Fig. 5

cAMP-GEF II ^−/− mice showed impaired reversal learning in the place preference learning task. a, Novel location recognition test. Left panel, experimental design. Right panel, no difference between genotypes in the discrimination index, which indicates that spatial memory is normal in cAMP-GEF II ^−/− mice (6-month-old male mice; wild-type (WT) = 10 mice; cAMP-GEF II ^−/− (KO) = 12 mice; unpaired t-test, p = 0.63). The discrimination index was calculated as follows: discrimination index = (contact duration for object B)/(total contact duration for objects). b, There were no differences in escape latency between genotypes in the Morris water maze test during training days from day 1 to 13 (6-month-old male mice; 12 mice per genotype; Two-way RM ANOVA, F_{(1, 22)} = 0.20, p = 0.66 for genotype; F_(13,286) = 11.28, p < 0.00001 for day; F_(13,286) = 0.60, p = 0.85 for genotype and day interaction). c, Stay time (WT = 34.54 ± 2.85 s; KO = 32.25 ± 2.69 s) in the initial target quadrant during a probe trial on day 14 showed that cAMP-GEF II ^−/− mice have similar spatial memory to wild-type mice (Two-way ANOVA; F_{(1, 66)} = 0.94, p = 0.33 for genotype; F_(2,66) = 17.76, p < 0.00001 for quadrant; F_(2,66) = 0.41, p = 0.66 for genotype and quadrant interaction). d, Wild-type and cAMP-GEF II ^−/− mice crossed more frequently the platform position in the target quadrant where the platform was located than pseudo-positions in other quadrants (Two-way ANOVA; F_(2,90) = 19.71, p < 0.00001 for position; F_(1,90) = 0.25, p = 0.62 for genotype; F_(2,90) = 1.28, p = 0.28 for genotype and position interaction; post-hoc Bonferroni test p = 0.001 between positions in WT and p = 0.004 between positions in KO) during a probe trial after initial learning. e, Escape latency to the new platform during reversal training was not different between genotypes (two-way RM ANOVA; F_{(1, 22)} = 0.27, p = 0.61 for genotype; F(4,88) = 14.92, p < 0.00001 for day; F_(4,88) = 0.95, p = 0.44 for genotype and day interaction). f, Stay time in the new target quadrant during a reversal probe trial on day 19. Wild-type and cAMP-GEF II ^−/− mice showed significant preference for the new target quadrant compared to opposite or adjacent quadrants, resulting in no difference between genotypes (Two-way ANOVA; F_{(1, 90)} = 0.1, p = 0.75 for genotype; F_(2,90) = 21.84, p < 0.00001 for quadrant; F_(2,90) = 1.27, p = 0.29 for genotype and quadrant interaction). g, cAMP-GEF II ^−/− mice crossed less frequently the platform position in the new target quadrant during the reversal probe trial (Two-way ANOVA; F_(1,92) = 5.48, p = 0.021 for position; F_(1,92) = 1.5, p = 0.22 for genotype; F_(1,92) = 1.24, p = 0.27 for genotype and position interaction; post-hoc Bonferroni test p = 0.015 between positions in WT and p = 0.39 between positions in KO). h, Experimental scheme for place preference and reversal learning test in IntelliCage. Performance was quantified as the percentage of correct corner visits (4-month-old female mice; 12 mice per genotype). i, There was no difference in spatial memory between the two genotypes in the place preference learning test (Two-way RM ANOVA; F_{(1, 22)} = 2.35, p = 0.14 for genotype; F_(2,44) = 71.2, p < 0.00001 for day; F_(2,44) = 0.55, p = 0.58 for genotype and day interaction). j, The percentage of correct corner visits in the reversal learning test was significantly reduced in cAMP-GEF II ^−/− mice, indicating a deficit in behavioral flexibility (Two-way RM ANOVA; F_{(1, 22)} = 6.03, p = 0.022 for genotype; F_(2,44) = 64.0, p < 0.0001 for day; F_(2,44) = 0.84, p = 0.43 for genotype and day interaction; post-hoc unpaired t-test, day 8 ( p = 0.0298), day 11 (p = 0.1376), day 14 (p = 0.0306). k and l, Learning speeds in the first days of the place preference (k, day 1) and reversal learning (l, day 8) tests. The slope of the learning curve in each mouse was determined by the least squares analysis. Black dashed lines indicate the chance level. Thin and thick blue lines represent wild-type mice and average, respectively. Thin and thick red lines represent cAMP-GEF II ^−/− mice and average, respectively. The slope was significantly decreased in cAMP-GEF II ^−/− mice in both place preference (PP) and place preference reversal (PPR) tests (for PP: WT, slope = 0.57 ± 0.03; cAMP-GEF II ^−/−, slope = 0.49 ± 0.022; Unpaired t-test, p = 0.048; for PPR, WT, slope = 0.64 ± 0.036; KO, slope = 0.53 ± 0.024; Unpaired t-test, p = 0.026). All data are shown as mean ± SEM