High-yield bacillus subtilis protease production by solid-state fermentation

A Bacillus subtilis isolate was shown to be able to produce extracellular protease in solid-state fermentations (SSF) using soy cake as culture medium. A significant effect of inoculum concentration and physiological age on protease production was observed. Maximum activities were obtained for inocula consisting of exponentially growing cells at inoculum concentrations in the range of 0.7–2.0 mg g⁻¹. A comparative study on the influence of cultivation temperature and initial medium pH on protease production in SSF and in submerged fermentation (SF) revealed that in SSF a broader pH range (5–10), but the same optimum temperature (37°C), is obtained when compared to SF. A kinetic study showed that enzyme production is associated with bacterial growth and that enzyme inactivation begins before biomass reaches a maximum level for both SF and SSF. Maximum protease activity and productivity were 960 U g⁻¹ and 15.4 U g⁻¹ h⁻¹ for SSF, and 12 U mL⁻¹ and 1.3 U mL⁻¹ h⁻¹ for SF. When SSF protease activity was expressed by volume of enzyme extract, the enzyme level was 10-fold higher and the enzyme productivity 45% higher than in SF. These results indicate that this bacterial strain shows a high biotechnological potential for protease production in solid-state fermentation.     

Effect of media composition and growth conditions on production of β-glucosidase by Aspergillus niger C-6

The hydrolytic activity of fungal originated β-glucosidase is exploited in several biotechnological processes to increase the rate and extent of saccharification of several cellulosic materials by hydrolyzing the cellobiose which inhibits cellulases. In a previous presentation, we reported the screening and liquid fermentation with Aspergillus niger, strain C-6 for β-glucosidase production at shake flask cultures in a basal culture medium with mineral salts, corn syrup liquor, and different waste lignocellulosic materials as the sole carbon source obtaining the maximum enzymatic activity after 5–6 d of 8.5 IU/mL using native sugar cane bagasse. In this work we describe the evaluation of fermentation conditions: growth temperature, medium composition, and pH, also the agitation and aeration effects for β-glucosidase production under submerged culture using a culture media with corn syrup liquor (CSL) and native sugar cane bagasse pith as the sole carbon source in a laboratory fermenter. The maximum enzyme titer of 7.2 IU/mL was obtained within 3 d of fermentation. This indicates that β-glucosidase productivity by Aspergillus niger C-6 is function of culture conditions, principally temperature, pH, culture medium conditions, and the oxygen supply given in the bioreactor. Results obtained suggest that this strain is a potential microorganism that can reach a major level of enzyme production and also for enzyme characterization.     

Invertase production on solid-state fermentation by Aspergillus niger strains improved by parasexual recombination

Invertase production by Aspergillus niger grown by solid-state fermentation was found to be higher than by conventional submerged fermentation. The haploid mutant strains Aw96-3 and Aw96-4 showed better productivity of various enzymes, as compared to wild-type parental strain A. niger C28B25. Here we use parasexual crosses of those mutants to increase further the productivity of invertase in solid-state fermentation. We isolated both a diploid (DAR2) and an autodiploid (AD96-4) strain, which were able to grow in minimal medium after mutation complementation of previously isolated haploid auxotrophic strains. Invertase production was measured in solid-state fermentation cultures, using polyurethane foam as an inert support for fungal growth. Water activity value (Aw) was adjusted to 0.96, since low Aw values are characteristic in some solid-state fermentation processes. Such diploid strains showed invertase productivity levels 5–18 times higher than levels achieved by the corresponding haploid strains. For instance, values for C28B25, Aw96-3, Aw96-4, DAR2, and AD96-4 were 441, 254, 62, 1324, and 2677 IU/(L·h), respectively. These results showed that genetic recombination, achieved through parasexual crosses in A. niger, results in improved strains with potential applications for solid-state fermentation processes.     

Statistical Optimization of Alpha-Amylase Production with Immobilized Cells of <Emphasis Type="Italic">Streptomyces erumpens</Emphasis> MTCC 7317 in <Emphasis Type="Italic">Luffa cylindrica</Emphasis> L. Sponge Discs

The purpose of this investigation was to study the effect of Streptomyces erumpens cells immobilized in various matrices, i.e., agar–agar, polyacrylamide, and luffa (Luffa cylindrica L.) sponge for production of α-amylase. Luffa sponge was found to be 21% and 51% more effective in enzyme yield than agar–agar and polyacrylamide, respectively. Response surface methodology was used to evaluate the effect of three main variables, i.e., incubation period, pH, and temperature on enzyme production with immobilized luffa cells. The experimental results showed that the optimum incubation period, pH, and temperature were 36h, 6.0, and 50 °C, respectively. The repeated batch fermentation of immobilized cells in shake flasks showed that S. erumpens cells were more or less equally physiologically active on the support even after three cycles of fermentation (3,830–3,575 units). The application of S. erumpens crude enzyme in liquefying cassava starch was studied. The maximum hydrolysis of cassava starch (85%) was obtained with the application of 4ml (15,200 units) of crude enzyme after 5 h of incubation.     

Biosorption of Lead,Copper, and Cadmium by <Emphasis Type="Italic">Phanerochaete chrysosporium</Emphasis> in Ternary Metal Mixtures: Statistical Analysis of Individual and Interaction Effects

Biosorption of three divalent metals, viz., lead, copper, and cadmium in ternary aqueous mixtures was studied using Phanerochaete chrysosporium in batch shake flasks. The mixtures were prepared containing the metals at their either varying optimum or equal initial concentration combinations in aqueous solution of pH optimum to each of the metals. Following were the optimum initial concentration ranges of the metals in mixture: lead, 60–100 mg/L; copper, 20–60 mg/L; and cadmium, 5–15 mg/L. And, for varying these optimum concentration levels of the metals, a 2³ full factorial design of experiments was employed. The results revealed that an increase in lead and cadmium concentrations helped in their better biosorption by the fungus, but an increase in initial copper concentration slightly diminished its removal. Statistical analysis of the results in the form of analysis of variance and Student t test gave a clear interpretation on the roles of both the individual metals and their interactions in the uptake of metals from mixture. Compared to the uptake of metals when presented individually, lead biosorption in mixture was found to be enhanced to a degree as high as 99%; on the other hand, copper and cadmium removals from mixtures were inhibited to the extent of 100% and 98%, respectively. However, this extent of inhibition or enhancement in the metal removals compared to the individual removals was less in mixtures containing all equal concentrations of the metals.     

Overproduction of Glucoamylase by a Deregulated Mutant of a Thermophilic Mould <Emphasis Type="Italic">Thermomucor indicae-seudaticae</Emphasis>

Thermomucor indicae-seudaticae, a glucoamylase-producing thermophilic mould, was mutagenised using nitrous acid and gamma (⁶⁰Co) irradiation in a sequential manner to isolate deregulated mutants for enhanced production of glucoamylase. The mutants were isolated on Emerson YpSs agar containing a non-metabolisable glucose analogue 2-deoxy-d-glucose (2-DG) for selection. The preliminary screening for glucoamylase production using starch–iodine plate assay followed by quantitative confirmation in submerged fermentation permitted the isolation of several variants showing varying levels of derepression and glucoamylase secretion. The mutant strain T. indicae-seudaticae CR19 was able to grow in the presence of 0.5 g l⁻¹ 2-DG and produced 1.8-fold higher glucoamylase. As with the parent strain, glucoamylase production by T. indicae-seudaticae CR19 in 250-ml Erlenmeyer flasks attained a peak in 48 h of fermentation, showing higher glucoamylase productivity (0.67 U ml⁻¹ h⁻¹) than the former (0.375 U ml⁻¹ h⁻¹). A large-scale cultivation in 5-l laboratory bioreactor confirmed similar fermentation profiles, though the glucoamylase production peak was attained within 36 h attributable to the better control of process parameters. Although the mutant grew slightly slow in the presence of 2-DG and exhibited less sporulation, it showed faster growth on normal Emerson medium with a higher specific growth rate (0.138 h⁻¹) compared to the parent strain (0.123 h⁻¹). The glucoamylase produced by both strains was optimally active at 60 °C and pH 7.0 and displayed broad substrate specificity by cleaving α-1,4- and α-1,6-glycosidic linkages in starch, amylopectin, amylose and pullulan. Improved productivity and higher specific growth rate make T. indicae-seudaticae CR19 a useful strain for glucoamylase production.     

Kinetic and Stoichiometric Parameters in the Production of Carotenoids by <Emphasis Type="Italic">Sporidiobolus salmonicolor</Emphasis> (CBS 2636) in Synthetic and Agroindustrial Media

With the objective of determining the kinetic behavior (growth, substrate, pH, and carotenoid production) and obtain the stoichiometric parameters of the fermentative process by Sporidiobolus salmonicolor in synthetic and agroindustrial media, fermentations were carried out in shaken flasks at 25°C, 180 rpm, and initial pH of 4.0 for 120 h in the dark, sampling every 6 h. The maximum concentrations of total carotenoids in synthetic (913 μg/L) and agroindustrial (502 μg/L) media were attained approximately 100 h after the start of the fermentative process. Carotenoid bioproduction is associated with cell growth and the ratio between carotenoid production and cell growth (Y _P/X) is 176 and 163 μg/g in the synthetic and agroindustrial media, respectively. The pH of the agroindustrial fermentation medium varied from 4.2 to 8.5 during the fermentation. The specific growth rate (μ _X) for S. salmonicolor in synthetic and agroindustrial media was 0.07 and 0.04 h⁻¹, respectively. The synthetic medium allowed for greater productivity, obtaining maximum cell productivity (P _x) of 0.08 g L⁻¹ h⁻¹ and maximum total carotenoid productivity (P _car) of 14.2 μg L⁻¹ h⁻¹. Knowledge of the kinetics of a fermentative process is of extreme importance when transposing a laboratory experiment to an industrial scale, as well as making a quantitative comparison between different culture conditions.     

Optimization of glucoamylase production by Aspergillus niger in solid-state fermentation

Glucoamylase production by Aspergillus niger in solid-state fermentation was optimized using factorial design and response surface techniques. The variables evaluated were pH and bed thickness in tray, having as response enzyme production and productivity. The bed thickness in tray was the most significant variable for both responses. The highest values for glucoamylase production occurred using pH 4.5 and bed thickness in the inferior limits at 2.0–4.2 cm. For productivity, the optimal conditions were at pH 4.5 as well and bed thickness from 4.4 to 7.5 cm. The optimal conditions for glucoamylase production while obtaining high activity without loss of productivity were pH 4.5 and bed thickness in tray from 4.0 to 4.5 cm, which resulted in an enzyme production of 695 U/g and productivity of 5791 U/h.     

Evaluation of solid and submerged fermentations for the production of cyclodextrin glycosyltransferase by Paenibacillus campinasensis H69-3 and characterization of crude enzyme

Cyclodextrin glycosyltransferase (CGTase) is an enzyme that produces cyclodextrins from starch by an intramolecular transglycosylation reaction. Cyclodextrins have been shown to have a number of applications in the food, cosmetic, pharmaceutical, and chemical industries. In the current study, the production of CGTase by Paenibacillus campinasensis strain H69-3 was examined in submerged and solid-state fermentations. P. campinasensis strain H69-3 was isolated from the soil, which grows at 45°C, and is a Gramvariable bacterium. Different substrate sources such as wheat bran, soybean bran, soybean extract, cassava solid residue, cassava starch, corn starch, and other combinations were used in the enzyme production. CGTase activity was highest in submerged fermentations with the greatest production observed at 48–72 h. The physical and chemical properties of CGTase were determined from the crude enzyme produced from submerged fermentations. The optimum temperature was found to be 70–75°C, and the activity was stable at 55°C for 1 h. The enzyme displayed two optimum pH values, 5.5 and 9.0 and was found to be stable between a pH of 4.5 and 11.0.     

Effect of aeration on lignin peroxidase production by Streptomyces viridosporus T7A

The effect of aeration on lignin peroxidase production by Streptomyces viridosporus T7A was studied in a bench-scale bioreactor using a previously optimized growth medium (0.65% yeast extract and 0.1% corn oil, pH7.0) at 37°C and natural pH. Airflow rates of 0.3, 1.0, and 1.5 vvm and a fixed agitation of 200 rpm were initially studied followed by 1.0 vvm and 200, 300, 400, and 500 rpm. The use of 1.0 vvm and 400 rpm increased enzyme concentration 1.8-fold (100–180 U/L) and process productivity 4.8-fold (1.4–6.7 U/[L·h]) in comparison with the use of 200 rpm and 0.3 vvm. The inexpensive corn oil, used as carbon source, besides its antifoam properties, proved to be nonrepressive for enzyme production.     

Repeated solid-phase fermentation and extraction for enzyme production

Solid-phase fermentation has been found to have a much higher productivity than the popular liquid submerged fermentation in producing cellulase enzymes. The highest reported productivity in the literature for cellulases by Trichoderma cultures in submerged fermentation is 158 filter paper units (FPU)/(h·L) of fermenting liquid. From preliminary experiments of solid-phase fermentation in 1000-m L flasks, a productivity of 234 FPU of cellulases/(h·L) of solid-bed volume was obtained. When two novel techniques—pressure pulsation and repeated extraction—were applied, a productivity of 806 FPU/(h·L) was achieved. The same techniques also greatly enhanced the productivity of other enzymes by fungal cultures in solid-phase fermentation.     

Effects of pH and Temperature on Recombinant Manganese Peroxidase Production and Stability

The enzyme manganese peroxidase (MnP) is produced by numerous white-rot fungi to overcome biomass recalcitrance caused by lignin. MnP acts directly on lignin and increases access of the woody structure to synergistic wood-degrading enzymes such as cellulases and xylanases. Recombinant MnP (rMnP) can be produced in the yeast Pichia pastoris αMnP1-1 in fed-batch fermentations. The effects of pH and temperature on recombinant manganese peroxidase (rMnP) production by P. pastoris αMnP1-1 were investigated in shake flask and fed-batch fermentations. The optimum pH and temperature for a standardized fed-batch fermentation process for rMnP production in P. pastoris αMnP1-1 were determined to be pH 6 and 30 °C, respectively. P. pastoris αMnP1-1 constitutively expresses the manganese peroxidase (mnp1) complementary DNA from Phanerochaete chrysosporium, and the rMnP has similar kinetic characteristics and pH activity and stability ranges as the wild-type MnP (wtMnP). Cultivation of P. chrysosporium mycelia in stationary flasks for production of heme peroxidases is commonly conducted at low pH (pH 4.2). However, shake flask and fed-batch fermentation experiments with P. pastoris αMnP1-1 demonstrated that rMnP production is highest at pH 6, with rMnP concentrations in the medium declining rapidly at pH less than 5.5, although cell growth rates were similar from pH 4–7. Investigations of the cause of low rMnP production at low pH were consistent with the hypothesis that intracellular proteases are released from dead and lysed yeast cells during the fermentation that are active against rMnP at pH less than 5.5.     

Continuous production of succinic acid by a fumarate-reducing bacterium immobilized in a hollow-fiber bioreactor

Enterococcus faecalis RKY1, a fumarate-reducing bacterium, was immobilized in an asymmetric hollow-fiber bioreactor (HFBR) for the continuous production of succinic acid. The cells were inoculated into the shell side of the HFBR, which was operated in transverse mode. Since the pH values in the HFBR declined during continuous operation to about 5.7, it was necessary to change the feed pH from 7.0 to 8.0 after 24 h of operation in order to enhance production of succinic acid. During continuous operation with a medium containing fumarate and glycerol, the productivity of succinate was 3.0–10.9 g/(L·h) with an initial concentration of 30 g/L of fumarate, 4.9–14.9 g/(L·h) with 50 g/L of fumarate, and 7.2–17.1 g/(L·h) with 80 g/L of fumarate for dilution rates between 0.1 and 0.4 h⁻¹. The maximum productivity of succinate obtained by the HFBR (17.1 g of succinate /[L·h]) was 1.7 times higher than that of the batch bioconversions (9.9 g of succinate /[L·h]) with 80 g/L of fumarate. Furthermore, the long-term stability of the HFBR was demonstrated with a continuously efficient production of succinate for more than 15 d (360 h).     

Influence of carbon and nitrogen sources and temperature on hyperproduction of a thermotolerant β-glucosidase from synthetic medium by Kluyveromyces marxianus

The effect of carbon source and its concentration, inoculum size, yeast extract concentration, nitrogen source, pH of the fermentation medium, and fermentation temperature on β-glucosidase production by Kluyveromyces marxianus in shake-flask culture was investigated. These were the independent variables that directly regulated the specific growth and β-glucosidase production rate. The highest product yield, specific product yield, and productivity of β-glucosidase occurred in the medium (pH 5.5) inoculated with 10% (v/v) inoculum of the culture. Cellobiose (20 g/L) significantly improved β-glucosidase production measured as product yield (Y _P/S ) and volumetric productivity (Q _P ) followed by sucrose, lactose, and xylose. The highest levels of productivity (144 IU/[L·h]) of β-glucosidase occurred on cellobiose in the presence of CSL at 35°C and are significantly higher than the values reported by other researchers on almost all other organisms. The thermodynamics and kinetics of β-glucosidase production and its deactivation are also reported. The enzyme was substantially stable at 60°C and may find application in some industrial processes.     

Oxygen uptake rate in production of xylitol by Candida guilliermondii with different aeration rates and initial xylose concentrations

The global oxygen uptake rate (OUR) and specific oxygen uptake rates (SOUR) were determined for different values of the volumetric oxygen mass transfer coefficient (15, 43, and 108 h⁻¹), and for varying initial xylose concentrations (50, 100, 150, and 200 g/L) in shaking flasks. The initial cell concentration was 4.0 g/L, and there was only significant growth in the fermentation with the highest oxygen availability. In this condition, OUR increased proportionally to cell growth, reaching maximum values from 2.1 to 2.5 g of O₂/(L·h) in the stationary phase when the initial substrate concentration was raised from 50 to 200 g/L, respectively. SOUR showed different behavior, growing to a maximum value coinciding with the beginning of the exponential growth phase, after which point it decreased. The maximum SOUR values varied from 265 to 370 mg of O₂/(g of cell·h), indicating the interdependence of this parameter and the substrate concentration. Although the volumetric productivity dropped slightly from 1.55 to 1.18 g of xylitol/(L·h), the strain producing capacity (γ _P/X ) rose from 9 to 20.6 g/g when the initial substrate concentration was increased from 50 to 200 g/L. As for the xylitol yield over xylose consumed (γ _P/S ), there was no significant variation, resulting in a mean value of 0.76 g/g. The results are of interest in establishing a strategy for controlling the dynamic oxygen supply to maximize volumetric productivity.     

Early phase process scale-up challenges for fungal and filamentous bacterial cultures

Culture pelleting and morphology has a strong influence on process productivity and success for fungal and filamentous bacterial cultures. This impact is particularly evident with early phase secondary metabolite processes with limited process definition. A compilation of factors affecting filamentous or pelleting morphology described in the literature indicates potential leads for developing process-specific control methodologies. An evaluation of the factors mediating citric acid production is one example of an industrially important application of these techniques. For five model fungal and filamentous bacterial processes in an industrial fermentation pilot plant, process development strategies were developed and effectively implemented with the goal of achieving reasonable fermentation titers early in the process development cycle. Examples of approaches included the use of additives to minimize pelleting in inoculum shake flasks, the use of large-volume frozen bagged inoculum obtained from agitated seed fermentors, and variations in production medium composition and fermentor operating conditions. Results were evaluated with respect to productivity of desired secondary metabolites as well as process scalability. On-line measurements were utilized to indirectly evaluate the cultivation impact of changes in medium and process development. Key laboratory to pilot plant scale-up issues also were identified and often addressed in subsequent cultivations.     

A new approach to improving the performance ofZymomonas in continuous ethanol fermentations

Although most fermentation ethanol is currently produced in traditional batch processes with yeast, the ethanologenic bacteriumZymomonas mobilis is recognized as an alternative process organism for fuel alcohol production. Different strategies for improving the productivity of ethanol fermentations are reviewed. In batch and open-type continuous fermentations the advantage of replacing yeast byZymomonas relates principally to the 10% higher fermentation efficiency (product yield), whereas in high cell density, closed-type continuous systems (operating with cell recycle or retention) the superior kinetic properties ofZymomonas can be exploited to affect about a five-fold improvement in volumetric productivity. Unlike yeast, the rate of energy supply (conversion of glucose to ethanol) inZymomonas is not strictly regulated by the energy demand and a nongrowing culture exhibits a maintenance energy coefficient that is at least 25 times higher than yeast. As an alternative to process improvement through genetic engineering of the process organism this investigation has taken a biochemical and physiological approach to increasing the kinetic performance ofZ. mobilis through manipulation and control of the chemical environment. Energetically “uncoupled” phenotypes with markedly increased specific rates of ethanol production were generated under conditions of nutritional limitation (nitrogen, phosphate, or potassium) in steady-state continuous culture. The pH was shown to influence energy coupling inZymomonas affecting the maintenance coefficient (m _e ) rather than the max growth yield coefficient (Y _x ^sάx ). Whereas the pH for optimal growth ofZ. mobilis (ATCC 29191) in a complex medium was 6.0–6.5, the specific rate of ethanol production in continuous fermentations was maximal in the range 4.0–4.5. Fermentation conditions are specified for maximizing the specific productivity of aZymomonas-based continuous ethanol fermentation where the potential exists for improving the volumetric productivity in dense culture fermentations with an associated 35–40% reduction in capital costs of fermentation equipment and an estimated savings of 10–15% on cost of product recovery (distillation), and 3–7% on overall production costs based on the projected use of inexpensive feedstocks.     

Hydrogen production by the thermophilic bacterium Thermotoga neapolitana

Virtually all members of the order Thermotogales have demonstrated the ability to produce hydrogen; however, some members of this order produce considerably greater quantities than others. With one representative of this order, Thermotoga neapolitana, we have consistently obtained accumulation of 25–30% hydrogen with 12–15% carbon dioxide as the only other prominent product in the batch reaction. In contradistinction to information widely disseminated in the literature, we have also found that most members of this order tolerate and appear to utilize the moderate amounts of oxygen present in the gaseous phase of batch reactors (6–12%), with no apparent decrease in hydrogen production. Hydrogen accumulation has been widely reported to inhibit growth of Thermotogales. While this may be true at very high hydrogen tensions, we have observed log phase bacterial morphology (rods) even in the presence of 25–35% hydrogen concentrations. To maximize hydrogen production and minimize production of hydrogen sulfide, inorganic sulfur donors are avoided and the cysteine concentration in the medium is increased. We and others have demonstrated that different members of the order Thermotogales utilize a wide variety of feedstocks, including complex carbohydrates and proteins. Thus, it appears that organisms within this order have the potential to utilize a variety of organic wastes and to cost-effectively generate hydrogen.     

Effect of starting xylose concentration on the microaerobic metabolism of Debaryomyces hansenii: the use of carbon material balances

Xylitol production by Debaryomyces hansenii NRRL Y-7426 was performed on synthetic medium varying the initial xylose concentration between 50 and 300 g/L. The experimental results of these tests were used to investigate the effect of substrate level on xylose consumption by this yeast. Satisfactory values of product yield on substrate (0.74–0.83 g/g) as well as volumetric productivity (0.481–0.694 g/L·h) were obtained over a wide range of xylose levels (90–200 g/L), while a worsening of kinetic parameters took place at higher concentration, likely due to a substrate inhibition phenomenon. The metabolic behavior of D. hansenii was studied, under these conditions, through a carbon material balance to estimate the fractions of xylose consumed by the cell for different activities (xylitol production, biomass growth, and respiration) during the lag, exponential, and stationary phases.     

Production of Raw Starch-Saccharifying Thermostable and Neutral Glucoamylase by the Thermophilic Mold <Emphasis Type="Italic">Thermomucor indicae-seudaticae</Emphasis> in Submerged Fermentation

Among physical and nutritional parameters optimized by “one variable at a time” approach, four cultural variables (sucrose, MgSO₄ ^.7H₂O, inoculum size, and incubation period) significantly affected glucoamylase production. These variables were, therefore, selected for optimization using response surface methodology. The p-values of the coefficients for linear effect of sucrose and inoculum size were less than 0.0001, suggesting them to be the key experimental variables in glucoamylase production. The enzyme production (34 U/ml) attained under optimized conditions (sucrose, 2%; MgSO₄ ^.7H₂O, 0.13%; yeast extract, 0.1%; inoculum size, 5 × 10⁶ spores per 50 ml production medium; incubation time, 48 h; temperature, 40°C; and pH 7.0) was comparable with the value predicted by polynomial model (34.2 U/ml). An over all 3.1-fold higher enzyme titers were attained due to response surface optimization. The experimental model was validated by carrying out glucoamylase production in shake flasks of increasing capacity (0.25–2.0 l) and 22-l laboratory bioreactors (stirred tank and airlift), where the enzyme production was sustainable. Furthermore, the fermentation time was reduced from 48 h in shake flasks to 32 h in bioreactors.