Evaluation of enzyme-constrained genome-scale model through metabolic engineering of anaerobic co-production of 2,3-butanediol and glycerol by Saccharomyces cerevisiae.

Enzyme-constrained genome-scale models (ecGEMs) have potential to predict phenotypes in a variety of conditions, such as growth rates or carbon sources. This study investigated if ecGEMs can guide metabolic engineering efforts to swap anaerobic redox-neutral ATP-providing pathways in yeast from alcoholic fermentation to equimolar co-production of 2,3-butanediol and glycerol. With proven pathways and low product toxicity, the ecGEM solution space aligned well with observed phenotypes. Since this catabolic pathway provides only one-third of the ATP of alcoholic fermentation (2/3 versus 2 ATP per glucose), the ecGEM predicted a growth decrease from 0.36 h-1 in the reference to 0.175 h-1 in the engineered strain. However, this <3-fold decrease would require the specific glucose consumption rate to increase. Surprisingly, after the pathway swap the engineered strain immediately grew at 0.15 h-1 with a glucose consumption rate of 29 mmol (g CDW)-1 h-1, which was indeed higher than reference (23 mmol (g CDW)-1 h-1) and one of the highest reported for S. cerevisiae. The accompanying 2,3-butanediol- (15.8 mmol (g CDW)-1 h-1) and glycerol (19.6 mmol (g CDW)-1 h-1) production rates were close to predicted values. Proteomics confirmed that this increased consumption rate was facilitated by enzyme reallocation from especially ribosomes (from 25.5 to 18.5 %) towards glycolysis (from 28.7 to 43.5 %). Subsequently, 200 generations of sequential transfer did not improve growth of the engineered strain, showing the use of ecGEMs in predicting opportunity space for laboratory evolution. The observations in this study illustrate both the current potential, as well as future improvements, of ecGEMs as a tool for both metabolic engineering and laboratory evolution.

Genome-scale metabolic modeling reveals metabolic trade-offs associated with lipid production in Rhodotorula toruloides.

Rhodotorula toruloides is a non-conventional, oleaginous yeast able to naturally accumulate high amounts of microbial lipids. Constraint-based modeling of R. toruloides has been mainly focused on the comparison of experimentally measured and model predicted growth rates, while the intracellular flux patterns have been analyzed on a rather general level. Hence, the intrinsic metabolic properties of R. toruloides that make lipid synthesis possible are not thoroughly understood. At the same time, the lack of diverse physiological data sets has often been the bottleneck to predict accurate fluxes. In this study, we collected detailed physiology data sets of R. toruloides while growing on glucose, xylose, and acetate as the sole carbon source in chemically defined medium. Regardless of the carbon source, the growth was divided into two phases from which proteomic and lipidomic data were collected. Complemental physiological parameters were collected in these two phases and altogether implemented into metabolic models. Simulated intracellular flux patterns demonstrated the role of phosphoketolase in the generation of acetyl-CoA, one of the main precursors during lipid biosynthesis, while the role of ATP citrate lyase was not confirmed. Metabolic modeling on xylose as a carbon substrate was greatly improved by the detection of chirality of D-arabinitol, which together with D-ribulose were involved in an alternative xylose assimilation pathway. Further, flux patterns pointed to metabolic trade-offs associated with NADPH allocation between nitrogen assimilation and lipid biosynthetic pathways, which was linked to large-scale differences in protein and lipid content. This work includes the first extensive multi-condition analysis of R. toruloides using enzyme-constrained models and quantitative proteomics. Further, more precise kcat values should extend the application of the newly developed enzyme-constrained models that are publicly available for future studies.

Detailed analysis of metabolism reveals growth-rate-promoting interactions between Anaerostipes caccae and Bacteroides spp.

INTRODUCTION: Human gut microbiota species which are next-generation probiotics (NGPs) candidates are of high interest as they have shown the potential to treat intestinal inflammation and other diseases. Unfortunately, these species are often not robust enough for large-scale cultivation, especially in maintaining diversity in co-culture production. OBJECTIVES: In this study, we describe interactions between human gut microbiota species in the cultivation process with unique substrates. We also demonstrated that it is possible to change the species ratio in co-culture by changing the ratio of carbon sources. METHODS: We screened 25 different bacterial species based on their metabolic capabilities. After evaluating unique substrate possibilities, we chose Anaerostipes caccae (A. caccae), Bacteroides thetaiotaomicron (B. thetaiotaomicron), and Bacteroides vulgatus (B. vulgatus) as subjects for further study. D-sorbitol, D-xylose, and D-galacturonic acid were selected as substrates for A. caccae, B. thetaiotaomicron, and B. vulgatus respectively. All three species were cultivated as both monocultures and in co-cultures in serial batch fermentations in an isothermal microcalorimeter. RESULTS: Positive interactions were detected between the species in both co-cultures (A. caccae + B. thetaiotaomicron; A. caccae + B. vulgatus) resulting in higher heat production compared to the sum of the monocultures. The same positive cross-feeding interactions took place in larger-scale cultivation experiments. We confirmed acetate and lactate cross-feeding between A. caccae and B. thetaiotaomicron with flux balance analysis (FBA). CONCLUSION: Changing the ratio of the selected carbon sources in the medium changed the species ratio accordingly. Such robustness is the basis for developing more efficient industrial co-culture processes including the production of NGPs.

A Conundrum of r-Protein Stability: Unbalanced Stoichiometry of r-Proteins during Stationary Phase in Escherichia coli.

Bacterial ribosomes are composed of three rRNA and over 50 ribosomal protein (r-protein) molecules. r-proteins are essential for ribosome assembly and structural stability and also participate in almost all ribosome functions. Ribosomal components are present in stoichiometric amounts in the mature 70S ribosomes during exponential and early stationary growth phases. Ribosomes are degraded in stationary phase; however, the stability and fate of r-proteins during stationary growth phase are not known. In this study, we report a quantitative analysis of ribosomal components during extended stationary-phase growth in Escherichia coli. We show that (i) the quantity of ribosomes per cell mass decreases in stationary phase, (ii) 70S ribosomes contain r-proteins in stoichiometric amounts, (iii) 30S subunits are degraded faster than 50S subunits, (iv) the quantities of 21 r-proteins in the total proteome decrease during 14 days (short-lived r-proteins) concomitantly with the reduction of cellular RNA, and (e) 30 r-proteins are stable and form a pool of free r-proteins (stable r-proteins). Thus, r-proteins are present in nonstoichiometric amounts in the proteome of E. coli during the extended stationary phase. IMPORTANCE Ribosome degradation has been extensively described from the viewpoint of its main component, rRNA. Here, we aim to complement our knowledge by quantitatively analyzing r-protein degradation and stability both in the ribosomes and in the whole-cell proteome during stationary phase in E. coli. r-proteins are considered to be very stable in the proteome. Here, we show that a specific set of r-proteins are rapidly degraded after release from the rRNA. The degradation of r-proteins is an intriguing new aspect of r-protein metabolism in bacteria.

Development of a dedicated Golden Gate Assembly Platform (RtGGA) for Rhodotorula toruloides.

Rhodotorula toruloides is a potential chassis for microbial cell factories as this yeast can metabolise different substrates into a diverse range of natural products, but the lack of efficient synthetic biology tools hinders its applicability. In this study, the modular, versatile and efficient Golden Gate DNA assembly system (RtGGA) was adapted to the first basidiomycete, an oleaginous yeast R. toruloides. R. toruloides CCT 0783 was sequenced, and used for the GGA design. The DNA fragments were assembled with predesigned 4-nt overhangs and a library of standardized parts was created containing promoters, genes, terminators, insertional regions, and resistance genes. The library was combined to create cassettes for the characterization of promoters strength and to overexpress the carotenoid production pathway. A variety of reagents, plasmids, and strategies were used and the RtGGA proved to be robust. The RtGGA was used to build three versions of the carotenoid overexpression cassette by using different promoter combinations. The cassettes were transformed into R. toruloides and the three new strains were characterized. Total carotenoid concentration increased by 41%. The dedicated GGA platform fills a gap in the advanced genome engineering toolkit for R. toruloides, enabling the efficient design of complex metabolic pathways.

Screening and Growth Characterization of Non-conventional Yeasts in a Hemicellulosic Hydrolysate.

Lignocellulosic biomass is an attractive raw material for the sustainable production of chemicals and materials using microbial cell factories. Most of the existing bioprocesses focus on second-generation ethanol production using genetically modified Saccharomyces cerevisiae, however, this microorganism is naturally unable to consume xylose. Moreover, extensive metabolic engineering has to be carried out to achieve high production levels of industrially relevant building blocks. Hence, the use of non-Saccharomyces species, or non-conventional yeasts, bearing native metabolic routes, allows conversion of a wide range of substrates into different products, and higher tolerance to inhibitors improves the efficiency of biorefineries. In this study, nine non-conventional yeast strains were selected and screened on a diluted hemicellulosic hydrolysate from Birch. Kluyveromyces marxianus CBS 6556, Scheffersomyces stipitis CBS 5773, Lipomyces starkeyi DSM 70295, and Rhodotorula toruloides CCT 7815 were selected for further characterization, where their growth and substrate consumption patterns were analyzed under industrially relevant substrate concentrations and controlled environmental conditions in bioreactors. K. marxianus CBS 6556 performed poorly under higher hydrolysate concentrations, although this yeast was determined among the fastest-growing yeasts on diluted hydrolysate. S. stipitis CBS 5773 demonstrated a low growth and biomass production while consuming glucose, while during the xylose-phase, the specific growth and sugar co-consumption rates were among the highest of this study (0.17 h-1 and 0.37 g/gdw*h, respectively). L. starkeyi DSM 70295 and R. toruloides CCT 7815 were the fastest to consume the provided sugars at high hydrolysate conditions, finishing them within 54 and 30 h, respectively. R. toruloides CCT 7815 performed the best of all four studied strains and tested conditions, showing the highest specific growth (0.23 h-1), substrate co-consumption (0.73 ± 0.02 g/gdw*h), and xylose consumption (0.22 g/gdw*h) rates. Furthermore, R. toruloides CCT 7815 was able to produce 10.95 ± 1.37 gL-1 and 1.72 ± 0.04 mgL-1 of lipids and carotenoids, respectively, under non-optimized cultivation conditions. The study provides novel information on selecting suitable host strains for biorefinery processes, provides detailed information on substrate consumption patterns, and pinpoints to bottlenecks possible to address using metabolic engineering or adaptive evolution experiments.

Potassium and Sodium Salt Stress Characterization in the Yeasts Saccharomyces cerevisiae, Kluyveromyces marxianus, and Rhodotorula toruloides.

Biotechnology requires efficient microbial cell factories. The budding yeast Saccharomyces cerevisiae is a vital cell factory, but more diverse cell factories are essential for the sustainable use of natural resources. Here, we benchmarked nonconventional yeasts Kluyveromyces marxianus and Rhodotorula toruloides against S. cerevisiae strains CEN.PK and W303 for their responses to potassium and sodium salt stress. We found an inverse relationship between the maximum growth rate and the median cell volume that was responsive to salt stress. The supplementation of K+ to CEN.PK cultures reduced Na+ toxicity and increased the specific growth rate 4-fold. The higher K+ and Na+ concentrations impaired ethanol and acetate metabolism in CEN.PK and acetate metabolism in W303. In R. toruloides cultures, these salt supplementations induced a trade-off between glucose utilization and cellular aggregate formation. Their combined use increased the beta-carotene yield by 60% compared with that of the reference. Neural network-based image analysis of exponential-phase cultures showed that the vacuole-to-cell volume ratio increased with increased cell volume for W303 and K. marxianus but not for CEN.PK and R. toruloides in response to salt stress. Our results provide insights into common salt stress responses in yeasts and will help design efficient bioprocesses. IMPORTANCE Characterization of microbial cell factories under industrially relevant conditions is crucial for designing efficient bioprocesses. Salt stress, typical in industrial bioprocesses, impinges upon cell volume and affects productivity. This study presents an open-source neural network-based analysis method to evaluate volumetric changes using yeast optical microscopy images. It allows quantification of cell and vacuole volumes relevant to cellular physiology. On applying salt stress in yeasts, we found that the combined use of K+ and Na+ improves the cellular fitness of Saccharomyces cerevisiae strain CEN.PK and increases the beta-carotene productivity in Rhodotorula toruloides, a commercially important antioxidant and a valuable additive in foods.

Benchmarking accuracy and precision of intensity-based absolute quantification of protein abundances in Saccharomyces cerevisiae.

Protein quantification via label-free mass spectrometry (MS) has become an increasingly popular method for predicting genome-wide absolute protein abundances. A known caveat of this approach, however, is the poor technical reproducibility, that is, how consistent predictions are when the same sample is measured repeatedly. Here, we measured proteomics data for Saccharomyces cerevisiae with both biological and inter-batch technical triplicates, to analyze both accuracy and precision of protein quantification via MS. Moreover, we analyzed how these metrics vary when applying different methods for converting MS intensities to absolute protein abundances. We demonstrate that our simple normalization and rescaling approach can perform as accurately, yet more precisely, than methods which rely on external standards. Additionally, we show that inter-batch reproducibility is worse than biological reproducibility for all evaluated methods. These results offer a new benchmark for assessing MS data quality for protein quantification, while also underscoring current limitations in this approach.

Metabolism Control in 3D-Printed Living Materials Improves Fermentation.

The three-dimensional (3D) printing of cell-containing polymeric hydrogels creates living materials (LMs), offering a platform for developing innovative technologies in areas like biosensors and biomanufacturing. The polymer material properties of cross-linkable F127-bis-urethane methacrylate (F127-BUM) allow reproducible 3D printing and stability in physiological conditions, making it suitable for fabricating LMs. Though F127-BUM-based LMs permit diffusion of solute molecules like glucose and ethanol, it remains unknown whether these are permissible for oxygen, essential for respiration. To determine oxygen permissibility, we quantified dissolved oxygen consumption by the budding yeast-laden F127-BUM-based LMs. Moreover, we obtained data on cell-retaining LMs, which allowed a direct comparison between LMs and suspension cultures. We further developed a highly reliable method to isolate cells from LMs for flow cytometry analysis, cell viability evaluation, and the purification of macromolecules. We found oxygen consumption heavily impaired inside LMs, indicating that yeast metabolism relies primarily on fermentation instead of respiration. Applying this finding to brewing, we observed a higher (3.7%) ethanol production using LMs than the traditional brewing process, indicating improved fermentation. Our study concludes that the present F127-BUM-based LMs are useful for microaerobic processes but developing aerobic bioprocesses will require further research.

Xylose Metabolism and the Effect of Oxidative Stress on Lipid and Carotenoid Production in Rhodotorula toruloides: Insights for Future Biorefinery.

The use of cell factories to convert sugars from lignocellulosic biomass into chemicals in which oleochemicals and food additives, such as carotenoids, is essential for the shift toward sustainable processes. Rhodotorula toruloides is a yeast that naturally metabolises a wide range of substrates, including lignocellulosic hydrolysates, and converts them into lipids and carotenoids. In this study, xylose, the main component of hemicellulose, was used as the sole substrate for R. toruloides, and a detailed physiology characterisation combined with absolute proteomics and genome-scale metabolic models was carried out to understand the regulation of lipid and carotenoid production. To improve these productions, oxidative stress was induced by hydrogen peroxide and light irradiation and further enhanced by adaptive laboratory evolution. Based on the online measurements of growth and CO2 excretion, three distinct growth phases were identified during batch cultivations. Majority of the intracellular flux estimations showed similar trends with the measured protein levels and demonstrated improved NADPH regeneration, phosphoketolase activity and reduced ß-oxidation, correlating with increasing lipid yields. Light irradiation resulted in 70% higher carotenoid and 40% higher lipid content compared to the optimal growth conditions. The presence of hydrogen peroxide did not affect the carotenoid production but culminated in the highest lipid content of 0.65 g/gDCW. The adapted strain showed improved fitness and 2.3-fold higher carotenoid content than the parental strain. This work presents a holistic view of xylose conversion into microbial oil and carotenoids by R. toruloides, in a process toward renewable and cost-effective production of these molecules.

Physical Confinement Impacts Cellular Phenotypes within Living Materials.

Additive manufacturing allows three-dimensional printing of polymeric materials together with cells, creating living materials for applications in biomedical research and biotechnology. However, an understanding of the cellular phenotype within living materials is lacking, which is a key limitation for their wider application. Herein, we present an approach to characterize the cellular phenotype within living materials. We immobilized the budding yeast Saccharomyces cerevisiae in three different photo-cross-linkable triblock polymeric hydrogels containing F127-bis-urethane methacrylate, F127-dimethacrylate, or poly(alkyl glycidyl ether)-dimethacrylate. Using optical and scanning electron microscopy, we showed that hydrogels based on these polymers were stable under physiological conditions, but yeast colonies showed differences in the interaction within the living materials. We found that the physical confinement, imparted by compositional and structural properties of the hydrogels, impacted the cellular phenotype by reducing the size of cells in living materials compared with suspension cells. These properties also contributed to the differences in immobilization patterns, growth of colonies, and colony coatings. We observed that a composition-dependent degradation of polymers was likely possible by cells residing in the living materials. In conclusion, our investigation highlights the need for a holistic understanding of the cellular response within hydrogels to facilitate the synthesis of application-specific polymers and the design of advanced living materials in the future.

Cell-Laden Hydrogels for Multikingdom 3D Printing.

Living materials are created through the embedding of live, whole cells into a matrix that can house and sustain the viability of the encapsulated cells. Through the immobilization of these cells, their bioactivity can be harnessed for applications such as bioreactors for the production of high-value chemicals. While the interest in living materials is growing, many existing materials lack robust structure and are difficult to pattern. Furthermore, many living materials employ only one type of microorganism, or microbial consortia with little control over the arrangement of the various cell types. In this work, a Pluronic F127-based hydrogel system is characterized for the encapsulation of algae, yeast, and bacteria to create living materials. This hydrogel system is also demonstrated to be an excellent material for additive manufacturing in the form of direct write 3D-printing to spatially arrange the cells within a single printed construct. These living materials allow for the development of incredibly complex, immobilized consortia, and the results detailed herein further enhance the understanding of how cells behave within living material matrices. The utilization of these materials allows for interesting applications of multikingdom microbial cultures in immobilized bioreactor or biosensing technologies.

C/N ratio and carbon source-dependent lipid production profiling in Rhodotorula toruloides.

Microbial oils are lipids produced by oleaginous microorganisms, which can be used as a potential feedstock for oleochemical production. The oleaginous yeast Rhodotorula toruloides can co-produce microbial oils and high-value compounds from low-cost substrates, such as xylose and acetic acid (from hemicellulosic hydrolysates) and raw glycerol (a byproduct of biodiesel production). One step towards economic viability is identifying the best conditions for lipid production, primarily the most suitable carbon-to-nitrogen ratio (C/N). Here, we aimed to identify the best conditions and cultivation mode for lipid production by R. toruloides using various low-cost substrates and a range of C/N ratios (60, 80, 100, and 120). Turbidostat mode was used to achieve a steady state at the maximal specific growth rate and to avoid continuously changing environmental conditions (i.e., C/N ratio) that inherently occur in batch mode. Regardless of the carbon source, higher C/N ratios increased lipid yields (up to 60% on xylose at a C/N of 120) but decreased the specific growth rate. Growth on glycerol resulted in the highest specific growth and lipid production (0.085 g lipids/gDW*h) rates at C/Ns between 60 and 100. We went on to study lipid production using glycerol in both batch and fed-batch modes, which resulted in lower specific lipid production rates compared with turbisdostat, however, fed batch is superior in terms of biomass production and lipid titers. By combining the data we obtained in these experiments with a genome-scale metabolic model of R. toruloides, we identified targets for improvements in lipid production that could be carried out either by metabolic engineering or process optimization.

Transcriptome analysis of the thermotolerant yeast Kluyveromyces marxianus CCT 7735 under ethanol stress.

The thermotolerant yeast Kluyveromyces marxianus displays a potential to be used for ethanol production from both whey and lignocellulosic biomass at elevated temperatures, which is highly alluring to reduce the cost of the bioprocess. Nevertheless, contrary to Saccharomyces cerevisiae, K. marxianus cannot tolerate high ethanol concentrations. We report the transcriptional profile alterations in K. marxianus under ethanol stress in order to gain insights about mechanisms involved with ethanol response. Time-dependent changes have been characterized under the exposure of 6% ethanol and compared with the unstressed cells prior to the ethanol addition. Our results reveal that the metabolic flow through the central metabolic pathways is impaired under the applied ethanol stress. Consistent with these results, we also observe that genes involved with ribosome biogenesis are downregulated and gene-encoding heat shock proteins are upregulated. Remarkably, the expression of some gene-encoding enzymes related to unsaturated fatty acid and ergosterol biosynthesis decreases upon ethanol exposure, and free fatty acid and ergosterol measurements demonstrate that their content in K. marxianus does not change under this stress. These results are in contrast to the increase previously reported with S. cerevisiae subjected to ethanol stress and suggest that the restructuration of K. marxianus membrane composition differs in the two yeasts which gives important clues to understand the low ethanol tolerance of K. marxianus compared to S. cerevisiae.

Improving the phenotype predictions of a yeast genome-scale metabolic model by incorporating enzymatic constraints.

Genome-scale metabolic models (GEMs) are widely used to calculate metabolic phenotypes. They rely on defining a set of constraints, the most common of which is that the production of metabolites and/or growth are limited by the carbon source uptake rate. However, enzyme abundances and kinetics, which act as limitations on metabolic fluxes, are not taken into account. Here, we present GECKO, a method that enhances a GEM to account for enzymes as part of reactions, thereby ensuring that each metabolic flux does not exceed its maximum capacity, equal to the product of the enzyme's abundance and turnover number. We applied GECKO to a Saccharomyces cerevisiae GEM and demonstrated that the new model could correctly describe phenotypes that the previous model could not, particularly under high enzymatic pressure conditions, such as yeast growing on different carbon sources in excess, coping with stress, or overexpressing a specific pathway. GECKO also allows to directly integrate quantitative proteomics data; by doing so, we significantly reduced flux variability of the model, in over 60% of metabolic reactions. Additionally, the model gives insight into the distribution of enzyme usage between and within metabolic pathways. The developed method and model are expected to increase the use of model-based design in metabolic engineering.

The yeast osmostress response is carbon source dependent.

Adaptation to altered osmotic conditions is a fundamental property of living cells and has been studied in detail in the yeast Saccharomyces cerevisiae. Yeast cells accumulate glycerol as compatible solute, controlled at different levels by the High Osmolarity Glycerol (HOG) response pathway. Up to now, essentially all osmostress studies in yeast have been performed with glucose as carbon and energy source, which is metabolised by glycolysis with glycerol as a by-product. Here we investigated the response of yeast to osmotic stress when yeast is respiring ethanol as carbon and energy source. Remarkably, yeast cells do not accumulate glycerol under these conditions and it appears that trehalose may partly take over the role as compatible solute. The HOG pathway is activated in very much the same way as during growth on glucose and is also required for osmotic adaptation. Slower volume recovery was observed in ethanol-grown cells as compared to glucose-grown cells. Dependence on key regulators as well as the global gene expression profile were similar in many ways to those previously observed in glucose-grown cells. However, there are indications that cells re-arrange redox-metabolism when respiration is hampered under osmostress, a feature that could not be observed in glucose-grown cells.

Absolute Quantification of Protein and mRNA Abundances Demonstrate Variability in Gene-Specific Translation Efficiency in Yeast.

Protein synthesis is the most energy-consuming process in a proliferating cell, and understanding what controls protein abundances represents a key question in biology and biotechnology. We quantified absolute abundances of 5,354 mRNAs and 2,198 proteins in Saccharomyces cerevisiae under ten environmental conditions and protein turnover for 1,384 proteins under a reference condition. The overall correlation between mRNA and protein abundances across all conditions was low (0.46), but for differentially expressed proteins (n = 202), the median mRNA-protein correlation was 0.88. We used these data to model translation efficiencies and found that they vary more than 400-fold between genes. Non-linear regression analysis detected that mRNA abundance and translation elongation were the dominant factors controlling protein synthesis, explaining 61% and 15% of its variance. Metabolic flux balance analysis further showed that only mitochondrial fluxes were positively associated with changes at the transcript level. The present dataset represents a crucial expansion to the current resources for future studies on yeast physiology.

Adaptation to different types of stress converge on mitochondrial metabolism.

Yeast cell factories encounter physical and chemical stresses when used for industrial production of fuels and chemicals. These stresses reduce productivity and increase bioprocess costs. Understanding the mechanisms of the stress response is essential for improving cellular robustness in platform strains. We investigated the three most commonly encountered industrial stresses for yeast (ethanol, salt, and temperature) to identify the mechanisms of general and stress-specific responses under chemostat conditions in which specific growth rate-dependent changes are eliminated. By applying systems-level analysis, we found that most stress responses converge on mitochondrial processes. Our analysis revealed that stress-specific factors differ between applied stresses; however, they are underpinned by an increased ATP demand. We found that when ATP demand increases to high levels, respiration cannot provide sufficient ATP, leading to onset of respirofermentative metabolism. Although stress-specific factors increase ATP demand for cellular growth under stressful conditions, increased ATP demand for cellular maintenance underpins a general stress response and is responsible for the onset of overflow metabolism.