Fig. 3: Cortical application of iSOS rescued memory deficits in HPC-lesioned mice.
From: Acquiring new memories in neocortex of hippocampal-lesioned mice
a–e Artificial iSOS in neocortex during training can rescue fear memory deficits in HPC-lesioned mice. a Scheme of the co-activation LEC fiber during encoding. The iSOS were induced by LED on cortical surface to activate oCHiEF-expressing widespread axons from LEC L5. Activation of cortical fibers but not LEC cell bodies is to avoid the unspecific activation of circuits from LEC to other brain regions. b Flow of the behavioral experiment. All four groups of mice received contextual fear conditioning (CFC) training. Two groups of HPC-lesioned mice were given iSOS during training trials. Long-term memory was tested 24 h later without iSOS. HPC intact group means no HPC lesion, no iSOS applied and no virus infected. c Travel distance of open field test upon given iSOS or not. (HPC-lesioned mice, N = 13, two-sided paired t-test, t(12) = 1.6, P = 0.1449). d Immediate freezing after foot shock (NHPC intact = 15, NHPC lesion = 19, NiSOS-10Hz = 8; NiSOS-30Hz = 8; ANOVA, F(3, 46) = 9.1, P < 0.0001; Bonferroni post-hoc test, PControl vs. HPC-lesion = 0.0005, PiSOS-10Hz vs. HPC-lesion = 0.0035, PiSOS-30Hz vs. HPC-lesion = 0.0005). e Memory test in the conditioned context (same mice as d, ANOVA, F (3, 46) = 13.5, P < 0.0001; Bonferroni post-hoc test, PControl vs. HPC-lesion < 0.0001, PiSOS-10Hz vs. HPC-lesion < 0.0001, PiSOS-30Hz vs. HPC-lesion < 0.0001). f–j Artificial iSOS rescued spatial memory deficit in HPC-lesioned mice. f Flow of the behavioral experiment. g Learning curves for the spatial memory task (mice number: NHPC-intact = 24, NHPC-lesion = 15, NiSOS (30Hz,EYFP) = 6, NiSOS (10Hz,oCHiEF) = 11; NiSOS (30Hz,oCHiEF) = 22; ANOVA, Time factor: F(3, 219) = 52.5, P < 0.0001; Group factor: F(4, 73) = 10.6, P < 0.0001; Interaction, F(12, 219) = 1.4, P = 0.1453; Bonferroni post-hoc test, PHPC-intact vs. HPC-lesion < 0.0001, P iSOS (30Hz,EYFP) vs. HPC-lesion > 0.9999, P iSOS (10Hz,oCHiEF) vs. HPC-lesion < 0.0001, P iSOS (30Hz,oCHiEF) vs. HPC-lesion < 0.0001). h Averaged occupancy maps for memory recall in day5. i Quantification of occupancy in food zone (same mice as g. ANOVA, F(4, 73) = 5.0, P = 0.0012; Bonferroni post-hoc test, PHPC-intact vs. HPC-lesion = 0.0048, P iSOS (30Hz,EYFP) vs. HPC-lesion > 0.9999, P iSOS (10Hz,oCHiEF) vs. HPC-lesion = 0.0243, P iSOS (30Hz,oCHiEF) vs. HPC-lesion = 0.0098). j Quantification of occupancy in four quadrants (same mice as g ANOVA for each quadrant, Target quadrant: F(4, 73) = 5.0, P = 0.0013; Bonferroni post-hoc test, PHPC-intact vs. HPC-lesion = 0.0159, P iSOS (30Hz,EYFP) vs. HPC-lesion = 0.9803, P iSOS (10Hz,oCHiEF) vs. HPC-lesion = 0.0102, P iSOS (30Hz,oCHiEF) vs. HPC-lesion = 0.0070, Opposite quadrant: F(4, 73) = 4.3, P = 0.0038, Bonferroni post-hoc test, P iSOS (30Hz,oCHiEF) vs. HPC-lesion = 0.0190; Left quadrant: F(4, 73) = 1.4, P = 0.2529; Right quadrant: F(4, 73) = 1.0, P = 0.4043). *P < 0.05, **P < 0.01, ***P < 0.001. Each dot represents one mouse. Error bar shows lines shows S.E.M. For all box plot, whiskers show min and max, box shows 25th, median and 75th percentile.
