Article Images (0)
×![]()
featured_play_listSave article to folder
X
Create new folder
Full Text
PDF
Table of Contents
Copyright © 2020 Fracassi et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license, which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed. Research Articles: Neurobiology of Disease Oxidative damage and antioxidant response in frontal cortex of demented and non-demented individuals with Alzheimer’s neuropathology https://doi.org/10.1523/JNEUROSCI.0295-20.2020 Cite as: J. Neurosci 2020; 10.1523/JNEUROSCI.0295-20.2020 Received: 5 February 2020 Revised: 5 October 2020 Accepted: 7 October 2020 This Early Release article has been peer-reviewed and accepted, but has not been through the composition and copyediting processes. The final version may differ slightly in style or formatting and will contain links to any extended data. Alerts: Sign up at www.jneurosci.org/alerts to receive customized email alerts when the fully formatted version of this article is published. 1 Oxidative damage and antioxidant response in frontal cortex of demented and non-demented 1 individuals with Alzheimer’s neuropathology 2 Abbreviated title: NDAN subjects show preserved antioxidant response 3 Anna Fracassi1, Michela Marcatti1, Olga Zolochevska1, Natalie Tabor2, Randall Woltjer3, 4 Sandra Moreno4*, Giulio Taglialatela1* 5 1. Mitchell Center for Neurodegenerative Diseases, Department of Neurology, University of Texas 6 Medical Branch (UTMB), 301 University Blvd, Route 0539, Galveston, Texas, 77550 USA. 7 2. Neuroscience Summer Undergraduate Program, University of Texas Medical Branch (UTMB), 8 Galveston, Texas, 77555, USA. 9 3. Department of Pathology, Oregon Health and Science University, 3181 SW Sam Jackson Park 10 Rd, Portland, OR, USA. 11 4. Department of Science, LIME, University Roma Tre, 446 Viale Marconi, 00146 Rome, Italy. 12 * These authors equally contributed to the work 13 14 Corresponding author email address: Giulio Taglialatela: gtaglial@utmb.edu; Sandra Moreno: 15 sandra.moreno@uniroma3.it 16 17 Number of pages: 42 18 Number of figures and tables: 10 figures, 5 extended data, and 1 table 19 Number of words: abstract (250 words), introduction (678 words), discussion (1,736 words) 20 21 Conflict of interest statement 22 The authors declare no competing financial interests. 23 24 Acknowledgments 25 This work was supported by NIH/NIA grants R01AG042890 and R01AG060718 (to GT) and 26 P30AG008017 (RW pathology core PI), and a grant from the Robert J. and Helen C. Kleberg 27 Foundation to GT, by a travelling fellowship from The Company of Biologists to AF, and by MIUR 28 PhD Fellowship to AF. The Grant of Excellence Departments, MIUR (Art. 1, L.232/2016), is 29 gratefully acknowledged. We also wish to acknowledge Dr. Fiorella Colasuonno, for figures and 30 text editing. 31 32 33 2 Abstract 34 Alzheimer’s disease (AD) is characterized by progressive neurodegeneration in the cerebral cortex, 35 histopathologically hallmarked by amyloid β (Aβ) extracellular plaques and intracellular 36 neurofibrillary tangles, constituted by hyperphosphorylated Tau protein. Correlation between these 37 pathological features and dementia has been challenged by the emergence of "Non-Demented with 38 Alzheimer’s Neuropathology" (NDAN) individuals, cognitively intact despite displaying 39 pathological features of AD. The existence of these subjects suggests that some unknown 40 mechanisms are triggered to resist Aβ-mediated detrimental events. Aβ accumulation affects 41 mitochondrial redox balance, increasing oxidative stress status, which in turn is proposed as a 42 primary culprit in AD pathogenesis. To clarify the relationship linking Aβ, oxidative stress and 43 cognitive impairment, we performed a comparative study on AD, NDAN and aged-matched human 44 post-mortem frontal cortices of either sex. We quantitatively analyzed immunofluorescence 45 distribution of oxidative damage markers, and of SOD2, PGC1α, PPARα, CAT as key factors in 46 antioxidant response, as well as the expression of miRNA-485, as a PGC1α upstream regulator. Our 47 results confirm dramatic redox imbalance, associated with impaired antioxidant defenses in AD 48 brain. By contrast, NDAN individuals display low oxidative damage, associated with high levels of 49 scavenging systems, possibly resulting from lack of PGC1α miRNA-485-related inhibition. 50 Comparative analyses in neurons and astrocytes further highlighted cell-specific mechanisms to 51 counteract redox imbalance. Overall, our data emphasize the importance of transcriptional and post-52 transcriptional regulation of antioxidant response in AD. This suggests that efficient PGC1α-53 dependent “safety mechanism” may prevent Aβ-mediated oxidative stress, supporting 54 neuroprotective therapies aimed at ameliorating defects in antioxidant response pathways in AD 55 patients. 56 Keywords 57 Alzheimer’s disease, NDAN, oxidative stress, superoxide dismutase, PGC1 alpha, PPAR alpha, 58 miRNA-485, amyloid 59 3 SIGNIFICANCE STATEMENT 60 The present study importantly contributes to clarifying the molecular events underlying age-related 61 AD pathology, emphasizing the role of antioxidant defenses against Aβ toxicity. Specifically, we 62 addressed the mechanisms whereby a particular group of individuals, referred to as Non-Demented 63 with AD Neuropathology (NDAN), resists dementia, despite displaying amyloid and Tau pathology 64 consistent with fully symptomatic AD. This study reveals the ability of these individuals to activate 65 an efficient antioxidant response to cope with oxidative stress possibly representing one of the 66 mechanisms by which they remain cognitively intact. Our work, besides advancing the knowledge 67 on the role of oxidative stress in AD, may lay the foundation for novel therapeutic approaches to the 68 disease, possibly based on the activation of PGC1α-mediated antioxidant pathway. 69 70 INTRODUCTION 71 Alzheimer’s disease (AD) is a progressive neurodegenerative disorder, histopathologically 72 characterized by extracellular amyloid β (Aβ) plaques, and intracellular neurofibrillary tangles 73 (Querfurth and LaFerla, 2010; DeTure and Dickson, 2019). Several mechanisms have been 74 proposed to explain AD pathogenesis, among which the so-called amyloid cascade, involving a 75 critical role of Aβ peptide (Hardy and Higgins, 1992; Selkoe and Hardy, 2016). However, the 76 correlation between Aβ accumulation and dementia (Holtzman et al., 2011; Musiek et al., 2015) has 77 been challenged by the emergence of a group of individuals recently classified as A+T+N- (Jack et 78 al., 2018), and here referred to as "Non-Demented with Alzheimer’s Neuropathology" (NDAN). 79 Despite harboring neuropathological features of AD (Bjorklund et al., 2012), they remain 80 cognitively intact (Zolochevska and Taglialatela, 2016). The existence of NDAN suggests that some 81 unknown mechanisms are triggered to resist the detrimental events that otherwise lead to cognitive 82 impairment in AD. Such mechanisms, while not impeding Aβ overproduction or aggregation, 83 possibly prevent neurotoxic effects of the peptide. Noteworthy, Aβ accumulation affects 84 4 mitochondrial redox balance increasing oxidative stress, which has been proposed to be a primary 85 culprit in AD pathogenesis (Mecocci et al., 1994; Hensley et al., 1995; Markesbery 1997; Smith et 86 al., 2000; Butterfield et al 2001; Lauderback et al., 2001; Sultana and Butterfield, 2013; Zhao and 87 Zhao, 2013; Bonda et al., 2014; Wang et al., 2014; Luca et al., 2015; Huang et al., 2016; Sanabria-88 Castro et al., 2017; Cheignon et al., 2018). Based on this evidence, we hypothesize that the 89 resistance to dementia in NDAN patients could be related to their ability to cope with reactive 90 oxygen species (ROS) overproduction, by activating an efficient antioxidant response. 91 Redox imbalance triggers an array of cellular mechanisms, including activation of transcription 92 factors, that regulate energy metabolism and antioxidant defenses (Clark and Simon, 2009). Among 93 these, peroxisome proliferator-activated receptor γ-coactivator 1α (PPARγ coactivator 1α, or 94 PGC1α) regulates genes involved in glucose and lipid metabolism, mitochondrial biogenesis, and 95 antioxidant response (Katsouri et al., 2012, Bagattin et al., 2013, Wenz et al., 2013). PGC1α also 96 coactivates the PPARα isotype, a major regulator of peroxisomal and mitochondrial biogenesis and 97 functions (Feige et al., 2006; Wenz, 2011). PPARα is known to be modulated in neurologic disease, 98 including AD (Fanelli et al., 2013; Porcellotti et al., 2015) and several studies emphasize the 99 possible treatment of AD based on PPARα natural or synthetic ligands (Santos et al., 2005; 100 Inestrosa et al., 2013; Fidaleo et al., 2014; D’Orio et al., 2018). 101 Our previous in vivo investigations (Cimini et al., 2009; Fanelli et al., 2013; Porcellotti et al., 2015), 102 conducted in the Tg2576 mouse model of AD (Hsiao et al., 1996), showed significant variations of 103 antioxidant enzymes expression levels and ensuing oxidative damage at the onset and during the 104 progression of disease. These changes were accompanied by altered expression of both PPARα and 105 PGC1α in mouse hippocampus and neocortex, starting from 3 months of age. 106 To transfer these observations to the human AD brain and clarify the relationship linking Aβ, 107 oxidative stress and cognitive impairment, we performed a comparative study on AD, NDAN and 108 normally-aged human post-mortem frontal cortices, focusing on possible differences concerning 109 antioxidant response mechanisms against oxidative stress. To assess the precise cellular localization 110 5 of oxidative damage, we evaluated the occurrence of 8-oxo-dG marker and 4-hydroxy-2-nonenal in 111 neuronal and astroglial cells, by quantitative double immunofluorescence. To gain information 112 about the specific antioxidant capacity of neurons and astrocytes in AD and NDAN individuals, we 113 studied the expression and distribution of superoxide dismutase 2 (SOD2). Considering the role of 114 PGC1α and PPARα as redox sensors and regulators of SOD2 transcription, we investigated the 115 localization of these factors in neurons and astrocytes. Furthermore, given the central role of 116 peroxisomes in ROS metabolism (Schrader and Fahimi, 2006; Pascual-Ahuir et al., 2017) we 117 investigated the expression and localization of catalase (CAT), whose levels are regulated by 118 PPARs and PGC1α (St-Pierre et al., 2006; Shin et al., 2016). 119 Understanding protective molecular and cellular processes underlying NDAN ability to resist Aβ-120 mediated detrimental effects should be of help to reveal novel targets for the development of 121 effective therapeutic approaches for AD. 122 123 MATERIALS AND METHODS 124 1. Human subjects and autopsy of brain tissues 125 Post-mortem brain tissues were obtained from the Oregon Brain Bank at Oregon Health and 126 Science University (OHSU), in Portland, OR (USA). Donor subjects of either sex were enrolled and 127 clinically evaluated in studies at the NIH-sponsored Layton Aging and AD Center (ADC) at OHSU, 128 in accordance with protocols that were approved by the OHSU Institutional Review Board (IRB). 129 Informed consent was obtained from all participants prior to their enrolment in the studies at the 130 ADC. Subjects were participants in brain aging studies at the ADC and received annual 131 neurological and neuropsychological evaluations, with a clinical dementia rating (CDR) assigned by 132 an experienced clinician. A neuropathological assessment was performed at autopsy, and in 133 compliance with IRB-approved protocols. A neuropathologist scored autopsy brain tissue for Aβ 134 plaques and neurofibrillary tangles, according to standardized CERAD criteria and Braak staging. 135 Participants were classified as AD when possessing a National Institute for Neurological and 136 6 Communicative Disorders and Stroke-Alzheimer’s Disease and Related Disorder Association 137 diagnostic criteria for clinical AD (CDR) including a mini-mental state exam (MMSE) score below 138 10. Control participants performed normally in cognitive examinations (MMSE of 29–30). NDAN 139 cases displayed little to no cognitive impairment (MMSE 27 or greater) though found at autopsy to 140 have amyloid plaques and neurofibrillary tangles comparable to fully symptomatic AD (Table 1). 141 Donor subject samples were de-identified by ADC prior to being provided to University of Texas 142 Medical Branch (UTMB), so that no approval was required from the UTMB IRB under CFR 143 §46.101(a)(1). The cases used in this study are described in Table 1. 144 To ensure that the variations in postmortem interval (PMI) did not affect the measurements, a 145 correlation analysis between PMI values and results obtained in the various assays presented here 146 was performed using a Pearson’s correlation test. No correlation was found between PMI values 147 and any of the elements/antigens assayed here (figure 1), and therefore observed differences could 148 not be attributed to differences in non-specific postmortem tissue degradation. However, even 149 though the results shown in figure 1 reassure on the data regarding the different antigens studied 150 here, it is nonetheless important to appreciate that brains obtained > 10h PMI might not necessarily 151 fully reflect freshly obtained brain tissue. 152 153 2. Tissue processing and immunofluorescence (IF) 154 Fresh frozen cortical tissue blocks (n=6/group) were removed from storage at −80 °C, equilibrated 155 at −20°C, embedded in O.C.T. compound (Cat# 4583, Tissue-Tek,Tokyo, Japan) and sectioned at 156 10 μm onto Superfrost/Plus slides (Cat# 12-550-15, Fisherbrand, ThermoFisher Scientific, 157 Waltham, MA, USA). Prepared slides were stored at −80 °C until use. Slides were fixed in 4% 158 paraformaldehyde in 0.1 M PBS, pH 7.4 for 30 min at room temperature (RT). Non specific binding 159 sites were blocked with 5% bovine serum albumin (BSA, CAT# A4503-100G, Sigma-Aldrich Inc., 160 Saint Louis, MO, USA)/10% normal goat serum (NGS, ThermoFisher Scientific) and sections were 161 permeabilized with 0.5% Triton X-100/0.05% Tween-20 for 1h at RT. Slides were incubated with 162 7 the following primary antibodies, diluted in PBS containing 1.5% NGS/0.25% Triton X-100 163 overnight at 4°C: rabbit anti-PPARα (1:200, Cat# ab8934, RRID:AB_306869, Abcam, Cambridge 164 Science Park, Cambridge, UK); rabbit anti-PGC1α (1:200, Cat# ab54481, RRID:AB_881987, 165 Abcam); rabbit anti-CAT (1:200, Cat# ab16731, RRID:AB_302482, Abcam); rabbit anti-SOD2 166 (1:200; Cat# GTX116093, RRID:AB_10624558, GeneTex, Irvine, CA, USA); mouse anti-8oxo-dG 167 (1:250; Cat# 4354-MC-050, RRID:AB_1857195, R and D Systems, Minneapolis, MN, USA); 168 rabbit anti 4-HNE (1:200, Cat# ab46545, RRID:AB_722490, Abcam); rabbit anti-Aβ (1:200, Cat# 169 ab201060, RRID:AB_2818982, Abcam); mouse anti-NeuN (1:200; Cat# MAB377, 170 RRID:AB_2298772 Millipore, Billerica, MA, USA); rabbit anti-NeuN (1:500, Cat# ABN78, 171 RRID:AB_10807945 Millipore); chicken anti-GFAP (1:500, Cat# GFAP, RRID:AB_2313547, 172 Aves Labs, Tigard, OR, USA). Slides were washed in PBS before incubation with the appropriate 173 Alexa-conjugated secondary antibodies (goat anti-rabbit Alexa Fluor 488, 1:400, Cat# A-11008, 174 RRID:AB_143165; goat anti-mouse Alexa Fluor 594, 1:400 Cat# A-11032, RRID:AB_2534091; 175 goat anti-mouse Alexa Fluor 488, 1:400 Cat# A-10680, RRID:AB_2534062; goat anti-chicken 176 Alexa Fluor 594, 1:400, Cat# A-11042, RRID:AB_2534099; ThermoFisher Scientific) in PBS 177 containing 1.5% NGS/0.25% Triton X-100 for 1 h at RT. Finally, slides were washed in PBS, 178 treated with 0.3% Sudan Black B (in 70% EtOH) for 10 minutes to block lipofuscin 179 autofluorescence, washed again with deionized water, and coverslipped using Fluoromount-G 180 containing 4′,6-diamidino-2-phenylindole (DAPI) (Cat# 0100-20, SouthernBiotech, Birmingham, 181 AL, USA) and sealed. 182 183 3. Quantitative microscopy 184 All immunoreacted sections were acquired with either a Nikon eclipse 80i (Nikon, Minato, Tokyo, 185 Japan) or a Keyence BZ-X800 (Keyence Corporation, Osaka, Japan) microscope, by using 20X 186 and immersion oil 60X objectives. For each subject, 4 sections were analyzed and 5 images per 187 section were captured. Quantitative analysis was performed using ImageJ software 188 8 (https://imagej.nih.gov/ij, NIH, Bethesda, MD, USA), analyzing the intensity of fluorescence for 189 each marker (Integrated Density, IntDen), when the overall distribution was studied. When the 190 colocalization of each marker with either NeuN or GFAP was addressed, the count of the positive 191 cells for each marker and either NeuN or GFAP positive cells over the number of total cells, was 192 made. Representative images were composed in an Adobe Photoshop CC2020 format. 193 194 4. Tissue processing and Western blot analyses 195 Fresh frozen cortical tissue blocks derived from control, AD, and NDAN subjects (n=7/group) were 196 removed from storage at −80 °C and used for Western blotting (Wb) analyses. RIPA buffer (Cat# 197 9806S, Cell Signaling Technology, Inc., Danvers, MA, USA) with 1% protease and phosphatase 198 cocktail inhibitors was used to lysate tissues and synaptosomes to obtain the total protein fraction 199 and the synaptosomal fraction, respectively. The synaptosomes were isolated from the cortical 200 tissues by using a method very well established in our laboratory (Franklin and Taglialatela, 2016; 201 Comerota et al., 2017; Franklin et al, 2019). Briefly, we lysed the cortical tissues by using the 202 SynPER lysis buffer (Cat #87793 ThermoFisher Scientifc) with 1% protease and phosphatase 203 cocktail inhibitors. The brain homogenates were centrifuged at 1,200 × g RCF for 10 min at 4 °C. 204 The supernatants (containing the synaptosomes) were collected and centrifuged at 15,000 × g RCF 205 for 20 min at 4 °C. The synaptosomal pellets were resuspended in HEPES-buffered Krebs-like 206 (HBK) buffer (143.3mM NaCl, 4.75mM KCl, 1.3mM MgSO47H2O, 1.2mM CaCl2, 20.1mM 207 HEPES, 0.1mM NaH2PO4, and 10.3mM D-glucose, pH 7.4). The cytosolic protein fraction was 208 obtained by using the Biovision Nuclear/cytosol fractionation kit (Cat# K266-100, Biovision, 209 Milpitas, CA, USA) according to the manufacture’s protocol. Briefly, the cortical tissues were 210 homogenized in 1-2ml of ice cold PBS and centrifuged at 500 x g for 2-3 min at 4°C. After adding 211 0.2 ml of the CEB-A mix the pellets were vortexed vigorously on the highest setting for 15 sec to 212 be fully resuspend. The suspensions were incubated on ice for 10 minutes and after adding 11 μl of 213 ice-cold Cytosol Extraction Buffer-B the samples were centrifuged for 5 min at maximal speed and 214 9 immediately the supernatants (cytosolic fraction) were transferred in clean pre-chilled tubes. All the 215 protein extracts prepared as above were quantified by using the Pierce™ BCA Protein Assay Kit 216 (Cat# 23225, ThermoFisher Scientific) and subjected to SDS-PAGE. Specifically, the protein 217 expression levels in the single individuals were analyzed by using 20μg of protein extracts. 218 Moreover, an equal amount of proteins extracted from each individual/group was pooled together to 219 obtain a total of three pools (control, AD, and NDAN) and a range of 70 μg to 100 μg of proteins 220 was used. Proteins were transferred to Amersham Protran nitrocellulose transfer membrane (Cat# 221 10600001, Sigma) at 85 V at 4 °C. Membranes were blocked using Odyssey blocking buffer (Cat# 222 927-60001, LI-COR, Lincoln, NE, USA) for 1 hour at RT and probed overnight at 4 °C with either 223 of the following primary antibodies: rabbit anti-PPARα (1:1000, Cat# ab24509, RRID:AB_448110, 224 Abcam); rabbit anti-PGC1α (1:1000, Cat# ab54481, RRID:AB_881987, Abcam); rabbit anti-CAT 225 (1:1000, Cat# ab16731, RRID:AB_302482, Abcam); rabbit anti-SOD2 (1:1000; Cat# GTX116093, 226 RRID:AB_10624558, GeneTex, Irvine, CA, USA); mouse anti-synaptophysin antibody (SYN) 227 (1:10000, Cat# ab8049, RRID:AB_2198854, Abcam), mouse anti-β actin (ACTB) (1:50000, Cat# 228 A1978, RRID:AB_476692, Sigma). All the primary antibodies were prepared in a solution of 1X 229 TBST and Odyssey blocking buffer (1:1). Membranes were than washed three times with 1X TBST 230 for 10 min each and incubated 1 hour with LI-COR secondary antibodies (1:10,000) in 1X 231 TBST/Odyssey blocking buffer at RT. The membranes were again washed three times for 10 min 232 each. Wb were imaged using LI-COR Odyssey infrared imaging system (LI-COR), application 233 software version 3.0.30. The band densities were analyzed using Image J software, normalizing 234 using the densities of the loading control obtained by reprobing the membranes either for ACTB or 235 SYN for total/cytosolic and synaptosomal fractions, respectively. Representative images were 236 composed in an Adobe Photoshop CC2020 format. 237 238 239 240 10 5. Quantitative RT-PCR of miRNAs 241 Total RNA was isolated using Trizol Reagent (Life technologies, Carlsbad, CA, USA) from post-242 mortem frozen human cortices of control, AD and NDAN (n=4/group). Approximately 100 mg of 243 tissue was placed in Trizol and homogenized using the Polytron homogenizer (ThermoFisher 244 Scientific, Waltham, MA, USA). Chloroform was then added and the samples were spun down at 245 12,000 rpm for 15 min at 4ºC. The aqueous phase was transferred to a new tube containing 246 isopropanol. The samples were centrifuged at 12,000 rpm for 10 min at 4ºC. Pellet was washed with 247 ice cold 80% ethanol and air-dried. The samples were resuspended in 40 μl nuclease free water. The 248 RNA concentration was measured using NanoDrop 2000c (ThermoFisher Scientific, Waltham, MA, 249 USA). 250 Reverse transcription was performed using miScript II RT Kit (Cat# 218160, Qiagen, Germantown, 251 MD, USA) according to manufacturer’s protocol. Briefly, 0.5 μg RNA was reverse-transcribed in 252 20 μl reaction volume containing 4 μl 5x HiSpec buffer, 2 μl 10x miScriptNucleics mix and 2 μl 253 miScript Reverse transcriptase. The mix was incubated at 37°C for 1h, then at 95°C for 5 min and 254 placed on ice. The reverse transcribed miRNA mix was diluted with nuclease free water to a final 255 concentration of 3 ng/μl. Real-time PCR was performed to quantitate miRNA in control, AD and 256 NDAN. miScript SYBR Green PCR Kit (Cat# 218073, Qiagen) was used according to 257 manufacturer’s protocol. Briefly, the reaction was performed in 25 μl final volume in each well 258 containing 3 ng reverse transcribed miRNA, 1x SYBR Green, Has_miR-485-5p_1 or Hs_RNU6-2 259 miScript primers (Qiagen). The reaction was performed in Mastercycler epgradient S (Eppendorf, 260 Hamburg, Germany). The samples were incubated at 95°C for 15 min to activate the polymerase 261 followed by 40 cycles of amplification: 94°C for 15 sec, 55ºC for 30 sec and 70°C for 30 sec. 262 Standard melting curve was performed at the end. The levels of miRNA-485 were normalized to U6 263 snRNA. The relative fold change in expression of target miRNAs was determined using the 264 comparative cycle threshold method (2-ΔΔCt), and the obtained values were then log2 transformed. 265 266 11 6. Statistical analysis 267 Statistical analyses were performed using Graphpad Prism 8.4.3 software. T-test, 1way ANOVA 268 with Tukey’s post hoc test or 2way ANOVA with Sudak’s multiple comparison test were used to 269 detect significant differences between groups. Data were then expressed as means ± SD and for all 270 statistical analyses p< 0.05 was considered as statistically significant. 271 272 RESULTS 273 OXIDATIVE DAMAGE AND ANTIOXIDANT RESPONSE 274 1. 8-oxo-dG and 4-HNE distribution in neurons and astrocytes 275 Considering the central role played by oxidative stress in AD pathogenesis and based on our 276 previous data collected on Tg2576 model (Fanelli et al., 2013; Porcellotti et al., 2015), we evaluated 277 oxidative damage, occurring in the frontal cortex of AD, NDAN, and control subjects utilizing 8-278 oxo-dG, as a marker of oxidative DNA/RNA modifications. Interestingly, immunofluorescent 279 staining predominantly localizes to the cytoplasmic compartment, indicating that such oxidative 280 modifications selectively affect mitochondrial and cytosolic nucleic acids, rather than nuclear DNA 281 (figure 2). When quantitatively evaluated by appropriate image analysis, 8-oxo-dG immunoreactive 282 levels appear significantly higher in AD vs. control or NDAN individuals (ctrl vs. AD p<0.0001; 283 AD vs. NDAN p<0.0001). These latter two brain samples indeed display consistently similar 8-oxo-284 dG immunoreactivity (ctrl vs. NDAN p=0.3866) (figure 2A’). To investigate the precise neural 285 localization of the DNA/RNA damage we performed double immunofluorescence of 8-oxo-dG in 286 combination with NeuN- as a neuronal marker -or GFAP, an astroglial marker (figure 2 B-C). In 287 AD frontal cortices, both neurons (figure 2B-B’) (ctrl vs. AD p<0.0001; AD vs. NDAN p<0.0001) 288 and astrocytes (figure 2C-C’) (ctrl vs. AD p<0.0001; AD vs. NDAN p<0.0001) display higher 8-289 oxo-dG immunoreactivity levels with respect to controls. Noteworthy, GFAP immunoreactivity 290 appears especially intense in AD samples, revealing ongoing astrogliosis. By contrast, in NDAN 291 frontal cortices, no astrogliosis was observed and 8-oxo-dG immunostaining was comparable to 292 12 control, in either neurons (ctrl vs. NDAN p=0.0991) (figure 2B-B’), or astrocytes (ctrl vs. NDAN 293 p=0.1018) (figure 2C-C’). Figure 2D summarizes the scenario in neurons and astrocytes in the three 294 conditions highlighting that oxidative damage predominantly occurs in neurons, while astrocytes 295 appear to be more resistant, showing a significantly fainter staining for 8-oxo-dG (ctrl vs. ctrl 296 p<0.0001; AD vs. AD p<0.0001; NDAN vs. NDAN p<0.0001). 297 To further confirm no effect of PMI length on the observed differences among groups, we evaluated 298 the expression of 8-oxo-dG as a representative antigen among those presented here, also using 299 brains samples from a different cohort with exceptionally short PMIs (figure 2-1). The quantitative 300 analysis showed significant differences among the three groups (ctrl vs. AD p=0.0029; AD vs. 301 NDAN p=0.0019; ctrl vs. NDAN p=0.9554) (figure 2-1 A-A’) similar to what observed in our 302 primary case cohort and the Pearson’s correlation test confirmed no correlation between PMI values 303 and the variation of the expression of the antigen tested (figure 2-1B). 304 To determine whether the levels of oxidative damage were associated with amyloid pathology we 305 analyzed the levels of 8-oxo-dG in relation to the accumulation of neurotoxic Aβ peptide (figure 3). 306 A double staining of 8-oxo-dG in combination with an anti-Aβ antibody was perfomed either 307 around or far Aβ plaques (figure 3 A-B). The quantitative analyses of the immunoreactivity (AD vs. 308 NDAN p<0.0001) and the count of 8-oxo-dG+ cells (AD vs. NDAN p<0.0001) showed a significant 309 increase of oxidative damage in AD patients as compared to NDAN in the proximity of Aβ plaques 310 (figure 3A-A’). Similarly, when areas far from plaques were considered, NDAN showed 311 significantly lower immunoreactivity levels of 8-oxo-dG, most comparable to control individuals 312 (figure 3B-B’) in both the analyses performed (IntDens: ctrl vs. AD p<0.0001; AD vs. NDAN 313 p<0.0001; ctrl vs. NDAN p=0.7978; count: ctrl vs. AD p<0.0001; AD vs. NDAN p<0.0001; ctrl vs. 314 NDAN p=0.8616). 315 As a further approach to evaluate oxidative damage, we analyzed the levels and localization of 4-316 hydroxy-2-nonenal (4-HNE), as a lipid peroxidation end-product. 4-HNE is one of the most 317 abundant and cytotoxic lipid-derived alkenal, able to readily react with various cellular components, 318 13 such as DNA, proteins and other molecules (Di Domenico et al., 2017). The immunofluorescent 319 staining revealed both cytoplasmic and nuclear localization, possibly indicating the formation of 4-320 HNE adducts with DNA and/or proteins with significantly higher levels in AD than controls and 321 NDAN individuals (ctrl vs. AD p=0.0003; AD vs. NDAN p=0.0002) (figure 4A-A’). Control and 322 NDAN frontal cortices consistently displayed comparable 4-HNE immunoreactivity levels 323 (p=0.9426) (figure 4A-A’), similarly to what observed for 8-oxo-dG staining. To investigate either 324 the neuronal or astroglial localization of 4-HNE, we performed double IF with NeuN and GFAP, 325 respectively. The quantitative analyses indicated that in AD frontal cortices, both neurons (figure 326 4B-B’ ) (ctrl vs. AD p<0.0001; AD vs. NDAN p<0.0001) and astrocytes (figure 4C-C’) (ctrl vs. AD 327 p<0.0001; AD vs. NDAN p<0.0001) displayed higher 4-HNE immunoreactivity levels with respect 328 to controls and NDAN. Also in this case GFAP immunoreactivity appeared especially strong in AD 329 samples, confirming the ongoing astrogliosis. By contrast, in NDAN frontal cortices, no astrogliosis 330 was observed and comparable levels of 4-HNE with controls were detected in both astrocytes (ctrl 331 vs. NDAN p=0.1862) and neurons (ctrl vs. NDAN p=0.9065). Figure 4D summarizes the scenario 332 in neurons and astrocytes in the three conditions, highlighting that the production of 4-HNE 333 following lipid peroxidation and the possible formation of adducts, are highly distributed in AD 334 neurons and astrocytes. Interestingly, in NDAN subjects the oxidative damage, even if at much 335 lower levels of AD, predominantly occurs in neurons, while astrocytes appear to be more resistant, 336 showing a significantly weaker staining for 4-HNE (ctrl vs. ctrl p=0.0041; AD vs. AD p=0.0300; 337 NDAN vs. NDAN p=0.0017). 338 339 2. SOD2 distribution in neurons and astrocytes. 340 The study of oxidative damage levels prompted us to investigate the antioxidant response status. 341 Particularly, given the well-established role of mitochondrial dysfunction as one of the central 342 cytopathologies of AD (Sweeney and Song, 2016; Swerdlow, 2018; Perez Ortiz and Swerdlow 343 2019), we analyzed the expression of the mitochondrial O2-.–scavenging enzyme SOD2 in frontal 344 14 cortices from AD, NDAN, and normally-aged individuals. Quantitative analysis of IF images 345 revealed that SOD2 was significantly downregulated in AD frontal cortex, as compared to control 346 patients (ctrl vs. AD p<0.0001). On the other hand, NDAN subjects showed overall normal levels of 347 SOD2 (ctrl vs. NDAN p=0.4712; AD vs. NDAN p<0.0001) (figure 5A-A’). In view of the synapses 348 as regions rich in mitochondria, we evaluated the expression of SOD2 in synaptosomes isolated 349 from controls, AD and NDAN frontal cortices. The analyses conducted either on protein extracts 350 from single individual synaptosomal fraction or on pooled extracts confirmed the morphological 351 observations showing significantly lower levels of SOD2 in AD vs. control and NDAN (figure 5-1). 352 To properly investigate its neuronal and astroglial distribution we performed double IF staining for 353 SOD2 in combination with either NeuN, or GFAP (figure 5B-C). Intriguingly, while in control and 354 AD brains the enzyme mainly localized to glial cells (figure 5C-C’) (ctrl vs. AD p<0.0001; ctrl vs. 355 NDAN p<0.0001; AD vs. NDAN p=0.0043), in NDAN samples, predominantly neuronal 356 localization was detected (figure 5B-B’) (ctrl vs. AD p=0.0428; ctrl vs. NDAN p<0.0001; AD vs. 357 NDAN p<0.0001). Figure 5D summarizes the expression and distribution of SOD2 in neurons and 358 astrocytes in all analyzed conditions. While in control and AD brain, astrocytes appear as more 359 protected than neurons against oxidative challenge, the reverse is true for NDAN, where neurons 360 are specifically endowed with high levels of SOD2, even higher than those detected in astrocytes. 361 Figure 5D highlights the higher expression of SOD2 in NDAN neurons suggesting that these 362 individuals could be endowed with a preserved antioxidant response able to counteract redox 363 imbalance (ctrl vs. ctrl p<0.0001; AD vs. AD p<0.0001; NDAN vs. NDAN p<0.0001). 364 365 3. Redox sensors: PGC1α and PPARα distribution in neurons and astrocytes. 366 We analyzed the expression of PGC1α, as a key regulator of antioxidant response, involved in the 367 transcriptional activity of several genes, i.e. SOD2 (St Pierre et al., 2006, Aquilano et al., 2013) . 368 Quantitative analysis of immunofluorescence microscopy images showed lower expression of 369 15 PGC1α in AD, whilst NDAN individuals displayed similar levels to control (ctrl vs. AD p=0.0003; 370 ctrl vs. NDAN p= 0.9785; AD vs. NDAN 0.0002) (figure 6A-A’). 371 The staining was mostly present in astroglial population in both control and AD patients (figure 6C-372 C’) (ctrl vs. AD p<0.0001; ctrl vs. NDAN p<0.0001; AD vs. NDAN p=0.1687). Conversely, in 373 NDAN subjects, PGC1α was mainly localized in neurons (figure 6B-B’) (ctrl vs. AD p<0.0001; ctrl 374 vs. NDAN (p=0.5076); AD vs. NDAN p<0.0001). Figure 6D summarizes the expression and 375 distribution of PGC1α in neurons and astrocytes in all analyzed conditions (ctrl vs. ctrl p=0.0001; 376 AD vs. AD p=0.0001; NDAN vs. NDAN p=0.0028). Wb analyses conducted on total protein 377 extracts confirmed the significant downregulation of PGC1α in AD frontal cortex as compared to 378 controls and NDAN individuals, the latter displaying similar levels to control subjects (figure 6-1). 379 Based on the role of PGC1α as a coactivator of PPARα, we further analyzed the expression and 380 distribution of the latter nuclear receptor, as an important oxidative stress sensor and a regulator of 381 energy metabolism. Extensive analysis of PPARα immunoreacted sections revealed a significantly 382 higher positivity in AD frontal cortex compared to control subjects (ctrl vs. AD p<0.0001). 383 Conversely, though similarly to what was observed for PGC1α, NDAN brains displayed levels of 384 PPARα comparable to controls (ctrl vs. NDAN p=0.5325; AD vs. NDAN p<0.0001) (figure 7A-A’). 385 Wb analyses carried out on controls, AD and NDAN total lysates, confirmed the morphological 386 observations, showing a significant increase of PPARα expression in AD as compared to controls 387 and NDAN individuals (figure 7-1). 388 Somewhat surprisingly, the localization of the nuclear receptor PPARα appeared as both nuclear 389 and cytosolic, irrespective of patients’ group (control, AD and NDAN) (figure 7A). Double 390 immunofluorescence demonstrated prevalent colocalization of PPARα with the astroglial marker 391 GFAP (figure 7C-C’), as compared to the neuron-specific marker NeuN (figure 7B-B’). While this 392 general trend was shared by all groups, a significant increase in AD astrocytes (ctrl vs. AD 393 p<0.0001) accompanied by a decrease in neurons was observed (ctrl vs. AD p<0.0001). Compared 394 to AD patients, NDAN individuals interestingly showed fainter glial immunoreactivity (ctrl vs. 395 16 NDAN p=0.2359; AD vs. NDAN p=0.0014) (figure 7C-C’) and higher neuronal expression (ctrl vs. 396 NDAN p=0.5549; AD vs. NDAN p<0.0001) (figure 7B-B’). Figure 7D displays the relative values 397 of colocalization of PPARα in neurons and astrocytes in all analyzed conditions (ctrl vs. ctrl 398 p=0.0107; AD vs. AD p<0.0001; NDAN vs. NDAN p=0.0003). 399 400 4. CAT distribution in neurons and astrocytes. 401 The increased expression of PPARα in AD patients prompted us to analyze the distribution of one 402 of the major peroxisomal protein, the scavenging enzyme CAT, whose transcription is driven by 403 both PPARα and PGC1α. We observed significantly higher expression of CAT in AD patients, 404 while no significant differences were detected between control and NDAN subjects (ctrl vs. AD 405 p=0.0229; ctrl vs. NDAN p=0.9823; AD vs. NDAN p=0.0161) (figure 8A-A’). Wb analyses 406 performed on cytosolic fractions showed the same trend of IF experiments, confirming a significant 407 increase of CAT in AD patients and no significant variations between controls and NDAN (figure 408 8-1). Interestingly, the highest levels of the peroxisomal enzyme in both AD and NDAN patients 409 were found in astrocytes (ctrl vs. AD p=0.1536; ctrl vs. NDAN p=0.0060; AD vs. NDAN p=0.2376) 410 (figure 8C-C’), whereas a significant downregulation of neuronal CAT was detected in AD 411 individuals (ctrl vs. AD p<0.0001; ctrl vs. NDAN 0.0032 AD vs. NDAN p<0.0001) (figure 8B-B’). 412 Figure 8D summarizes the expression and distribution of CAT in neurons and astrocytes in all the 413 analyzed conditions (ctrl vs. ctrl p=0.0044; AD vs. AD p<0.0001; NDAN vs. NDAN p<0.0001). 414 Somewhat surprisingly, a prominent nuclear rather than the canonical cytosolic localization of the 415 scavenger enzyme was detected in NDAN patients as shown in figures 8B and C. 416 417 REGULATION OF PGC1 VIA MIRNA-485. 418 Given the importance of PGC1α as a key modulator of antioxidant responses and its levels 419 differentially down-regulated in AD and preserved in NDAN, we wanted to further investigate its 420 upstream regulators in the frontal cortices of controls, AD and NDAN subjects. To that end, we 421 measured the tissue levels of miRNA-485 that has been shown, although in non-neuronal tissue, to 422 17 negatively modulate the transcription and expression of PGC1α (Lou et al., 2016). As shown in 423 figure 9, using quantitative RT-PCR we found that the expression of miRNA-485 was significantly 424 increased in AD cortex as compared to control (ctrl vs. AD p=0.007) and NDAN subjects (AD vs. 425 NDAN p=0.003). On the other hand, no significant differences were detected between NDAN 426 individuals and controls (ctrl vs. NDAN p= 0.0909). 427 428 DISCUSSION 429 The aim of this work was to investigate the relationship among amyloid overload, oxidative stress 430 and cellular response elicited by this status. To this purpose, AD, NDAN and normally-aged 431 individuals were comparatively analyzed, focusing on the frontal cortex as one brain region most 432 vulnerable to dementia. This study, highlighting AD-related alterations to pathways regulating 433 cellular redox homeostasis, also sheds light onto the mechanisms allowing NDAN subjects to 434 preserve cognitive functions, despite Aβ toxic insult. 435 Concerning AD patients, redox imbalance was demonstrated by the increased immunofluorescent 436 distribution of the DNA/RNA oxidative damage marker 8-oxo-dG and lipid peroxidation end-437 product 4-HNE. Interestingly, we observed high levels of oxidative damage to nucleic acids in AD 438 patients in the proximity of Aβ plaques, where oligomers are particularly abundant. These findings 439 are consistent with the well-established notion that oligomers are the most toxic species in AD 440 (Selkoe et al., 2008; Sengupta et al., 2016). These results, supporting the current concept that 441 oxidative stress is a major and early causative factor in AD (Mecocci et al., 1994; Hensley et al., 442 1995; Markesbery 1997; Smith et al., 2000; Butterfield et al 2001; Lauderback et al., 2001; Sayre et 443 al., 2008; Sultana and Butterfield, 2013; Zhao and Zhao, 2013; Bonda et al., 2014; Wang et al., 444 2014; Luca et al., 2015; Huang et al., 2016; Jiang et al., 2106; D’Orio et al., 2018; Sanabria-Castro 445 et al., 2017; Cheignon et al., 2018), also correlates with our previous data on Tg2576 mice (Fanelli 446 et al., 2013; Porcellotti et al., 2015). The substantial localization of 8-oxo-dG in the neuronal 447 cytoplasm, already observed in the mouse model (Porcellotti et al., 2015) indicates a prevalence of 448 18 modifications to mitochondrial nucleic acids or cytosolic RNA, consistent with the well-established 449 mitochondrial abnormalities as prominent features of AD (Cai and Tammineni, 2017). On the other 450 hand, the cytosolic and nuclear localization of 4-HNE suggests that this lipid peroxidation end-451 product actively reacts with either cytoplasmic or nuclear proteins, thus increasing the oxidative 452 stress status. Interestingly, the relatively scarce 8-oxo-dG and 4-HNE immunoreactivity in 453 astrocytes suggests that this cell type could be consistently protected from oxidative stress in AD, 454 reflecting cell population specificity, in terms of antioxidant defences. Indeed, the greater resistance 455 of astroglia possibly relates to their higher content in SOD2, consistent with our findings in aged 456 Tg2576 mice (Porcellotti et al., 2015). Noteworthy, and consistent with the dramatic oxidative 457 damage, we found a significant decrease of total SOD2 levels in AD vs. control brains, especially 458 sharp in neurons, thus supporting the idea that altered expression of this mitochondrial enzyme is 459 crucial in the progression of AD pathology (Cimini et al., 2009; Massaad et al., 2009; Fanelli et al., 460 2013; Flynn and Melov, 2013; Hroudova et al., 2014; Porcellotti et al., 2015; Majd and Power 461 2018; Swerdlow, 2018; Perez Ortiz and Swerdlow 2019). In relation to these changes, the levels and 462 distribution of the transcription factor PGC1α was investigated. In agreement with the literature 463 (Qin et al., 2009), we found a significant downregulation of PGC1α in AD brains compared to 464 control, exactly reflecting the expression and localization of its target gene product SOD2. Indeed, 465 PGC1α was mainly localized in astrocytes, in accordance with the localization observed in aged 466 Tg2576 mice (Porcellotti et al., 2015), and supporting the above hypothesis of cell type-based 467 antioxidant response ability. As PGC1α regulates mitochondrial and peroxisomal biogenesis 468 (Austin and St-Pierre, 2012), a correlation between the oxidative stress observed in AD frontal 469 cortices and dysfunctions of these organelles, likely due to PGC1α downregulation, could be 470 hypothesized (Demarquoy and Le Borgne 2015; Sweeney and Song, 2016; Wanders et al., 2016). 471 The reason for such decreased expression may well relate to enhanced levels of miRNA-485, as we 472 assessed by quantitative RT-PCR. This molecule has indeed recently been demonstrated to 473 negatively regulate PGC1α (Lou et al., 2016). 474 19 We analyzed the expression of PPARα, not only for its synergism with PGC1α, but also in view of 475 its roles in energy metabolism and in modulation of neuroinflammation in AD (Feige et al., 2006; 476 Fidaleo et al., 2014). Increased levels of this receptor in AD brains, as compared to controls, were 477 detected, suggesting a possible activation of mechanisms to compensate for mitochondrial 478 dysfunction, likely through stimulation of peroxisomal-based energy metabolism. Indeed, in AD 479 patients, the significant increase of CAT, whose transcription is driven by PPARα (Shin et al., 480 2016), well correlates with this hypothesis. The higher levels of the H2O2 scavenging enzyme might 481 represent an abortive attempt to cope with the oxidative stress activating peroxisomal detoxyfiyng 482 pathways. The predominat astroglial localization of CAT is consistent with the higher oxidative 483 damage found in AD neurons confirming the hypothesis that astrocytes could be more resistant to 484 oxidative damage being characterized of higher levels of antioxidant enzymes. However, the greater 485 amount of antioxidant enzymes results to be unsuccessful due the multiple sources of intracellular 486 and extracellular ROS production (Tönnies and Trushina, 2017; Bodega et al; 2019). Further studies 487 employing appropriate markers and biochemical assays are needed to ascertain putative 488 peroxisomal proliferation and/or activation. 489 Compelling evidence demonstrates that the expression and activity of PPARα are influenced by 490 oxidative stress (Kim and Yang, 2013) and its activation is likely due to oxidized lipids, which may 491 act as specific ligands (Yeldandi et al., 2000). In this context, 4-HNE has been described to act as a 492 PPARα endogenous agonist (Manea et al., 2015). This event, possibly explaining the higher levels 493 of PPARα in AD patients, could however represent a double-edge sword. If on the one hand the 494 activation of PPARα could result in peroxisomal proliferation, on the other hand the resulting 495 increased fatty acid -oxidation might trigger an excessive formation of ROS (Del Rio et al., 2016; 496 Liu et al., 2019; Lismont et al., 2019). Moreover, the predominantly cytosolic immunostaining and 497 astroglial localization could reflect a nongenomic action of PPARα (Feige et al., 2006), particularly 498 related to its anti-inflammatory role in response to Aβ toxicity. Such interpretation is consistent 499 with the remarkable astrogliosis and consequent neuroinflammation occurring at advanced AD 500 20 stages (Birch, 2014; Osborn et al., 2016; González-Reyes et al., 2017). It is possible, however, that 501 the augmented expression of PPARα in AD cortex is not sufficient per se to exert efficient 502 neuroprotective and anti-inflammatory actions, since the available endogenous ligands may be 503 present at very low concentrations (Roy et al., 2016). Supplementing the brain with appropriate 504 dosage of PPARα agonists may thus be of therapeutic value, especially as a supporting treatment, 505 possibly in combination with antioxidants. Among PPARα ligands, naturally occurring substances 506 (e.g., oleoylethanolamide and palmitoylethanolamide), as well as synthetic molecules (e.g., 507 fibrates), have been proven effective in rescuing neurodegeneration, while promoting 508 neuroregeneration, in a number of in vitro and in vivo models of neuropathologies (Fidaleo et al., 509 2014; D’Orio et al., 2018). Feasibility of such trials is encouraged by the current use of several 510 fibrates, in therapeutic protocols against hypercholesterolemia and hyperlipidemia, both of which 511 are recognized as risk factors in AD (Xue-shan et al., 2016; Wong et al., 2017). 512 A completely different scenario from AD patients emerged from the study of NDAN individuals, 513 characterized by lesser susceptibility to oxidative damage, associated with more efficacious 514 antioxidant response. Indeed, analyses performed on NDAN samples revealed remarkable 515 similarities with control brain, rather than AD. In this context, cognitive reserve could play a key 516 role in contributing to the synaptic resilience observed in NDAN individuals, which may display 517 brain flexibility and adaptability leading to cognitive networks that resist or compensate for the 518 effects of AD or aging-related changes (Stern et al., 2019). 519 Interestingly, the very low levels of 8-oxo-dG, found also in the proximity of Aβ plaques, where Aβ 520 oligomers are especially abundant, might be consistent with the idea that NDAN synapses are more 521 resistant to Aβ oligomers (Bjorklund et al., 2012), strengthening the concept of synaptic resilience. 522 Moreover, the relatively low levels of 8-oxo-dG and 4-HNE in NDAN frontal cortex detected in 523 both neurons and astrocytes may well result from preserved antioxidant response mechanisms, 524 including not only SOD2-based scavenging system, but also its major regulator PGC1α. Indeed, 525 unchanged total levels of PGC1α, compared to control subjects, hallmark NDAN frontal cortex. 526 21 This peculiarity is likely due to lack of inhibition by miRNA-485, the expression of which is low in 527 these individuals. It is relevant to point out that PGC1α pattern is mostly due to neuronal 528 contribution, suggesting a specific ability of this cell type in NDAN to activate a compensatory 529 response against superoxide-mediated damage. Accordingly, SOD2 is selectively induced in NDAN 530 neurons, as confirmed also by the high content of SOD2 in NDAN synaptosomes, suggesting 531 maintenance of mitochondrial homeostasis and integrity, despite Aβ insult. Therefore, energy 532 metabolism is likely preserved in NDAN and no PPARα activation is required to induce 533 peroxisomal biogenesis, compensating for mitochondrial dysfunction. Consistent with this 534 hypothesis, NDAN individuals and controls display comparable PPARα patterns. Consistent with 535 these data, no significant differences were found between control and NDAN patients in terms of 536 CAT expression supporting the hypothesis that NDAN subjects are characterized by a preserved 537 antioxidant response even against the accumulation of H2O2 leading to significantly low levels of 538 oxidative stress. The preserved capability to resist the oxidative damage may also correlate with the 539 low levels of 4-HNE, which, as such low levels, has been described to act as a defense mechanism 540 promoting cell survival and proliferation, as well as antioxidant response via an NRF2-mediated 541 pathway (Chen et al., 2005; Breitzig et al., 2016; Di Domenico et al., 2019). 542 Our results on NDAN thus could demonstrate for the first time that an efficient antioxidant 543 response, possibly involving PGC1α might represent a major mechanism by which these individuals 544 resist the detrimental burden of Aβ, thus preventing cognitive impairment. It should be here 545 mentioned that NDAN display lower levels of amyloid oligomers at postsynaptic sites (Bjorklund et 546 al., 2012). Whether this feature is a cause or a consequence of the preserved antioxidant defence is 547 yet to be addressed. It is possible that a vicious cycle linking Aβ, ROS, oxidative damage and 548 antioxidant response (Sayre et al., 2008; Zhao and Zhao, 2013; Sultana and Butterfield, 2013; 549 Bonda et al., 2014; Wang et al., 2014; Luca et al., 2015; Huang et al., 2016; Cheignon et al., 2018) 550 is not triggered in NDAN. This aspect, together with the proposed feature of NDAN individuals to 551 22 display a preserved neurogenesis (Briley et al., 2016), could explain intactness of cognitive abilities 552 in this subjects, while helping to identify novel targets for AD cure. 553 554 CONCLUSIONS 555 The present study brings new evidence to confirm the crucial role of redox imbalance in AD 556 pathogenesis, emphasizing the importance of effective antioxidant defences to cope with A -557 mediated insult. In this context, the comparative analysis of AD vs. NDAN individuals proved 558 especially enlightening in clarifying the role of specific factors. In AD patients, low efficacy of 559 antioxidant response, possibly involving PGC1α as a regulator and SOD2 as an effector, likely 560 allows the vicious cycle linking Aβ, ROS, and oxidative damage to occur, leading to dementia 561 progression. By contrast, the low oxidative damage, correlated with the high content of scavenging 562 systems, observed in NDAN frontal cortex suggests that the ability to activate a PGC1α-dependent 563 “safety-mechanisms” to resist oxidative imbalance might be crucial to prevent Aβ-mediated 564 detrimental effects (figure 10). The analyses conducted in a comparative manner in neurons and 565 astrocytes further highlight cell-specific processes to counteract redox imbalance. 566 These emerging concepts may help envisioning neuroprotective therapies aimed at ameliorating 567 defects in antioxidant response in AD patients. Such treatments might either involve PGC1α 568 induction, possibly by modulation of its inhibitor miRNA-485, or direct PPARα activation by 569 synthetic or natural ligands. This approach would not only improve ROS metabolism, but more 570 generally, could induce mitochondrial and/or peroxisomal biogenesis and functions. 571 572 AUTHOR CONTRIBUTIONS 573 AF made substantial contribution to design of the study, performed morphological experiments, 574 quantitative analyses and prepared draft of the manuscript. MM performed Wb experiments. OZ 575 performed RT-PCR experiments. NT performed morphological experiments and quantitative 576 analyses. RW provided human samples needed for the study, contributed to interpretation of data, 577 23 and critically revised the manuscript. SM coordinated the study, contributed to study design and 578 data interpretation, and critically revised the manuscript and figures. GT conceived, designed and 579 funded the study, and gave final approval of the version to be published. All authors read and 580 approved the final manuscript. 581 582 REFERENCES 583 Aquilano K, Baldelli S, Pagliei B, Cannata SM, Rotilio G, Ciriolo MR (2013) p53 orchestrates 584 the PGC-1alpha-mediated antioxidant response upon mild redox and metabolic imbalance. 585 Antioxid Redox Signal; 18: 386–399. 586 Austin S, St-Pierre J (2012) PGC1α and mitochondrial metabolism--emerging concepts and 587 relevance in ageing and neurodegenerative disorders. J Cell Sci 125:4963-71. 588 Bagattin A, Hugendubler L, Mueller E (2010) Transcriptional coactivator PGC-1alpha promotes 589 peroxisomal remodeling and biogenesis. Proc Natl Acad Sci U S A 107:20376-81. 590 Birch AM (2014) The contribution of astrocytes to Alzheimer's disease. BiochemSocTrans 591 42:1316-20. 592 Bjorklund NL, Reese LC, Sadagoparamanujam VM, Ghirardi V, Woltjer RL, Taglialatela G 593 (2012) Absence of amyloid β oligomers at the postsynapse and regulated synaptic Zn2+ in 594 cognitively intact aged individuals with Alzheimer's disease neuropathology. Mol Neurodegener 595 7:23. 596 Bonda DJ, Wang X, Lee HG, Smith MA, Perry G, Zhu X (2014) Neuronal failure in 597 Alzheimer's disease: a view through the oxidative stress looking-glass. Neurosci Bull 30:243-598 52. 599 Bodega G, Alique M, Puebla L, Carracedo J, Ramírez RM (2019) Microvesicles: ROS 600 scavengers and ROS producers. J Extracell Vesicles 8(1):1626654. 601 24 Breitzig M, Bhimineni C, Lockey R, Kolliputi N (2016) 4-Hydroxy-2-nonenal: a critical target 602 in oxidative stress? Am J Physiol Cell Physiol 311(4):C537‐C543. 603 Briley D, Ghirardi V, Woltjer R, Renck A, Zolochevska O, Taglialatela G, Micci MA (2016) 604 Preserved neurogenesis in non-demented individuals with AD neuropathology. Sci Rep 605 6:27812. 606 Butterfield DA, Drake J, Pocernich C, Castegna A (2001) Evidence of oxidative damage in 607 Alzheimer's disease brain: central role for amyloid beta-peptide. Trends Mol Med. 7(12):548–608 554. 609 Cai Q, Tammineni P (2017) Mitochondrial Aspects of Synaptic Dysfunction in Alzheimer's 610 Disease. J Alzheimers Dis 57:1087-1103. 611 Cheignon C, Tomas M, Bonnefont-Rousselot D, Faller P, Hureau C, Collin F (2018) Oxidative 612 stress and the amyloid beta peptide in Alzheimer's disease. Redox Biol. 14:450-464. 613 Chen ZH, Saito Y, Yoshida Y, Sekine A, Noguchi N, Niki E (2005) 4-Hydroxynonenal induces 614 adaptive response and enhances PC12 cell tolerance primarily through induction of thioredoxin 615 reductase 1 via activation of Nrf2. J Biol Chem. 280(51):41921-7. 616 Cimini A, Moreno S, D'Amelio M, Cristiano L, D'Angelo B, Falone S, Benedetti E, Carrara P, 617 Fanelli F, Cecconi F, Amicarelli F, Cerù MP (2009) Early biochemical and morphological 618 modifications in the brain of a transgenic mouse model of Alzheimer's disease: a role for 619 peroxisomes. J Alzheimers Dis18:935-52. 620 Clark J, Simon DK (2009) Transcribe to survive: transcriptional control of antioxidant defense 621 programs for neuroprotection in Parkinson's disease. Antioxid Redox Signal 11:509-28. 622 Comerota MM, Krishnan B, Taglialatela G (2017) Near infrared light decreases synaptic 623 vulnerability to amyloid beta oligomers. Sci Rep 7(1):15012. 624 25 Del Río LA, López-Huertas E (2016) ROS Generation in Peroxisomes and its Role in Cell 625 Signaling. Plant Cell Physiol 57(7):1364-1376. 626 Demarquoy J, Le Borgne F (2015) Crosstalk between mitochondria and peroxisomes. World J 627 Biol Chem 6:301-9. 628 DeTure MA, Dickson DW (2019) The neuropathological diagnosis of Alzheimer's disease. Mol 629 Neurodegener 14(1):32. 630 D’Orio B, Fracassi A, Cerù MP, Moreno S (2018) Targeting PPARalpha In Alzheimer`s 631 Disease. Curr Alzheimer Res 14:1-10. 632 Di Domenico F, Tramutola A, Butterfield DA (2017) Role of 4-hydroxy-2-nonenal (HNE) in 633 the pathogenesis of Alzheimer disease and other selected age-related neurodegenerative 634 disorders. Free Radic Biol Med 111:253-261. 635 Fanelli F, Sepe S, D'Amelio M, Bernardi C, Cristiano L, Cimini A, Cecconi F, Cerù MP, 636 Moreno S (2013) Age-dependent roles of peroxisomes in the hippocampus of a transgenic 637 mouse model of Alzheimer's disease. Mol Neurodegener 8:8. 638 Feige JN, Gelman L, Michalik L, Desvergne B, Wahli W (2006) From molecular action to 639 physiological outputs: peroxisome proliferator-activated receptors are nuclear receptors at the 640 crossroads of key cellular functions. Prog Lipid Res 45:120-59. 641 Fidaleo M, Fanelli F, Cerù MP, Moreno S (2014) Neuroprotective properties of peroxisome 642 proliferator-activated receptor alpha (PPARα) and its lipid ligands. Curr Med Chem 21:2803-21. 643 Flynn JM, Melov S (2013) SOD2 in mitochondrial dysfunction and neurodegeneration. Free 644 Radic Biol Med 62:4-12. 645 Franklin W, Taglialatela G (2016) A method to determine insulin responsiveness in 646 synaptosomes isolated from frozen brain tissue. J Neurosci Methods 261:128-34. 647 26 Franklin W, Krishnan B, Taglialatela G (2019) Chronic synaptic insulin resistance after 648 traumatic brain injury abolishes insulin protection from amyloid beta and tau oligomer-induced 649 synaptic dysfunction. Sci Rep 9(1):8228. 650 González-Reyes RE, Nava-Mesa MO, Vargas-Sánchez K, Ariza-Salamanca D, Mora-Muñoz L 651 (2017) Involvement of Astrocytes in Alzheimer's Disease from a Neuroinflammatory and 652 Oxidative Stress Perspective. Front Mol Neurosci 19;10:427. 653 Hardy JA, Higgins GA (1992) Alzheimer's disease: the amyloid cascade hypothesis. Science 654 256:184-5. 655 Hensley K, Hall N, Subramaniam R, Cole P, Harris M, Aksenov M, Aksenova M, Gabbita SP, 656 Wu JF, Carney JM, Lovell M, Markesbery WR, Butterfield DA (1995) Brain regional 657 correspondence between Alzheimer's disease histopathology and biomarkers of protein 658 oxidation. J Neurochem. 65(5):2146-56. 659 Holtzman DM, Morris JC, Goate AM (2011) Alzheimer’s disease: the challenge of the second 660 century. Sci Transl Med 3:77sr1. 661 Hroudová J, Singh N, Fišar Z (2014) Mitochondrial dysfunctions in neurodegenerative diseases: 662 relevance to Alzheimer's disease. Biomed Res Int 2014:175062. 663 Huang WJ, Zhang X, Chen WW (2016) Role of oxidative stress in Alzheimer's disease. Biomed 664 Rep 4:519-522. 665 Inestrosa NC, Carvajal FJ, Zolezzi JM, Tapia-Rojas C, Serrano F, Karmelic D, Toledo EM, 666 Toro S, Toro J, Santos MJ (2013) Peroxisome proliferators reduce spatial memory impairment, 667 synaptic failure, and neurodegeneration in brains of a double transgenic mice model of 668 Alzheimer's disease. J Alzheimers Dis 33:941-59. 669 Jack CR Jr, Bennett DA, Blennow K, Carrillo MC, Dunn B, Haeberlein SB, Holtzman DM, 670 Jagust W, Jessen F, Karlawish J, Liu E, Molinuevo JL, Montine T, Phelps C, Rankin KP, Rowe 671 27 CC, Scheltens P, Siemers E, Snyder HM, Sperling R; Contributors (2018) NIA-AA Research 672 Framework: Toward a biological definition of Alzheimer's disease. Alzheimers Dement 673 14(4):535-562. 674 Jiang T, Sun Q, Chen S (2016) Oxidative stress: A major pathogenesis and potential therapeutic 675 target of antioxidative agents in Parkinson's disease and Alzheimer's disease. Prog Neurobiol 676 147:1-19. 677 Katsouri L, Blondrath K, Sastre M (2012) Peroxisome proliferator-activated receptor-γ 678 cofactors in neurodegeneration. IUBMB Life. 64:958-64. 679 Kim GH, Kim JE, Rhie SJ, Yoon S (2015) The Role of Oxidative Stress in Neurodegenerative 680 Diseases. Exp Neurobiol 24:325-40. 681 Kim T, Yang Q (2013) Peroxisome-proliferator-activated receptors regulate redox signaling in 682 the cardiovascular system. World. J Cardiol 5:164-74. 683 Lauderback CM, Hackett JM, Huang FF, Keller JN, Szweda LI, Markesbery WR, Butterfield 684 DA (2001) The glial glutamate transporter, GLT-1, is oxidatively modified by 4-hydroxy-2-685 nonenal in the Alzheimer's disease brain: the role of Abeta1-42 J Neurochem. 78(2):413–416 686 Lismont C, Revenco I, Fransen M (2019) Peroxisomal Hydrogen Peroxide Metabolism and 687 Signaling in Health and Disease Int J Mol Sci. 20(15):3673 688 Liu J, Lu W, Shi B, Klein S, Su X (2019) Peroxisomal regulation of redox homeostasis and 689 adipocyte metabolism. Redox Biol 24:101167. 690 Lou C, Xiao M, Cheng, S, Lu X, JiaS, Ren Y, Li Z (2016) MiR-485-3p and miR-485-5p 691 suppress breast cancer cell metastasis by inhibiting PGC-1α expression. Cell Death Dis. 692 7:e2159. 693 28 Luca M, Luca A, Calandra C (2015) The Role of Oxidative Damage in the Pathogenesis and 694 Progression of Alzheimer's Disease and Vascular Dementia. Oxid Med Cell Longev 695 2015:504678. 696 Majd S, Power JHT (2018) Oxidative Stress and Decreased Mitochondrial Superoxide 697 Dismutase 2 and Peroxiredoxins 1 and 4 Based Mechanism of Concurrent Activation of AMPK 698 and mTOR in Alzheimer's Disease. Curr Alzheimer Res 15(8):764–776. 699 Manea A, Manea SA, Todirita A, Albulescu IC, Raicu M, Sasson S, Simionescu M (2015) 700 High-glucose-increased expression and activation of NADPH oxidase in human vascular 701 smooth muscle cells is mediated by 4-hydroxynonenal-activated PPARα and PPARβ/δ. Cell 702 Tissue Res 361(2):593-604. 703 Markesbery WR (1997) Oxidative stress hypothesis in Alzheimer's disease. Free Radic Biol 704 Med. 23(1):134–147. 705 Massaad CA, Pautler RG, Klann E (2009) Mitochondrial superoxide: a key player in 706 Alzheimer's disease. Aging US 1:758-61. 707 Mecocci P, MacGarvey U, Beal MF (1994). Oxidative damage to mitochondrial DNA is 708 increased in Alzheimer's disease. Ann Neurol. 36(5):747-51. 709 Musiek ES, Holtzman DM (2015) Three dimensions of the amyloid hypothesis: time, space and 710 'wingmen'. Nat Neurosci 18:800-6. 711 Osborn LM, Kamphuis W, Wadman WJ, Hol EM (2016) Astrogliosis: An integral player in the 712 pathogenesis of Alzheimer's disease. Prog Neurobiol 144:121-41. 713 Pascual-Ahuir A, Manzanares-Estreder S, Proft M (2017) Pro- and Antioxidant Functions of the 714 Peroxisome-Mitochondria Connection and Its Impact on Aging and Disease. Oxid Med Cell 715 Longev 2017;2017:9860841. 716 29 Perez Ortiz JM, Swerdlow RH (2019) Mitochondrial dysfunction in Alzheimer's disease: Role 717 in pathogenesis and novel therapeutic opportunities. Br J Pharmacol 176(18):3489–3507. 718 Porcellotti S, Fanelli F, Fracassi A, Sepe S, Cecconi F, Bernardi C, Cimini A, Cerù MP, Moreno 719 S (2015) Oxidative Stress during the Progression of β-Amyloid Pathology in the Neocortex of 720 the Tg2576 Mouse Model of Alzheimer’s Disease. Oxid Med Cell Longev 2015:967203. 721 Qin W, Haroutunian V, Katsel P, Cardozo CP, Ho L, Buxbaum JD, Pasinetti GM (2009) PGC-722 1alpha expression decreases in the Alzheimer disease brain as a function of dementia. Arch 723 Neurol 66:352-61. 724 Querfurth HW, LaFerla FM (2010) Alzheimer's disease. New Engl J Med 362:329-44. 725 Reddy PH (2014) Misfolded proteins, mitochondrial dysfunction, and neurodegenerative 726 diseases. Biochim Biophys Acta 1842:1167. 727 Roy A, Kundu M, Jana M, Mishra RK, Yung Y, Luan CH, Gonzalez FJ, Pahan K (2016) 728 Identification and characterization of PPARα ligands in the hippocampus. Nat Chem Biol 729 12:1075-1083. 730 Sanabria-Castro A, Alvarado-Echeverría I, Monge-Bonilla C (2017) Molecular Pathogenesis of 731 Alzheimer's Disease: An Update. Ann Neurosci 24:46-54. 732 Santos MJ, Quintanilla RA, Toro A, Grandy R, Dinamarca MC, Godoy JA, Inestrosa NC, 733 (2005) Peroxisomal proliferation protects from beta-amyloid neurodegeneration. J Biol Chem 734 280:41057-68. 735 Sayre LM, Perry G, Smith MA (2008) Oxidative stress and neurotoxicity. Chem Res Toxicol 736 21:172-88. 737 Schrader M1, Fahimi HD (2006) Peroxisomes and oxidative stress. Biochim Biophys Acta 738 1763(12):1755-66. 739 30 Selkoe DJ (2008) Soluble oligomers of the amyloid beta-protein impair synaptic plasticity and 740 behavior. Behav Brain Res 192:106-13. 741 Selkoe DJ, Hardy J (2016) The amyloid hypothesis of Alzheimer's disease at 25 years. EMBO 742 Mol Med 8:595-608. 743 Sengupta U, Nilson AN, Kayed R (2016) The Role of Amyloid-β Oligomers in Toxicity, 744 Propagation, and Immunotherapy. EBioMedicine 6:42-49. 745 Shin MH, Lee SR Kim MK Shin CY Lee DH Chung JH (2016) Activation of Peroxisome 746 Proliferator-Activated Receptor Alpha Improves Aged and UV-Irradiated Skin by Catalase 747 Induction. PLoS One 11(9):e0162628. 748 Smith MA, Rottkamp CA, Nunomura A, Raina AK, Perry G (2000) Oxidative stress in 749 Alzheimer's disease. Biochim Biophys Acta 1502(1):139–144. 750 Stern Y, Barnes CA, Grady C, Jones RN, Raz N (2019) Brain reserve, cognitive reserve, 751 compensation, and maintenance: operationalization, validity, and mechanisms of cognitive 752 resilience. Neurobiol Aging 83:124-129. 753 St-Pierre J, Drori S, Uldry M, Silvaggi JM, Rhee J, Jäger S, Handschin C, Zheng K, Lin J, Yang 754 W, Simon DK, Bachoo R, Spiegelman BM (2006) Suppression of reactive oxygen species and 755 neurodegeneration by the PGC-1 transcriptional coactivators. Cell 127(2):397-408. 756 Sultana R, Butterfield DA (2013) Oxidative modification of brain proteins in Alzheimer's 757 disease: perspective on future studies based on results of redox proteomics studies. J Alzheimers 758 Dis 33 Suppl 1:S243-51. 759 Sweeney G, Song J (2016) The association between PGC-1α and Alzheimer’s disease. Anat 760 Cell Biol 49:1-6. 761 31 Swerdlow RH (2018) Mitochondria and Mitochondrial Cascades in Alzheimer’s Disease. J 762 Alzheimers Dis 62(3):1403–1416. 763 Tönnies E, Trushina E (2017) Oxidative Stress, Synaptic Dysfunction, and Alzheimer's Disease. 764 J Alzheimers Dis 2017;57(4):1105-1121. 765 Wanders RJ, Waterham HR, Ferdinandusse S (2016) Metabolic Interplay between Peroxisomes 766 and Other Subcellular Organelles Including Mitochondria and the Endoplasmic Reticulum. 767 Front Cell Dev Biol 3:83. 768 Wang X, Wang W, Li L, Perry G, Lee HG, ZhuX (2014) Oxidative stress and mitochondrial 769 dysfunction in Alzheimer's disease. Biochim Biophys Acta 1842:1240-7. 770 Wenz T (2011) Mitochondria and PGC-1α in Aging and Age-Associated Diseases. J Aging Res 771 2011:810619. 772 Wenz T (2013) Regulation of mitochondrial biogenesis and PGC-1α under cellular stress. 773 Mitochondrion 13:134-142. 774 Wong MW, Braidy N, Poljak A, Pickford R, Thambisetty M, Sachdev PS (2017) Dysregulation 775 of lipids in Alzheimer’s disease and their role as potential biomarkers. Alzheimers Dement 776 13:810-827. 777 Xue-shan Z, Juan P, Qi W, Zhong R, Li-hong P, Zhi-han T, Zhi-sheng J, Gui-xue W, Lu-shan L 778 (2016) Imbalanced cholesterol metabolism in Alzheimer's disease. Clin Chim Acta 456:107-14. 779 Yeldandi AV, Rao MS, Reddy JK (2000), Hydrogen peroxide generation in peroxisome 780 proliferator-induced oncogenesis. Mutat Res 448:159-77. 781 Zhao Y, Zhao B (2013) Oxidative stress and the pathogenesis of Alzheimer's disease. Oxid Med 782 Cell Longev 2013:316523. 783 32 Zolochevska O, Taglialatela G (2016) Non-Demented Individuals with Alzheimer's Disease 784 Neuropathology: Resistance to Cognitive Decline May Reveal New Treatment Strategies. Curr 785 Pharm Des 22:4063-8. 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 33 TABLES 804 809 34 FIGURE LEGENDS 810 Figure 1. Correlation analysis between each of the studied parameters and PMI values across all the 811 assayed speciemens. A Pearson’s correlation test was performed for each measurement against the 812 PMI. Correlation coefficient (r) and p values are noted in the individual plots showing no significant 813 correlation with PMI values. 814 expressed as mean ± SD. **** p<0.0001. (B-D) 8-oxo-dG expression and distribution in frontal 820 cortex neurons and astrocytes of control, AD and NDAN subjects. (B-B’) Double IF of 8-oxo-821 dG (green) in combination with NeuN (red) shows high levels of oxidative damage in AD neurons. 822 NDAN neurons demonstrate low levels of oxidative damage marker. Magnification 60X, scale bar 823 30 μm. The quantitative analysis of IF images shows a significantly higher levels of oxidative 824 damage marker in AD neurons. Statistical analyses were made using 1way ANOVA (F(2,15) = 825 48.47, p<0.0001) following Tukey’s multiple comparisons test. Values are expressed as mean ± SD. 826 **** p<0.0001. (C-C’) Double IF of 8-oxo-dG (green) in combination with GFAP (red) showing 827 high levels of oxidative damage to astrocytes in AD compared to control and NDAN, although less 828 than in neurons. Magnification 60X, scale bar 30 μm. The quantitative analysis of IF images shows 829 a significantly higher levels of the oxidative damage marker in AD, while NDAN and control 830 astrocytes displayed comparable levels of damage. Statistical analyses were made using 1way 831 ANOVA (F(2,15) = 37.64, p<0.0001) following Tukey’s multiple comparisons test. Values are 832 expressed as mean ± SD. **** p<0.0001. (D) The analysis demonstrates relatively higher resistance 833 of astrocytes to oxidative damage, compared to neurons, which appear more prone to AD-834 35 associated oxidative damage. Statistical analyses were made using 2way ANOVA (F (2,30)=80, 835 p<0.0001). Values are expressed as mean ± SD. **** p< 0.0001. 836 Figure 3. 8-oxo-dG expression and distribution in relation to Aβ accumulation. (A-A’) Double 837 IF of 8-oxo-dG (green) and Aβ (red) showing the oxidative damage to nucleic acids around Aβ 838 plaques in AD and NDAN subjects. The quantitative analyses in terms of intensity of fluorescence 839 (t(10)=13.06, p<0.0001, unpaired t-test) and number of 8-oxo-dG+ cells (t(10)=15.02, p<0.0001, 840 unpaired t-test) show increased levels of oxidative damage around amyloid plaques in AD as 841 compared to NDAN individuals. Original magnification 60X, scale bar 30 μm. Values are 842 expressed as mean ± SD.**** p<0.0001. (B-B’) Immunostaining of 8-oxo-dG and Aβ showing 843 significant high levels of oxidative damage in AD subjects as compared to controls and NDAN 844 subjects even far from Aβ plaques. Statistical analyses were made using 1way ANOVA (IntDens: 845 F(2,15)=122.1, p<0.000; count: F (2,15)=42.34, p<0.0001) following Tukey’s multiple comparisons 846 test. Original magnification 60X, scale bar 30 μm. Values are expressed as mean ± SD. **** 847 p<0.0001. 848 Values are expressed as mean ± SD. *** p< 0.001. (B-D) 4-HNE expression and distribution in 854 frontal cortex neurons and astrocytes of control, AD and NDAN subjects. (B’B’) Double IF of 855 4-HNE (green) in combination with NeuN (red) shows high levels of oxidative damage in AD 856 neurons. NDAN neurons demonstrate low levels of lipid peroxidation marker. Magnification 60X, 857 scale bar 30 μm. The quantitative analysis of IF images shows a significantly higher levels of 858 oxidative damage marker in AD neurons. Statistical analyses were made using 1way ANOVA 859 (F(2,15)=45.43, p<0.0001) following Tukey’s test multiple comparisons test. Values are expressed 860 36 as mean ± SD. **** p< 0.0001. (C) Double IF of 4-HNE (green) in combination with GFAP (red) 861 showing high levels of oxidative damage to astrocytes in AD as compared to control and NDAN, 862 although less than in neurons. Magnification 60X, scale bar 30 μm. (C’) The quantitative analysis 863 of IF images shows a significantly higher levels of the oxidative damage marker in AD, while 864 NDAN and control astrocytes displayed comparable levels of damage. Statistical analyses were 865 made using 1way ANOVA (F(2,15)=407.9, p<0.0001) following Tukey’s test multiple comparisons 866 test. Values are expressed as mean ± SD. **** p< 0.0001. (D) The analysis demonstrates the 867 significant slightely higher distribution of lipid peroxidation end-product in AD neurons vs. 868 astrocytes, and the relatively higher resistance of astrocytes to oxidative damage in NDAN. 869 Statistical analyses were made using 2way ANOVA (F (2,30)=172.5, p<0.0001). Values are 870 expressed as mean ± SD. * p< 0.05; ** p< 0.01. 871 Figure 5. (A-A’) SOD2 expression in frontal cortex of control, AD and NDAN subjects. IF 872 images and quantitative analyses showing a significant downregulation of SOD2 in AD patients and 873 preserved levels of SOD2 in NDAN individuals, as compared to controls individuals. Magnification 874 60X, scale bar 30 μm. Statistical analyses were made using 1way ANOVA (F(2,15)=30.82, 875 p<0.0001) following Tukey’s test multiple comparisons test. Values are expressed as mean ± SD. 876 ****p< 0.0001. (B-D) SOD2 expression and distribution in frontal cortex neurons and 877 astrocytes of control, AD and NDAN subjects. (B-B’) Double IF of SOD2 (green) in combination 878 with NeuN (red) and quantitative analysis showing significant low levels of the antioxidant enzyme 879 in AD neurons and significantly higher levels in NDAN neurons, as compared to control. 880 Magnification 60X, scale bar 30 μm. Statistical analyses were made using 1way ANOVA 881 (F(2,15)=151.8, p<0.0001) following Tukey’s test multiple comparisons test. Values are expressed 882 as mean ± SD. * p< 0.05; **** p< 0.0001. (C-C’) Double IF of SOD2 (green) in combination with 883 GFAP (red) and quantitative analysis of images showing the downregulation of the antioxidant 884 enzyme in AD and NDAN, while in AD brains SOD2 mainly localizes to astrocytes. Magnification 885 60X, scale bar 30 μm. Statistical analyses were made using 1way ANOVA (F(2,15)=68.34, 886 37 p<0.0001) following Tukey’s test multiple comparisons test. Values are expressed as mean ± SD. ** 887 p< 0.01; **** p<0.0001. (D) The diagram shows an impairment of the antioxidant response in AD 888 and a preserved scavenging system in NDAN. Significantly higher levels of SOD2 in neurons and 889 astrocytes of NDAN and AD, respectively, are highlighted. Statistical analyses were made using 890 2way ANOVA (F (2,30)=39, p<0.0001). Values are expressed as mean ± SD. **** p< 0.0001. 891 are expressed as mean ± SD. *** p< 0.001. (B-D) PGC1 expression and distribution in frontal 896 cortex neurons and astrocytes of control, AD and NDAN subjects. (B-B’) Double IF of PGC1α 897 (green) in combination with NeuN (red) showing significantly lower levels of the transcription 898 factor in AD neurons and preserved levels in NDAN. Magnification 60X, scale bar 30 μm. 899 Quantitative analysis of IF images shows significantly higher levels of PGC1α in NDAN neurons as 900 compared to AD. Statistical analyses were made using 1way ANOVA (F(2,15)=134, p<0.0001) 901 following Tukey’s test multiple comparisons test. Values are expressed as mean ± SD. **** p< 902 0.0001. (C-C’) Double IF of PGC1α (green) in combination with GFAP (red) and quantitative 903 analysis showing the downregulation of the transcription factor in AD and NDAN astrocytes as 904 compared to controls. Magnification 60X, scale bar 30 μm. Statistical analyses were made using 905 1way ANOVA (F(2,15)=54.69, p<0.0001) following Tukey’s test multiple comparisons test. Values 906 are expressed as mean ± SD. **** p< 0.0001. (D) The analysis shows a downregulation of PGC1 907 in AD frontal cortex although with a prevalent localization in astrocytes as compared to neurons. 908 Conversely, NDAN and control display comparable levels of PGC1 , with a significant increase in 909 neurons. Statistical analyses were made using 2way ANOVA (F (2,30)=134.8, p<0.0001). Values 910 are expressed as mean ± SD. ** p<0.01; *** p< 0.001. 911 38 Figure 7. (A-A’) PPAR expression in frontal cortex of control, AD and NDAN subjects. The 912 quantitative analyses of the IF images showing upregulation of PPAR in AD compared to control. 913 NDAN and control show comparable levels of the nuclear receptor. Magnification 60X, scale bar 914 30 μm. Statistical analyses were made using 1way ANOVA (F(2,15)=52.78, p<0.0001) following 915 Tukey’s test multiple comparisons test. Values are expressed as mean ± SD. **** p< 0.0001. (B-D) 916 PPAR expression and distribution in frontal cortex neurons and astrocytes of control, AD 917 and NDAN subjects. (B-B’) Double IF of PPARα (green) in combination with NeuN (red) 918 showing significant downregulation of the nuclear receptor in AD neurons. Magnification 60X, 919 scale bar 30 μm. Quantitative analysis of IF images showing a similar neuronal localization of 920 PPARα in NDAN compared to AD. Statistical analyses were made using 1way ANOVA 921 (F(2,15)=31.94, p<0.0001) following Tukey’s test multiple comparisons test. Values are expressed 922 as mean ± SD. **** p< 0.0001. (C-C’) Double IF of PPARα (green) in combination with GFAP 923 (red) and quantitative analysis showing a predominant localization of the nuclear receptor in AD 924 astrocytes, while NDAN and control astrocytes display comparable levels of PPAR . Magnification 925 60X, scale bar 30 μm. Statistical analyses were made using 1way ANOVA (F(2,15)=19.85, 926 p<0.0001) following Tukey’s test multiple comparisons test. Values are expressed as mean ± SD. ** 927 p<0.01; **** p< 0.0001. (D) The analysis shows the significant upregulation of PPAR in AD 928 astrocytes. NDAN and control subjects display comparable levels of PPAR in both neurons and 929 astrocytes. Statistical analyses were made using 2way ANOVA (F (2,30)=42.54, p<0.0001). Values 930 are expressed as mean ± SD. * p<0.05; *** p< 0.001; **** p< 0.0001. 931 Figure 8. (A-A’) CAT expression in frontal cortex of control, AD and NDAN subjects. (a) The 932 quantitative analyses of the IF images showing upregulation of CAT in AD compared to control. 933 NDAN and control show comparable levels of the antioxidant enzyme. Magnification 60X, scale 934 bar 30 μm. Statistical analyses were made using 1way ANOVA (F(2,15)=6.387, p=0.0099) 935 following Tukey’s test multiple comparisons test. Values are expressed as mean ± SD. * p< 0.05. 936 (B-D) CAT expression and distribution in frontal cortex neurons and astrocytes of control, 937 39 AD and NDAN subjects. (B-B’) Double IF of CAT (green) in combination with NeuN (red) and 938 quantitative analysis showing significant lower levels of the antioxidant enzyme in AD neurons as 939 compared to control, and significantly higher levels in NDAN neurons vs. AD. Magnification 60X, 940 scale bar 30 μm. Statistical analyses were made using 1way ANOVA (F(2,15)=50.94, p<0.0001) 941 following Tukey’s test multiple comparisons test. Values are expressed as mean ± SD. ** p<0.01; 942 **** p<0.0001. (C-C’) Double IF of CAT (green) in combination with GFAP (red) showing 943 predominat nuclear localization of the H2O2 scavenging-enzymes in NDAN patients. Quantitative 944 analysis of images showing no significant changes of CAT expression in astrocytes. Magnification 945 60X, scale bar 30 μm. Statistical analyses were made using 1way ANOVA (F(2,15)=6.752, ** 946 p=0.0081) following Tukey’s test multiple comparisons test. Values are expressed as mean ± SD. ** 947 p<0.01. (D) The diagram shows a significantly predominat localization of CAT in astrocytes rather 948 than in neurons in all the three considered conditions. Statistical analyses were made using 2way 949 ANOVA (F (2,30)=28, p<0.0001). Values are expressed as mean ± SD. ** p<0.01; **** p< 0.0001. 950 expressed as mean ± SD. ** p<0.01; *** p< 0.001. 955 Figure 10. Antioxidant response and oxidative stress in AD and NDAN frontal cortex. a) Aβ 956 plays a critical role in AD pathogenesis leading to mitochondrial alterations in terms of biogenesis 957 and functions. Downregulation of PGC1 , possibly inhibited by high levels of miRNA-485, and its 958 target gene SOD2 contribute to energy dysmetabolism. Mitochondrial dysfunction and antioxidant 959 response impairment lead to ROS increaseand oxidative stress, affecting both mitochondria and 960 peroxisomes (yellow lightning). PPAR increase, in response to redox imbalance, may activate a 961 peroxisomal based energy metabolism, as well a ROS detoxifying mechanism (dotted lines), 962 compensating for mitochondrial dysfunction. b) In the frontal cortex of NDAN subjects the lack of 963 40 PGC1 miRNA-485-related inhibition results in unchanged levels of PGC1α and SOD2, and thus 964 preserved antioxidant response and mitochondrial integrity, blunting oxidative damage. This 965 suggests that the activation of a PGC1α-dependent response, to cope with the redox imbalance, is 966 crucial to prevent Aβ-mediated toxicity. The unchanged levels of PPARα keep peroxisomes at a 967 physiological level. Based on this, both mitochondria and peroxisomes cooperate in ROS and 968 energy metabolism. 969 970 EXTENDED DATA LEGENDS 971 Figure 2-1. 8-oxo-dG expression and distribution in low PMIs brains. (A, A’) 8-oxo-dG 972 immunostaining and quantitative analyses of controls, AD and NDAN frontal cortices (n=4) with 973 low PMIs (2-5 hours) showing increased oxidative damage in AD as compared to control and no 974 significant differences between controls and NDAN. Original magnification 60X, scale bar 30 μm. 975 Statistical analyses were made using 1way ANOVA (F(2, 9)=15.71, p=0.0012), following Tukey’s 976 multiple comparisons test. Values are expressed as mean ± SD. ** p<0.01. (B) A Pearson’s 977 correlation test was performed for each measurement against the PMI. Correlation coefficient (r) 978 and p values are noted in the individual plots showing no significant correlation with PMI values. 979 Figure 5-1. SOD2 protein levels in synaptosomes. Wb analyses performed on synaptosomal 980 fractions showing significantly decreased levels of SOD2 in AD patients as compared to NDAN 981 and control subjects. The analyses conducted either on protein extracts from single individual 982 synaptosomal fraction (A) (ctrl vs. AD p=0.0367; ctrl vs. NDAN p=0.9987; AD vs. NDAN 983 p=0.0404) or on pooled extracts (B) (ctrl vs. AD p= 0.0011; ctrl vs. NDAN p=0.9988; AD vs. 984 NDAN p=0.0012) confirmed the preserved antioxidant content in NDAN individuals. Statistical 985 analyses were made using 1way ANOVA (A: F (2, 18) = 4.799, p=0.0214; B: F (2, 9) = 19.21, 986 p=0.0006 ) following Tukey’s test multiple comparisons test. Values are expressed as fold change 987 as function of SYN ± SD (n=7/group; 3 technical replicates). ± SD. * p< 0.05; ** p< 0.01. 988 41 SD (n=7/group; 3 technical replicates). * p< 0.05. 997 Figure 7-1. PPARα expression in cortical total protein extracts. Wb analyses performed on total 998 protein extracts showing significant increase of PPARα in AD cortical samples as compared to 999 controls and NDAN. The analyses conducted either on total protein extracts from single individuals 1000 (A) (ctrl vs. AD p=0.0121; ctrl vs. NDAN p=0.1106; AD vs. NDAN p=0.5266) or on pooled total 1001 protein extracts (B) (ctrl vs. AD p=0.0223 ctrl vs. NDAN p=0.9999; AD vs. NDAN p=0.0227) 1002 indicate higher levels of PPARα in AD patients. Statistical analyses were made using 1way 1003 ANOVA (A: F (2, 18)=5.412 p=0.0144; B: F (2, 6)=9.317 p=0.0144) following Tukey’s test 1004 multiple comparisons test. Values are expressed as fold change as function of ACTB ± SD 1005 (n=7/group; 3 technical replicates) ± SD. * p< 0.05. 1006 Figure 8-1. CAT expression levels in cytosolic fraction. Wb analyses performed on cytosolic 1007 fraction showing significant increase of CAT in AD frontal cotices vs. controls and NDAN. The 1008 analyses conducted either on cytosolic protein extracts from single individuals (A) (ctrl vs. AD 1009 p=0.0297; ctrl vs. NDAN p=0.5300; AD vs. NDAN p=0.2265) or on pooled cytosolic protein 1010 extracts (B) (ctrl vs. AD p=0.0089 ctrl vs. NDAN p=0.1711; AD vs. NDAN p=0.1867) indicate 1011 higher levels of CAT in AD patients. Statistical analyses were made using 1way ANOVA (A: F (2, 1012 18)=4.012 p=0.0362; B: F (2, 9)=7.659 p=0.0114) following Tukey’s test multiple comparisons test. 1013 42 Values are expressed as fold change as function of ACTB ± SD (n=7/group; 4 technical replicates). 1014 * p< 0.05; ** p<0.01. 1015 1 TABLES
Article Images (0)
×![]()
Bioz is the world’s most comprehensive AI search engine for scientific experimentation. The Bioz search engine offers researchers billions of data-driven product, technique, and protocol recommendations. Bioz taps into the latest advances in Artificial Intelligence (AI) – including Natural Language Processing (NLP), Machine Learning (ML) and Generative AI to mine and structure hundreds of millions of pages of complex and unstructured scientific papers. By harnessing the power of AI, Bioz provides researchers with an unprecedented amount of summarized scientific experimentation knowledge right at their fingertips, ultimately helping to speed up drug discovery and the rate of success in finding cures for diseases. Bioz is used by over 15 Million researchers from over 17,000 different universities and companies in 196 countries.
