Figure 3.

Odor-Evoked Responses in MBON Dendrites in the β′ Lobe Are Bidirectionally Altered by Conditioning

(A) Schematic of the imaging plane and area of interest of the M4β′/MBON-β′2mp neuron.

(B) Example pseudocolored traces of calcium transients measured in the MBON dendrites in the β′ lobe in a naive fly exposed to MCH or OCT, the odors used in conditioning. Scale bar is 10 μm.

(C and D) Time courses of odor-evoked GCaMP responses (ΔF/F) collected at the level of the M4β′ neuron dendrites (C) (n = 18, nine animals) or presynaptic boutons (D) (n = 9, nine animals). Traces represent mean odor responses (solid line) and standard deviation (gray shading). Arrows indicate onset of odor presentation.

(E and F) Blocking M4β′/MBON-β′2mp and M6/MBON-γ5β′2a neurons significantly impairs 2 hr appetitive (E) (n ≥ 16, p < 0.05) and aversive memory retrieval (F) (n ≥ 19, p < 0.05). Statistics are one-way ANOVA followed by a Tukey’s HSD post-hoc test. Data shown are the mean ± SEM.

(G and H) Single example traces of calcium transients evoked by trained odors recorded from MBON dendrites in the β′ lobe 1–2 hr after (G) appetitive and (H) aversive conditioning. In these examples CS+ is the MCH responses and CS− is the OCT response. Arrows indicate onset of odor presentation.

(I) Difference of responses evoked by the CS+ (MCH in red, OCT in blue) and CS− following appetitive training relative to the mean transients of mock trained flies (also see Figures S6 and S7).

(J) Difference of responses evoked by the CS+ (MCH in red, OCT in blue) and CS− following aversive training. Shock training shifts the curve toward a relative increase of the CS+ response, while sugar training shifts the curve in the opposite direction. Data shown are the mean ± SEM. Light gray boxes indicate the time of the odor exposure.

(K and L) Bar graphs illustrate peak ± 0.5 s values of the odor response difference curves for trained and mock trained animals expressed as a percentage difference to the mean of the mock (see Experimental Procedures, Equation 2), for (K) appetitive or (L) aversive paradigms. Data are mean ± SEM; for MCH as CS+: n (appetitively trained) = 22, 11 animals, n (mock) = 19, 11 animals, p < 0.05; n (aversively trained) = 24, 13 animals, n (mock) = 19, 11 animals, p < 0.05; for OCT as CS+: n (appetitively trained) = 59, 32 animals, n (mock) = 58, 31 animals, p < 0.05; n (aversively trained) = 37, 20 animals, n (mock) = 29, 16 animals, p < 0.05; statistics are Mann-Whitney U-test.