1. Volume 26, Issue 17, Pages 2351-2357 (September 2016)
Hippocampal Activation of Rac1 Regulates the Forgetting of Object Recognition Memory 
Yunlong Liu, Shuwen Du, Li Lv, Bo Lei, Wei Shi, Yikai Tang, Lianzhang Wang, Yi Zhong 
Current Biology 
Volume 26, Issue 17, Pages (September 2016)
DOI: /j.cub
Copyright © 2016 Elsevier Ltd Terms and Conditions

2. Figure 1
Decay of Object Recognition Memory Is Bi-directionally Regulated by Rac1 Activity
(A) Top: diagram of mice that were injected in the hippocampus with AAV expressing EGFP alone (Ctrl), dominant-negative form of Rac1 (Rac1-DN), or constitutively active form of Rac1 (Rac1-CA) fused with EGFP driven by CaMKIIα promoter. Bottom: representative image of brain slice from mice injected with AAV expressing EGFP in the hippocamus. Scale bar, 1 mm.
(B) Representative blots and group data of the AAV Rac1 mutants in the hippocampus. Rac1 activity was increased by injection with Rac1-CA (t test, p < 0.01, n = 3–4) and decreased by Rac1-DN (t test, p < 0.01, n = 3–4). See also Figure S1 for effects of manipulating Rac1 activity on spine morphology in hippocampal neuronal culture using these AAVs.
(C) Top: depiction of the novel object recognition task. After AAV injection for 2 weeks, mice were habituated for 10 min, and sampling was conducted after 24 hr for 5 min and then tested for 5 min at various time intervals (4 hr, 24 hr, 72 hr, and 120 hr) to calculate the memory retention curve shown in the bottom panel. Compared with the control animals, naive animals showed comparable memory retention, while mice that were injected with Rac1-DN showed a slower memory decay (4 hr, ANOVA, p = , n = 16–19; 24 hr, ANOVA, p = , n = 17–19; 72 hr, p = , n = 9–16; 120 hr, ANOVA, p = , n = 12–13), whereas mice that were injected with Rac1-CA showed dramatically accelerated forgetting (Ctrl versus Rac1-CA, 4 hr, ANOVA, p = , n = 9–18 and 24 hr, ANOVA, p = 0.0012, n = 9–15). Only the above error bar is shown. See also Figure S2 for influence of neuronal or glial Rac1 manipulation on fear memory or novel object recognition memory and effects of injecting AAV-Rac1 mutants on basal behavior.
All plots depict mean ± SEM.
Current Biology  , DOI: ( /j.cub )
Copyright © 2016 Elsevier Ltd Terms and Conditions

3. Figure 2
Temporal Effects of Retroactive Interference on Recognition Memory Are Correlated with Rac1 Activation
(A) Top: diagram of retroactive interference (RI) introduced 8 hr or 22 hr after sampling by exposure to two extra objects in the same chamber as the 24-hr object memory test. Forgetting of the original memory was induced by 22-hr RI (t test, p < ), but not 8-hr RI (t test, p = ). n = 19, n = 14, and n = 19 for No RI, 8-hr RI, and 22-hr RI, respectively.
(B) Top: diagram of RI introduced 22 hr after sampling by exposure to two extra objects in the same chamber as the 48-hr object memory test. Forgetting of the original memory was induced by 22-hr RI (t test, p = , n = 16 and n = 20 for No RI and 22-hr RI, respectively).
(C) Top: diagram of RI introduced 8 hr or 22 hr after sampling by presenting a loud tone (1 min, 12,000 Hz, 85 db) in a different chamber from the 24-hr object memory test. Memories were comparable among the groups (n = 14–16).
(D) Top: diagram of RI introduced 2 hr after sampling by exposure to two extra objects in the same chamber as the 4-hr object memory test. Short-term memory was not affected by the 2-hr RI (t test, p = , n = 25 and n = 17 for No RI and 2-hr RI, respectively). Significance (∗p < 0.05; ∗∗p < 0.01) is indicated.
(E) Top: experimental design; hippocampal tissues were isolated for the western blot analysis 30 min after sampling or RI, which was introduced either 8 hr or 22 hr after sampling. Middle and bottom: representative blots and group data showing Rac1 activation in the hippocampus of mice subjected to 22-hr RI (t test, p = ), but not 8-hr interference (t test, p = ). n = 5, n = 4, and n = 4 for No RI, 8-hr RI, and 22-hr RI, respectively.
(F) Top: experimental design; hippocampal tissues were isolated for the western blot analysis from young (12 weeks) and aging (50 weeks) mice with or without 22-hr RI. RI induced Rac1 activation in both groups (young group, t test, p = , n = 3–4; aging group, t test, p = , n = 3–4).
All plots depict mean ± SEM.
Current Biology  , DOI: ( /j.cub )
Copyright © 2016 Elsevier Ltd Terms and Conditions

4. Figure 3
Suppression of RI-Induced Forgetting by Inhibiting Rac1 Activity
(A) Top: experimental design. Two weeks after AAV injection, the mice were subjected to a 24-hr novel object recognition test with 22-hr RI. RI-induced forgetting was blocked by the neuronal expression of a CaMKIIα promoter-driven dominant-negative Rac1 protein (t test, p =  compared with CaMKIIα-EGFP, p =  compared with naive, n = 12–14), but not the glial expression of Rac1-DN (t test, p = 0.7151, n = 10 for each group).
(B) Top: experimental design. The Rac1 inhibitor Ehop016 was i.p. injected at a dose of 20 mg/kg immediately after sampling, RI was introduced 22 hr later, and the test was performed 2 hr after RI. RI-induced forgetting was also blocked by the inhibitor (t test, p = , n = 13–14). See also Figure S3 for effects of ehop016 injection on Rac1 activity in the hippocampus and basal behavior.
All plots depict mean ± SEM.
Current Biology  , DOI: ( /j.cub )
Copyright © 2016 Elsevier Ltd Terms and Conditions

5. Figure 4
Stability of LTP Is Bi-directionally Regulated by Rac1 Activity
(A) Top: typical fEPSP trace at baseline and during the LTP recordings. One train of TBS was given 15 min after the baseline recording; LTP was recorded in the Schaffer collateral pathway for 2 hr. Rac1 inhibition slowed the decay of LTP within 2 hr, whereas Rac1 activation accelerated the LTP decay compared with the control, as indicated by the fEPSP slope. The average fEPSP slope in the last hour of the LTP recording was shown on the right. Rac1-DN slices show higher fEPSP slopes than the ctrl slices (t test, p = , n = 8 in each group), whereas the Rac1-CA slices generate lower fEPSP slopes (t test, p = , n = 8 in each group).
(B) Similar to (A), except that PP-LFS (50-ms interval) was introduced 1 hr after the LTP recordings; PP-LFS failed to reverse LTP in the Rac1-DN slices but significantly induced LTD in the Rac1-DN slices compared with the baseline. The average fEPSP slope after PP-LFS is shown on the bottom right. The fEPSP slopes in the Rac1-DN slices remain higher than the ctrl (t test, p = 0.0115, n = 9–10), whereas the fEPSP slopes are significantly reduced in the Rac1-CA slices after PP-LFS compared with the ctrl slices (t test, p = 0.0001, n = 9–10).
All plots depict mean ± SEM.
Current Biology  , DOI: ( /j.cub )
Copyright © 2016 Elsevier Ltd Terms and Conditions