Yeast batch fermentation is widely used in industrial biotechnology, yet its performance is strongly influenced by temperature and nitrogen availability, which affect growth kinetics and metabolite production. The development of predictive models that accurately describe these effects is essential for automating and optimizing fermentation design, reducing trial-and-error experimentation, and improving process efficiency and product quality. However, most mathematical models focus on primary metabolism and lack a systematic approach to integrate the effects of temperature. Existing models often rely on empirical corrections with limited predictive power beyond specific experimental conditions. Furthermore, there is no unified framework for optimizing fermentation processes while accounting for the temperature-dependent metabolic responses. We addressed these gaps by developing a temperature-dependent kinetic model for nitrogen-limited batch fermentation by Saccharomyces cerevisiae. The modeling approach is based on advanced systems identification, integrating identifiability analyses (structural and practical), multi-experiment parameter estimation, and automated model selection to determine the most appropriate temperature dependencies for key metabolic processes. Validated across five industrial S. cerevisiae strains in an illustrative example related to wine fermentation, the model exhibited strong predictive performance (NRMSE <10.5%, median R2>0.95) and enabled simulation-based process optimization, including nitrogen-supplementation strategies and strain selection for improved fermentation outcomes. By providing a systematic modeling framework that accounts for temperature effects, this work bridges a critical gap in predictive modeling and advances the rational design and control of industrial fermentation processes.

ABSTRACT During batch fermentation, a variety of compounds are synthesized, as microorganisms undergo distinct growth phases: lag, exponential, growth-no-growth transition, stationary, and decay. A detailed understanding of the metabolic pathways involved in these phases is crucial for optimizing the production of target compounds. Dynamic flux balance analysis (dFBA) offers insight into the dynamics of metabolic pathways. However, explaining secondary metabolism remains a challenge. A multiphase and multi-objective dFBA scheme (MPMO model) has been proposed for this purpose. However, its formulation is discontinuous, changing from phase to phase; its accuracy in predicting intracellular fluxes is hampered by the lack of a mechanistic link between phases; and its simulation requires considerable computational effort. To address these limitations, we combine a novel model with a genome-scale model to predict the distribution of intracellular fluxes throughout batch fermentation. This integrated multiphase continuous model (IMC) has a unique formulation over time, and it incorporates empirical regulatory descriptions to automatically identify phase transitions and incorporates the hypotheses that yeasts might vary their cellular objective over time to adapt to the changing environment. We validated the predictive capacity of the IMC model by comparing its predictions with intracellular metabolomics data for Saccharomyces uvarum during batch fermentation. The model aligns well with the data, confirming its predictive capabilities. Notably, the IMC model accurately predicts trehalose accumulation, which was enforced in the MPMO model. We further demonstrate the generalizability of the IMC model, explaining the dynamics of primary and secondary metabolism of three Saccharomyces species. The model provides biological insights consistent with the literature and metabolomics data, establishing it as a valuable tool for exploring the dynamics of novel fermentation processes. IMPORTANCE This work presents an integrated multiphase continuous dynamic genome-scale model (IMC model) for batch fermentation, a crucial process widely used in industry to produce biofuels, enzymes, pharmaceuticals, and food products or ingredients. The IMC model integrates a continuous kinetic model with a genome-scale model to address the critical limitations of existing dynamic flux balance analysis schemes, such as the difficulty of explaining secondary metabolism, the lack of mechanistic links between growth phases, or the high computational demands. The model also introduces the hypothesis that cells adapt the FBA objective over time. The IMC improves the accuracy of intracellular flux predictions and simplifies the implementation process with a unique dFBA formulation over time. Its ability to predict both primary and secondary metabolism dynamics in different Saccharomyces species underscores its versatility and robustness. Furthermore, its alignment with empirical metabolomics data validates its predictive power, offering valuable insights into metabolic processes during batch fermentation. These advances pave the way for optimizing fermentation processes, potentially leading to more efficient production of target compounds and novel biotechnological applications.

The species Saccharomyces uvarum and Saccharomyces kudriavzevii have gained popularity in recent decades due to their interesting oenological properties. However, although it plays a crucial role in yeast fermentation performance and compound synthesis, our understanding of nitrogen metabolism in these species remains limited. Therefore, we compared how three strains of Saccharomyces cerevisiae, Saccharomyces uvarum and Saccharomyces kudriavzevii use relevant nitrogen sources by combining quantitative analysis approaches based on isotopic tracing and modelling. The model we have developed aims to facilitate the calculation and interpretation of stable isotope data for other experiments, by providing easy visualisation of the results and predicting the kinetics of isotope incorporation beyond the sampling points. The three species exhibit significant variations in their nitrogen assimilation profile. They differ in the timing of uptake of ammonium, arginine and glutamine: Saccharomyces cerevisiae prefers glutamine, Saccharomyces kudriavzevii ammonium and Saccharomyces uvarum arginine. This contributes to a different pattern of nitrogen redistribution towards proteinogenic amino acids between strains at the start of the exponential phase, which fades on entering the stationary phase. Additionally, we found that the contribution of leucine and valine to isoamyl alcohol production varies between species; also, Saccharomyces kudriavzevii activates the synthesis of volatile compounds earlier.

Addressing the challenge of achieving a global circular bioeconomy requires efficient and robust bio-based processes operating at different scales. These processes should also be competitive replacements for the production of chemicals currently obtained from fossil resources, as well as for the production of new-to-nature compounds. To that end, genetic circuits can be used to control cellular behavior and are instrumental in developing efficient cell factories. Whole-cell biosensors harbor circuits that can be based on allosteric transcription factors (TFs) to detect and elicit a response depending on the target molecule concentrations. By modifying regulatory elements and testing various genetic components, the responsive behavior of genetic biosensors can be finely tuned and engineered. While previous models have described and characterized the behavior of naringenin biosensors, additional data and resources are required to predict their dynamic response and performance in different contexts, such as under various gene expression regulatory elements, media, carbon sources, or media supplements. Tuning these conditions is pivotal in optimizing biosensor design for applications operating in varying conditions, such as fermentation processes. In this study, we assembled a library of FdeR biosensors, characterized their performance under different conditions, and developed a mechanistic model to describe their dynamic behavior under reference conditions, which guided a machine learning-based predictive model that accounts for context-dependent dynamic parameters. Such a Design-Build-Test-Learn (DBTL) pipeline allowed us to determine optimal condition combinations for the desired biosensor specifications, both for automated screening and dynamic regulation. The findings of this work contribute to a deeper understanding of whole-cell biosensors and their potential for precise measurement, screening, and dynamic regulation of engineered production pathways for valuable molecules.

Batch fermentation is a biotechnological dynamic process that produces various products by employing microorganisms that undergo different growth phases: lag, exponential, growth-non-growth, stationary, and decay. Genome-scale constrained-based models are commonly used to explore the phenotypic potential of these microorganisms.Previous studies have primarily used dynamic Flux Balance Analysis (dFBA) to elucidate the metabolism during the exponential phase. However, this approach falls short in addressing the multi-phase nature of the process and secondary metabolism, posing significant challenges to our understanding of batch fermentation. A recent attempt at a solution was a discontinuous, multi-phase, multi-objective dFBA implementation.However, this approximation lacks the mechanistic connection between phases, limiting its applicability in predicting intracellular fluxes during batch fermentation.To overcome these limitations, we combined a novel continuous model with a genome-scale model to predict the distribution of intracellular fluxes throughout the batch fermentation process. The proposed model includes empirical descriptions of regulation that automatically identify the transition between phases. Its application to explain primary and secondary metabolism of Saccharomyces species in batch fermentation results in biological insights that are in good agreement with the previous literature. The ability to account for all process phases and explain secondary metabolism makes this model a valuable and easy-to-use tool for exploring novel fermentation processes.IMPORTANCE This research proposes a novel dynamic genome-scale modelling approach for batch fermentation, a crucial process widely used to produce a diverse range of products such as biofuels, enzymes, pharmaceuticals, and food products or ingredients.The proposed approach automatically accounts for the transitions between different phases of the fermentation process (lag, exponential, growth-no-growth, and stationary). This is a significant advancement over previous methods that required different model formulations for different phases.We have successfully applied this modelling approach to explore the primary and secondary metabolism of three yeast species under batch fermentation conditions. The model accurately explained experimental data and provided biological insights consistent with previous research findings, instilling confidence in its reliability and accuracy.The ability of this modelling approach to explain primary and secondary metabolism makes it a valuable tool for designing novel, more efficient, and effective fermentation processes, which could have far-reaching implications in industrial biotechnology.

Abstract Crabtree-positive yeast species such as Saccharomyces cerevisiae produce most of their ATP through alcoholic fermentation in the presence of oxygen and abundant glucose. This seemingly inefficient growth strategy leads to substantial industrial losses. Thus, many research efforts have focused on formulating hypotheses and computational growth models to explain this phenomenon, but have not given much attention to the development of dynamic models that integrate growth and mechanistic regulation of protein expression. From a regulatory point of view, this phenotype implies rapid growth and repression of respiratory genes. When intracellular glucose levels are low, SNF1 and the HAP complex control respiration activation. Among the proteins in the HAP complex, only HAP4 is believed not to be constitutively expressed and is considered the signal transduction element in this pathway. The transcription of genes from the translation machinery is primarily associated with TORC1 and PKA. Here, we have formulated an integrated dynamic model of respiro-fermentative growth and regulation of S. cerevisiae using two proteomic data sets obtained during diauxic shift experiments. After analysing the model and data, we concluded that HAP5 is also up-regulated in response to decreasing glucose levels, and this regulation occurs downstream of the TORC1 and PKA pathways. These two hypotheses were validated using independent transcriptomics data sets. Given its architecture, we postulate that the HAP complex integrates information from the PKA/TORC1 and SNF1 pathways. Our findings provide valuable information on the evolution of the Crabtree effect and can help design strategies to reduce its impact.

Genome-scale models (GEMs) are structured representations of a target organism’s metabolism based on existing genetic, biochemical, and physiological information. These models store the available knowledge of the physiology and metabolic behaviour of organisms and summarise this knowledge in a mathematical description. Flux balance analysis uses GEMs to make predictions about cellular metabolism through the solution of a constrained optimisation problem. The gene inactivity moderated by metabolism and expression (GIMME) approach further constrains FBA by means of transcriptomics data. The underlying idea is to deactivate those reactions for which transcriptomics is below a given threshold. GIMME uses a unique threshold for the entire cell. Therefore, non-essential reactions can be deactivated, even if they are required to meet the production of a certain external metabolite, because of their low associated transcript expression values. Here, we propose a new approach to enable the selection of different transcriptomics thresholds for different cell compartments or modules, such as cellular organelles and specific metabolic pathways. The approach was compared with the original GIMME in the analysis of a number of examples related to yeast batch fermentation for the production of ethanol from glucose or xylose. In some cases, the original GIMME results in biological unfeasibility, while the compartmentalised version successfully recovered flux distributions. The method is implemented in the python-based toolbox MEWpy and can be applied to other metabolic studies, opening the opportunity to obtain more refined and realistic flux distributions, which explain the connections between genotypes, environment and phenotypes.

Fermentation employing Saccharomyces cerevisiae has produced alcoholic beverages and bread for millennia. More recently, S. cerevisiae has been used to manufacture specific metabolites for the food, pharmaceutical, and cosmetic industries. Among the most important of these metabolites are compounds associated with desirable aromas and flavors, including higher alcohols and esters. Although the physiology of yeast has been well-studied, its metabolic modulation leading to aroma production in relevant industrial scenarios such as winemaking is still unclear. Here we ask what are the underlying metabolic mechanisms that explain the conserved and varying behavior of different yeasts regarding aroma formation under enological conditions? We employed dynamic flux balance analysis (dFBA) to answer this key question using the latest genome-scale metabolic model (GEM) of S. cerevisiae. The model revealed several conserved mechanisms among wine yeasts, for example, acetate ester formation is dependent on intracellular metabolic acetyl-CoA/CoA levels, and the formation of ethyl esters facilitates the removal of toxic fatty acids from cells using CoA. Species-specific mechanisms were also found, such as a preference for the shikimate pathway leading to more 2-phenylethanol production in the Opale strain as well as strain behavior varying notably during the carbohydrate accumulation phase and carbohydrate accumulation inducing redox restrictions during a later cell growth phase for strain Uvaferm. In conclusion, our new metabolic model of yeast under enological conditions revealed key metabolic mechanisms in wine yeasts, which will aid future research strategies to optimize their behavior in industrial settings.

Saccharomyces non-cerevisiae yeasts are gaining momentum in wine fermentation due to their potential to reduce ethanol content and achieve attractive aroma profiles. However, the design of the fermentation process for new species requires intensive experimentation. The use of mechanistic models could automate process design, yet to date, most fermentation models have focused on primary metabolism. Therefore, these models do not provide insight into the production of secondary metabolites essential for wine quality, such as aromas. In this work, we formulate a continuous model that accounts for the physiological status of yeast, that is, exponential growth, growth under nitrogen starvation and transition to stationary or decay phases. To do so, we assumed that nitrogen starvation is associated with carbohydrate accumulation and the induction of a set of transcriptional changes associated with the stationary phase. The model accurately described the dynamics of time series data for biomass and primary and secondary metabolites obtained for various yeast species in single culture fermentations. We also used the proposed model to explore different process designs, showing how the addition of nitrogen could affect the aromatic profile of wine. This study underlines the potential of incorporating yeast physiology into batch fermentation modelling and provides a new means of automating process design.

Nonconventional yeasts promise to provide new metabolic pathways for producing industrially relevant compounds and tolerating specific stressors such as cold temperatures. The mechanisms behind the cold tolerance of S. kudriavzevii or its sympatric relationship with S. cerevisiae in Mediterranean oaks are not well understood.

In recent years, the winemaking sector is facing new challenges due to the consumers' demands of lower alcohol wines with improved aroma profiles. Yeasts play a critical role in responding to the challenge. However the design of novel wine fermentation processes with alternative yeast species requires a full understanding of the fermentation dynamics, taking into account the different phases (lag, exponential, nitrogen-limited, stationary and decay). In this work, we present a mechanistic model to explain wine fermentation. The proposed model accounts for the dynamics of yeast biomass, the uptake of sugars and assimilable nitrogen, and the production of several relevant metabolites including ethanol, glycerol, acetate, succinate and higher alcohols. The model was calibrated using experimental data obtained during fermentation of natural must by a commercial strain, Saccharomyces cerevisiae T73. The model successfully explained the data, thus showing the potential to be applied to the design of novel wine fermentations with alternative yeast species.

Today, the most relevant food fermentation processes -such as beer, wine, or bread making- are conducted in batch conditions and use yeasts as starters. However, the metabolism of different yeast species in food fermentation is often not fully understood. The integration of multi-omics data into mathematical genome-scale models (GEMs) might help to fill this gap. However, modelling food fermentation at genome-scale is far from trivial and there are not software tools that may help non-experts to deal with such task. This work addresses the implementation of an open source dynamic GEM of yeast fermentation in Python 3. The model accounts for the relevant phases and mechanisms, including primary and secondary metabolism. The software combines the SciPy ordinary differential equation solver to explain the dynamics of the exometabolites and the COBRApy to solve the flux balance analysis problem. The software was successfully tested for the simulation of wine fermentation using three strains of yeast of the Saccharomyces genus.

The yeast Saccharomyces cerevisiae is an essential microorganism in food biotechnology; particularly, in wine and beer making. During wine fermentation, yeasts transform sugars present in the grape juice into ethanol and carbon dioxide. The process occurs in batch conditions and is, for the most part, an anaerobic process. Previous studies linked limited-nitrogen conditions with problematic fermentations, with negative consequences for the performance of the process and the quality of the final product. It is, therefore, of the highest interest to anticipate such problems through mathematical models. Here we propose a model to explain fermentations under nitrogen-limited anaerobic conditions. We separated the biomass formation into two phases: growth and carbohydrate accumulation. Growth was modelled using the well-known Monod equation while carbohydrate accumulation was modelled by an empirical function, analogous to a proportional controller activated by the limitation of available nitrogen. We also proposed to formulate the fermentation rate as a function of the total protein content when relevant data are available. The final model was used to successfully explain experiments taken from the literature, performed under normal and nitrogen-limited conditions. Our results revealed that Monod model is insufficient to explain biomass formation kinetics in nitrogen-limited fermentations of S. cerevisiae. The goodness-of-fit of the herewith proposed model is superior to that of previously published models, offering the means to predict, and thus control fermentations. Importance Problematic fermentations still occur in the winemaking industrial practise. Problems include sluggish rates of fermentation, which have been linked to insufficient levels of assimilable nitrogen. Data and relevant models can help anticipate poor fermentation performance. In this work, we proposed a model to predict biomass growth and fermentation rate under nitrogen-limited conditions and tested its performance with previously published experimental data. Our results show that the well-known Monod equation does not suffice to explain biomass formation.

Yeasts constitute over 1,500 species with great potential for biotechnology. Still, the yeast Saccharomyces cerevisiae dominates industrial applications, and many alternative physiological capabilities of lesser-known yeasts are not being fully exploited. While comparative genomics receives substantial attention, little is known about yeasts' metabolic specificity in batch cultures. Here, we propose a multiphase multiobjective dynamic genome-scale model of yeast batch cultures that describes the uptake of carbon and nitrogen sources and the production of primary and secondary metabolites. The model integrates a specific metabolic reconstruction, based on the consensus Yeast8, and a kinetic model describing the time-varying culture environment. In addition, we proposed a multiphase multiobjective flux balance analysis to compute the dynamics of intracellular fluxes. We then compared the metabolism of S. cerevisiae and Saccharomyces uvarum strains in a rich medium fermentation. The model successfully explained the experimental data and brought novel insights into how cryotolerant strains achieve redox balance. The proposed model (along with the corresponding code) provides a comprehensive picture of the main steps occurring inside the cell during batch cultures and offers a systematic approach to prospect or metabolically engineering novel yeast cell factories. IMPORTANCE Nonconventional yeast species hold the promise to provide novel metabolic routes to produce industrially relevant compounds and tolerate specific stressors, such as cold temperatures. This work validated the first multiphase multiobjective genome-scale dynamic model to describe carbon and nitrogen metabolism throughout batch fermentation. To test and illustrate its performance, we considered the comparative metabolism of three yeast strains of the Saccharomyces genus in rich medium fermentation. The study revealed that cryotolerant Saccharomyces species might use the γ-aminobutyric acid (GABA) shunt and the production of reducing equivalents as alternative routes to achieve redox balance, a novel biological insight worth being explored further. The proposed model (along with the provided code) can be applied to a wide range of batch processes started with different yeast species and media, offering a systematic and rational approach to prospect nonconventional yeast species metabolism and engineering novel cell factories.

Yeasts constitute over 1500 species with great potential for biotechnology. Still, the yeast Saccharomyces cerevisiae dominates industrial applications and many alternative physiological capabilities of lesser-known yeasts are not being fully exploited. While comparative genomics receives substantial attention, little is known about yeasts’ metabolic specificity in batch cultures. Here we propose a multi-phase multi-objective dynamic genome-scale model of yeast batch cultures that describes the uptake of carbon and nitrogen sources and the production of primary and secondary metabolites. The model integrates a specific metabolic reconstruction, based on the consensus Yeast8, and a kinetic model describing the time-varying culture environment. Besides, we proposed a multi-phase multi-objective flux balance analysis to compute the dynamics of intracellular fluxes. We then compared the metabolism of S. cerevisiae and S. uvarum strains in wine fermentation. The model successfully explained the experimental data and brought novel insights into how cryotolerant strains achieve redox balance. The proposed modeling captures the dynamics of metabolism throughout the batch and offers a systematic approach to prospect or engineer novel yeast cell factories.

Yeasts constitute over 1500 species with great potential for biotechnology. Still, the yeast Saccharomyces cerevisiae dominates industrial applications and many alternative physiological capabilities of lesser-known yeasts are not being fully exploited. While comparative genomics receives substantial attention, little is known about yeasts’ metabolic specificity in batch cultures. Here we propose a multi-phase multi-objective dynamic genome-scale model of yeast batch cultures that describes the uptake of carbon and nitrogen sources and the production of primary and secondary metabolites. The model integrates a specific metabolic reconstruction, based on the consensus Yeast8, and a kinetic model describing the time-varying culture environment. Besides, we proposed a multi-phase multi-objective flux balance analysis to compute the dynamics of intracellular fluxes. We then compared the metabolism of S. cerevisiae and S. uvarum strains in wine fermentation. The model successfully explained the experimental data and brought novel insights into how cryotolerant strains achieve redox balance. The proposed modeling captures the dynamics of metabolism throughout the batch and offers a systematic approach to prospect or engineer novel yeast cell factories.

8th IFAC Conference on Foundations of Systems Biology in Engineering, Valencia, 15-18 October 2019

Wineries face unprecedented challenges due to new market demands and climate change effects on wine quality. New yeast starters including non-conventional Saccharomyces species, such as S. kudriavzevii, may contribute to deal with some of these challenges. The design of new fermentations using non-conventional yeasts requires an improved understanding of the physiology and metabolism of these cells. Dynamic modeling brings the potential of exploring the most relevant mechanisms and designing optimal processes more systematically. In this work we explore mechanisms by means of a model selection, reduction and cross-validation pipeline which enables to dissect the most relevant fermentation features for the species under consideration, Saccharomyces cerevisiae T73 and Saccharomyces kudriavzevii CR85. The pipeline involved the comparison of a collection of models which incorporate several alternative mechanisms with emphasis on the inhibitory effects due to temperature and ethanol. We focused on defining a minimal model with the minimum number of parameters, to maximize the identifiability and the quality of cross-validation. The selected model was then used to highlight differences in behavior between species. The analysis of model parameters would indicate that the specific growth rate and the transport of hexoses at initial times are higher for S. cervisiae T73 while S. kudriavzevii CR85 diverts more flux for glycerol production and cellular maintenance. As a result, the fermentations with S. kudriavzevii CR85 are typically slower; produce less ethanol but higher glycerol. Finally, we also explored optimal initial inoculation and process temperature to find the best compromise between final product characteristics and fermentation duration. Results reveal that the production of glycerol is distinctive in S. kudriavzevii CR85, it was not possible to achieve the same production of glycerol with S. cervisiae T73 in any of the conditions tested. This result brings the idea that the optimal design of mixed cultures may have an enormous potential for the improvement of final wine quality.