Skip to main page content

• HOME
• Current Issue
• Archive
• Alerts
• About ASM
• Contact us
• Tech Support
• Journals.ASM.org

This Article Abstract Full Text (PDF) Supplemental material Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Bideaux, C. Articles by Guillouet, S. E. Search for Related Content PubMed PubMed Citation Articles by Bideaux, C. Articles by Guillouet, S. E. Agricola Articles by Bideaux, C. Articles by Guillouet, S. E.

 Previous Article  |  Next Article 

Applied and Environmental Microbiology, March 2006, p. 2134-2140, Vol. 72, No. 3
0099-2240/06/$08.00+0     doi:10.1128/AEM.72.3.2134-2140.2006
Copyright © 2006, American Society for Microbiology. All Rights Reserved.

Minimization of Glycerol Production during the High-Performance Fed-Batch Ethanolic Fermentation Process in Saccharomyces cerevisiae, Using a Metabolic Model as a Prediction Tool

Carine Bideaux, Sandrine Alfenore, Xavier Cameleyre, Carole Molina-Jouve, Jean-Louis Uribelarrea, and Stéphane E. Guillouet^*

Biotechnology and Bioprocess Laboratory, UMR-CNRS 5504, UMR-INRA 792, Département de Génie Biochimique et Alimentaire, Institut National des Sciences Appliquées, 135 Avenue de Rangueil, 31077 Toulouse Cedex, France

Received 21 July 2005/ Accepted 29 November 2005

Top Abstract Introduction Materials and Methods Results Discussion References

 
On the basis of knowledge of the biological role of glycerol in the redox balance of Saccharomyces cerevisiae, a fermentation strategy was defined to reduce the surplus formation of NADH, responsible for glycerol synthesis. A metabolic model was used to predict the operating conditions that would reduce glycerol production during ethanol fermentation. Experimental validation of the simulation results was done by monitoring the inlet substrate feeding during fed-batch S. cerevisiae cultivation in order to maintain the respiratory quotient (RQ) (defined as the CO₂ production to O₂ consumption ratio) value between 4 and 5. Compared to previous fermentations without glucose monitoring, the final glycerol concentration was successfully decreased. Although RQ-controlled fermentation led to a lower maximum specific ethanol production rate, it was possible to reach a high level of ethanol production: 85 g · liter^–1 with 1.7 g · liter^–1 glycerol in 30 h. We showed here that by using a metabolic model as a tool in prediction, it was possible to reduce glycerol production in a very high-performance ethanolic fermentation process.

Top Abstract Introduction Materials and Methods Results Discussion References

 
It is well-known that glycerol is a major by-product during alcoholic fermentation in the yeast Saccharomyces cerevisiae. The role of glycerol in the cellular redox balance (5) and in osmoregulation (4, 16) have been reported in the literature. Participating in the cell redox balance, glycerol is produced by Saccharomyces cerevisiae to reoxidize surplus NADH, formed in the synthesis of biomass and secondary fermentation products, to NAD⁺ (5). Two different strategies for modulating the production of glycerol have been reported in the literature. One strategy is to change the operating conditions of the fermentation process (i.e., aeration levels, stress conditions [salts, acids, osmotic stress, heat shock ...]) (3, 4, 8, 11, 12, 19, 31, 40, 42); another is to use genetically modified strains. Engineered strains were constructed by overexpressing and repressing components of the glycerol synthesis pathway (14, 18, 19, 21, 25).

Complementing biochemical and metabolic engineering experimental approaches, mathematical models may also be useful for interpretation and prediction of biochemical processes. Process models consist in coupling mass balance equations at the reactor scale and at the cellular scale (29). Cells can be modeled at different levels: unstructured models on the population level (15, 27) or structured models on the cellular level. The latter can be a compartmental model (41) or a metabolic regulator model (28). Constraints on carbon, energy, and redox potential linked to the metabolic network can be taken into account by a metabolic model that contains a complete set of metabolic pathways involved in biomass synthesis, energy production, substrate degradation, and cometabolite production (30, 35).

Metabolic flux analysis is generally used to quantify intracellular fluxes in the central metabolism of microorganisms (7, 9, 11, 22, 24, 43), but it has had a limited impact due to the lack of regulatory information in model formulation (10). Kinetic information can be introduced by the description of enzymatic rates with mathematical equations whose parameters are estimated with experimental data (26, 38). Another way of accounting for the regulatory aspects of cell metabolism consists in applying cybernetic principles (37) by using neural networks (34) or fuzzy logic-based models (13) when mechanistic details are missing.

Metabolic models were used in a few studies to predict (36) and to optimize (32) growth and by-product secretion. Metabolic flux analysis, combined with a black box kinetic model, then permits the introduction of dynamic information keeping the constraints imposed by metabolism.

With the aim of increasing the production of ethanol for use as a biofuel, glycerol production has to be minimized in order to better monitor the fate of carbon during the fermentation process and to limit technological problems in distillation units and/or separation processes after the fermentation stage. Reducing by-product yields is indeed a general focus in developing a competitive bioprocess, since the price of raw materials is a very important parameter in determining the overall economy of the process.

In this paper, the adopted strategy of reducing the production of glycerol during ethanolic fermentation was based on the combination of "process" (simulation, prediction, and process control) and metabolic approaches. On the basis of knowledge of the biological role of glycerol in redox balances, a strategy of fermentation monitoring was defined to reduce the need of surplus NADH oxidation responsible for glycerol synthesis during ethanol fermentation. A metabolic model was used to predict the operating conditions that would reduce glycerol production during ethanol fermentation in fed-batch process. Experimental validation of the simulation results was done by monitoring the inlet substrate feeding in order to maintain the respiratory quotient (RQ) (defined as the CO₂ production to O₂ consumption ratio) value between 4 and 5. Fermentation data were then discussed and compared in terms of growth, cell viability, and the production of ethanol and glycerol with previous highly productive ethanolic fermentations.

Top Abstract Introduction Materials and Methods Results Discussion References

 
Microorganism, media, and growth conditions.
The Saccharomyces cerevisiae CBS8066 strain was from the Centraalbureau voor Schimmelcultures (Baarn, The Netherlands). The culture conditions and medium composition used have been described previously (2).

Fermentations.
Fed-batch experiments were performed in a 20-liter fermentor using the Braun Biostat E fermenting system without oxygen limitation. The temperature was maintained at 30°C, and the pH was maintained at 4 with the addition of a 14% (vol/vol) NH₃ solution. All fermentations were performed with an exponential feeding of vitamins on the basis of the growth profile (2). The fermentor was connected to a computer. A homemade software program enabled the online acquisition of the controlled parameters (stirring rate, pH, temperature, partial pressure of the dissolved oxygen, NH₃ and antifoam additions) and the monitoring and regulation of these parameters. The pressure in the bioreactor was regulated at 0.2 bar (relative pressure). The fermentor was flushed continuously with air through the sparger placed at the bottom of the reactor. The airflow rate was 313 normal liters h^–1. The stirring rate was fixed at 400 rpm until the concentration of dissolved oxygen reached 20% of saturation, and then it was increased in order to avoid any oxygen limitation in the culture.

The bioreactor was also equipped with a detection system of foam and controlled addition of Struktol (Schill & Seilacher, Germany) antifoam. The maximum amount of added Struktol was 1.2 ml. The fermentor was fed with a sterile concentrated glucose solution (about 700 g liter^–1) using a peristaltic pump (Masterflex). The glucose concentration in the reactor was automatically calculated by the software program on the basis of all the inlet and outlet volumes (base addition, sampling, and antifoam addition), the added mass (estimated instantly by weighing), and density, and the glucose concentration in the feeding solution was measured by high-performance liquid chromatography. Glucose density was measured by weighing a known volume.

Both the instantaneous CO₂ production and the O₂ consumption rates, and thus, the respiratory quotient, were calculated by gas analysis. As a consequence of the model prediction, the RQ value had to be regulated between 4 and 5 by controlling glucose flux. Thus, a PID (proportional-integral-derivative) controller was built to command the glucose pump in order to maintain the RQ value.

Gas analysis.
Outlet gas analysis was performed by mass spectroscopy after the condenser. The mass spectrometer (PRIMA 600s; VG Gas, Manchester, United Kingdom) was used for its accuracy to measure CO₂, O₂, N₂, and Ar. The gas flow was ionized by electronic bombardment, and the gases were then separated according to their mass/charge ratio (m/e) by the application of a magnetic field and then collected by a detector. Gas analysis was conducted on the outlet flow of the reactor every 20 seconds, and gas analysis was performed on the inlet air every hour by commutation of the valves. The O₂ consumption rate and the CO₂ production rate were calculated from the differences between the inlet and outlet gas compositions, taking into account the evolution of the liquid volume in the reactor, inlet airflow (measured by a mass flowmeter), temperature, and pressure.

Analytical methods.
Biomass and concentrations of substrate and product were measured by the previously described analytical methods (2).

Determination of cell viability.
The viability of cells was determined by the methylene blue technique (23). A 200-µl sterile solution of methylene blue (0.3 mM in 68 mM sodium citrate) was mixed with 200 µl of a yeast suspension diluted to reach an optical density at 620 nm of 0.4 to 0.7. The mixture was shaken and placed in a Thoma counting chamber after 5 min of incubation. The numbers of stained (nonactive) or unstained (active) cells and the number of budding cells were counted in five different fields with a total of at least 200 to 300 cells. The percentage of viable cells was the number of unstained cells (lively cells) divided by the total number of cells (stained and unstained). Under our conditions, the mean viability (v) is estimated with an accuracy of 10% (17), which means that the interval v ± 0.1v contains the true value of viability with a probability of 95%.

Chemicals.
Chemical products (glucose, salts, oligo-elements, orthophosphoric acid, and NH₃) were from Prolabo, the vitamins were from Sigma, and the sodium glutamate was from Merck. All products were of the highest analytical grade available.

Top Abstract Introduction Materials and Methods Results Discussion References

 
One of the known major physiological roles of glycerol involves the cell redox balance. Under anaerobic conditions, glycerol is produced by Saccharomyces cerevisiae to reoxidize the surplus NADH formed in the synthesis of biomass and secondary fermentation products, mainly organic acids. Under aerobic conditions, glycerol synthesis and respiration participate concomitantly to the redox balance. On the basis of this knowledge, a strategy was formulated to reduce the surplus formation of NADH by controlling the specific rate for glucose consumption (q_glucose). If the glucose feeding is controlled so that it fulfills the requirements for anabolism and ethanol production, this strategy should lead to a significant decrease in glycerol formation. In order to dictate the experimental conditions to carry out, a metabolic model named metabolic descriptor was used.

Simulation by the metabolic descriptor model.
The metabolic descriptor model was composed in two equation systems: a metabolic component describing the cell metabolism through stoichiometric equations and classical mass balance equations describing the interactions between the cells and the bioreactor environment.

A metabolic matrix was constructed on the basis of the mass balance and on the pseudo-steady-state hypothesis on the intracellular intermediate metabolites. The formulation resulted in a set of linear equations expressed as a stoichiometric matrix S (m x n) with vectors for reaction rates or fluxes r (m x 1) and vectors for consumption or secretion rates of metabolites R (1 x n). Thus, S x r = R. This linear system was undetermined, and calculation of a unique solution was possible by considering a set of rates whose number was equal to the degrees of freedom of the system. For the major reactions of catabolism, anabolism, and energetics that we considered, see the supplemental material. The metabolic matrix construction resulted in a system containing 110 metabolites (n = 110) and 99 intracellular rates (m = 99) and a degree of freedom equal to 6.

Starting from stoichiometric equations of the metabolic pathway of Saccharomyces cerevisiae, the metabolic descriptor model allowed us to obtain expression of intracellular fluxes and extracellular specific rates as a function of six specific rates, such as the specific growth rate (µ), specific rate for glucose consumption (q_glucose), specific rate for ethanol production (q_ethanol), specific rate for acetate production (q_acetate), specific rate for glycerol production (q_glycerol), and specific rate for succinate production (q_succinate). In order to verify the possibility of producing ethanol during the fermentation process without any other by-products (particularly glycerol), the sign of metabolic fluxes and extracellular specific rates was first studied when the specific rates of production of glycerol, acetate, and succinate were null. A set of inequality functions of µ, q_glucose, and q_ethanol was obtained. The resolution of these inequalities allowed us to determine a variation domain for q_glucose [q_glucose-min; q_glucose-max], each being a function of µ and q_ethanol in such a way that q_glycerol, q_acetate, and q_succinate were null: q_glucose-min(µ, q_ethanol q_glucose q_glucose-max (µ, q_ethanol).

The ordinary differential equations (ODE) of the bioreactor (as a function of the specific rates) were solved by the Runge-Kutta method using Matlab software. The specific rates were determined each time ODE were computed with the metabolic descriptor model by constrained optimization. Aiming to minimize by-product production and concomitantly to maximize the ethanol production, the following objective function had to be minimized:

(1)

There were four constraints. (i) All intracellular rates, the opposite of the specific rate of oxygen consumption, and the specific rate of carbon dioxide production had to be positive. (ii) The specific growth rate had to be equal to the maximum value of the specific growth rate reachable at the present ethanol concentration (P) in the medium. This maximum specific growth rate was given by an equation taking into account ethanol inhibition with parameters determined from previous experimental data (2):

(2)

with the maximum specific growth rate µ_max being 0.42 h^–1 and the critical ethanol concentration P_c (when growth stops) being 100 g liter^–1

(3)

where Y_{(X, ATP)} is the global ATP/biomass yield expressed in grams of dried biomass per mole of ATP. (iii) The specific rate of glucose consumption had to be in the interval [q_glucose-min q_glucose-max] determined previously. The substrate feeding rate was calculated in order to be equal to the substrate need (q_glucose). (iv) The specific rate of ethanol production had to be lower than or equal to the maximum value of the ethanol production specific rate reachable at the present specific growth rate. This maximum specific rate was given by the equation

(4)

The parameters (coupling constant for the ethanol production specific rate) and ß (uncoupling constant for the ethanol production specific rate) were determined from experimental data. ( = 0.0612 and ß = 0.0063) (2).

The simulation results of the metabolic descriptor model showed that ethanol production without any by-product synthesis was stoichiometrically possible, provided that a specific glucose feeding profile was applied during the fed-batch culture. This simulated profile for glucose feeding led us to monitor the glucose-limited condition in the medium during the culture. As a result, calculation of the respiratory quotient showed that its value had to be lower than 6.2.

The estimated biomass and ethanol concentrations after 30 h of culture were 66 g · liter^–1 and 100 g · liter^–1, respectively.

Experimental validation. (i) Fermentation parameters.
In order to validate the results of the simulation carried out by the metabolic descriptor model, two fed-batch fermentations were performed in a 20-liter fermentor in a mineral medium. Fermentation conditions were identical to those used for very high-performance ethanolic fermentation (2) except for the glucose feeding. As a result of the simulation, the RQ value was regulated within the range 4 to 5 by monitoring the glucose feeding rate using a PID controller.

The mass of ethanol and glycerol and biomass produced during the fed-batch fermentation process are shown in Fig. 1. The RQ value was kept constant at 4.5. As a result, the residual glucose concentration remained limiting (as required by the predictive model) during the first 17 hours and then increased slightly up to 5 g · liter^–1. Concomitant growth and production of glycerol and ethanol was observed during the 30 h of fermentation. Final concentrations of 18 g liter^–1 biomass, 85 g liter^–1 ethanol, and 1.7 g liter^–1 glycerol were reached. The fermentation led to global yields on glucose of 0.071 g g^–1, 0.34 g g^–1, and 0.006 g g^–1, for biomass, ethanol, and glycerol, respectively (Table 1).

View larger version (23K): [in this window] [in a new window]

FIG. 1. Evolution of the biomass and ethanol, glycerol, and residual glucose concentrations (A) and respiratory quotient (B) in the RQ-controlled aerobic batch-fed fermentation.

View this table: [in this window] [in a new window]

TABLE 1. Comparison of ethanolic fermentation under full aeration, microaeration, and RQ-controlled conditionsa

The RQ-controlled fermentation was duplicated and gave similar results. Carbon and redox balances were calculated for each fermentation. All fermentations showed a small carbon balance deficit of 6% and a 10% redox balance deficit.

(ii) Fermentation kinetic parameters.
The specific growth rate, the specific ethanol production rate, and the specific glycerol production rate were drawn versus the ethanol concentration along the fermentation (Fig. 2). µ and q_ethanol were coupled and decreased gradually, likely due to inhibition by the ethanol accumulated during the process. Maximum values were obtained at the beginning of the fermentation and reached 0.42 h^–1 for growth and 0.62 g g^–1 h^–1 for ethanol production. The specific glycerol production rate showed a different pattern, as its specific production rate remained constant within the range of 20 to 70 g liter^–1 ethanol.

View larger version (17K): [in this window] [in a new window]

FIG. 2. Specific growth rate, specific ethanol production rate (p), and glycerol production rates versus ethanol (EtOH) concentration for the RQ-controlled batch-fed fermentation.

(iii) Cell viability.
Cell viability was also quantified during the fermentations. As shown in Fig. 3, a drastic drop in cell viability occurred at the end of the fermentation when the ethanol concentration reached 80 g liter^–1.

View larger version (15K): [in this window] [in a new window]

FIG. 3. Comparison of the percentage of methylene blue-stained cells (% MB staining, a measure of viability) as a function of ethanol concentration between RQ-controlled fermentation (open symbols) and high-performance ethanol (EtOH) fermentation (black symbols).

Top Abstract Introduction Materials and Methods Results Discussion References

 
The aim of this work was to minimize glycerol production during the ethanol production fermentation process. We used a dual mass balance and stoichiometric model based on the metabolic network of Saccharomyces cerevisiae to predict the operating conditions for reaching the objectives described above. The results of the simulation dictated to control the RQ value within the range of 4 to 5 and to limit the accumulation of residual glucose in the medium. The global ATP/biomass yield was fixed within the range of 10 to 11 according to the literature (3, 11). Data on the inhibitory effect of ethanol concentration on growth and ethanol production were included in the model. These parameters were selected from previous fed-batch fermentations leading to very high-performance ethanolic fermentation (2), i.e., an ethanol concentration for growth arrest of 100 g liter^–1 and an ethanol concentration for ethanol production arrest of 150 g liter^–1. The results of the simulation also gave the following data: final glycerol concentration equal to zero with a concomitant biomass production of 66 g liter^–1 and ethanol concentration of 100 g liter^–1 in 30 h of fermentation.

The validation was experimentally tested in glucose-limited fed-batch cultivation monitored by glucose feeding in order to control the RQ value at 4.5. The fermentation led to a final glycerol concentration of 1.7 g liter^–1, biomass concentration of 18 g liter^–1, and ethanol concentration of 85 g liter^–1. These data were compared with a fully aerated fed-batch fermentation with a sequenced glucose feeding (3) (Table 1). This reference high-performance ethanolic fermentation allowed us to reach an ethanol concentration of 132 g liter^–1 in 30 h under nonlimiting aeration conditions.

Here, the experimental results showed that the inhibition of the specific growth rate by ethanol was obtained for a lower concentration than in the reference experiment used for the prediction (85 g liter^–1 versus 100 g liter^–1). It was also observed that the oxygen consumption specific rate and oxygen to biomass yield (Table 1) were greater in this case than in the reference fermentation, leading to a lower expected value of the yield of ATP (Y_ATP).

A study of the sensitivity of all the symbolic variables and parameters of the prediction model (P_c, Y_ATP, phosphorus versus oxygen atom ratio, biomass macromolecular composition, ...) showed that only ATP yield and P_c induced a significant change in the biomass production (Fig. 4). By integrating into the model the ATP yield calculated from experimental data and a P_c value of 85 g · liter^–1, the predicted final biomass concentration can be decreased from 66 g liter^–1 to 20 g liter^–1 concomitantly with a final ethanol concentration of 85 g liter^–1 in 30 h of fermentation.

View larger version (39K): [in this window] [in a new window]

FIG. 4. Sensitivity analysis of variables and parameters of the prediction model (µmax, maximum specific growth rate; Pc, critical ethanol concentration; , coupling constant for the ethanol production specific rate; ß, uncoupling constant for the ethanol production specific rate; Y(X;ATP), global ATP/biomass yield; P/O, phosphorus versus oxygen atom ratio).

Besides differences between model prediction and experimental validation, the use of the Metabolic Descriptor model allowed us to effectively produce a high ethanol concentration (85 g liter^–1) with a concomitant low glycerol concentration (1.7 g liter^–1) in a short period of time (30 h). This performance can be compared to those of previous fermentations carried out in the same conditions of strain, medium, and sequenced glucose feeding but with different aeration strategies (Table 1). Using the Metabolic Descriptor model, the final glycerol concentration was successfully decreased by factors of 7 and 2 compared to fermentations carried out under microaeration and full aeration conditions, respectively (3). It was confirmed by a decrease in the apparent yield of glycerol production on glucose (0.006 versus 0.042 and 0.010 g g^–1, respectively).

Maintaining cell viability is one of the most important parameters in order to reach high ethanol concentrations in ethanolic fermentation. Comparing the results between our high-performance fed-batch process (2) and this RQ-controlled fed-batch process, a drop in cell viability was observed when glycerol production was minimized. This observation is unlikely to be explained by some potential energetic or redox imbalance resulting from the reduction in glycerol synthesis. The lower glycerol formation could have led to a perturbation in the redox balance, since anabolic and catabolic NADH production increased due to the higher biomass yield and lower glycerol yield. However, this deficit of NADH oxidation was likely compensated for by the observed increase in the oxygen consumption (Table 1). Such an increase in oxygen consumption is in agreement with reported data where overexpressing the transcription factor HAP4 gene in Saccharomyces cerevisiae led to a 3.7-fold decrease in q_glycerol concomitantly with a twofold increase in q_O2, a 25% reduction of q_ethanol, and a 2.2-fold increase in q_acetate during aerobic batch culture compared to the values for the wild-type strain (6).

From an energetic point of view, the reduction of glycerol production in the RQ-controlled fermentation could also have resulted in an imbalance for ATP, since ATP production decreased due to a higher biomass yield and a lower ethanol yield. As previously, the increasing oxygen consumption may have balanced the cell energetics. Our data could then suggest a protective role of glycerol against ethanolic stress during ethanolic fermentation. It is well-known that glycerol formation in yeast is a response to the osmotic pressure of the medium (4, 16). Increasing glycerol production was also reported after submitting yeast cells to heat shock (12, 20). A growth-coupled glycerol formation was also reported to be enhanced during aerobic ethanolic fermentation with increasing temperatures within the range of 27 to 33°C, whereas a decoupling phenomenon occurred above 33°C, pointing out a possible role of glycerol in yeast thermal protection (1).

The industrial ethanol production industry is increasingly interested in the reduction of fermentation by-products, such as glycerol, since the price of the sugar source is a very important criterion for evaluation of the economic process. The hope is to increase the ethanol yield by minimizing glycerol production during ethanol fermentation. The data here showed that reducing the glycerol yield by 50% by glucose feeding monitoring led to a 20% decrease in the ethanol yield to the benefit of biomass (30%), acetate, and succinate compared to the reference fermentation (Table 1). When the formation of glycerol was reduced in a microaerobic ethanolic fermentation in continuous culture by a carefully controlled oxygenation, a decrease in the ethanol yield was also observed (8). RQ-controlled fermentation led to a lower maximum specific ethanol production rate and average ethanol productivity; however, it was possible to reach high ethanol production (85 g · liter^–1 in 30 h). The obtained mass ratio of ethanol production to glycerol production of 50 was the highest among the highly productive ethanol production processes reported in the literature (within 10 to 20) (33, 39). Previous attempts at reducing the glycerol formation by construction of mutant strains were reported in the literature (18, 19). The authors succeeded in increasing the ethanol yield by 10% with a 38% lower glycerol yield in anaerobic cultivations by changing the cofactor requirement in amino acid synthesis (18). Mutant strains impaired in glycerol synthesis were also obtained by deleting the glycerol dehydrogenase isoenzymes GPD1 (glycerol 3-phosphate dehydrogenase isoform 1) and GPD2 (19). Under aerobic conditions, this deletion resulted in a 12% increase in the ethanol yield with no glycerol production and higher acetate and succinate yields, but it also resulted in a dramatic reduction in the maximum specific growth rate and biomass formation. However, these two types of mutant strains led to poor ethanol production, since 10 g liter^–1 ethanol were produced in 16 h with no glycerol under aerobic conditions (19) and in 30 h with 1.3 g liter^–1 glycerol in anaerobic cultivations (18).

We showed here that it was possible to reduce glycerol production in very high-performance ethanolic fermentation process by monitoring the glucose feeding and that metabolic models may be valuable tools for predicting optimal experimental conditions. However, the microbiological engineering and metabolic engineering approaches cited above showed that managing glycerol production during ethanol fermentation remains challenging. Strategies will have to deal with metabolic and stress management issues in highly productive systems, since viable biomass formation is critical in achieving high-performance ethanolic fermentation.

 
* Corresponding author. Mailing address: Biotechnology and Bioprocess Laboratory, UMR-CNRS 5504, UMR-INRA 792, Département de Génie Biochimique et Alimentaire, Institut National des Sciences Appliquées, 135 Avenue de Rangueil, 31077 Toulouse Cedex, France. Phone: (33) 5 61 55 94 47. Fax: (33) 5 61 55 94 00. E-mail: guillouet{at}insa-toulouse.fr.

Supplemental material for this article may be found at http://aem.asm.org/.

These two authors worked equally on this project.

Top Abstract Introduction Materials and Methods Results Discussion References

 

1
1. Aldiguier, A. S., S. Alfenore, X. Cameleyre, G. Goma, J. L. Uribelarrea, S. E. Guillouet, and C. Molina-Jouve. 2004. Synergistic temperature and ethanol effect on Saccharomyces cerevisiae dynamic behaviour in ethanol bio-fuel production. Bioprocess Biosyst. Eng. 26:217-222.[Medline]
2
2. Alfenore, S., C. Molina-Jouve, S. E. Guillouet, J. L. Uribelarrea, G. Goma, and L. Benbadis. 2002. Improving ethanol production and viability of Saccharomyces cerevisiae by a vitamin feeding strategy during fed-batch process. Appl. Microbiol. Biotechnol. 60:67-72.[CrossRef][Medline]
3
3. Alfenore, S., X. Cameleyre, L. Benbadis, C. Bideaux, J. L. Uribelarrea, G. Goma, C. Molina-Jouve, and S. E. Guillouet. 2004. Aeration strategy: a need for very high ethanol performance in Saccharomyces cerevisiae fed-batch process. Appl. Microbiol. Biotechnol. 63:537-542.[CrossRef][Medline]
4
4. Andre, L., A. Hemming, and L. Adler. 1991. Osmoregulation in Saccharomyces cerevisiae: studies on the osmotic induction of glycerol production and glycerol 3-phosphate dehydrogenase (NAD⁺). FEBS Lett. 286:13-17.[CrossRef][Medline]
5
5. Bakker, B. M., K. M. Overkamp, A. J. Van Maris, P. Kotter, M. A. Luttik, J. P. Van Dijken, and J. T. Pronk. 2001. Stoichiometry and compartmentation of NADH metabolism in Saccharomyces cerevisiae. FEMS Microbiol. Rev. 25:15-37.[CrossRef][Medline]
6
6. Blom, J., M. Joost Teixeira de Mattos, and L. A. Grivell. 2000. Redirection of the respiro-fermentative flux distribution in Saccharomyces cerevisiae by overexpression of the transcription factor Hap4p. Appl. Environ. Microbiol. 66:1970-1973.[Abstract/Free Full Text]
7
7. Follstad, B. D., R. R. Balcarcel, G. Stephanopoulos, and D. I. C. Wang. 1999. Metabolic flux analysis of hybridoma continuous culture steady state multiplicity. Biotechnol. Bioeng. 63:675-683.[CrossRef][Medline]
8
8. Franzen, C. J. 2003. Metabolic flux analysis of RQ controlled microaerobic ethanol production by Saccharomyces cerevisiae. Yeast 20:117-132.[CrossRef][Medline]
9
9. Goel, A., J. Lee, M. M. Domach, and M. M. Ataai. 1999. Metabolic fluxes, pools, and enzyme measurements suggest a tighter coupling of energetics and biosynthetic reactions associated with reduced pyruvate kinase flux. Biotechnol. Bioeng. 64:129-134.[CrossRef][Medline]
10
10. Gombert, A. K., and J. Nielsen. 2000. Mathematical modelling of metabolism. Curr. Opin. Biotechnol. 11:180-186.[CrossRef][Medline]
11
11. Grosz, R., and G. Stephanopoulos. 1990. Physiology, biochemical and mathematical studies of microaerobic continuous ethanol fermentation by Saccharomyces cerevisiae. I. Hysteresis, oscillations and maximum specific ethanol productivities in chemostat culture. Biotechnol. Bioeng. 36:1006-1019.[CrossRef]
12
12. Kajiwara, Y., K. Ogawa, H. Takashita, and T. Omori. 2000. Enhanced glycerol production in Shochu yeast by heat-shock treatment is due to prolonged transcription of GPD1. J. Biosci. Bioeng. 90:121-123.[Medline]
13
13. Lee, B., J. Yen, L. Yang, and J. C. Liao. 1999. Incorporating qualitative knowledge in enzyme kinetic models using fuzzy logic. Biotechnol. Bioeng. 62:722-729.[CrossRef][Medline]
14
14. Michnick, S., J. L. Roustan, F. Remize, P. Barre, and S. Dequin. 1997. Modulation of glycerol and ethanol yields during alcoholic fermentation in Saccharomyces cerevisiae strains overexpressed or disrupted for GPD1 encoding glycerol 3-phosphate dehydrogenase. Yeast 13:783-793.[CrossRef][Medline]
15
15. Monod, J. 1958. Recherches sur la croissance des cultures bactériennes. Hermann et Cie, Paris, France.
16
16. Nevoigt, E., and U. Stahl. 1997. Osmoregulation and glycerol metabolism in the yeast Saccharomyces cerevisiae. FEMS Microbiol. Rev. 21:231-241.[CrossRef][Medline]
17
17. Nielsen, L. K., G. K. Smyth, and P. F. Greenfield. 1991. Hematocytometer cell count distributions: implications of non-Poisson behavior. Biotechnol. Prog. 7:560-563.[CrossRef]
18
18. Nissen, T. L., M. C. Kielland-Brandt, J. Nielsen, and J. Villadsen. 2000. Optimization of ethanol production in Saccharomyces cerevisiae by metabolic engineering of the ammonium assimilation. Metab. Eng. 2:69-77.[CrossRef][Medline]
19
19. Nissen, T. L., C. W. Hamann, M. C. Kielland-Brandt, J. Nielsen, and J. Villadsen. 2000. Anaerobic and aerobic batch cultivations of Saccharomyces cerevisiae mutants impaired in glycerol synthesis. Yeast 16:463-474.[CrossRef][Medline]
20
20. Omori, T., K. Ogawa, Y. Umemote, K. Yuki, Y. Kajihara, M. Shimoda, and H. Wada. 1996. Enhancement of glycerol production by brewing yeast (Saccharomyces cerevisiae) with heat shock treatment. J. Ferment. Bioeng. 82:187-190.[CrossRef]
21
21. Overkamp, K. M., B. M. Bakker, P. Kötter, M. A. H. Luttik, J. P. Van Dijken, and J. T. Pronk. 2002. Metabolic engineering of glycerol production in Saccharomyces cerevisiae. Appl. Environ. Microbiol. 68:2814-2821.[Abstract/Free Full Text]
22
22. Pitkanen, J.-P., A. Aristidou, L. Salusjarvi, L. Ruohonen, and M. Penttila. 2003. Metabolic flux analysis of xylose metabolism in recombinant Saccharomyces cerevisiae using continuous culture. Metab. Eng. 5:16-31.[CrossRef][Medline]
23
23. Postgate, J. P. 1967. Viability measurements and the survival of microbes under minimum stress. Adv. Microb. Physiol. 1:1-23.
24
24. Pramanik, J., P. L. Trelstad, A. J. Schuler, D. Jenkins, and J. D. Keasling. 1999. Development and validation of a flux-based stoichiometric model for enhanced biological phosphorus removal metabolism. Water Res. 33:462-476.[CrossRef]
25
25. Remize, F., J. L. Roustan, J. M. Sablayrolles, P. Barre, and S. Dequin. 1999. Glycerol overproduction by engineered Saccharomyces cerevisiae wine yeast strains leads to substantial changes in by-product formation and to a stimulation of fermentation rate in stationary phase. Appl. Environ. Microbiol. 65:143-149.[Abstract/Free Full Text]
26
26. Rizzi, M., M. Baltes, U. Theobald, and M. Reuss. 1997. In vivo analysis of metabolic dynamics in Saccharomyces cerevisiae. II. Mathematical model. Biotechnol. Bioeng. 55:592-608.[CrossRef]
27
27. Roels, J. A. 1983. Energetics and kinetics in biotechnology. Elsevier Biomedical Press, Amsterdam, The Netherlands.
28
28. Savageau, M. A. 1976. Biochemical systems analysis: a study of function and design in molecular biology. Addison Wesley, London, United Kingdom.
29
29. Schügerl, K. 2001. Progress in monitoring, modelling and control of bioprocesses during the last 20 years. J. Biotechnol. 85:149-173.[CrossRef][Medline]
30
30. Stephanopoulos, G. 1999. Metabolic fluxes and metabolic engineering. Metab. Eng. 1:1-11.[CrossRef][Medline]
31
31. Taherzadeh, M. J., L. Adler, and G. Liden. 2002. Strategies for enhancing fermentative production of glycerol: a review. Enz. Microb. Technol. 31:53-66.[CrossRef]
32
32. Takiguchi, N., H. Shimizu, and S. Shhioya. 1997. An on-line physiological state recognition system for the lysine fermentation process based on a metabolic reaction model. Biotechnol. Bioeng. 55:170-181.[CrossRef]
33
33. Thomas, K. C., S. H. Hynes, A. M. Jones, and W. M. Ingledew. 1993. Production of fuel alcohol from wheat by VHG technology. Appl. Biochem. Biotechnol. 43:211-226.
34
34. Van Cal, H. J. L., H. A. B. Braake, C. Hellinga, K. C. A. M. Luyben, and J. J. Heijnen. 1997. An efficient model development strategy for bioprocesses based on neural networks in macroscopic balances. Biotechnol. Bioeng. 54:549-566.[CrossRef]
35
35. Varma, A., and B. O. Palsson. 1994. Metabolic flux balancing: basic concepts, scientific and practical use. Bio/Technology 12:994-998.[CrossRef]
36
36. Varma, A., and B. O. Palsson. 1994. Stoichiometric flux balance models quantitatively predict growth and metabolic by-product secretion in wild-type Escherichia coli W3110. Appl. Environ. Microbiol. 60:3724-3731.[Abstract/Free Full Text]
37
37. Varner, J., and D. Ramkrishna. 1998. Application of cybernetic models to metabolic engineering: investigation of storage pathways. Biotechnol. Bioeng. 58:282-291.[CrossRef][Medline]
38
38. Vaseghi, S., A. Baumeister, M. Rizzi, and M. Reuss. 1999. In vivo dynamics of the pentose phosphate pathway in Saccharomyces cerevisiae. Metab. Eng. 1:128-140.[CrossRef][Medline]
39
39. Wang, S., K. C. Thomas, W. M. Ingledew, K. Sosulski, and F. W. Sosulski. 1998. Production of fuel ethanol from rye and triticale by very-high-gravity (VHG) fermentation. Appl. Biochem. Biotechnol. 69:157-175.
40
40. Wang, Z. X., J. Zhuge, H. Fang, and B. A. Prior. 2001. Glycerol production by microbial fermentation: a review. Biotechnol. Adv. 19:201-223.[CrossRef][Medline]
41
41. Williams, E. M. 1967. A method of cell growth dynamics. J. Theor. Biol. 15:190-207.[CrossRef][Medline]
42
42. Wojda, I., R. Alonso-Monge, J. P. Bebelman, W. H. Mager, and M. Siderius. 2003. Response to high osmotic conditions and elevated temperature in Saccharomyces cerevisiae is controlled by intracellular glycerol and involves coordinate activity of MAP kinase pathways. Microbiology 149:1193-1204.[Abstract/Free Full Text]
43
43. Yang, Y. T., G. N. Bennett, and K. Y. San. 1999. Effect of inactivation of nuo and ackA-pta on redistribution of metabolic fluxes in Escherichia coli. Biotechnol. Bioeng. 65:291-297.[CrossRef][Medline]

---

Applied and Environmental Microbiology, March 2006, p. 2134-2140, Vol. 72, No. 3
0099-2240/06/$08.00+0     doi:10.1128/AEM.72.3.2134-2140.2006
Copyright © 2006, American Society for Microbiology. All Rights Reserved.

This Article Abstract Full Text (PDF) Supplemental material Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Bideaux, C. Articles by Guillouet, S. E. Search for Related Content PubMed PubMed Citation Articles by Bideaux, C. Articles by Guillouet, S. E. Agricola Articles by Bideaux, C. Articles by Guillouet, S. E.