OPTIMIZATION OF CITRIC ACID PRODUCTION FROM GLUCOSE SYRUP BY YARROWIA LIPOLYTICA USING RESPONSE SURFACE METHODOLOGY

Waldemar Rymowicz¹, Edmund Cibis^2
1 Department of Biotechnology and Food Microbiology, Wrocław University of Environmental and Life Sciences, Poland
² Department of Bioprocess Engineering, Wrocław University of Economics, Poland

ABSTRACT

A new criterion was proposed to evaluate the process of citric acid biosynthesis by the acetate negative mutant strain Yarrowia lipolytica AWG-7 cultivated on glucose syrup by the repeated-batch method. This criterion imparts the same weight to the overall amount of the citric acid production and to the trend in citric acid concentration. To determine the optimum levels of bacto-peptone, ammonium chloride and potassium dihydrogen phosphate which maximize the proposed objective function, a central composite design was developed using 50 ml repeated-batch fermentation. The design involved 20 processes conducted with various combinations of the five values of these three parameters. The experiments produced empirical values of the proposed criterion which were then approximated with the fourth-order polynomial. It was found that the optimal concentrations of bacto-peptone, ammonium chloride, and potassium dihydrogen phosphate equaled 8.5; 159.8 and 65 mgl^-1, respectively.

Key words: Yarrowia lipolytica, glucose syrup, citric acid, response surface methodology, acetate negative mutant.

INTRODUCTION

Citric acid, a popular component made use of in food and pharmaceutical industries, is synthesized by the filamentous fungus Aspergillus niger in an industrial-scale submerged fermentation process. The annual global production of citric acid is estimated at 880,000 metric tons [8]. Because of the large and ever-increasing demand for citric acid, the use of alternative fermentation processes involving yeast strains, especially in continuous operations, seems to be desirable for its production [9,13,22,23]. The available literature includes references to the chemostat cultivation of Y. lipolytica for citric acid production from hydrocarbons, glucose and ethanol [1,9,13]. There are also reports on attempts to develop a continuous process of citrate production from glucose with immobilized cells and with a membrane bioreactor [22,23].

Some major drawbacks of conventional mould fermentation, such as high sensitivity to trace metals and low production rates, might be overcome by using yeast strains. Many yeasts that grow on carbohydrate substrates have the ability to accumulate high concentrations of citric acid during tricarboxylic acid cycle respiration. However, Yarrowia lipolytica is the only species known for its capability of maximizing citric acid production from various substrates such as carbohydrate, hydrocarbon, ethanol, methanol, fatty acids, acetate and glycerol; glucose and glucose syrup having attracted particular interest [3,4,10,18-21,27,28]. Y. lipolytica has a noticeable ability to produce extracellular citric acid and isocitric acid from glucose syrup. The concentration of isocitrate can be effectively decreased by modification of the production media either by the addition of inhibitors of aconitate hydratase (e.g. monofluoroacetate or fluorocitrate) or omission of Fe ions, which are constituents of the aconitate protein molecule [18]. Alternatively, mutant strains that give improved yields of citrate without concomitant accumulation of isocitric acid have been developed [26]. Because aconitate hydratase is the enzyme responsible for the conversion of citrate to isocitrate, either mutants of low aconitate activity or mutants with an enzyme sensitive to monofluoroacetate have been isolated. An acetate negative mutant of the Y. lipolytica strain is advantageous in that less isocitric acid is obtained as a by-product from the glucose medium [27]. High productivity and concentration of citric acid (up to 120 gl^-1) have been achieved from glucose syrup with the selected acetate negative mutant Y. lipolytica AWG-7 [22].

In the literature, there is a considerable amount of data on citric acid production by the Y. lipolytica yeast. However, only a few papers include mathematical descriptions of citrate production by this microorganism in batch and continuous cultivation [12]. The response surface methodology (RSM) is preferred for the optimization of culture media, because it enables simultaneous consideration of several factors at many different levels and the determine the corresponding interaction among these factors using a smaller number of experimental observations. Recently, the RSM has been employed to solve multivariate problems and optimize several responses in many types of experimentation [15,17]. Central composite design has been used to optimize citric acid production by Candida lipolytica Y-1095 [6] and bacteriocin synthesis by Lactococcus lactis [5].

In this study a new criterion for the evaluation of citric acid biosynthesis was proposed and the RSM approach was adopted in order to locate the optimum levels of the bacto-peptone, ammonium chloride and potassium phosphate concentrations, which maximized the proposed criterion. The chosen factors are very important in the citric acid formation by the acetate mutant Y. lipolytica AWG-7.

MATERIALS AND METHODS

Microorganisms. The acetate negative mutant Y. lipolytica AWG-7 (smooth strain) used in this study came from collection of Department of Biotechnology and Food Microbiology, Agricultural University of Wroclaw, Poland. The microorganism was grown on yeast malt extract agar slants for 2 days and was stored at 4°C. New agar slants were inoculated every 1-2 months.

Substrate. Glucose syrup G95 used as the carbon source was obtained from Cargill Co., Bielany Wrocławskie, Poland. The syrup contained 95% glucose in dry matter.

Growth medium. Biomass yeast was produced in a medium containing (gl^-1): glucose syrup, 125.0; NH₄Cl, 3.0; KH₂PO₄, 0.2; MgSO₄·7H₂O, 1.0; and YE, 1.0 in tap water. Cultivation was carried out for 2 days in shake-flask fermentation in a 300 ml Erlenmeyer flask containing 50 ml of growth medium, at 160 rpm and 29°C; pH was maintained at 5.5 by addition of 10 gl^-1 CaCO₃.

Production medium. Statistical experimental design and empirical modeling were used for the optimization of the citric acid production medium which contained (gl^-1): 125 gl^-1 glucose syrup in tap water. Different concentrations of bacto-peptone, ammonium chloride and potassium dihydrogen phosphate were chosen as independent variables for the series of five cycles of repeated-batch fermentations. Repeated-batch fermentations were carried out in a 300 ml Erlenmeyer flask containing 50 ml of each medium and 0.8 g dry weight of yeast cells, at 160 rpm and 29°C. The production medium was replaced every 7 days. The biomass yeast was separated via 10 minute centrifugation (4000 g) at room temperature and washed with distilled water each time before transfer into the fresh medium. The pH was maintained at 5.5 using 10 gl^-1 CaCO₃. The supernatant was withdrawn for further analysis of citric acid and glucose concentrations.

ANALYTICAL METHODS

Biomass concentration. Yeast concentration was determined by dry cell weight analysis. 50 ml samples of the fermentation broth were centrifuged (4000 g) at room temperature for 10 min. The biomass yeasts were harvested by filtration through a pre-weighed membrane filter (cellulose nitrate filter, 1.2 µm pore size, Millipore) and dried at 80°C to a constant weight.

Citric acid and glucose concentrations. Concentrations of citric acid and glucose were determined using HPLC on an Aminex HPX87H Organic Acids Column coupled either to a UV detector (at 210 nm) or RI detector, respectively. The column was eluted with 20 mM H₂SO₄ at room temperature and a flow rate of 0.6 mlmin^-1. Citric acid and glucose were identified and quantified with reference to authentic standards.

Protein concentration. Concentration of protein in yeast biomass was determined according to Stewart [24].

Experimental design. The central composite rotatible experimental design was used to optimize the composition of the production medium for the maximization of the proposed criterion in order to evaluate the process of citric acid biosynthesis. Bacto-peptone (X₁, gl^-1), ammonium chloride (X₂, gl^-1) and potassium dihydrogen phosphate (X₃, gl^-1), concentrations were chosen as independent variables for the series of repeated-batch fermentations (Table 1).

Because of the three independent variables, it is necessary to use the 2³ factorial experimental design with eight star points (a=1.682) and six replicates at the central point, which leads to the total number of twenty experiments (Table 1).

RESULTS AND DISCUSSION

Criterion for the evaluation of citric acid biosynthesis, and optimization of the medium components

In our earlier studies, the medium composition constituents and the process parameters were optimized by single factor optimization, the other factors being kept constant [22]. As bacto-peptone, ammonium chloride and potassium dihydrogen phosphate were found to exert the strongest effect on citrate production from a complex medium, it seemed advisable to select them for statistical optimization by RSM. In order to find the optimum combination of the major components of the medium, experiments were performed according to the central composite rotatible design experiment plan (Table 1). Biosynthesis of citric acid by the acetate negative mutant Y. lipolytica AWG-7 from glucose syrup was conducted in repeated-batch fermentations. Arzumanov et al. [8] have shown that the use of repeated-batch cultivation for citric acid production by Y. lipolytica produced better results than those obtained with batch cultivation. What is more, in the repeated-batch cultivation system the activity of the culture remained stable for a long period of time (more than 700 h).

The criterion selected for assessing the course of citric acid biosynthesis was a factor that included not only the total amount of the citric acid produced during five successive cycles of the repeated-batch process but also the trend in the citric acid concentration determined for these five cycles. The factor takes the form of the following relations:

(1)

(2)

(3)

where
K_i = overall concentration of citric acid for five cultivations in the ith experiment;
K_max = maximal value chosen from all K_i;
K_min = minimal value chosen from all K_i;
A_max = maximal value chosen from all A_i;
A_i = slope of the straight line of the regression approximating the citric acid concentrations
in the ith experiment; A_min = minimal value chosen from all A_i ;
A_max = maximal value chosen from all A_i .

The objective function proposed imparts the same weight to the overall amount of the citric acid produced and to the trend in citric acid concentration. Thus, the higher the value of the function Y (equation 3), the more advantageous the course of the process. Hence, there was a need to maximize this function.

Using the data in Table 2 and considering the relations of (1) to (3), the experimental values of Y were calculated. As for their approximation the use of a second-order polynomial resulted in an inadequate function, we made use of the fourth-order polynomial (for α = 0.05) described by the following equation (4):

Y= a₁₁₁₁·X₁⁴ + a₂₂₂₂·X₂⁴ + a₃₃₃₃·X₃⁴ + a₁₁₃·X₁²·X₃ + a₁₂₂·X₁·X₂² + a₁₁·X₁² + a₂₂·X₂² + a₁₂·X₁·X₂ + a₁·X₁ + a₂·X₂ + a₃·X₃ + a₀    (4)

where with an appropriate subscript stands for the parameters of the polynomial, X₁, X₂, and X₃ are bacto-peptone, NH₄Cl, and KH₂PO₄, respectively, expressed by coded values. The parameters of the polynomial were estimated using the MicrosoftŽExcel 2000 software. The same software was used to calculate the values of the statistical factors which describe the accuracy of fitting the results calculated with the said polynomial to the experimental results. The estimated values of the polynomial parameters, as well as the statistical factors of estimation, are listed in Table 3.

The value of the coefficient of determination, which is 0.9026, indicates that 90.26% of variability in the experimental response is explained by the polynomial model, the remainder being residual variability. The value in the last row of the column in Table 3 (0.2377) shows that there is no reason, even at the significance level of α < 0.2377, for rejecting the hypothesis about the adequacy of the model (4). Function Y was maximized using the Optimization Toolbox of MATLAB (version 4.2c.1). The optimal composition of the production medium for the biosynthesis of citric acid from glucose syrup by Y. lipolytica AGW-7 in the repeated-batch cultivation system is characterized in Table 4. The contour graphs in Fig.1 relate the values of the polynomial (4) to the concentrations of two production medium components, the concentration of the third component being kept constant at the optimal level. The red lines show the boundary of the variability area for the concentrations of the medium components. The arrows indicate the position of the optimal point.

Characterization of Y. lipolytica AWG-7 biomass in repeated-batch citric acid fermentations

Fig. 2 and 3 show the experimental data obtained during five cycles of repeated-batch fermentations of Y. lipolytica. Biomass concentration at the end of the repeated-batch process ranged from 15 to 21 gl^-1. In the cultures involving production media which contained bacto-peptone and ammonium chloride the biomass amount was greater than in the control culture (No. 0) where the production medium contained glucose syrup alone (125 gl^-1). The presence of the two compounds accounted for an increment in the biomass as high as 28% at the end of the process. Protein concentration in the biomass of Y. lipolytica AWG-7 varied between 14.8 and 23%. The process carried out with a production medium containing 0.5 gl^-1 bacto-peptone and 0.3 gl^-1 ammonium chloride (No. 12) was found to display the highest stability. The Y. lipolytica AWG-7 yeast had a 23% protein content and produced 36.7 gl^-1 citric acid in the final cycle. According to many investigators, it is the protein concentration in the yeast biomass that determines the efficiency of citric acid biosynthesis by the Y. lipolytica strain. Yeast with a 19 to 25% protein content produced the greatest amounts of citric acid [14, 22].

Einesele et al. [7] have reported that nitrogen limitation leads to the reduction of the intracellular protein content and influences the formation of nucleic acid and well wall polymers. Kozlova et al. [14] have found a strong reduction in the protein content in C. lipolytica from 43% in the exponential phase to 17% during the stationary growth phase. A similar decrease in nitrogen content (from 8.5% in the trophophase to 4% at the end of the exponential phase) has been reported by Briffaud and Engasser for Saccharomycopsis lipolytica D1805 [4]. Moresi has observed a reduction in the intracellular nitrogen fraction from 7-8% to 2.3-4.4% in Y. lipolytica ATCC 20346 [16]. According to Anastassiadis et al. [2], intracellular nitrogen limitation and the increase in intracellular NH₄⁺ concentration are the most important factors influencing the formation of citric acid in yeast.

CONCLUSION

The results obtained in this study made it possible to construct empirical models which are based on statistical analysis and describe the effect of bacto-peptone concentration, ammonium chloride concentration and potassium dihydrogen phosphate concentration on citric acid production by Y. lipolytica AWG-7 under conditions of repeated-batch fermentation. Considering the data established in this way, the following conclusions can be drawn:

1. Bacto-peptone, ammonium chloride and potassium dihydrogen phosphate concentrations had a significant influence on the production of citric acid, the concentration of protein in the biomass of Y. lipolytica AWG-7 and the stability of the process.

2. The results obtained in this study will be used to develop a new process for the production of citric acid in continuous membrane fermentation.

ACKNOWLEDGEMENTS

This research was supported by a grant from State Committee for Scientific Research within the project 3P06T 03624.

REFERENCES

1. Aiba S., Matsuoka M., 1978, Citrate production from n-alkane by Candida lipolytica in reference to carbon fluxes in vivo. Eur. J. Appl. Microbiol. Biotechnol. 5: 247-261.

2. Anastassiadis S.. Aivasidis A.. Wandrey C., 2002, Citric acid production by Candida strains under intracellular nitrogen limitation. Appl. Microbiol. Biotechnol. 60: 81-87.

3. Arzumanov T. E., Shiskanova N. V., Finogenova T. V., 2000, Biosynthesis of citric acid by Yarrowia lipolytica repeat-batch culture on ethanol. Appl. Microbiol. Biotechnol. 53: 525-529.

4. Briffaud I., Engasser I.M., 1979, Citric acid production from glucose. II Growth and excretion kinetics in trickle flow fermentor. Biotechnol. Bioeng. 21: 2093-2111.

5. Chan L., Jinghua B., Zhaoling C., Fan O., 2002, Optimization of cultural medium for bacteriocin production by Lactococcus lactis using response surface methodology. J. Biotechnol. 93: 27-34.

6. Crolla A., Kennedy K.J., 2001, Optimization of citric acid production from Candida lipolytica Y-1095 using n-paraffin. J. Biotechnol. 89: 27-40.

7. Einelese A., Finn A., Samhaber W., 1985, Mikrobiologische und biochemische Verfahrenstechnik. Verlag Chemie, Weiheim.

8. Esker T., Janshekar H., Sakuma Y., 1999, CEH report citric acid. http://ceh.sric.com/Enframe/Report.html?report=636.5000&show=Navigation.html.15H, 17/10/2002.

9. Finogenova T.V., Kamzolova S.V., Shishkanova N.V., Ilchenko A.P., Dedynkhina E.G., 1996, Influence of the specific growth rate and Zn ions on the synthesis of citric acid and isocitric acid on the biomass composition of Y.lipolytica N1 yeast. Appl. Biochem. Microbiol. 32(1): 30-34.

10. Kamzolova S.V., Shishkanova N.V., Morgunov I.G., Finogenova T.V., 2003, Oxygen requirements for growth and citric acid production of Yarrowia lipolytica. FEMS Yeast Research 3: 217-222.

11. Kautola H., Rymowicz W., Linko Y.Y., Linko P., 1991, Production of citric acid with immobilized Yarrowia lipolytica. Appl. Microbiol. Biotechnol. 35: 447-449.

12. Kilic M., Bayraktar E., Ates S., Mehmetoglu U., 2002, Investigation of extractive citric acid fermentation using response surface methodology. Process Biochem. 37: 759-767.

13. Klasson T.K., Clausen E.C., Gaddy J.L., 1989, Continuous fermentation for the production of citric acid from glucose. Appl. Biochem. Biotechnol. 20/21: 491-509.

14. Kozlova T.M., Medvedeva G.A., Glazunova L.M., Finogenova T.V., 1981, Structural changes in the cells of Candida lipolytica in the biosynthesis of citric acid. Microbiology, 5: 508-514.

15. Medeiros A. B. P., Pandey A., Freitas R. J. S., Christen P., Soccol C. R., 2000, Optimization of the production of aroma compounds by Kluyveromyces marxianus in solid-state fermentation using factorial design and response surface methodology. Biochem. Eng. J. 6: 33-39.

16. Moresi M., 1994, Effect of glucose concentration on citric acid production by Yarrowia lipolytica. J. Chem. Tech. Biotechnol. 60: 387-395.

17. Musiał I., Rymowicz W., Cibis E., 2004, Optimization of single-cell-biomass production by Yarrowia lipolytica using response surface methodology and pulse method. EJPAU. Biotechnol. Vol. 7, Issue 1.

18. Nakanishi T., Yamamoto M., Kimura K., Tanaka K., 1972, Fermentative production of citric acid from n-paraffin by yeasts. J. Ferm. Technol. 50(12): 855-867.

19. Papanikolaou S., Muniglia L., Chevalot I., Aggelis G., Marc I., 2002, Yarrowia lipolytica as a potential producer of citric acid from raw glycerol. J. Appl. Microbiol. 92: 737-744.

20. Pazouki M., Felse P.A., Sinha J., Panda T., 2000, Comparative studies on citric acid production by Aspergillus niger and Candida lipolytica using molasses and glucose. Bioprocess Eng. 22: 353- 361.

21. Rane K.D., Sims K.A., 1993, Production of citric acid by Candida lipolytica Y1095: Effect of glucose concentration on yield and productivity. Enzyme Microbiol. Technol., 15: 646-651.

22. Rymowicz W., Rywińska A., 2003, Cišgła produkcja kwasu cytrynowego z syropu glukozowego przez mutanta Yarrowia lipolytica w reaktorze membranowym. Acta Scient. Polon. Biotechnol. 2(1-2): 21-30.

23. Rymowicz W., Kautola H., Wojtatowicz M., Linko Y.Y., Linko P., 1993, Studies on citric acid production with immobilized Yarrowia lipolytica in repeated batch and continuous air- lift bioreactors. Appl. Microbiol. Biotechnol. 39: 1-4.

24. Stewart K., 1975, Methods in Cell Biology. Ed. Prescott D. M. New York, 12 (8): 112-118.

25. Tisandjaja D., Gutierrez N.A., Maddox J.S., 1996, Citric acid production in a bubble- column reactor using cells of the yeast Candida guilliermondii immobilized by adsorption onto sawdust. Enzyme Microbiol. Technol. 19: 343-347.

26. Wojtatowicz M., Marchin G.L., Erickson L.E., 1993, Attempts to improve strain A-101 of Yarrowia lipolytica for citric acid production from n-paraffins. Process Biochem. 28(7): 453-460.

27. Wojtatowicz M., Rymowicz W., Kautola H., 1991, Comparison of different strains of the yeast Yarrowia lipolytica for citric acid production from glucose hydrol. Appl. Biochem. Biotechnol. 31: 165- 174.

28. Żarowska B., Wojtatowicz M., Rymowicz W., Robak M., 2001, Production of citric acid on sugar beet molasses by single and mixed cultures of Yarrowia lipolytica. EJPAU. Biotechnol.. Vol. 4. Issue 2.

Responses to this article, comments are invited and should be submitted within three months of the publication of the article. If accepted for publication, they will be published in the chapter headed 'Discussions' and hyperlinked to the article.