A similar result was obtained with an OGT inhibitor (Figure 3c,d; = 7, number of independent experiments). in the plasma membrane by using immunocytochemistry and discontinuous sucrose gradients method. By using the biotinylation method, we also found that APP preferentially underwent endocytosis from lipid rafts and Pitolisant that the amount of internalized APP from lipid rafts was specifically reduced by O-GlcNAcylation. These results indicate that O-GlcNAcylation can regulate lipid raft-dependent APP endocytosis via translocation of APP into non-raft microdomains. Our findings showed a new functional role of O-GlcNAcylation for the regulation of APP trafficking, offering new mechanistic insight for A production. = 40, number of cells) in insulin-treated cells compared to control cells, indicating that insulin decreased APP localization in lipid rafts. As another marker for lipid rafts, we used cholera toxin B (CTB), which binds to a component of lipid rafts called ganglioside GM1 [62]. Insulin significantly decreased the colocalization of APP and CTB (Figure S2), which was consistent with the result in Figure 1. Open in Pitolisant a separate window Figure 1 Insulin reduces APP Pitolisant localization in lipid raft microdomains. SH-SY5Y-APP/BACE1 cells were incubated with 1 M insulin for 2 h. Akt inhibitor (Akti, 10 M) or OGT inhibitor (OSMI-1, 50 M) was also used. After washing, cells were incubated with 6E10 antibody at 4 C to label APP at the plasma membrane. Cells were then fixed, permeabilized, and incubated with caveolin antibody to label lipid rafts. Fluorescent conjugated secondary antibodies were used. (a) Typical immunoreactivities of APP (green) and caveolin (red) are shown. (b) Coefficients of APP and caveolin were quantified as arbitrary units using Zen software (= 40, number of cells were randomly selected from four independent experiments). The threshold Pitolisant was automatically set as the 60100% value. (Manders M1 coefficient: control, 0.4025 0.01967; insulin, 0.30248 0.01868; Akti, 0.49505 0.01759; OSMI-1, 0.55038 0.02195; fraction of A overlapping B, image A: APP, image B: caveolin). Scale bar is 10 m. One-way ANOVA: ***, 0.001. All values represent the mean SEM. Effects of insulin on APP O-GlcNAcylation and APP endocytosis are mediated by Akt signaling and OGT [29]. Thus, we tested the effect of insulin on APP localization in lipid rafts in the presence of Akt inhibitor (Akti) or OGT inhibitor (OSMI-1). Inhibitory effects of insulin on colocalization of APP and caveolin were completely blocked in the presence of Akti or OSMI-1, indicating that the insulin effect was mediated by Akt and OGT signaling (Figure 1a,b). It was noteworthy that levels of colocalization were higher in the presence of these inhibitors than in control cells, indicating basal activities of Akt or OGT in control conditions as we observed previously [29]. The immunocytochemical method we used to study the localization of APP in lipid rafts may generate ambiguous results due to the insufficient resolution for LIFR the small size of lipid raft microdomains. To confirm the colocalization results by using the biochemical method, lipid raft fractions were isolated in detergent-free conditions as described previously [58]. For this purpose, the same amount of proteins from cell lysates were loaded on discontinuous sucrose gradients. After obtaining 12 fractions, equal volumes of each fraction Pitolisant were loaded for Western blot. A typical result for APP distribution is shown in Figure S3. The majority of APP was localized in non-raft fractions. However, a small but significant amount of APP was localized in lipid raft fractions, which was consistent with the previous report [58]. When levels of cholesterol and protein from each fraction were measured, cholesterol-high lipid raft fractions (fractions 4 to 6 6) were clearly separated from protein-high non-raft fractions (fractions 8 to 12). Moreover, caveolin was enriched in fractions 4 to 6 6. Protein levels in each fraction did not differ between control and insulin-treated cells. Total APP level was not changed by insulin, as observed previously [29]. For a better quantification of APP.