Ecdysone acts through cortex glia to regulate sleep in Drosophila

  1. Yongjun Li
  2. Paula Haynes
  3. Shirley L Zhang
  4. Zhifeng Yue
  5. Amita Sehgal

  • Curated by eLife

    eLife logo

    eLife assessment

    Li and colleague report observations that constitute a potentially fundamental advance, pointing to a mechanism by which non-neural cells can influence sleep regulation by neurons, The authors provide evidence in Drosophila showing that ecdysone synthesised outside the brain regulates sleep via ecdysone receptors in cortex glia. It further suggests that steroid signalling in glia can act on sleep through lipid droplet mobilization.

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

Steroid hormones are attractive candidates for transmitting long-range signals to affect behavior. These lipid-soluble molecules derived from dietary cholesterol easily penetrate the brain and act through nuclear hormone receptors (NHRs) that function as transcription factors. To determine the extent to which NHRs affect sleep:wake cycles, we knocked down each of the 18 highly conserved NHRs found in Drosophila adults and report that the ecdysone receptor (EcR) and its direct downstream NHR Eip75B (E75) act in glia to regulate the rhythm and amount of sleep. Given that ecdysone synthesis genes have little to no expression in the fly brain, ecdysone appears to act as a long-distance signal and our data suggest that it enters the brain more at night. Anti-EcR staining localizes to the cortex glia in the brain and functional screening of glial subtypes revealed that EcR functions in adult cortex glia to affect sleep. Cortex glia are implicated in lipid metabolism, which appears to be relevant for actions of ecdysone as ecdysone treatment mobilizes lipid droplets (LDs), and knockdown of glial EcR results in more LDs. In addition, sleep-promoting effects of exogenous ecdysone are diminished in lsd-2 mutant flies, which are lean and deficient in lipid accumulation. We propose that ecdysone is a systemic secreted factor that modulates sleep by stimulating lipid metabolism in cortex glia.

Article activity feed

  1. Version published to 10.7554/elife.81723 on eLife
  2. eLife

    Author Response

    Reviewer #2 (Public Review):

    This is an interesting manuscript establishing a role for Ecdysone signaling in the control of sleep. The authors show that the Ecdysone receptor EcR is required primarily in cortex glia for the control of sleep and that its target E75 is also involved in sleep regulation. This is a novel function for both cortex glia and steroid signaling in Drosophila. The authors also present evidence that Ecdysone signaling would be important for response to starvation, and that lipid droplet mobilization would mediate the effect of ecdysone on sleep. This work is certainly innovative. However, the main conclusions need to be strengthened. In particular: variability in sleep amounts in certain strains could complicate interpretation, the idea that ecdysone modulates sleep response to starvation is not sufficiently well supported, and genetic evidence for mobilization of lipid droplets being the mechanism linking steroid signaling to sleep is currently quite weak.

    Major concerns:

    1. I have concerns with the variability observed with the GS drivers (whether nSyb or repo). This is particularly striking in figure S3 when comparing experiments conducted with EcR-c and the Ecl RNAi. Daytime is most affected, but even nighttime looks significantly different. Definitely, nighttime quantification should be shown in addition to total sleep in figure S3. However, I feel that confirming the key results of this study with an additional driver would be reassuring. Could repo-GAL4 combined with GAL80ts be used to drive EcR RNAi, instead of repo-GS? The same combination could help determine whether glia is responsible for the 20E-mediated increase in sleep after starvation (figure S4A).

    We have updated the old Figure S3 source data (now Figure 2 - source data 5) with both daytime and nighttime sleep and the conclusion is similar, please also see our response to essential revision question 1. Regarding the GAL80ts experiment, as noted in our detailed response to essential revision question 1, we conducted this experiment and confirmed that adult-specific knockdown of EcR in glia affects sleep. We also tried to do this experiment under starvation conditions (Figure 3 – figure supplement 1A), but this is more challenging to conduct and interpret as it requires temperature shifts, ecdysone treatment and starvation. In particular, high temperature coupled with starvation turned to be an extreme stressor for Repo-Gal4; TublinGal80ts>EcR RNAi #1 flies, as 8 of 12 flies died after 1 day in our first run; thus, we did not proceed with this experiment.

    1. The idea that ecdysone might suppress the response to starvation is interesting, but the results are not convincing. First, there is an important control missing. It is important to test the effect of Ecdysone on fed flies, to ensure that Ecdysone does not simply make flies sleepy. Second, it is not clear that EcR RNAi has a specific effect on starved flies. Starvation reduces sleep, but is this reduction really exaggerated in flies expressing EcR RNAi than in control flies? It seems to me that starvation reduces sleep by the same amount when comparing results in panels 3D and E. The effect of EcRNAi and starvation might be simply additive, which would suggest that 20E impacts sleep independently of starvation.

    We now show effects of exogenous ecdysone on fed flies. As expected, and previously, shown, ecdysone promotes sleep in fed and starved flies (Figures 3 and 6). We agree with the reviewer that 20E impacts sleep independently of starvation. The major point we made with this experiment was that robust effects of starvation on sleep are maintained in RepoGS-EcR RNAi flies. The fact that these two manipulations together virtually eliminate sleep suggests that glial ecdysone signaling is required for the sleep that remains during starvation.

    1. The material and method section needs to be improved. In particular, it is not clear to me how the starvation/ecdysone feeding assay was done. There are some additional explanations in the figure legend, but the approach is still not clear to me. Indicate clearly when the flies were starved, and when they were exposed to Ecdysone.

    We rewrote the ecdysone treatment and starvation assay section with more details in Methods. We hope it is now clear.

    1. I am not convinced that the Lsd2 results necessarily support the idea that this gene is required for the effect of 20E on sleep. Sleep is dramatically reduced during the day in the Lsd2 mutant. This is actually an interesting observation, but this strong effect on baseline sleep might be masking the ability of 20E to modulate sleep.

    Thanks so much for this great comment. As noted in our response to essential revision question 4, we now demonstrate that lsd2 mutants respond effectively to GABA, showing that their sleep can be modulated.

  3. eLife

    eLife assessment

    Li and colleague report observations that constitute a potentially fundamental advance, pointing to a mechanism by which non-neural cells can influence sleep regulation by neurons, The authors provide evidence in Drosophila showing that ecdysone synthesised outside the brain regulates sleep via ecdysone receptors in cortex glia. It further suggests that steroid signalling in glia can act on sleep through lipid droplet mobilization.

  4. eLife

    Reviewer #1 (Public Review):

    In this manuscript, Li et al identify sleep and circadian regulatory role for ecdysone signaling in cortex glia. Prior to this report, numerous studies have linked ecdysone to sleep regulation, though these have primarily focused on its function in neurons. The manuscript is of high interest for a number of reasons. First, it provides a systematic analysis of how NHRs impact sleep. Second, the identification of ecdysone as a critical regulator of both sleep and circadian neurons provides new avenues to study how glia regulate sleep. Finally, the link to lipid accumulation is interesting, but perhaps preliminary. The manuscript is well written, and the data are clearly presented with appropriate controls. Overall this is an exciting manuscript that opens up new directions for the field.

  5. eLife

    Reviewer #2 (Public Review):

    This is an interesting manuscript establishing a role for Ecdysone signaling in the control of sleep. The authors show that the Ecdysone receptor EcR is required primarily in cortex glia for the control of sleep and that its target E75 is also involved in sleep regulation. This is a novel function for both cortex glia and steroid signaling in Drosophila. The authors also present evidence that Ecdysone signaling would be important for response to starvation, and that lipid droplet mobilization would mediate the effect of ecdysone on sleep. This work is certainly innovative. However, the main conclusions need to be strengthened. In particular: variability in sleep amounts in certain strains could complicate interpretation, the idea that ecdysone modulates sleep response to starvation is not sufficiently well supported, and genetic evidence for mobilization of lipid droplets being the mechanism linking steroid signaling to sleep is currently quite weak.

    Major concerns:

    1. I have concerns with the variability observed with the GS drivers (whether nSyb or repo). This is particularly striking in figure S3 when comparing experiments conducted with EcR-c and the Ecl RNAi. Daytime is most affected, but even nighttime looks significantly different. Definitely, nighttime quantification should be shown in addition to total sleep in figure S3. However, I feel that confirming the key results of this study with an additional driver would be reassuring. Could repo-GAL4 combined with GAL80ts be used to drive EcR RNAi, instead of repo-GS? The same combination could help determine whether glia is responsible for the 20E-mediated increase in sleep after starvation (figure S4A).
    2. The idea that ecdysone might suppress the response to starvation is interesting, but the results are not convincing. First, there is an important control missing. It is important to test the effect of Ecdysone on fed flies, to ensure that Ecdysone does not simply make flies sleepy. Second, it is not clear that EcR RNAi has a specific effect on starved flies. Starvation reduces sleep, but is this reduction really exaggerated in flies expressing EcR RNAi than in control flies? It seems to me that starvation reduces sleep by the same amount when comparing results in panels 3D and E. The effect of EcRNAi and starvation might be simply additive, which would suggest that 20E impacts sleep independently of starvation.
    3. The material and method section needs to be improved. In particular, it is not clear to me how the starvation/ecdysone feeding assay was done. There are some additional explanations in the figure legend, but the approach is still not clear to me. Indicate clearly when the flies were starved, and when they were exposed to Ecdysone.
    4. I am not convinced that the Lsd2 results necessarily support the idea that this gene is required for the effect of 20E on sleep. Sleep is dramatically reduced during the day in the Lsd2 mutant. This is actually an interesting observation, but this strong effect on baseline sleep might be masking the ability of 20E to modulate sleep.
  6. Version published to 10.1101/2022.08.24.505204v1 on bioRxiv