Scale-up from shake flasks to bioreactor, based on power input and Streptomyces lividans morphology, for the production of recombinant APA (45/47 kDa protein) from Mycobacterium tuberculosis

Culture conditions in shake flasks affect filamentous Streptomyces lividans morphology, as well the productivity and O-mannosylation of recombinant Ala-Pro-rich O-glycoprotein (known as the 45/47 kDa or APA antigen) from Mycobacterium tuberculosis. In order to scale up from previous reported shake flasks to bioreactor, data from the literature on the effect of agitation on morphology of Streptomyces strains were used to obtain gassed volumetric power input values that can be used to obtain a morphology of S. lividans in bioreactor similar to the morphology previously reported in coiled/baffled shake flasks by our group. Morphology of S. lividans was successfully scaled-up, obtaining similar mycelial sizes in both scales with diameters of 0.21 ± 0.09 mm in baffled and coiled shake flasks, and 0.15 ± 0.01 mm in the bioreactor. Moreover, the specific growth rate was successfully scaled up (0.09 ± 0.02 and 0.12 ± 0.01 h^?1, for bioreactors and flasks, respectively), and the recombinant protein productivity measured by densitometry, as well. More interestingly, the quality of the recombinant glycoprotein measured as the amount of mannoses attached to the C-terminal of APA was also scaled- up; with up to five mannose residues in cultures carried out in shake flasks; and six in the bioreactor. However, final biomass concentration was not similar, indicating that although the process can be scaled-up using the power input, others factors like oxygen transfer rate, tip speed or energy dissipation/circulation function can be an influence on bacterial metabolism.     

Is an organic nitrogen source needed for cellulase production by Trichoderma reesei Rut-C30?

The effect of organic and inorganic nitrogen sources on Trichoderma reesei Rut-C30 cellulase production was investigated in submerged cultivations. Stirred tank bioreactors and shake flasks, with and without pH control, respectively, were employed. The experimental design involved the addition of individual organic nitrogen sources (soy peptone, glutamate, glycine and alanine) within a basal medium containing Avicel (i.e. micro crystalline cellulose) and ammonium sulphate. It was found that in the shake flask experiments, the highest cellulase activities (~0.1 ± 0.02 FPU ml^?1) were obtained with media containing soy peptone (3–6 g l^?1) and glutamate (3.6 g l^?1). However, these improvements in the cellulase titers in the presence of the organic nitrogen sources appeared to be related to smaller changes in the pH of the medium. This was confirmed using stirred tank bioreactors with pH control. No significant differences were observed in the highest cellulase titers and the protein pattern (according to the SDS-PAGE) of supernatants from pH controlled stirred tank bioreactor cultivations, when different nitrogen sources were used in the medium. Here the cellulase activities (~1.0 ± 0.2 FPU ml^?1) were also much greater (8–150 times) than in shake flask cultivation. Consequently, the addition of ammonium sulphate as sole nitrogen source to Avicel basal medium is recommended when performing cultivations in stirred tank bioreactors with strict pH controlled conditions.     

Improving the production yield and productivity of 1,3-dihydroxyacetone from glycerol fermentation using <Emphasis Type="Italic">Gluconobacter oxydans</Emphasis> NL71 in a compressed oxygen supply-sealed and stirred tank reactor (COS-SSTR)

In this study, a compressed oxygen gas supply was connected to a sealed aerated stirred tank reactor (COS-SSTR) bio-system, leading to a high-oxygen pressure bioreactor used to improve the bio-transformative performance in the production of 1,3-dihydroxyacetone (DHA) from glycerol using Gluconobacter oxydans NL71. A concentration of 301.2 ± 8.2 g L^?1 DHA was obtained from glycerol after 32 h of fed-batch fermentation in the COS-SSTR system. The volumetric productivity for this process was 9.41 ± 0.23 g L^?1 h^?1, which is presently the highest obtained level of glycerol bioconversion into DHA. These results show that the application of this bioreactor would enable microbial production of DHA from glycerol at the industrial scale.     

Improved Biomass and Hydrocarbon Productivity of Botryococcus braunii by Periodic Ultrasound Stimulation

Ultrasonic treatment was firstly found to accelerate both biomass and hydrocarbon productivities of Botryococcus braunii algal cells cultured in the shake flasks. The most effective sonication strategy was to subject the cells to three 5-min ultrasonic treatments at a 4-day interval using a fixed frequency of 40 kHz and power of 240 W, and the ultrasound-treated algal cells showed the highest biomass productivity of 0.043 g L^?1 day^?1 and the highest hydrocarbon productivity of 13.1 mg L^?1 day^?1 among all ultrasound treatments tested. The improved productivity was proved to be mainly due to the enhancement of both endogenous indole-3-acetic acid (IAA) biosynthesis and membrane permeability in the ultrasound-stimulated algal cells. The efficient ultrasonic stimulation strategy also showed good performance for the algal culture in a 2-L airlift bioreactor. Together, these results not only illustrate the immense potential for an enhanced understanding on ultrasound-stimulated algal cells but also as provide a powerful process intensification method to improve the biomass and hydrocarbon productivity of B. braunii in practice.     

Effect of sugar concentration on ginsenoside biosynthesis in hairy root cultures of Panax quinquefolium cultivated in shake flasks and nutrient sprinkle bioreactor

The study assessed the influence of sugar concentration (10, 20, 30, 50, 70, 100, 120 g l^?1) on growth and ginsenoside biosynthesis in Panax quinquefolium hairy roots cultivated in shake flasks and a nutrient sprinkle bioreactor. The highest growth rate was achieved in medium containing 3–5 % sucrose. More than 70 g l^?1 or less than 20 g l^?1 sugar content in the medium induces significant inhibition of root growth when cultivated in shake flasks. The saponin content was determined using HPLC. The maximum yield (above 9 mg g^?1 d.w.) of the sum of six examined ginsenosides (Rb₁, Rb₂, Rc, Rd, Re and Rg₁) in hairy roots cultivated in shake flasks was obtained with 30 g l^?1 sucrose in the medium. The sucrose concentration in the medium was found to correlate with saponin content in bioreactor-cultured specimens. A higher level of protopanaxadiol derivatives was found for lower (20 and 30 g l^?1) sucrose concentrations; higher sucrose concentrations (50 and 70 g l^?1) in the medium stimulated a higher level of Rg group saponins.     

Mathematical modelling and process optimization of a continuous 5-stage bioreactor cascade for production of poly[-(R)-3-hydroxybutyrate] by Cupriavidus necator

A multistage system for poly(hydroxyalkanoate) (PHA) production consisting of five continuous stirred tank reactors in series (5-CSTR) with Cupriavidus necator DSM 545 as production strain was modelled using formal kinetic relations. Partially growth-associated production of PHA under nitrogen limited growth was chosen as modelling strategy, thus the Luedeking-Piret’s model of partial growth-associated product synthesis was applied as working hypothesis. Specific growth rate relations adjusted for double substrate (C and N source) limited growth according to Megee et al. and Mankad-Bungay relation were tested. The first stage of the reactor cascade was modelled according to the principle of nutrient balanced continuous biomass production system, the second one as two substrate controlled process, while the three subsequent reactors were adjusted to produce PHB under continuous C source fed and nitrogen deficiency. Simulated results of production obtained by the applied mathematical models and computational optimization indicate that PHB productivity of the whole system could be significantly increased (from experimentally achieved 2.14 g L^?1 h^?1 to simulated 9.95 g L^?1 h^?1) if certain experimental conditions would have been applied (overall dilution rate, C and N source feed concentration). Additionally, supplemental feeding strategy for switching from batch to continuous mode of cultivation was proposed to avoid substrate inhibition.     

Inulinolytic activity of broths of Aspergillus niger ATCC 204447 cultivated in shake flasks and stirred tank bioreactor

It is the first detailed study of an inulinolytic fungus Aspergillus niger ATCC 204447 since its discovery, covering submerged cultivations both in shake flasks and a stirred tank bioreactor. Various carbon sources were applied to induce the inulinolytic activity in shake flask cultures. The highest volumetric and specific (per gram of biomass) activities (respectively 0.68 U/mL and 184 U g/X) were observed for the initial inulin and sucrose concentrations equal to 20 g/L. The fungus grew as large (>3 mm) spherical pellets. The influence of inoculum density and application of microparticle‐enhanced cultivation (MPEC) were studied in the batch bioreactor cultivations. Inoculum density moderately affected the inulinolytic activities, whose highest values were 0.7 U/mL and 165 U g/X at the lowest studied spore density of 3.33·10⁸ L^?1. Dispersed hyphae evolved in the bioreactor made the broth difficult to aerate due to high apparent viscosity (exceeding 200 Pa sⁿ at shear rate about 0.05 s^?1) and shear thinned properties (flow behavior index below 0.2). In MPEC (10 μm talc microparticles) the pellets of diameter between 1 and 2 mm were formed, which facilitated the aeration of the broth and increased the specific inulinolytic activity 3.5‐fold.     

Nitrogen and phosphorus as the factors affecting ginsenoside production in hairy root cultures of <Emphasis Type="Italic">Panax quinquefolium</Emphasis> cultivated in shake flasks and nutrient sprinkle bioreactor

The study assesses the influence of different concentrations of nitrogen and phosphorus sources on ginsenoside biosynthesis in Panax quinquefolium hairy roots cultivated in shake flasks and a nutrient sprinkle bioreactor. The saponin content was determined using HPLC. The maximum yield (12.45 mg g^?1 dw) of the sum of six examined ginsenosides (Rb1, Rb2, Rc, Rd, Re and Rg1) in hairy roots cultivated in shake flasks was achieved in modified Gamborg B-5 medium containing 0.83 mM l^?1 phosphate, 12.4 mM l^?1 nitrate and 0.5 mM l^?1 ammonium. The yield itself was 1.93 times higher than that achieved in standard Gamborg medium. The modified medium also favourably influenced the biosynthesis of studied saponins in bioreactor cultures. The saponin content (35.11 mg g^?1 d.w.) was 2.75-times higher than that achieved in control medium.     

Effects of flashing light-emitting diodes on algal biomass productivity

To reduce power consumption and enhance algal biomass productivity in a thin flat-plate bioreactor (called a sliver tank bioreactor), flashing (pulsing) light was used. Biomass productivity and power consumption were monitored in controlled experiments using various photon flux levels, including a constant (non-flashing) flux of 75 μmol photons m^?2 s^?1 and three flashing experiments with photon fluxes of 375, 275, and 175 μmol photons m^?2 s^?1. Flashing experiments were performed at 10 kHz and a duty cycle of 20 %. A sliver tank bioreactor with a chamber width of 6.4 mm was used for its short optical path. Data from the experiments where light was flashed with a photon flux of 375 μmol photons m^?2 s^?1 indicated 9.6 % less power and 2.86 times the biomass productivity compared to the constant photon flux experiments. Similar results were obtained for the other flashing light regimes, which had lower biomass yields but also less input power per unit biomass produced, indicating that a large fraction of the continuously applied photons are shed or wasted, even at levels approximately 1/30th the intensity of full sun.     

Oxygen mass transfer and scale-up studies in baffled roller bioreactors

Oxygen mass transfer was studied in conventional, bead mill and baffled roller bioreactors. Using central composite rotational design, impacts of size, rotation speed and working volume on the oxygen mass transfer were evaluated. Baffled roller bioreactor outperformed its conventional and bead mill counterparts, with the highest k _L a obtained in these configurations being 0.58, 0.19, 0.41 min^?1, respectively. Performances of the bead mill and baffled roller bioreactor were only comparable when a high bead loading (40 %) was applied. Regardless of configuration increase in rotation speed and decrease in working volume improved the oxygen mass transfer rate. Increase in size led to enhanced mass transfer and higher k _L a in baffled roller bioreactor (0.49 min^?1 for 2.2 L and 1.31 min^?1 for 55 L bioreactors). Finally, the experimentally determined k _L a in the baffled roller bioreactors of different sizes fit reasonably well to an empirical correlation describing the k _L a in terms of dimensionless numbers.     

From field sampling to pneumatic bioreactor mycelia production of the ectomycorrhizal mushroom Laccaria trichodermophora

In order to increase survival rates of greenhouse seedlings destined for restoration and conservation programs, successful mycorrhization of the seedlings is necessary. To reforest forest ecosystems, host trees must be inoculated with ectomycorrhizal fungi and, in order to guarantee a sufficient supply of ectomycorrhizal inoculum, it is necessary to develop technologies for the mass production of ectomycorrhizal fungi mycelia. We selected the ectomycorrhizal fungus Laccaria trichodermophora, due to its ecological traits and feasible mycelia production in asymbiotic conditions. Here, we report the field sampling of genetic resources, as well as the highly productive nutritional media and cultivation parameters in solid cultures. Furthermore, in order to achieve high mycelial production, we used strain screening and evaluated pH, carbon source concentration, and culture conditions of submerged cultures in normal and baffled shake flasks. The higher productivity culture conditions in shake flasks were selected for evaluation in a pneumatic bioreactor, using modified BAF media with a 10 g/L glucose, pH 5.5, 25 °C, and a volumetric oxygen transfer coefficient (K_La) of 36 h⁻¹. Under those conditions less biomass (12–37 %) was produced in the pneumatic bioreactor compared with the baffled shake flasks. This approach shows that L. trichodermophora can generate a large biomass concentration and constitute the biotechnological foundation of its mycelia mass production.     

Process Engineering for High-Cell-Density Cultivation of Lipid Rich Microalgal Biomass of <Emphasis Type="Italic">Chlorella</Emphasis> sp. FC2 IITG

In the present study, process engineering strategy was applied to achieve lipid-rich biomass with high density of Chlorella sp. FC2 IITG under photoautotrophic condition. The strategy involved medium optimization, intermittent feeding of limiting nutrients, dynamic change in light intensity, and decoupling growth and lipid induction phases. Medium optimization was performed using combinations of artificial neural network or response surface methodology with genetic algorithm (ANN-GA and RSM-GA). Further, a fed-batch operation was employed to achieve high cell density with intermittent feeding of nitrate and phosphate along with stepwise increase in light intensity. Finally, mutually exclusive biomass and lipid production phases were decoupled into two-stage cultivation process: biomass generation in first stage under nutrient sufficient condition followed by lipid enrichment through nitrogen starvation. The key findings were as follows: (i) ANN-GA resulted in an increase in biomass titer of 157 % (0.95 g L^?1) in shake flask and 42.8 % (1.0 g L^?1) in bioreactor against unoptimized medium at light intensity of 20 μE m^?2 s^?1; (ii) further optimization of light intensity in bioreactor gave significantly improved biomass titer of 5.6 g L^?1 at light intensity of 250 μE m^?2 s^?1; (iii) high cell density of 13.5 g L^?1 with biomass productivity of 675 mg L^?1 day^?1 was achieved with dynamic increase in light intensity and intermittent feeding of limiting nutrients; (iv) finally, two-phase cultivation resulted in biomass titer of 17.7 g L^?1 and total lipid productivity of 313 mg L^?1 day^?1 which was highest among Chlorella sp. under photoautotrophic condition.     

Scale-up from shake flasks to pilot-scale production of the plant growth-promoting bacterium Azospirillum brasilense for preparing a liquid inoculant formulation

Azospirillum brasilense has industrial significance as a growth promoter in plants of commercial interest. However, there is no report in the literature disclosing a liquid product produced in pilot-scale bioreactors and is able to be stored at room temperature for more than 2 years. The aim of this work was to scale up a process from a shake flask to a 10-L lab-scale and 1,000-L pilot-scale bioreactor for the production of plant growth-promoting bacterium A. brasilense for a liquid inoculant formulation. Furthermore, this work aimed to determine the shelf life of the liquid formulation stored at room temperature and to increase maize crops yield in greenhouses. Under a constant oxygen mass transfer coefficient (K _L a), a fermentation process was successfully scaled up from shake flasks to 10- and 1,000-L bioreactors. A concentration ranging from 3.5 to 7.5?×?10⁸ CFU/mL was obtained in shake flasks and bioreactors, and after 2 years stored at room temperature, the liquid formulation showed one order of magnitude decrease. Applications of the cultured bacteria in maize yields resulted in increases of up to 95 % in corncobs and 70 % in aboveground biomass.     

Growth and biomass production with enhanced β‐glucan and dietary fibre contents of Ganoderma australe ATHUM 4345 in a batch‐stirred tank bioreactor

In this study we maximized biomass production by the basidiomycete Ganoderma australe ATHUM 4345, a species of pharmaceutical interest as it is a valuable source of nutraceuticals, including dietary fibers and glucans. We used the Biolog FF MicroPlate to screen 95 different carbon sources for growth monitoring. The pattern of substrate catabolism forms a substrate assimilation fingerprint, which is useful in selecting components for media optimization of maximum biomass production. Response surface methodology, based on the central composite design was applied to explore the optimum concentrations of carbon and nitrogen sources of culture medium in shake flask cultures. When the improved culture medium was tested in a 20‐L stirred tank bioreactor, using 13.7 g/L glucose and 30.0 g/L yeast extract, high biomass yields (10.1±0.4 g/L) and productivity of 0.09 g L^?1 h^?1 were obtained. The yield coefficients for total glucan and dietary fibers on biomass formed were 94.82±6 and 341.15±12.3 mg/g mycelium dry weight, respectively.     

Kinetic modeling and scale up of lipoic acid (LA) production from Saccharomyces cerevisiae in a stirred tank bioreactor

Scale up studies for production of lipoic acid (LA) from Saccharomyces cerevisiae have been reported in this paper for the first time. LA production in batch mode was carried out in a stirred tank bioreactor at varying agitation and aeration with maximum LA production of 512 mg/L obtained at 350 rpm and 25 % dissolved oxygen in batch culture conditions. Thus, LA production increased from 352 mg/L in shake flask to 512 mg/L in batch mode in a 5 L stirred tank bioreactor. Biomass production under these conditions was mathematically explained using logistic equation and data obtained for LA production and substrate utilization were successfully fitted using Luedeking–Piret and Mercier’s models. The kinetic studies showed LA production to be growth associated. Further enhancement of LA production was carried out using fed-batch (variable volume) and semi-continuous modes of fermentation. Semi-continuous fermentation with three feeding cycles of sucrose effectively increased the production of LA from 512 to 725 mg/L.     

Effect of bioreactor configuration on the growth and maturation of Picea sitchensis somatic embryo cultures

Somatic embryo suspension cultures of Picea sitchensis (Sitka spruce) derived from two cell lines, SS03 and SS10, were grown in shake flasks, air-lift, bubble, stirred tank and hanging stirrer bar bioreactors. Cell line SS03 yielded freely suspended and individual stage 1 embryos, while the embryos of SS10 were present in large aggregates. Compared to shake flasks, proliferation in bioreactors resulted in increased biomass; however, cell line morphology influenced the effect of different bioreactor configurations on growth and maturation of embryo cultures. Somatic embryos grown in shake flasks and bioreactors were matured on gelled solid medium and in submerged culture where gelled solid medium was covered with a layer of liquid medium. The number of stage 3 (mature) embryos produced from SS03 in the bubble bioreactor was significantly higher than those from stirred tank and hanging stirrer bar bioreactors with both solid medium and submerged culture. Submerged culture was unsuitable for SS10 embryo maturation. This revised version was published online in June 2006 with corrections to the Cover Date.     

Reaction engineering analysis of cellulase production with Trichoderma reesei RUT-C30 with intermittent substrate supply

The effects of varying initial concentrations of microcrystalline cellulose on cellulase production with Trichoderma reesei RUT-C30 as well as the effects of varying lactose and ammonium sulfate concentrations in the feed medium were studied simultaneously in parallel-operated shake flasks and, alternatively, in parallel-operated stirred-tank bioreactors on a 10-mL scale. Fifteen experiments were performed as triplicates in shake flasks as well as in stirred-tank bioreactors in parallel to identify the parameters of second-order polynomials for the estimation of the final filter paper activity of T. reesei RUT-C30 after a process time of 96 h. Even though parameter estimation was not possible based on the results of the shake flasks due to final enzyme activities at or below the detection limit (with the exception of one shake flask), the identification of the second-order polynomial was successful with the results of the parallel-operated stirred-tank bioreactors on a 10-mL scale. Reaction conditions with 53.3 g L^?1 microcrystalline cellulose in the initial medium, no lactose feeding and 3.3 g L^?1 day^?1 intermittent ammonium sulfate addition were estimated to be optimal. The final experimental validation of the optimum substrate supply on a L-scale resulted in the production of 4.88 filter paper units (FPU) mL^?1 with T. reesei RUT-C30 after 96 h. This is an improvement by a factor of 3.6 compared to the reference batch process (1.35 FPU mL^?1).     

Production of alginate by Azotobacter vinelandii grown at two bioreactor scales under oxygen-limited conditions

The oxygen transfer rate (OTR) was evaluated as a scale-up criterion for alginate production in 3- and 14-L stirred fermentors. Batch cultures were performed at different agitation rates (200, 300, and 600 rpm) and airflow rates (0.25, 0.5, and 1 vvm), resulting in different maximum OTR levels (OTR_max). Although the two reactors had a similar OTR_max (19 mmol L^?1 h^?1) and produced the same alginate concentration (3.8 g L^?1), during the cell growth period the maximum molecular weight of the alginate was 1,250 kDa in the 3-L stirred fermentor and 590 kDa in 14-L stirred fermentor. The results showed for the first time the evolution of the molecular weight of alginate and OTR profiles for two different scales of stirred fermentors. There was a different maximum specific oxygen uptake rate between the two fermenters, reaching 8.3 mmol g^?1 h^?1 in 3-L bioreactor and 10.6 mmol g^?1 h^?1 in 14-L bioreactor, which could explain the different molecular weights observed. These findings open the possibility of using $ q_{{{\text{O}}_{ 2} }} $ instead of OTR_max as a scaling criterion to produce polymers with similar molecular weights during fermentation.     

Biochemical investigation of kraft lignin degradation by Pandoraea sp. B-6 isolated from bamboo slips

Kraft lignin (KL) is the major pollutant in black liquor. The bacterial strain Pandoraea sp. B-6 was able to degrade KL without any co-substrate under high alkaline conditions. At least 38.2 % of chemical oxygen demand and 41.6 % of color were removed in 7 days at concentrations from 1 to 6 g L^?1. The optimum pH for KL degradation was 10 and the optimum temperature was 30 °C. The greatest activities of 2,249.2 U L^?1 for manganese peroxidase and 1,120.6 U L^?1 for laccase were detected on the third and fifth day at pH 10, respectively. Many small molecules, such as cinnamic acid, ferulic acid, 2-hydroxy benzyl alcohol, and vanillyl methyl ketone, were formed during the period of KL degradation based on GC–MS analysis. These results indicate that this strain has great potential for biotreatment of black liquor.     

Induction of laccases in Trametes versicolor by aqueous wood extracts

The induction of laccase isoforms in Trametes versicolor HEMIM-9 by aqueous extracts (AE) from softwood and hardwood was studied. Samples of sawdust of Pinus sp., Cedrela sp., and Quercus sp. were boiled in water to obtain AE. Different volumes of each AE were added to fungal cultures to determine the amount of AE needed for the induction experiments. Laccase activity was assayed every 24 h for 15 days. The addition of each AE (50 to 150 μl) to the fungal cultures increased laccase production compared to the control (0.42 ± 0.01 U ml^?1). The highest laccase activities detected were 1.92 ± 0.15 U ml^?1 (pine), 1.87 ± 0.26 U ml^?1 (cedar), and 1.56 ± 0.34 U ml^?1 (oak); laccase productivities were also significantly increased. Larger volumes of any AE inhibited mycelial growth. Electrophoretic analysis revealed two laccase bands (lcc1 and lcc2) for all the treatments. However, when lcc2 was analyzed by isoelectric focusing, inducer-dependent isoform patterns composed of three (pine AE), four (oak AE), and six laccase bands (cedar AE) were observed. Thus, AE from softwood and hardwood had induction effects in T. versicolor HEMIM-9, as indicated by the increase in laccase activity and different isoform patterns. All of the enzymatic extracts were able to decolorize the dye Orange II. Dye decolorization was mainly influenced by pH. The optimum pH for decolorization was pH 5 (85 %), followed by pH 7 (50 %) and pH 3 (15 %). No significant differences in the dye decolorizing capacity were detected between the control and the differentially induced laccase extracts (oak, pine and cedar). This could be due to the catalytic activities of isoforms with pI 5.4 and 5.8, which were detected under all induction conditions.