Activation of specific mushroom body output neurons inhibits proboscis extension and feeding behavior

The ability to modify behavior based on prior experience is essential to an animal’s survival. For example, animals may become attracted to a previously neutral odor or reject a previously appetitive food source upon learning. In Drosophila, the mushroom bodies (MBs) are critical for olfactory associative learning and conditioned taste aversion, but how the output of the MBs affects specific behavioral responses is unresolved. In conditioned taste aversion, Drosophila shows a specific behavioral change upon learning: proboscis extension to sugar is reduced after a sugar stimulus is paired with an aversive stimulus. While studies have identified MB output neurons (MBONs) that drive approach or avoidance behavior, whether the same MBONs impact proboscis extension behavior is unknown. Here, we tested the role of MB pathways in modulating proboscis extension and identified 10 MBON split-GAL4 lines that upon activation significantly decreased proboscis extension to sugar. Activating several of these lines also decreased sugar consumption, revealing that these MBONs have a general role in modifying feeding behavior beyond proboscis extension. Although the MBONs that decreased proboscis extension and ingestion are different from those that drive avoidance behavior in another context, the diversity of their arborizations demonstrates that a distributed network influences proboscis extension behavior. These studies provide insight into how the MB flexibly alters the response to taste compounds and modifies feeding decisions.


59
In support of the model that individual MBONs encode a positive or negative valence, 60 the behavioral roles of MBONs have been investigated through direct activation (15,17).    Fig 1B). Importantly, the decrease in proboscis extension was not due to fly paralysis, as 138 determined by measuring walking speed in a locomotor assay ( Fig S1). The 10 MBON split-Gal4 139 lines that reduced proboscis extension are not localized to any single compartment or lobe.

167
Because the strong Shibire effector has been reported to produce phenotypes at the 168 permissive temperature (15), we repeated these experiments using the weaker 1xUAS-Shi ts1 . We 169 also altered the behavioral paradigm to stimulate with 50 mM sucrose instead of 100 mM 170 sucrose, as PER to 100 mM sucrose under control conditions was high, creating the possibility of 9 171 ceiling effects. Under these conditions, 1 of the 10 MBON split-Gal4 lines (MB310C) showed 172 increased proboscis extension upon neural silencing ( Fig 2B).

173
Finally, we tested an additional acute silencing strategy that provides rapid light-triggered 227 lines showed decreased PER to sucrose (Fig 4).

241
In addition to these lines whose activation decreased PER, we also found one DAN split-242 Gal4 line whose activation caused spontaneous PER: MB296B, a split-Gal4 for PPL-γ2α'1 ( Fig.   243 4). We re-tested this line with the effector CsChrimson and found robust PER to red light (Fig.   244 S2A). Chronic silencing of these neurons with the inward rectifying potassium channel Kir2.1 245 resulted in increased PER compared to genetic controls (Fig S2B). Acute silencing with gtACR1 246 did not have a significant effect (Fig S2C). MB296B labels some neurons outside PPL-γ2α'1 in 247 the SEZ where gustatory sensory axons terminate and proboscis extension motor neurons are 248 located (Fig. S2D). To address the contribution of non-PPL-γ2α'1 in MB296B, we used an 249 intersectional strategy to restrict CsChrimson expression to the SEZ using a Hox gene promoter 250 that overlaps with the expression of MB296B (Fig. S2D). In flies that express the red-light 264 PER to sucrose and identified 3 lines that decreased PER (Fig 5). These studies provide insight 265 into how MBONs influence taste responses and encode behavior.

307
The finding that many MBONs reduce the probability of proboscis extension is consistent 308 with the view that the MB is a complex, distributed circuit. The lines with the greatest PER 309 suppression phenotype cover 7 different cell types: γ4>γ1γ2, α1, β'1, γ2α'1, α'2, α2sc, and the 310 calyx, demonstrating that multiple compartments can influence this behavior (Fig. 5). In 311 addition, 3 dopaminergic inputs into the MBs were also sufficient to regulate PER. These 312 innervate the PAM-β'2(amp) and PAM-α1 compartments as well as weakly label PAM-γ5 and

366
In the CsChrimson screen, we simultaneously activated MBONs while presenting 100 367 mM sucrose to the tarsi. This concentration was chosen because it is a moderately appetitive 368 stimulus that results in proboscis extension ~50% of the time in control flies. Flies were water-369 satiated before the experiment and between trials, and presented with the tastant and red light 370 until proboscis extension was observed, for up to 5 seconds. Flies were given a score of 0 (for no 371 extension) or 1 (for full extension), and the average was taken across two trials.

375
For subsequent optogenetic (Chrimson88, gtACR1) PER assays with both MBONs and 376 DANs, flies were assayed using 10 mM sucrose because 50 mM sucrose elicited close to 100% 377 PER in the dark condition for some lines.

379
Temporal consumption assays were performed as described (23). Flies were glued onto 380 glass slides using nail polish or UV glue, then allowed to recover in a humidified chamber for 2-381 4 hours. Each fly was water-satiated, then presented with 100 mM sucrose on the proboscis and 382 forelegs. Cumulative drinking time over 10 consecutive presentations was recorded.