Water sensor ppk28 modulates Drosophila lifespan and physiology through AKH signaling

Abstract

Sensory perception modulates lifespan across taxa, presumably due to alterations in physiological homeostasis after central nervous system integration. The coordinating circuitry of this control, however, remains unknown. Here, we used the Drosophila melanogaster gustatory system to dissect one component of sensory regulation of aging. We found that loss of the critical water sensor, pickpocket 28 (ppk28), altered metabolic homeostasis to promote internal lipid and water stores and extended healthy lifespan. Additionally, loss of ppk28 increased neuronal glucagon-like adipokinetic hormone (AKH) signaling, and the AKH receptor was necessary for ppk28 mutant effects. Furthermore, activation of AKH-producing cells alone was sufficient to enhance longevity, suggesting that a perceived lack of water availability triggers a metabolic shift that promotes the production of metabolic water and increases lifespan via AKH signaling. This work provides an example of how discrete gustatory signals recruit nutrient-dependent endocrine systems to coordinate metabolic homeostasis, thereby influencing long-term health and aging.

Keywords: adipokinetic hormone signaling; taste.

Fig. 1.

Loss of ppk28 function increases healthy lifespan in D. melanogaster. (A and B) Kaplan–Meier survival curves for female ppk28 deletion mutants in both w^1118-VDRC (w) [n = 246 (w); n = 248 (w,ppk28^Δ); mean lifespan increase of 20.68 d (43.55%)] (A) and yw [n = 244 (yw); n = 248 (yw,ppk28^Δ); mean lifespan increase of 13.48 d (24.78%)] (B) background and their corresponding background controls. (C) Analysis of vertical distance climbed in the longitudinal negative geotaxis assay of ppk28 mutant (w,ppk28^Δ) and background control (w) female flies (n = 10 groups of 20 flies per genotype per time point). P_age < 1 × 10⁻¹⁵, P_genotype = 1.13 × 10⁻¹⁰, and P_interaction = 0.031 for ANCOVA. Error bars indicate ±SEM. (D) Kaplan–Meier survival curves for approximately 2-wk old female ppk28 mutant (w,ppk28^Δ) and background control (w) female flies [n = 149 (w); n = 150 (w,ppk28^Δ)] under starvation conditions. (E and F) Survival curves for female ppk28 deletion mutants (w,ppk28^Δ) and background controls (w) [n = 245 (w); n = 247 (w,ppk28^Δ); mean lifespan increase of 17.50 d (30.54%)] (E) compared with survival curves representing the addition of a genomic region containing the endogenous ppk28 locus into both ppk28 mutant [w,ppk28^Δ;+;Dp(320)] and control [w;+;Dp(320)] backgrounds (n = 180 [w;+;Dp(320)]; n = 137 [w,Δppk28;+;Dp(320)]; mean lifespan increase of 1.14 d [1.65%]) (F) on SY5% (wt/vol) food. Pairwise comparisons between statistically significantly different genotypes in survival analyses yielded P < 1 × 10⁻⁶ by log-rank test for all cases.

Fig. 2.

ppk28 modulates lifespan through nutrient signaling. (A and B) Longitudinal measures of whole-fly TAG (A) and glucose (B) in ppk28 deletion mutant female flies (w,ppk28^Δ) and in mutant animals also containing a ppk28 genomic rescue construct [w,ppk28^Δ;+;Dp(320)], as well as their appropriate genetic background controls [w and w;+;Dp(320), respectively]. n = 8–12 groups of five flies per genotype per time point. **P < 0.01; ***P < 0.001 for the interaction term of two-way ANOVA. Error bars indicate ±SEM. (C and D) Kaplan–Meier survival curves for female ppk28 deletion mutant flies (w,ppk28^Δ) and background controls (w) [n = 162 (w); n = 249 (w,ppk28^Δ); mean lifespan increase of 20.68 d (36.96%)] (C) and the same backgrounds containing loss of function of AKHR (w;AkhR−) and (w,ppk28^Δ;AkhR–) [n = 253 (w;AkhR−); n = 248 (w,ppk28^Δ;AkhR−); mean lifespan increase of 0.67 d (0.99%)] (D). (E and F) Survival curves for female ppk28 deletion mutant flies (w,ppk28^Δ) and their controls (w) [n = 221 (w); n = 243 (w,ppk28^Δ); mean lifespan increase of 14.16 d (24.73%)] (E) and same backgrounds containing deletion of the FoxO transcription factor (w;dFoxO− and w,ppk28^Δ;dFoxO−) [n = 248 (w;dFoxO−); n = 80 (w,ppk28^Δ;dFoxO−); mean lifespan increase of 2.4 d (4.28%)] (F). Pairwise comparisons between statistically significantly different genotypes yielded P < 1 × 10⁻⁶ by log-rank test for all cases.

Fig. 3.

AKH signaling mediates ppk28 effects on lifespan and physiology. (A) Quantification of AKH levels in axonal projections of adult CC. Preparations from fully fed female ppk28 deletion mutant flies (w,ppk28^Δ) as well as fed and starved control animals (w) were stained with α-dAKH and categorized as having high, low, or no observable AKH. A representative image from each category is shown for reference [n = 19 (w); n = 15 (w,ppk28^Δ); n = 6 (w,starved)]. (B) Kaplan–Meier survival curves for female flies with activated Akh-ergic neurons at both an activating (29 °C) and control nonactivating (23 °C) temperature (neuronal activation: w;UAS-dTRPA1/+;Akh-Gal4/+; Gal4 construct control: w;;Akh-Gal4/+; UAS construct control: w;UAS-dTRPA1/+) on SY5% food [23 °C: n = 202 (w;;Akh-Gal4/+); n = 235 (w;UAS-dTRPA1/+); n = 237 (w;UAS-dTRPA1/+;Akh-Gal4/+); mean lifespan decrease of 3 d (2.50%) and 3.19 d (3.65%) compared with Gal4 and UAS construct controls, respectively; 29 °C: n = 207 (w;;Akh-Gal4/+); n = 237 (w;UAS-dTRPA1/+); n = 235 (w;UAS-dTRPA1/+;Akh-Gal4/+); mean lifespan extension of 3.80 d (8.21%) and 3.74 d (8.07%) compared with Gal4 and UAS construct controls, respectively]. Pairwise comparisons between statistically significantly different genotypes yielded P < 1 × 10⁻⁶ by log-rank test for all cases.

Fig. 4.

Perceived scarcity of external water promotes longevity through metabolic water production. (A) Difference in wet and dry mass for approximately 2-wk-old ppk28 deletion mutant background control (w) and ppk28 deletion mutant (w,ppk28^Δ) female flies (n = 10 groups of 10 flies per genotype). *P < 0.05 two-sided Student’s t test. Error bars indicate ±SEM. (B) Kaplan–Meier survival curves for ppk28 deletion mutant background control (w) and ppk28 deletion mutant (w,ppk28^Δ) female flies under desiccating conditions (n = 100 flies for both genotypes). Pairwise comparisons between genotypes yielded P < 1 × 10⁻⁶ by log-rank test. (C) Model for control of metabolic homeostasis and lifespan by ppk28-mediated gustatory inputs. FFA, free fatty acid; SOG, subesophageal ganglion.