Decision letter | Metabolite exchange between microbiome members produces compounds that influence Drosophila behavior

Open accessCopyright infoDownload PDFDownload figures

Metabolite exchange between microbiome members produces compounds that influence Drosophila behavior

Decision letter

Affiliation details

Yale University, United States; University of Georgia, United States; University of Connecticut, United States
Kristin Scott, Reviewing editor, University of California, Berkeley, United States

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

Thank you for submitting your article "Metabolite exchange within the microbiome produces compounds that influence Drosophila behavior" for consideration by eLife. Your article has been favorably evaluated by Wendy Garrett (Senior Editor) and three reviewers, one of whom is a member of our Board of Reviewing Editors. The following individual involved in review of your submission has agreed to reveal his identity: Michael B Eisen (Reviewer #2).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.


The manuscript examines olfactory attraction to microbes found in the Drosophila microbiome. The authors show that a co-culture of yeast and bacteria (S. cerevisiae and A. malorum) grown together is more attractive than when the two cultures are mixed before testing. Attraction is reduced in Or42b and OrCo mutants with olfactory defects. Ethanol reduction and acetic acid production are unique to the co-culture. The preference toward the co-culture can largely be explained by acetic acid and esters production. The general point – that microbes interact with each other to affect the culture they are growing in, and that this in turn affects the way the culture is perceived by other organisms – is almost certainly true, and is interesting and important. We think that this study is of interest to the broad readership of eLife and support the publication of a streamlined and improved version which addresses the major points raised.

Essential revisions:

1) The authors change the statistical analyses throughout the manuscript without clarifying why they do so. It is not clear why the authors sometimes choose to compare if the value differs from 0 versus an ANOVA with a post-hoc test. In general, when making statements comparing conditions (to say that effect increases etc.) this cannot be based on a comparison vs. 0 and based on the magnitude of the p-value. In general, the statistical tests should be non-parametric when dealing with behavioural indices.

2) The interchange between microbiome and microbes is confusing. The authors are testing olfactory preference to microbes found in the microbiome but they are not studying how the microbiome affects behavior. It is not clear that S. cerevisiae and bacteria reside in the same niche or interact in the host. Other studies have shown that there is no matching between yeast and microbes associated with wild Drosophila. The title and text should be changed to use the word microbe rather than microbiome. The authors should be careful to distinguish what this work shows about microbial interactions versus the microbiome.

3) The descriptions in the Methods are incomplete. It is not clear how the cultures are presented to the flies, whether there was air flow, the possibility that the flies could touch the substrate etc.

4) Why was the comparison of co-cultures to mixtures of individual species made after they had reached stationary phase? Drosophila prefer exponentially growing cultures to stationary phase cultures, and their preference varies considerably as a function of cell density and growth rate. These are very difficult experiments to get right, but why didn't the authors didn't compare the growing co-cultures to growing mono-cultures?

5) The GC-MS data is surprising, especially the report that isoamyl acetate is unique to the co-culture. Previous studies found that this compound is produced in large quantities by many different strains (wild and lab) of S. cerevisiae. It's obviously possible that this result is a product of the different strains and culture conditions used in this paper. But the failure of yeast alone to produce this here warrants some discussion.

6) The preference of the flies for the co-culture seems to depend on the presence of specific olfactory receptors and it is proposed that OrCo and Or42b are key for mediating this attraction. Unfortunately, these findings require validation by more stringent genetic experiments. Currently, the relevant genetic background controls are missing (e.g. heterozygotes). Also, the involvement of Or42b and OrCo should be validated using different approaches (for example, by neuronal silencing of the neurons expressing these receptors or rescue experiments).

7) What is the logic of the survival experiments? Why look at adults? If there really is a selective advantage for flies in calibrating the specific balance of ethanol and acetic acid, then surely the relevant life stage to look at is eggs and larvae, not adults. Adults can come in and lay eggs and then leave (as they often do) – eggs and larvae don't have that option. And as the authors point out here, cultures like this are not stable. If selection has indeed driven flies to choose ideal ethanol and acetic acid concentrations, then the flies should be selecting egg laying conditions that will be good for eggs and larvae over the coming days, not those that are good for adults at that exact concentration.

8) The authors go to a great length to describe the interesting findings buried in the supplemental figures to Figure 6. We suggest that they restructure the manuscript to include them as main figures in order to ensure that the reader can properly follow the text. The metabolic reconstitution experiments are interesting and highly relevant. Currently they are difficult to grasp as they follow a lot of data and the data are not easily accessible.