Octopamine and Tyramine Contribute Separately to the Counter-Regulatory Response to Sugar Deficit in Drosophila

Abstract

All animals constantly negotiate external with internal demands before and during action selection. Energy homeostasis is a major internal factor biasing action selection. For instance, in addition to physiologically regulating carbohydrate mobilization, starvation-induced sugar shortage also biases action selection toward food-seeking and food consumption behaviors (the counter-regulatory response). Biogenic amines are often involved when such widespread behavioral biases need to be orchestrated. In mammals, norepinephrine (noradrenalin) is involved in the counterregulatory response to starvation-induced drops in glucose levels. The invertebrate homolog of noradrenalin, octopamine (OA) and its precursor tyramine (TA) are neuromodulators operating in many different neuronal and physiological processes. Tyrosine-ß-hydroxylase (tßh) mutants are unable to convert TA into OA. We hypothesized that tßh mutant flies may be aberrant in some or all of the counter-regulatory responses to starvation and that techniques restoring gene function or amine signaling may elucidate potential mechanisms and sites of action. Corroborating our hypothesis, starved mutants show a reduced sugar response and their hemolymph sugar concentration is elevated compared to control flies. When starved, they survive longer. Temporally controlled rescue experiments revealed an action of the OA/TA-system during the sugar response, while spatially controlled rescue experiments suggest actions also outside of the nervous system. Additionally, the analysis of two OA- and four TA-receptor mutants suggests an involvement of both receptor types in the animals' physiological and neuronal response to starvation. These results complement the investigations in Apis mellifera described in our companion paper (Buckemüller et al., 2017).

Keywords: biogenic amines; insects; proboscis extension response; starvation; starvation resistance.

Figure 1

Reduced sugar response in tßh^nM18 mutant flies compared to wild-type flies after 20 h of starvation. (A) Fraction of flies that responded to each concentration of sucrose. (B) Total number of positive responses. Boxplots represent the median (bar), the 75- and 25%-quartiles (box) and data within 1.5 times the interquartile range (whiskers). Data outside 1.5 times the interquartile range are considered as outliers (black dots). Numbers indicate sample size, asterisk denotes significant difference between genotypes (Wilcoxon rank sum test, p = 1.8 × 10⁻¹²).

Figure 2

Change in hemolymph glucose and trehalose after starvation is smaller in tßh^nM18 mutants than in wild type. (A) Concentration of trehalose and glucose in the hemolymph of fed and 20 h starved flies, which was calculated from absorbance at 540 nm compared to calibration solutions, is shown in boxplots. Numbers indicate sample size. No statistical test was applied. (B) Index of the change (difference over the sum of the two numbers) in absorbance between the starved and the fed fly, paired per day. Numbers indicate sample size, asterisk denotes significant difference between genotypes (paired Wilcoxon rank sum test, p = 0.03223).

Figure 3

Longer survival of tßh^nM18 mutants under starvation conditions. (A) Kaplan Meier survival curve for the two genotypes. We ran 16 experiments with about 35 flies per vial. The difference between the curves was statistically significant, while there was also an effect of the different trials (Cox proportional hazards regression model, p = 0.029 for trials, p = 2 × 10⁻⁹ for genotypes).

Figure 4

Effects of temporally controlled (ubiquitous) expression of Tßh in tßh^nM18 mutant background on flies sugar response. (A) Temperature shift 3 h before the test. Flies with a rescue construct showed an intermediate PER level, significantly different from both the mutant and the control flies, Wilcoxon rank sum test with Bonferroni correction (uncorrected p = 0.00019). (B) Temperature shifts during the starvation period, but not immediately before the test. Flies with a rescue construct behaved similarly as mutant flies. Total number of proboscis extension responses is represented in boxplots (see Figure 1). Numbers indicate sample size, asterisks denote significant difference between groups (paired Wilcoxon rank sum test, p = 0.00041 and 0.00781).

Figure 5

Effect of starvation on taste neuron sensitivity: electrophysiological recording from different gustatory sensilla on the labellum. (A) In sated and starved tßh^nM18 mutants and their respective controls and (B) in sated and starved usual wild-type flies. Extracellular action potentials within 1 s after stimulation onset were counted and plotted as boxplots. Numbers represent the sample size of the recorded sensilla, Different letters denote significant differences (paired Wilcoxon rank sum test, (A): p = 0.037 and 0.048, with Bonferroni correction, (B): w1118 p = 0.03938, CS, p = 0.00174).

Figure 6

Spatially controlled Tßh expression in tßh^nM18 mutant background. Ubiquitous (actin), pan-neuronal (nSyb) and non-neuronal TDC (Tdc1) drivers significantly increased sugar responsiveness. Neuronal TDC (tdc2) and OA (NP7088) specific drivers did not alter sugar responsiveness. Boxplots depict total number of proboscis extensions in hemizygous mutant males with or without a UAS-tßh construct, and heterozygous for the Gal4 driver. Numbers indicate sample sizes, asterisks denote significant difference between the mutant its respective rescue group (Wilcoxon rank sum test, Actin p = 0.01621, Tdc1 p = 0.02782, nSyb p = 0.01341).

Figure 7

Starvation resistance and sugar responsiveness in TA- and OA-receptor mutants. Total number of proboscis extensions (A) and Kaplan Meier survival curve (B, see Figure 3) in different female mutants and their respective control strains. See text for fly strain labels. The three different groups were tested independently and are therefore statistically treated as different experiments. (A) Numbers indicate sample sizes, asterisks denote significant difference between the mutant and its respective control group (Wilcoxon rank sum test, honoka p = 0.0117, Tyr p = 0.002911 and 0.007432). (B) 5 to 8 experiments were run per genotype, with about 35 flies per vial. Cox proportional hazards regression model was used to test statistical differences between mutants and their control, (different: honoka p = 1.4 × 10⁻¹⁴, TyrR,TyrII p = 10⁻⁵, TyrII p = 2.1 × 10⁻⁶, oamb²⁸⁶ p = 0.00029, not different: TyR p = 0.067, oamb⁵⁸⁴ p = 0.069, before bonferroni correction).