In Press

Journal of Alzheimer’s Disease xx (20xx) x–xxDOI 10.3233/JAD-142052IOS Press

1

Lymphocytic Mitochondrial AconitaseActivity is Reduced in Alzheimer’s Diseaseand Mild Cognitive Impairment

Francesca Mangialaschea,b, Mauro Baglionia, Roberta Cecchettia, Miia Kivipeltob,Carmelinda Ruggieroa, Danilo Piobbicoc, Lothar Kussmauld, Roberto Monasteroe,Stefano Brancorsinic,∗ and Patrizia Mecoccia,∗aSection of Gerontology and Geriatrics, Department of Medicine, University of Perugia, Perugia, ItalybAging Research Center, Karolinska Institutet-Stockholm University, Stockholm, SwedencDepartment of Experimental Medicine - Section of Terni, University of Perugia, ItalydDepartment of Central Nervous System Research, Boehringer-Ingelheim Pharma GmbH & Co,Biberach an der Riss, GermanyeSection of Neurology, Department of Experimental Biomedicine and Clinical Neuroscience, University of Palermo,Palermo, Italy

Handling Associate Editor: Domenico Pratico

Accepted 16 September 2014

Abstract.Background: Specific mechanisms behind the role of oxidative/nitrosative stress and mitochondrial dysfunction in Alzheimer’sdisease (AD) pathogenesis remain elusive. Mitochondrial aconitase (ACO2) is a Krebs cycle enzyme sensitive to free radical-mediated damage.Objective: We assessed activity and expression of ACO2 extracted from blood lymphocytes of subjects with AD, mild cognitiveimpairment (MCI), older adults with normal cognition (OCN, age ≥65 years), and younger adults with normal cognition (YCN,age <65 years). Plasma levels and activities of antioxidants were also measured.Methods: Blood samples were collected from 28 subjects with AD, 22 with MCI, 21 OCN, and 19 YCN. ACO2 activity wasevaluated in a subsample before and after in vitro exposure to free radicals.Results: ACO2 activity was significantly lower in AD and MCI cases than controls: ACO2 median activity was 0.64 ± 0.21U/mg protein for AD, 0.93 ± 0.28 U/mg protein for MCI, 1.17 ± 0.78 U/mg protein for OCN subjects, and 1.23 ± 0.43 U/mgprotein for YCN individuals. In subjects with AD and MCI, ACO2 expression was lower than OCN subjects, and ACO2activity correlated with vitamin E plasma levels (rho: 0.64, p < 0.001) and Mini-Mental State Examination total score (rho: 0.82,p < 0.001). Furthermore, free radicals exposure reduced ACO2 activity more in individuals with AD than in OCN subjects.Conclusion: Our results suggest that ACO2 activity is reduced in peripheral lymphocytes of subjects with AD and MCI andcorrelates with antioxidant protection. Further studies are warranted to verify the role of ACO2 in AD pathogenesis and itsimportance as a marker of AD progression.

Keywords: Aconitase (aconitate hydratase), Alzheimer disease, antioxidants, free radicals, lymphocyte, mild cognitive impair-ment, mitochondria, oxidative stress, reactive nitrogen species, reactive oxygen species

∗Correspondence to: Stefano Brancorsini, PhD, Department ofExperimental Medicine – Section of Terni, University of Peru-gia, Via Mazzieri, 3, 05100 Terni, Italy. Tel.: +39 0744 202820;Fax: +39 0744 202804; E-mail: stefano.brancorsini@unipg.it. andPatrizia Mecocci, MD PhD, Section of Gerontology and Geri-atrics, University of Perugia, Ospedale S. Maria della MisericordiaPerugia, Block C Floor 4, S.Andrea delle Fratte, 06156 Perugia,Italy. Tel.: +39 075 578 3270; Fax: +39 075 578 3878; E-mail:patrizia.mecocci@unipg.it.

ISSN 1387-2877/15/$27.50 © 2015 – IOS Press and the authors. All rights reserved

In Press2 F. Mangialasche at al. / Mitochondrial Aconitase in Alzheimer’s Disease

INTRODUCTION

Alzheimer’s disease (AD) is the main cause ofdementia in older adults and is often preceded bya predementia phase of mild cognitive impairment(MCI) [1]. Mechanisms responsible for selective neu-ronal death in AD are still unclear, but oxidative andnitrosative stress (OS/NS) are believed to play key roles[2]. Increased levels of oxidative/nitrosative damagehave been reported in subjects with AD, both in thecentral nervous system (CNS) and in peripheral tis-sues, including blood lymphocytes [3–7]. In addition,reduced plasma levels of non-enzymatic antioxidantshave been found in individuals with AD, indicat-ing inadequate antioxidant activity in persons withthis disorder [8, 9]. Both increased levels of markersof oxidative/nitrosative damage and depleted antioxi-dants have been reported in subjects with MCI [3, 10,11], which suggests that OS/NS are early events in ADpathogenesis.

Proteomic studies have shown that oxidative andnitrosative damage in AD preferentially involves spe-cific proteins, which can be more susceptible becauseof sequence motifs, environmentally exposed resid-uals, and interaction with redox-metal groups [4,12–14]. Mitochondrial proteins involved in energy pro-duction are key targets for OS/NS, as mitochondria arethe principal intracellular source of free radicals. Freeradical-dependent mitochondrial damage is believed tobe crucial in neurodegeneration [15]; yet, the role ofmitochondrial dysfunction in AD pathogenesis has notbeen fully elucidated.

Mitochondrial aconitase (ACO2) is an iron-sulphurdehydratase containing a [4Fe-4S] cluster that cat-alyzes the interconversion of citrate and isocitrate inthe Krebs cycle, a reaction essential to normal cellmetabolism [16, 17]. Because of the presence of theiron-sulphur group, ACO2 is sensitive to reactive oxy-gen and nitrogen species (ROS/RNS), which can affectACO2 structure and activity. Studies in animal and cel-lular models have shown that free radicals and productsof oxidative damage can promote oxidation of ACO2,resulting in impaired enzymatic activity [18–20]. Thus,loss of ACO2 activity can reflect increased levels ofOS/NS and could be relevant to AD pathogenesis.In support of this hypothesis, some neuropathologi-cal studies reported reduced levels of ACO2 activity inthe brains of subjects with familial [21] or sporadic AD[12]. In another postmortem study, decreased expres-sion of ACO2 mRNA was observed in hippocampaltissues from AD brains [22]. However, definite evi-dence on the role of ACO2 in AD is lacking, and

there have been no studies done in vivo or in periph-eral tissues. In this cross-sectional study, we assessedthe activity and expression of ACO2 extracted fromperipheral blood lymphocytes of subjects with AD,subjects with MCI, and cognitively normal (CN) indi-viduals. Enzymatic activity was correlated to plasmalevels/activities of several antioxidants. Finally, changein ACO2 activity after in vitro exposure to free radicalswas evaluated to investigate the presence of a differentsusceptibility to ROS/RNS in the context of diseasecondition.

MATERIALS AND METHODS

Subjects

Twenty-eight subjects with mild AD (ClinicalDementia Rating scale [CDR] = 1) [23], 22 with MCI,21 older adults with normal cognition (OCN, age≥65 years), and 19 younger adults with normal cog-nition (YCN, age <65 years) were enrolled at theMemory Clinic of the Section of Gerontology andGeriatrics, University of Perugia, Italy (Table 1). Allwere evaluated in accordance with a standard proto-col that included a detailed anamnesis and clinical andneuropsychological examinations. Global cognitionwas assessed with the Mini-Mental State Examination(MMSE) [24], and dementia severity was assessed withthe CDR scale [23]. Subjects with anxiety or depres-sion (evaluated by the Hamilton Anxiety Scale andthe Geriatric Depression Scale, respectively) [25, 26],history of smoking/alcohol abuse, major organ fail-ure, malnutrition, dyslipidemia, alteration of proteinmetabolism, or taking iron or antioxidant supplementswere excluded from the study. Blood cell count of allsubjects was in the normal range according to the lab-oratory cutoff values and did not differ among groups.To exclude secondary causes of dementia, blood levelsof vitamin B12, folic acid, and thyroid hormones werealso determined.

Dementia was diagnosed on the basis of the Diag-nostic and Statistical Manual of Mental Disorders,Fourth Edition (DSM-IV) criteria [27], and probableAD was diagnosed following the National Instituteof Neurological and Communicative Disorders andStroke-Alzheimer’s Disease and Related DisordersAssociation (NINCDS-ADRDA) criteria [28].

MCI was defined in accordance with the MayoClinic AD Research Center Criteria, [29, 30]: 1)absence of dementia according to the DSM-IV [27];2) normal general cognitive function; 3) memory com-plaint: self- or informant-reported memory problem;

In PressF. Mangialasche at al. / Mitochondrial Aconitase in Alzheimer’s Disease 3

Table 1Main characteristics of the study population by diagnosis

AD subjects (n = 28) MCI subjects (n = 22) OCN subjects (n = 21) YCN subjects (n = 19)

Age, years, mean (SD)∗ 74.9 ± 6.9 76.5 ± 6.6 79.1 ± 7.7 41.4 ± 9.9Gender (M/F)∗∗ 13/15 10/12 10/11 11/8CDR 1 0.5 0 0MMSE, Median (IQR)∗∗∗ 21.5 (3)‡,†,§ 25 (2)¶,# 30 (1) 30 (0)∗ANOVA with Bonferroni post-hoc correction: p < 0.001 for YCN versus AD, MCI and OCN. ∗∗Chi-square test p = 0.849. ∗∗∗Mann-WhitneyU-test: AD versus YCN: ‡p < 0.001 AD versus OCN: †p < 0.001 AD versus MCI: §p < 0.001 MCI versus YCN: ¶p < 0.001 MCI versus OCN:#p < 0.001. AD, Alzheimer’s disease; CDR, Clinical Dementia Rating scale; MCI, mild cognitive impairment; MMSE, Mini Mental StateExamination; OCN, older adults with normal cognition; YCN, younger adults with normal cognition.

4) objective memory impairment: pathological score(below normality cutoff) in at least one standardizedmemory test, with normal performance in the othercognitive tasks; and 5) normal activities of daily living.

The investigation conformed to the principles out-lined in the Declaration of Helsinki, and informedconsent was obtained from the subjects or their rel-atives.

Blood sampling and antioxidant measurements

Venous blood samples were taken in EDTA/K3tubes after an overnight fast. Blood was immediatelycentrifuged and plasma stored at −80◦C until anal-ysis. To preserve vitamin C, an aliquot of plasmawas deproteinized with 10% metaphosphoric acid andthe supernatant was kept at –80◦C until analysis.All measurements were performed blind to clinicalinformation. Vitamin C and uric acid were detectedby high-performance liquid chromatography (HPLC)with electrochemical detection (ESA, Chelmsford,MA, USA) according to Kutnink et al. [31] withSupelco C 18 column (250 mm × 4.6 mm i.d.) andSupelco C 18 guard column (20 mm × 4.6 mm i.d.)(Sigma-Aldrich, Milan, Italy). After extraction withethanol and hexane as described elsewhere [32],vitamin A and vitamin E (�-tocopherol) were mea-sured by HPLC with UV detection at 280 nm witha Waters Simmetry C 8 column (150 mm × 4.6 mmi.d.) (Waters, Milan, Italy) [33]. Superoxide dismu-tase SOD3 (pSOD, U/ml) and glutathione peroxidase(GPX, mol NADPH/min/ml) activities in plasma weremeasured spectrophotometrically in accordance withdescribed methods [34, 35].

Isolation of mitochondria

Mitochondria were isolated from peripheral bloodlymphocytes as previously described [36]. Briefly,

freshly obtained blood was layered on Lymphoprep(Gibco, BRL, Bethesda MD, USA), centrifuged, andwashed twice. Pellet was then suspended in 400 �lof ice-cold PBS, 8 �l of 2.5% digitonin were added,and samples kept on ice for 5 min, inverting gentlyevery 30 s. Samples were then sonicated for 1 min andcentrifuged at 600 g for 10 min at 4◦C to eliminatenuclei and unbroken cells. Supernatants were collectedand centrifuged at 14,000 g for 10 min at 4◦C. Pel-lets were then resuspended in 400 �l ice-cold PBS andcentrifuged at 7,000 g for 10 min at 4◦C. The pelletobtained was resuspended in 400 �l ice-cold PBS andcentrifuged at 3,500 g for 10 min at 4◦C. Pelleted puri-fied mitochondria obtained were immediately frozenand kept at -80◦C until analysis. The amount of proteinin the mitochondrial pellets was measured by MicroBCA Protein Assay Reagent Kit (Pierce, Rockford IL,USA) following the protocol recommended by manu-facturer.

Assessment of ACO2 activity

For each subject, and in all experiments, ACO2activity was measured in amounts of pelleted purifiedmitochondria corresponding to 80 �g of mitochondrialproteins. Assessment of ACO2 activity was conductedin 2 ml of reaction mixture containing 13 mM Tris(pH 7.4), 5 mM sodium citrate, 0.2 mM Beta NADP,0.6 mM MnCl2, 2U/ml isocitrate dehydrogenase, andpelleted purified mitochondria. ACO2 activity wasmeasured spectrophotometrically by following the lin-ear absorbance change at 340 nm, at +37◦C and pH 7.4for 30 min.

For 10 AD, 10 MCI, and 10 OCN subjects ran-domly selected from each group, ACO2 activity wastested both in basal conditions and after incubationin media containing a superoxide (O2−) generatingsystem or the nitric oxide (NO) donor sodium nitro-prusside (SNP) [37]. Since previous studies suggested

In Press4 F. Mangialasche at al. / Mitochondrial Aconitase in Alzheimer’s Disease

an age-related decline in ACO2 activity [18, 38], weexcluded YCN subjects in order to perform the exper-iment in age-matched subjects. For each subject, 3samples of mitochondrial pellet were used: one samplewas used to assess basal ACO2 activity as describedabove; one sample was incubated for 30 min at +37◦Cwith 100 �l xanthine 10 mM plus 1 �l xanthine oxidase25 U/ml. This treatment promotes the production of theROS superoxide anion. The third sample was incubatedfor 30 min at +37◦ C with 100 �l of SNP 10 mM. In allexperiments, ACO2 activity was assessed spectropho-tometrically by following the linear absorbance changeat 340 nm at +37◦C and pH 7.4 for 30 min.

Western blot analysis

Western blot analyses were performed in samplesfrom 10 AD, 10 MCI, and 10 OCN subjects. For thesesamples, amounts of mitochondrial pellets contain-ing 80 �g of proteins were homogenized in boilingLaemmli buffer, analyzed by 9% sodium dodecylsulphate polyacrylamide gel electrophoresis (SDS-PAGE), and transferred to a nitrocellulose membrane(Sigma-Aldrich, Milan, Italy) [39]. Membranes wereprobed using monoclonal antibodies against humanACO2 from rat (Boehringer-Ingelheim Pharma GmbH& Co laboratories, Biberach an der Riss, Germany) ormonoclonal antibodies raised against purified humanmitochondrial complex IV (COX4) from mouse (SantaCruz Biotechnology, Santa Cruz, CA), which wasused as internal control. After blocking, membraneswere incubated with antibody diluted in blockingbuffer (0.5% milk/BSA) overnight at 4◦C. The fol-lowing day, membranes were washed and incubatedwith horseradish peroxidase-conjugated anti-rat oranti-mouse antibodies. Immunodetection of bands wasvisualized by ECL chemiluminescence assay (GEHealthcare, Amersham Bioscience, Piscataway, NJ,USA). Imagej software from http://imagej.nih.gov/ij(National Institutes of Health, USA) was used for den-sitometry.

RNA isolation and quantitative PCR

Total RNA was isolated from lymphocytes of10 AD, 10 MCI, and 10 OCN subjects. TRIzolreagent (Invitrogen, Carlsbad, CA, USA) was usedfor RNA extraction following the protocol rec-ommended by manufacturer. 3 �g of total RNAwere retrotranscribed using RevertAidTM H MinusM-MuLV Reverse Transcriptase (Fermentas LifeSciences, Burlington, Toronto, CAN) and random

hexamer primers. Real-Time PCR amplificationswere performed with Mx3000PTM®Real TimePCR System using Brilliant SYBR Green QPCRMaster Mix (Stratagene, La Jolla, CA, USA) andROX as reference dye. ACO2 specific primers were:forward 5’-ACATCCATTATGACCTGCTAG-3’,reverse 5’-TGTCCATACACAATCTTCTCC-3’(Invitrogen, Carlsbad, CA, USA). Housekeepingcontrol was human hypoxanthine phosphorybosil-transferase (HPRT) gene. All the experiments wereperformed in triplicate.

Statistical analysis

Gender-related differences between groups (AD,MCI, OCN, and YCN) were examined with the chi-square test, and ANOVA with Bonferroni post-hoccorrection was used to compare age. As the values oftotal MMSE score, ACO2 activity, and plasma antioxi-dants had a non-Gaussian distribution, non-parametrictests were applied to compare all these measuresamong groups, and also to compare ACO2 activity inbasal conditions and after in vitro exposure to free radi-cals. The Kruskal-Wallis test was used to compare eachblood measure among groups, and the Mann-WhitneyU test was used for pairwise group comparisons.

Rank-transformed ANCOVA was used to investi-gate the association between diagnosis (AD, MCI,CN) and ACO2 activity and plasma antioxidants whileadjusting for age and gender. This analysis is basedon the rank transformation of Conover and Iman,through which data are replaced by their ranks [40], andthe normal score transformation is then applied. Thisapproach permits the performance of a non-parametricand robust analysis because it is related to a compari-son of the median instead of the mean values. On theother hand, the relative magnitude of the data is lost.Thus, ANCOVA parameter estimates should not beinterpreted in the usual sense and can only be usedto identify the importance of associated predictors.For interpretation, we will refer to descriptive statis-tics based on the median. In the ANCOVA models wefitted, CN subjects were the reference group.

Spearman’s rank order correlation coefficient (rho)was used for correlations between aconitase activityand levels of plasma non-enzymatic antioxidants andcorrelation between aconitase activity and total MMSEscore.

For all analysis, the level of significance was setto p ≤ 0.05. Statistical analysis was performed withthe program IBM SPSS Statistics Version 20 (IBMCorporation, Armonk, NY, USA).

In PressF. Mangialasche at al. / Mitochondrial Aconitase in Alzheimer’s Disease 5

RESULTS

Lymphocyte ACO2 activity and plasmaantioxidants

Gender distribution did not vary among AD, MCI,OCN, and YCN subjects (p = 0.849). Mean age ofAD, MCI, and OCN subjects did not differ signif-icantly, but all three groups were older than YCNsubjects (p < 0.001 versus all groups) (Table 1). Asexpected, the median total MMSE score was lowerin subjects with AD than in MCI and CN subjects,and it was also lower in MCI subjects than in OCNand YCN subjects (Table 1). For all blood parame-ters examined, results differed among the four groups(Kruskal-Wallis: p < 0.001) as reported in Table 2.Notably, median levels of lymphocyte ACO2 activ-ity were lower in AD than in MCI, OCN, and YCNsubjects. ACO2 activity was also significantly lowerin MCI subjects than both OCN and YCN subjects.Finally, OCN subjects had a median level of ACO2activity lower than that of YCN subjects, but the differ-ence was not statistically significant (Table 2). Plasmalevels of vitamin A, vitamin E, vitamin C, and uric acidwere consistently lower in AD and MCI than YCN andOCN subjects. MCI subjects had antioxidant valuesbetween those of the AD and CN groups, and only themedian value of vitamin E was significantly higher inMCI than AD subjects (Table 2). Plasma SOD activ-ity in AD and MCI subjects was lower than in YCNsubjects, and a small but significant difference in GPXactivity was observed in MCI subjects compared tothe other groups (Table 2). In summary, lymphocyteACO2 activity was reduced in subjects with MCI, andto a larger extent, in individuals with AD. Both ADand MCI were characterized by a similar depletion ofplasma non-enzymatic antioxidants, though the trend

for plasma enzymatic antioxidant activities was lessclear.

To evaluate the possible confounding effects of ageand gender, we compared all blood parameters amongAD, MCI, and CN subjects with the rank-transformANCOVA analyses, setting CN subjects as the ref-erence group. After controlling for age and gender,diagnoses of AD and MCI were still associated withreduced ACO2 activity (AD versus CN: p < 0.001;MCI versus CN: p = 0.001) (Table 3). Additionally, therank-transform ANCOVA analyses confirmed the neg-ative association between AD diagnosis and plasmalevels of non-enzymatic antioxidants (Table 3). Simi-lar results were obtained for MCI, except for vitamin A.For the association between cognitive status and vita-

Table 3Age- and gender-adjusted association between cognitive status andACO2 activity, plasma enzymatic antioxidants, and plasma non-

enzymatic antioxidants

AD � (Std.Error) MCI � (Std.Error)[p-value] [p-value]

ACO2 −1.91 (0.22) −0.82 (0.24)[<0.001] [0.001]

Vitamin A −1.10 (0.25) −0.44 (0.28)[<0.001] [0.121]

Vitamin E −1.71 (0.18) −0.52 (0.20)[<0.001] [0.011]

Vitamin C −1.10 (0.25) −0.87 (0.28)[<0.001] [0.003]

Uric acid −1.46 (0.28) −1.01 (0.31)[<0.001] [0.002]

SOD −0.19 (0.31) −0.14 (0.34)[0.54] [0.69]

GPX 0.08 (0.31) −0.63 (0.35)[0.80] [0.07]

Rank-transform ANCOVA analyses on ACO2 activity and plasmaenzymatic and non-enzymatic antioxidants. CN subjects were thereference category. ACO2, mitochondrial aconitase; SOD, super-oxide dismutase; GPX, glutathione peroxidase; AD Alzheimer’sdisease; MCI, mild cognitive impairment.

Table 2Activity of ACO2 from peripheral blood lymphocytes; plasma levels of vitamin A, vitamin E (�-tocopherol), vitamin C, and uric acid; and

plasma SOD and GPX activities in AD, MCI, YCN, and OCN subjects

AD subjects (n = 28) MCI subjects (n = 22) OCN subjects (n = 21) YCN subjects (n = 19)

ACO2 (U/mg prot.) 0.64 (0.21)∗,†,§ 0.93 (0.28)¶,# 1.17 (0.78) 1.23 (0.43)Vitamin A (�mol/l) 2.17 (0.36)∗,†† 2.32 (0.07)¶,## 2.39 (0.16)‡ 2.65 (0.24)Vitamin E (�mol/l) 26.35 (4.3)∗,†,§ 44.11 (7.1)¶,## 46.20 (4.9)‡ 63.85 (3.7)Vitamin C (�mol/l) 23.6 (4.7)∗,†† 24.05 (3.6)¶,## 25.9 (3.9)‡ 38.10 (3.0)Uric acid (�mol/l) 186.0 (73)∗,†† 198.5 (17)¶,# 259.0 (102)‡‡ 301.0 (29)SOD (U/ml) 24.1 (4.8)∗ 24.25 (6.2)¶ 23.4 (3.9)‡ 29.8 (2.0)GPX (�M NAPH/min/ml) 0.10 (0.002)§ 0.09 (0.011)¶,## 0.10 (0.003) 0.10 (0.002)

Data are given as median (interquartile range). Mann-Whitney U-test: AD versus YCN: ∗p < 0.001 AD versus OCN: †p < 0.001; ††p < 0.01 ADversus MCI: §p < 0.001 MCI versus YCN: ¶p < 0.001 MCI versus OCN: #p<0.001; ##p < 0.01 OCN versus YCN: ‡p < 0.001; ‡‡p < 0.05. ACO2,mitochondrial aconitase; SOD, superoxide dismutase; GPX, glutathione peroxidase; AD Alzheimer’s disease; MCI, mild cognitive impairment;OCN, older adults with normal cognition; YCN, younger adults with normal cognition.

In Press6 F. Mangialasche at al. / Mitochondrial Aconitase in Alzheimer’s Disease

min E levels, we performed additional analyses inwhich vitamin C was entered as covariate. This wasdone to account for the fact that the antioxidant effi-ciency of �-tocopherol is affected by the availability ofvitamin C. When vitamin C was added to the model,the association of AD and MCI diagnosis with vita-min E levels was reduced, but it was still significant inboth cases [AD versus CN: � (St. Error): −1.35 (0.14),p < 0.001; MCI versus CN: � (St. Error): −0.28 (0.14),p = 0.046].

In the age- and gender-adjusted analyses, enzymaticantioxidant activities were not associated with MCIand AD diagnoses (Table 3). Thus, age- and gender-adjusted analyses confirmed reduced lymphocyteACO2 activity and depleted plasma non-enzymaticantioxidants in AD and MCI subjects. We performedadditional rank-transform ANCOVA analyses in whichMCI subjects were used as reference category. Aftercontrolling for age and gender, ACO2 activity waslower in AD than MCI subjects [� (St. Error): −1.09(0.25), p < 0.001]. AD diagnosis was also associatedwith reduced levels of vitamin A [� (St. Error):−0.66 (0.29), p = 0.03] and vitamin E [� (St. Error):−1.19 (0.21), p < 0.001]. For vitamin E the associa-tion remained significant also after further adjustmentfor vitamin C [� (St. Error): −1.07 (0.13), p < 0.001].The difference between the levels of other antioxidantsin AD and MCI subjects was not significant (data notshown).

To test the hypothesis that lymphocyte ACO2activity is related to the concentration of plasma antiox-idants, we assessed the correlations between ACO2activity and non-enzymatic antioxidant levels. In theoverall study population, ACO2 activity was signifi-cantly correlated with vitamin E (rho: 0.74, p < 0.001),vitamin A (rho: 0.46, p < 0.001), vitamin C (rho: 0.52,p < 0.001), and uric acid (rho: 0.53, p < 0.001) levels.Further, we measured the correlations by stratifyingsubjects in two groups: those with cognitive impair-ment (MCI and AD) and those with normal cognition(OCN and YCN). This stratification was done to takeinto account the possible confounding effect of ADpathology in subjects with dementia and MCI. In sub-jects with cognitive impairment, ACO2 activity wassignificantly correlated with vitamin E levels (rho:0.64, p < 0.001) (Fig. 1). To account for possible effectsof vitamin C on such association, we further strati-fied cognitively impaired subjects by median levelsof vitamin C. We found that the correlation betweenACO2 and vitamin E remained significant (subjectswith vitamin C levels below the median: rho: 0.57,p = 0.002; subjects with vitamin C levels above the

1.40

1.20

1.00

0.80

0.60

0.40

20.0 25.0 30.0 35.0 40.0 45.0 50.0

MCIAD

Vitamin E (umol/l)A

CO

2 (U

/mg

prot

.)

Fig. 1. Correlation between plasma vitamin E (�-tocopherol) lev-els and mitochondrial aconitase (ACO2) activity in subjects withAlzheimer’s disease (AD) and mild cognitive impairment (MCI).Scatter plot and linear prediction plot; Spearman’s correlation coef-ficient: 0.64, p < 0.001.

median: rho: 0.72, p < 0.001). The correlation of ACO2with other non-enzymatic antioxidants was not sig-nificant. In subjects with normal cognition, there wasno significant correlation between ACO2 activity andlevels of plasma non-enzymatic antioxidants.

We also evaluated the correlation between ACO2activity and total MMSE score. The correlation waspositive and statistically significant in the total sam-ple (rho: 0.82, p < 0.001). In the stratified analysis,ACO2 activity correlated with total MMSE score (rho:0.82, p < 0.001) in subjects with cognitive impairment(Fig. 2). The correlation was not significant (rho: 0.05,p = 0.766) in those with normal cognition.

ACO2 activity under OS/NS conditions

To test whether the susceptibility of ACO2 to OS orNS varied by cognitive status, ACO2 activity was eval-uated in basal conditions and after exposure to mediacontaining an O2−generating system or the NO donorSNP in a randomly selected subsample of 10 OCN, 10MCI, and 10 AD subjects. Superoxide anions caused adecrease of ACO2 activity by 14% in OCN [% of base-line activity, median (interquartile range; i.e., IQR):86% (6.5)], by 19% in MCI [median (IQR): 81 (6.7)],and by 23% in AD subjects [median (IQR): 77 (8.8)](Fig. 3). The reduction was significantly greater for ADthan OCN subjects (p = 0.01). Exposure to NO led to an

In PressF. Mangialasche at al. / Mitochondrial Aconitase in Alzheimer’s Disease 7

MCIAD

28

26

24

22

20

18

0.40 0.60 0.80

ACO2 (U/mg prot.)

1.00 1.20 1.40

MM

SE

tota

l sco

re

Fig. 2. Correlation between mitochondrial aconitase (ACO2) activ-ity and Mini-Mental State Examination (MMSE) total score insubjects with Alzheimer’s disease (AD) and mild cognitive impair-ment (MCI). Scatter plot and linear prediction plot; Spearman’scorrelation coefficient: 0.82, p < 0.001.

8% reduction of ACO2 activity in OCN [median (IQR):92 (5.7)], 11% in MCI [median (IRQ): 89 (9.3)], and13% in AD subjects [median (IQR): 87 (6.4)] (Fig. 3).Also in this case, the decrease was significantly higherfor AD than CN subjects (p = 0.04) but did not differsignificantly between MCI and OCN subjects.

ACO2 expression

To evaluate whether the low levels of ACO2 activityobserved in AD and MCI subjects could be induced byaltered enzyme expression, we analyzed mRNA and

protein expression of mitochondrial ACO2 extractedfrom lymphocytic cells of a subsample of AD, MCI,and OCN subjects randomly selected from each group(Fig. 4). As in the experiment described previously,we excluded YCN subjects to compare age-matchedsubjects. ACO2 mRNA was detected in all sam-ples by quantitative PCR (Fig. 4A). Median ACO2mRNA levels did not differ significantly among groups(Kruskal-Wallis: p = 0.288). However, there was atrend toward lower ACO2 mRNA expression in ADand MCI subjects than in OCN subjects. Indeed,the median (and IQRs) of the relative quantities ofACO2 mRNA clearly separated AD from OCN sub-jects, whereas the median and IQRs of MCI subjectsoverlapped with those of the AD and OCN sub-jects (Fig. 4A). Analysis of mitochondrial proteinsby Western blot, including densitometry, showed lessexpression of ACO2 in MCI and AD than OCN sub-jects (Fig. 4B).

DISCUSSION

The main finding of this study is that mitochondrialACO2 activity is lower in peripheral blood lympho-cytes of AD and MCI than CN subjects. To the best ofour knowledge, this is the first report on ACO2 activ-ity in peripheral tissues of subjects with AD and MCI.Previous investigations have focused on ACO2 activ-ity in the brain. One neuropathological study showedreduced activity in the frontal cortex of subjects withfamilial AD. The same study also found that ACO2activity was reduced in the frontal cortex of patientswith sporadic AD, but the difference with controlswas not statistically significant [21]. In another post-mortem examination of the prefrontal cortex, ACO2

86% 81% 89%92%77%∗ 87%∗

ROS exposure RNS exposure

100%

80%

60%

40%

20%

0%

AD vs OCN: ∗ p<0.05

% A

CO

2 ac

tivity OCN

MCIAD

Fig. 3. ACO2 activity (median percentage of basal value) after 30 minutes of exposure to superoxide generating system or after 30 minutes ofexposure to the nitric oxide donor sodium nitroprusside, in AD, MCI, and OCN subjects. ACO2, mitochondrial aconitase; AD, Alzheimer’sdisease; MCI, mild cognitive impairment; OCN, older adults with normal cognition; RNS, reactive nitrogen species, ROS, reactive oxygenspecies.

In Press8 F. Mangialasche at al. / Mitochondrial Aconitase in Alzheimer’s Disease

0.5A B

0.4

0.3

0.2

0.1

OCN MCI AD

OCN

AC02—

COX4—

MCI AD

OCN OCN MCI MCI AD AD

—100 kD

—60 kD

—20 kD

—15 kD

100

80

60

40

20R

elat

ive

AC

O2

leve

ls (

%)

AC

O2

mR

NA

rel

ativ

e qu

antit

y (d

Rn)

LegendData are given as median (interquartile range)

OCN

AC02—

COX4—

MCI AD

100

80

60

40

20R

elat

ive

AC

O2

leve

ls (

%)

Fig. 4. A) ACO2 mRNA analysis by quantitative RT-PCR is reported as relative quantities from three independent analyses. Data are givenas median (black line) with interquartile range (grey box). B) Western blot analysis of ACO2 in lymphocyte mitochondria (see Materialand Methods). Immunodetection with anti-COX4 monoclonal antibody was used to guarantee the same amount of proteins in each sample.Representative Western blots of two individuals within each group are shown. Bottom, densitometry was shown for relative ACO2 expressionlevels; the first control was assumed as 100%. ACO2, mitochondrial aconitase; AD, Alzheimer’s disease; MCI, mild cognitive impairment;COX4, mitochondrial complex IV; OCN, older adults with normal cognition.

activity was not significantly lower in AD than in con-trols [41], whereas a third study reported 50% lowerACO2 activity in the hippocampus of people with ADthan in controls [12].

Reduction of the activity of an enzymatic protein canbe due to several factors, including transcriptional andpost-transcriptional modifications, post-translationalregulation, and increased protein degradation. In oursample, less ACO2 protein was found in AD andMCI subjects, who also exhibited a trend toward lowerACO2 mRNA than OCN subjects. We can reasonablyassume that both factors may contribute to the reduc-tion of mitochondrial ACO2 activity we observed inAD and MCI subjects. Our findings are in line withprevious microarray analysis results, which showedreduced ACO2 mRNA expression in postmortem hip-pocampal samples of subjects with AD [22]. A trendtoward down-regulation of the ACO2 gene in the ADbrain was also reported by Brooks et al. [42], althoughin that study the difference with controls was not sta-tistically significant.

Reduced levels of ACO2 protein could be also aconsequence of increased removal. This study did notevaluate ACO2 degradation, but previous investiga-tions have shown that increased OS/NS and depletionof antioxidants can perturb the tertiary structure of

mitochondrial proteins, which are then removed byenzymatic degradation by proteases such as Lon [43,44]. Both ACO2 susceptibility to ROS/RNS [19, 20,38] and increased OS/NS in AD and MCI [3] havebeen well documented, and ACO2 has been identi-fied as a preferential substrate of Lon protease [45].Therefore we cannot exclude the possibility that thereduced amount of ACO2 protein observed in AD andMCI subjects could have been due to increased proteindegradation by mitochondrial proteases.

Post-translational modifications due to OS/NS canalso contribute to the reduction of ACO2 activity.Examination of ACO2 post-translational modificationswas not the objective of our study; however, we showedthat exposure of equal amounts of mitochondrial pro-teins to ROS and RNS led to a greater loss of ACO2activity in subjects with AD than in controls. We canthus hypothesize that the reduced levels of ACO2 pro-tein in AD subjects could amplify the reduction of theenzymatic activity caused by free radical damage. Insupport of this hypothesis, several studies have shownthat free radicals can alter ACO2 structure and decreaseits activity. ACO2 contains a [4Fe-4S] cluster, essen-tial for its catalytic activity, in which the fourth iron isonly loosely bound to the protein and can easily disso-ciate from the cluster, converting the enzyme into an

In PressF. Mangialasche at al. / Mitochondrial Aconitase in Alzheimer’s Disease 9

inactive form. Free radicals such as superoxide anionand NO are produced in vivo, and ACO2 is sensitiveto superoxide anion-mediated inactivation by selec-tive reaction with Fe-S group [46]. Additionally, NOreacts avidly with superoxide anions to form perox-ynitrite, which can also inactivate ACO2 by reactingwith the Fe-S cluster [46, 47]. Furthermore, free radi-cals can promote oxidative modifications of the proteinitself (e.g., thiol oxidation, carbonylation) which canreduce its activity [19, 48, 49]. Consistent with thisobservation, Perluigi et al. reported increased levelsof ACO2-bound 4-hydroxy-2-nonenal (4-HNE) in thehippocampus of AD patients, and such oxidative mod-ification was associated with a reduction of ACO2activity [12]. Notably, increased levels of protein-bound 4-HNE have been found also in mitochondriaderived from peripheral lymphocytes of subjects withAD [36]. It is thus plausible that free radical-mediatedpost-translational changes can contribute to the reduc-tion of ACO2 activity in subjects with AD and MCI.

The reduced availability of antioxidants contributesto the redox unbalance observed in AD and MCIand might further boost the loss of ACO2 activity.In our population, subjects with AD and MCI hadlower levels of plasma non-enzymatic antioxidantsthan those with normal cognition, as reported in otherstudies [8–10, 50]. Furthermore, in subjects with cog-nitive impairment, ACO2 activity was correlated withplasma levels of vitamin E (even when taking intoaccount vitamin C concentrations) and global cogni-tion. Vitamin E, the main lipid-soluble non-enzymaticantioxidant in the human body, is essential for main-taining the functional integrity of mitochondria [47].A study on rat brain mitochondria reported that lowlevels of vitamin E exacerbated OS-mediated damageand decreased oxidative phosphorylation, possibly viainhibition of ACO2 activity, pointing at vitamin E ascritical for the normal functioning of ACO2 and theKrebs cycle [51]. It is thus possible that the reducedavailability of vitamin E observed in AD and MCIcontributes to the decrease in ACO2 activity.

OS/NS and mitochondrial dysfunction are believedto play a key role in AD pathogenesis [2]. How-ever, the mechanisms are still unclear and the overlapbetween early biological changes related to AD patho-physiology and markers of normal aging makes ADpathogenesis difficult to unravel. Our data suggestthat ACO2 activity differs between cognitively healthyolder adults and subjects in the early stage of AD, withintermediate levels in individuals who have MCI, indi-cating that this enzyme could be involved in diseaseprogression. In support of this hypothesis, it has been

shown that even partial inhibition of ACO2 can resultin dysfunction of the Krebs cycle, impacting energyproduction and neuronal viability [52]. Another studyreported that under OS conditions, ACO2 releaseslabile iron in primary ventral mesencephalic cells fromrats, an event followed by mitochondrial dysfunctionand cell death [53]. The reduction of ACO2 activityin mild AD, already present in the predementia phase(i.e. MCI), might promote energy unbalance and fosterthe increase in OS/NS, which then promotes A� accu-mulation, synaptic dysfunction, and neuronal death.

Strengths of this study include the accurate clinicalevaluation of the study participants, the measurementof ACO2 enzymatic activity both in basal conditionsand after in vitro exposure to free radicals; the assess-ment of ACO2 protein levels, mRNA expression, and adetailed panel of enzymatic and non-enzymatic antiox-idants; and the control group that consisted of bothYCN and OCN subjects. Further, robust statisticalmethods were used.

Our study has also some limitations. It is knownthat, among Krebs cycle enzymes, ACO2 exhibits themost significant age-associated decline in activity [18,38]. In our sample, ACO2 activity in OCN subjectswas lower than in YCN individuals, but the differencewas not statistically significant, which could be due tothe relatively small sample size and limited statisticalpower of the analyses. However, we detected a signif-icant difference in ACO2 activity between those withnormal cognition and those with cognitive impairment.Our results need to be confirmed in a larger sample,which should include also subjects with other CNSdisorders characterized by increased OS/NS.

Second, an extensive investigation of the possiblemechanisms underlying the reduction of ACO2 activ-ity was not performed in this study; however, our datasuggest that the decreased enzymatic activity in ADand MCI could be due, at least in part, to mRNAderegulation and free radical-mediated damage. Amore comprehensive investigation of transcriptionaland post-transcriptional changes in ACO2 in AD iswarranted to elucidate this point. Also, proteomicstudies will be crucial in clarifying the role of freeradical-mediated mechanisms (i.e., redox modifica-tions, iron disassembly) in determining the change inACO2 activity in AD. Finally, an increasing body ofscientific evidence supports the role of different vita-min E congeners in MCI and AD [11, 54–56]. In thisstudy only �-tocopherol was investigated, and furtherstudies are needed to explore the association amongdifferent vitamin E forms and ACO2 activity in relationto AD.

In Press10 F. Mangialasche at al. / Mitochondrial Aconitase in Alzheimer’s Disease

In conclusion, the present investigation suggests thatACO2 plays a role in the early stage of AD. This find-ing, together with the possibility of evaluating ACO2through a minimally invasive procedure (blood sam-pling), makes this enzyme a potential candidate foruse in monitoring disease progression. Further longi-tudinal studies on large samples are needed to explorewhether ACO2 has a causative role in AD pathogene-sis and to clarify its potential utility as marker for ADprogression.

ACKNOWLEDGMENTS

The authors thank Alessio Farcomeni, PhD, from theDepartment of Public Health and Infectious Diseases,Faculty of Medicine, University “La Sapienza,” Rome,Italy, for statistical support and Lara Macchioni, PhD,and Lanfranco Corazzi from the Department of Exper-imental Medicine, University of Perugia, for supportin laboratory analyses.

This study was supported by a grant of the Ital-ian Ministry of University and Research PRIN 2007prot.20078TC4E5; the American Alzheimer Associa-tion (HAT-10-173121); the Foundation for GeriatricDiseases at Karolinska Institutet (Sweden); and theSwedish foundations Gamla Tjanarinnor, Loo ochHans Ostermans, Ragnhild och Einar LundstromsMinne, Lindhes, and Stohnes. MB was supported by agrant from Regione Umbria, Italy.

The funding sources did not play any role in thedesign and conduct of the study; collection, man-agement, analysis, or interpretation of data; or thepreparation of the manuscript.

Editor Kimberly Kane revised the language in themanuscript.

Authors’ disclosures available online (http://www.j-alz.com/disclosures/view.php?id=2557 ).

REFERENCES

[1] Winblad B, Palmer K, Kivipelto M, Jelic V, Fratiglioni L,Wahlund LO, Nordberg A, Backman L, Albert M, AlmkvistO, Arai H, Basun H, Blennow K, de Leon M, DeCarli C, Erk-injuntti T, Giacobini E, Graff C, Hardy J, Jack C, Jorm A,Ritchie K, van Duijn C, Visser P, Petersen RC (2004) Mildcognitive impairment–beyond controversies, towards a con-sensus: Report of the International Working Group on MildCognitive Impairment. J Intern Med 256, 240-246.

[2] Sayre LM, Perry G, Smith MA (2008) Oxidative stress andneurotoxicity. Chem Res Toxicol 21, 172-188.

[3] Mangialasche F, Polidori MC, Monastero R, Ercolani S,Camarda C, Cecchetti R, Mecocci P (2009) Biomarkers ofoxidative and nitrosative damage in Alzheimer’s disease andmild cognitive impairment. Ageing Res Rev 8, 285-305.

[4] Calabrese V, Sultana R, Scapagnini G, Guagliano E, SapienzaM, Bella R, Kanski J, Pennisi G, Mancuso C, Stella AM, But-terfield DA (2006) Nitrosative stress, cellular stress response,and thiol homeostasis in patients with Alzheimer’s disease.Antioxid Redox Signal 8, 1975-1986.

[5] Mecocci P, Polidori MC, Ingegni T, Cherubini A, Chionne F,Cecchetti R, Senin U (1998) Oxidative damage to DNA inlymphocytes from AD patients. Neurology 51, 1014-1017.

[6] Morocz M, Kalman J, Juhasz A, Sinko I, McGlynn AP,Downes CS, Janka Z, Rasko I (2002) Elevated levels ofoxidative DNA damage in lymphocytes from patients withAlzheimer’s disease. Neurobiol Aging 23, 47-53.

[7] Sultana R, Baglioni M, Cecchetti R, Cai J, Klein JB, BastianiP, Ruggiero C, Mecocci P, Butterfield DA (2014) Lymphocytemitochondria: Toward identification of peripheral biomarkersin the progression of Alzheimer disease. Free Radic Biol Med65, 595-606.

[8] Pratico D, Sung S (2004) Lipid peroxidation and oxidativeimbalance: Early functional events in Alzheimer’s disease.J Alzheimers Dis 6, 171-175.

[9] Mecocci P, Polidori MC, Cherubini A, Ingegni T, MattioliP, Catani M, Rinaldi P, Cecchetti R, Stahl W, Senin U, BealMF (2002) Lymphocyte oxidative DNA damage and plasmaantioxidants in Alzheimer disease. Arch Neurol 59, 794-798.

[10] Rinaldi P, Polidori MCC, Metastasio a, Mariani E, MattioliP, Cherubini a, Catani M, Cecchetti R, Senin U, MecocciP (2003) Plasma antioxidants are similarly depleted in mildcognitive impairment and in Alzheimer’s disease. NeurobiolAging 24, 915-919.

[11] Mangialasche F, Xu W, Kivipelto M, Costanzi E, Ercolani S,Pigliautile M, Cecchetti R, Baglioni M, Simmons A, Soini-nen H, Tsolaki M, Kloszewska I, Vellas B, Lovestone S,Mecocci P (2012) Tocopherols and tocotrienols plasma lev-els are associated with cognitive impairment. Neurobiol Aging33, 2282-2290.

[12] Perluigi M, Sultana R, Cenini G, Di Domenico F, Memo M,Pierce WM, Coccia R, Butterfield DA (2009) Redox pro-teomics identification of 4-hydroxynonenal-modified brainproteins in Alzheimer’s disease: Role of lipid peroxidationin Alzheimer’s disease pathogenesis. Proteomics Clin Appl 3,682-693.

[13] Butterfield DA (2004) Proteomics: A new approach to inves-tigate oxidative stress in Alzheimer’s disease brain. Brain Res1000, 1-7.

[14] Sultana R, Perluigi M, Butterfield DA (2006) Redox pro-teomics identification of oxidatively modified proteins inAlzheimer’s disease brain and in vivo and in vitro modelsof AD centered around Abeta(1-42). J Chromatogr B AnalytTechnol Biomed Life Sci 833, 3-11.

[15] Lin MT, Beal MF (2006) Mitochondrial dysfunction andoxidative stress in neurodegenerative diseases. Nature 443,787-795.

[16] Gardner PR, Raineri I, Epstein LB, White CW (1995)Superoxide radical and iron modulate aconitase activity inmammalian cells. J Biol Chem 270, 13399-13405.

[17] Robbins AH, Stout CD (1989) Structure of activated aconi-tase: Formation of the [4Fe-4S] cluster in the crystal. ProcNatl Acad Sci U S A 86, 3639-3643.

[18] Yarian CS, Toroser D, Sohal RS (2006) Aconitase is the mainfunctional target of aging in the citric acid cycle of kidneymitochondria from mice. Mech Ageing Dev 127, 79-84.

[19] Bulteau A-LL, Ikeda-Saito M, Szweda LI (2003) Redox-dependent modulation of aconitase activity in intactmitochondria. Biochemistry 42, 14846-14855.

In PressF. Mangialasche at al. / Mitochondrial Aconitase in Alzheimer’s Disease 11

[20] Raza H, John A, Brown EM, Benedict S, Kambal A (2008)Alterations in mitochondrial respiratory functions, redoxmetabolism and apoptosis by oxidant 4-hydroxynonenal andantioxidants curcumin and melatonin in PC12 cells. ToxicolAppl Pharmacol 226, 161-168.

[21] Raukas M, Rebane R, Mahlapuu R, Jefremov V, Zilmer K,Karelson E, Bogdanovic N, Zilmer M (2012) Mitochondrialoxidative stress index, activity of redox-sensitive aconitaseand effects of endogenous anti- and pro-oxidants on its activ-ity in control, Alzheimer’s disease and Swedish FamilialAlzheimer’s disease brain. Free Radic Res 46, 1490-1495.

[22] Blalock EM, Geddes JW, Chen KC, Porter NM, MarkesberyWR, Landfield PW (2004) Incipient Alzheimer’s disease:Microarray correlation analyses reveal major transcriptionaland tumor suppressor responses. Proc Natl Acad Sci U S A101, 2173-2178.

[23] Hughes CP, Berg L, Danziger WL, Coben LA, Martin RL(1982) A new clinical scale for the staging of dementia. Br JPsychiatry 140, 566-572.

[24] Folstein MF, Folstein SE, McHugh PR (1975) “Mini-mentalstate". A practical method for grading the cognitive state ofpatients for the clinician. J Psychiatr Res 12, 189-198.

[25] Hamilton M (1959) The assessment of anxiety states by rating.Br J Med Psychol 32, 50-55.

[26] Sheikh JI, Yesavage JA (1986) Geriatric Depression Scale(GDS): Recent evidence and development of a shorter ver-sion. In Clinical Gerontology: A Guide to Assessment andIntervention, Brink TL, ed. Haworth Press, New York,165-173.

[27] American Psychiatric Association (1994) Diagnostic andStatistical Manual of Mental Disorders, Fourth Edition.American Psychiatric Association, Washington, DC.

[28] McKhann G, Drachman D, Folstein M, Katzman R, PriceD, Stadlan EM (1984) Clinical diagnosis of Alzheimer’s dis-ease: Report of the NINCDS-ADRDA Work Group under theauspices of Department of Health and Human Services TaskForce on Alzheimer’s Disease. Neurology 34, 939-944.

[29] Petersen RC, Doody R, Kurz A, Mohs RC, Morris JC, RabinsPV, Ritchie K, Rossor M, Thal L, Winblad B (2001) Currentconcepts in mild cognitive impairment. Arch Neurol 58, 1985-1992.

[30] Petersen RC, Smith GE, Waring SC, Ivnik RJ, Tangalos EG,Kokmen E (1999) Mild cognitive impairment: Clinical char-acterization and outcome. Arch Neurol 56, 303-308.

[31] Kutnink MA, Hawkes WC, Schaus EE, Omaye ST (1987)An internal standard method for the unattended high-performance liquid chromatographic analysis of ascorbic acidin blood components. Annal Biochem 166, 424-430.

[32] Mangialasche F, Kivipelto M, Mecocci P, Rizzuto D, PalmerK, Winblad B, Fratiglioni L (2010) High plasma levels of vita-min E forms and reduced Alzheimer’s disease risk in advancedage. J Alzheimers Dis 20, 1029-1037.

[33] Nierenberg DW, Nann SL (1992) A method for determiningconcentrations of retinol, tocopherol, and five carotenoids inhuman plasma and tissue samples. Am J Clin Nutr 56, 417-426.

[34] Flohe L, Gunzler WA (1984) Assays of glutathione peroxi-dase. Methods Enzymol 105, 114-121.

[35] L’Abbe MR, Fischer PW (1990) Automated assay of super-oxide dismutase in blood. Methods Enzymol 186, 232-237.

[36] Sultana R, Mecocci P, Mangialasche F, Cecchetti R, BaglioniM, Butterfield DA (2011) Increased protein and lipid oxida-tive damage in mitochondria isolated from lymphocytesfrom patients with Alzheimer’s disease: Insights into therole of oxidative stress in Alzheimer’s disease and initial

investigations into a potential biomarker for this dementingdisorder. J Alzheimers Dis 24, 77-84.

[37] Andersson U, Leighton B, Young ME, Blomstrand E, New-sholme EA (1998) Inactivation of aconitase and oxoglutaratedehydrogenase in skeletal muscle in vitro by superoxideanions and/or nitric oxide. Biochem Biophys Res Commun249, 512-516.

[38] Yan LJ, Levine RL, Sohal RS (1997) Oxidative damage duringaging targets mitochondrial aconitase. Proc Natl Acad SciU S A 94, 11168-11172.

[39] Hames BD, Rickwood D (1990) Gel Electrophoresis of Pro-teins. A Practical Approach. IRL Press at Oxford UniversityPress, New York.

[40] Conover WJ, Iman RL (1982) Analysis of covariance usingthe rank transformation. Biometrics 38, 715-724.

[41] Bubber P, Haroutunian V, Fisch G, Blass JP, Gibson GE (2005)Mitochondrial abnormalities in Alzheimer brain: Mechanisticimplications. Ann Neurol 57, 695-703.

[42] Brooks WM, Lynch PJ, Ingle CC, Hatton A, Emson PC,Faull RLM, Starkey MP (2007) Gene expression profiles ofmetabolic enzyme transcripts in Alzheimer’s disease. BrainRes 1127, 127-135.

[43] Baker BM, Haynes CM (2011) Mitochondrial protein qualitycontrol during biogenesis and aging. Trends Biochem Sci 36,254-261.

[44] Luce K, Weil AC, Osiewacz HD (2010) Mitochondrial proteinquality control systems in aging and disease. Adv Exp MedBiol 694, 108-125.

[45] Bota DA, Davies KJ (2002) Lon protease preferentiallydegrades oxidized mitochondrial aconitase by an ATP-stimulated mechanism. Nat Cell Biol 4, 674-680.

[46] Tortora V, Quijano C, Freeman B, Radi R, Castro L, Tortora V(2007) Mitochondrial aconitase reaction with nitric oxide,S-nitrosoglutathione, and peroxynitrite: Mechanisms and rel-ative contributions to aconitase inactivation. Free Radic BiolMed 42, 1075-1088.

[47] Vatassery GT (2004) Impairment of brain mitochondrialoxidative phosphorylation accompanying vitamin E oxida-tion induced by iron or reactive nitrogen species: A selectivereview. Neurochem Res 29, 1951-1959.

[48] Bulteau AL, Planamente S, Jornea L, Dur A, Lesuisse E,Camadro JM, Auchere F (2012) Changes in mitochondrialglutathione levels and protein thiol oxidation in yfh1 yeastcells and the lymphoblasts of patients with Friedreich’s ataxia.Biochim Biophys Acta 1822, 212-225.

[49] Sorolla MA, Reverter-Branchat G, Tamarit J, Ferrer I, Ros J,Cabiscol E (2008) Proteomic and oxidative stress analysis inhuman brain samples of Huntington disease. Free Radic BiolMed 45, 667-678.

[50] Polidori MC, Mattioli P, Aldred S, Cecchetti R, Stahl W, Grif-fiths H, Senin U, Sies H, Mecocci P (2004) Plasma antioxidantstatus, immunoglobulin g oxidation and lipid peroxidationin demented patients: Relevance to Alzheimer disease andvascular dementia. Dement Geriatr Cogn Disord 18, 265-270.

[51] Vatassery GT, DeMaster EG, Lai JC, Smith WE, Quach HT(2004) Iron uncouples oxidative phosphorylation in brainmitochondria isolated from vitamin E-deficient rats. BiochimBiophys Acta 1688, 265-273.

[52] Fariss MW, Chan CB, Patel M, Van Houten B, Orrenius S(2005) Role of mitochondria in toxic oxidative stress. MolInterv 5, 94-111.

[53] Cantu D, Schaack J, Patel M (2009) Oxidative inactivation ofmitochondrial aconitase results in iron and H2O2-mediatedneurotoxicity in rat primary mesencephalic cultures. PLoSOne 4, e7095.

In Press12 F. Mangialasche at al. / Mitochondrial Aconitase in Alzheimer’s Disease

[54] Mangialasche F, Kivipelto M, Mecocci P, Rizzuto D, PalmerK, Winblad B, Fratiglioni L (2010) High plasma levels of vita-min E forms and reduced Alzheimer’s disease risk in advancedage. J Alzheimers Dis 20, 1029-1037.

[55] Mangialasche F, Solomon A, Kareholt I, Hooshmand B, Cec-chetti R, Fratiglioni L, Soininen H, Laatikainen T, Mecocci P,Kivipelto M (2013) Serum levels of vitamin E forms and riskof cognitive impairment in a Finnish cohort of older adults.Exp Gerontol 48, 1428-1435.

[56] Mangialasche F, Westman E, Kivipelto M, Muehlboeck JS,Cecchetti R, Baglioni M, Tarducci R, Gobbi G, Floridi P,Soininen H, Kloszewska I, Tsolaki M, Vellas B, SpengerC, Lovestone S, Wahlund LO, Simmons A, Mecocci P(2013) Classification and prediction of clinical diagnosis ofAlzheimer’s disease based on MRI and plasma measures ofalpha-/gamma-tocotrienols and gamma-tocopherol. J InternMed 273, 602-621.