Conclusion

We used genetic alterations of OA and TA to elucidate the role of OA and TA in survival and sugar responsiveness of fruit flies. Our data suggest complex and multiple central and peripheral actions of these amines on physiology and behavior.

OA/TA and proboscis extension response
The tßhnM18 mutant proboscis extension response to sucrose after 20 h of starvation was reduced compared to controls. Previous studies did not report a sugar sensitivity phenotype of the tßhnM18 and oamb286 mutants (Schwaerzel et al., 2003; Kim et al., 2013), probably because the T-maze assay they were using is lacking sensitivity (Colomb et al., 2009). Alternatively, the phenotype might not be visible in this assay because of the locomotion phenotype of tßhnM18 mutants (Saraswati et al., 2004; Fox et al., 2006; Koon et al., 2011; Damrau et al., in preparation). Indeed, a role of OA and TA in sugar sensitivity has been demonstrated before in Drosophila (Scheiner et al., 2014) and Apis mellifera (Buckemüller et al., in preparation) using locomotion-independent assays. Because appetitive conditioning is dependent on prior starvation, these data suggest that previous reports on the role of OA during appetitive conditioning (Schwaerzel et al., 2003; Kim et al., 2013) may have to be re-evaluated.

OA/TA and metabolism

Sugar response is dependent on starvation (Colomb et al., 2009; Damrau et al., 2014). A decreased sugar response as found in tßhnM18 mutants could be due to a decreased sensitivity to our starvation treatment. Corroborating this hypothesis, we found that the levels of carbohydrates in the hemolymph of tßhnM18 mutant flies are higher after starvation than in control flies. Since trehalose constitutes the energy store of a fly and its hemolymph concentration reflects starvation level (Thompson, 2003), it is reasonable to argue that the mutant flies were affected less by the starvation treatment than the controls even though they were deprived of food for the same amount of time. This interpretation is also supported by longer survival of tßhnM18 mutants under starvation conditions, a result which was recently reproduced (Damrau et al., 2012; Scheiner et al., 2014). Complementing our analysis in flies, injection of the OA-receptor antagonist epinastine in honey bees also prolonged survival (Buckemüller et al., in preparation). Taken together, these results suggest that the absence of OA-signaling saves the mutant animal’s energy, making them less sensitive to starvation, a conclusion in line with previous reports on the role of OA in trigylceride (Woodring et al., 1989; Erion et al., 2012) and carbohydrate (Blau et al., 1994; Park and Keeley, 1998) metabolism.

Acute OA/TA-effect rescues mutant sugar responsiveness.
From the results thus far, one would expect that increasing tßh synthesis would increase sugar response and decrease hemolymph carbohydrates.
Our temporal rescue experiments revealed that tßh expression exclusively during starvation and not the test did not increase the PER (Fig. 1B). We performed the complementary experiment showing that acute tßh expression is sufficient to increase the PER (Fig. 1A) even if tßh was not available during starvation, i.e., hemolymph carbohydrate levels have been high during the entire starvation time until the day before testing. This result suggests that the decrease of carbohydrate levels is not the only starvation-induced alteration that leads to a normal PER but that probably a second OA/TA-dependent mechanism has to be engaged during test. In addition to the internal state that is altered by starvation, the likelihood to extend the proboscis or the sensory input, or both may be modified.

OA/TA and gustatory receptor sensitivity
We recorded from taste sensilla in order to test whether the sensory input from sugar stimuli was altered in tßhnM18 mutant flies and found that tßhnM18 mutants showed a lower sensillar response to sucrose than control flies after 20 h of starvation (Fig. 1A). In fact, one can even observe a decrease in the spike frequency of the sensilla in the mutants, while the spike frequency of the control animals remained constant. Starvation can lead to an increased sensitivity of the gustatory receptors (Meunier et al., 2007; Nishimura et al., 2012), and OA/TA is able to modulate sensory neuron activity (Braun and Bicker, 1992; Erber and Kloppenburg, 1995; Kloppenburg and Erber, 1995; Pophof, 2000; de Haan et al., 2012). Here, control flies increased their proboscis extension (Fig. 1) but not their sensillar sensitivity (Fig. 1A) after starvation. That indicates that the sensillar sensitivity is not the only modulated pathway by the OA/TA-system. We rather think that there has to be a second OA/TA-independent pathway that underlies genetic variance.

Receptor mutant analysis uncoupled physiological from behavioral effects of starvation
The TßH enzyme converts TA into OA such that tßhnM18 mutants not only lack OA but also accumulate TA (Monastirioti et al., 1996). To disentangle the roles of the two amines, we tested known and newly generated OA- or TA-receptor mutants in our assays (Fig. 1). Our companion paper (Buckemüller et al., in prep.) demonstrates a PER reduction after OA-receptor antagonist injection in starved honeybees. However, we find no effect of the tested OA-receptor mutations on sugar response, suggesting an exclusive contribution of other OA-receptors, e.g. Octß3R which was found to be involved in larval feeding response (Zhang et al., 2013).
Our results show that flies can exhibit starvation resistance comparable to the wild type controls but with a lower sugar response (TyrRIIΔ24), or a higher starvation resistance than wild type controls but with a normal sugar response (Octß2RΔ3.22 and Octß2RΔ4.3). These findings strongly suggest that starvation resistance and sugar response are not mediated by the same OA/TA-cells and receptors, but by different sub-populations. In addition, the data show that both OA and TA play a role in starvation-induced sugar response. The performance trend of OA- and TA-receptor mutants goes into the same direction, such that they are probably not counteracting each other, as previously suggested for crawling behavior (Saraswati et al., 2004).

Where is the site of OA/TA-action?
In order to define responsible cells contributing to the behavioral response to sucrose and the metabolic response to starvation, we rescued tßh in different OA/TA-subsets inside or outside the nervous system, using the GAL4-UAS system (Fig. 1). The Tdc1-GAL4-driver is tßh-negative (Monastirioti et al., 1995) and expresses non-neuronally e.g. in the crop and the hind gut (Cole et al., 2005; Chintapalli et al., 2007; Blumenthal, 2009). Driving the UAS-tßh expression by Tdc1-GAL4 rescues the tßhnM18 mutant phenotype. Because OA-action in non-tßh cells is unlikely this rescue is probably due to decreased TA-levels, indicating a role of non-neuronal, TA-specific cells in peripheral tissues on eliciting a wild type-like PER. TA-signaling may be involved in mediating the metabolic signal of starvation in non-neuronal tissues is probably independent of the OA/TA-system in the brain acting on receptor sensitivity.

Conclusions
Taken together with the experiments from our accompanying paper (Buckemüller et al., in prep.), our results suggest that the OA/TA-system is involved in both the physiological and the behavioral changes that follow starvation, and that these changes are regulated independently. They also show that the behavioral change is due not only to a modulation of the taste neuron activity and to action of TA-specific cells in peripheral organs, but that a more central effect is additionally in play.