Microbes use transporters to regulate the release of metabolites based on value

Curation statements for this article:
  • Curated by eLife

    eLife logo

    eLife Assessment

    The manuscript lays out an investigation into an explanatory principle to describe the secretion of metabolites by microorganisms. The authors propose a hypothesis based on the concept of metabolite value and provide compelling computational and experimental evidence to support their hypothesis. The work generates important insights that could be beneficial for researchers across a range of disciplines.

This article has been Reviewed by the following groups

Read the full article See related articles

Abstract

Microbial interactions are shaped by the exchange of metabolites, ranging from intermediates in central carbon metabolism to amino acids, vitamins and fermentation by-products 1 . While specific metabolic processes such as overflow metabolism are well documented 2 , a general principle explaining the chemical diversity of metabolite release and the variation in individual release rates remains elusive. Here, using a combination of computational and experimental approaches on seven different microbes, we show that release rates are negatively correlated with the value of metabolites, and that this relationship explains more variability in release rates than other frequently associated factors. By measuring metabolite release from 67 E. coli mutants lacking individual transport genes, we find that transporters both cause and counteract net metabolite release. These findings suggest that metabolite release constitutes a fitness cost that microbes have evolved to reduce, while revealing the underlying transport mechanism. This unifying explanation reshapes our understanding of how microbial interactions emerge and persist.

Article activity feed

  1. eLife Assessment

    The manuscript lays out an investigation into an explanatory principle to describe the secretion of metabolites by microorganisms. The authors propose a hypothesis based on the concept of metabolite value and provide compelling computational and experimental evidence to support their hypothesis. The work generates important insights that could be beneficial for researchers across a range of disciplines.

  2. Reviewer #1 (Public review):

    Summary:

    This manuscript addresses how and why intracellular-specific metabolites associated with biosynthesis end up outside microbial cells. Using a variety of approaches (analytical chemistry, metabolic models, comparative analyses of mutants) and incorporating their own data as well as data from previous studies, the authors find an inverse relationship between metabolite externalization and the cost of producing those metabolites, and that both uptake and efflux transporters are major participants in determining extracellular metabolite levels.

    Strengths:

    (1) Presents much useful raw information that will benefit the broader community interested in microbial physiology, ecology, evolution, and biotechnology.

    (2) Compares the authors' own data with data from previous studies.

    (3) Considers and addresses alternative hypotheses.

    (4) Varied approaches and data sets without losing sight of the core motivating questions - this is a large and impressive amount of work.

    (5) Distils many diverse data sets to deliver clear and insightful interpretations.

    Weaknesses:

    No weaknesses were identified by this reviewer. There is always more to learn, but this manuscript already delivers more than I would typically expect; this is an important advance in an area that has been of much recent interest and debate.

  3. Reviewer #2 (Public review):

    The manuscript addresses the lack of a general explanatory framework for the secretion of metabolites by microorganisms and proposes a compelling hypothesis - the concept of "metabolite value" - which is tested both computationally and experimentally. The hypothesis is clearly formulated and represents a promising attempt to offer a unified perspective on this long-standing question.

    Previous major comments (from review at another journal):

    (1) L58ff: Only net accumulations of one average cell in a potentially heterogenous cell population can be measured. More "valuable" metabolites would also be preferentially taken up by other cells. The authors should comment with respect to the working hypothesis.

    (2) L90ff: There are some open questions regarding the selected metabolome data. Of the 126 metabolites measured, up to 56% were released during exponential growth, and 34 of these were used to calculate the release rates. Although the amount of data is considerable, it remains unclear why these particular metabolites were selected. For example, why was metanephrine included as a non-native metabolite of E. coli? Why not just focus on the metabolites that are mapped by the applied GEM?

    (3) L105ff: To substantiate the hypothesis regarding metabolite value, the use of data sets from different strains and conditions, as well as different GEMs, is compelling. However, if the hypothesis is correct, one could question the validity of models that currently do not take metabolite release into account. In this way, the single-factor approach, which examines the effect of the release of a specific metabolite on the growth rate, could lead to distorted results. The authors should comment on this.

    (4) L127ff.: The statement that "43% of the variability" is explained by metabolite value is difficult to interpret without further context - it may represent a substantial effect, or it may be of limited explanatory power. This requires clearer framing. It is also surprising that intracellular concentration appears to have such little influence. While the data may support this conclusion, the result is counterintuitive and would benefit from further discussion or contextualization.

    (5) L193ff: Without Figure S20B, it is impossible to understand the effects of the various knockouts in the cell. Why was DlacA used as a negative control? Furthermore, it is to be expected that the substrate of an enzyme whose gene has been deleted will accumulate. To what extent does this approach, which is based on such clear genetic disturbances, support the general metabolite value hypothesis, which is primarily intended to explain native cell metabolism?

    (6) Figure S25: Based on the raw data shown, it is not entirely convincing that a true chemostat steady state was achieved. In a properly operated chemostat, a metabolic steady state is established, meaning that extracellular metabolite concentrations reflect a balance of secretion, uptake, and dilution. Under such conditions, it is not possible to determine whether specific exometabolites are being re-assimilated; one might only conclude that their net secretion rates are constant.

    In addition, several extracellular metabolite profiles are difficult to reconcile with metabolic expectations. For example, the increase of F6P in a pgi knockout strain and the lack of accumulation of AKG-derived amino acids in a sucB mutant are surprising. The authors should provide more explanations.

    (7) L253ff: The statement that "our findings so far suggest that microbes can modulate metabolite release rates" is difficult to reconcile with the central hypothesis. The manuscript initially proposes that the abstract parameter "metabolite value" determines secretion levels, rather than active modulation by the cell. This conceptual distinction requires clarification.

    (8) The dataset of 66 knockout strains growing on glucose is extensive, but some methodological details are not sufficiently described in the "Materials and Methods" section. For example, there is no explanation as to why 8,700 measurements were performed on 2,100 samples and whether the samples were injected multiple times. Furthermore, the criteria for selecting the 10 time points and the approach for data normalization (e.g., TIC normalization) are not specified.

    (10) From a conceptual point of view, the energy costs of metabolite transport are not taken into account. If the "metabolite value" is also related to active import as a compensation mechanism, the costs associated with metabolite transport and transporter synthesis should be included in the discussion.

    Comments on latest version:

    The authors have thoroughly addressed my previous concerns.

  4. Reviewer #3 (Public review):

    Summary:

    This manuscript by Sulheim et al. aims to find general explanations for why some metabolites are found in higher abundance in the exometabolome of bacteria than others. Starting with a correlation of value of a metabolite with release, they can demonstrate that the more valuable a metabolite, the less likely it is to be released. This pattern holds over several bacterial species and carbon sources in minimal medium. The authors claim that transporters are involved in shaping the exometabolome and show that knock-outs of certain transporters affect the release significantly.

    Strengths:

    The collected data largely underscores the hypothesis by the authors that metabolite value is linked to the release rate. The big dataset generated about external metabolite concentrations upon deletion of single transporter genes is especially valuable and exciting.

    Weaknesses:

    The authors overclaim some of their findings, such as that in the evolution experiment, they would expect evolution to drive the cells to a "WT-like" state, while here, they are selecting for fast growth instead of efficiency, which would probably be closer to what the WT is showing. Moreover, the experiment showing the ecological relevance of transporter knock-outs with amino acid auxotrophs does not clearly show what they claim, as even mutants with a negative expected effect increase growth, sometimes at higher levels than the WT.

  5. Amazing dataset describing the metabolic profiles of 66 E. coli transporter knockouts by @sulheim.bsky.social with Peter Doubleday and @nzamboni.bsky.social. The tool, set up by @lambdapp.bsky.social and Eric Ulrich allows you to explore for yourself. Enjoy! Pre-print here: doi.org/10.1101/2025...

  6. This project wouldn't been possible without Peter Doubleday and @nzamboni.bsky.social at @imsb-eth.bsky.social. Peter conducted >10,000 mass spec analyses 🤯. So grateful for this opportunity to team up with the best of the best in metabolomics 😍 Read more: www.biorxiv.org/content/10.1... #microsky

  7. Authors published a paper in Biorxiv, they used COBRApy in the study. Including #RRIDs will make this less ambiguous. SciScore made a table with this resource, see “Automated Services” module (download as csv, xml or #jats) #methodsmatter #OpenScience

  8. Microbes release lower-value metabolites at higher rates | bioRxiv https://www.biorxiv.org/content/10.1101/2025.08.19.671024v1?rss=1

  9. Microbes release lower-value metabolites at higher rates | bioRxiv https://www.biorxiv.org/content/10.1101/2025.08.19.671024v1?rss=1

  10. Particularly enjoyed the lovely figures www.biorxiv.org/content/10.1...

  11. Excited to share the first results from my postdoc at @dmf-unil.bsky.social with @saramitri.bsky.social www.biorxiv.org/content/10.1...

  12. Microbes release lower-value metabolites at higher rates https://www.biorxiv.org/content/10.1101/2025.08.19.671024v1

  13. Microbes release lower-value metabolites at higher rates https://www.biorxiv.org/content/10.1101/2025.08.19.671024v1