Figure 1.

NSCs targeted to the fimbria fornix improve hippocampal-dependent cognition in APP/PS1 mice. Percent time exploring familiar and novel objects in the NOR test, preoperatively at 8 weeks (a) and postoperatively at 16 weeks (b). All groups performed the task normally at 8 weeks (a); however, at 16 weeks (4 weeks post-NSC/vehicle transplant), the sham group was unable to perform the task, whereas WT and NSC groups recognized the novel object and spent significantly more time exploring it (b; **p < 0.005 WT, ***p < 0.0001 NSC, novel vs. familiar; t-test). MWM at 28 weeks (16 weeks post-NSC/vehicle transplant); escape latency during 5 d hidden platform training (c; 6 trials per day) and 24 h probe trial quadrant statistics (d), escape latency (e) and swim speed (f). At days 4 and 5 of training, sham controls exhibited a deficit in the latency to find the hidden platform compared to WT (^#p < 0.05; repeated measures ANOVA), while at these same time points, NSC mice exhibited improved performance compared to sham mice (*p < 0.05; repeated measures ANOVA) (c). In a 24 h probe trial, WT and NSC-treated mice exhibited a significant preference for the TQ over the other quadrants (AL, AR, and OP) (d; *p < 0.05; ANOVA) and over chance (d; ^#p < 0.05; vs. chance; t-test with a hypothetical mean of 25%, indicated by dashed line), and reached the former platform location almost twice as fast as sham controls (e; *p < 0.05 NSC vs. sham; ^#p < 0.05 WT vs. sham; ANOVA), demonstrating a strong memory for the former platform location. These data indicate that NSCs improved short-term non-associative memory as well as spatial reference and working memory. Data are mean ± SEM (*^#p < 0.05, **p < 0.005, ***p < 0.0001). Sample size: WT n = 5 (NOR) and n = 14 (MWM in 2 cohorts), NSC n = 10, and sham n = 10.