Compared with wild-type mice, the increase of BDNF levels in Bdnf+/− mice and that of KIF1A levels in Kif1a+/− mice was attenuated following 3 weeks of enrichment exposure (enriched/nonenriched ratio: Bdnf+/− mice, BDNF, 1.15 ± 0.03, p = 0.0039; Kif1a+/− mice, KIF1A, 1.13 ± 0.02, p = 0.0056, two-tailed t test) ( Figure 1D). Importantly, no significant increase in KIF1A levels was observed in enriched Bdnf+/− mice (1.06 ± 0.03, p = 0.1501, two-tailed t test), while enriched Kif1a+/− Target Selective Inhibitor Library in vitro mice showed a moderate increase of BDNF levels (1.37 ± 0.04, p < 0.001, two-tailed t test) ( Figure 1D). These results indicate the possibility that BDNF works as an upstream signal in enrichment-induced

KIF1A upregulation. Next, to examine the Idelalisib cell line possible role of KIF1A upregulation in spatial learning ability, we performed the Morris water maze test for wild-type, Bdnf+/−, and Kif1a+/− mice after enrichment. After enrichment for 3 weeks, wild-type mice performed significantly better in the hidden platform trial (nonenriched versus enriched: latency,

F(1,22) = 13.50, p = 0.0013, two-way repeated-measures ANOVA) ( Figure 2A), spent significantly more time in the target quadrant (TQ) in the probe test (non-enriched versus enriched: 31.4% ± 3.5% versus 45.6% ± 3.2%, p = 0.0069, two-tailed t test) ( Figure 2B), and crossed the targeted area more frequently (nonenriched versus enriched: 3.08 ± 0.36 versus 5.08 ± 0.48, p = 0.0030, two-tailed t test) than nonenriched wild-type mice ( Figure S2A). These data are consistent with previous reports ( Kempermann et al., 1997). Compared with nonenriched littermate control mice, nonenriched Bdnf+/− mice exhibited impaired spatial learning (littermate control versus Bdnf+/−: latency, F(1,22) = 7.724, p = 0.0109, two-way repeated-measures ANOVA; time in TQ in probe test, 42.6% ± 3.6% versus 28.4% ± 3.9%, p = 0.0134, two-tailed t test) ( Figures S2M–S2O), as previously reported ( Linnarsson et al., 1997). Nonenriched Kif1a+/− mice showed

an Thiamine-diphosphate kinase intact spatial learning ability (littermate control versus Kif1a+/−: latency, F(1,22) = 0.3175, p = 0.5788, two-way repeated-measures ANOVA; time in TQ in probe test, 42.9% ± 3.5% versus 41.2% ± 3.8%, p = 0.7465, two-tailed t test) ( Figures S2T–S2V). Significantly, in contrast to wild-type mice, no enhancement of spatial learning was found in enriched Bdnf+/− or Kif1a+/− mice, compared with respective nonenriched mice (nonenriched versus enriched Bdnf+/−: latency, F(1,22) = 0.2391, p = 0.6297, two-way repeated-measures ANOVA; time in TQ in probe test: 39.1% ± 3.3% versus 40.1% ± 3.4%, p = 0.8460, two-tailed t test; number of platform crossings: 3.83 ± 0.51 versus 3.92 ± 0.49, p = 0.9076, two-tailed t test; non-enriched versus enriched Kif1a+/−: latency, F(1,22) = 1.103, p = 0.3050, two-way repeated-measures ANOVA; time in TQ in probe test: 42.0% ± 3.1% versus 44.1% ± 3.0%, p = 0.6424, two-tailed t test; number of platform crossings: 4.25 ± 0.44 versus 4.42 ± 0.40, p = 0.