Fig. 3: Encoding-recruited acute and repeated cocaine ensembles are reactivated during the retrieval of cocaine-context associative memories.
From: Cocaine-context memories are transcriptionally encoded in nucleus accumbens Arc ensembles

a Experimental design for CPP ensemble tagging in Arc-CreERT2::Ai14 mice. 4-OHT 4-Hydroxytamoxifen, 1āP 1 Pairing, 5āP 5 Pairings. b Time spent in the paired vs unpaired chamber (left) and CPP score (right). nā=ā7, weak; nā=ā6, strong. Left: Two-way RM-ANOVA: interaction conditioning x session F1,24ā=ā6.23, *pā=ā0.0199; main effect of conditioning F1,24ā=ā4.89, *pā=ā0.0368; main effect of session F1,24ā=ā9.05, **pā=ā0.0061; followed by Å idĆ”k post-hoc tests. Right: Two-tailed unpaired t-test: t12ā=ā2.791 *pā=ā0.0163. c Representative images of cells activated during conditioning (Tom+ā, red), test (ARC+ā, green) or both (arrows indicated). Scalebar, 50āµM. d Reactivation ratio (as āRecallā, over total number of Tom+ cells). nā=ā4, 1āP/homecage; nā=ā5, 5āP/homecage; nā=ā8, 1āP/test; nā=ā10, 5āP/test. Two-way ANOVA: interaction conditioning x test, F1,19ā=ā3.559, pā=ā0.0746; main effect of conditioning, F1,19ā=ā8.783, **pā=ā0.008; main effect of test, F1,19ā=ā104.9, ***pā<ā0.0001; followed by Å idĆ”k post-hoc tests. e Chance-adjusted probabilities of cell activation during both conditioning (Tom+ā, red) and test (ARC+ā, green). nā=ā8595 cells in nā=ā4 mice, 1āP/homecage; nā=ā16538 cells in nā=ā7 mice, 1āP/test; nā=ā9951 cells in nā=ā5 mice, 5āP/homecage; nā=ā16130 cells in nā=ā7 mice, 5āP/test. Baseline-Category Logit Mixed Model (BCLogMM), equation for Tom+/ARC+ counts: interaction conditioning x test, zā=āā2.093, *pā=ā0.0363; main effect of conditioning, zā=āā1.002, pā=ā0.3163; main effect of test, zā=ā8.984, ***pā<ā0.0001; followed by Å idĆ”k post-hoc tests. f Correlation between the reactivation ratio and CPP score. 1-pairing (left): nā=ā7; Pearsonās rā=ā0.8926, **pā=ā0.0068. 5-pairing (right): nā=ā7, Pearsonās rā=āā 0.8912, **pā=ā0.0071. Bar graphs are expressed as meanā±āSEM. Regression lines and probability estimates are shown with their 95% confidence interval. Source data are provided as a Source Data File. Created in BioRender. Parise, E. (2025) https://BioRender.com/voi7l54.