Neuronal cell culture
Rat neuronal primary cultures were prepared as described previously . Hippocampal or cortical neurons at 21-days in vitro (DIV) were treated for 24 h with 100 nM insulin, 10 ng/ml TNFα, or a combination of both to induce IR in vitro. Glial release of TNFα induces neuroinflammation . Therefore during the preparation of neuronal primary culture, the addition of Ara-c (cytosine β-D-arabinofluranoside), a suppressor of glial proliferation, was omitted. Before application, half of the conditioned Neurobasal Medium (NB, Gibco, Gaithersburg, MD) was replaced with fresh serum-free NB medium. To assess responsiveness to insulin, cell were washed with conditioned NB medium for 1 h and either 100 nM insulin or vehicle was applied for 15 min.
TBA2.1 mice and MetS
Mice were housed in groups of up to 4 in individually ventilated cages (IVCs, Green line system, Tecniplast, Lugano, Switzerland) under controlled environmental conditions (22 ± 2 °C, 55% ± 10% humidity, 12 h light/dark cycle, with lights on at 06:00). Animals had free access to food and water. Male wild-type (WT, +/+) and heterozygous (+ /Tg) TBA2.1 mice  were fed with normal chow until the age of 8–9 weeks. After this period, they were fed with either a regular diet (RD) or a high-fat/high-calorie chow (High-fat diet [HFD], Ssniff #E15126-34, Soest, Germany). Mice fed with HFD were selected for further experiments when they reached a 70% body weight increase, which took on average 4–5 months. Mice used in the experiments were 6–8 months old.
Aβ3(pE)-42 oligomer and drug treatment
Aβ3(pE)-42 oligomers (Anaspec, #AS‐20,276; Fremont, CA) were prepared according to a previously established protocol . The lyophilized peptide was dissolved in 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP) to 0.5 mg/ml and the aliquots were stored at − 80 °C. HFIP was evaporated for 24 h at room temperature. The peptide was dissolved in 0.1 M NaOH, diluted in NB medium buffered with 0.1 M HCl, and incubated for 24 h. The oligomers were added directly to cultures at a final concentration of 500 nM . The neural precursor cell-expressed developmentally down-regulated gene 8 (NEDD8) inhibitor MLN-4924 (#A-1139, Active Biochem, Hong Kong, China) was directly applied to the culture medium at a final concentration of 1 μM . MG-132 (#M-1157, AG Scientific, San Diego, CA) was used in cell culture experiments at a final concentration of 20 µM.
Expression constructs and Adeno-associated virus 9 (AAV9) infection
Expression constructs are listed in Additional file 1: Table S1. Cortical neurons cultured in T-75 flasks (#156499, Thermo Scientific, Freiburg im Breisgau, Germany) were infected at DIV10 with 10,000 vg/cell of AAV9 expressing either HA-NEDD8 or the HA tag alone for 14 days. Thereafter, neuronal cell lysates were used for tag-specific immunoprecipitation with anti-HA microbeads (µMACS™ HA Isolation Kit, #130-091-122, Miltenyi Biotec, Bergisch Gladbach, Germany).
Glucose tolerance test
A small drop of blood (< 5 μl) was taken from the tail vein of the animals and placed on the test strip of the blood glucose meter Contour Next (Bayer, Leverkusen, Germany). Glucose in the blood measured after 6-h fastening was considered as the basal level (t = 0). Immediately after the first measurement, D(+)glucose solution of 0.25 g/ml was intraperitoneally injected into the animals at a final concentration of 1 g/kg. Blood glucose levels were measured at 15, 30, 60 and 120 min (t = 15, t = 30, t = 60 and t = 120) postinjection.
Enzyme-linked immunosorbent assays (ELISA)
Animals were anesthetized with ketamine (100 mg/kg) and xylazine (10 mg/kg). Blood was collected directly from the heart ventricle. Following blood coagulation at room temperature, samples were centrifuged for 10 min at 1500 g and the blood serum was collected. Serum insulin levels were measured using a mouse insulin ELISA kit (#EZRMI-13 K; Merck Millipore, Darmstadt, Germany). To determine the TNFα level in brain tissue, frozen cortices from WT and heterozygous TBA2.1 mice fed with either RD or HFD were homogenized in ice-cold PBS supplemented with protease inhibitor cocktail. After two freeze–thaw cycles, the homogenates were centrifuged for 5 min at 5000 g at 4 °C. The supernatant was assayed immediately after isolation using a mouse TNFα ELISA kit (#MHSTA50; R&D Systems, Wiesbaden, Germany) according to the manufacturer’s instruction.
Novel location recognition and novel object recognition experiments were performed in a square arena (50 × 50 × 50 cm3) under mild light conditions as described previously . The task consisted of 4 sessions: habituation, training, novel location recognition and novel object recognition. On the first day, the animals were habituated to the empty arena for 20 min. The training session took place 24 h later where the mice were left free to explore for 20 min a pair of similar objects (made of plastic mounting bricks), positioned in the arena. Twenty-four hours later, one of the identical objects was moved to a new position, and mice were left for 20 min in the arena for exploration. After the last 24-h interval, a novel object recognition test was performed, in which a familiar object was replaced by a novel one and free exploration was observed for 20 min. In studies of the effect of MLN-4924 in HFD-fed mice, the exploration time in all sessions was set to 10 min. All four sessions were video-recorded, and behavior was analyzed offline using ANY-maze software (ANY-MazeTM Video Tracking System, version 4.50/4.99, Stoelting Co., Wood Dale, IL). Exploration was considered only when the animal touched or reached the objects with nose at a distance of less than 2 cm. The time mice spent exploring the objects was recorded, and the discrimination index was calculated, taking into account the difference of time spent exploring the new and the familiar object locations ((Tnew − Tfamiliar)/(Tnew + Tfamiliar) × 100%). Chambers and objects were cleaned with 10% ethanol before and after each animal was tested.
MLN-4924 administration in mice
MLN-4924 was solubilized in DMSO at a concentration of 25 mg/ml. To increase the solubility of MLN-4924 in aqueous buffer, the solubilized drug was mixed with a solution of 10% non-toxic hydrophilic solubilizer 2-hydroxypropyl-β-cyclodestrin (#12446-35-5; Sigma-Aldrich Chemie, Hamburg, Germany) to a final concentration of 0.5 mg/ml and sterile filtered. A dose of 2 mg/kg of MLN-4924 or vehicle was injected intraperitoneally once a day for 14 days.
Acute hippocampal slice preparation and electrophysiology
Hippocampal slices from male WT (+/+) and heterozygous TBA2.1 (Tg) mice were prepared according to previously described protocols [28, 29]. Field excitatory postsynaptic potentials (fEPSPs) were recorded with glass capillary microelectrodes (3–5 MΩ) filled with artificial cerebrospinal fluid (aCSF), amplified by an Extracellular Amplifier (EXT-02B, NPI electronic, Germany) and digitized at a sampling frequency of 5 kHz by Digidata 1401plus AD/DA converter (CED, England). Stimulation strength was adjusted to 40%–50% (long-term depression, LTD) of the maximum fEPSP-slope values. A single stimulus with 0.1 ms width was applied every 30 s (at 0.0333 Hz) and values were averaged every 3 min. Following 30-min stable baseline recording, a low-frequency stimulus (900 stimuli at 1 Hz frequency) was applied for induction of LTD. For bath application of MLN-4924, the drug was solubilized in DMSO and diluted in aCSF at a final concentration of 1 μM. Following 10-min baseline recording, MLN-4924 or DMSO was applied 20 min before LTD induction and kept for the entire duration of recording.
Immunoprecipitation of phosphatidylinositide 3-kinase γ (PI3Kγ) and PI3Kγ ELISA
PI3Kγ was immunoprecipitated either from cultured primary cortical neurons or hippocampi of TBA2.1 mice. Cultured cells were collected in 20 mM Tris–HCl, pH 7.4, 137 mM NaCl, 1 mM CaCl2 1 mM MgCl2, and 1 mM Na3VO4, and pelleted at 500 g. Dissected mice hippocampi were snap-frozen in liquid nitrogen, mechanically homogenized using a pellet pestle motor (Kimble Kontes, Vineland, NJ) and pelleted at 500 g. Both types of samples were lysed for 1 h in lysis buffer (20 mM Tris–HCl, pH 7.4, 137 mM NaCl, 1 mM CaCl2 1 mM MgCl2, 1 mM Na3VO4, 1% NP-40, 1 mM PMSF) and then centrifuged at 14,000 g to sediment the insoluble material. The supernatant was incubated with 2 µg of rabbit polyclonal anti-PI3Kγ antibody (#5405; Cell Signaling Technology, Frankfurt, Germany) with gentle rotation for 1 h at 4 °C. Secondary antibody-bound dynabeads protein G (#10003D; Thermo Fisher Scientific, Schwerte, Germany) was incubated with the lysate containing the primary antibody-bound PI3Kγ for 1 h at 4 °C. The beads were collected with the use of DYNAL magnet (Invitrogen, Schwerte, Germany) and subsequently washed with the lysis buffer followed by wash buffer (0.1 M Tris–HCl, pH 7.4, 5 mM LiCl, 1 mM Na3VO4), TNE buffer (0.1 M Tris–HCl, pH 7.4, 5 mM LiCl, 1 mM Na3VO4) and an ELISA-reaction buffer (20 mM Tris–HCl, 4 mM MgCl2, 10 mM NaCl, ATP 150 μM). While still bound to the beads, PI3Kγ reaction was set up by adding the reaction buffer supplemented with 10 μM of PI(3,4)P2 and incubation at 37 °C for 2 h. In the negative control samples, 10 nM of AS-605240 was added to inhibit PI3Kγ activity. The kinase reaction was stopped using the kinase stop solution (reaction buffer supplemented with 4 mM EDTA). Reaction buffer containing the product of the enzymatic reaction phosphatidylinositol (3,4,5)-triphosphate (PI(3,4,5)P3) was collected and used for a competitive ELISA (PI3-Kinase Activity ELISA: Pico, #K-1000S, Echelon Biosciences, Göttingen, Germany) according to the manufacturer’s instructions. Thereafter, PI3Kγ was extracted from the beads using 2× SDS buffer and used for immunoblotting.
Synaptosomes were prepared from mouse cortices or primary rat cortical culture (21 DIV) according to previously published protocols [31, 32]. The cortex from mice was used in toto. To assess synaptic responsiveness to insulin, isolated synaptosomes from +/+ and +/Tg TBA2.1 mice on RD or HFD were stimulated with 100 nM of insulin or vehicle for 10 min at 37 °C. The reaction was performed in HEPES-buffered Krebs-like buffer (HBK, containing 308 mM NaCl, 308 mM KCl, 154 mM MgSO4, 1 M CaCl2, 100 mM Na2HPO4, 87 mM HEPES/Tris, pH 7.4, 0.48 g D(+)-glucose) supplied with 8 mM of ATP (#A2383; Sigma-Aldrich Chemie, Hamburg, Germany). Synaptosomes were centrifuged at 10,000 g, washed twice with fresh buffer and finally lysed in 2× SDS buffer.
Neurons were fixed in 4% paraformaldehyde (PFA) at room temperature. After permeabilization with 0.1% Triton-X/PBS, cells were blocked in blocking buffer (2% glycine, 2% BSA, 0.2% gelatin and 50 mM NH4Cl) and subsequently probed with primary and secondary antibodies (Additional file 1: Table S1) as described previously . Coverslips were mounted with Mowiol (# 17951500; Polysciences Inc., Hirschberg an der Bergstraße, Germany).
Mice were anaesthetized with isofluran (Baxter Deutschland GmbH, Unterschleißheim, Germany) and then perfused with 0.9% NaCl followed by fixation with 4% PFA in PBS. Brains were processed as described previously . Briefly, slices were incubated in immunohistochemistry blocking buffer (10% goat serum, 0.3% Triton-X 100 in PBS) and subsequently with primary antibody (Additional file 1: Table S1) diluted in blocking buffer. Hippocampi of both hemispheres from mouse brain sections were imaged on a SP5 CLSM system (Leica Mycroystem, Mannheim, Germany). Maximum projections of the scans, comprised of 5 z-stacks (0.39 μm z-step), were created using ImageJ software (NIH, Bethesda, MD) for subsequent analysis. The region of interest (ROI) was defined by positioning a rectangular frame of 220 × 300 μm2 onto the pyramidal cell layer of CA1 adjacent to the fasciola cinereum at low-power magnification (20×). Three to four brain sections from each animal were used for quantification and the number of positive cells for each marker (Iba1, GFAP, NeuN) was counted.
Fluorescence microscopy and image analysis
Image analysis was carried out using a Zeiss Axio Imager A2 fluorescent microscope (Zeiss, Jena, Germany) with Cool Snap EZ camera (Visitron System, Puchheim, Germany) and MetaMorph Imaging software (MDS Analytical Technologies, Ismaning, Germany). Up to 3 coverslips were treated individually and processed per group. For each coverslip, the same exposure time and intensity were taken among the different groups. After background subtraction, the fluorescence intensity of the immunosignal was measured along dendrites right after the first branching point using ImageJ software. The synaptic immunofluorescence intensities of pan-AKT and phospho-AKT were assessed in a region of 400 nm × 400 nm square set by the mask generated based on synaptic marker Shank3. The Shank3 mask was created semi-automatically using OpenView software .
Confocal laser scan microscopy
Images were acquired using Leica SP8 TCS STED 3X confocal microscopy, equipped with a pulsed White Laser (WLL) and a diode 405 nm laser. To quantify changes of pAKT and AKT in the spines, dendrites were scanned sequentially with detection of Alexa fluorophore (AF) 488 for MAP2 or GFP, AF 568 and AF 633 for Shank3. The optical sections were acquired along Z-axis with 0.27-μm Z resolution. After background subtraction, the fluorescent intensity was measured within ROI defined by Shank3 mask using ImageJ software.
Heterologous co-immunoprecipitation, pull-down assays and immunoblotting
Human embryonic kidney-293-T (HEK293T) cells were transfected using published protocols  and heterologous co-immunoprecipitation was performed as described previously . For insulin receptor substrate 1 (IRS1) ubiquitination analysis, cells were preincubated overnight with MLN-4924 (1 μM) or vehicle. They were then harvested in ice-cold PBS supplemented with protease inhibitor cocktail. For pull-down assay, cells were pelleted and lysed in RIPA buffer (50 mM Tris–HCl, pH 8.0, 150 mM NaCl, 1% NP40, 0.5% Na-deoxycholate, 0.1% SDS) containing protease inhibitor cocktail and the de-ubiquitination inhibitor PR-619 (#662141; Sigma-Aldrich, Hamburg, Germany). Following sonication and centrifugation for 10 min at 14,000 g, the supernatant was incubated with 20 μl of Chitin resin (#S6651L; New England Biolabs, Frankfurt am Main, Germany) while rotating for 2 h at 4 °C. The lysate was then transferred to BioSpin Chromotographic columns (#732-6204; Bio-Rad, Feldkirchen, Germany) and washed with RIPA buffer (50 mM Tris–HCl, pH 8.0, 500 mM NaCl, 1% NP40, 0.5% Na-deoxycholate, 0.1% SDS). Protein complexes were eluted with 2× SDS buffer. Both 20 μl of input and the eluate were loaded onto SDS-PAGE gradient gels as described previously . Images were acquired using Intas ECL Chemocam Imager (Li-cor, Bad Homburg vor der Höhe, Germany). Protein band intensities were measured using Gel Analyzer plugin ImageJ software (NIH, Bethesda, USA) and normalized to the loading control.
Statistical analysis was carried out with GraphPad Prism software (GraphPad software Inc., San Diego, CA). Data are presented as mean ± SEM. Student’s t-test (two experimental groups) or one-way ANOVA followed by Tukey’s multiple comparison test was used for comparison as denoted in figure legends. For biochemical and behavioral analysis, a two-way ANOVA was employed followed by Bonferroni's multiple comparison test. For LTD experiments, averages from 180 to 210 min after LTD induction were compared by two-way ANOVA followed by Bonferroni's multiple comparison test.