Tau regulates hippocampal response to insulin. (A) Hippocampal LTD induced by 1 µM insulin (30 min) in tau KO and WT mice. Each point represents mean ± SEM normalized to baseline values preceding the application of insulin (***, P < 0.001, two-way ANOVA). (B) Akt phosphorylation in hippocampal slices from tau KO mice and littermate controls, 10 min after 200 nM insulin treatment (*, P < 0.05, Student’s t test). (C) Akt phosphorylation in the hippocampus of tau KO mice and littermate controls, 1h after an icv injection of 2 µl at 5 mg/ml insulin (*, P < 0.05, Student’s t test). Tau KO mice were 7–12 mo old. (D) hTau expression and phosphorylation and (E) hTau detection in sarkosyl-soluble and -insoluble fractions from N1E115 cells overexpressing WT 1N4R human tau isoform. (F) Akt phosphorylation in N1E115 overexpressing WT hTau or transfected by empty vector, 10 min after insulin treatment at 200 nM (*, P < 0.05, Student’s t test). Controls are indicated as open circles/bars and tau KO/overexpressing cells as black circles/bars. Data in A–C show mean ± SEM from six (A), six to eight (B), and six or seven (C) mice per group from six (A), two (B), or three (C) independent experiments. Data in D and E show results from one independent experiment. Data in F show mean ± SEM from eight wells per group from five independent experiments. Ins, insulin. Molecular mass is indicated in kilodaltons.

Impact of tau deletion upon IR and IRS-1. (A) IRs from hippocampal slice extracts (400 µg protein) were precipitated on wheat-germ lectin (WGL) agarose beads, eluted with Laemmli sample buffer, and submitted to SDS-PAGE followed by Western blotting. Insulin-induced phosphorylation of the IR on the three tyrosines of the kinase domain was evaluated using anti–IRpY1158-1162-1163 antibody. The amount of IR present in each track was evaluated by reprobing the membrane with an anti-IRβ antibody. (B) Tyrosine phosphorylation of IRS-1 in hippocampal slices from tau KO mice and littermate controls 10 min after 200 nM insulin treatment (*, P < 0.05, Student’s t test). Total IRS-1 remains constant. (C) Tyrosine phosphorylation of IRS-1 in the hippocampus of KO tau mice and littermate controls 1 h after icv injection of 2 µl insulin at 5 mg/ml (*, P < 0.05, Student’s t test). Total IRS-1 remains constant. (D) Serine phosphorylation of IRS-1 in tau KO mice and littermate controls. (*, P < 0.05, Student's t test). (E) Coimmunoprecipitation experiments between hTau and different members of the insulin-signaling pathway in N1E115 cells transfected with WT 1N4R hTau. Tau interacted with PTEN only (). Ø indicates negative control of immunoprecipitation (IP; without primary antibody; only mouse anti-IgG beads); Ab indicates immunoprecipitation with the primary antibody and mouse anti-IgG beads. Unbound indicates the immunoprecipitation supernatant samples, whereas eluate indicates the immunoprecipitation complexes. Controls are indicated as open bars and tau KO animals as black bars. KO mice were 7–12 mo old. Data in A–D show mean ± SEM from four (A), three (B), three (C), and five (D) mice per group from two independent experiments. Coimmunoprecipitations were obtained from at least two independent experiments. Molecular mass is indicated in kilodaltons.

Tau interacts with PTEN and modulates its lipid phosphatase activity. (A and B) Coimmunoprecipitation of hTau (WT 1NR4) overexpressed in N1E115 cell (A) and mouse tau in the hippocampus of WT mice (B) with PTEN. Data are representative of two independent experiments from independent samples. (C) Bimolecular fluorescence complementation (BiFC) assay confirms the interaction between WT 1N4R tau and PTEN in the cytoplasm of HEK-293 cells. Data are representative of three independent experiments (***, P < 0.001 vs. VN-PTEN; one-way ANOVA, LSD Fisher’s post-hoc test). Bars, 30 μm. (D) In cell-free conditions, lipid phosphatase activity of PTEN is decreased in the presence of recombinant hTau protein (control, open circles; WT 1N4R, black circles; ***, P < 0.001, two-way ANOVA). Note that PTEN activity is not modulated in the presence of an irrelevant protein (albumin; not depicted). Data show mean ± SEM from six to eight assays per condition from three independent experiments. (E) Modulation of PtdIns(3,4,5)P3 production by hTau (WT 1N4R) using BRET. In each experiment, delta BRET was determined by removing mean basal BRET values from mean BRET values obtained for each experimental condition (20 BRET measurements per experimental condition). Inset: typical BRET experiment showing real-time insulin effect on PIP3 production is presented (open circle, control/vehicle; closed circles, hTau/vehicle; open square, control/insulin; closed square, hTau/insulin; *, P < 0.05; ***, P < 0.001 vs. control; #, P < 0.05 vs. insulin condition; one-way ANOVA, LSD Fisher’s post-hoc test). Data show mean ± SEM from five assays per condition from two independent experiments. Molecular mass is indicated in kilodaltons.

Tau deficiency inhibits an anorexigenic effect of brain insulin administration leading to metabolic disturbances. (A) Cumulative food intake for 24 h measured using metabolic cages in tau KO and littermate controls intracerebroventricularly injected first with vehicle and then 2 µl insulin (5 mg/ml; ***, P < 0.001 vs. WT-PBS; °°°, P < 0.001 vs. tau KO/insulin; two-way ANOVA). (B) Cumulated food intake in WT and tau KO mice 24 h after vehicle or insulin brain injection (same animals as in A; *, P < 0.05; **, P < 0.01 vs. WT-PBS; #, P < 0.05 vs. WT/insulin; °°°, P < 0.001 vs. tau KO/insulin; one-way ANOVA, LSD Fisher’s post-hoc test). (C) 48-h body weight variation after vehicle or insulin brain injection in tau KO and WT littermates (same animals as in A; °°°, P < 0.001 vs. tau KO/insulin; two-way ANOVA). Data in A–C show mean ± SEM from 5–10 mice per group from three independent experiments. (D) Mean food intake (*, P < 0.05, two-way ANOVA). (E) Ambulatory activity (*, P < 0.05, two-way ANOVA). (F) Body weight gain (***, P < 0.05, two-way ANOVA). (G) Plasma leptin levels (**, P < 0.01, Student’s t test). (H and I) Adipose tissue weight (*, P < 0.05; **, P < 0.01, Student’s t test). (J) Glycemia. (K) Insulinemia (*, P < 0.05, Student’s t test). (L) Intraperitoneal glucose tolerance test (***, P < 0.05, two-way ANOVA) in tau KO mice and littermate controls. Quantifications represent mean ± SEM. Controls are indicated as open circles/bars, tau KO as black circles/bars. Dashed lines/bars represent insulin-treated animals. Mice were 6–8 mo old at time of experiments and sacrifice. Metabolic data in C–L show mean ± SEM from 10–12 (D), 9–12 (E), 13–27 (F), 4–5 (G), 11–14 (H and I), 15–17 (J), 20–23 (K), and 12–14 (L) mice per group acquired from three independent experiments.