{"paper_id":"248610ab-cb0b-4f54-94ca-ae1b0ce52482","body_text":"Microbes release lower-value metabolites \nat higher rates \nSnorre Sulheim1,2*, Gunn Broli3, Alisson Gillon1, Julien S. Luneau1, Andrew Quinn1, Eric Ulrich1, \nMargaret A. Vogel1, Philipp Engel1 & Sara Mitri1* \nAffilia&ons \n1. Department of Fundamental Microbiology, University of Lausanne \n2. Department of Biological Sciences, University of Bergen \n3. Department of Biotechnology and Nanomedicine, SINTEF Industry \n*Corresponding author. Email: snorre.sulheim@unil.ch; sara.mitri@unil.ch  \n \n \nAbstract \nMicrobes release a wide range of metabolites into their environment, yet the reasons for this \nrelease and the factors that influence release rates are not well understood. Here, using a \ncombinaUon of computaUonal and experimental approaches on seven different microbes we \nshow that release rates are negaUvely correlated with the value of metabolites, and that this \nrelaUonship explains more variability in release rates than other frequently associated factors. \nThese findings are in accordance with the idea that metabolite release generally consUtutes a \nfitness cost that microbes have evolved to reduce. This new conceptual explanaUon for why \nmicrobes release metabolites has important implicaUons for how we think about the \nemergence of cross-feeding interacUons. \nMain \nIntroduc)on \nMetabolites are frequently exchanged between microbes in natural communiUes (1). Such \ncross-feeding plays a key role in maintaining community diversity and funcUon (2–6). While \ncertain occurrences of cross-feeding arise from extracellular metabolism such as polymer \ndegradaUon by extracellular enzymes (6, 7), many observaUons of cross-feeding stem from the \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nrelease of intracellular metabolites into the environment by microbial community members (8–\n10). Recent in vitro quanUficaUons of microbial cultures have demonstrated that this release \nincludes high-energy glycolyUc intermediates, amino acids and organic acids (11–14). The \nubiquity of metabolite release may help explain the high frequency of auxotrophs isolated from \nnatural environments that may be supported by co-culUvated partners (15–18). However, why \nso many diverse intracellular metabolites are being released and what explains differences in \nrelease rates remains unanswered. \n \nPrevious efforts to answer why microbes release intracellular metabolites have revealed \nfundamental principles of metabolism, including rate-yield trade-offs and constraints on \nproteome allocaUon (19–23), mechanisms of pathway regulaUon and maintenance of \nhomeostasis (24), opUmal catabolic pathway lengths in different environments (25, 26) and \nredox balancing (27, 28).  These explanaUons are context-, metabolite- or species-specific. \nHowever, to fully understand metabolite release, we need to consider broader principles that \napply across diverse microbial taxa and environments. Cell lysis and passive diffusion across the \ncell membrane are two such commonly proposed broad mechanisms (29–32). \n \nHere we move beyond these mechanisUc explanaUons and ask whether metabolite release \ngenerally confers a benefit or imposes a cost on the source organism. While recent work has \nexplored how metabolite release can be beneficial, e.g. by alleviaUng costs associated with \nmolecular noise (32, 33), we here propose and test an alternaUve hypothesis that metabolite \nrelease comes with a cost that microbes have evolved to reduce.  \n \nMetabolite release rates are nega)vely correlated with metabolite value \nIf metabolite release generally imposes a fitness cost, and natural selecUon acts to minimize \nsuch costs, then the release rates of specific metabolites should be negaUvely correlated with \ntheir relaUve importance, or value, to the microbe (Fig. 1A). Release of more valuable \nmetabolites – that require more energy to produce – should have a more severe negaUve \nimpact on fitness and be more strongly selected against.  \n \nTo quanUfy metabolite release rates across different species, we first analysed an exisUng Ume-\nseries exometabolome dataset with high temporal resoluUon and absolute quanUficaUon of 19 \nto 37 metabolites collected from batch cultures of E. coli, B. licheniformis, S. cerevisiae and C. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nglutamicum in high glucose concentraUons (12). We esUmated net uptake and release rates for \nall metabolites using simple linear regression, modelling extracellular metabolite concentraUon \nas a linear funcUon of the area under the growth curve (AUC), unUl the point where the \nextracellular concentraUon saturated or the exponenUal growth phase ended (Methods, Fig. 1B, \nand Figs. S1-S4). The iniUal dynamics of 44-95% of metabolites were well described by this \nsimple linear model (two-sided Wald test, FDR < 0.05, R2 > 0.5, Fig. S1-S4).  \n \nTo esUmate metabolite value, we used previously reconstructed enzyme-constrained genome-\nscale metabolic models (ecGEMs) (34–36). Genome-scale metabolic models (GEMs) are \nmathemaUcal representaUons of a species’ metabolic capacity that can for example be used to \npredict opUmal metabolic phenotypes across various environments (37, 38). The addiUonal \nenzyme constraints account for the total enzyme pool limitaUon in cells (20), enabling \npredicUon of relevant trade-offs and temporal dynamics, e.g. acetate overflow in E. coli (35). \nVarious measures of metabolite value have been proposed in the past (39–41). Here we define \nmetabolite value as the negaUve change in growth rate on metabolite release close to opUmal \ngrowth as predicted by enzyme-constrained GEMs (Fig. 1C). This measure explicitly reflects the \nfitness penalty on growth rate associated with the marginal release of a metabolite, and it can \neasily be used for different species across different environmental contexts.  \n \nWith these two measures we could now test our hypothesis that release rate should be \nnegaUvely correlated with metabolite value. Indeed, this holds for E. coli, S. cerevisiae, B. \nlicheniformis in glucose medium (Pearson ρ, P < 0.001, Fig. 1D). This did not hold for C. \nglutamicum (Fig. 1D), possibly because the engineering of its central carbon metabolism for \nenhanced lysine producUon has also affected the release of other metabolites (42). \n \nTo test whether our hypothesis held beyond simple glucose environments, we conducted new \nwell-controlled batch culUvaUons of E. coli in minimal media with galactose, L-malate or L-\nalanine as the single carbon source. These carbon sources were chosen to span the variability in \nE. coli metabolism to avoid biasing the results towards a parUcular class of compounds or \nmetabolic flux pakerns (Methods, Fig. S5), since this has been shown to produce similar \nexometabolome pakerns (43). Triplicate batch bioreactor experiments were performed under \nstrictly controlled aerobic condiUons with acUve regulaUon of pH and dissolved oxygen (Fig. S6). \nRelaUve extracellular concentraUons of 126 metabolites were then quanUfied at three \nexponenUal-phase samples and one sample from staUonary phase. Of these metabolites, 41-\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n56% were released at significant rates during exponenUal phase (two-sided Wald test, FDR < \n0.05, Figs. S7-S10), consistent with previous findings showing that microbes generate rich \nexometabolomes (12, 13, 44). From this set, we were able to esUmate absolute release rates for \n34 metabolites that we compared with esUmated metabolite values. Again, we found significant \nnegaUve correlaUons for all three condiUons (Pearson ρ, P < 0.001, Fig. 1E). \n \nFinally, we analysed a third dataset that in addiUon to E. coli and Pseudomonas puGda contained \nenvironmental isolates from the genera Enterobacter and Pseudomonas (43). Unlike previous \ndatasets, these experiments were conducted without pH and oxygen control, offering an \nopportunity to test the hypothesis under more variable condiUons. Despite these differences, \nwe again found consistent negaUve correlaUons (Fig. 1F), suggesUng that the relaUonship \nbetween metabolite value and release rate is robust across species, nutrient condiUons, and \nexperimental planorms. Importantly, the ecGEMs used here were reconstructed by different \nresearch groups using different pipelines, suggesUng that our results are not dependent on the \nchoices made during model reconstrucUon. This conclusion was also supported by consistent \nresults when metabolite values were esUmated using four other E. coli GEMs (Fig. S11).  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nFigure 1: Correla'on between microbial metabolite release rates and metabolite values. A) Hypothesis: if metabolite release \ngenerally cons:tutes a fitness cost that microbes have evolved to reduce, metabolite release rates should be nega:vely \ncorrelated with metabolite values. Three metabolites are used to conceptually illustrate their rela:ve posi:on based on their \nes:mated release rate and metabolite value, see panels B and C. B) Metabolite release rates were es:mated from :me-series \nexometabolomics data from batch cultures by linear regression of extracellular concentra:ons vs area under the curve (AUC) of \nbiomass. This panel shows extracellular concentra:ons of 37 metabolites measured during exponen:al phase of E. coli cul:vated \nin glucose medium by Paczia et al. (2012), and the linear regressions used to es:mate net release rates. Linear regressions with \nan R2 < 0.5 or P > 0.05 (two-sided Wald test) are drawn in red. Full :me-series, including sta:onary phase and linear regression \nfor E. coli, B. sub:lis, C. glutamicum and S. cerevisiae are shown in Figs. S1-S4. Metabolite abbrevia:ons: α-KG: alpha-\nketoglutarate, DHAP: dihydroxyacetone phosphate, FBP: fructose 1,6-bisphosphate, F6P: fructose 6-phosphate, G6P: glucose 6-\nphosphate, 2/3PG: 2/3-phosphoglycerate, GA3P: glyceraldehyde 3-phosphate, R5P: ribose 5-phosphate, PEP: \nphosphoenolpyruvate, RU5P/X5P: ribulose 5-phosphate/xylulose 5-phosphate, E4P: erythrose-4-phosphate. C) We used enzyme-\nconstrained genome-scale metabolic models to es:mate metabolite values, quan:fied as the nega:ve change in growth rate \n(Δy) on metabolite release (Δx) close to op:mal growth. D-F) Three different datasets were used to test our hypothesis: D) Paczia \net al. (2012), including S. cerevisiae, B. licheniformis, C. glutamicum and E. coli on glucose (12), E) E. coli on three other carbon \nsources (this work), and F) Vila et al. (2023) covering a smaller number of metabolites on E. coli, P . pu:da and two environmental \nstrains of the genera Enterobacter and Pseudomonas on up to 8 different carbon sources (43). The scaeer plots show mean \nes:mated rates +- standard errors vs es:mated metabolite values. Black-filled circles are datapoints where the lower error bar is \nomieed because mean minus standard error is less than 0 and cannot be included on a log-scale. Axis labels and colour legend \nare shared for panel D-F. Annota:ons in panel F show the Pearson correla:on and associated P value in parentheses.  \n \nMetabolite value is the most important factor explaining variability in metabolite release \nrates \nThe consistent negaUve correlaUon between metabolite release rates and metabolite values \nencouraged us to ask how well this factor explains release variability compared to other \nmetabolic or physiochemical properUes previously associated to metabolite release (33, 45). We \nfocused on the data from bioreactor batch cultures of E. coli with glucose, galactose, L-malate \nor L-alanine as the carbon source, where condiUons were well-defined, the measured \nmetabolites were diverse and reference values for intracellular metabolite concentraUons were \navailable (46–49). We first built univariate linear models to quanUfy the importance of these \nfactors: metabolite value, intracellular concentraUon, compound class, solubility in oil vs water \n(log P), molecular weight, topological polar surface area, charge, hydrogen bond acceptor and \ndonor count, rotatable bond count, turnover and carbon source (Fig. 2A). Metabolite value \nexplained 43% of the variability, more than any other factor (Fig. 2B). The structure-based \nclassificaUon of compound class also had good predicUve power (29% explained), primarily \nbecause carboxylic and keto/hydroxy acids were released at significantly higher rates than \namino acids and other compounds (Fig. 2A). However, these two factors are clearly confounded \nas class-level differences mirrored significant differences in metabolite values (Fig. 2C). The \nnumber of rotatable bonds, solubility in oil vs water (log P) and the number of hydrogen bonds, \neach related to membrane permeability (50, 51), were also predicUve of metabolite release, \nsuggesUng passive diffusion as a relevant mechanism. Metabolite turnover is predicted to \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\ncorrelate negaUvely with release rates, according to the noise-averaging cooperaUon hypothesis \n(low turnover metabolites are more suscepUble to noise and therefore more useful to share \nwithin the populaUon) (33), but our dataset shows likle support for this. \n \nWhen we asked how well the best linear model could predict metabolite release rates, we \nfound that 63% of the variability could be explained by metabolite value, compound class, \nintracellular concentraUon and charge (Fig. 2B). However, predicUve accuracy declined when \napplied to out-of-sample condiUons (45%, P = 0.04, 52.9 ± 3.8% expected from random model, \nFig. S12) or out-of-sample metabolites (50%, P < 0.001, 55.2 ± 0.7% expected from random \nmodel, Fig. S13), likely due to correlaUons among the factors, which warrants cauUon when \ninterpreUng factors’ relaUve importance (Fig. S14). Nevertheless, excluding metabolite value as \na factor reduced model performance on out-of-sample condiUons (10% decline in median R2, P \n= 0.003, one-sided Mann-Whitney U, N = 20, Fig. S15), and model quality was largely dependent \non including metabolite value or compound class (Fig. S16). Furthermore, we found differences \nin metabolite value between condiUons to be negaUvely correlated with differences in release \nrate (Pearson ρ, P = 2e-3, Fig. S17). This not only supports the relevance of this factor in \npredicUng out-of-sample rates but also raises the quesUon of whether microbes adapt their \nrelease rates to context-dependent costs. Finally, when we esUmated each factor’s importance \nseparately for each microbe in each dataset (except C. glutamicum), metabolite value sUll \nemerged as the most robust explanatory factor (Fig. 2D). \n \nCell lysis is commonly considered a key mechanism for metabolite release (31, 32). To measure \nits importance in our experiments, we used E. coli batch cultures with paired sampling of intra- \nand extracellular metabolites, as well as flow cytometry with live/dead staining to quanUfy the \nfracUon of lysed cells. In general, less than 10% of extracellular concentraUons could be \nexplained by the release of intracellular metabolites following cell lysis, and osen much less \n(Fig. 3A). We then extended this analysis by comparing our collecUon of E. coli extracellular \nmetabolite concentraUons to published intracellular concentraUons (46–49). To avoid asserUng \nequal lysis rates across condiUons, we asked how large a fracUon of the cell populaUon would \nneed to be lysed to account for the extracellular levels, considering only the contribuUon from \nintracellular metabolite pools (Fig. 3B). Only 10.4% (157/1508) of the data points fall within the \nrange of measured lysis fracUon (0.2-2.3%, mean ± std across condiUons = 0.6 ± 0.6%) or below, \nand more than 56.3% (849/1508) are above the limit where more than the whole cell \npopulaUon is required. Glutamate, glutamine and NAD are the only metabolites where most \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\ndata points can be accounted for by cell lysis, due to high intracellular concentraUons and \nmoderate to low extracellular levels (Fig. S18). \n \nDegradaUon of cell debris may also contribute to extracellular metabolite pools (52). In E. coli, \nthis degradaUon is dependent on the presence of Lon protease in the cell lysate (53). However, \nthe benefit of cell debris degradaUon by Lon was only significant well beyond exponenUal \nphase, and only 19 ± 6% of oligopepUdes (6-10 amino acids) in the lysate were catabolised by \nLon over 20 hours (53). Thus, in our dataset where 90% of the samples were collected before 26 \nhours (67% in exponenUal phase), we expect only a fracUon of the proteome from lysed cells to \ncontribute to the extracellular concentraUons of amino acids. When we included 10% proteome \ndegradaUon in our analysis, only the extracellular levels of tryptophan and threonine \naddiUonally became fully explained by cell lysis (Fig. S19). However, apart from amino acids, the \nmetabolites in Fig. 3B are not typical degradaUon products. Overall, these results suggest that \ncell lysis and proteome degradaUon are insufficient to explain extracellular levels of most \nmetabolites – consistent with previous work (12, 31). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n \nFigure 2: The importance of different factors in explaining metabolite release rates. A) Among the tested chemical and \nmetabolic factors, log-scaled metabolite value explains the most variability in log-scaled E. coli release rates. The tested factors \nare shown in order of how much variability (R2) in log-scaled metabolite value they explain as a univariate linear model, shown \nin panel B. Compound class abbrevia:ons: CA: Carboxylic acids, AA: Amino acids, KA: Keto/hydroxy acids, OC: Organooxygen \ncompounds, O: Other. Sta:s:cal test for compound classes and carbon source: Kruskal-Wallis H-test followed by Conover’s test \nwith BH correc:on. For the other factors the reported P values are from Pearson correla:ons. B) The best linear model, ranked \nby BIC score, is a four-factor model that explains 63.4% of the variability in log-scaled release rates. C) Differences in metabolite \nvalues between compound classes. Abbrevia:ons and sta:s:cal test as in Fig. 2A. D) Log-scaled metabolite value consistently \nranks among the most important factors when evaluated separately for each species in each dataset. Factor abbrevia:ons: RBC: \nRotatable Bond Count, HBDC: Hydrogen Bond Donor Count, TPSA: Topological Polar Surface Area, HBAC: Hydrogen Bond \nAcceptor Count. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n \n \nFigure 3: Contribu'on from cell lysis and associated release of intracellular metabolites to extracellular concentra'ons. A) \nUsing paired intra- and extracellular metabolite samples from the exponen:al phase of E. coli batch cultures we quan:fy what \nfrac:on of extracellular metabolite concentra:ons can be explained by cell lysis. Point size reflects order-of-magnitude increase \nin extracellular concentra:on from the 2-hour sample to the late-exponen:al phase sample (Fig. S30), and all values except \nleucine in the L-malate condi:on were in the μΜ range. B) We expanded this analysis to all metabolite measurements in our E. \ncoli datasets by using literature values for intracellular concentra:ons (Methods). For each metabolite, we es:mated the frac:on \nof cell popula:on that would need to be lysed for the intracellular concentra:ons released by cell lysis to explain the measured \nextracellular concentra:ons. The green region covers the range of measured frac:ons of lysed cells from A), and the red region is \nwhere more than the whole cell popula:on is required to explain extracellular concentra:ons. Abbrevia:ons: 2PG: 2-\nphosphoglycerate, 3PG: 3-phosphoglycerate, α-KG: Alpha-ketoglutarate, DHAP: Dihydroxyacetone phosphate, E4P: Erythrose 4-\nphosphate, FBP: Fructose 1,6-bisphosphate, F6P: Fructose 6-phosphate, G6P: Glucose 6-phosphate, PEP: Phosphoenolpyruvate, \nR5P: Ribose 5-phosphate, RU5P: Ribulose 5-phosphate, XU5P: Xylulose 5-phosphate.  \n \nMetabolic disrup)on and evolu)on change metabolite release rates \nGiven the effect of gene deleUons on intracellular metabolite levels (54), we assumed that the \nlack of negaUve correlaUon between metabolite value and release rate in C. glutamicum (Fig. \n1D) could be a result of its engineered metabolism (42). To test the sensiUvity of metabolite \nrelease pakerns to modificaUons in core metabolism, we measured the exometabolome of \neight E. coli knockout (KO) strains in late exponenUal phase in galactose medium (Fig. 4A). These \ngene knockouts target different parts of core metabolism, were previously shown to affect \nintracellular metabolite pools (54), and were predicted to change intracellular flux pakerns (Fig. \nS20). In line with our expectaUons, 14-38% (16-42 of 111) of the extracellular metabolite \nconcentraUons differed significantly from the wild type across the KO strains (Welch’s t-test, FDR \n< 0.05, Fig. 4B). In contrast, only 2.7% of metabolite concentraUons were different in the \nnegaUve control strain ΔlacA. Each KO strain displayed disUnct exometabolome pakerns (Fig. \n4C), and we observed the most pronounced differences for metabolites in proximity of the \ndeleted reacUon (Figs. S21-S23), consistent with intracellular effects of enzyme-deleUons (54). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nThe WT reference for these KO strains from the KEIO collecUon (55) - E. coli BW25113 - had \nlargely comparable rates as E. coli MG1655 used in the experiments presented in Fig. 1 (Fig. \nS24).  \n \nOur central hypothesis for the negaUve correlaUon between metabolite value and release rate is \nthat bacteria face a trade-off: the cost of prevenUng metabolite release versus the energy lost \nthrough that release. If this trade-off is under strong selecUon, then disrupUng the balance (e.g. \nvia a gene knockout) should trigger rapid evoluUonary compensaUon to restore it. To test this, \nwe evolved the ΔaceE and ΔsucB mutants – chosen for their severe growth defects and disUnct \nexometabolome characterisUcs – for ~100 (replicates M2-M4) and ~200 generaUons (M5-M7) in \ngalactose medium, respecUvely (Fig. S25). The strains evolved in chemostats, where byproducts \nare conUnuously removed to avoid accumulaUon and later uptake of metabolites. From each \nchemostat endpoint, we isolated three strains and selected from each replicate the fastest \ngrowing clone (Fig. 4, D and E). These clones showed 94-146% (ΔaceE) and 50-82% (ΔsucB) \ngrowth rate increases relaUve to their ancestors (Fig. 4, F and G). To assess changes in \nmetabolite release, we measured extracellular concentraUons of 5 metabolites at the end of \nexponenUal growth (Fig. 4, D and E). These metabolites were selected because their release was \nstrongly disturbed in the ancestors (Figs. S21-S23), and either near the deleted reacUon \n(pyruvate and lactate for the ΔaceE; citrate, isocitrate and cis-aconitate for ΔsucB) where we \nexpected the strongest effects, or located elsewhere in the metabolic network and of different \nchemical classes to probe more widespread effects. Lactate levels were significantly reduced in \n2/3 ΔaceE isolates (Fig. 4F), which also exhibited the highest growth rates. The reducUons were, \nhowever, not sufficient to reach WT concentraUons. In contrast, all three ΔaceE isolates showed \nincreased pyruvate release, contrary to expectaUons. These results were qualitaUvely invariant \nto growth rate normalizaUon (Fig. S26). For the ΔsucB isolates, we found in general small \nreducUons in extracellular concentraUons (Fig. 4G), but some of these differences change \nqualitaUvely upon growth rate normalizaUon (Fig. S27). For the metabolites elsewhere in the \nmetabolic network, the outcomes were variable (Figs. S28 and S29). \n \nTo esUmate the fitness contribuUon of reducing metabolite release, we used ecGEMs to predict \nrelaUve growth rate penalUes associated with the observed release rates, using the ancestors’ \nmean growth rates as references. In the ΔaceE ancestor, lactate release is predicted to reduce \nfitness by 23% whereas pyruvate release has negligible impact (0.02%). The reduced release of \nlactate in the isolates from chemostat M2 and M4 is predicted to reduce the negaUve fitness \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\neffect to only 16% and 7%, respecUvely. These contribuUons are small compared to the \nobserved increase (94-146%). The pakern is similar for the ΔsucB isolates: the esUmated net \nfitness reducUon of citrate, isocitrate and cis-aconitate in the ancestor is only 0.37%, far from \nsufficient to explain the improved growth rates.  \n \nTo beker understand the reasons for increased growth and changed release rates, we idenUfied \nsequence variants in the selected fast-growing strains and mutaUons that had fixed in the \nchemostat populaUons (Table S1-S4). Aside from the ΔaceE-M4 isolate, which likely \nhypermutated due to a mutT mutaUon, the evolved ΔaceE isolates have few sequence variants, \nand either related to galactose uptake (e.g. galS in M2, M4 and M7, regulaUng galactose uptake) \nor modulaUng fluxes near the deleted reacUon (e.g. poxB in M2, ldhA in M4, and icd, aceK and \nsdhA in M5, M6 & M7 and M7, respecUvely). Fixed non-synonymous variants in the chemostat \npopulaUons support these observaUons, e.g. ldhA and galS were fixed in M2 and ilvG was fixed \nin M4. The ΔsucB populaUons have many more fixed mutaUons, but among the ones shared \nacross replicates we find again galS and aceK. Together, these results show that metabolite \nrelease rates change as microbes evolve, but in our experiments, these changes were likely \nsecondary to other, more dominant drivers of fitness – leading to inconsistent changes in \nmetabolite release across replicates.  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n \nFigure 4: Effect of gene knockouts on metabolite release rates and the effect of subsequent evolu'on to higher fitness. A) \nGrowth curves of selected KO mutants, with sampling :me-points indicated by open circles. B) Comparison of extracellular \nmetabolite concentra:ons in late exponen:al (OD600~1.2-1.3) between KO mutants and WT. Significant differences are marked \nby solid black outlines (Welch’s t-test, FDR < 0.05). C) A PCA plot of exometabolome profiles, showing that each strain has a \ndis:nct paeern. D & E) Batch culture growth curves and sampling :mepoints for isolates from the endpoint of the evolu:on \nexperiment for ΔaceE and ΔsucB, respec:vely. F & G) Growth rates and extracellular metabolite concentra:ons for metabolites \nnear the deleted reac:on, for ΔaceE and ΔsucB, respec:vely. Sta:s:cal significance was assessed using Welch’s t-test. \n \nDiscussion \nTo beker understand metabolic interacUons between microbes, it is necessary to develop and \ntest hypotheses for the fundamental principles that govern them. Here we propose that the \nbroad release of intracellular metabolites is in general a loss of energy and that for each \nmetabolite there is a trade-off between the cost of its release and the cost of prevenUng it. How \nmicrobes act to prevent release remains poorly understood, but one possibility is expressing \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\ntransporters that immediately re-consume released metabolites (56, 57). This hypothesis \nconnects to previous work linking fitness costs to microbial traits, including amino acid uptake \nrates (14), proteome composiUon (10, 39), properUes of extracellular enzymes (58) and \nmetabolic strategies (59).  \n \nWe find consistent support for our hypothesis: (i) Metabolite values are negaUvely correlated \nwith metabolite release rates across different species and nutrient environments, (ii) Metabolite \nvalue explains the most variability in release rates among all tested factors, (iii) Release rates \ncannot be explained by cell lysis alone. We also find that release rates are sensiUve to changes \nin intracellular metabolism, both caused by gene knockouts and subsequent evoluUon. But on \nthe Umescale of our evoluUonary experiments, changes in release rates were inconsistent \nacross replicates and did not always match our expectaUons. This may be because alternaUve \npaths towards increased fitness were favoured that were independent of metabolite release, or \nthat increased growth rate was selected for more than increased yield.  \n \nBeyond enabling order-of-magnitude predicUons of release rates using metabolic models, these \nfindings impact how we think about metabolite release and the evoluUon of microbial cross-\nfeeding. While it is commonly assumed that cross-feeding either emerges from extracellular \nenzymes, from the release of waste products or is an act that benefits other species at the \nexpense of the releaser (2, 60–63), our results suggest another alternaUve: Microbes release \nmetabolites because using, retaining or recapturing them is not worth the investment. This \nincreases the scope for commensal cross-feeding, which in this scenario can arise without the \nneed to invoke the evoluUon of cooperaUon or being vulnerable to cheaters (64).  \n \n \n \n \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nReferences \n1.  C. Kost, K. R. PaUl, J. Friedman, S. L. Garcia, M. Ralser, Metabolic exchanges are ubiquitous \nin natural microbial communiUes. Nat. Microbiol., 1–9 (2023). \n2.  E. J. Culp, A. L. Goodman, Cross-feeding in the gut microbiome: Ecology and mechanisms. \nCell Host Microbe 31, 485–499 (2023). \n3.  A. E. Douglas, The microbial exometabolome: ecological resource and architect of \nmicrobial communiUes. Philos. Trans. R. Soc. B Biol. Sci. 375, 20190250 (2020). \n4.  G. D’Souza, S. Shitut, D. Preussger, G. Yousif, S. Waschina, C. Kost, Ecology and evoluUon of \nmetabolic cross-feeding interacUons in bacteria. Nat. Prod. Rep. 35, 455–488 (2018). \n5.  L. M. Solden, A. E. Naas, S. Roux, R. A. Daly, W. B. Collins, C. D. Nicora, S. O. Purvine, D. W. \nHoyt, J. Schückel, B. Jørgensen, W. Willats, D. E. Spalinger, J. L. Firkins, M. S. Lipton, M. B. \nSullivan, P . B. Pope, K. C. Wrighton, Interspecies cross-feeding orchestrates carbon \ndegradaUon in the rumen ecosystem. Nat. Microbiol. 3, 1274–1284 (2018). \n6.  S. Pontrelli, R. Szabo, S. Pollak, J. Schwartzman, D. Ledezma-Tejeida, O. X. Cordero, U. \nSauer, Metabolic cross-feeding structures the assembly of polysaccharide degrading \ncommuniUes. Sci. Adv. 8, eabk3076 (2022). \n7.  M. Gralka, R. Szabo, R. Stocker, O. X. Cordero, Trophic InteracUons and the Drivers of \nMicrobial Community Assembly. Curr. Biol. 30, R1176–R1188 (2020). \n8.  M. Daniels, S. van Vliet, M. Ackermann, Changes in interacUons over ecological Ume scales \ninfluence single-cell growth dynamics in a metabolically coupled marine microbial \ncommunity. ISME J. 17, 406–416 (2023). \n9.  Y.-C. Chuang, N. W. Haas, R. Pepin, M. G. Behringer, Y . Oda, B. LaSarre, C. S. Harwood, J. B. \nMcKinlay, Bacterial adenine cross-feeding stems from a purine salvage bokleneck. ISME J. \n18, wrae034 (2024). \n10.  M. T. Mee, J. J. Collins, G. M. Church, H. H. Wang, Syntrophic exchange in syntheUc \nmicrobial communiUes. Proc. Natl. Acad. Sci. 111, E2149–E2156 (2014). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n11.  K. Campbell, J. Vowinckel, M. Mülleder, S. Malmsheimer, N. Lawrence, E. Calvani, L. Miller-\nFleming, M. T. Alam, S. Christen, M. A. Keller, M. Ralser, Self-establishing communiUes \nenable cooperaUve metabolite exchange in a eukaryote. eLife 4, e09943 (2015). \n12.  N. Paczia, A. Nilgen, T. Lehmann, J. Gätgens, W. Wiechert, S. Noack, Extensive \nexometabolome analysis reveals extended overflow metabolism in various \nmicroorganisms. Microb. Cell Factories 11, 122 (2012). \n13.  J. S. L. Yu, C. Correia-Melo, F. Zorrilla, L. Herrera-Dominguez, M. Y . Wu, J. Hartl, K. Campbell, \nS. Blasche, M. Kreidl, A.-S. Egger, C. B. Messner, V. Demichev, A. Freiwald, M. Mülleder, M. \nHowell, J. Berman, K. R. PaUl, M. T. Alam, M. Ralser, Microbial communiUes form rich \nextracellular metabolomes that foster metabolic interacUons and promote drug tolerance. \nNat. Microbiol. 7, 542–555 (2022). \n14.  M. Zampieri, M. Hörl, F. Hotz, N. F. Müller, U. Sauer, Regulatory mechanisms underlying \ncoordinaUon of amino acid and glucose catabolism in Escherichia coli. Nat. Commun. 10, \n3354 (2019). \n15.  G. D’Souza, S. Waschina, S. Pande, K. Bohl, C. Kaleta, C. Kost, Less Is More: SelecUve \nAdvantages Can Explain the Prevalent Loss of BiosyntheUc Genes in Bacteria. EvoluGon 68, \n2559–2570 (2014). \n16.  R. Gregor, G. T. Vercelli, R. E. Szabo, M. Gralka, R. C. Reynolds, E. B. Qu, N. M. Levine, O. X. \nCordero, Vitamin auxotrophies shape microbial community assembly in the ocean. bioRxiv \n[Preprint] (2024). hkps://doi.org/10.1101/2023.10.16.562604. \n17.  B. Ryback, M. Borneld-Miller, J. A. Vorholt, Metabolic adaptaUon to vitamin auxotrophy by \nleaf-associated bacteria. ISME J. 16, 2712–2724 (2022). \n18.  G. Yousif, F. Zorrilla, S. Dash, L. Oña, A. Shekhar, S. Giri, R. Guan, S. Harshe, M. Itermann, D. \nWelter, V. Benes, K. R. PaUl, C. Kost, Obligate cross-feeding of metabolites is common in soil \nmicrobial communiUes. bioRxiv [Preprint] (2025). \nhkps://doi.org/10.1101/2025.01.29.635426. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n19.  V. Baldazzi, D. Ropers, J.-L. Gouzé, T. Gedeon, H. de Jong, Resource allocaUon accounts for \nthe large variability of rate-yield phenotypes across bacterial strains. eLife 12, e79815 \n(2023). \n20.  M. Basan, S. Hui, H. Okano, Z. Zhang, Y . Shen, J. R. Williamson, T. Hwa, Overflow \nmetabolism in Escherichia coli results from efficient proteome allocaUon. Nature 528, 99–\n104 (2015). \n21.  M. Huelsmann, O. T. Schubert, M. Ackermann, A framework for understanding collecUve \nmicrobiome metabolism. Nat. Microbiol. 9, 3097–3109 (2024). \n22.  C. Malina, R. Yu, J. Björkeroth, E. J. Kerkhoven, J. Nielsen, AdaptaUons in metabolism and \nprotein translaUon give rise to the Crabtree effect in yeast. Proc. Natl. Acad. Sci. 118, \ne2112836118 (2021). \n23.  T. Pfeiffer, S. Schuster, S. Bonhoeffer, CooperaUon and CompeUUon in the EvoluUon of ATP-\nProducing Pathways. Science 292, 504–507 (2001). \n24.  M. L. Reaves, B. D. Young, A. M. Hosios, Y .-F. Xu, J. D. Rabinowitz, Pyrimidine homeostasis is \naccomplished by directed overflow metabolism. Nature 500, 237–241 (2013). \n25.  R. Heinrich, S. Schuster, H.-G. Holzhüker, MathemaUcal analysis of enzymic reacUon \nsystems using opUmizaUon principles. Eur. J. Biochem. 201, 1–21 (1991). \n26.  E. Costa, J. Pérez, J.-U. Kres, Why is metabolic labour divided in nitrificaUon? Trends \nMicrobiol. 14, 213–219 (2006). \n27.  M. J. van Hoek, R. M. Merks, Redox balance is key to explaining full vs. parUal switching to \nlow-yield metabolism. BMC Syst. Biol. 6, 22 (2012). \n28.  G. N. Vemuri, M. A. Eiteman, J. E. McEwen, L. Olsson, J. Nielsen, Increasing NADH oxidaUon \nreduces overflow metabolism in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. 104, \n2402–2407 (2007). \n29.  A. Bar-Even, E. Noor, A. Flamholz, J. M. Buescher, R. Milo, Hydrophobicity and Charge \nShape Cellular Metabolite ConcentraUons. PLoS Comput. Biol. 7 (2011). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n30.  K. Campbell, L. Herrera-Dominguez, C. Correia-Melo, A. Zelezniak, M. Ralser, Biochemical \nprinciples enabling metabolic cooperaUvity and phenotypic heterogeneity at the single cell \nlevel. Curr. Opin. Syst. Biol. 8, 97–108 (2018). \n31.  J. B. McKinlay, Are Bacteria Leaky? Mechanisms of Metabolite ExternalizaUon in Bacterial \nCross-Feeding. Annu. Rev. Microbiol. 77, 277–297 (2023). \n32.  S. Gude, G. J. Pherribo, M. E. Taga, Emergence of Metabolite Provisioning as a By-Product \nof Evolved Biological FuncUons. mSystems 5, 10.1128/msystems.00259-20 (2020). \n33.  J. G. Lopez, N. S. Wingreen, Noisy metabolism can promote microbial cross-feeding, eLife \n(2022). hkps://doi.org/10.7554/eLife.70694. \n34.  B. J. Sánchez, C. Zhang, A. Nilsson, P .-J. Lahtvee, E. J. Kerkhoven, J. Nielsen, Improving the \nphenotype predicUons of a yeast genome-scale metabolic model by incorporaUng \nenzymaUc constraints. Mol. Syst. Biol. 13, 935 (2017). \n35.  P . S. Bekiaris, S. Klamt, AutomaUc construcUon of metabolic models with enzyme \nconstraints. BMC BioinformaGcs 21, 19 (2020). \n36.  Z. Mao, J. Niu, J. Zhao, Y . Huang, K. Wu, L. Yun, J. Guan, Q. Yuan, X. Liao, Z. Wang, H. Ma, \nECMpy 2.0: A Python package for automated construcUon and analysis of enzyme-\nconstrained models. Synth. Syst. Biotechnol. 9, 494–502 (2024). \n37.  J. D. Orth, I. Thiele, B. Ø. O. Palsson, What is flux balance analysis? Nat Biotech 28, 245–\n248 (2010). \n38.  C. Gu, G. B. Kim, W. J. Kim, H. U. Kim, S. Y . Lee, Current status and applicaUons of genome-\nscale metabolic models. Genome Biol. 20, 121 (2019). \n39.  H. Akashi, T. Gojobori, Metabolic efficiency and amino acid composiUon in the proteomes \nof Escherichia coli and Bacillus subUlis. Proc. Natl. Acad. Sci. 99, 3695–3700 (2002). \n40.  M. Mori, C. Cheng, B. R. Taylor, H. Okano, T. Hwa, FuncUonal decomposiUon of metabolism \nallows a system-level quanUficaUon of fluxes and protein allocaUon towards specific \nmetabolic funcUons. Nat. Commun. 14, 4161 (2023). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n41.  A. R. Pacheco, M. Moel, D. Segrè, Costless metabolic secreUons as drivers of interspecies \ninteracUons in microbial ecosystems. Nat. Commun. 10, 103 (2019). \n42.  T. Georgi, D. Rikmann, V. F. Wendisch, Lysine and glutamate producUon by \nCorynebacterium glutamicum on glucose, fructose and sucrose: Roles of malic enzyme and \nfructose-1,6-bisphosphatase. Metab. Eng. 7, 291–301 (2005). \n43.  J. C. C. Vila, J. Goldford, S. Estrela, D. Bajic, A. Sanchez-GorosUaga, A. Damian-Serrano, N. \nLu, R. Marsland, M. Rebolleda-Gomez, P . Mehta, A. Sanchez, Metabolic similarity and the \npredictability of microbial community assembly. bioRxiv [Preprint] (2023). \nhkps://doi.org/10.1101/2023.10.25.564019. \n44.  R. Baran, E. L. Brodie, J. Mayberry-Lewis, E. Hummel, U. N. Da Rocha, R. Chakraborty, B. P . \nBowen, U. Karaoz, H. Cadillo-Quiroz, F. Garcia-Pichel, T. R. Northen, Exometabolite niche \nparUUoning among sympatric soil bacteria. Nat. Commun. 6, 8289 (2015). \n45.  P . Wendering, Z. Nikoloski, COMMIT: ConsideraUon of metabolite leakage and community \ncomposiUon improves microbial community reconstrucUons. PLOS Comput. Biol. 18, \ne1009906 (2022). \n46.  B. D. Bennek, J. Yuan, E. H. Kimball, J. D. Rabinowitz, Absolute quanUtaUon of intracellular \nmetabolite concentraUons by an isotope raUo-based approach. Nat. Protoc. 3, 1299–1311 \n(2008). \n47.  J. O. Park, S. A. Rubin, Y .-F. Xu, D. Amador-Noguez, J. Fan, T. Shlomi, J. D. Rabinowitz, \nMetabolite concentraUons, fluxes and free energies imply efficient enzyme usage. Nat. \nChem. Biol. 12, 482–489 (2016). \n48.  T. Sajed, A. Marcu, M. Ramirez, A. Pon, A. C. Guo, C. Knox, M. Wilson, J. R. Grant, Y . \nDjoumbou, D. S. Wishart, ECMDB 2.0: A richer resource for understanding the \nbiochemistry of E. coli. Nucleic Acids Res. 44, D495–D501 (2016). \n49.  L. B. Thorfinnsdo‡r, L. García-Calvo, G. H. Bø, P . Bruheim, L. M. Røst, OpUmized Fast \nFiltraUon-Based Sampling and ExtracUon Enables Precise and Absolute QuanUficaUon of \nthe Escherichia coli Central Carbon Metabolome. Metabolites 13, 150 (2023). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n50.  G. Falcón-Cano, C. Molina, M. Á. Cabrera-Pérez, Reliable PredicUon of Caco-2 Permeability \nby Supervised Recursive Machine Learning Approaches. PharmaceuGcs 14, 1998 (2022). \n51.  X. Liu, B. Testa, A. Fahr, Lipophilicity and Its RelaUonship with Passive Drug PermeaUon. \nPharm. Res. 28, 962–977 (2011). \n52.  S. Pontrelli, K. B. Villi, A. Sichert, J. Trouillon, A. Rutz, Z. C. Landry, S. H. Rüdisser, R. Stocker, \nU. Sauer, DegradaUon of extracellular polymeric substances shapes microbial community \ndiversity. PLOS Biol. 23, e3003287 (2025). \n53.  S. E. R. Gibson, I. Frost, S. J. Hierons, T. Moses, W. C. K. Poon, S. A. West, M. J. Cann, \nBacteria encode post-mortem protein catabolism that enables altruisUc nutrient recycling. \nNat. Commun. 16, 1400 (2025). \n54.  T. Fuhrer, M. Zampieri, D. C. Sévin, U. Sauer, N. Zamboni, Genomewide landscape of gene–\nmetabolome associaUons in Escherichia coli. Mol. Syst. Biol. 13, 907 (2017). \n55.  T. Baba, T. Ara, M. Hasegawa, Y . Takai, Y . Okumura, M. Baba, K. A. Datsenko, M. Tomita, B. L. \nWanner, H. Mori, ConstrucUon of Escherichia coli K-12 in-frame, single-gene knockout \nmutants: the Keio collecUon. Mol. Syst. Biol. 2, 2006.0008 (2006). \n56.  D. E. Rancourt, J. T. Stephenson, G. A. Vickell, J. M. Wood, Proline excreUon by Escherichia \ncoli K12. Biotechnol. Bioeng. 26, 74–80 (1984). \n57.  B. Enjalbert, P . Millard, M. Dinclaux, J.-C. Portais, F. LéUsse, Acetate fluxes in Escherichia coli \nare determined by the thermodynamic control of the Pta-AckA pathway. Sci. Rep. 7, 42135 \n(2017). \n58.  D. R. Smith, M. R. Chapman, Economical EvoluUon: Microbes Reduce the SyntheUc Cost of \nExtracellular Proteins. mBio 1, e00131-10 (2010). \n59.  A. Flamholz, E. Noor, A. Bar-Even, W. Liebermeister, R. Milo, GlycolyUc strategy as a \ntradeoff between energy yield and protein cost. Proc. Natl. Acad. Sci. 110, 10039–10044 \n(2013). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n60.  R. K. Friks, A. L. McCully, J. B. McKinlay, Extracellular Metabolism Sets the Table for \nMicrobial Cross-Feeding. Microbiol. Mol. Biol. Rev. 85, 10.1128/mmbr.00135-20 (2021). \n61.  J. A. Lopez, B. Liu, Z. Li, M. S. Donia, N. S. Wingreen, When is microbial cross-feeding \nevoluUonarily stable? bioRxiv [Preprint] (2025). \nhkps://doi.org/10.1101/2025.05.16.654511. \n62.  S. Pande, H. Merker, K. Bohl, M. Reichelt, S. Schuster, L. F. de Figueiredo, C. Kaleta, C. Kost, \nFitness and stability of obligate cross-feeding interacUons that emerge upon gene loss in \nbacteria. ISME J. 8, 953–962 (2014). \n63.  D. Preussger, S. Giri, L. K. Muhsal, L. Oña, C. Kost, Reciprocal Fitness Feedbacks Promote \nthe EvoluUon of MutualisUc CooperaUon. Curr. Biol. 30, 3580-3590.e7 (2020). \n64.  A. R. Zomorrodi, D. Segrè, Genome-driven evoluUonary game theory helps understand the \nrise of metabolic interdependencies in microbial communiUes. Nat. Commun. 8, 1563 \n(2017). \n65.  K. A. Datsenko, B. L. Wanner, One-step inacUvaUon of chromosomal genes in Escherichia \ncoli K-12 using PCR products. Proc. Natl. Acad. Sci. 97, 6640–6645 (2000). \n66.  N. E. Lewis, K. K. Hixson, T. M. Conrad, J. A. Lerman, P . CharusanU, A. D. PolpiUya, J. N. \nAdkins, G. Schramm, S. O. Purvine, D. Lopez-Ferrer, K. K. Weitz, R. Eils, R. König, R. D. Smith, \nB. Ø. Palsson, Omic data from evolved E. coli are consistent with computed opUmal growth \nfrom genome-scale models. Mol. Syst. Biol. 6, 390 (2010). \n67.  R. J. G. B. Campello, D. Moulavi, J. Sander, “Density-Based Clustering Based on Hierarchical \nDensity EsUmates” in Advances in Knowledge Discovery and Data Mining, J. Pei, V. S. Tseng, \nL. Cao, H. Motoda, G. Xu, Eds. (Springer Berlin Heidelberg, Berlin, Heidelberg, 2013), pp. \n160–172. \n68.  H. Gallart-Ayala, I. Konz, F. Mehl, T. Teav, A. Oikonomidi, G. Peyratout, V. van der Velpen, J. \nPopp, J. Ivanisevic, A global HILIC-MS approach to measure polar human cerebrospinal \nfluid metabolome: Exploring gender-associated variaUon in a cohort of elderly cogniUvely \nhealthy subjects. Anal. Chim. Acta 1037, 327–337 (2018). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n69.  V. van der Velpen, T. Teav, H. Gallart-Ayala, F. Mehl, I. Konz, C. Clark, A. Oikonomidi, G. \nPeyratout, H. Henry, M. Delorenzi, J. Ivanisevic, J. Popp, Systemic and central nervous \nsystem metabolic alteraUons in Alzheimer’s disease. Alzheimers Res. Ther. 11, 93 (2019). \n70.  J. A. Myers, B. S. CurUs, W. R. CurUs, Improving accuracy of cell and chromophore \nconcentraUon measurements using opUcal density. BMC Biophys. 6, 4 (2013). \n71.  Q. Ren, B. Henes, M. Fairhead, L. Thöny-Meyer, High level producUon of tyrosinase in \nrecombinant Escherichia coli. BMC Biotechnol. 13, 18 (2013). \n72.  A. Goelzer, J. Muntel, V. Chubukov, M. Jules, E. Prestel, R. Nölker, M. Mariadassou, S. \nAymerich, M. Hecker, P . Noirot, D. Becher, V. Fromion, QuanUtaUve predicUon of genome-\nwide resource allocaUon in bacteria. Metab. Eng. 32, 232–243 (2015). \n73.  Y . Zhang, J. Cai, X. Shang, B. Wang, S. Liu, X. Chai, T. Tan, Y . Zhang, T. Wen, A new genome-\nscale metabolic model of Corynebacterium glutamicum and its applicaUon. Biotechnol. \nBiofuels 10, 169 (2017). \n74.  S. L. McGill, Y . Yung, K. A. Hunt, M. A. Henson, L. Hanley, R. P . Carlson, Pseudomonas \naeruginosa reverse diauxie is a mulUdimensional, opUmized, resource uUlizaUon strategy. \nSci. Rep. 11, 1457 (2021). \n75.  J. Kuepper, J. Dickler, M. Biggel, S. Behnken, G. Jäger, N. Wierckx, L. M. Blank, Metabolic \nEngineering of Pseudomonas puUda KT2440 to Produce Anthranilate from Glucose. Front. \nMicrobiol. 6, 1310 (2015). \n76.  J. Niu, Z. Mao, Y . Mao, K. Wu, Z. Shi, Q. Yuan, J. Cai, H. Ma, ConstrucUon and Analysis of an \nEnzyme-Constrained Metabolic Model of Corynebacterium glutamicum. Biomolecules 12, \n1499 (2022). \n77.  K. Wu, Z. Mao, Y . Mao, J. Niu, J. Cai, Q. Yuan, L. Yun, X. Liao, Z. Wang, H. Ma, ecBSU1: A \nGenome-Scale Enzyme-Constrained Model of Bacillus subUlis Based on the ECMpy \nWorkflow. Microorganisms 11, 178 (2023). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n78.  J. Guo, H. Zhang, C. Wang, J.-W. Chang, L.-L. Chen, ConstrucUon and analysis of a genome-\nscale metabolic network for Bacillus licheniformis WX-02. Res. Microbiol. 167, 282–289 \n(2016). \n79.  J. Nogales, J. Mueller, S. Gudmundsson, F. J. Canalejo, E. Duque, J. Monk, A. M. Feist, J. L. \nRamos, W. Niu, B. O. Palsson, High-quality genome-scale metabolic modelling of \nPseudomonas puUda highlights its broad metabolic capabiliUes. Environ. Microbiol. 22, \n255–269 (2020). \n80.  Z. Mao, J. Niu, J. Zhao, Y . Huang, K. Wu, L. Yun, J. Guan, Q. Yuan, X. Liao, Z. Wang, H. Ma, \nECMpy 2.0: A Python package for automated construcUon and analysis of enzyme-\nconstrained models. Synth. Syst. Biotechnol. 9, 494–502 (2024). \n81.  N. E. Lewis, K. K. Hixson, T. M. Conrad, J. A. Lerman, P . CharusanU, A. D. PolpiUya, J. N. \nAdkins, G. Schramm, S. O. Purvine, D. Lopez-Ferrer, K. K. Weitz, R. Eils, R. König, R. D. Smith, \nB. Ø. Palsson, Omic data from evolved E. coli are consistent with computed opUmal growth \nfrom genome-scale models. Mol. Syst. Biol. 6, 390 (2010). \n82.  A. Ebrahim, J. A. Lerman, B. O. Palsson, D. R. Hyduke, COBRApy: COnstraints-Based \nReconstrucUon and Analysis for Python. BMC Syst. Biol. 7, 74 (2013). \n83.  A. Varma, B. W. Boesch, B. O. Palsson, Stoichiometric interpretaUon of Escherichia coli \nglucose catabolism under various oxygenaUon rates. Appl. Environ. Microbiol. 59, 2465–\n2473 (1993). \n84.  E. Reznik, P . Mehta, D. Segrè, Flux Imbalance Analysis and the SensiUvity of Cellular Growth \nto Changes in Metabolite Pools. PLOS Comput. Biol. 9, e1003195 (2013). \n85.  Y . Ram, E. Dellus-Gur, M. Bibi, K. Karkare, U. Obolski, M. W. Feldman, T. F. Cooper, J. \nBerman, L. Hadany, PredicUng microbial growth in a mixed culture from growth curve data. \nProc. Natl. Acad. Sci. 116, 14698–14707 (2019). \n86.  J. L. Reed, T. D. Vo, C. H. Schilling, B. O. Palsson, An expanded genome-scale model of \nEscherichia coli K-12 (iJR904 GSM/GPR). Genome Biol. 4, R54 (2003). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n87.  A. M. Feist, C. S. Henry, J. L. Reed, M. Krummenacker, A. R. Joyce, P . D. Karp, L. J. Broadbelt, \nV. HatzimanikaUs, B. Ø. Palsson, A genome-scale metabolic reconstrucUon for Escherichia \ncoli K-12 MG1655 that accounts for 1260 ORFs and thermodynamic informaUon. Mol. Syst. \nBiol. 3, 121 (2007). \n88.  J. D. Orth, T. M. Conrad, J. Na, J. A. Lerman, H. Nam, A. M. Feist, B. Ø. Palsson, A \ncomprehensive genome-scale reconstrucUon of Escherichia coli metabolism—2011. Mol. \nSyst. Biol. 7, 535 (2011). \n89.  J. M. Monk, C. J. Lloyd, E. Brunk, N. Mih, A. Sastry, Z. King, R. Takeuchi, W. Nomura, Z. \nZhang, H. Mori, A. M. Feist, B. O. Palsson, iML1515, a knowledgebase that computes \nEscherichia coli traits. Nat. Biotechnol. 35, 904–908 (2017). \n90.  Y . Djoumbou Feunang, R. Eisner, C. Knox, L. Chepelev, J. HasUngs, G. Owen, E. Fahy, C. \nSteinbeck, S. Subramanian, E. Bolton, R. Greiner, D. S. Wishart, ClassyFire: automated \nchemical classificaUon with a comprehensive, computable taxonomy. J. CheminformaGcs 8, \n61 (2016). \n91.  S. Seabold, J. Perktold, “Statsmodels: Econometric and StaUsUcal Modeling with Python” \n(AusUn, Texas, 2010; hkps://doi.curvenote.com/10.25080/Majora-92bf1922-011), pp. 92–\n96. \n92.  M. Lukačišinová, B. Fernando, T. Bollenbach, Highly parallel lab evoluUon reveals that \nepistasis can curb the evoluUon of anUbioUc resistance. Nat. Commun. 11, 3105 (2020). \n93.  S. Chen, Ultrafast one-pass FASTQ data preprocessing, quality control, and deduplicaUon \nusing fastp. iMeta 2, e107 (2023). \n94.  Simon Andrews, FastQC: A Quality Control Tool for High Throughput Sequence Data, \nBabraham InsUtute (2010). \n95.  H. Li, Minimap2: pairwise alignment for nucleoUde sequences. BioinformaGcs 34, 3094–\n3100 (2018). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n96.  P . Danecek, J. K. Bonfield, J. Liddle, J. Marshall, V. Ohan, M. O. Pollard, A. Whitwham, T. \nKeane, S. A. McCarthy, R. M. Davies, H. Li, Twelve years of SAMtools and BCFtools. \nGigaScience 10, giab008 (2021). \n97.  P . Cingolani, A. Plaks, L. L. Wang, M. Coon, T. Nguyen, L. Wang, S. J. Land, X. Lu, D. M. \nRuden, A program for annotaUng and predicUng the effects of single nucleoUde \npolymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-\n2; iso-3. Fly (AusGn) 6, 80–92 (2012). \n98.  E. Garrison, G. Marth, Haplotype-based variant detecUon from short-read sequencing. \narXiv arXiv:1207.3907 [Preprint] (2012). hkps://doi.org/10.48550/arXiv.1207.3907. \n99.  S. D. Burz, A.-L. Abraham, F. Fonseca, O. David, A. Chapron, F. Béguet-Crespel, S. Cénard, K. \nLe Roux, O. Patrascu, F. Levenez, C. Schwintner, H. M. Blo‡ère, C. Béra-Maillet, P . Lepage, \nJ. Doré, C. Juste, A Guide for Ex Vivo Handling and Storage of Stool Samples Intended for \nFecal Microbiota TransplantaUon. Sci. Rep. 9, 8897 (2019). \n100.  B. D. Bennek, E. H. Kimball, M. Gao, R. Osterhout, S. J. Van Dien, J. D. Rabinowitz, Absolute \nmetabolite concentraUons and implied enzyme acUve site occupancy in Escherichia coli. \nNat. Chem. Biol. 5, 593–599 (2009). \n101.  M. Mori, S. Schink, D. W. Erickson, U. Gerland, T. Hwa, QuanUfying the benefit of a \nproteome reserve in fluctuaUng environments. Nat. Commun. 8, 1–8 (2017). \n102.  B. Volkmer, M. Heinemann, CondiUon-Dependent Cell Volume and ConcentraUon of \nEscherichia coli to Facilitate Data Conversion for Systems Biology Modeling. PLoS ONE 6, \ne23126 (2011). \n103.  V. Simensen, C. Schulz, E. Karlsen, S. Bråtelund, I. Burgos, L. B. Thorfinnsdo‡r, L. García-\nCalvo, P . Bruheim, E. Almaas, Experimental determinaUon of Escherichia coli biomass \ncomposiUon for constraint-based metabolic modeling. PLOS ONE 17, e0262450 (2022). \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nAcknowledgements \nWe thank all members of the Mitri lab at the University of Lausanne for discussions and \nfeedback. We also thank people at the Department of Biotechnology and Nanomedicine at \nSINTEF for technical support and advice on the experimental design, and Vassily HatzimanikaUs \nand the Laboratory of ComputaUonal Systems Biotechnology at EPFL for feedback on the \ncomputaUonal analyses. Furthermore, we would like to thank Simon van Vliet, Olga Schubert, \nGuilhem Panneau, Jean C. C. Vila and Uwe Sauer for good discussions, Olga Schubert and Kim \nSchlegel for providing the KEIO strains, SebasUan Burz, You Zheng Teo and Clément Vulin for \nsupport with the live/dead staining protocol, and Vladimir Sentchilo for support with the flow \ncytometry and sampling for metabolomics analyses. We also thank the Metabolomics unit at \nUNIL, and specifically Hector G. Ayala and Julijana Ivanisevic, not only for conducUng the LC-MS \nanalyses, but also for providing guidance and recommendaUons on related makers. Special \nthanks to Florent Mazel and Daniel Segrè for feedback on the dras manuscript. Finally, we thank \nLilja B. Thorfinnsdo‡r, Jean C. C. Vila and Stephan Noack for sharing of experimental data. \n \nFunding \nSwiss NaUonal Science FoundaUon Swiss Postdoctoral Fellowship TMPFP3_217172 (SS),  \nNaUonal Center of Competence in Research Microbiomes grant SNF 51NF40_180575 (SM, SS, \nAG, MAV, JL) \nFaculty of Biology and Medicine, University of Lausanne (EU) \nSwiss NaUonal Science FoundaUon Eccellenza grant PCEGP3_181272 (SM, MAV) \nSINTEF Industry (GB) \nResearch Council of Norway, SFI Industrial Biotechnology, grant no 309558 (GB) \n \nAuthor contribu)ons \nConceptualizaUon: SS, SM; Methodology: SS, AQ, JSL, GB, MAV, EU; InvesUgaUon: SS, GB, AG, \nEU; VisualizaUon: SS; Funding acquisiUon: SS, SM, PE; Project administraUon: SS, SM; \nSupervision: SS, SM; WriUng – original dras: SS; WriUng – review & ediUng: SS, AQ, GB, JSL, PE, \nEU, MAV, SM, AG \n \nCompe)ng interests \nAuthors declare that they have no compeUng interests. \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nData and materials availability \nStrains and isolates used in this study are available on request. Formaked data and code for \ndata analyses and visualizaUon are available on GitHub at hkps://github.com/Mitri-\nlab/metabolite-release. A permanent archive of this repository will be deposited to Zenodo \nupon publicaUon. The raw sequencing data and metadata have been deposited to the NCBI SRA \ndatabase under the BioProject ID PRJNA1270783. \n \nMaterials and Methods \nStrains  \nAll experiments were conducted with either E. coli K-12 MG1655, E. coli K12 BW25113 (wild-type \nancestor for KEIO KO strains) or the following KO mutants from the KEIO collecUon (55): ΔaceE \n(JW0110), ΔcyoD (JW0419), ΔlacA (JW0333), ΔnuoA (JW2283), Δpgi (JW3985), Δrpe (JW3349), \nΔsucB (JW0716), Δzwf (JW1841). The KEIO KO strains were verified by PCR with one primer \nbinding inside the KanR casseke inserted during gene knock-out (Keio_K1_Rev for genes on the (-\n) strand; Keio_KT_RvComp_Fw for genes on the (+) strand) and one gene-specific primer binding \ndownstream of the knocked-out gene (primers are listed in Table S5) (55, 65). For all experiments, \nthe KEIO KO strains were used with the kanamycin resistance casseke in place. \n \nPrecultures and growth media \nFor all experiments detailed below the strains were precultured as described here unless \notherwise stated: From glycerol stocks stored at -70 °C each strain was streaked out onto LB \nagar plates and incubated overnight (37 °C). Then, one colony of each strain was picked and \nused to inoculate liquid precultures (10 mL LB in 50 mL Erlenmeyer flasks) that were incubated \novernight (37 °C, 200 rpm). The kanamycin-resistant KEIO KO strains and strains evolved from \nthese were always precultured in liquid LB and streaked out onto LB agar plates with 25 µg/mL \nkanamycin, while E. coli K-12 MG1655 and E. coli K12 BW25113 were always precultured in LB \nand streaked onto LB agar plates without selecUon. Unless otherwise noted, precultures were \nwashed three Umes by centrifugaUon, with the supernatant removed aser each step, and the \ncells resuspended in M9 medium with no carbon source to eliminate residual LB before \ninoculaUon. Unless otherwise stated, M9 medium was always prepared by mixing (per litre): \n200 mL 5X M9 salts (M6030, Sigma-Aldrich, St. Louis, MO, USA), 2 mL 1M MgSO4, 0.1 mL 1M \nCaCl2, 1 mL 0.5 g/L FeSO4·7H2O, MQ water and appropriate amounts of a carbon source stock \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nsoluUon. The final M9 medium was then pH-adjusted to pH = 7.0, filter sterilized (0.22 μm) and \nstored at 4°C unUl use.  \n \nSelec)ng carbon sources for E. coli K-12 MG1655 batch cultures \nTo complement the exisUng exometabolome data from batch cultures of E. coli on glucose (12), \nwe aimed to select three carbon sources that differed in chemical class and caused variaUon in \nmetabolic flux pakerns. We used the ecGEM eciJO1366 of E. coli K-12 MG1655 without any \nmodificaUons and parsimonious FBA to simulate opUmal flux pakerns across 104 different \ngrowth supporUng metabolites (35, 66). We then performed a PCA analysis of the predicted \nfluxes (as Boolean values) to compare flux pakerns between the carbon sources (Fig. S5). We \nused HDBSCAN (67) to idenUfy clusters and selected from the different clusters in total 10 \ncompounds that also differed in chemical class (sugars, amino acids and organic acids) and their \nentry point into central carbon metabolism (Fig. S5). We then screened the growth of the \nselected carbon sources in triplicates over 48 hours in 96-well plates (Fig. S5). For this \nexperiment, we prepared M9 medium with each of the 10 different carbon sources to a \nconcentraUon scaled to 120 mM carbon atoms. From washed LB precultures, E. coli K-12 MG155 \nwas inoculated to OD600 = 0.05 in 200 μL of media and culUvated with conUnuous shaking \n(double orbital, 425 cpm) in a BioTek Synergy H1 (Agilent Technologies, Winooski, VT, USA) plate \nreader at 37 °C for 48 hours. We ulUmately chose galactose, L-alanine and L-malate, as they \nsupported rapid growth, differed in chemical class, entered central metabolism at different \nlocaUons and were grouped into different clusters based on flux pakerns (Fig. S5). \n \nE. coli K-12 MG1655 batch cultures in bioreactors with different carbon sources \nE. coli K-12 MG1655 was culUvated in lab-scale bioreactors using the three chosen carbon \nsources. Their concentraUons were normalized by the number of carbon atoms to a carbon \natom concentraUon of 120 mM, corresponding to 20 mM galactose (G0625, Sigma-Aldrich), 30 \nmM L-malic acid (02290, Sigma-Aldrich) and 40 mM L-alanine (A7627, Sigma-Aldrich). M9 \nmedium with these three carbon sources were prepared as described above using 25X carbon \nsource stock soluUons.  \n \nThe strain was streaked from a frozen glycerol vial onto an LB agar plate and incubated at 37 °C \novernight. One colony was transferred to a 250 mL baffled shake flask containing 40 mL LB \nmedium. Aser 7 hours, a culture volume corresponding to a start OD600 = 0.05 was inoculated \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nto a 500 mL baffled shake flask containing 100 mL M9 + 40 mM L-alanine. This flask was \nincubated for 36 h (200 rpm, 37 °C). For the two other carbon sources, the strain was incubated \nin LB for 17 h before inoculaUng a culture volume corresponding to OD600 = 0.05 in a 500 mL \nbaffled shake flask containing 100 mL M9 + 20 mM L-galactose or M9 + 30 mM L-malic acid, \nrespecUvely. The flasks were incubated for 24 h (200 rpm, 37 °C).  Before inoculaUon to \nbioreactors, M9 precultures were centrifuged (3220 rcf, 10 min, 4 °C), supernatant removed, \nand pellet resuspended in M9 medium without carbon. \n \nBatch culUvaUons were performed in 1 L DASGIP bioreactors (Eppendorf DASGIP, Jülich, \nGermany) with an iniUal volume of 600 mL M9 medium supplemented with either 20 mM \ngalactose, 30 mM L-malic acid or 40 mM L-alanine, three replicates per condiUon. The reactors \nwere equipped with two Rushton impellers and probes for measurements of dissolved oxygen \n(DO) and pH. Submerged aeraUon was maintained at 0.5 vvm, and agitaUon was cascaded \nbetween 400 and 800, controlled by a DO set point of 30 %. pH set point was 7.0 and controlled \nusing 2 M NaOH and 2 M HCl. Start OD 600 was 0.05. \n  \nSampling was performed regularly during the exponenUal growth phase with addiUonal samples \ncollected during staUonary phase. OD600 was measured at all Ume points and supernatant \nstored: 2 mL culture was centrifuged (3220 rcf, 10 min, 4 °C), and supernatant frozen in -20 °C. \nAt the end of exponenUal phase and at the end of the experiment, samples were collected for \ncell dry weight (CDW). 50 mL culture was centrifuged (3220 rcf, 10 min, 4 °C), supernatant \nremoved, cell pellet resuspended in 50 mL MQ water, centrifuged, supernatant removed, cell \npellet resuspended in 5-10 mL MQ water and transferred to a pre-weighed aluminium beaker \nand dried at 105 °C for 24 hours. Beakers were then weighed again and CDW calculated. \n \nExometabolome analyses of )me-series samples from E. coli K-12 MG1655 batch cultures \nwith different carbon sources \nFour samples from each of the nine batch cultures (three replicates of three different media) \nplus media samples were analysed with LC-MS and GC-MS to measure the extracellular \nconcentraUon of metabolites at different Ume points throughout the experiment. The Ume \npoints were selected based on Ume-series culUvaUon data to cover the exponenUal growth \nphase (three samples) plus a staUonary phase sample (Fig. S6). Samples from the different \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nreplicates were aligned based on the Uming of the end of the exponenUal growth phase (Fig. \nS6).  \n \nLC-MS of selected bioreactor samples was conducted by the Metabolomics facility at UNIL using \ntheir method “h igh-coverage targeted analysis of polar metabolites”, following this procedure: \nMedia samples were thawed on ice. An aliquot (20 µL) was extracted with the addiUon of ice-cold \nmethanol (80 µL), vortexed (30 sec) and centrifuged (15000 rpm, 15 min, 4°C). The supernatants \nwere transferred to vials for liquid chromatography – tandem mass spectrometry (LC-MS/MS) \nanalyses. The extracts were analyzed by hydrophilic interacUon chromatography coupled to \ntandem mass spectrometer (6496 iFunnel , Agilent Te chnologies) using mulUple reacUon \nmonitoring - MRM approach in both, posiUve and negaUve ionizaUon modes, to maximize the \npolar metabolome coverage. The analyUcal condiUons have been described in detail elsewhere \n(68, 69). Raw LC-MS/MS metabolome data were processed using the Mass Hunter QuanUtaUve \nanalysis sosware (Agilent Technologies). The peak areas (or extracted ion chromatograms (EICs) \nfor the monitored MRM transiUons) were used for relaUve comparison of metabolit e levels \nbetween different condiUons or groups of samples. Data quality assessment, including signal dris \ncorrecUon, was performed using pooled quality control (QC) samples analyzed periodically \nthroughout the enUre batch.  \n \nMetabolite quanUficaUon was performed using stable isotope labeled internal standards and \ncalibraUon curves following the signal dris correcUon. Data processing was done using \nMassHunter QuanUtaUve analysis. Peak area integraUon was manually curate d, and \nconcentraUons were reported by selecUng a 6-point porUon of the linear standard curves relevant \nfor the observed concentraUons. \n \nThe relaUve LC-MS data was standardized before subsequent analyses and data processing. \nBoth the relaUve and absolute LC-MS data were cleaned of obvious outliers. For the absolute \ndata the outliers were caused by carry-over from previous samples leading to too high malate \n(six samples) and succinate (one sample) values. For the relaUve LC-MS data there were \naddiUonal outliers of unknown reasons (39 values in total) related to in total 14 metabolites. For \nseveral metabolites there were more missing values (not detected/not quanUfied) in the \nabsolute than in the relaUve data. For a few metabolites (phenylalanine, proline, creaUne) we \nleveraged good linear relaUonships between relaUve and absolute data (R2>0.8) to esUmate the \nmissing absolute values using a simple linear transformaUon.  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n \nGC-MS analysis of the same samples was conducted in the following way: AnalyUcal standards \nwere prepared as M9 medium with increasing concentraUons of acetate, lactate, pyruvate, \nformate and propionate and stored at -70°C unUl use. Samples from the E. coli bioreactor batch \ncultures and analyUcal standards were thawed on iced. Then, 100 μL was transferred to an \nEppendorf tube and kept on a cold block (-20 °C) while adding 5 μL 11% HCl and 500 μL diethyl \nether. Tubes were capped, vortexed on a thermoblock (2000 rpm, 10 min, 1°C) and centrifuged \n(13 000 rcf, 5 min, 4°C). The top organic phase layer was pipeked to a glass vial and derivaUzed \nwith 20 μL N-(t-butyldimethylsilyl)-N-methyltrifluoroacetamide (MTBSTFA) and vortexed briefly. \nSamples were then placed on a heat block (90 min, 35°C) and kept at 16°C unUl analysis. The \nsamples were injected (1 μL) by a Pal3 autosampler onto an Agilent 8890-5977B GC-MSD \n(Agilent Technologies) with a VF-5MS (30 m x 0.25 mm x 0.25 mm) column. The samples were \ninjected with a split raUo of 15:1, helium flow rate of 1 mL/min and inlet temperature of 230 °C. \nThe temperature was held for 2 min at 50 °C, raised at 25 °C/min to 175 °C, 30 °C/min to 280 °C \nand held for 3.5 min. The MSD was run in scan mode from 40-500 Da. Analyte abundances were \ncalculated using the MassHunter QuanUtaUve Analysis sosware (Agilent Technologies). \nAbsolute concentraUons were calculated using standard curves.  \nGalactose concentraUons were measured using the L-Arabinose/D-Galactose assay kit (K-ARGA, \nMegazyme, Bray, Ireland), following their rapid protocol for analyses in 96-well plates.  \n \nEs)ma)on of metabolite release rates of E. coli, B. licheniformis, C. glutamicum and S. \ncerevisiae in glucose medium \nTime-series data on biomass (in OD600) and absolute extracellular metabolite concentraUons \nfrom batch culUvaUons of E. coli, B. licheniformis, C. glutamicum and S. cerevisiae in high \nglucose concentraUons (10-20 g/L) (12) was kindly provided by the authors (mean and standard \ndeviaUons). For metabolites reported as a sum of two metabolites we akributed equal amounts \nto those metabolites to enable direct comparison with metabolite values esUmated for \nindividual metabolites (described below). This includes 2-phosphoglycerate/3-phosphoglycerate \nand ribulose 5-phosphate/xylulose 5-phosphate. To esUmate the metabolite release and uptake \nrates of each species we first converted the OD600 values to gDW/L. For E. coli we used the \nconversion factor we esUmated from our own cell dry weight measurements in M9 galactose \nmedium (0.346 gDW/L/OD600), which is comparable, but on the lower end of values found in the \nliterature (0.36-0.515 gDW/L/OD600) (70, 71). For the three other species we used literature \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nvalues: B. licheniformis (data from B. subUlis) 0.48 gDW/L/OD600 (72); C. glutamicum 0.27 \ngDW/L/OD600 (73); S. cerevisiae (mean of literature values) 0.746 gDW/L/OD600 (70). We then \nintegrated the translated biomass Ume-series data to obtain biomass AUC data that we \nleveraged together with the mean absolute exometabolome data to esUmate the uptake or \nrelease rates for each measured metabolite for each species using linear regression (Figs. S1-\nS4). Rates were only esUmated for the exponenUal growth phase, and only unUl each \nmetabolite either saturated or was clearly being re-consumed.  \n \nTheore&cal jus&fica&on of es&ma&ng uptake or release rates from extracellular concentra&ons \nWe start out from a simple assumpUon that the change in extracellular concentraUon 𝐶! of \nmetabolite i can be explained by the net release or uptake 𝑎!(𝑡) (by any mechanism) by the \nmicrobes with biomass density 𝑋(𝑡) in the batch culture: \n𝑑𝐶!\n𝑑𝑡 = 𝑋(𝑡) ⋅ 𝑎!(𝑡) \nThe concentraUon 𝐶!(𝑡)\tat a given Umepoint t is thus: \n𝐶!(𝑡) = \t + 𝑋(𝑡) ⋅ 𝑎! (𝑡)𝑑𝑡\n\"\n\"#\n+ 𝐶!(𝑡#) \nIf we assume that the uptake or release rate 𝑎!(𝑡) is constant, we see that the concentraUon \n𝐶! (𝑡) is described by a linear funcUon of the constant release rate and the integrated biomass \n(area under the curve, AUC): \n𝐶!(𝑡) = \t 𝑎! + 𝑋(𝑡)𝑑𝑡\n\"\n\"#\n+ 𝐶! (𝑡#) = 𝑎! ⋅ 𝑋$%& (𝑡) + 𝐶# \nHence, when metabolites are released at constant rates the dynamics should be well explained \nby a linear regression of extracellular metabolite concentraUons on biomass AUC, and the \nspecific net rate (uptake or release) is the slope of the curve.  \n \nEs)ma)on of metabolite release rates of E. coli in galactose, L-malate and L-alanine \nFor our bioreactor batch culUvaUons of E. coli in galactose, L-malate and L-alanine, we used \npaired cell dry weight measurements and OD600 readings to calculate conversion factors used to \ntranslate OD600 values for all Ume points (galactose: 0.346 ± 0.016; L-malate: 0.279 ± 0.017; L-\nalanine: 0.296 ± 0.011; mean ± std in gDW/L/OD600). The translated Ume-series data was used \nto calculate the biomass area under the curve used to esUmate metabolite release and uptake \nrates further described below. \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nWe esUmated the metabolite uptake or release rate by performing for each metabolite and \neach carbon source a linear regression of measured concentraUon (relaUve or absolute) vs \nbiomass AUC using the three samples (Umepoints T1-T3) acquired from the three replicate \nbatch cultures during exponenUal phase plus the iniUal M9 medium (T0; Figs. S7-S9). To avoid \nunderesUmaUng release rates for metabolites that potenUally were being re-consumed at the \nend of the exponenUal phase we discarded the last Umepoint from the exponenUal phase in \ncases where the rate was posiUve (indicaUng metabolite release) but there was a significant \nreducUon in extracellular concentraUon from T2-T3 (P < 0.05, one-sided t-test). For 17 (of 126) \nmetabolites we chose to not include T0 for at least one of the three condiUons in the linear \nregression, either because of very large variability of T0 or because the trend from T0 to T1 was \nclearly different from the general trend from T1-T3. We only esUmated the rate for metabolites \nwhere we had at least three data points from at least two different Ume points. For the \nmetabolites where we had both relaUve and absolute data, we used the absolute data if there \nwere enough absolute data points to get adequate slope esUmates. For seven metabolites \n(glutamine, alpha-aminoadiapate, serine, lysine, lactate, NAD and isocitrate) where both \nabsolute and relaUve quanUficaUon was performed, we could esUmate the rate more accurately \non the relaUve data set (because of more missing data points in the absolute data set). \nHowever, there were sufficient absolute measurements to esUmate the absolute spread \n(standard deviaUon) of these data points, enabling translaUon of the slope value from Z-score to \nμmol/gDW/h.  \n \nEs)ma)on of metabolite release rates of E. coli, P.  p u 9 d a, an Enterobacter sp. and a \nPseudomonas sp. in various carbon source environments \nThe dataset from Vila et al. (2023) was kindly provided by the authors. The dataset contains \nexometabolome data from three Umepoints (16, 28, and 48 hours) of E. coli MG1655, P.  p u G d a  \nKT2440, and two environmental isolates of the genera Enterobacter and Pseudomonas, \nrespecUvely (43). All species were culUvated in M9 medium with one of five carbon sources (D-\nglucose, D-fructose, glycerol, pyruvate or L-malate). AddiUonally, the dataset includes \nexometabolome data from two Umepoints (28 and 48 hours) of E. coli and the Enterobacter sp. \non D-ribose, L-arabinose and D-galactose, and of P.  p u G d a  and the Pseudomonas sp. on acetate, \nfumarate and succinate. To esUmate metabolite release rates, we first used the plate reader \ngrowth curves to calculate biomass AUC aser 16, 28 and 48 hours and to idenUfy the end of \nexponenUal growth of each organism in each condiUon. The end of exponenUal phase was \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nautomaUcally idenUfied as the first peak in the second derivaUve of the smoothed growth curve \nwith an OD600 above 60% of maximum. We converted the AUC from OD600 to gDW/L using the \nsame conversion factor for all the different carbon source environments: 0.36 gDW/L/ OD600 for \nE. coli and the Enterobacter sp., mean of measured values and literature values (70, 71); 0.476 \ngDW/L/ OD600 for P.  p u G d a  and the Pseudomonas sp, mean of literature values (74, 75). Finally, \nwe performed a linear regression between extracellular metabolite concentraUons and AUC on \nboth duplicates of each strain/carbon source combinaUon including all datapoints before the \nend of exponenUal growth + 2 hours (a buffer to account for some inaccuracy in the automaUc \ndetecUon). We only included rate esUmates for metabolites that were quanUfied at least twice \nfor each strain. As the dataset did not include an early Umepoint or media measurement we \nassumed the iniUal concentraUon of each metabolite to be 0. If a metabolite was detected more \nat more than one Umepoint we discarded the esUmated rate if there was a qualitaUve \ndiscrepancy in slope value when the slope was esUmated with or without the T0 concentraUon \nassumed to be 0. If the carbon source was among the measured metabolites, this metabolite \nwas excluded from rate esUmaUon in those specific condiUons.  \n \nGenome-scale metabolic models \nFor E. coli, S. cerevisiae and C. glutamicum we used the previously developed enzyme-\nconstrained genome-scaled metabolic models (GEMs) eciJO1366 (35), ecCGL1 (76) and \necYeastGEM 8.3.4 (34), respecUvely. For B. licheniformis there was no available enzyme-\nconstrained GEM, and we therefore opted to use the enzyme-constrained GEM ecBSU1 of the \nclosely related species B. subGlis (77), which was available at github.com/Ubbdc/ecBSU1. A \nnormal GEM for B. licheniformis is available (78), but with this model we could not achieve any \nfeasible flux balance soluUon. To simulate the environmental isolate of the genus Enterobacter \nwe used the E. coli GEM eciJO1366. For the environmental isolate of the genus Pseudomonas \nand for P.  p u G d a  we used the enzyme-constrained version of the P. puGda GEM iJN1463 (79) \nautomaUcally generated by the ECMpy2 method (80), which we here refer to as eciJN1463.  \n \nWe made a few minor changes to the published GEMs before predicUng metabolite values. In \nthe E. coli GEM eciJO1366 the proteome weight of the forward reacUon GALKr (galactokinase) \nwas extremely high (0.016), limiUng the growth rate on galactose to much less than the growth \nrate we measured in M9 galactose medium. To enable eciJO1366 to reach the measured growth \nrate with galactose as the only carbon source we replaced the proteome weight of the forward \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nGALKr reacUon with the proteome weight of the reverse GALKr reacUon which was much \nsmaller (0.00037). For the ecYeastGEM we only merged exchange reacUons that were previously \nsplit to simplify its use. Both ecBSU1 and eciJN1463 were translated from json to sMOMENT \nformat in xml (35), modified by merging split exchange reacUons and fixing metabolite \ncompartment names and annotaUons. For ecBSU1 we also removed the enzyme constraint on \nthe reacUon “PPAm_num1” (inorganic diphosphatase) as it caused unrealisUc uptake and \nrelease rates of phosphate and diphosphate, respecUvely. The C. glutamicum GEM ecCGL1 was \nalso converted from the ECMpy format (36) to the sMOMENT format (35) but otherwise kept \nunchanged. All GEMs were used with their default biomass equaUons as objecUve when \nmaximizing growth in all FBA or pFBA analyses (81). Reframed 1.5.3 \n(github.com/cdanielmachado/reframed) or COBRApy 0.29.1 (82) were used to load, manipulate \nand run constraint-based analyses with Gurobi 10 (Gurobi OpUmizaUon, LLC) as the solver.  \n \nCalcula)on of metabolite values and metabolite turnover using genome-scale metabolic \nmodels \nIn constraint-based modelling a shadow price is an esUmate of how sensiUve the objecUve \nfuncUon in a flux balance analysis (FBA) is to increased influx or efflux of a metabolite (83, 84). \nHence, one can interpret the shadow price as metabolite value as it quantifies the impact on \ngrowth caused by the release of a metabolite (to have positive values we here define \nmetabolite value as the shadow price multiplied by -1). A similar approach to measure the cost \nof metabolite release has been used to identify costless secretions (41). The metabolite value \n(or the shadow price) can be predicted for a specific organism in a specific context by using a \nGEM of the corresponding organism (or closely related) and then constraining this GEM to the \nspecific context (i.e. primarily defining metabolite uptake rates). While shadow prices are \nusually estimated directly by the solver used to run FBA (37), the robustness of these values is \noften questioned because of numerical issues. To get more robust shadow price estimates we \nsolved for each metabolite in each context a separate FBA, where we constrained the GEM to \nhave an (additional) release of 0.01 mmol/gDW/h of that metabolite and quantified the change \nin the objective function, i.e. the change Δy in growth rate caused by the forced metabolite \nrelease. The metabolite value was then quantified as Δy/0.01. If the metabolite was predicted \nto be released by FBA, we added 0.01 mmol/gDW/h to the FBA-predicted value to estimate the \nrate. This was only relevant for acetate for E. coli in the L-malate condition, for ethanol, acetate, \nformate and pyruvate for S. cerevisiae, and for pyruvate for P. putida in the L-malate condition.  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n \nUsing this approach, we esUmated metabolite values for E. coli in batch cultures with glucose, \ngalactose, L-malate or L-alanine as the carbon source by constraining the uptake of the \ncorresponding exchange reacUon in eciJO1366 to the uptake rate of the carbon source as \nesUmated from the experimental data. The same approach was used for C. glutamicum, B. \nlicheniformis and S. cerevisiae using the corresponding models. For E. coli, P.  p u G d a  and the two \nenvironmental isolates of the genera Enterobacter and Pseudomonas we first extracted \nmaximum growth rates for each organism in each environment from growth curves (using \ncurveball (85)). Then, we used the respecUve GEMs to calculate the uptake rates of the \ncorresponding carbon sources required to sustain the maximum growth rates. These uptake \nrates were then used to constrain the GEMs when esUmaUng the metabolite values as \ndescribed above. The turnover (i.e. the sum of posiUve fluxes producing an intracellular \nmetabolite) was esUmated by running parsimonious FBA (66), using the same GEMs constrained \nin the same way as described for above for metabolite values.  \n \nTo validate that the negaUve correlaUon between log-transformed metabolite values and \nrelease rates we compared esUmates from eciJO1366 with esUmates from the 4 benchmark E. \ncoli K-12 MG1655 GEMs iJR904 (86), iAF1260 (87), iJO1366 (88) and iML1515 (89) that are of \nincreasing complexity. We then compared esUmated metabolite values and esUmated release \nrates using data from E. coli in bioreactors with glucose, galactose, L-malate or L-alanine as the \nsole carbon source, the same data used in Fig. 2A (Fig. S11). In the L-malate condiUon for the \niJR904 model we discarded the metabolite value for formate as an outlier because it was \nextremely small (<10-12), below the solver tolerance. Note that no metabolite values were \nnegaUve and hence no other values were discarded upon log-transformaUon. \n \nMetabolite classifica)on, chemical proper)es and intracellular concentra)ons \nThe charge and molecular weight of metabolites were obtained from eciJO1366 (35). All other \nchemical properUes and InChIKeys were obtained from PubChem using the python interface \nPubChemPy (github.com/mcs07/PubChemPy). The InChIKeys were then used to classify \nmetabolites into a chemical taxonomy using ClassyFire (90). To get an appropriate level of detail \nof metabolic classificaUon we used the “Class” level category. However, we separated the class \n“Carboxylic acids and derivaUves” into “Amino acids” and “Carboxylic acids” based on the \n“Subclass” level. We also merged the classes “Hydroxy acids and derivaUves” and “Keto acids \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nand derivaUves” into the joint class “Keto / hydroxy acids”. Finally, metabolites that did not fall \ninto either of the abovemenUoned categories nor the class “Organooxygen compounds” were \njoined into the category “Other”.  \n \nLinear models for evalua)ng the importance of different factors for predic)ng metabolite \nrelease rates \nLinear models were created and analysed using statsmodels (91). Compound class and carbon \nsource was incorporated as categorical variables. Metabolites predicted to have zero turnover \nwas given a value of 10-4, less than the minimum predicted turnover. Release rates where \ncorresponding values were missing in any of the compared factors were discarded before model \nfi‡ng. To evaluate the staUsUcal significance of the out-of-sample predicUons we performed 104 \npermutaUons of model fi‡ng and out-of-sample esUmaUons where the labels (either \nmetabolite or condiUon) was shuffled on each permutaUon. We then calculated the fracUon of \nout-of-sample R2 values from these permutaUons below the observed R2 value to esUmate the P \nvalue.  \n \nCul)va)on and exometabolome analyses of KEIO knockout strains \nOne objecUve was to study how disrupUon of key metabolic fluxes would affect extracellular \nmetabolite concentraUons and how this would change if these strains were allowed to evolve in \na simple nutrient environment. The KEIO collecUon is a collecUon of non-essenUal E. coli KO \nstrains that we could use for this purpose (55). To idenUfy relevant KEIO KO strains we used \neciJO1366 and parsimonious FBA (81) to predict the effect of a gene knockout on opUmal \nmetabolic fluxes, and previous metabolomics data (54) of these strains to see the effect on \nintracellular concentraUons (Fig. S20A). We eventually chose seven strains (ΔaceE, ΔcyoD, \nΔnuoA, Δpgi, Δrpe, ΔsdhB, ΔsucB) that were predicted to have different opUmal flux \ndistribuUons, had many significantly changed intracellular metabolite concentraUons, and \ntargeted different part of key metabolic pathways (Fig. S20). We included ΔlacA as a negaUve \ncontrol expected to behave like the WT as done previously (92).  \n \nThe selected strains from the KEIO collecUon were precultured on LB agar and liquid LB as \npreviously described, and glycerol stocks used for experiments described below were prepared \nby adding glycerol to a final concentraUon of 25 %. \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nFor exometabolome sampling, the KEIO strains and E. coli BW25113 were precultured as \ndetailed above. They were then re-inoculated to OD600 = 0.05 in 10 mL M9 + 40 mM galactose (+ \n25 µg/mL kanamycin for KEIO strains) in 50 mL Erlenmeyer flasks (200 rpm, 37 °C, 31 h). 1 mL \nculture was sampled into Eppendorf tubes and washed twice with M9 without carbon source by \ncentrifugaUon (6000 rcf, 6 min, 4°C). UlUmately, samples were resuspended in 0.5 mL M9 \nwithout carbon source and used to inoculate 200 μL M9 + 40 mM galactose (+25 µg/mL \nkanamycin for KEIO strains) in a flat bokom 96-well plate to OD600 = 0.05 and 6 mL M9 + 40 mM \ngalactose (+25 µg/mL kanamycin for KEIO strains) to straight glass tubes (15.8 cm tall, 1.5 cm \ndiameter) to OD600 = 0.005. The 96-well plate was incubated in a Synergy H1 plate reader with \nconUnuous shaking (double orbital, 425 cpm, 37 °C), and the glass tubes in a shaking incubator \n(200 rpm, 37 °C, 45° angle), both for 108 hours. The experiment in the well plate was used to \nmonitor more closely the growth (as OD600) of these strains and to help with the Uming of the \nexometabolome sampling. All strains were culUvated in three replicates. Two samples were \ncollected during mid-exponenUal growth phase for all cultures, mostly to ensure that at least \none sample from each replicate was obtained before the end of exponenUal phase. 1 mL culture \nvolume was centrifuged (14 000 rpm, 10 min, 4 °C), and supernatant stored in -70 °C for \nexometabolome analysis. The second sample for each replicate (Fig. 4A) was analysed using LC-\nMS at the UNIL Metabolomics facility, following their protocol described above. AddiUonally, we \nalso analysed three samples of pooled inoculums to verify that any carry-over with the \ninoculum was much lower (or below detecUon limit) of than the sample metabolite \nconcentraUons. Three replicate pooled samples were obtained by inoculaUng 6 mL M9 + 20 mM \ngalactose medium with all the strains, each strain to an OD600 = 0.005, and subsequently \nhandled as described above for the culture samples. \n \nChemostat evolu)on experiment with ΔaceE and ΔsucB in bioreactors \nTwo of the KEIO KO strains, ΔaceE and ΔsucB, were selected for the evoluUon experiment with \nconUnuous culUvaUon (chemostats) in bioreactors based on their reduced growth compared to \ntheir ancestor and difference in exometabolome pakerns (Figs. 4A and S21-S23). The strains \nwere streaked from glycerol vials onto LB + 25 µg/mL kanamycin agar plates and incubated at 37 \n°C overnight. 5-6 colonies were transferred to 500 mL baffled shake flasks containing 75 mL LB + \n25 µg/mL kanamycin and incubated for 22 h (200 rpm, 37 °C). \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nChemostat culUvaUons were performed in 1 L DASGIP bioreactors with a constant volume of \n300 mL M9 + 20 mM galactose medium + 25 µg/mL kanamycin (prepared as previously \ndescribed, but without pH adjustment because of pracUcal challenges with the large volumes \nrequired, so pH ≈ 7.1). CulUvaUon was performed at 37 °C with a submerged airflow maintained \nat 0.5 vvm, and a cascaded agitaUon of 400-700 rpm, controlled by a DO set point of 30 %. pH \ncontrol was started at day 6 (ΔaceE) or day 20 (ΔsucB) with a set point of 7.1, controlled using 2 \nM NaOH (Fig. S25A).  \n \nThe reactors (three replicates per strain) were inoculated to OD600 = 0.1. During the late stage of \nexponenUal growth, pumps were started to feed fresh culture medium and remove used \nmedium. Medium was fed using a pre-calibrated peristalUc pump and used medium was \nremoved by a steel pipe placed right above the culture surface. The steel pipe was coupled to a \nperistalUc pump operaUng at a flow rate approximately 1.5x of the medium in flow, to avoid \nvolume accumulaUon. The set diluUon rate was adjusted several Umes during the experiment to \ncompensate for (someUmes surprising) changes in growth dynamics (from 0-0.1 h-1 for ΔaceE, 0-\n0.3 for ΔsucB, Fig. S25B). The actual diluUon rate was monitored by keeping the flasks with fresh \nculture medium on scales. OD600 was measured daily, and weekly, larger samples were \ncollected. Culture was centrifuged (3220 rcf, 10 min, 4 °C), and supernatant and pellet stored at \n-70 °C. In addiUon, cultures were preserved at -70 °C as glycerol stocks. The number of \ngeneraUons were esUmated from the measured diluUon rates across the experiment plus the \ngeneraUons during the iniUal batch culture phase before the culture diluUon was started.  \n \nCul)va)on of single colonies from evolu)on experiment and sampling of exometabolome \nCulture samples from different bioreactors and Umepoints collected in the chemostat \nexperiment were streaked out onto LB + 25 µg/mL kanamycin agar plates and incubated \novernight at 37 °C. Three colonies were picked and individually inoculated in 20 mL LB + 25 \nµg/mL kanamycin in 100 mL Erlenmeyer flasks and incubated (37 °C, 200 rpm, 16 h). 12 mL of \nthe culture was then centrifuged (3220 rcf, 6 min, RT), and the pellets stored at –20 °C for \nsequencing. The isolates were stored as glycerol stocks at -70 °C made from these cultures. \n \n2 mL samples from the precultures were washed as previously described, resuspended in M9 \nwithout carbon source, and used to inoculate a flat bokom 96-well plate to OD600 = 0.05 with \n200 µL M9 + 25 mM galactose + 25 µg/mL kanamycin. The M9 medium was prepared as \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\npreviously described, but to a pH of 7.1 as in the chemostats. The well plate was incubated in a \nSynergy H1 plate reader with conUnuous shaking (double orbital, 425 cpm, 37 °C, 120 h) and the \nOD600 was read every 20 minutes. \n \nBased on these growth measurements we selected the fastest growing isolates from the last \nUmepoint of each of the six bioreactors (three bioreactors per strain). These isolates, the iniUal \nKEIO KO strains ΔaceE and ΔsucB, and the KEIO ancestor E. coli BW25113 were precultured and \nsubsequently washed as described above. The washed preculture were used to inoculate 10 mL \nM9 + 20 mM galactose to OD600 = 0.01 in straight glass tubes (15.8 cm tall, 1.5 cm diameter) in \ntriplicates and incubated for 20-84 hours (200 rpm, 37 °C, 45° angle). 1 mL samples for \nexometabolome analysis were collected at OD600 ≈ 1 (Fig. 4, D and E), filtered using a 0.22 μm \nMillex-GV PVDF syringe filter (SLGVR33RS, MilliporeSigma, Darmstadt, Germany) into a 1.5 mL \nEppendorf tube stacked in a cold block (-20 °C), and then stored at -70 °C unUl it was analysed. \nThe exometabolome LC-MS analyses were conducted by the UNIL Metabolomics facility, \nfollowing the procedure described above.  \n \nDNA extrac)on of samples from selected isolates and chemostat culture samples \nThe DNA extracUon of 14 different samples from the chemostat cultures (#37A D44 M2 Pellet, \n#38A D44 M3 Pellet, #39A D44 M4 Pellet, #28B D30 M5 Pellet, #29B D30 M6 Pellet, #30B D30 \nM7 Pellet) and the selected isolates (sucB-M5-D30-4, sucB-M6-D30-6, sucB-M7-D30-4, sucB-\nAncestor, aceE-M2-D44-2, aceE-M3-D44-3, aceE-M4-D44-1, aceE-Ancestor) was performed by \nadapUng the protocol from the FastPure Bacteria DNA IsolaUon Mini Kit (DC103-01, Vazyme, \nNanjing, China). 1 mL GA buffer was added to the thawed pellet samples, and 230 µL of these \nresuspensions were kept in Eppendorf tubes for DNA extracUon. The rest was centrifuged (4000 \nrpm, 6 min, RT) and the pellet stored at –20 °C. Genomic DNA (gDNA) was extracted from the \n230 µL sample following the manufacturer’s protocol. In the final step, eluUon of the DNA was \nperformed twice with fresh eluUon buffer, eluUng a total of 100 µL DNA. Extracted gDNA was \nstored at –70 °C unUl sequencing. \n \nGenomic sequencing processing \nGenomic DNA from the selected isolates and chemostat samples was sequenced by Novogene \n(Planegg, Germany) using the Illumina NovaSeqX PE 150. The isolates and the bioreactor \nsamples were sequenced to 1 Gb and 10 Gb sequencing depth, respecUvely. Good quality of \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nboth datasets was ensured using fastqc v. 0.12.1 before trimming the reads with fastp v. 0.24.0 \n(--qualified_quality_phred 20 --cut_front --cut_tail --average_qual 20 --length_required 50) (93, \n94). Both read sets were mapped against the E. coli BW25113 reference genome (NCBI RefSeq \nassembly GCF_000750555.1) with minimap2’s v. 2.28 default “sr” parameters (95). The resulUng \nalignments were filtered with samtools view v. 1.20 (-b -f 3 -q 60) to filter out reads not properly \nmapped in pairs or mapped to the reference with a MAPQ lower than 60 (96). Average coverage \nwas assessed for both datasets showing a minimum average coverage for the chemostat culture \nsamples of 2017 and 190 for the sequenced isolates. \nFor both datasets, variants were idenUfied with freebayes v. 1.3.6 (--min-alternate-count 3 -p 1 -\n-min-alternate-fracUon 0.05 --pooled-conUnuous --haplotype-length 0) and annotated with \nsnpEff v. 5.0 using the BW25113 reference genome (97, 98). A custom python script was used to \nfilter the annotated variants. For the isolates, variants with a frequency lower than 0.95 and a \ncoverage lower than 100 were filtered out. For the chemostat culture samples, variants with a \nfrequency lower than 0.05 and coverage lower than 100 were filtered out. A variant was \nconsidered fixed if its alle frequency was at least 0.9.  \nPaired intracellular and extracellular metabolomics and live/dead staining \nM9 medium with 20 mM galactose, 30 mM malate or 40 mM L-alanine was prepared as \ndetailed above. E. coli K-12 MG1655 was precultured on LB agar plates and in liquid LB as \ndetailed above. 1 mL preculture was washed as described above and used to start a second \npreculture in 20 mL M9 + 40 mM L-alanine, inoculated to OD600 = 0.05. Another LB preculture of \nE. coli K-12 MG1655 was used to inoculate a second M9 + 20 mM galactose and a M9 + 30 mM \nmalate preculture in the same way 12 hours later. Aser 36/24 hours, 1 mL was collected from \neach of the three precultures and washed twice as detailed above. These washed precultures \nwere used to inoculate 10 mL of the same media (M9 + either 20 mM galactose, 30 mM malate \nor 40 mM L-alanine) in straight glass tubes (15.8 cm tall, 1.5 cm diameter). These glass tubes \nwere incubated for two days (200 rpm, 37 °C, 45° angle). Growth was monitored by measuring \nOD600 directly in the glass tubes (NANOCOLOR VIS II, MACHEREY-NAGEL, Düren, Germany) at 21 \nUmepoints. The glass tubes were vortexed briefly before any OD reading or sampling.  \nA reference exometabolome 0.5 mL sample was collected aser 2 hours following the procedure \ndetailed below. Then, 2 mL samples for paired intra- and extracellular metabolome and \nlive/dead analyses were collected in the late exponenUal phase (at OD600 ≈ 1.1, Fig. S30). \nExometabolome samples were obtained by filtering 0.5 mL culture using a 0.22 μm Millex-GV \nPVDF syringe filter (SLGVR33RS, MilliporeSigma) into a 1.5 mL Eppendorf tube stacked in a cold \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nblock (-20 °C). The samples for intracellular metabolomics, i.e. cells, were obtained immediately \naser following this procedure: A 25 mm 0.22 μm Durapore PVDF filter (GVWP02500, \nMilliporeSigma) was placed in a new filter holder. A 5 mL syringe was connected to the inlet of \nthe filter holder and a vacuum pump (250 mbar) was connected to the outlet of the filter \nholder. Then, 2 mL of 37 °C MQ water was added to pre-wet the filter. Immediately aser, 1 mL \nof culture sample was added to the syringe and allowed to pass through the filter. If the culture \ndid not pass through solely based on the vacuum pump, light addiUonal pressure was added by \nusing the syringe itself to push the culture through the filter. This process took 10-30 seconds. \nSUll using the same set-up, the filter was finally rinsed with 5 mL of 37 °C MQ water as \npreviously recommended (49). Two tweezers were quickly rinsed in MQ water and used to \ncarefully transfer the filter from the filter holder to 2 mL lysis tubes filled with 1.4 mL of 80:20 \nmethanol:water (v/v), pre-chilled to -20 °C and kept on a cold block also pre-chilled to -20 °C. All \nsamples were then quickly transferred to -70 °C for storage. Live/dead staining was performed \nusing a similar approach as previous work (99): Briefly, the live sample was made by diluUng 10 \nμL culture in 990 μL PBS and the dead sample was made by mixing 100 μL culture with 1 mL \n70% isopropanol. Both samples were vortexed and incubated at room temperature for 1 h. \nThen, the isopropanol was removed from the dead sample by centrifugaUon (8000 rcf, 4 min, \nRT), pipe‡ng off the supernatant, resuspending the pellet in 1 mL PBS and vortexing. This \nprocedure was repeated once. Then, both the live and the dead sample were stained with \npropidium iodide (PI; P4170, Sigma-Aldrich) and SYBR green (Invitrogen S7563, Thermo Fisher \nScienUfic, Carlsbad, CA, USA) with the following procedure: 48 μL sample (live or dead) was \nmixed in a 96-well plate with 49 μL PBS, 2 μL 0.5 mg/mL PI and 1 μL 100X SYBR green and \nincubated for 15 min in the dark. Then, a 0.1X samples of both the stained live and the stained \ndead samples were created by diluUng 10 μL of each sample in 90 μL of PBS. The stained \nlive/dead samples were then analyzed immediately on a CytoFLEX S flow cytometer (Beckman \nCoulter, Brea, CA, USA) and the CytExpert sosware (v2.4.0.28). The dead samples were used to \nmake the gaUngs required to quanUfy the fracUon of dead cells in the live samples (Table S6). \nThe extra- and intracellular metabolomics samples were analysed by the metabolomics facility \nat UNIL to quanUfy the five metabolites leucine, glutamate, aspartate, fructose 6-phosphate and \ncis-aconitate. These metabolites were chosen because they were chemically and metabolically \ndiverse and likely to be quanUfiable both intracellularly and extracellularly based on previous \nexperience with similar samples. Also, as glutamate is the most abundant intracellular \nmetabolite (49, 100), it serves as a best-case scenario in our akempt to test how much of the \nextracellular metabolite levels that can be explained by cell lysis and the associated release of \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\nintracellular metabolites. Intracellular metabolite samples were first normalized by the amount \nof protein detected in the same sample and subsequently converted to absolute intracellular \nconcentraUons using the esUmated protein density in E. coli (13.5 ⋅ 10'( µg µm)⁄ ) (101). The \nfructose 6-phosphate value for sample 1B, Umepoint 1, was discarded as an outlier as it was one \norder of magnitude larger than any other fructose 6-phosphate measurements.  \n \nCalcula)on of contribu)on from intracellular metabolites and cell lysis to extracellular \nconcentra)ons \nApproximate cell volume was esUmated from measured OD600 and a previous esUmate of OD to \ncell volume for E. coli of 3.6 μL/ΟD600/mL (102). From approximate cell volume, measured \nabsolute intracellular concentraUons and cell lysis fracUons we then calculated the \ncorresponding change in extracellular concentraUons. This potenUal contribuUon was then \ncompared to the net change in extracellular concentraUons between the T0 reference sample \n(aser 2-hours) and the late exponenUal phase sample. If all T0 values were below the detecUon \nlimit, we assumed the iniUal concentraUon of this metabolite to be 0. If only some values at T0 \nwere below the detecUon limit we imputed these values to the mean of the detected values, \ngiving conservaUve esUmates for the net change in extracellular concentraUons. To esUmate a \npotenUal contribuUon from protein depolymerizaUon we esUmated cell density in gDW/L from \nOD600 (0.346 gDW/L/OD600) and from this specific amino acid density (in g/L) from recent \nesUmates of E. coli amino acid mass fracUons (103). Where mass fracUons were akributed to \npairs of amino acids (glutamate/glutamine and aspartate/asparagine) we assumed equal \ncontribuUons. Amino acid density was converted from g/L to μΜ by dividing by the amino acids’ \nmolecular weight subtracted the weight of a water molecule to account for the condensaUon \nreacUon in protein polymerizaUon. A similar procedure as described above was followed when \nwe extended this analysis by using literature values for intracellular concentraUons. When \nextracellular concentraUons were reported as two indisUnguishable molecules, we akributed \nequal amounts to the two metabolites. This includes 2-phosphoglycerate/3-phosphoglycerate \nand ribulose 5-phosphate/xylulose 5-phosphate. To idenUfy typical biomass degradaUon \nproducts we used the metabolites consumed by the biomass equaUon of the E. coli model \niML1515 (89), although excluding the soluble pool (as categorized in (87)) since this should \ncorrespond to the intracellular concentraUons already accounted for. Of the metabolites in our \ndataset (Fig. 3B), only the amino acids were among the consumed biomass components and \ntherefore considered to be typical biomass degradaUon products.  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint \n\n \n \n \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted August 19, 2025. ; https://doi.org/10.1101/2025.08.19.671024doi: bioRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}