Figure 3. Odd neurons receive input from PNs and respond to odor. All images are second instar larval brains. A, Confocal stack of a larval brain expressing the split-GFP (GRASP; green) in PNs (GH146lexA) and Odd neurons (Odd-gal4). Reconstitution of GFP is restricted to the calyx (dotted circles). B, Confocal stack of a larval brain expressing the split-GFP in Kenyon cells (MB247lexA driver) and Odd neurons (Odd-gal4 driver). A few GRASP puncta localize to the calyx (dotted circle). C, Confocal stack, high-magnification (63×) image of the calyx showing split-GFP localization between the PNs and Odd neurons (green) colabeled with nc82, which labels presynaptic densities. GRASP puncta are distributed throughout most of the calyx. D, Single confocal image (63×, 5× zoom) of the calyx showing split-GFP localization between the PNs and Odd neurons (green) in the background of nc82 staining labeling presynaptic sites. Part of the GRASP puncta colocalizes with nc82 (examples labeled by arrows), showing that these puncta are likely synapses. E, Response of Odd neurons expressing GCaMP6 to ACV. E, Before odor stimuli. E′, During odor stimuli. E″, After odor stimuli. Some cells respond to odor (arrow). F, Response of a single cell to repeated cycles of odor presentation. Gray boxes represent odor stimuli. G, Dose–response curve showing the response profile of individual Odd neurons to different concentrations of ACV. Gray boxes represent odor stimuli. H, Lowering the concentration 10-fold significantly decreases the average maximum responses of the Odd neurons to ACV, (Kruskal–Wallis test, Bonferroni post-test, p < 0.05) but responses are similar between 1:10 and 1:20 dilutions of ACV. I, Response profile of PNs in larvae in which the Odd neurons are either silenced using Kir2.1 or made more excitable using NaChBac. Gray boxes represent odor stimuli. J, Neither manipulation affects the average maximum response of PNs (Kruskal–Wallis test, Bonferroni post-test). Scale bars: A, B, F–I, 20 μm; C, E, 10 μm; D, 5 μm.