To evaluate Aβ phagocytosis, we used

a previously described ex vivo brain slice assay (Bard et al., 2000) (Figure 6A), where microglial cells are added to unfixed brain slices from aged, plaque-depositing APP transgenic CDK inhibitor mice. Microglia then surround and engulf Aβ (Figure 6B), and its removal can be measured immunohistochemically or by ELISA. While control BV2 cells efficiently cleared Aβ deposits from these brain sections, beclin 1-deficient cells were significantly less efficient at clearing the aggregates in both the hippocampus (Figures 6C and 6D) and cortex (Figure 6E). These histological findings were supported by independent measurements of Aβ content in brain slices by ELISA (Figures 6F and 6G). Moreover, primary microglia isolated from beclin 1+/− mice showed similar impairments in Aβ phagocytosis when compared to wild-type littermates ( Figures S4A and S4B). Importantly, Aβ phagocytosis by beclin 1-deficient BV2 cells could be rescued by recovering UMI-77 in vivo beclin 1 levels ( Figures S4C and S4D). To determine whether beclin 1 has a role in the removal of extracellular Aβ in vivo, we used beclin 1+/− mice and injected fibrillar Aβ into the frontal

cortex ( Figure 7A). In agreement with our ex vivo phagocytosis assay, the removal of Aβ was significantly impaired in beclin 1-deficient mice, resulting in twice as much Aβ remaining in the brains of beclin 1+/− mice compared with wild-type littermates ( Figures 7B and 7C). We then tested whether reduced phagocytic capacity could explain impairments in the removal of Aβ in vivo by injecting pH-sensitive beads, which fluoresce when internalized by microglial cells (insert of Figure 7D), into the frontal cortex of beclin 1+/− mice or wild-type littermates. Dichloromethane dehalogenase Quantification of fluorescence as an indicator of phagocytosed beads showed that beclin 1+/− mice phagocytosed almost 4-fold fewer beads than wild-type mice ( Figures 7D and 7E). Because the fluorescence readout of our

pH-sensitive beads could be confounded by changes in phagosomal or lysosomal pH, we tested whether reduced beclin 1 levels might affect phagosomal or lysosomal pH. To do this, we isolated primary microglia from beclin 1+/− mice and analyzed lysosomal pH in these cells using LysoSensor Yellow/Blue dextran ( Lee et al., 2010) ( Figure 7F). In addition, we analyzed phagosomal and lysosomal pH using FITC-conjugated beads ( Figure S5). Phagosomal and lysosomal pH in beclin 1-deficient cells were not significantly different than control cells ( Figure 7F; Figure S5), suggesting that the diminished fluorescent signal in beclin 1+/− mice injected with pH-sensitive beads is likely due to reduced uptake of the pH-sensitive beads. Together, these ex vivo and in vivo studies support our cell culture experiments and demonstrate that beclin 1 is necessary for efficient phagocytosis.