of Alzheimer's disease. Neurobiol. Dis. 8,890–899.

Roberson, E.D., Scearce-Levie, K., Palop, J.J., Yan, F., Cheng, I.H.,Wu, T., Gerstein, H., Yu, G.Q., Mucke, L., 2007. Reducingendogenous tau ameliorates amyloid beta-induced deficits inan Alzheimer's disease mouse model. Science 316,750–754.

Selkoe, D.J., 2002. Alzheimer's disease is a synaptic failure. Science298, 789–791.

Sparks, D.L., Connor, D.J., Sabbagh, M.N., Peterson, R.B., Lopez, J.,Browne, P., 2006. Circulating cholesterol levels, apolipoprotein

170 B R A I N R E S E A R C H 1 3 7 1 ( 2 0 1 1 ) 1 6 1 – 1 7 0

E genotype and dementia severity influence the benefit ofatorvastatin treatment in Alzheimer's disease: result of theAlzheimer's Disease Cholesterol-Lowering Treatment (ADCLT)trial. Acta Neurol. Scand. 114 (suppl 185), 3–7.

Suzuki, H., Aoki, T., Tamaki, T., Sato, F., Kitahara, M., Saito, Y., 1999.Hypolipidemic effect of NK-104, a potent HMG-CoA reductaseinhibitor, in guinea pigs. Atherosclerosis 146, 259–270.

Tomiyama,T.,Matsuyama, S., Iso,H.,Umeda,T., Takuma,H.,Ohnishi,K., Ishibashi, K., Teraoka, R., Sakama, N., Yamashita, T., Nishitsuji,K., Ito, K., Shimada, H., Lambert, M.P., Klein, W.L., Mori, H., 2010.A mouse model of amyloid β oligomers: Their contribution to

synaptic alteration, abnormal tau phosphorylation, glialactivation, and neuronal loss in vivo. J. Neurosci. 30, 4845–4856.

Tong, X.K., Nicolakakis, N., Fernandes, P., Ongali, B., Brouillette, J.,Quirion, R., Hamel, E., 2009. Simvastatin improvescerebrovascular function and counters soluble amyloid-beta,inflammation and oxidative stress in aged APP mice.Neurobiol. Dis. 35, 406–414.

Wolozin, B., Kellman, W., Ruosseau, P., Celesia, G.G., Siegel, G.,2000. Decreased prevalence of Alzheimer disease associatedwith 3-hydroxy-3-methyglutaryl coenzyme A reductaseinhibitors. Arch. Neurol. 57, 1439–1443.