Ecdysone-mediated intestinal growth contributes to microbiota-driven developmental plasticity under malnutrition

  1. Longwei Bai
  2. Stéphanie Bellemin
  3. Elodie Guillemot
  4. Maura Strigini
  5. Benjamin Gillet
  6. Cathy Isaura Ramos
  7. François Leulier

Reviewed by

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

Log in to save this article

Abstract

Organ and systemic growth must remain coordinated during development, even under nutritional stress. In Drosophila larvae, the intestinal microbiota contributes to this coordination by promoting growth and maturation under chronic undernutrition. Using gnotobiotic models, we show that association with Lactiplantibacillus plantarum ( Lp ) selectively enhances midgut growth relatively to other organs, providing an adaptive mechanism that buffers the impact of dietary restriction. Transcriptomic profiling of larval midguts revealed a strong Ecdysone signaling signature upon Lp association. Functional analyses showed that local conversion of Ecdysone to its active form, 20-hydroxyecdysone, by the cytochrome P450 enzyme Shade, together with enterocyte Ecd receptor activity, is required for Lp -dependent intestinal and systemic growth. Pharmacological activation of Ecd signaling partially mimicked the bacterial effect, confirming its sufficiency to drive adaptive midgut expansion. Our results uncover an unexpected role of intestinal Ecd signaling in microbiota-driven developmental plasticity, revealing how commensal bacteria modulate local steroid signaling to fine-tune organismal growth and maturation.

Article activity feed

  1. Review Commons

    Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    We would like to warmly thank all the reviewers for their helpful and fair comments which will increase the quality of our manuscript.

    We would like to inform the reviewers that changes have been made concerning the Figures numbers as follows :

    Figure number in old version

    Figure number in revised manuscript

    1B

    S1C

    S1C

    S1D

    1C

    S2A

    S1D

    S2B

    S1E

    S2C

    1D

    1B

    S2

    S3

    S3

    S4

    S4

    S5

    1. Description of the planned revision

    Reviewer #1

    Major comments

    1. Upon food supplementation with 20E the authors could not measure a significant effect on systemic growth or midgut maturation (Fig. S3), whereas the dose of 20E they fed (20µg/ml) was already much higher than endogenous 20E level they measured in the midgut (Fig. 2B).

    We thank the reviewer #1 for this comment.

    Fig. S3 is now Fig. S4

    First, the concentration of 20µg/mL is the final concentration in the fly food and is different from the levels of 20HE we measured in the organs and in the haemolymph, due to the different cell absorption and degradation of the product.

    This concentration of 20µg/mL corresponds to a molar concentration of approximately 0.04mM which is less than the common concentration of 20HE used in the literature in the food (1mM).

    Tiffany V. Roach, Kari F. Lenhart; Mating-induced Ecdysone in the testis disrupts soma-germline contacts and stem cell cytokinesis. Development 1 June 2024; 151 (11): dev202542. doi: https://doi.org/10.1242/dev.202542

    Ahmed, S.M.H., Maldera, J.A., Krunic, D. et al. Fitness trade-offs incurred by ovary-to-gut steroid signalling in Drosophila. Nature 584, 415-419 (2020). https://doi.org/10.1038/s41586-020-2462-y

    The authors should consider to feed larvae with RH5849 (Dr. Ehrenstorfer), which is an insecticide functioning as an ecdysone agonist and was designed for high stability (Wing et al, 1988). RH5849 was already successfully fed to adult Drosophila to investigate the impact of Ecdysone signalling on the adult midgut (Neophytou et al, 2023; Zipper et al, 2025; Zipper et al, 2020) and elicits 20E response. Furthermore, uptake of RH5849 is not limited by the levels of EcI.

    We thank the reviewer #1 for this comment. We ordered that compound and the experiment should be performed in July since the sending date is expected in late June.

    1. The authors should include a discussion of how Ecdysone signalling in postmitotic EC is regulating midgut size, which may include recent data from Edgar and Reiff labs (Ahmed et al, 2020; Zipper et al., 2025; Zipper et al., 2020).

    We thank the reviewer #1 for this comment. We would like to target a format of report for the journal, thus there are some constraints about the number of words. Of course, if the editor allows us to bypass that limit, we would be delighted to cite and discuss these papers.

    1. There are several recent publications showing a role for gut microbiota in regulating oestrogen metabolism in humans, and implications in oestrogen-related diseases such as endometriosis (Baker et al, 2017; Xholli et al, 2023). More precisely bacteria including Lactobacilli strains produce gut microbial β-glucuronidase enzymes, which reactivate oestrogens (Ervin et al, 2019; Hu et al, 2023). As Drosophila ecdysone is the functional equivalent of mammalian oestrogens (Aranda & Pascual, 2001; Martinez et al, 1991; Oberdörster et al, 2001) these publications should be discussed by the authors.

    We thank the reviewer #1 for this comment. We would like to target a format of report for the journal, thus there are some constraints about the number of words. Also, the topics of these papers seem a little bit out of the scope of our manuscript which is focused on the microbiota impact on midgut growth.

    Reviewer #2 Minor Comments

    Figure S2: columns A and B are box plots, while columns C and D are columns with error bars. Presentation of quantitative data should be uniform and ideally as box plots throughout.

    The authors thank the reviewer #2 for this advice and the figure will be further revised.

    Fig. S2 is now Fig. S3

    __Reviewer #3 __

    Major comments:

    The study relies on loss-of-function experiments to manipulate ecdysone signaling; gain-of-function experiments would provide an informative complement. Does feeding ecdysone phenocopy Lp association in GF larvae? Would ecdysone feeding have an additive effect with Lp association? Given the pleiotropic effects of ecdysone on larval phenotypes, a more targeted approach could be used to overexpress transgenes to augment ecdysone signaling.

    We thank the reviewer #3 for this comment. This thought is shared with reviewer #1 and this experiment will be repeated with RH5849. The results are expected in July.

    Minor comments:

    For gut and carcass length analysis, the EcR-RNAi and shd-RNAi conditions look slightly smaller in both GF and Lp conditions. Is there a genetic background effect on larval size? It would be helpful to calculate the interaction score between genotype and microbiome status via a 2-way ANOVA with post hoc tests.

    The authors thank the reviewer #3 for this comment. We will further analyse statistically that differences.

    1. In Fig. 3 the authors added the values for numbers of biological replica within the graphs. In Fig. 4 M-P they added the values for number of technical replicas. They should apply adding these two types of values to all graphs and I would suggest to make the difference between biological replica 'n' and technical replica 'N' obvious in the figure.

    The authors thank the reviewer #3 for this comment. We will modify these numbers in the Figures and/or we will clarify these numbers in the legends to not overwrite the Figures.

    The scope of the bibliography seems limited in scope. As one example, Shin et al., 2011 seems quite relevant for this study.

    We thank the reviewer #1 for this comment. We would like to target a format of report for the journal, thus there are some constraints about the number of words. Of course, if the editor allows us to bypass that limit we would be delighted to cite and discuss this paper.

    2. Description of the revisions that have already been incorporated in the transferred manuscript

    All changes are visible in red in the text of the revised manuscript.

    __Reviewer #1 __

    __Major remarks __

    1. In Fig.2 E - G there is a remarkable difference between controls in D compared to F and E compared to G. The difference between the controls in E and G is stronger than the shown significant difference of EcRRNAi to the control in E. How do the authors explain such a difference of the two (basically equal) controls and the high variance in control values shown in G?

    We thank the reviewer #1 for this comment. As mentioned in the material and methods, the controls are different due to the different RNAi construct. Thus, this can generate variability in such type of developmental experiment.

    Line 253: "UAS-EcRRNAi (BDSC 9327), UAS-dsmCherryRNAI (BDSC 35785), UAS-shadeRNAi (VDRC 108911), and respective RNAi control lines (KK60101)."

    Are the comparisons of control and EcRRNAi shown in D significantly different?

    As mentioned in the figure panel, the EcRRNAi GF and control GF are significantly different and this is discussed in the text as follows in Line 154: "This phenomenon could be explained by genetic background and/or by additional deleterious effect of germ-freeness, as well as a putative contribution of EcR to intestinal functions that are important for systemic growth independently of the contributions of microbiota to adaptive growth."

    1. Lines 167-169: the authors state that 'Size-matched Lp associated larvae, controlRNAi or EcRRNAi, show longer midguts than their relative GF condition (Fig. 3A, B)', but there are no significant statistics shown for this comparison in Fig. 3A, B.

    We thank the reviewer #1 for this comment and we agree that the sentence can be misleading. Thus, we reformulated it as : "Size-matched Lp-associated EcRRNAi larvae show longer midguts than their relative GF controls (Fig. 3A, B)."

    1. Fig. S4 is not mentioned at all in the manuscript.

    We thank the reviewer #1 for this comment and we added the reference to the supplementary Figure 4, now Figure S5 on Line 202 : "In the anterior part, the cells and nuclei are bigger in Lp-associated than GF animals (Fig. 4M-N, Fig.S5). For the posterior part, the cell area was significantly increased in *Lp- *monoassociated animals compared to GF cell while no change was shown for the nucleus area (Fig. 4O-P, Fig.S5)."

    Minor comments: • The authors are inconsistent in indicating their experimental groups. One example is Fig. S3: In A and B they write the GF groups non-italic, whereas the L.p. groups are written italic. In C - E they only partially write the L.p. groups italic. Furthermore, in A, C - E they write 'L.p.', whereas its written 'Lp' and missing the 'WJL' in B.

    We thank the reviewer #1 for this comment and we corrected that mistake in Fig. S3.

    Fig. S3 is now Fig. S4

    Line 52: The last 'i' in 'Lactobacilli' is not italic.

    We thank the reviewer #1 for this comment and we corrected that mistake. • Line 122: Spelling error in 'Surpringsinly'

    We thank the reviewer #1 for this comment and we corrected that mistake. • Line 151: Spelling error in 'progenies'. Needs to read 'progeny'.

    We thank the reviewer #1 for this comment and we corrected that mistake. • Lines 231-235: Last part of the sentence is repetitive

    We thank the reviewer #1 for this comment and we corrected that mistake as "Our work paves the way to deciphering the signals delivered by the bacteria that are sensed at the host cellular level and to understand how this microbe-mediated Ecd-dependent midgut growth contributes to the Drosophila larval growth upon malnutrition."

    Reviewer #2 Minor Comments

    1. Figure 1 is interesting but challenging to follow. The fonts are very small and challenging to read. Pink on blue background is particularly hard to read and doesn't seem necessary. As the entire manuscript follows from data in Figure 1, I would encourage the authors to revise it with a vie3w to making the results more accessible.

    The authors thank the reviewer #2 for this advice and the Figure 1 has been revised.

    Figure 4 is impressive and important for the overall manuscript. The authors should provide representative images to show how they measured cell area and nucleus area.

    The authors thank the reviewer #2.

    How cell area and nucleus area were measured is described in Figure S4. The reference to this supplementary Figure was missing in the initial manuscript and we deeply apologize for that.

    Reviewer #1 also pointed out that the reference of Figure S4 covering that point was missing in the text and we corrected that point.

    I struggled to follow this sentence (line 215): "Also, it will be interesting to test, beyond their shared growth phenotype, whether they respond differently at the mechanistical level to the presence of bacteria in the anterior compartment." I would encourage the authors to consider alternative formulations.

    The authors thank the reviewer #2 and revised that sentence as follows :

    "Also, it will be interesting to investigate whether the midgut comprises sub-populations of enterocytes that differ in their physiological functions. Indeed, these sub-populations could be differently distributed along the midgut and be localized on anterior and/or posterior parts. Thus, they could present varied responses to the presence of the bacteria."

    __Reviewer #3 __

    Major comments

    Figure 4 title is misleading. No manipulations of ecdysone signaling are performed to demonstrate whether scaling relationships across tissues differ depending on ecdysone. The same experiment should be performed using mex>EcR-RNAi larvae and/or mex>shd-RNAi larvae.

    We thank the reviewer #3 for this comment.

    We agree with the reviewer and the title has been changed as follows and mentioned in red in the manuscript : Midgut-specific adaptive growth promoted by Lp in Drosophila larvae.

    Minor comments:

    It is notable that mex>EcR-RNAi in germ-free larvae exacerbates developmental delay. A possible interpretation is that ecdysone signaling in the germ-free context promotes increased growth rate. Could the authors comment?

    We thank reviewer #3 for this comment.

    Since we described a local effect at the intestine level for Ecd it is unlikely but not totally excluded that intestinal Ecd promotes systemic growth.

    Our comments are here in the text :

    "This phenomenon could be explained by genetic background and/or by additional deleterious effect of germ-freeness, as well as a putative contribution of EcR to intestinal functions that are important for systemic growth independently of the contributions of microbiota to adaptive growth."

    Experimental variation is substantial between the control conditions of the EcR and Shd knockdown experiments; median control + Lp D50 in the EcR experiment is ~6 days whereas in the shade experiment it is ~9 days. Can the authors comment on this between-experiment variation, which seems substantial (similar to the effect size between control + Lp and control GF)?

    We thank reviewer #3 for this comment which was also highlighted by the reviewer #1 and we answered as follows :

    As mentioned in the material in methods, the controls are different due to the different RNAi construct. Thus, this can generate variability in such type of developmental experiment.

    Line 253: "UAS-EcRRNAi (BDSC 9327), UAS-dsmCherryRNAI (BDSC 35785), UAS-shadeRNAi (VDRC 108911), and respective RNAi control lines (KK60101)."

    As mentioned in the figure panel, the EcRRNAi GF and control GF are significantly different and this is discussed in the text as follows in Line 154: "This phenomenon could be explained by genetic background and/or by additional deleterious effect of germ-freeness, as well as a putative contribution of EcR to intestinal functions that are important for systemic growth independently of the contributions of microbiota to adaptive growth."

    The methods detail an ecdysone feeding protocol that I could not find used in the experiments. Please clarify.

    We thank reviewer #3 for this comment.

    We would like to highlight that this protocol is related to an experiment described in Fig. S3 (now Fig.S4) and that supplementary Figure was cited here in the text of the manuscript Line 179 as follows "While the systemic growth of animals is not affected by addition of 20E, a slight trend to faster midgut maturation of GF larvae is observed through the measurements of longer guts (Fig. S4)."

    Also, in supplementary data :

    Fig. S3 : Feeding larvae with 20E does not impact the gut growth.

    (A-B) Addition of 20E has no impact on larval developmental timing (DT) and their D50. From size-matched animals (C), Lp promotes intestinal growth compare to GF (D) but no significant difference is shown in the gut/carcass ratio (E). Animals receiving 20E are represented with color filled circles +Lp (blue), GF (black) and controls without 20E supplementation with empty circles.

    The manuscript would benefit from additional proofreading. The text contains spelling errors throughout. The in-text reference formatting is inconsistent. Figure legends could be improved to better describe the data.

    We thank reviewer #3 for this comment and following the different reviewers comments we improved the manuscript in that way.

    3. Description of analyses that authors prefer not to carry out

    Reviewer #1

    __Major remarks __

    1. The authors should consider investigating an EcIRNAi in addition to EcRRNAi. EcR functions as activator, but also as suppressor in the absence of Ecdysone and a EcRRNAi suppresses both functions of EcR. By knocking down EcI the authors would prevent uptake of Ecdysone and thus interfere only with the ligand-induced activating function of EcR.

    We thank reviewer #1 for this comment.

    This experiment has been performed using EcI RNAi but not shown here because in our hands the genetic tool was not efficient (RNA interference does not work effectively) and thus the experiment was not conclusive.

    No phenotype was observed in our study (see Figure attached). Also, the others Oatp family members were tested for their expression in midgut and were found close to null expression.

    1. Why are the authors comparing the carcass length of GF shade RNAi with L.p. control in Fig. 3 D?

    We thank reviewer #1 for this comment. For transparency of the results, these statistics were added. Because in these conditions GF larvae were difficult to rise at the same size than their relative *Lp *monoassociated. Hence, the carcass length was used to normalize the data.

    1. In Fig. S3C the authors compared L.p. WJL 20E with the GF EtOH control, where is the comparison to the corresponding L.p. WJL EtOH control? The L.p. WJL EtOH control is compared to GF 20E instead.

    We thank reviewer #1 for this comment that will help to clarify our experiment.

    Fig. S3 is now Fig. S4

    For the Fig. S4C, it is a larval size that allows to compare sizes in all conditions independently. That explains that statistics are shown between all conditions. To not overload the Figure the p values not different are not mentioned.

    Reviewer #2 Minor Comments

    1. Figure S3 confuses me. It seems that addition of 20E to GF larvae leads to a significant reduction of larval size, and that mono-association with Lp also significantly shortens larval size. Data in Figure 4G suggest that Lp should not affect larval body length relative to GF larvae. Can the authors explain the apparent discrepancy?

    The authors thank the reviewer #2 for this question. Fig. S3 is now Fig. S4.

    This difference could be explained as follows :

    • The developmental experiment in Fig. S3B shows no difference between the two GF conditions. Thus, at the end of the is larval development, systemic growth is similar in both conditions.

    Because performed earlier during development, the larval size experiment shows higher variability in measurements of larval size. Moreover, less larvae are present in the GF 20E condition that could explained that difference.

    • We have previously shown that Lp mono-associated larvae grow faster than GF. Thus, to collect size-matched larvae on the same day, GF or Lp animals come from a different initial day of experiment. Due to biological variability, some differences in timing could be observed between GF and Lp animals.

    Reviewer #3

    Major comments

    The authors conclude that intestinal ecdysone signals are not required for Lp-promoted systemic growth. However, their data shows that circulating 20E titer increases in an Lp-dependent manner, and this circulating 20E presumably affects multiple tissues throughout the organism. Since EcR is broadly expressed, can the authors examine how EcR knockdown in other tissues influences systemic growth in Lp-associated larvae? Fat body-specific EcR knockdown seems particularly of interest here given the established relationship between fat body ecdysone signaling and growth (Delanoue et al., 2010). This additional analysis would help clarify whether ecdysone signaling in non-intestinal tissues mediates the Lp-associated growth phenotype.

    We thank reviewer #3 for this comment that will help to clarify our manuscript.

    We would like to emphasize that we never mention in this manuscript that intestinal ecdysone signals are not required for systemic growth. Nevertheless, we highlighted that it is required for Lp-related midgut growth and not rate limiting for Lp-promoted systemic growth:

    Line 179 : "While the systemic growth of animals is not affected by addition of 20E, a slight trend to faster midgut maturation of GF larvae is observed through the measurements of longer guts (Fig. S3). Thus, the intestinal Ecd signaling is required for the midgut growth effect mediated by Lp in a context of malnutrition."

    Line 227: "Specifically, intestinal Ecd signaling is not rate-limiting for Lp-mediated adaptive growth."

    While it will be very interesting to study the effects of Ecd modulation from Fat Body, we feel this is out of the scope of our manuscript that focused on the Lp-based intestinal growth.

    The experimental design compares larvae associated with live Lp versus germ-free larvae provided sterile PBS. Since Lp cells constitute a potential nutrient source for developing larvae, it's unclear whether gene expression differences arise from larvae digesting Lp cells as a nutrient source or from active, microbe-host signaling interactions. To conclusively address this ambiguity, the authors should perform RNA-seq on larvae inoculated with live versus heat-killed Lp. Alternatively, qPCR could be used to provide evidence for the extent to which changes in ecdysone-related gene expression specifically require live Lp.

    We thank reviewer #3 for this comment.

    We (the lab) previously showed that the systemic growth phenotype is supported by bacteria during development and that bacteria have to be alive to support optimal effects (Storelli et al 2018, PMID: 29290388; Consuegra et al 2020a, PMID: 32196485; Consuegra et al 2020b, PMID: 32563155). This topic of bacteria viability has also been directly addressed independently by colleagues and reported recently (da Silva Soares NF, PMID: 37488173). Hence, we did not design our RNAseq with inactivated bacteria. However, if the editor believes this is essential to provide qPCR results on Ecd-related gene expression in live vs inactivated bacteria associations, we shall provide them but at this stage we believe this notion is not core to our message.

    Shade is expressed in the larval midgut, however the larval fat body is thought to be a major site of 20E to 20HE conversion. Can the authors test how Shd knockdown in the fat body affects systemic growth in the Lp-associated condition?

    We thank reviewer #3 for this comment. Nevertheless, we think this is out of the scope of our manuscript that focused on the Lp-based intestinal growth.

    In the knockdown experiments, body size is not measured for larvae/pupae. Given that ecdysone signaling impacts pupal volume (Delanoue et al., 2010) and controls metamorphosis timing, D50 plots by pupal volume would be informative to give a rough estimate of growth rate. For example, do germ-free EcR-RNAi larvae, which develop slower, have an equivalent body size to germ-free control larvae?

    We thank the reviewer #3 for this comment.

    All experiments were done with size-matched larvae because the aim of this manuscript is to detail what is the impact of Lp on the relative midgut vs systemic growth. Hence, we are using animals of similar systemic size to study their midgut size and identify allometry changes (midgut/larval size ratios) at a similar developmental point, which is same larval systemic growth (here L3). Thus, we feel that focusing on growth rates and systemic sizes in different genetic conditions, while interesting in general, is out of the scope of the study since we focus our study on midgut/larval size allometry.

    __Minor comments __

    The number of pupae in the EcR-RNAi and shd-RNAi experiments (Fig 2D, F) differ. Were larval densities controlled during development?

    I could not find this mentioned in the methods, and it is an important control parameter as larval density impacts developmental growth. Presenting this data as % viability of a known number of larvae deposited in food would be preferable.

    We thank the reviewer #3 for this comment.

    As mentioned in the material and methods, 40 eggs from axenic animals were deposited on each tube. It is true that the final number of pupae is different and could come from differential viability of the genetic backgrounds used. It would be difficult to follow from the same tube the larval development because of the manipulation of gnotobiotics animals. Nevertheless, in all experiments more than 25% of initial eggs deposited in tubes emerged as adults.

  2. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #3

    Evidence, reproducibility and clarity

    Bai and colleagues present a detailed characterization of the larval midgut transcriptome, comparing size-matched germ-free and Lactobacillus plantarum (Lp)-associated L3 larvae. The authors identify Lp-mediated changes in ecdysone signaling genes, show increased ecdysone titer in Lp-associated larval hemolymph, and explore the functional role of intestinal ecdysone signaling and Lp in gut-specific vs systemic growth. This work offers insights into microbiome-intestinal EcR signaling that will attract broad interest in the Drosophila community.

    Major comments:

    1. The authors conclude that intestinal ecdysone signals are not required for Lp-promoted systemic growth. However, their data shows that circulating 20E titer increases in an Lp-dependent manner, and this circulating 20E presumably affects multiple tissues throughout the organism. Since EcR is broadly expressed, can the authors examine how EcR knockdown in other tissues influences systemic growth in Lp-associated larvae? Fat body-specific EcR knockdown seems particularly of interest here given the established relationship between fat body ecdysone signaling and growth (Delanoue et al., 2010). This additional analysis would help clarify whether ecdysone signaling in non-intestinal tissues mediates the Lp-associated growth phenotype.
    2. The experimental design compares larvae associated with live Lp versus germ-free larvae provided sterile PBS. Since Lp cells constitute a potential nutrient source for developing larvae, it's unclear whether gene expression differences arise from larvae digesting Lp cells as a nutrient source or from active, microbe-host signaling interactions. To conclusively address this ambiguity, the authors should perform RNA-seq on larvae inoculated with live versus heat-killed Lp. Alternatively, qPCR could be used to provide evidence for the extent to which changes in ecdysone-related gene expression specifically require live Lp.
    3. Shade is expressed in the larval midgut, however the larval fat body is thought to be a major site of 20E to 20HE conversion. Can the authors test how Shd knockdown in the fat body affects systemic growth in the Lp-associated condition?
    4. In the knockdown experiments, body size is not measured for larvae/pupae. Given that ecdysone signaling impacts pupal volume (Delanoue et al., 2010) and controls metamorphosis timing, D50 plots by pupal volume would be informative to give a rough estimate of growth rate. For example, do germ-free EcR-RNAi larvae, which develop slower, have an equivalent body size to germ-free control larvae?
    5. Figure 4 title is misleading. No manipulations of ecdysone signaling are performed to demonstrate whether scaling relationships across tissues differ depending on ecdysone. The same experiment should be performed using mex>EcR-RNAi larvae and/or mex>shd-RNAi larvae.
    6. The study relies on loss-of-function experiments to manipulate ecdysone signaling; gain-of-function experiments would provide an informative complement. Does feeding ecdysone phenocopy Lp association in GF larvae? Would ecdysone feeding have an additive effect with Lp association? Given the pleiotropic effects of ecdysone on larval phenotypes, a more targeted approach could be used to overexpress transgenes to augment ecdysone signaling.

    Minor comments:

    1. For gut and carcass length analysis, the EcR-RNAi and shd-RNAi conditions look slightly smaller in both GF and Lp conditions. Is there a genetic background effect on larval size? It would be helpful to calculate the interaction score between genotype and microbiome status via a 2-way ANOVA with post hoc tests.
    2. It is notable that mex>EcR-RNAi in germ-free larvae exacerbates developmental delay. A possible interpretation is that ecdysone signaling in the germ-free context promotes increased growth rate. Could the authors comment?
    3. The number of pupae in the EcR-RNAi and shd-RNAi experiments (Fig 2D, F) differ. Were larval densities controlled during development? I could not find this mentioned in the methods, and it is an important control parameter as larval density impacts developmental growth. Presenting this data as % viability of a known number of larvae deposited in food would be preferable.
    4. Experimental variation is substantial between the control conditions of the EcR and Shd knockdown experiments; median control + Lp D50 in the EcR experiment is ~6 days whereas in the shade experiment it is ~9 days. Can the authors comment on this between-experiment variation, which seems substantial (similar to the effect size between control + Lp and control GF)?
    5. The methods detail an ecdysone feeding protocol that I could not find used in the experiments. Please clarify.
    6. Figure S4 is not called out in the text.
    7. The scope of the bibliography seems limited in scope. As one example, Shin et al., 2011 seems quite relevant for this study.
    8. The manuscript would benefit from additional proofreading. The text contains spelling errors throughout. The in-text reference formatting is inconsistent. Figure legends could be improved to better describe the data.

    Significance

    The authors identify Lp-mediated changes in ecdysone signaling genes, show increased ecdysone titer in Lp-associated larval hemolymph, and explore the functional role of intestinal ecdysone signaling and Lp in gut-specific vs systemic growth. This work offers insights into microbiome-intestinal EcR signaling that will attract broad interest in the Drosophila community.

  3. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    Summary

    Gut-resident bacteria promote growth and development in the host intestine. With remarkable genetic accessibility and broad physiological relevance, flies are an excellent experimental system to ask how intestinal bacteria modify growth. Previous studies established a role for Lactobacillus planetarium in larval fly growth, particularly during periods of nutrient deprivation. Growth is principally supported by signals from insulin-like peptides and the melting hormone ecdysone. In this manuscript, the authors present transcriptomic data that Lp association has large effects on transcription in larval hosts raised on low-protein diets, including signatures of ecdysone signalling activation. Through a series of follow up studies the authors present evidence that ecdysone signalling is particularly important for midgut growth in Lp-associated larvae raised on low protein food.

    Major Comments

    I have no major comments. The manuscript is well written, easy to follow, and data are interpreted appropriately. Most of my remarks are minor in nature and can be addressed by simple changes to the text or figures.

    Minor Comments

    1. Figure 1 is interesting but challenging to follow. The fonts are very small and challenging to read. Pink on blue background is particularly hard to read and doesn't seem necessary. As the entire manuscript follows from data in Figure 1, I would encourage the authors to revise it with a vie3w to making the results more accessible.
    2. Figure S2: columns A and B are box plots, while columns C and D are columns with error bars. Presentation of quantitative data should be uniform and ideally as box plots throughout.
    3. Figure S3 confuses me. It seems that addition of 20E to GF larvae leads to a significant reduction of larval size, and that mono-association with Lp also significantly shortens larval size. Data in Figure 4G suggest that Lp should not affect larval body length relative to GF larvae. Can the authors explain the apparent discrepancy?
    4. Figure 4 is impressive and important for the overall manuscript. The authors should provide representative images to show how they measured cell area and nucleus area.
    5. I struggled to follow this sentence (line 215): "Also, it will be interesting to test, beyond their shared growth phenotype, whether they respond differently at the mechanistical level to the presence of bacteria in the anterior compartment." I would encourage the authors to consider alternative formulations.

    Significance

    Strengths and limitations

    This is a well written manuscript that adds to our understanding of the effects of intestinal bacteria on host development. Except for labeling in Figure 1, all data are accessibly presented, and accurately interpreted.

    Advance

    The study accurately links gut microbes to endocrinological control of organ growth in the fly. It also advances our understanding of Drosophila intestinal development by showing how ecdysone is important for gut development.

    Audience

    The work is focussed in scope and will likely have greater appeal to groups that work primarily with the Drosophila model. However, the work will likely have a broader appeal to those that study microbial control of development in other animal models.

    My expertise

    This study aligns with my area of expertise.

  4. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    The manuscript entitled 'Drosophila larval gut transcriptome reveals a microbe-mediated intestinal tissue growth via Ecdysone during adaptive growth' identifies endocrine ecdysone signalling to regulate Lactobacilli-mediated adaptive growth during Drosophila larval development in context of malnutrition. The authors performed transcriptomic analysis to pin down genes that are deregulated in Lactiplantibacillus plantarum (L.p.) associated animals compared to germ-free raised animals. Besides the newly highlighted ecdysteroid kinase-like genes the other isolated deregulated candidates will be of interest for the audience specialized on gut microbiomes. Furthermore, by knockdown of the ecdysone receptor and the converting enzyme shade in midgut enterocytes of L.p. associated and GF raised larvae Bai et al. validated a requirement for ecdysone signalling in midgut growth, but not systemic growth of malnourished animals.

    Although the manuscript is mostly well-written and concise, I have major remarks (and suggestions) that the authors need to consider during a major revision process to sustain their claims. In addition, in several occasions the shown graphs do not support the text statements statistically, which I point out in the following major remarks. In addition, an entire Figure is not referred to in the manuscript. Overall, that leaves a bit of a 'premature' impression of the manuscript.

    Major remarks:

    1. In Fig.2 E - G there is a remarkable difference between controls in D compared to F and E compared to G. The difference between the controls in E and G is stronger than the shown significant difference of EcRRNAi to the control in E. How do the authors explain such a difference of the two (basically equal) controls and the high variance in control values shown in G? Are the comparisons of control and EcRRNAi shown in D significantly different?
    2. The authors should consider investigating a EcIRNAi in addition to EcRRNAi. EcR functions as activator, but also as suppressor in the absence of Ecdysone and a EcRRNAi suppresses both functions of EcR. By knocking down EcI the authors would prevent uptake of Ecdysone and thus interfere only with the ligand-induced activating function of EcR.
    3. Upon food supplementation with 20E the authors could not measure a significant effect on systemic growth or midgut maturation (Fig. S3), whereas the dose of 20E they fed (20µg/ml) was already much higher than endogenous 20E level they measured in the midgut (Fig. 2B). The authors should consider to feed larvae with RH5849 (Dr. Ehrenstorfer), which is an insecticide functioning as an ecdysone agonist and was designed for high stability (Wing et al, 1988). RH5849 was already successfully fed to adult Drosophila to investigate the impact of Ecdysone signalling on the adult midgut (Neophytou et al, 2023; Zipper et al, 2025; Zipper et al, 2020) and elicits 20E response. Furthermore, uptake of RH5849 is not limited by the levels of EcI.
    4. Lines 167-169: the authors state that 'Size-matched Lp associated larvae, controlRNAi or EcRRNAi, show longer midguts than their relative GF condition (Fig. 3A, B)', but there are no significant statistics shown for this comparison in Fig. 3A, B.
    5. Why are the authors comparing the carcass length of GF shade RNAi with L.p. control in Fig. 3 D?
    6. In Fig. 3 the authors added the values for numbers of biological replica within the graphs. In Fig. 4 M-P they added the values for number of technical replicas. They should apply adding these two types of values to all graphs and I would suggest to make the difference between biological replica 'n' and technical replica 'N' obvious in the figure.
    7. In Fig. S3C the authors compared L.p. WJL 20E with the GF EtOH control, where is the comparison to the corresponding L.p. WJL EtOH control? The L.p. WJL EtOH control is compared to GF 20E instead.
    8. The authors should include a discussion of how Ecdysone signalling in postmitotic EC is regulating midgut size, which may include recent data from Edgar and Reiff labs (Ahmed et al, 2020; Zipper et al., 2025; Zipper et al., 2020).
    9. There are several recent publications showing a role for gut microbiota in regulating oestrogen metabolism in humans, and implications in oestrogen-related diseases such as endometriosis (Baker et al, 2017; Xholli et al, 2023). More precisely bacteria including Lactobacilli strains produce gut microbial β-glucuronidase enzymes, which reactivate oestrogens (Ervin et al, 2019; Hu et al, 2023). As Drosophila ecdysone is the functional equivalent of mammalian oestrogens (Aranda & Pascual, 2001; Martinez et al, 1991; Oberdörster et al, 2001) these publications should be discussed by the authors.
    10. Fig. S4 is not mentioned at all in the manuscript.

    Minor comments:

    • The authors are inconsistent in indicating their experimental groups. One example is Fig. S3: In A and B they write the GF groups non-italic, whereas the L.p. groups are written italic. In C - E they only partially write the L.p. groups italic. Furthermore, in A, C - E they write 'L.p.', whereas its written 'Lp' and missing the 'WJL' in B.
    • Line 52: The last 'i' in 'Lactobacilli' is not italic.
    • Line 122: Spelling error in 'Surpringsinly'
    • Line 151: Spelling error in 'progenies'. Needs to read 'progeny'.
    • Lines 231-235: Last part of the sentence is repetitive

    References

    Ahmed SMH, Maldera JA, Krunic D, Paiva-Silva GO, Penalva C, Teleman AA, Edgar BA (2020) Fitness trade-offs incurred by ovary-to-gut steroid signalling in Drosophila. Nature 584: 415-419

    Aranda A, Pascual A (2001) Nuclear hormone receptors and gene expression. Physiol Rev 81: 1269-1304

    Baker JM, Al-Nakkash L, Herbst-Kralovetz MM (2017) Estrogen-gut microbiome axis: Physiological and clinical implications. Maturitas 103: 45-53

    Ervin SM, Li H, Lim L, Roberts LR, Liang X, Mani S, Redinbo MR (2019) Gut microbial β-glucuronidases reactivate estrogens as components of the estrobolome that reactivate estrogens. J Biol Chem 294: 18586-18599

    Hu S, Ding Q, Zhang W, Kang M, Ma J, Zhao L (2023) Gut microbial beta-glucuronidase: a vital regulator in female estrogen metabolism. Gut Microbes 15: 2236749

    Martinez E, Givel F, Wahli W (1991) A common ancestor DNA motif for invertebrate and vertebrate hormone response elements. The EMBO journal 10: 263-268

    Neophytou C, Soteriou E, Pitsouli C (2023) The Sterol Transporter Npc2c Controls Intestinal Stem Cell Mitosis and Host-Microbiome Interactions in Drosophila. Metabolites 13

    Oberdörster E, Clay MA, Cottam DM, Wilmot FA, McLachlan JA, Milner MJ (2001) Common phytochemicals are ecdysteroid agonists and antagonists: a possible evolutionary link between vertebrate and invertebrate steroid hormones. J Steroid Biochem Mol Biol 77: 229-238

    Wing KD, Slawecki RA, Carlson GR (1988) RH 5849, a Nonsteroidal Ecdysone Agonist: Effects on Larval Lepidoptera. Science 241: 470-472

    Xholli A, Cremonini F, Perugi I, Londero AP, Cagnacci A (2023) Gut Microbiota and Endometriosis: Exploring the Relationship and Therapeutic Implications. Pharmaceuticals (Basel) 16 Zipper L, Corominas-Murtra B, Reiff T (2025) Steroid hormone-induced wingless ligands tune female intestinal size in Drosophila. Nature Communications 16: 436

    Zipper L, Jassmann D, Burgmer S, Görlich B, Reiff T (2020) Ecdysone steroid hormone remote controls intestinal stem cell fate decisions via the PPARγ-homolog Eip75B in Drosophila. eLife 9

    Significance

    Brief general assessment before a revision: The study provides important new insights into organ versus systemic growth and show that this is regulated by a central steroid hormonal pathway making this study interesting for a broad audience.

  5. Version published to 10.1101/2025.02.07.637066 on bioRxiv