Controlling the feed rate of propanol to optimize erythromycin fermentation by on-line capacitance and oxygen uptake rate measurement

The aim of the present study was to optimize the feeding proportion of glucose and propanol for erythromycin biosynthesis by real-time monitoring and exploring its limited ratio by the on-line multi-frequency permittivity measurement. It was found that the capacitance values were sensitive to the variation of biomass concentration and microbial morphology as well as the true state of cell growth. It was most favorable to both cell growth and secondary metabolism to keep the ratio of glucose to propanol at 4.3 (g/g). The specific growth rate calculated by the capacitance measurement correctly and accurately reflected the cell physiological state. An appropriate feed rate of propanol was crucial for cell growth and secondary metabolism, as well as to improve the quality of erythromycin-A. In addition, the erythromycin production titer (10,950 U/mL) was further enhanced by 4 % when the propanol feed was regulated by step-down strategy based on both OUR (oxygen uptake rate) and the on-line monitoring capacitance.     

The membraneless bioelectrochemical reactor stimulates hydrogen fermentation by inhibiting methanogenic archaea

The membraneless bioelectrochemical reactor (Ml-BER) is useful for dark hydrogen fermentation. The effect of the electrochemical reaction on microorganisms in the Ml-BER was investigated using glucose as the substrate and compared with organisms in a membraneless non-bioelectrochemical reactor (Ml-NBER) and bioelectrochemical reactor (BER) with a proton exchange membrane. The potentials on the working electrode of the Ml-BER and BER with membrane were regulated to ?0.9 V (versus Ag/AgCl) to avoid water electrolysis with a carbon electrode. The Ml-BER showed suppressed methane production (19.8?±?9.1 mg-C·L^?1·day^?1) and increased hydrogen production (12.6?±?3.1 mg-H·L^?1·day^?1) at pH_out 6.2?±?0.1, and the major intermediate was butyrate (24.9?±?2.4 mM), suggesting efficient hydrogen fermentation. In contrast, the Ml-NBER showed high methane production (239.3?±?17.9 mg-C·L^?1·day^?1) and low hydrogen production (0.2?±?0.0 mg-H·L^?1·day^?1) at pH_out 6.3?±?0.1. In the cathodic chamber of the BER with membrane, methane production was high (276.3?±?20.4 mg-C·L^?1·day^?1) (pH_out, 7.2?±?0.1). In the anodic chamber of the BER with membrane (anode-BER), gas production was low because of high lactate production (43.6?±?1.7 mM) at pH_out 5.0?±?0.1. Methanogenic archaea were not detected in the Ml-BER and anode-BER. However, Methanosarcina sp. and Methanobacterium sp. were found in Ml-NBER. Prokaryotic copy numbers in the Ml-BER and Ml-NBER were similar, as were the bacterial community structures. Thus, the electrochemical reaction in the Ml-BER affected hydrogenotrophic and acetoclastic methanogens, but not the bacterial community.     

Effects of carbon sources on growth and extracellular polysaccharide production of Nostoc flagelliforme under heterotrophic high-cell-density fed-batch cultures

Dissociated cells separated from a natural colony of Nostoc flagelliforme were cultivated heterotrophically in the darkness on glucose under fed-batch culture conditions. The effects of carbon sources (glucose, fructose, xylose, and sucrose) and concentrations on cell growth and extracellular polysaccharide (EPS) production were investigated. At harvest, the culture contained 1.325 g L^?1 of biomass and 117.2 mg L^?1 of EPS, respectively. The gravimetric EPS production rate was 16.7 mg g^?1 cell dry weight day^?1, which was 2.1 times higher than previously reported. Using sigmoid model, batch fermentation of N. flagelliforme was kinetically simulated to obtain equations including substrate consumption, biomass growth, and EPS accumulation. Results from a simulation correlated well with the experimental ones, indicating that this method could be useful in studying EPS production from batch and fed-batch cultures.     

Integrated production of lactic acid and biomass on distillery stillage

The possibilities of parallel lactic acid and biomass production in batch and fed-batch fermentation on distillery stillage from bioethanol production were studied. The highest lactic acid yield and productivity of 92.3 % and 1.49 g L^?1 h^?1 were achieved in batch fermentation with initial sugar concentration of 55 g L^?1. A significant improvement of the process was achieved in fed-batch fermentation where the concentration of lactic acid was increased to 47.6 % and volumetric productivity for 21 % over the batch process. A high number of Lactobacillus rhamnosus ATCC 7469 viable cells of 10⁹ CFU ml^?1 was attained at the end of fed-batch fermentation. The survival of 92.9 % of L. rhamnosus cells after 3 h of incubation at pH 2.5 validated that the fermentation media remained after lactic acid removal could be used as a biomass-enriched animal feed thus making an additional value to the process.     

Kinetics of growth and ethanol formation from a mix of glucose/xylose substrate by Kluyveromyces marxianus UFV-3

The fermentation of both glucose and xylose is important to maximize ethanol yield from renewable biomass feedstocks. In this article, we analyze growth, sugar consumption, and ethanol formation by the yeast Kluyveromyces marxianus UFV-3 using various glucose and xylose concentrations and also under conditions of reduced respiratory activity. In almost all the conditions analyzed, glucose repressed xylose assimilation and xylose consumption began after glucose had been exhausted. A remarkable difference was observed when mixtures of 5 g L^?1 glucose/20 g L^?1 xylose and 20 g L^?1 glucose/20 g L^?1 xylose were used. In the former, the xylose consumption began immediately after the glucose depletion. Indeed, there was no striking diauxic phase, as observed in the latter condition, in which there was an interval of 30 h between glucose depletion and the beginning of xylose consumption. Ethanol production was always higher in a mixture of glucose and xylose than in glucose alone. The highest ethanol concentration (8.65 g L^?1) and cell mass concentration (4.42 g L^?1) were achieved after 8 and 74 h, respectively, in a mixture of 20 g L^?1 glucose/20 g L^?1 xylose. When inhibitors of respiration were added to the medium, glucose repression of xylose consumption was alleviated completely and K. marxianus was able to consume xylose and glucose simultaneously.     

Knocking out of tailoring genes eryK and eryG in an industrial erythromycin-producing strain of Saccharopolyspora erythraea leading to overproduction of erythromycin B, C and D at different conversion ratios

Aims: To overproduce erythromycin C, B or D and evaluate the effect of disruption of tailoring genes eryK and eryG in an industrial erythromycin producer. Methods and Results: The tailoring genes eryG and eryK were inactivated individually or simultaneously by targeted gene disruption in an industrial strain Saccharopolyspora erythraea HL3168 E3, resulting in the overproduction of erythromycin C (2·48 g l^?1), B (1·70 g l^?1) or D (2·15 g l^?1) in the mutant strain QL‐G, QL‐K or QL‐KG, respectively. Analysis of the erythromycin congeners throughout the fermentation indicated that, at the end of fermentation, comparatively large amount of erythromycin D (0·67 g l^?1) was accumulated in QL‐G, whereas only small amount of erythromycin D (0·10 g l^?1) was produced in QL‐K. Conclusions: Inactivation of tailoring genes eryG and eryK in the high producer did not affect the biosynthesis of erythromycin. However, erythromycin D could be more efficiently methylated by EryG than be hydroxylated by EryK. Significance and Impact of the Study: Development of the mutant strains provides a method for the economical large‐scale production of potent lead compounds. The information about the accumulation and conversion of erythromycins in the industrial strains may contribute to further improving erythromycin production.     

Screening and characterization of Isochrysis strains and optimization of culture conditions for docosahexaenoic acid production

Isochrysis is a genus of marine unicellular microalgae that produces docosahexaenoic acid (DHA, C22:6), a very long chain polyunsaturated fatty acid (PUFA) of significant health and nutritional value. Mass cultivation of Isochrysis for DHA production for human consumption has not been established due to disappointing low DHA productivity obtained from commonly used Isochrysis strains. In this study, 19 natural Isochrysis strains were screened for DHA yields and the results showed that the cellular DHA content ranged from 6.8 to 17.0 % of total fatty acids with the highest DHA content occurring in the exponential growth phase. Isochrysis galbana #153180 exhibited the greatest DHA production potential and was selected for further investigation. The effects of different light intensities, forms, and concentrations of nitrogen, phosphorus, and salinity on growth and DHA production of I. galbana #153180 were studied in a bubble column photobioreactor (PBR). Under favorable culture conditions, I. galbana #153180 contained DHA up to 17.5 % of total fatty acids or 1.7 % of cell dry weight. I. galbana #153180 was further tested in outdoor flat-plate PBRs varying in light path length, starting cell density (SCD), and culture mode (batch versus semicontinuous). When optimized, record high biomass and DHA productivity of I. galbana #153180 of 0.72 g L^?1?day^?1 and 13.6 mg?L^?1?day^?1, or 26.4 g?m^?2?day^?1 and 547.7 mg?m^?2?day^?1, respectively, were obtained, suggesting that I. galbana #153180 may be a desirable strain for commercial production of DHA.     

Metabolic threshold and sulfide-buffering in diffusion controlled marine sediments impacted by continuous organic enrichment

The effects of organic enrichment on sediment biogeochemistry was studied in diffusion controlled sediment mesocosms, where labile organic matter (OM) (fish feed) pulses were added once a week to the sediment surface. Two types of sediments, differing mainly in content of reactive Fe, were used. The aim of this experiment was two-fold, (1) to evaluate the importance of Fe-driven sulfide buffering for sulfide accumulation in surface enriched sediments, and (2) to estimate the diagenetic capacity for degradation of labile OM near the sediment surface. The simulated OM loading rate of 375 mmol C m^?2 day^?1 led to a 5–6 times increase in CO₂-production and a 4–5 times increase in O₂-uptake. Sulfate reduction estimated by radiotracer experiments and CO₂-release was 105–131 mmol m^?2 day^?1, but accumulation of porewater sulfide was low in both sediment types. Instead 99% of sulfide was oxidized with O₂ at the sediment water interface in the low Fe treatment, whereas 46% of produced sulfide precipitated as Fe-S compound in the high Fe treatment resulting in significantly lower O₂-uptake. Furthermore, the accumulation of up to 30% of added OM by the end of the experiment indicated a saturation of the heterotrophic microbial communities in the upper enriched surface layer. These results suggest a maximum diagenetic capacity for OM degradation in the range of ~25 μmol C cm^?3 day^?1 or 260 mmol m^?2 day^?1 for the present sediment types.     

Effect of Seawater–Sewage Cross-Transplants on Bacterial Metabolism and Diversity

Bioassays experiments were conducted to determine the metabolic and community composition response of bacteria to transplants between relatively pristine coastal seawater and sewage-impacted seawater. There were four treatments: (1) pristine seawater bacteria?+?pristine seawater (Pb?+?Pw), (2) sewage-impacted bacteria?+?sewage-impacted water (Sb?+?Sw), (3) pristine seawater bacteria?+?sewage-impacted water (Pb?+?Sw), and (4) sewage-impacted bacteria?+?pristine seawater (Sb?+?Pw). Sewage-derived DOC was more labile and readily utilized by bacteria, which favored the growth of high nucleic acid (HNA) bacteria, resulting in high bacterial production (BP, 113?±?4.92 to 130?±?15.8 μg C l^?1?day^?1) and low respiration rate (BR, <67?±?11.3 μg C l^?1?day^?1), as well as high bacterial growth efficiency (BGE, 0.68?±?0.09 to 0.71?±?0.05). In contrast, at the relatively pristine site, bacteria utilized natural marine-derived dissolved organic matter (DOM) at the expense of lowering their growth efficiency (BGE, <0.32?±?0.02) with low BP (<62?±?6.3 μg C l^?1?day^?1) and high BR 133?±?14.2 μg C l^?1?day^?1). Sewage DOM input appeared to alter the partitioning of carbon between respiration and production of bacteria, resulting in a shift toward higher BGE, which would not enhance oxygen consumption. Taxonomic classification based on 454 pyrosequencing reads of the 16S rRNA gene amplicons revealed that changes in bacterial community structure occurred when seawater bacteria were transferred to the eutrophic sewage-impacted water. Sewage DOM fueled the growth of Gammma-proteobacteria and Epsilson-proteobacteria and reduced the bacterial richness, but the changes in the community were not apparent when sewage-impacted bacteria were transferred to pristine seawater.     

Emerging biotechnologies for production of itaconic acid and its applications as a platform chemical

Recently, itaconic acid (IA), an unsaturated C5-dicarboxylic acid, has attracted much attention as a biobased building block chemical. It is produced industrially (>80 g L^?1) from glucose by fermentation with Aspergillus terreus. The titer is low compared with citric acid production (>200 g L^?1). This review summarizes the latest progress on enhancing the yield and productivity of IA production. IA biosynthesis involves the decarboxylation of the TCA cycle intermediate cis-aconitate through the action of cis-aconitate decarboxylase (CAD) enzyme encoded by the CadA gene in A. terreus. A number of recombinant microorganisms have been developed in an effort to overproduce it. IA is used as a monomer for production of superabsorbent polymer, resins, plastics, paints, and synthetic fibers. Its applications as a platform chemical are highlighted. It has a strong potential to replace petroleum-based methylacrylic acid in industry which will create a huge market for IA.     

Upper limits of photosynthetic productivity and problems of scaling

Some 1,370 W m^?2 of light energy reaches the outer atmosphere of earth and on average only 240 W m^?2 reaches the earth’s surface. Only a fraction of this is used to fix CO₂ through photosynthesis, and efficiencies ranging from 0.1?8% for total irradiance have been reported. The theoretical maximum quantum efficiency of carbon fixation is 0.125 mol C (mol quanta)^?1 which relates to a maximum productivity of about 12 g C m^?2 day^?1 or 29.8 g(dw) m^?2 day^?1. This could increase to a maximum of 200 g(dw) m^?2 day^?1 in intermittent light of high frequencies, which is on average eight times higher than the average measured under field conditions where rates approaching 25 g(dw) m^?2 day^?1 are considered high. Several possibilities exist for achieving higher yields and photosynthetic efficiencies, such as limiting the antennae sizes and pulsing light at frequencies equivalent to electron turnover in the electron transport chains of photosynthesis. Scaling from laboratory experimental conditions to large commercial photobioreactors is a major stumbling block and may be the single most important factor responsible for the overall low reported areal production rates.     

Start-up and operation strategies on the liquefied food waste anaerobic digestion and a full-scale case application

Batch anaerobic digestion was employed to investigate the efficient start-up strategies for the liquefied food waste, and sequencing batch digestion was also performed to determine maximum influent organic loading rate (OLR) for efficient and stable operation. The results indicated that the start-up could be well improved using appropriate wastewater organic load and food-to-microorganism ratios (F/M). When digestion was initialized at low chemical oxygen demand (COD) concentration of 20.0 gCOD L^?1, the start-up would go well using lower F/M ratio of 0.5–0.7. The OLR 7.0 gCOD L^?1 day^?1 was recommended for operating the ASBR digestion, in which the COD conversion of 96.7 ± 0.53 % and biomethane yield of 3.5 ± 0.2 L gCOD^?1 were achieved, respectively. The instability would occur when OLR was higher than 7.0 gCOD L^?1 day^?1, and this instability was not recoverable. Lipid was suggested to be removed before anaerobic digestion. The anaerobic digestion process in engineering project ran well, and good performance was achieved when the start-up and operational strategies from laboratory study were applied. For case application, stable digestion performance was achieved in a digester (850 m³ volume) with biogas production of 1.0–3.8 m³ m^?3 day^?1.     

Effect of controlled oxygen limitation on Candida shehatae physiology for ethanol production from xylose and glucose

Carbon distribution and kinetics of Candida shehatae were studied in fed-batch fermentation with xylose or glucose (separately) as the carbon source in mineral medium. The fermentations were carried out in two phases, an aerobic phase dedicated to growth followed by an oxygen limitation phase dedicated to ethanol production. Oxygen limitation was quantified with an average specific oxygen uptake rate (OUR) varying between 0.30 and 2.48 mmolO₂ g dry cell weight (DCW)^?1 h^?1, the maximum value before the aerobic shift. The relations among respiration, growth, ethanol production and polyol production were investigated. It appeared that ethanol was produced to provide energy, and polyols (arabitol, ribitol, glycerol and xylitol) were produced to reoxidize NADH from assimilatory reactions and from the co-factor imbalance of the two-first enzymatic steps of xylose uptake. Hence, to manage carbon flux to ethanol production, oxygen limitation was a major controlled parameter; an oxygen limitation corresponding to an average specific OUR of 1.19 mmolO₂ g DCW^?1 h^?1 allowed maximization of the ethanol yield over xylose (0.327 g g^?1), the average productivity (2.2 g l^?1 h^?1) and the ethanol final titer (48.81 g l^?1). For glucose fermentation, the ethanol yield over glucose was the highest (0.411 g g^?1) when the specific OUR was low, corresponding to an average specific OUR of 0.30 mmolO₂ g DCW^?1 h^?1, whereas the average ethanol productivity and ethanol final titer reached the maximum values of 1.81 g l^?1 h^?1 and 54.19 g l^?1 when the specific OUR was the highest.     

Nisin production in a chitin-including continuous fermentation system with Lactococcus lactis displaying a cell wall chitin-binding domain

The limiting factors in the continuous production of nisin are high amount of biomass loss and low dilution rate application. In this study, a chitin-including continuous nisin fermentation system (CICON-FER) was constructed for high volumetric nisin production using nisin producer L. lactis displaying cell wall chitin-binding domain (ChBD) together with chitin in the reactor. In this respect, the highest binding conditions of relevant L. lactis cells to chitin were determined. Then the chitin flakes carrying nisin-producing L. lactis cells were used within the CICON-FER system at different dilution rates (0.1–0.9 h^?1) and initial glucose concentrations (20–60 g l^?1). The results revealed that the pH 7 conditions and the use of 100 mM sodium phosphate buffer with 0.1 % Tween 20 and Triton X-100 significantly increased the binding capacity of ChBD displaying L. lactis cells to chitin. The constructed CICON-FER system maintained the presence of the ChBD surface displaying L. lactis cells in the reactor system until 0.9 h^?1 dilution rate that resulted in a considerably high level of volumetric nisin production and productivity (10,500 IU ml^?1 and 9,450 IU ml^?1 h^?1, respectively) with the combination of a 0.9-h^?1 dilution rate and a 40-g l^?1 initial glucose concentration. In conclusion, an innovative nisin fermentation system that yielded the highest nisin production thus far and that was feasible for industrial application was created.     

Microbial production of 2,3-butanediol by a newly-isolated strain of Serratia marcescens

A newly-isolated strain of Serratia marcescens, G12, was characterized for 2,3-butanediol (2,3-BD) production. In shake-flask and batch fermentations, 2,3-BD reached 48.5 and 51 g l^?1, respectively. Low amounts of (~8 g l^?1) of acetoin were also formed. In fed-batch fermentations, strain G12 produced 72.8 g 2,3-BD l^?1 with glucose initially at 130 g l^?1. When aeration rate was increased to 2.5 vvm for the fermentation process, 2,3-BD reached 87.8 g l^?1 and the highest productivity was 1.6 g l^?1 h^?1. Acetoin was at 6.2 g l^?1. G12 therefore may be a suitable candidate strain for large-scale production of 2,3-BD.     

Enhanced riboflavin production by recombinant Bacillus subtilis RF1 through the optimization of agitation speed

Dissolved oxygen is one of the most important bioprocess parameters that could affect cell growth and product formation, and it is easy to control by changing agitation speed. In this work, the effects of agitation speed on the performance of riboflavin production by recombinant Bacillus subtilis RF1 was investigated in fed-batch fermentation. The lower agitation speed (600 rpm) was beneficial for cell growth and riboflavin biosynthesis in the initial phase of fermentation process. While, during the later phase, higher agitation speed (900 rpm) was favor for cell growth and riboflavin biosynthesis. Thus, a two-stage agitation speed control strategy was proposed based on kinetic analysis, in which the agitation speed was controlled at 600 rpm in the first 26 h and then switched to 900 rpm to maintain high μ for cell growth and high q _p for riboflavin production during the entire fermentation process. However, it was observed that a sharp increase of agitation speed resulted in an adverse effect on cell growth and riboflavin synthesis within a short time. To avoid this phenomenon, a multi-stage agitation speed control strategy was set up based on the two-stage control strategy, the maximum concentration of riboflavin reached 9.4 g l^?1 in 48 h with the yield of 0.051 g g^?1 by applying this strategy, which were 20.5 and 21.4 % over the best results controlled by constant agitation speeds.     

The effect of light availability on the photosynthetic activity and productivity of outdoor cultures of Arthrospira platensis (Spirulina)

The basic requirement for establishing economically viable large-scale production of algal biomass, be it for food, feed, high-value product, or energy, is the ability to produce the biomass at a low price. To achieve this goal, an efficient production protocol is needed that ensures that the potential productivity is obtained at any given time. When productivity is defined by the ability to utilize the available solar radiation that drives photosynthesis, the production protocol must be optimized to meet this requirement. In the current study, we demonstrate that by modifying the light available to Arthrospira platensis cells cultured outdoors by a variety of options like modifying the standing biomass concentration, changing the mixing rate, or shading can change the potential photosynthetic activity and apparent activity. By optimizing the light available to algae cells under outdoor conditions, productivity can be increased by approximately 50 %, from 15.6 g m^?2 day^?1 in a culture that suffers from overexposure to light to 22.4 g m^?2 day^?1 in a culture in which light downregulation is minimized. Therefore, by using a variety of methodologies to estimate photosynthetic activity, we demonstrate that overexposing the cells to light may result in downregulation of the photosynthetic activity leading to photoinhibition and lower biomass productivity.     

Two-stage fermentation process for enhanced mannitol production using Candida magnoliae mutant R9

Mutants of Candida magnoliae NCIM 3470 were generated by treatment of ultra-violet radiations, ethyl methyl sulphonate and N-methyl-N′-nitro-N-nitrosoguanidine. Mutants with higher reductase activity were screened by means of 2,3,5-triphenyl tetrazolium chloride agar plate assay. Among the screened mutants, the mutant R9 produced maximum mannitol (i.e. 46 g l^?1) in liquid fermentation medium containing 250 g l^?1 glucose and hence was selected for further experiments. Preliminary optimization studies were carried out on shake-flask level which increased the mannitol production to 60 g l^?1 in liquid fermentation medium containing 300 g l^?1 glucose. A two-stage fermentation process comprising of growth phase and production phase was employed. During the growth phase, glucose was supplemented and aerobic conditions were maintained. Thereafter, the production phase was initiated by supplementing fructose and switching to anaerobic conditions by discontinuing aeration and decreasing the speed of agitation. The strategy of two-stage fermentation significantly enhanced the production of mannitol up to 240 g l^?1, which is the highest among all fermentative production processes and corresponds to 81 % yield and 4 g l^?1 h^?1 productivity without formation of any by-product.     

Batch and semi-continuous microalgal TAG production in lab-scale and outdoor photobioreactors

Microalgal triglycerides (TAGs) represent a sustainable feedstock for food, chemical and biofuel industries. The operational strategy (batch, semi-continuous, continuous cultivations) has an impact on the TAG productivity. In this study, semi-continuous (i.e. with fixed harvesting frequency) and batch cultivations were compared on TAG production both at lab-scale and in outdoor cultivations. At lab-scale, the semi-continuous TAG productivity was highest for a cycle time of 2 days (SC1; 0.21 g L^?1 day^?1) and similar to the maximum obtained with the batch (optimal harvest time; 0.23 g L^?1 day^?1). Although TAG content was lower for SC1 (22 %) than for the batch (35 %), higher biomass productivities were obtained with SC1. Outdoors, semi-continuous cultivations were subjected to a lower degree of stress (i.e. higher amount of nitrogen present in the system relative to the given irradiance) compared to lab-scale. This yielded low and similar TAG contents (10–13 %) in the different semi-continuous runs that were outdone by the batch on both TAG content (15–25 %) and productivity (batch, 0.97–2.46 g m^?2 day^?1; semi-continuous, 0.35–0.85 g m^?2 day^?1). The lab-scale experiments showed that semi-continuous strategies, besides leading to similar TAG productivities compared to the batch, could make TAG production cost effective by valorising also non-TAG compounds. However, optimization of outdoor semi-continuous cultivations is still required. For instance, the nitrogen supply and the harvest frequency should be adjusted on the total irradiance. Additionally, future research should focus on recovery metabolism upon nitrogen resupply.     

Effects of pulse feeding of beet molasses on recombinant benzaldehyde lyase production by Escherichia coli BL21(DE3)

The effect of fed-batch operation (FBO) strategy was investigated using pretreated-beet molasses, containing galactose that induces the lac promoter, on benzaldehyde lyase (BAL) production by recombinant Escherichia coli BL21(DE3)pLySs. After batch cultivation with 30 g l^?1 pretreated-beet molasses consisting of 7.5 g l^?1 glucose and 7.5 g l^?1 fructose, three FBO strategies were applied at dissolved oxygen (=40%) cascade to air-flow rate. In FBO1 when air-flow rate decreased considerably, feed was given to the system in pulses in such a way that pretreated-beet molasses concentration increased by 10 kg m^?3 (containing 2.5 g l^?1 glucose+2.5 g l^?1 fructose); however, decrease in air-flow rate demonstrated only the absence of glucose but not fructose. Thus, in FBO2 when fructose and glucose were completely utilized, pretreated-beet molasses was pulse-fed and its concentration increased by 10 g l^?1. In FBO3 with the decreased amount of pretreated-beet molasses (6 g l^?1), shift response time from glucose to fructose consumption was avoided, and glucose and fructose consumptions were well correlated with air-flow rate, and the highest C _X (8.04 g l^?1) and BAL (2,315 U ml^?1) production were obtained (t?=?24 h) with the highest substrate yield on cell and product formation.