a, Rapamycin improves learning in PDAPP mice. While learning in both transgenic groups was impaired with respect to wild-type littermates' [**, P<0.001 for both comparisons, Bonferroni's post hoc test applied to a significant effect of genotype and treatment, F(3,120) = 29.46, P<0.0001, repeated measures two-way ANOVA], performance of rapamycin-fed PDAPP mice was improved with respect to the control-fed transgenic group only in the last day of training (^#P = 0.036 for the comparison of performance between transgenic groups, Student's t test), indicating improved learning of rapamycin-fed PDAPP mice at day 4. No significant interaction was observed between day number and genotype (P = 0.96), indicating that genotype had roughly the same effect at all times during training. Although no significant interaction was observed between day number and treatment for control-treated animals (P = 0.91), a significant interaction was observed between day number and treatment for rapamycin-treated groups. The effect of rapamycin treatment became more pronounced as training progressed, as indicated by the slopes for the learning curves (m = −5.14 for rapamycin-treated as compared to m = −3.58 for control-treated PDAPP transgenic mice; m = −4 for rapamycin-treated as compared to m = −2.95 for control-treated non-transgenic mice). A trend to improved learning was observed in rapamycin-treated non-Tg mice, but this difference was not significant. Overall learning was effective in all groups [F(3,120) = 10.29, P<0.0001, repeated measures two-way ANOVA]. Inset, learning was effective in all experimental groups during cued training. b, Rapamycin restores spatial memory in PDAPP mice. While retention in control-fed PDAPP mice was impaired with respect to all other groups, as previously described[11], [31], [35], [50] [P values are indicated, Tukey's multiple comparisons test applied to a significant effect of genotype (P<0.0001) in one-way ANOVA], memory in rapamycin-fed PDAPP mice was indistinguishable from that of control- or rapamycin-fed non-Tg groups. A trend to improved retention was observed in rapamycin-treated non-Tg mice, but this difference did not reach statistical significance. c and d, Rapamycin treatment does not affect non-cognitive components of behavior. c, Although transgenic groups spent more time engaged in thigmotactic swim, as described[31] (** P<0.001, Bonferroni's post hoc test applied to a significant effect of genotype [F(3,440) = 15.04, P<0.0001, two-way ANOVA], no significant difference in percent time spent in thigmotactic swim was observed between transgenic groups. d, No significant difference in floating was observed between groups. Data are mean ± SEM.

a, b and f, representative immunoblots of whole brain lysates from control- and rapamycin-treated PDAPP transgenic and non-transgenic littermate mice; c, g–k, quantitative analyses of protein or phosphoprotein levels. a–c, Levels of phosphorylated (activated) p70 were decreased in brains of rapamycin-treated non-transgenic (a) and transgenic PDAPP (b) mice (c, **, P = 0.006 and *, P = 0.01 respectively). d, rapamycin did not alter Aβ₄₀ levels but significantly decreased soluble Aβ₄₂ levels in the brains of transgenic PDAPP mice *, P = 0.02. Homogenates were measured at 100 mg brain tissue/ml. e, rapamycin did not alter levels of endogenous mouse Aβ₄₀ levels in brains of non-transgenic mice. Aβ₄₂ levels were below the detection limit of the ELISA (not shown). f, representative immunoblots of PDAPP mouse brain extracts. g–k, Quantitative analyses of APP, C99 and C83, NEP and IDE immunoreactivity in lysates of brains from control- and rapamycin-treated PDAPP mice. Data were normalized to β-actin levels. Student's t test was used to determine significance of differences between means. Data are means ± SEM.

a, f and h, representative immunoblots of hippocampal lysates from control- and rapamycin-treated transgenic PDAPP mice and non-transgenic littermate controls. b, g and i, quantitative analyses. a and b, LC3-II levels are decreased in hippocampi of rapamycin-treated transgenic PDAPP mice (*, P = 0.0009), but not in hippocampi of rapamycin-treated non-transgenic littermates. c and d, representative epifluorescent (c, 200×) and higher-magnification confocal (d, 600×) images of hippocampal CA1 (e, green box, region of epifluorescent images; blue box, region of confocal images) in control- and rapamycin-fed transgenic PDAPP mice stained with an anti-LC3 antibody. An increase in LC3-immunoreactive puncta was observed in CA1 projections of transgenic PDAPP mice following rapamycin administration. f and g, levels of the autophagic substrate p62SQSTM are decreased (*, P = 0.0015) in hippocampi of rapamycin-treated PDAPP transgenic mice. f, representative Western blots; g, quantitative analyses of p62SQSTM levels. h and i, Levels of phosphorylated (activated) p70 were decreased in brains of rapamycin-treated PDAPP and non-transgenic mice (*, P = 0.001 and P = 0.04 respectively). Significance of differences between group means were determined using two-tailed unpaired Student's t test. Data are means ± SEM.