Juvenile hormone regulates body size and perturbs insulin signaling in Drosophila

Abstract

The role of juvenile hormone (JH) in regulating the timing and nature of insect molts is well-established. Increasing evidence suggests that JH is also involved in regulating final insect size. Here we elucidate the developmental mechanism through which JH regulates body size in developing Drosophila larvae by genetically ablating the JH-producing organ, the corpora allata (CA). We found that larvae that lack CA pupariated at smaller sizes than control larvae due to a reduced larval growth rate. Neither the timing of the metamorphic molt nor the duration of larval growth was affected by the loss of JH. Further, we show that the effects of JH on growth rate are dependent on the forkhead box O transcription factor (FOXO), which is negatively regulated by the insulin-signaling pathway. Larvae that lacked the CA had elevated levels of FOXO activity, whereas a loss-of-function mutation of FOXO rescued the effects of CA ablation on final body size. Finally, the effect of JH on growth appears to be mediated, at least in part, via ecdysone synthesis in the prothoracic gland. These results indicate a role of JH in regulating growth rate via the ecdysone- and insulin-signaling pathways.

Keywords: developmental hormones; insect physiology; size control.

Fig. 1.

Ablation of the CAX does not influence critical weight in Drosophila. (A) The time to pupariation for CAX and control larvae starved at different weights. The inflection in the regression line indicates the point at which larvae have reached critical weight. (B) Critical weight in CAX and control larvae is not significantly different (permutation test), as determined by a breakpoint analysis of the bisegmental regression line in A. (C) CAX larvae attain critical weight later than controls (permutation test), and pupariate slightly later (ANOVA). (D) Allatectomy does not appear to influence the time to pupariation from critical weight in starved larvae, but does shorten the TTP_CW in fed larvae (permutation test). Error bars are 95% confidence intervals. Sample sizes: permutation tests, n = 73 (CAX) and 97 (control); ANOVAs, n = 150 (CAX) and 390 (control).

Fig. 2.

Allatectomized (CAX) larvae grow more slowly than controls. Growth in CAX larvae is significantly slower than in control larvae (ANCOVA_genotype*age, P < 0.001). CAX larvae are significantly smaller than controls at ecdysis to the third instar (t test, P < 0.001). The addition of pyriproxyfen, a JH mimic, to the food increases growth rate in CAX larvae (ANCOVA_{treatment*age}, P = 0.0023), although not to the same rate as controls (ANCOVA_{treatment*age}, P = 0.0103). Pyriproxyfen has no effect on the growth rate of control larvae (ANCOVA_{treatment*age}, P = 0.5972). Error bars are 95% confidence intervals and are obscured by the data points in some cases. Lines are from linear regression. EtOH, ethanol. Sample sizes: ANCOVAs, n = 36 (CAX + EtOH), 42 (CAX + JH), 41 (control + EtOH), and 43 (control + JH); t test, n = 16 (CAX) and 14 (control).

Fig. 3.

Loss of JH down-regulates the activity of the IIS pathway and up-regulates ecdysone signaling. (A and B) The expression profile of Inr and 4E-BP throughout larval development in CAX and control larvae (n = 5 biological replicates for each data point). (C and D) Expression of Inr and 4E-BP is significantly up-regulated from 0 to 32 h AL3E in CAX larvae relative to control (ANOVA, P < 0.001 for both; n = 35 for CAX and control), consistent with a systemic decrease in insulin signaling and activation of FOXO. (E) FOXO activity is up-regulated in CAX larvae relative to control (ANOVA, P = 0.02; n = 7 for CAX and 6 for control). (F) Levels of ecdysone are significantly higher from 0 to 32 h AL3E in CAX larvae relative to controls [ANOVA, P < 0.001; five replicates per time point (n = 20) for both CAX and control]. (G) The effect of allatectomy on body size interacts significantly with the presence or absence of FOXO [ANOVA, P < 0.001; n = 28 (CAX), 19 (control), 8 (CAX – FOXO), 13 (control – FOXO)], such that allatectomy reduces final body size in genetically wild-type flies but not in flies mutant for FOXO [Tukey's Honest Significant Difference (HSD) test]; columns with different letters are significantly different at P < 0.05). Error bars are 95% confidence intervals.

Fig. 4.

JH acts on the prothoracic gland to regulate ecdysone synthesis and modify growth. (A) Mutations in the JH receptor Met reduce both mean pupal and adult body size, and these reductions can be rescued by ubiquitous overexpression of Met [ANOVA, P < 0.001; n = 30 (v¹; Act-GAL4), 41 (Met^w3; Act-GAL4), 29 (Met^w3; UAS-Met/Act-GAL4), Tukey HSD; columns with different letters are significantly different at P < 0.01]. (B) There is a significant reduction in body size when Met is knocked-down in the PG [t test, P < 0.001; n = 20 (phm>GFP), 20 (phm>Met.RNAi)], but not when Met is knocked down in the PTTH-producing cells or the central nervous system [t test, P > 0.05 for both; n = 24 (PTTH,II>GFP), 25 (PTTH,II>Met.RNAi), 26 (PTTH, III>GFP), 26 (PTTH,III>Met.RNAi), 22 (elav>GFP), 23 (elav>Met.RNAi)]. **P < 0.001. (C–E) Reducing JH signaling in the PG reduces larval growth rates (ANCOVA_{genotype*time}, P = 0.0113; n = 214), but does not alter development time [t test, P = 0.637, n = 83 (phm>GFP), 136 (phm>Met.RNAi)] or minimal viable weight [nominal logistic regression, P = 0.196; n = 69 (phm>GFP), 70 (phm>Met.RNAi)]. Lines are from linear regression. Error bars are 95% confidence intervals.