Predicting growth rates and growth boundary of Listeria monocytogenes - An international validation study with focus on processed and ready-to-eat meat and seafood.

The performance of six predictive models for Listeria monocytogenes was evaluated using 1014 growth responses of the pathogen in meat, seafood, poultry and dairy products. The performance of the growth models was closely related to their complexity i.e. the number of environmental parameters they take into account. The most complex model included the effect of nine environmental parameters and it performed better than the other less complex models both for prediction of maximum specific growth rates (micro(max) values) and for the growth boundary of L. monocytogenes. For this model bias and accuracy factors for growth rate predictions were 1.0 and 1.5, respectively, and 89% of the growth/no-growth responses were correctly predicted. The performance of three other models, including the effect of five to seven environmental parameters, was considered acceptable with bias factors of 1.2 to 1.3. These models all included the effect of acetic acid/diacetate and lactic acid, one of the models also included the effect of CO(2) and nitrite but none of these models included the effect of smoke components. Less complex models that did not include the effect of acetic acid/diacetate and lactic acid were unable to accurately predict growth responses of L. monocytogenes in the wide range of food evaluated in the present study. When complexity of L. monocytogenes growth models matches the complexity of foods of interest, i.e. the number of hurdles to microbial growth, then predicted growth responses of the pathogen can be accurate. The successfully validated models are useful for assessment and management of L. monocytogenes in processed and ready-to-eat (RTE) foods.

Lactic acid production from lime-treated wheat straw by Bacillus coagulans: neutralization of acid by fed-batch addition of alkaline substrate.

Conventional processes for lignocellulose-to-organic acid conversion requires pretreatment, enzymatic hydrolysis, and microbial fermentation. In this study, lime-treated wheat straw was hydrolyzed and fermented simultaneously to lactic acid by an enzyme preparation and Bacillus coagulans DSM 2314. Decrease in pH because of lactic acid formation was partially adjusted by automatic addition of the alkaline substrate. After 55 h of incubation, the polymeric glucan, xylan, and arabinan present in the lime-treated straw were hydrolyzed for 55%, 75%, and 80%, respectively. Lactic acid (40.7 g/l) indicated a fermentation efficiency of 81% and a chiral L(+)-lactic acid purity of 97.2%. In total, 711 g lactic acid was produced out of 2,706 g lime-treated straw, representing 43% of the overall theoretical maximum yield. Approximately half of the lactic acid produced was neutralized by fed-batch feeding of lime-treated straw, whereas the remaining half was neutralized during the batch phase with a Ca(OH)2 suspension. Of the lime added during the pretreatment of straw, 61% was used for the neutralization of lactic acid. This is the first demonstration of a process having a combined alkaline pretreatment of lignocellulosic biomass and pH control in fermentation resulting in a significant saving of lime consumption and avoiding the necessity to recycle lime.

In vivo kinetics with rapid perturbation experiments in Saccharomyces cerevisiae using a second-generation BioScope.

We present a robust second-generation BioScope: a system for continuous perturbation experiments. Firstly, the BioScope design parameters (i.e., pressure drop, overall oxygen (O2) and carbon dioxide (CO2) mass transfer, mean residence time distribution and plug flow characteristics) were evaluated. The average overall mass transfer coefficients were estimated to be 1.8E-5 m s(-1) for O2 and 0.34E-5 m s(-1) for CO2. It was determined that the O2/CO2 permeable membrane accounted for 75% and 95% of the overall resistance for O2 and CO2, respectively. The Peclet number (Pe) of the system was found to be >500 for liquid flow rates between 1 and 4 ml min(-1), ensuring plug flow characteristics. Secondly, steady-state intracellular metabolite concentrations obtained using direct rapid sampling from the fermentor were compared with those obtained by rapid sampling via the pre-perturbation sample port of the BioScope. With both methods the same metabolite levels were obtained. Thirdly, glucose perturbation experiments were carried out directly in the fermentor as well as in the BioScope, whereby steady-state Saccharomyces cerevisiae cells from a glucose/ethanol limited chemostat were perturbed by increasing the extracellular glucose concentration from 0.11 to 2.8 mM. Intracellular and extracellular metabolite levels were measured within a time window of 180 s. It was observed that the dynamic metabolite concentration profiles obtained from both perturbations were nearly the same, with the exception of the C4 metabolites of the TCA cycle, which might be due to differences in culture age.

Optimal re-design of primary metabolism in Escherichia coli using linlog kinetics.

This paper examines the validity of the linlog approach, which was recently developed in our laboratory, by comparison of two different kinetic models for the metabolic network of Escherichia coli. The first model is a complete mechanistic model; the second is an approximative model in which linlog kinetics are applied. The parameters of the linlog model (elasticities) are derived from the mechanistic model. Three different optimization cases are examined. In all cases, the objective is to calculate the enzyme levels that maximize a certain flux while keeping the total amount of enzyme constant and preventing large changes of metabolite concentrations. For an average variation of metabolite levels of 10% and individual changes of a factor 2, the predicted enzyme levels, metabolite concentrations and fluxes of both models are highly similar. This similarity holds for changes in enzyme level of a factor 4-6 and for changes in fluxes up to a factor 6. In all three cases, the predicted optimal enzyme levels could neither have been found by intuition-based approaches, nor on basis of flux control coefficients. This demonstrates that kinetic models are essential tools in Metabolic Engineering. In this respect, the linlog approach is a valuable extension of MCA, since it allows construction of kinetic models, based on MCA parameters, that can be used for constrained optimization problems and are valid for large changes of metabolite and enzyme levels.

Analysis of in vivo kinetics of glycolysis in aerobic Saccharomyces cerevisiae by application of glucose and ethanol pulses.

This article presents the dynamic responses of several intra- and extracellular components of an aerobic, glucose-limited chemostat culture of Saccharomyces cerevisiae to glucose and ethanol pulses within a time window of 75 sec. Even though the ethanol pulse cannot perturb the glycolytic pathway directly, a distinct response of the metabolites at the lower part of glycolysis was found. We suggest that this response is an indirect effect, caused by perturbation of the NAD/NADH ratio, which is a direct consequence of the conversion of ethanol into acetaldehyde. This effect of the NAD/NADH ratio on glycolysis might serve as an additional explanation for the observed decrease of 3PG, 2PG, and PEP during a glucose pulse. The responses measured during the ethanol pulse were used to evaluate the allosteric regulation of glycolysis. Our results confirm that FBP stimulates pyruvate kinase and suggest that this effect is pronounced. Furthermore, it appears that PEP does not play an important role in the allosteric regulation of phosphofructo kinase.

Dynamic simulation and metabolic re-design of a branched pathway using linlog kinetics.

This paper presents a new mathematical framework for modeling of in vivo dynamics and for metabolic re-design: the linlog approach. This approach is an extension of metabolic control analysis (MCA), valid for large changes of enzyme and metabolite levels. Furthermore, the presented framework combines MCA with kinetic modeling, thereby also combining the merits of both approaches. The linlog framework includes general expressions giving the steady-state fluxes and metabolite concentrations as a function of enzyme levels and extracellular concentrations, and a metabolic design equation that allows direct calculation of required enzyme levels for a desired steady state when control and response coefficients are available. Expressions giving control coefficients as a function of the enzyme levels are also derived. The validity of the linlog approximation in metabolic modeling is demonstrated by application of linlog kinetics to a branched pathway with moiety conservation, reversible reactions and allosteric interactions. Results show that the linlog approximation is able to describe the non-linear dynamics of this pathway very well for concentration changes up to a factor 20. Also the metabolic design equation was tested successfully.

Rapid sampling for analysis of in vivo kinetics using the BioScope: a system for continuous-pulse experiments.

In this article we present a novel device, the BioScope, which allows elucidation of in vivo kinetics of microbial metabolism via perturbation experiments. The perturbations are carried out according to the continuous-flow method. The BioScope consists of oxygen permeable silicon tubing, connected to the fermentor, through which the broth flows at constant velocity. The tubing has a special geometry (serpentine channel) to ensure plug flow. After leaving the fermentor, the broth is mixed with a small flow of perturbing agent. This represents the start of the perturbation. The broth is sampled at different locations along the tubing, corresponding to different incubation times. The maximal incubation time is 69 s; the minimally possible time interval between the samples is 3-4 s. Compared to conventional approaches, in which the perturbation is carried out in the fermentor, the BioScope offers a number of advantages. (1) A large number of different perturbation experiments can be carried out on the same day, because the physiological state of the fermentor is not perturbed. (2) In vivo kinetics during fed-batch experiments and in large-scale reactors can be investigated. (3) All metabolites of interest can be measured using samples obtained in a single experiment, because the volume of the samples is unlimited. (4) The amount of perturbing agent spent is minimal, because only a small volume of broth is perturbed. (5) The system is completely automated. Several system properties, including plug-flow characteristics, mixing, oxygen and carbon dioxide transfer rates, the quenching time, and the reproducibility have been explored, with satisfactory results. Responses of several glycolytic intermediates in Saccharomyces cerevisiae to a glucose pulse, measured using a conventional approach are compared to results obtained with the BioScope. The agreement between the results demonstrates that the BioScope is indeed a promising device for studying in vivo kinetics.

The mathematics of metabolic control analysis revisited.

In this minireview, several different approaches to derivation of the theorems and relationships of Metabolic Control Analysis (MCA) are discussed and an alternative approach is presented. This new approach consists of solving the steady-state mass balances for the intracellular metabolites using linearized kinetics. The application of linearized kinetics reflects the fact that MCA is based on linearization of the system equations in a reference steady state. Our derivation is valid for metabolic networks of arbitrary complexity, including those containing conserved moieties and branches. The value of our approach is its simplicity: the derivation is straightforward and therefore easy to follow. It can serve as a compact introduction to the mathematical basis of MCA.