Steroid Hormone Signaling Is Essential for Pheromone Production and Oenocyte Survival

Abstract

Many of the lipids found on the cuticles of insects function as pheromones and communicate information about age, sex, and reproductive status. In Drosophila, the composition of the information-rich lipid profile is dynamic and changes over the lifetime of an individual. However, the molecular basis of this change is not well understood. To identify genes that control cuticular lipid production in Drosophila, we performed a RNA interference screen and used Direct Analysis in Real Time and gas chromatography mass spectrometry to quantify changes in the chemical profiles. Twelve putative genes were identified whereby transcriptional silencing led to significant differences in cuticular lipid production. Amongst them, we characterized a gene which we name spidey, and which encodes a putative steroid dehydrogenase that has sex- and age-dependent effects on viability, pheromone production, and oenocyte survival. Transcriptional silencing or overexpression of spidey during embryonic development results in pupal lethality and significant changes in levels of the ecdysone metabolite 20-hydroxyecdysonic acid and 20-hydroxyecdysone. In contrast, inhibiting gene expression only during adulthood resulted in a striking loss of oenocyte cells and a concomitant reduction of cuticular hydrocarbons, desiccation resistance, and lifespan. Oenocyte loss and cuticular lipid levels were partially rescued by 20-hydroxyecdysone supplementation. Taken together, these results identify a novel regulator of pheromone synthesis and reveal that ecdysteroid signaling is essential for the maintenance of cuticular lipids and oenocytes throughout adulthood.

Fig 1. Schematic of RNA interference screen used to identify genes expressed in the oenocytes that contribute to cuticular lipid production.

(A) Tissue-specific knockdown of the expression of 80 candidate genes was performed using 2 different oenocyte-expressing drivers, oeno-Gal4 and dsx-Gal4. (B) Progeny from each cross were assayed individually in the DART MS source (arrowhead). (C) Chemical profiles obtained from female and male cuticles. The intensity index for molecules of interest (in blue) is calculated by normalizing signal intensity to an internal reference peak (in red: 7,11-HD for females, 7-T for males); 7-T: (Z)-7-tricosene; 7-P: (Z)-7-pentacosene; 7,11-PD: (Z, Z)-7,11-pentacosadiene; 7,11-HD: (Z, Z)-7,11-heptacosadiene; 7,11-ND: (Z, Z)-7,11-nonacosadiene. (D) Transgenic lines displaying intensity indices which are 2 standard deviations (SD) above or below the average value are subjected to a secondary DART MS screen and quantification by GCMS. (E) Chemical structures of the major CHC species characteristic of male or female cuticular profiles.

Fig 2. Spidey expression during development is essential for adult survival.

(A) Silencing spidey expression from 0–72 hours after egg laying (AEL) results in pharate lethality of males and females. Male survivorship significantly improves when spidey is suppressed from 96 hours AEL but few females successfully eclose. Survival % is calculated as the number of successfully eclosed adults relative to the total number of male and female embryos; E: embryo; L1: 1^st instar larvae; L2: 2^nd instar larvae; L3: 3^rd instar larvae; P: pupae. N = 50 embryos for each stage. See S6 Table for full genotypes. (B) Top: oeno>spidey^RNAi flies fail to emerge from the pupal case. Bottom left: malformed legs (arrowheads) are shorter in oeno>spidey^RNAi males and females compared to genetic controls (bottom right). (C) Representative DART MS spectra from oeno>spidey^RNAi males reveals an overall decrease of major cuticular lipid signals (blue, arrows) relative to tricosene (peak 1, red). In spectra from control lines, pentacosene (peak 4) and tricosene (peak 1) exhibit similar relative intensity levels. (D) Continuous overexpression of spidey at 29°C during development results in significant pre-pupal and pharate lethality compared to controls raised at 19°C (p<0.0001 for oeno>spidey and oeno>spidey.HA, compared to controls, Chi-square test, n = 189). (E) Top: oeno>spidey and oeno>spidey.HA overexpression lines exhibit pre-pupal lethality and defects in head (arrowheads) and spiracle (*) eversion. Bottom: controls raised at 19°C. (F) Knockdown or overexpression of spidey at 29°C alters Cyp18a1 and spidey transcript levels relative to control levels measured at 19°C. Rp49 was used as an internal control for normalization. Data represent the average of 3 experimental replicates ± SEM.

Fig 3. Quantification of ecdysteroids and ecdysonic acids metabolites by LC-MS/MS.

(A) Top row: 20HE-oic acid levels were significantly elevated in spidey^KD L3 larvae, resulting in a lower 20HE to 20HE-oic acid ratio. 20HE-oic acid was not detected in L1 and L2 animals. Bottom row: makisterone A (MA) and makisteronoic acid (MA-oic acid) levels were lower in spidey^KD animals compared to spidey^control at mid-pupal stage (P). However, the ratio of MA to MA-oic acid in pupal stage was not significantly different. (B) Following overexpression of spidey, L3 larvae exhibit a significant decrease of 20HE and MA compared to controls. Data represent the average log-transformed values of 3 experimental replicates ± SEM; Student’s t-test.

Fig 4. Spidey suppression leads to compromised fitness, longevity, and “Spider-Man” phenotype.

(A) Spidey^KD males and females (dashed lines) exhibit poorer desiccation resistance compared to genetically identical spidey^control flies (solid lines); N = 50 flies for all treatments, P-values comparing experimental to control flies were obtained by log-rank test. See S6 Table for full genotypes. (B) Spidey^KD males are less resistant to starvation compared to genetically identical control flies. N = 50 flies for all treatments, P-values comparing experimental to control flies were obtained by log-rank test. (C) Spidey^KD males and females exhibit poorer oxidation resistance compared to genetically identical control flies; N = 50 flies for all treatments, P-values comparing experimental to control flies were obtained by log-rank test. (D) Spidey^KD males and females exhibit a shorter lifespan compared to genetically identical control flies; N = 100 flies for all treatments, P-values comparing experimental to control flies were obtained by log-rank test. (E) Flies exhibiting the sticky “Spider-man” phenotype are found dead on the vertical wall of the vial and do not detach despite vigorous shaking and knocking. The sticky phenotype was observed in 50–60% of males with suppressed spidey expression (spidey^KD, 14–19 days old; oeno>spidey^RNAi, 20–27 days old) or genetically ablated oenocytes (oeno>hid, 13–17 days old). Only 1 fly from the age-matched control groups (spidey^control, 47–56 days old; oeno/+, 53–62 days old) exhibited the sticky phenotype; one tailed Fisher’s exact test, ****p<0.0001; sample size is shown above each bar. (F) Representative electron micrographs showing the accumulation of food residue (indicated by arrowheads) on the legs of flies exhibiting the sticky phenotype (spidey^KD, oeno>hid). Control (spidey^control) and wildtype Canton-S flies do not appear to have residue on the legs.

Fig 5. Spidey is essential for cuticular lipid synthesis in adult flies.

(A) Two different feeding regimes of adult flies are used to produce an immediate or delayed knockdown of spidey expression. Gene suppression is achieved by feeding RU-486 to GSoeno>spidey^RNAi flies; suppression is removed by placing flies on standard food. Cuticular lipids are extracted at 8 days old. A 65–85% decrease in CHC levels is apparent when spidey is suppressed using the immediate knockdown regime but not the delayed regime. Data represent the mean intensity of 3 experimental replicates (8 flies each) ± SEM; Student’s t-test, *p<0.05; a.u.: arbitrary units. See S6 Table for full genotypes. (B) Immediate and delayed knockdown of spidey over 15 days. The CHC levels do not recover at 15 days when spidey is suppressed from age 0–2 days. Spidey suppression between age 10–12 days produces an intermediate effect on CHC levels. Data shown represent the mean ± SEM from 3 experimental replicates with 8 flies each); Student’s t-test, *p<0.05. (C) Representative GCMS chromatograms of cuticular lipid extract from 15 day old spidey^control and spidey^KD males. Major peaks are labelled with the number of carbon atoms followed by number of double bonds; T: tricosene; P: pentacosene; STD: C26:0 standard spiked into the hexane solvent.

Fig 6. Spidey is necessary for oenocyte viability in aged adults.

(A) Fluorescent microscope images of GFP-labeled oenocytes in males reveal that cellular morphology and cell numbers change upon silencing of spidey in adult flies. At 8 and 15 days old, oenocytes appear swollen in spidey-suppressed flies compared to controls. At 20 days old, very few GFP-positive cells were observed. See S6 Table for full genotypes. (B) Quantification of GFP intensity in oenocytes reveals a significant increase at 8 and 15 days old in spidey-suppressed flies followed by a dramatic decrease at 20 days old. Sample sizes are indicated above each bar; one-way ANOVA with Tukey’s multiple comparisons test; **: p<0.01, ****: p<0.0001; ns: not significant. (C) Fluorescent microscope images of GFP-labeled oenocytes in spidey^KD females reveal similar morphological changes as observed with males. (D) Quantification of GFP intensity in oenocytes reveals a slight but significant increase at 8 and 15 days old in spidey suppressed females followed by a dramatic loss at 20 days old; samples sizes are indicated above each bar. One-way ANOVA with Tukey’s multiple comparisons test; *: p<0.05, ****: p<0.0001; ns: not significant.

Fig 7. Partial rescue of spidey defects by 20HE.

(A) At 8 days old, 20HE supplementation restores total CHC levels to near control conditions in male but not female spidey^KD flies. Data shown are the mean ± SEM from 3 experimental replicates with 8 flies each; one-way ANOVA with Tukey’s multiple comparison test, *p<0.05, ns: not significant; a.u.: arbitrary units. See S6 Table for full genotypes. (B) At 20 days old, GFP signal intensity of oenocytes in oeno>spidey^RNAi males remains low and aberrantly distributed despite 20HE supplementation. (C) Females exhibit a significant increase in GFP intensity with 20HE feeding. Mean GFP intensity ± SEM shown; samples sizes are indicated above; Student’s t-test, **: p<0.01; ns: not significant.