High-cell-density cultivation of Vibrio natriegens in a low-chloride chemically defined medium.

Vibrio natriegens is a halophilic bacterium with the fastest generation time of non-pathogenic bacteria reported so far. It therefore has high potential as a production strain for biotechnological production processes or other applications in biotechnology. Culture media for V. natriegens typically contain high sodium chloride concentrations. The corresponding high chloride concentrations can lead to corrosion processes on metal surfaces in bioreactors. Here we report the development of a low-chloride chemically defined medium for V. natriegens. Sodium chloride was completely replaced by the sodium salts disodium hydrogen phosphate, disodium sulfate, and sodium citrate, while keeping the total concentration of sodium ions constant. The use of citrate prevents the occurrence of precipitates, especially of ammonium magnesium phosphate. With this defined medium, high-cell-density fed-batch cultivations in laboratory-scale bioreactors using exponential feeding yielded biomass concentrations of more than 60 g L-1. KEY POINTS: A defined medium for V. natriegens that only contains traces of chloride was developed Corrosion processes on metal surfaces in industrial bioreactors can thus be prevented High yields of biomass can be achieved in fed-batch cultivation with this medium.

Optimized production strategy of the major capsid protein HPV 16L1 non-assembly variant in E. coli.

The capsid of human papillomavirus (HPV) consists of two capsid proteins - the major capsid protein L1 and the minor capsid protein L2. Assembled virus-like particles, which only consist of L1 proteins, are successfully applied as prophylactic vaccines against HPV infections. The capsid subunits are L1-pentamers, which are also reported to protect efficiently against HPV infections in animals. The recombinant production of L1 has been previously shown in E. coli, yeast, insect cells, plants and mammalian cell culture. Principally, in E. coli-based expression system L1 shows high expression yields but the protein is largely insoluble. In order to overcome this problem reported strategies address fusion proteins and overexpression of bacterial chaperones. However, an insufficient cleavage of the fusion proteins and removal of co-purified chaperones can hamper subsequent down streaming. We report a significant improvement in the production of soluble L1-pentamers by combining (I) a fusion of a N-terminal SUMO-tag to L1, (II) the heterologous co-expression of the chaperon system GroEL/ES and (III) low expression temperature. The fusion construct was purified in a 2-step protein purification including efficient removal of GroEL/ES and complete removal of the N-terminal SUMO-tag. The expression strategy was transferred to process-controlled high-cell-density fermentation with defined media according to the guidelines of good manufacturing practice. The produced L1 protein is highly pure (>95%), free of DNA (260:280 = 0.5) and pentameric. The production strategy yielded 5.73 mg of purified L1-pentamers per gram dry biomass. The optimized strategy is a suitable alternative for high yield L1-pentamer production and purification as a cheaper process for vaccine production.

Development of cultivation strategies for friulimicin production in Actinoplanes friuliensis.

Actinoplanes friuliensis is a rare actinomycete which produces the highly potent lipopeptide antibiotic friulimicin. This lipopeptide antibiotic is active against a broad range of multi-resistant gram-positive bacteria such as methicillin-resistant Enterococcus sp. and Staphylococcus aureus (MRE, MRSA) strains. Antibiotic biosynthesis and regulation in actinomycetes is very complex. In order to study the biosynthesis of these species and to develop efficient production processes, standardized cultivation conditions are a prerequisite. For this reason a chemically defined production medium for A. friuliensis was developed. With this chemically defined medium it was possible to analyze the influence of medium components on growth and antibiotic biosynthesis. These findings were used to develop process strategies for friulimicin production. The focus of the project presented here was to develop cultivation strategies which included fed-batch and continuous cultivation processes. In fed-batch processes, volumetric productivities for friulimicin of 1-2 mg/l h were achieved. In a perfusion process, a very simple cell retention system, which works via sedimentation of the mycelial cell pellets, was used. With this system, stable continuous cultivations with cell retention were dependent on the dilution rate. With a dilution rate of 0.05 h(-1), cell retention worked well and volumetric productivity of friulimicin was enhanced to 3-5 mg/l h. With a higher dilution rate of 0.1 h(-1), friulimicin production ceased because cell retention was not possible any longer with this simple cell retention system. In order to support process development, cultivation data were used to characterize metabolic fluxes in the developed friulimicin production processes.

Complete genome sequence of the actinobacterium Actinoplanes friuliensis HAG 010964, producer of the lipopeptide antibiotic friulimycin.

Actinoplanes friuliensis HAG 010964 (DSM 7358) was isolated from a soil sample from the Friuli region in Italy and characterized as a producer of the antibiotic friulimycin. The complete genome sequence includes genomic information of secondary metabolite biosynthesis and of its lifestyle. Genbank/EMBL/DDBJ Accession Nr: CP006272 (chromosome).

Metabolic flux estimation in mammalian cell cultures.

Metabolic flux analysis with its ability to quantify cellular metabolism is an attractive tool for accelerating cell line selection, medium optimization, and other bioprocess development activities. In the stoichiometric flux estimation approach, unknown fluxes are determined using intracellular metabolite mass balance expressions and measured extracellular rates. The simplicity of the stoichiometric approach extends its application to most cell culture systems, and the steps involved in metabolic flux estimation by the stoichiometric method are presented in detail in this chapter. Specifically, overdetermined systems are analyzed since the extra measurements can be used to check for gross measurement errors and system consistency. Cell-specific rates comprise the input data for flux estimation, and the logistic modeling approach is described for robust-specific rate estimation in batch and fed-batch systems. A simplified network of mammalian cell metabolism is used to illustrate the flux estimation procedure, and the steps leading up the consistency index determination are presented. If gross measurement errors are detected, a technique for determining the source of gross measurement error is also described. A computer program that performs most of the calculation described in this chapter is presented, and references to flux estimation software are provided. The procedure presented in this chapter should enable rapid metabolic flux estimation in any mammalian cell bioreaction network by the stoichiometric approach.

Calorimetric control of the specific growth rate during fed-batch cultures of Saccharomyces cerevisiae.

The specific growth rate of a Saccharomyces cerevisiae strain with glucose as limiting C-source was estimated from the measured heat flow produced by the cells. For the cultivation a standard 30 l laboratory bioreactor was used, which was extended in such a way that heat balancing is possible. The feed rate was adjusted by a feedforward/feedback controller such that the specific growth rate was kept on the desired set-point value. On the basis of experimental investigations it was demonstrated that the specific growth rate can be controlled at a given set point value below the critical value to prevent the production of growth-inhibitory ethanol due to the Crabtree effect. With this control strategy high biomass concentrations of more than 110 g l(-1) can be obtained.

Calorimetric control for high cell density cultivation of a recombinant Escherichia coli strain.

In order to achieve maximum productivity of recombinant proteins in Escherichia coli high cell density cultivation (HCDC) strategies have been the subject of many studies. The aim of this work was the application of calorimetric methods to HCDC. The specific growth rate of a recombinant E. coli strain producing green fluorescent protein (GFP) was controlled during fed-batch cultivations by estimating the specific growth rate from the measured heat flow produced by the cells. For the cultivation a standard 30 l laboratory bioreactor was used, which was extended in such a way that heat balancing is possible. The feed rate was adjusted by an adaptive controller such that the specific growth rate was kept on the desired set point value. On the basis of experimental investigations with a recombinant E. coli strain using glucose as limiting C-source it was demonstrated that the specific growth rate can be kept on a given set point value and biomass concentrations of up to 120 g l(-1) can be obtained, reproducibly.

Metabolic flux analysis of CHO cells in perfusion culture by metabolite balancing and 2D [13C, 1H] COSY NMR spectroscopy.

The physiological state of CHO cells in perfusion culture was quantified by determining fluxes through the bioreaction network using (13)C glucose and 2D-NMR spectroscopy. CHO cells were cultivated in a 2.5L perfusion bioreactor with glucose and glutamine as the primary carbon and energy sources. The reactor was inoculated at a cell density of 8 x 10(6)cells/mL and operated at approximately 10 x 10(6)cells/mL using unlabeled glucose for the first 13 days. The second phase lasted 12 days and the medium consisted of 10% [U-(13)C]glucose, 40% labeled [1-(13)C]glucose with the balance unlabeled. After the culture attained isotopic steady state, biomass samples from the last 3 days of cultivation were considered representative and used for flux estimation. They were hydrolyzed and analyzed by 2D [(13)C, (1)H] COSY measurements using the heteronuclear single quantum correlation sequence with gradients for artifacts suppression. Metabolic fluxes were determined using the 13C-Flux software package by minimizing the residuals between the experimental and the simulated NMR data. Normalized residuals exhibited a Gaussian distribution indicating good model fit to experimental data. The glucose consumption rate was 5-fold higher than that of glutamine with 41% of glucose channeled through the pentose phosphate pathway. The fluxes at the pyruvate branch point were almost equally distributed between lactate and the TCA cycle (55% and 45%, respectively). The anaplerotic conversion of pyruvate to oxaloacetate by pyruvate carboxylase accounted for 10% of the pyruvate flux with the remaining 90% entering the TCA cycle through acetyl-CoA. The conversion of malate to pyruvate catalyzed by the malic enzyme was 70% higher than that for the anaplerotic reaction catalyzed by pyruvate carboxylase. Most amino acid catabolic and biosynthetic fluxes were significantly lower than the glycolytic and TCA cycle fluxes. Metabolic flux data from NMR analysis validated a simplified model where metabolite balancing was used for flux estimation. In this reduced flux space, estimates from these two methods were in good agreement. This simplified model can routinely be used in bioprocess development experiments to estimate metabolic fluxes with much reduced analytical investment. The high resolution flux information from 2D-NMR spectroscopy coupled with the capability to validate a simplified metabolite balancing based model for routine use make (13)C-isotopomer analysis an attractive bioprocess development tool for mammalian cell cultures.

Error propagation from prime variables into specific rates and metabolic fluxes for mammalian cells in perfusion culture.

Error propagation from prime variables into specific rates and metabolic fluxes was quantified for high-concentration CHO cell perfusion cultivation. Prime variable errors were first determined from repeated measurements and ranged from 4.8 to 12.2%. Errors in nutrient uptake and metabolite/product formation rates for 5-15% error in prime variables ranged from 8-22%. The specific growth rate, however, was characterized by higher uncertainty as 15% errors in the bioreactor and harvest cell concentration resulted in 37.8% error. Metabolic fluxes were estimated for 12 experimental conditions, each of 10 day duration, during 120-day perfusion cultivation and were used to determine error propagation from specific rates into metabolic fluxes. Errors of the greater metabolic fluxes (those related to glycolysis, lactate production, TCA cycle and oxidative phosphorylation) were similar in magnitude to those of the related greater specific rates (glucose, lactate, oxygen and CO(2) rates) and were insensitive to errors of the lesser specific rates (amino acid catabolism and biosynthesis rates). Errors of the lesser metabolic fluxes (those related to amino acid metabolism), however, were extremely sensitive to errors of the greater specific rates to the extent that they were no longer representative of cellular metabolism and were much less affected by errors in the lesser specific rates. We show that the relationship between specific rate and metabolic flux error could be accurately described by normalized sensitivity coefficients, which were readily calculated once metabolic fluxes were estimated. Their ease of calculation, along with their ability to accurately describe the specific rate-metabolic flux error relationship, makes them a necessary component of metabolic flux analysis.

Comparative analysis of transcriptional activities of heterologous promoters in the rare actinomycete Actinoplanes friuliensis.

Manipulation of secondary metabolite production in the rare actinomycete Actinoplanes friuliensis, the producer of the lipopeptide antibiotic friulimicin, is hampered by the lack of sophisticated genetic tools. Since no expression vectors have been developed from endogenous Actinoplanes plasmids and expression signals, engineering of antibiotic biosynthesis relies on the use of vector systems derived from Streptomyces. While PhiC31 derived vectors were shown to integrate efficiently into the chromosome of Actinoplanes, information on promoter activity is missing. The manuscript describes the investigation of several different promoter systems which are widely used in Streptomyces in A. friuliensis by promoter probe experiments using eGFP as a reporter. These experiments indicated that promoter strength in A. friuliensis did not correlate to activity in Streptomyces lividans. The ermE* promoter regarded as one of the strongest promoter in Streptomyces has only low activity in A. friuliensis. In contrast, the promoter of the apramycin resistance gene aac(3)IV, originating from the Gram-negative Escherichia coli had the highest activity. By real-time RT-PCR experiments the transcription activity of ermE* promoter in comparison to a native promoter of the friulimicin biosynthetic gene cluster was analysed. This confirmed the results of the promoter probe experiments that indicated quite weak promoter activity of P-ermE* in Actinoplanes.

Analysis of RegA, a pathway-specific regulator of the friulimicin biosynthesis in Actinoplanes friuliensis.

The rare actinomycete Actinoplanes friuliensis is the producer of the lipopeptide antibiotic friulimicin, which is active against a broad range of Gram-positive bacteria such as methicillin-resistant Enterococcus spec. and Staphylococcus aureus (MRE, MRSA) strains. Friulimicin consists of a decapeptide core and an acyl residue linked to an exocyclic amino acid. The complete biosynthetic gene cluster consisting of 24 open reading frames was characterized by sequence analysis and the transcription units were subsequently determined by RT-PCR experiments. In addition to several genes for biosynthesis, self-resistance and transport four different regulatory genes (regA, regB, regC and regD) were identified within the cluster. To analyse the role of the pathway-specific regulatory protein RegA in the friulimicin biosynthesis, the corresponding gene was inactivated resulting in friulimicin non-producing mutants. Furthermore, several protein-binding sites within the friulimicin gene cluster were identified by gel retardation assays. By real-time RT-PCR experiments, it was shown that the majority of the friulimicin biosynthetic genes is positively regulated by RegA.

Towards industrial application of quasi real-time metabolic flux analysis for mammalian cell culture.

Cellular physiology and metabolism were monitored using a quasi real-time combination of on-line and off-line data to estimate metabolic fluxes in an established bioreaction network. The utility of this approach towards optimizing bioreactor operation was demonstrated for CHO cells cultivated in 15 L perfusion reactors at 20 x 10(6) cells/mL. Medium composition and dilution rates were changed to obtain several steady states with varying glucose and glutamine concentrations. When cells were restored to initial culture medium and perfusion rate conditions after being exposed to lower glucose and glutamine concentrations, the pyruvate flux into the TCA cycle was increased 30% while the pyruvate flux through lactate was decreased 30%, suggesting steady-state multiplicity. By appropriately altering cellular metabolism, perfusion bioreactors can operate at lower perfusion rates without significant accumulation of inhibitory metabolites such as lactate. Changes in glucose, lactate and glutamine uptake/production rates had significant effects on the calculation of other fluxes in the network. Sensitivity analysis of these key metabolic fluxes highlighted the need for accurate and reliable real-time sensors. Overall, rapid observation of metabolic fluxes can be a valuable tool for bioprocess development, monitoring and control. The framework presented in this study offers a convenient means for quasi real-time estimation of metabolic fluxes and represents a step towards realizing the potential of metabolic flux analysis for accelerated bioprocess optimization.