Fig. 2
Nlgn3−/y rats display reduced classic freezing behaviour in an auditory fear conditioning paradigm. A Schematic of the auditory fear conditioning protocol. B Nlgn3−/y rats show less classic freezing behaviours during the conditioning phase (p < 0.0001, F(1, 22) = 6.61, repeated measures two-way ANOVA, WT n = 12, KO n = 12). PT: Pre-tone. C Nlgn3−/y rats show less classic freezing behaviours during the recall and extinction phase (p = 0.001, F(1, 22) = 13.36, post hoc two-way ANOVA, WT n = 12, KO n = 12). D Nlgn3−/y rats show reduced classic freezing behaviours during the second extinction phase (p = 0.0009, F(1, 22) = 14.61, repeated measures two-way ANOVA, WT n = 12, KO n = 12). E When analysed as “immobility response” (all four paws unmoving but allowing for movement of head and neck) Nlgn3−/y rats show significantly increased response to CS in comparison with classic freezing scoring (p = 0.004, F(1, 22) = 13.31, post hoc two-way ANOVA, KO n = 12). WT rats also show significantly increased paw immobility response in comparison with classic freezing behaviour (p = 0.019, F(1, 22) = 7.58, post hoc two-way ANOVA, WT n = 12). Expression of paw immobility response behaviour is significantly lower in Nlgn3−/y rats in comparison with WT (p < 0.0001, F(1, 22) = 3.26, post hoc two-way ANOVA, WT n = 12, KO n = 12). F Percentage time exhibiting a fear response (defined as either classic freezing (black, purple) or immobility of paws (grey, pink) for pre-tone and average of tones 1–5 of recall shows a significant interaction between genotype, method of scoring, and presence of CS (p = 0.012, F(1, 22) = 7.52, three-way ANOVA, WT n = 12, KO n = 12). Both WT and Nlgn3−/y rats display significant response to the CS (WT classic freezing: p < 0.0001, WT paw immobility: p < 0.0001, KO classic freezing: p = 0.008, KO paw immobility: p < 0.0001, post hoc Bonferroni-corrected paired t-tests). Scoring method does not affect fear response behaviour during recall for WT rats (p = 0.24, post hoc Bonferroni-corrected paired t-test) however a significantly higher paw immobility response is expressed by Nlgn3−/y rats in comparison with classic freezing behaviour (p < 0.0001, post hoc Bonferroni-corrected paired t-test). Data represented as mean ± SEM