Optimization of human serum albumin production in methylotrophic yeast Pichia pastoris by repeated fed-batch fermentation

An optimization method for repeated fed-batch fermentation was established with the aim of improving the recombinant human serum albumin (rHSA) production in Pichia pastoris. A simulation model for fed-batch fermentation was formulated and the optimal methanol-feeding policy calculated by dynamic programming method using five different methanol-feeding periods. The necessary state variables were collected from the calculated results and used for further optimization of repeated fed-batch fermentation. The optimal operation policy was investigated using the pre-collected state variables by estimating the overall profit per total methanol-feeding time. The calculated results indicated that the initial cell mass from the 2nd fed-batch fermentation on should be set at 35 or 40 g and methanol-feeding time at 264 h. In repeated fed-batch fermentation using the optimal operation policy, actual culture volume was in good agreement with the values simulated by model equations, but some discrepancy was observed in rHSA production. Minimum experiments were therefore carried out to re-evaluate rHSA production levels, which were then applied in re-calculations to determine the optimal operation policy. The optimal policy for repeated fed-batch fermentation established in the present study (i.e., 4-times-repeated fed-batch fermentation) achieved a 47% increase in annual rHSA production. Optimization of the culture period also brought about a 28% increase in annual rHSA production even in simple (not repeated) fed-batch fermentation.     

Cephalosporin C production by immobilized Cephalosporium acremonium cells in a repeated batch tower bioreactor

The industrial production of antibiotics with filamentous fungi is usually carried out in conventional aerated and agitated tank fermentors. Highly viscous non-Newtonian broths are produced and a compromise must be found between convenient shear stress and adequate oxygen transfer. In this work, cephalosporin C production by bioparticles of immobilized cells of Cephalosporium acremonium ATCC 48272 was studied in a repeated batch tower bioreactor as an alternative to the conventional process. Also, gas-liquid oxygen transfer volumetric coefficients, k(L)a, were determined at various air flow-rates and alumina contents in the bioparticle. The bioparticles were composed of calcium alginate (2.0% w/w), alumina ( < 44 micra), cells, and water. A model describing the cell growth, cephalosporin C production, oxygen, glucose, and sucrose consumption was proposed. To describe the radial variation of oxygen concentration within the pellet, the reaction-diffusion model forecasting a dead core bioparticle was adopted. The k(L)a measurements with gel beads prepared with 0.0, 1.0, 1.5, and 2.0% alumina showed that a higher k(L)a value is attained with 1.5 and 2.0%. An expression relating this coefficient to particle density, liquid density, and air velocity was obtained and further utilized in the simulation of the proposed model. Batch, followed by repeated batch experiments, were accomplished by draining the spent medium, washing with saline solution, and pouring fresh medium into the bioreactor. Results showed that glucose is consumed very quickly, within 24 h, followed by sucrose consumption and cephalosporin C production. Higher productivities were attained during the second batch, as cell concentration was already high, resulting in rapid glucose consumption and an early derepression of cephalosporin C synthesizing enzymes. The model incorporated this improvement predicting higher cephalosporin C productivity.     

Online detection of feed demand in high cell density cultures of Escherichia coli by measurement of changes in dissolved oxygen transients in complex media

A starvation-based dissolved oxygen (DO) transient controller was developed to supply growth-limiting substrate to high cell density fed-batch cultures of recombinant Escherichia coli. The algorithm adjusted a preexisting feed rate in proportion to the culture's oxygen demand, which was estimated from transients in the DO concentration after short periods of feed interruption. In this manner, the addition of glucose feed was precisely controlled at a rate that did not exceed the acetate production threshold, thus preventing acetate accumulation. In comparison to exponential feed algorithms commonly used in industry, the implementation of the new feeding strategy increased the final cell density from 32 to 44 g (dry cell weight).L(-1), with less than 16 mM acetate accumulated, producing an ideal culture for subsequent induction. Despite a constant starvation level and relatively low levels of acetate, experimental cultivations still tended to produce acetate towards the end of the process. The use of a simple Monod model provided an explanation as to why this may occur in high cell density cultivations and suggests how it may be overcome.     

Transient response of aerobic and anoxic activated sludge activities to sudden substrate concentration changes

The state-of-the-art understanding of activated sludge processes as summarized in activated sludge models (ASMs) predicts an instantaneous increase in the biomass activity (which is measured, e.g., by the corresponding respiration rate OUR, NUR, etc.) under sudden substrate concentration changes. Experimental data (e.g., short-term batch respiration experiments under aerobic or anoxic conditions) collected for the calibration of the dynamic models (ASMs) often exhibit a transient phenomenon while attaining maximum activity, which cannot be explained by the current understanding of the activated sludge process. That transient phenomenon exhibits itself immediately upon addition of a substrate source to an endogenously respiring activated sludge sample and it usually takes a few minutes until the activated sludge reaches its maximum possible rate under given environmental conditions. This discrepancy between the state-of-the-art model and the experimental data is addressed in detail in this investigation. It is shown that the discrepancy is not caused by an error in the experimental set-up/data but it is rather due to model inadequacy. Among the hypotheses proposed, it appears that this transient response of the activated sludge most likely results from the sequence of intracellular reactions involved in substrate degradation by the activated sludge. Results from studies performed elsewhere with pure cultures (S. cerevisae and E. coli) support the hypothesis. The transient phenomenon can be described by a dynamic metabolic network model or by a simple first-order model, as adopted in this study. The transient phenomenon occurring in short-term batch respiration experiments is shown to interfere severely with parameter estimation if not modeled properly (2.8%, 11.5%, and 16.8% relative errors [average of three experiments] on Y(H), micro(maxH), and K(S), respectively). Proper modeling of this transient phenomenon whose time constant is on the order of minutes (1 to 3 min) is expected to contribute fundamentally to a better understanding and modeling of Orbal, carousel, and SBR-type treatment plants with fast-alternating process conditions, although such studies are beyond the scope of this report.     

Change of extracellular cAMP concentration is a sensitive reporter for bacterial fitness in high-cell-density cultures of Escherichia coli

Guanosine-3',5'-tetraphosphate (ppGpp) and sigmaS, two regulators of the starvation response of Escherichia coli, have received increasing attention for monitoring cell physiological changes in production processes, although both are difficult to quantify. The kinetics of cAMP formation and degradation were not yet investigated in such processes, although the complex regulation of cAMP by synthesis, release, and degradation in connection with straightforward methods for analysis renders it a highly informative target. Therefore, we followed the cAMP concentration in various nonrecombinant and in four different recombinant glucose-limited fed-batch processes in different production scales. The intracellular cAMP concentration increases strongly at the end of the batch phase. Most cAMP is released to the cultivation medium. The rates of accumulation and degradation of extracellular cAMP are growth-rate-dependent and show a distinct maximum at a growth rate of about 0.35 h(-1). At very low growth rates, below 0.05 h(-1), extracellular cAMP is not produced but rather degraded, independent of whether this low growth rate is caused by glucose limitation or by the high metabolic load of recombinant protein production. In contrast to intracellular cAMP, which is highly unstable, analysis of extracellular cAMP is simpler and the kinetics of accumulation and degradation reflect well the physiological situation, including unlimited growth, limitation, and severe starvation of a production host.     

A strategy for modelling dynamic responses in metabolic samples characterized by GC/MS

A multivariate strategy for studying the metabolic response over time in urinary GC/MS data is presented and exemplified by a study of drug-induced liver toxicity in the rat. The strategy includes the generation of representative data through hierarchical multivariate curve resolution (H-MCR), highlighting the importance of obtaining resolved metabolite profiles for quantification and identification of exogenous (drug related) and endogenous compounds (potential biomarkers) and for allowing reliable comparisons of multiple samples through multivariate projections. Batch modelling was used to monitor and characterize the normal (control) metabolic variation over time as well as to map the dynamic response of the drug treated animals in relation to the control. In this way treatment related metabolic responses over time could be detected and classified as being drug related or being potential biomarkers. In summary the proposed strategy uses the relatively high sensitivity and reproducibility of GC/MS in combination with efficient multivariate curve resolution and data analysis to discover individual markers of drug metabolism and drug toxicity. The presented results imply that the strategy can be of great value in drug toxicity studies for classifying metabolic markers in relation to their dynamic responses as well as for biomarker identification.     

基于遗传算法和支持向量机的乳杆菌发酵过程建模

由于生化反应过程的复杂性和高度非线性,多数简单的数学模型不能准确描述。该文基于Matlab软件,利用改进的支持向量机(υ-SVR)对植物乳酸杆菌发酵这一典型生化过程进行研究,应用遗传算法估计模型最优参数,建立植物乳杆菌的菌体密度预测模型。同时建立传统的logistic动力学模型以进行比较。结果表明,采用结合遗传算法的υ-SVR预测模型拟合误差小,皮尔森相关系数(R)更高,可以较好地预测乳酸杆菌的发酵过程,为其优化控制及放大提供依据。     

Plasmid‐free T7‐based Escherichia coli expression systems

In order to release host cells from plasmid‐mediated increases in metabolic load and high gene dosages, we developed a plasmid‐free, T7‐based E. coli expression system in which the target gene is site‐specifically integrated into the genome of the host. With this system, plasmid‐loss, a source of instability for conventional expression systems, was eliminated. At the same time, system leakiness, a challenging problem with recombinant systems, was minimized. The efficiency of the T7 RNA polymerase compensates for low gene dosage and provides high rates of recombinant gene expression without fatal consequences to host metabolism. Relative to conventional pET systems, this system permits improved process stability and increases the host cell's capacity for recombinant gene expression, resulting in higher product yields. The stability of the plasmid‐free system was proven in chemostat cultivation for 40 generations in a non‐induced and for 10 generations in a fully induced state. For this reason plasmid‐free systems benefit the development of continuous production processes with E. coli. However, time and effort of the more complex cloning procedure have to be considered in relation to the advantages of plasmid‐free systems in upstream‐processing. Biotechnol. Bioeng. 2010. 105: 786–794. © 2009 Wiley Periodicals, Inc.     

Profiling of N‐glycosylation gene expression in CHO cell fed‐batch cultures

One of the goals of recombinant glycoprotein production is to achieve consistent glycosylation. Although many studies have examined the changes in the glycosylation quality of recombinant protein with culture, very little has been done to examine the underlying changes in glycosylation gene expression as a culture progresses. In this study, the expression of 24 genes involved in N‐glycosylation were examined using quantitative RT PCR to gain a better understanding of recombinant glycoprotein glycosylation during production processes. Profiling of the N‐glycosylation genes as well as concurrent analysis of glycoprotein quality was performed across the exponential, stationary and death phases of a fed‐batch culture of a CHO cell line producing recombinant human interferon‐γ (IFN‐γ). Of the 24 N‐glycosylation genes examined, 21 showed significant up‐ or down‐regulation of gene expression as the fed‐batch culture progressed from exponential, stationary and death phase. As the fed‐batch culture progressed, there was also an increase in less sialylated IFN‐γ glycoforms, leading to a 30% decrease in the molar ratio of sialic acid to recombinant IFN‐γ. This correlated with decreased expression of genes involved with CMP sialic acid synthesis coupled with increased expression of sialidases. Compared to batch culture, a low glutamine fed‐batch strategy appears to need a 0.5 mM glutamine threshold to maintain similar N‐glycosylation genes expression levels and to achieve comparable glycoprotein quality. This study demonstrates the use of quantitative real time PCR method to identify possible “bottlenecks” or “compromised” pathways in N‐glycosylation and subsequently allow for the development of strategies to improve glycosylation quality. Biotechnol. Bioeng. 2010;107: 516–528. © 2010 Wiley Periodicals, Inc.     

Model‐based high cell density cultivation of Rhodospirillum rubrum under respiratory dark conditions

The potential of facultative photosynthetic bacteria as producers of photosynthetic pigments, vitamins, coenzymes and other valuable products has been recognized for decades. However, mass cultivation under photosynthetic conditions is generally inefficient due to the inevitable limitation of light supply when cell densities become very high. The previous development of a new cultivation process for maximal expression of photosynthetic genes under semi‐aerobic dark conditions in common bioreactors offers a new perspective for utilizing the facultative photosynthetic bacterium Rhodospirillum rubrum for large‐scale applications. Based on this cultivation system, the present study aimed in determining the maximal achievable cell density of R. rubrum in a bioreactor, thereby providing a major milestone on the way to industrial bioprocesses. As a starting point, we focus on aerobic growth due to higher growth rates and more facile process control under this condition, with the option to extend the process by an anaerobic production phase. Process design and optimization were supported by an unstructured computational process model, based on mixed‐substrate kinetics. Key parameters for growth and process control were determined in shake‐flask experiments or estimated by simulation studies. For fed‐batch cultivation, a computer‐controlled exponential feed algorithm in combination with a pH‐stat element was implemented. As a result, a maximal cell density of 59 g cell dry weight (CDW) L^?1 was obtained, representing so far not attainable cell densities for photosynthetic bacteria. The applied exponential fed‐batch methodology therefore enters a range which is commonly employed for industrial applications with microbial cells. The biochemical analysis of high cell density cultures revealed metabolic imbalances, such as the accumulation and excretion of tetrapyrrole intermediates of the bacteriochlorophyll biosynthetic pathway. Biotechnol. Bioeng. 2010. 105: 729–739. © 2009 Wiley Periodicals, Inc.