Skip to main content
. 2016 Jun 1;90(5):1086–1099. doi: 10.1016/j.neuron.2016.04.034

Figure 2.

Figure 2

Aversive Learning Drives Persistent Depression of Conditioned Odor Drive to MVP2 Neurons yet Output Is Only Required to Express Short-Term Aversive Memory

(A) Schematic of MVP2 neuron showing imaging plane (dotted line).

(B) Example pseudocolor images of baseline and MCH-evoked GCaMP fluorescence recorded from MVP2 dendrites in a living fly. ROI indicated by white ellipse.

(C) Aversive conditioning depresses the relative CS+ to CS− odor-drive to MVP2 neurons. CS+ and CS− odor-evoked calcium transients were imaged 30–60 min after mock or regular shock conditioning (red curves: OCT, blue curves: MCH). Data are mean [solid line] ± SEM [shaded area] normalized curves (see Experimental Procedures).

(D) Bar graphs represent percent difference to the mean mock integrated peak response (4.5 ± 1.5 s after odor delivery, see methods) (Mann-Whitney U-test; OCT is CS+ (top): n(mock) = 7, n(trained) = 11, p < 0.05. MCH is CS+ (bottom): n(mock) = 8, n(trained) = 11, p < 0.05).

Difference of responses evoked by CS+ and CS− after aversive conditioning relative to the mean responses after mock training (red curve: OCT is CS+, blue curve: MCH is CS+). Data are mean [solid line] ± SEM [shaded area] normalized curves (see Experimental Procedures).

(E and F) Same as in (C) and (D), but odor-responses were imaged 3 to 4 hr after training. Bar graphs represent percent difference to the mean mock integrated peak response (4.5 ± 1.5 s after odor delivery) (Mann-Whitney U-test; OCT is CS+ (top): n(mock) = 5, n(trained) = 5, p < 0.05. MCH is CS+ (bottom): n(mock) = 11, n(trained) = 8, p < 0.05).

(G) Blocking output from R83A12 or MB112C neurons during testing impaired 30 min aversive memory performance compared to the relevant controls (Kruskal-Wallis, n = 18–22, p < 0.001 and ANOVA, n = 10–13, p < 0.01, respectively).

(H) Blocking output from R83A12 or MB112C neurons during testing did not impair 3 hr aversive memory (Kruskal-Wallis, n = 25, p > 0.9 and ANOVA, n = 9 to 10, p > 0.2, respectively).

(I) Activating R83A12 or MB112C neurons during testing impaired 30 min aversive memory (ANOVA, n = 9 to 10, p < 0.001 and ANOVA, n = 13 to 14, p < 0.01, respectively).

(G–I) Schematics depict temperature protocols. All flies were trained at 23°C and tested at 33°C. Data are mean ± SEM. See Figure S2 for permissive control. Asterisks indicate statistical significance.