Kinetics of multiproduct fermentations

A simple model for biomass, product, and substrate evolution proposed previously for batch polysaccharide fermentations is extended to multiproduct fermentations. The examples involve Clostridium thermocellum, (ATCC 27405) fermentations of glucose to four products (ethanol, acetic, formic, and lactic acid), of fructose to two products (ethanol and acetic acid), and of cellobiose to two products (ethanol and acetic acid). In all cases, parameter evaluation was carried out in a serial deterministic procedure.     

PHOSPHATE AND SILICATE GROWTH AND UPTAKE KINETICS OF THE DIATOMS ASTERIONELLA FORMOSA AND CYCLOTELLA MENEGHINIANA IN BATCH AND SEMICONTINUOUS CULTURE1

Information on the nutrient kinetics of Asterionella formosa Hass. and Cyclotella meneghiniana Kutz. under either phosphate or silicate limitation was obtained for use in a Monod model and in a variable internal stores model of growth. Short-term batch culture growth experiments were fit to the Monod model and long-term semicontinuous culture experiments and short-term uptake experiments were fit to the variable internal stores model. Mathematical analysis indicates that the parameters of the 2 models may be expressed in terms of each other at steady state. The qualitative results of both batch and steady state culture methods agree. For limiting phosphate experiments. A. formosa is better able to grow at low PO₄-P concentrations than C. meneghiniana, as shown by its lower K for PO₄-P limited growth. The k_Q of A. formosa compared to C. meneghiniana found in long-term semicontinuous culture indicates that A. formosa is almost an order of magnitude more efficient at using internal phosphate for growth. The qualitative results under silicate-limited growth of C. meneghiniana is less than that of A. formosa. The k_Q from semicontinuous culture experiments indicates that C. meneghiniana is the more efficient at using internal silicate for growth. Nutrient uptake experiments showed more variability from a Michaelis-Menten relationship than short-term growth experiments. There were no significant differences between the 2 species in half saturation constants for either phosphate or silicate uptake. We observed a marked dependence of the coefficient of luxury consumption (R) of phosphate on the steady state growth rate. A. formosa has a higher R than C. meneghiniana.     

Comparison between in vivo and in vitro enzyme activities in continuous and batch fermentations of Clostridium acetobutylicum

Summary In vitro activities of key enzymes and related parameters (ATP and ADP concentrations, intracellular pH (pH _i ), cell volume and the transmembrane pH) in various continuous and batch fermentations of Clostridium acetobutylicum were studied in order to investigate the regulation (genetic vs. enzyme level) of the solventogenesis process. In vitro activities varied significantly among an acidogenic (glucose limited) and three solventogenic (an iron limited, a CO gassed and a biomass recycle) continuous fermentations. However, in vitro enzyme activities did not correlate with in vivo specific production rates in continuous cultures indicating that solvent formation is regulated primarily at the enzyme level. Carbon monoxide (CO) gassing of an acidogenic continuous culture resulted in butyrate uptake without acetone formation due to inactivation of the acetoacetate decarboxylase by CO. In continuous, and to some extent in batch cultures, butyrate can be taken up via the reversal of the butyrate kinase and phosphotransbutyrylase pathway. Solvent formation in batch fermentations is both a result of enzyme induction and regulation. Acetone formation and the induction of acetoacetate decarboxylase occur simultaneously whereas both alcohol dehydrogenases are induced several hours before initiation of alcohol production. Finally, the levels of intracellular and related cell parameters (pH _i , pH, ATP and ADP concentrations) are discussed and related to the possible mechanisms of solventogenesis.     

Kinetic metabolic modelling for the control of plant cells cytoplasmic phosphate

A previously developed kinetic metabolic model for plant metabolism was used in a context of identification and control of intracellular phosphate (Pi) dynamics. Experimental data from batch flask cultures of Eschscholtiza californica cells was used to calibrate the model parameters for the slow dynamics (growth, nutrition, anabolic pathways, etc.). Perturbation experiments were performed using a perfusion small-scale bioreactor monitored by in vivo³¹P NMR. Parameter identification for Pi metabolism was done by measuring the cells dynamic response to different inputs for extracellular Pi (two pulse-response experiments and a step-response experiment). The calibrated model can describe Pi translocation between the cellular pools (vacuole and cytoplasm). The effect of intracellular Pi management on ATP/ADP and phosphomonoesters concentrations is also described by the model. The calibrated model is then used to develop a control strategy on the cytoplasmic Pi pool. From the identification of the systems dynamics, a proportional-integral controller was designed and tuned. The closed-loop control was implemented in the small-scale NMR bioreactor and experimental results were in accordance with model predictions. Thus, the calibrated model is able to predict cellular behaviour for phosphate metabolism and it was demonstrated that it is possible to control the intracellular level of cytoplasmic Pi in plant cells.     

Metabolic flux analysis for optimizing the specific growth rate of recombinant <Emphasis Type="Italic">Aspergillus niger</Emphasis>

A comprehensive metabolic network comprising three intracellular compartments (cytoplasm, mitochondrion and peroxisome) was developed for Aspergillus niger. The metabolic flux network includes carbohydrate and amino acid metabolism in both anabolic and catabolic reactions. Linear programming was used for the optimization of the specific growth rates in combination with 37 measured input and output fluxes of the key metabolites to evaluate corresponding intracellular flux distributions throughout the batch fermentations. Logarithmic sensitivity analysis revealed that the addition of proline, alanine and glutamate benefited growth in defined media. The experimental observations and flux analysis showed that tyrosine was a potential candidate for biomass production improvement. Model predictions was verified by conducting batch and fed-batch fermentations and it was found that the addition of the four amino acids according to the predetermined schedule resulted in a 44 and 41% improvements in biomass and recombinant protein productions, respectively.     

Fed-batch propionic acid production by Propionibacterium acidipropionici

The batch fermentations were conducted using lactose as the substrate at pH 6.5 and temperature 30°C. Average batch kinetic data was eventually used to develop an unstructured mathematical model. The kinetic parameters of the model were determined by non-linear regression technique using the batch experimental results. Parametric sensitivity analysis showed the maximum specific substrate consumption rate (r_S^max) and the maintenance energy constant (m_S) to be the most sensitive parameters. The experimental observations in batch fermentation were close to the model predictions. The batch model was extrapolated to identify nutrient feeding strategies, which were tested successfully for two different fed-batch fermentations. It demonstrated enhanced propionic acid productivity. The developed model was found suitable for the design of feeding strategies to increase propionic acid production in fed-batch mode of reactor operation.     

Modeling of growth and lactate fermentation by Lactococcus lactis subsp. lactis biovar. diacetylactis in batch culture

The kinetic behaviour of Lactococcus lactis subsp. lactis biovar. diacetylactis was studied in batch culture under non-limiting conditions that allow high growth and product formation. A model based on laboratory results is proposed for growth and l-lactate fermentation. It shows the necessity for differentiating biomass into three physiological states, two active, X_g (growth + acidification) and X_ng (acidification), and one inactive, X_i. The kinetic theory of the model demonstrates the non-competitive nature of fermentation end-product inhibition on growth and acidification, and describes the passage from one physiological state to another. Satisfying simulations were obtained for batch fermentations, and the use of this type of model for determining and optimizing fermentation parameters is discussed. Correspondence to: C. Diviès     

Growth of Escherichia coli in phosphate limited cultures : Measurements and modelling

The batch cultivation ofEscherichia coli limited by inorganic phosphate shows an exponential growth phase followed by a linear phase when the phosphate in the medium is exhausted. Data analysis and modelling of batch cultivation with different initial concentrations of the limiting factor give information about an exchangeable phosphate store, with RNA as the main component, and a non-exchangeable store.     

A model of xylitol production by the yeast Candida mogii

Production of xylitol from xylose in batch fermentations of Candida mogii ATCC 18364 is discussed in the presence of glucose as the cosubstrate. Various initial ratios of glucose and xylose concentrations are assessed for their impact on yield and rate of production of xylitol. Supplementation with glucose at the beginning of the fermentation increased the specific growth rate, biomass yield and volumetric productivity of xylitol compared with fermentation that used xylose as the sole carbon source. A mathematical model is developed for eventual use in predicting the product formation rate and yield. The model parameters were estimated from experimental observations, using a genetic algorithm. Batch fermentations, which were carried out with xylose alone and a mixture of xylose and glucose, were used to validate the model. The model fitted well with the experimental data of cell growth, substrate consumption and xylitol production.     

Growth and bacteriocin production by Enterococcus faecium DPC1146 in batch and continuous culture

Production of the bacteriocin enterocin 1146 (E1146) by Enterococcus faecium DPC1146 was studied in batch and continuous fermentation. Growth was strongly inhibited by lactic acid. In batch fermentations maximum E1146 activity (2.8 MBU L⁻¹) was obtained in 9 h with 20 g L⁻¹ glucose. Increase in initial glucose concentration did not lead to a proportional increase in E1146 activity. A simple linear model was found to be adequate to explain the relationship between specific bacteriocin production rate and specific growth rate in batch fermentations with initial glucose concentration higher than 20 g L⁻¹. Maximum bacteriocin activity (2.9–3.2 MBU L⁻¹) was obtained in continuous fermentations at dilution rates between 0.12 and 0.17 h⁻¹ and specific bacteriocin production rate increased linearly with dilution rate. Received 31 July 1996/ Accepted in revised form 01 November 1996     

Inhibition of yeast by lactic acid bacteria in continuous culture: nutrient depletion and/or acid toxicity?

Lactic acid was added to batch very high gravity (VHG) fermentations and to continuous VHG fermentations equilibrated to steady state with Saccharomyces cerevisiae. A 53% reduction in colony-forming units (CFU) ml^–1 of S. cerevisiae was observed in continuous fermentation at an undissociated lactic acid concentration of 3.44% w/v; and greater than 99.9% reduction was evident at 5.35% w/v lactic acid. The differences in yeast cell number in these fermentations were not due to pH, since batch fermentations over a pH range of 2.5–5.0 did not lead to changes in growth rate. Similar fermentations performed in batch showed that growth inhibition with added lactic acid was nearly identical. This indicates that the apparent high resistance of S. cerevisiae to lactic acid in continuous VHG fermentations is not a function of culture mode. Although the total amount of ethanol decreased from 48.7 g l^–1 to 14.5 g l^–1 when 4.74% w/v undissociated lactic acid was added, the specific ethanol productivity increased ca. 3.2-fold (from 7.42×10^–7 g to 24.0×10^–7 g ethanol CFU^–1 h^–1), which indicated that lactic acid stress improved the ethanol production of each surviving cell. In multistage continuous fermentations, lactic acid was not responsible for the 83% (CFU ml^–1) reduction in viable S. cerevisiae yeasts when Lactobacillus paracasei was introduced to the system at a controlled pH of 6.0. The competition for trace nutrients in those fermentations and not lactic acid produced by L. paracasei likely caused the yeast inhibition.     

A mathematical model of growth and alkaloid production in the submerged culture of Claviceps purpurea

A mathematical model of the batch cultivation of Claviceps purpurea was formulated. The main attention was devoted to the effect of exocellular and intracellular phosphate on the growth of the mycelium and production of clavine alkaloid under experimental conditions without limitation by carbon and nitrogen sources. The method of nonlinear regression was used ot predict the optional technological regime of the phosphate addition in the batch culture at different time intervals of additions.     

Improved production and stability ofE. coli recombinants expressing transketolase for large scale biotransformation

Summary TheE.coli tkt gene has been subcloned into high copy number vectors. In fed batch fermentations up to 4gL^–1 of soluble intracellular transketolase was produced representing 43% of the total cell protein. Increased plasmid stability during fed-batch fermentations was obtained by using kanamycin resistant pBGS vectors rather than the ampicillin resistant pUC vectors. Plasmid stability was maintained throughout growth in a complex medium without any selective pressure by incorporating thecer region fromColE1 into the expression construct.     

Development and application of a flexible controller in yeast fermentations using pO2 cascade control

The development and application of a flexible process controller in fed‐batch yeast fermentations using pO₂ cascade control was performed. A new algorithm for fed‐batch fermentations using pO₂ cascade control was developed, the concept of which could be used as a realizable solution in fermentation systems equipped according to the basic configuration. The algorithm is based on the combined influence of pO₂ and pH on the substrate feeding intensity. To test and develop this algorithm, Saccharomyces cerevisiae DY 7221 and Candida tropicalis CK‐4 fermentations were carried out. As a result of the use of the combined algorithm, the specific growth rate and productivity grew in both fermentations. In this case, the effect of the use of the algorithm was most pronounced in the C. tropicalis fermentation.     

Modeling of mini-proinsulin production in Pichia pastoris using the AOX promoter

An unstructured model based on mass balance equations for a recombinant methylotrophic yeast Pichia pastoris Mut^S (Methanol Utilization Slow) strain expressing the mini-proinsulin (MPI), was successfully established in quasi-steady state fed-batch fermentations with varying total quantity of biomass in a 7 l fermenter. The model describes the relationships between the total biomass and induction time, both in the batch and fed-batch phases. In addition, good correlations were obtained when the total quantity of MPI was correlated with the total biomass, demonstrating that the product of interest is associated with growth in the methanol phase. The parameters of the fermentation model obtained are similar to those reported by other researchers.     

Kinetics of Bifidobacterium longum ATCC 15707 fermentations: effect of the dilution rate and carbon source

The effect of the dilution rate on biomass and product synthesis in fermentations of glucose, fructose and a commercial mixture of fructooligosaccharides (FOS) by Bifidobacterium longum ATCC 15707 was studied. Kinetic parameters (maximum specific growth rate, Monod constant, maintenance, and yield coefficients) in the mathematical model of the fermentation were estimated from experimental data. In the FOS mixture fermentations, approximately 12% of the total reducing sugars (mainly fructose) in the feed were not metabolized by the bacterium. In fermentations of fructose and the FOS mixture, biomass concentration increased as the dilution rate increased and, once maximum values were reached [3.90 (D=0.20 h^–1) and 2.54 g l^–1 (D=0.15 h^–1), respectively], decreased rapidly as the culture was washed out. Formic acid was detected at low dilution rates in glucose and fructose fermentations. The main products in fermentations of the three carbon sources were lactic and acetic acids. Average values of the molar ratio between acetic and lactic acids of 1.18, 1.21 and 0.83 mol mol^–1 were obtained in glucose, fructose and FOS mixture fermentations, respectively. In batch fermentations carried out without pH control this molar ratio was lower than 1.5 only when fructose was used as the carbon source.     

Hydrocarbon fermentations using Candida lipolytica. I. Basic growth parameters for batch and continuous culture conditions

Candida lipolytica (strain ATCC 8661) was grown on a simple defined medium with n-dodecane as sole carbon source under batch and continuous fermentation conditions. The composition of cellular material recovered from the fermentations, the oxygen demand of the cells, and the effect of operating conditions on cell growth were evaluated experimentally. These basic data are presented and discussed.     

Modelling of E.coli fermentations: comparison of multicompartment and variable structure models

Two novel approaches for modelling processes that can be described by a sequence of phases (metabolic states) are suggested and applied to Escherichia Coli fermentations. The first approach uses a multi-compartment model framework, coupled with knowledge-based logic. In the second approach the multi-compartment model is reduced into the Variable Structure Model consisting of a battery of alternative submodels, each of which qualitatively represents one of the process steps. Furthermore, simulated intracellular process variables are compared with the output of a multi-wavelength fluorosensor and excitation-emission pairs that predict best these variables are identified.     

The Amino Acid Content in Gamma-irradiated Rice†

High phosphate accumulating bacteria were isolated by autoradiography. One isoate, Arthrobacter globiformis PAB-6 accumulated phosphate intracellularly at 20% of dry cell mass in a simple synthetic medium. This amount was 3~7 times higher than type cultures examined. Almost no phosphate was released into the medium after cessation of growth. Fifty percent of total intracellular phosphate was fractionated as nucleic acids, while 20% each was recovered from cold PCA soluble fractions and polyphosphate fractions. The large content of nucleic acids in this bacterium appeared due to increased RNA content, specifically 4 S RNA fraction.