Bovine serum albumin fraction V (BSA), rabbit anti-human Haptoglobin IgG, amino acids, dopamine, acetylcholine, tyramine, γ-aminobutyric acid (GABA), DL-lysine-4,4,5,5-d4 dihydrochloride (d4-lysine), formic acid, salts and buffers were purchased from Sigma-Aldrich (St. Louis, MO, USA). Nε-carboxymethyllysine (CML) and Nε-carboxyethyllysine (CEL) were obtained from Iris Biotech (Marktredwitz, Germany). Acetonitrile, water and ammonium formate of mass spectrometry grade were obtained from Merck (Darmstadt, Germany).
The dye reagent for protein titration was from Bio-Rad (Hercules, CA); polyvinylidene difluoride (PVDF) and nitrocellulose membranes were from GE Healthcare (Milan, Italy). Horseradish peroxidase (HRP)-conjugated secondary antibodies (goat anti-rabbit, GAR-HRP; goat anti-mouse, GAM-HRP) were from Immunoreagents (Raleigh, NC, USA). Fuji Super RX 100 film was from Laboratorio Elettronico Di Precisione (Naples, Italy).
All experiments were carried out with male Wistar rats (30 days old, P30) purchased from Charles River; Calco, Como, Italy). Housing, treatments and euthanasia of animals were performed as previously published . The rats were divided into two groups, one fed a fructose rich diet (F group), the other one fed a control diet (C group) for 3 weeks. At the end of treatment, half of the rats from each group was euthanized, while the other half received a control diet (FR and CR groups) for further 3 weeks (P72). The composition of both control and fructose rich diet is reported in Supplementary Table 1.
The animals were then euthanized, and frontal cortex was harvested and dissected as previously described . Mitochondrial oxygen consumption was immediately assessed in little aliquots of tissue. Pieces of each sample were fixed for immunofluorescence analysis, while the remaining samples were snap frozen in liquid nitrogen and stored at − 80°C for further analyses.
Frontal cortex samples were homogenized (1:1000, w/v) in Mir05 medium containing 110 mM sucrose, 60 mM K-lactobionate, 20 mM Hepes, 20 mM taurine, 10 mM KH2PO4, 6 mM MgCl2, 0.5 mM EGTA, and 0.1% w/v fatty acid-free BSA, pH 7.0.
Homogenates (2mg) were transferred into calibrated Oxygraph-2k (O2k, Oroboros Intruments, Innsbruck, Austria) 2-mL chambers. Oxygen polarography was performed at 37 ± 0.001°C (electronic Peltier regulation), and oxygen concentration (µM) and oxygen flux (pmol O2 s−1 mL− 1) were real-time recorded and corrected automatically for instrumental background by DatLab software (Oroboros Intruments, Innsbruck, Austria).
After addition of the homogenates, the O2 flux was allowed to stabilize. A substrate, uncoupler, inhibitor titration (SUIT) protocol was applied to assess qualitative and quantitative mitochondrial changes . After stabilization, leak respiration supported primarily by electron flow through complex I of the respiratory chain was evaluated by adding the substrates malate (0.5 mM), pyruvate (5 mM), and glutamate (10 mM). Electron transfer was coupled to phosphorylation by the addition of 2.5 mM ADP, assessing phosphorylating respiration with electron transfer supported by complex I. Succinate (10 mM) was added to the chamber to induce maximal phosphorylating respiration with parallel electron input from complexes I and II. Oligomycin (2.5 mM) was added to assess leak respiration when substrates and ADP were provided, but ATP synthase is inhibited. Maximum capacity of the electron transport chain was obtained by addition of the uncoupler carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP, 0.5 mM). Rotenone (0.5 µM) was added to inhibit complex I; hence, the maximal capacity supported by complex II alone was determined. Residual oxygen consumption was established by addition of the inhibitor antimycin A (2.5 mM) and the resulting value was subtracted from the fluxes in each run, to correct for non-mitochondrial respiration. All samples were run in duplicates and the mean was used for analysis.
Procedures to test mitochondrial integrity were routinely carried out at the beginning of each measurement, by evaluating the stimulating effect of 10 mM exogenous cytochrome c on mitochondrial respiration in the presence of complex I- linked substrates and ADP.
Metabolic parameters assay
The amount of fructose and uric acid in frontal cortex samples was measured by colorimetric enzymatic methods, using commercial kits according to the manufacturer’s instruction (Sigma Aldrich, St. Louis, MO, USA for fructose, and GS Diagnostics SRL, Guidonia Montecelio, Rome, Italy for uric acid).
Aliquots of frontal cortex (about 50 mg) were homogenized in seven volumes (w/v) of cold RIPA buffer, as previously published , and protein concentration was titrated by the colorimetric Bio-Rad Bradford protein assay, using a commercial kit (Bio-Rad, Hercules, CA, USA), according to the manufacturer’s instruction. Then protein extracts were used for titrating the markers reported below by ELISA or Western blotting.
Analysis of tumor necrosis factor alpha (TNF-alpha)
TNF-alpha concentration was evaluated by sandwich ELISA in frontal cortex homogenates diluted 1:20  using the TNF-alpha Duo-Set kit (R&D, DBA Italia). Data were reported as pg of TNF-alpha per mg of total proteins.
Aliquots (30 µg) of cortex proteins were resolved by electrophoresis, under denaturing and reducing conditions , on 12.5% (to quantify glucose transporter-5, Glut-5; glial fibrillary acidic protein, GFAP; synaptophysin; brain derived neurotrophic factor, BDNF; Nerve growth factor, NGF; extracellular signal-regulated kinase, Erk1/2) or 10% (post-synaptic density protein 95, PSD-95; synaptotagmin I; Peroxisome proliferator-activated receptor gamma coactivator 1-alpha, PGC-1α; Tropomyosin receptor kinase B, TrkB; Tropomyosin receptor kinase A, TrkA; Thyroxine hydroxylase, TH) polyacrylamide gels. Proteins were then blotted onto PVDF or nitrocellulose membrane , and the following washing and blocking steps were performed according to . Immunodetection was carried out as follows:
Glut-5: rabbit anti Glut-5 IgG (Invitrogen, Carlsbad, CA, USA, 0.5 µg/mL in T-TBS containing 3% w/v BSA; overnight, 4°C), followed by goat anti-rabbit horseradish peroxidase-conjugated IgG (GAR-HRP; 1: 45,000 dilution in T-TBS containing 2% w/v BSA; 1 h, 37°C); GFAP: rabbit anti-human GFAP IgG (Cell Signaling, MA, USA; 1: 1,000 in T-TBS containing 1% v/v non-fat milk; overnight, 4°C), followed by GAR-HRP IgG (1:150,000 in 5% v/v non-fat milk; 1 h, 37°C); Synaptophysin: rabbit anti-Synaptophysin IgG (Merk Millipore, Milan, Italy; 1:150,000 in 3% w/v BSA; overnight, 4°C) followed by GAR-HRP IgG (1:40,000 dilution in 1% v/v non-fat milk; 1 h, 37°C); Synaptotagmin I: rabbit anti-Synaptotagmin I IgG (Cell Signaling; 1:1,000 dilution in 3% v/v non-fat milk; overnight, 4°C) followed by GAR-HRP IgG (1:200,000 dilution in 3% v/v non-fat milk; 1 h, 37°C); PSD-95: rabbit anti-PSD 95 IgG (Cell Signaling; 1:1,000 in 3% v/v non-fat milk; overnight, 4°C) followed by GAR-HRP IgG (1:70,000 dilution in 3% v/v non-fat milk; 1 h, 37°C); BDNF: rabbit anti-human BDNF IgG (Santa Cruz Biotechnology, CA, USA; 1:500 dilution in 0.25% v/v non-fat milk; overnight, 4°C), followed by GAR-HRP IgG (1:70,000 dilution in 1% v/v non-fat milk; 1 hour, 37°C); PGC-1α: rabbit anti-human PGC-1α IgG (Millipore, MA, USA; 1:2,000 dilution in 3% w/v BSA; overnight, 4°C), followed by GAR-HRP IgG (1:20,000 dilution in 3% w/v BSA; 1 h, 37°C); TrkB: rabbit anti-TrkB (Santa Cruz Biotechnology, CA, USA; 1:2000 dilution in 2% w/v BSA; overnight, 4°C) followed by GAR-HRP IgG (1: 140 000 dilution in 2% w/v BSA; 1 h, 37°C); pErk1/2: rabbit anti-pErk1/2 IgG (Cell Signaling; 1:1,000 dilution in 2% w/v BSA; overnight, 4°C) followed by GAR-HRP IgG (1:150,000 dilution in 2% w/v BSA; 1 h, 37°C); Erk 1/2: rabbit anti-Erk1/2 (Cell Signaling; 1:1,000 dilution in T-TBS containing 2% BSA; overnight, 4°C) followed by GAR-HRP IgG (1:170,000 dilution in 2% w/v BSA; 1 h, 37°C); TH: mouse anti-TH (Santa Cruz; 1:1,000 dilution in 5% v/v non-fat milk; overnight, 4°C) followed by goat anti-mouse (GAM-HRP) IgG (1:10,000 dilution in T-TBS; 1 h, room temperature); TrkA: rabbit anti-TrkA (Santa Cruz; 1:1,000 dilution in 5% v/v non-fat milk; overnight, 4°C) followed by GAR-HRP IgG (1:10,000 dilution in T-TBS; 1 h, room temperature); NGF mouse anti-NGF (Santa Cruz; 1:1,000 dilution in 5% v/v non-fat milk; overnight, 4°C) followed by GAM-HRP IgG (1:10,000 dilution in T-TBS; 1 h, room temperature).
For loading control, β-actin or vinculin was revealed after detection of each marker. To this aim, the membranes were stripped , and then treated with mouse anti-β-actin IgG (1:1,000 in 0.25% v/v non-fat milk; overnight, 4°C) followed by GAM-HRP IgG (1:30,000 in 0.25% v/v non-fat milk; 1 h, 37°C) or with mouse anti-vinculin IgG (Sigma Aldrich, Milan, Italy; (1:40,000 dilution in 5% v/v non-fat milk; overnight, 4° C) followed by GAM-HRP IgG (1:10,000 dilution in T-TBS; 1 h, room temperature).
The Excellent Chemiluminescent detection Kit (Cyanagen s.r.l., Bologna, Italy) was used for detection. Chemidoc or digital images of X-ray films exposed to immunostained membranes were used for densitometric analysis and quantification was carried out by Un-Scan-It gel software (Silk Scientific, UT, USA).
Haptoglobin (Hpt) evaluation
Hpt concentration in frontal cortex samples was measured by ELISA as previously reported . Samples were diluted (1: 3,000, 1:10,000, 1:30,000) with coating buffer (7 mM Na2CO3, 17 mM NaHCO3, 1.5 mM NaN3, pH 9.6), and aliquots (50 µl) were then incubated in the wells of a microtiter plate (Immuno MaxiSorp; overnight, 4°C). Washing and blocking were carried out as previously reported , then the wells were incubated (1 h, 37°C) with 50 µl of rabbit anti-human haptoglobin, (1:500 in 130 mM NaCl, 20 mM Tris-HCl, 0.05% Tween, pH 7.4, containing 0.25% BSA), followed by 60 µl of GAR-HRP IgG (1:5000 dilution, 1 h, 37°C). Peroxidase-catalyzed color development from o-phenylenediamine was measured at 492 nm.
Evaluation of nitro-tyrosine levels, acetylcholinesterase (AChE) and monoamine oxidase (MAO) activities
Nitro-Tyrosine (N-Tyr) titration was carried out by ELISA in frontal cortex homogenates as previously described . Samples were diluted (1:1,500, 1:3,000, 1:6,000) with coating buffer, and aliquots (50 µl) were then incubated in the wells of a microtiter plate (overnight, 4°C). After washing and blocking, the wells were incubated (1 h, 37°C) with 50 µl of rabbit anti-N-Tyr IgG (Covalab, distributed by VinciBiochem, Vinci, Italy; 1: 1000 dilution in T-TBS containing 0.25% w/v BSA) followed by 60 µl of GAR-HRP (1:9,000 dilution; 1 h, 37°C). Peroxidase-catalyzed color development from o-phenylenediamine was measured at 492 nm. Data were reported as OD per mg of total proteins.
The AChE activity was measured in frontal cortex samples as previously described . Enzyme activity was expressed as nmol/min mg protein.
The monoamine oxidase (MAO) activity was measured spectrophotometrically following the conversion of benzylamine to benzaldehyde, as previously described .
Paraffin embedded sections of frontal cortex from all the groups were stained with the cAMP response element-binding protein (p-CREB) specific monoclonal antibody (p-CREB antibody, Cell Signaling, 1:1,000 in dilution in PBS containing 2% w/v BSA; overnight, 4°C), and DAPI (Sigma Aldrich, Saint Louis, MO, USA). For the analysis, images were acquired with x40 magnification and 3 random field/section per rat were analyzed using ImageJ (National Institutes of Health, Bethesda, MD, USA). Images were captured and visualized using a Nikon Eclipse E1000 microscope.
Liquid chromatography-electrospray-high-resolution tandem mass spectrometry (LC-MS/MS)
Amino acids and their derivatives were analyzed by liquid chromatography-electrospray-high-resolution tandem mass spectrometry (LC-MS/MS) as previously reported , with minor modifications. Frontal cortex samples (25 ± 10 mg) were dissolved in 0.390 mL of 0.1% formic acid along with 10 µL of 10 µg/mL lysine d-4; suspensions were accurately homogenized by using a stainless-steel disperser (IKA T10, Staufen, Germany, 3 passes, 30 s) in an ice bath. Supernatants (0.1 mL) were further purified by using 0.3 mL of 0.1% formic acid in acetonitrile by a protein precipitation and phospholipids removal cartridge (Phree, 1 mL, Phenomenex, Torrance, CA); eluates were collected and dried by using a centrifugal evaporator (SpeedVac, Thermo Fisher Scientific, Bremen, Germany). Dried samples were dissolved in 0.1 mL of a mixture acetonitrile: water: formic acid (50:49.9:0.1, v/v/v) and 5 µL injected into the LC-MS/MS system consisting in a linear ion trap with Orbitrap detector (LTQ Orbitrap XL) interfaced to an Ultimate 3000 RS (Thermo Fisher Scientific, Bremen, Germany). Analytes were separated through hydrophilic interaction chromatography in data dependent scan positive ions mode for identification and quantitation of amino acids, catecholamines, indole derivatives and polar markers of oxidation.
Chromatographic separation was achieved through a silica sulfobetaine zwitterionic modified HILIC column (100 x 2.1 mm, 1.7 µm, Syncronis HILIC, Thermo Fisher, Bremen, Germany) at 35°C. Mobile phases consisted in 0.1% formic acid in acetonitrile:water 95:5 (v/v, solvent A) and 0.1% formic acid in water (solvent B) both with 5 mM ammonium formate. Samples (4°C) were separated through the following gradient of solvent B (minutes/%B): (0/3), (3.5/3), (15.5/75), (17.5/75) at a flow rate of 0.25 mL/min. Electrospray interface (ESI) parameters were the following: spray voltage 5.0 kV, capillary voltage 21.0 V, capillary temperature 300°C, sheath gas flow and auxiliary gas flow were 25 and 4 arbitrary units, respectively. Profile data type were acquired in full scan FTMS mode (Fourier transformed) in the mass range 75–750 m/z. For data dependent scanning mode, MS/MS normalized collision energy was set to 20, activation Q 0.25, activation time 25 ms, with a 1 m/z isolation window, while a reject mass list was generated by injecting blank samples consisting in a mixture of acetonitrile:water:formic acid (50:49.9:0.1, v/v/v).
For compound identification, differential analysis and metabolic pathways, raw data were loaded in Compound Discoverer (v. 3.2, Thermo Fisher Scientific). The workflow included the identification of both expected and unknown metabolites; briefly, each node performed retention time alignment, expected compound detection, biotransformation, dealkylation and dearylation products formation. Resolution and isotope pattern matching with unknown compounds detection were used across all samples with a mass accuracy below 5 ppm. FISh (fragment ion searching) scoring was applied to all expected compounds with automatic fragment annotations based on targeted and untargeted compound chemical behavior outlined in Human Metabolome Database (https://hmdb.ca/). According to the background in blank samples, the procedure predicted elemental compositions for all unknown compounds, while quality control samples (QC, consisting in pooled samples spiked with amino acid standards) corrected signal intensities each 10 runs. Hierarchical clustering was obtained through filtering procedures based on technical replicates coefficient of variation (CV%), retention time and mass accuracy. The workflow included differential analysis (p-values, adjusted p-values, ratios, fold change), Euclidean distance and complete linkage method without data normalization to enhance differences among frontal cortex of groups fed different diets.
For targeted analyte quantitation, a calibration curve of the compounds listed in Supplementary Table 2 was built in the range 100–5000 ng/mL. Linearity and the responses of intraday and interday assays were monitored by using Xcalibur 2.1 with a mass accuracy fixed at 5 ppm (Thermo Fisher Scientific, Bremen).
Data were expressed as mean values ± SEM. The program GraphPad Prism 6 (GraphPad Software, San Diego, CA, USA) was used to verify that raw data had normal distribution and to perform one-way ANOVA followed by Bonferroni post-test. P < 0.05 was considered significant in the reported analyses.