Journal of Biotechnology 76 (2000) 147 – 155 www.elsevier.com/locate/jbiotec
Optimization of gibberellic acid production by immobilized Gibberella fujikuroi mycelium in fluidized bioreactors Eleazar M. Escamilla S. *, Luc Dendooven, Ignacio P. Magan˜a, R. Parra S., M. De la Torre Department of Biotechnology and Bioengineering, CINVESTAV-IPN, Mexico D.F., Mexico Received 26 March 1999; received in revised form 13 July 1999; accepted 30 July 1999
Abstract An orthogonal experimental design L9 (34) was used to investigate effects of temperature, pH, C:N ratio (glucoseC, NH4ClN) and concentrations of rice flour on production of gibberellic acid by Gibberella fujikuroi in 3.5 l fluidized bioreactors. The gibberellic acid production in a fluidized bioreactor could reach 3.90 g l − 1, more than 3-times greater than previously reported for submerged and solid fermentations. pH, rice flour concentration and C:N ratio were the factors that most influenced the production of gibberellic acid; pH being the most important. The response surface of gibberellic acid production to changes in pH and C:N ratio or rice flour concentration indicated that greatest production was found with a C:N ratio of 36.8 and pH 5 while the optimum concentration for rice flour was 2 g l − 1 and production increased with increased pH. The effect of temperature on the production of gibberellic acid was also significant and greatest production was at 30°C. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Orthogonal design; Fluidized bioreactor; Gibberellin; Gibberellic acid
1. Introduction The industrial production of plant growth promoters, such as, gibberellins by Gibberella fujikuroi, in batch submerged fermentators can help to increase agricultural production. The chemistry, biosynthesis, action mode, relationship be* Corresponding author. Present address: Chemical Department of Technological Institute of Celaya, Av. Tecnolo´gico y A. Garcı´a Cubas S/N, C.P. 38010, Celaya Gto., Me´xico. Tel.: + 52-461-17575, ext. 130; fax: +52-461-17744. E-mail address:
[email protected] (E.M. Escamilla S.)
tween structure and activity and the use of gibberellins has been investigated extensively (e.g. Bru¨ckner and Blechsmidt, 1991; Hollmann et al., 1995; Lu et al., 1995) but little is known about the fermentation production of gibberellin. A profound knowledge of mycelial physiology is necessary for the design and control of industrial fermentation processes (Fredickson et al., 1979; Kossen and Oosterhuis, 1985; Nielsen and Villadsen, 1992). The most important nutrients have to be applied in specific concentrations for maximum mould growth and production of secondary metabolites (Gutke, 1980).
0168-1656/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 1 6 5 6 ( 9 9 ) 0 0 1 8 2 - 0
148
E.M. Escamilla S. et al. / Journal of Biotechnology 76 (2000) 147–155
The viscosity of a suspension of free fungal cells increases during fermentations, reducing gas and liquid mass transfer rates (Karel et al., 1985). Immobilisation of fungi in pellets overcomes this problem (Karel et al., 1985; Knorr, 1987) and pellets can be re-used extending the production of secondary metabolites (Cheetham, 1980). The aim of this work was to find trough an orthogonal experimental array the culture medium that produced the largest concentration of gibberellic acid in a fluidized bioreactor using immobilized mycelia of G. fujikuroi in Capolygalacturonate.
2.3. Preparation of immobilized mycelium
G. fujikuroi (Sawada) strain CDBB H-984 conserved in potato glucose slants at 4°C and subcultured every two months was used in this experiment (Culture collection of the Department of Biotechnology and Bioengineering, CINVESTAV-IPN, Mexico).
Inside a sterilized laminar flow chamber mycelia were immobilized in Ca-polygalacturonate (Modified from Montes and Magan˜a, 1991). Briefly, a suspension of G. fujikuroi mycelium was mixed with an equal volume of 8% (w/v) polygalacturonic acid solution. The mixture was homogenized in a Mixer (Osterizer junior of 230 ml at 100 rpm) for 10 min. The solution was adjusted with polygalacturonic acid to a final concentration of 4%. The mixture was forced through a multi-needle template (gauge 18 for 5 mm beads and 21 for 3 mm ones) with a peristaltic pump (CRODE, Mexico) flowing at 10 ml min − 1 and the droplets were collected in a sterile 3.5% CaCl2 solution. After soaking for 3 h, the liquid was decanted and the spherical beads were washed with sterile distilled H2O and stored at 4°C for 24 h. The diameter of the beads was measured in a microscope (Leica, LMDS) with a micrometer grid. Pellets of 3 mm or 5 mm ( 90.15) were selected and tested for production of gibberellic acid.
2.2. Culture medium
2.4. Batch culture in fluidized bioreactor
In a preliminary laboratory experiment effects of different concentrations of glucose, NH4Cl, KH2PO4 and MgSO4 on gibberellic acid production were investigated. An inoculum of G. fujikuroi (Sawada) was added to a culture medium (ratio of 1:10, v/v) in 500 ml shake flasks rotating at 300 rpm and incubated at 29°C while pH was not adjusted during the fermentation process. The different concentrations of glucose used in the culture medium were 80 g l − 1, 100 g l − 1 or 120 g l − 1, those for NH4Cl 1 g l − 1, 3 g l − 1 or 5 g l − 1, for KH2PO4 3 g l − 1, 5 g l − 1, or 7 g l − 1, for MgSO4 0.5 g l − 1, 1.5 g l − 1 or 2 g l − 1, and for rice flour 0.5 g l − 1, 2 g l − 1 or 3.5 g l − 1. The rice flour had a water content of 11%, contained 80% carbohydrates, 7% proteins, 1.1% fat, 0.4% crude fibre and 0.3% ash. Its organic C content was 340 mg C g − 1 of dry rice flour and assuming that all nitrogen is in the form of proteins than the total organic N content was 11.8 mg N g − 1 of dry rice flour.
An orthogonal experimental design L9 (34) in duplicate (Wu and Hobbs, 1987; Peace, 1991) was used to investigate effects of temperature, pH, C:N ratio (glucose, NH4Cl) and rice flour concentrations (Table 1) on gibberellic acid production in two, 3.5-l bioreactors (CRODE, Mexico, Fig. 1) in a random sequence. In design of experiments, orthogonal means balanced, separable, not mixed or confounded. The symbol La(bc) is used to represent the orthogonal array where ‘a’ is the number of experimental runs, ‘b’ the number of levels for each factor or variable and ‘c’ the number of factors investigated. Concentration for KH2PO4 were 3 g l − 1 and 1.5 g l − 1 for MgSO4. Rice flour was used as G. fujikuroi easily infects and grows abundantly in rice plants (Coolbaugh, 1983). pH tested for were 2, 3.5 and 5, because G. fujikuroi starts to produce gibberellins GA4 and GA7 above pH 5.5 and we wanted to investigate production of gibberellic acid (GA3). The culture media were sterilized in situ and the immobilized
2. Materials and methods
2.1. Micro-organism
E.M. Escamilla S. et al. / Journal of Biotechnology 76 (2000) 147–155
149
Table 1 Orthogonal experimental design L9 (34) run in duplicate to assess effects of temperature, pH, C:N ratio (glucose, NH4Cl) and rice flour concentration on production of gibberellic acid with 3 g l−1 KH2PO4 and 1.5 g l−1 MgSO4 Experiment
Temperature (°C)
PH
C:N ratio
Rice flour concentration (g l−1)
Mean gibberellic acid production (g l−1)
1 2 3 4 5 6 7 8 9
35 30 30 35 25 30 35 25 25
3.5 2.0 5.0 5.0 3.5 3.5 2.0 2.0 5.0
36.8 51.0 36.8 51.0 51.0 122.0 122.0 36.8 122.0
3 3 2 1 2 1 2 1 3
0.367 0.144 2.862 1.179 0.335 0.260 0.270 0.193 0.869
mycelium beads added. Per batch, 7070 (9 707) pellets were added and each pellet had a wet weight of 1.4 ×10 − 4 g mycelium so that approximately a total of 1 g of mycelium was added. Aeration rates of 3 volumes of air per volume of
medium per minute (vvm) were used to maintain the fluidized state during fermentation as they a gave larger production of gibberellic acid than an aeration rate of 1 and 5 vvm. An aeration rate of 1 and 5 vvm resulted in a gibberellic acid produc-
Fig. 1. Fermentation system.
150
E.M. Escamilla S. et al. / Journal of Biotechnology 76 (2000) 147–155
tion B600 mg l − 1 while it was \900 mg l − 1 with an aeration rate of 3 mm. During the fermentation of 192 h at 29°C, sterile H2O was added at 10 h or 12 h intervals to compensated for evaporation loss while pH was kept constant in each run by adding HCl or NaOH. Each day, a sub-sample of 30 ml was taken from the culture medium and analysed for biomass dry weight, reductive sugars, NH+ 4 N, pH and gibberellic acid.
2.5. Analytical methods Ten beads were taken from the 30 ml sub-sample, disintegrated by adding 25 ml 0.5 M EDTAphosphate solution and stirred for 25 min at 250 rpm. The released mycelium was filtered through a pre-weighted Millipore 0.45 mm membrane and free mycelium biomass was measured by drying them to constant weight in an oven at 90°C. NH+ 4 N in the medium was determined by Kjeldahl (A.O.A.C., 1990) and reducing sugars, i.e. glucose, by the dinitrosalycilic acid method (Miller, 1959). A sub-sample of 10 ml filtered culture medium was acidified to pH 2.5 with 10% HCl and extracted three times with ethyl acetate. The organic phase was vacuum dried and re-dissolved in methanol. The sample was filtered three times through a Sep Cartridge-Pak Millipore, C-18 part 51910 and analysed for gibberellic acid on a high resolution chromatograph (HPLC) using a Bondapak C-18 analytical reverse-phase HPLC column at 25°C and a UV detector at 204 nm (LC 1150, GBC Scientific Equipment, Australia). The mobile phase was a mixture of 75% methanol and 25% KH2PO4 buffer (75:25) flowing at 1.8 ml min − 1.
3. Results and discussions The greatest production of gibberellic acid in shake flasks, i.e. 1.1 g l − 1, was obtained with 100 g glucose l − 1, 1 g NH4Cl l − 1, and 2 g of rice flour l − 1 while concentrations of KH2PO4 and MgSO4 tested for, had no significant effect on gibberellic acid production. Consequently, effects of glucose,
NH4Cl and rice flour concentrations on production of gibberellic acid were tested for in a fluidized bioreactor while concentrations of KH2PO4 and MgSO4 were not varied. Preliminary experiments indicated that pellets of 3 mm gave a production of gibberellic acid \900 mg l − 1 while pellets of 5 mm gave B 600 mg l − 1. Pellets of 3 mm were thus used in further experiments. In each of the bioreactor runs concentrations of glucose and NH+ 4 N decreased sharply while microbial biomass dry weight increased immediately (Fig. 2). The production of gibberellic acid sometimes showed a lag. These dynamics found in a fluidized bioreactor were similar to those reported by Borrow et al. (1964) for shake flasks and submerged fermentations. The gibberellic acid production ranged from 0.144 to 2.862 g l − 1 (Table 1). Early work by Rehm (1980) using surface cultivation and low nutrient concentrations reported values ranging from 40 to 60 mg l − 1 of gibberellic acid. Production of gibberellic acid in shake flasks using the same media was 200 mg l − 1 (Bru¨ckner and Blechsmidt, 1991). Further increases up to 1 g l − 1 were obtained in batch techniques when the medium composition was optimized, especially the initial nitrogen and carbon concentration (Holme and Zacharias, 1965). In solid fermentation the gibberellic acid production was 1.6-times greater when wheat bran was used than in submerged fermentation (0.8 g l − 1) but control of factors affecting the production in these systems is extremely difficult (Kumar and Lonsane, 1987). Kumar and Lonsane (1988) using a non-conventional bioreactor and immobilized mycelia of G. fujikuroi reported a production of 79 mg l − 1 of gibberellic acid. In some patents, difficult to verify productions of 1–5 g l − 1 are reported, but we obtained productions of 2.862 g l − 1 gibberellic acid in 8 days. pH, rice flour concentration and C:N ratio were the factors that most explained the production of gibberellic acid; pH being the most important (Table 2). At high pH the acid produced was neutralized, i.e. Le Chaˆtelier principle, and the mould could thus continue to produce acid. The changes in production of gibberellic acid with changes in pH and C:N ratio indicated that the
E.M. Escamilla S. et al. / Journal of Biotechnology 76 (2000) 147–155
151
−1 Fig. 2. Dynamics of () gibberellic acid (mg l − 1), ( ) biomass (g l − 1), ( ) residual glucose (g l − 1) and () NH+ ). 4 N (g l
greatest production was found at low C:N ratio and large pH (Fig. 3). The relation between pH and C:N ratio is important not only for the amount but also for the type of gibberellin produced, i.e. GA4 and GA7, and should thus be controlled during fermentation (Pitel et al., 1971). Initial pH values are often reported, e.g. 5.5, but they were not controlled in those experiments (Borrow et al., 1964; Holme and Zacharias, 1965; Jefferys, 1970). The reason, presumably, why these authors obtained lower production of gibberellic acid than we obtained.
Rice flour concentration was the second most important factor explaining gibberellic acid production but it was difficult to indicate a mechanism that could affect its production. The amount of C and N added with the rice flour is small in comparison to the C added with the glucose and the N added with the NH4Cl so its effect on the C:N ratio or total C and N available is minimal. It might be that rice flour contains a carotenoid or another precursor of gibberellin. It has been reported that meal and flour of vegetal origin favours gibberellin
Table 2 Statistical analysis of different factors used in the optimisation study for the production of gibberellic acid in a fluidized bioreactor Factor
Sum of squares
df
Mean square
F-value
Pr\F
Bioreactor Temperature pH C:N ratio Rice flour Residual df: degree of freedom
0.03 15.95 973.17 52.13 88.24 4.04
1 2 2 2 2 8
0.03 7.97 486.59 26.07 44.12 0.50
0.06 15.79 963.71 51.62 87.39
8.073×10−1 1.668×10−3 2.919×10−10 2.674×10−5 3.671×10−6
152
E.M. Escamilla S. et al. / Journal of Biotechnology 76 (2000) 147–155
and bikaverin production but nothing was mentioned about a possible mechanism (Balan et al., 1970; Bru¨ckner and Blechsmidt, 1991). The response surface of gibberellic acid production showed an optimum concentration for rice flour of 2 g l − 1 and production increased with increased pH (Fig. 4).
C:N ratio was the third most important factor explaining gibberellic acid production. The C:N ratio affects the growth of G. fujikuroi and the production of gibberellic acid in batch fermentation in different ways (Borrow et al., 1964; Jefferys, 1970; Bu’Lock et al., 1974). The initial active growth of the mycelial takes place in a
Fig. 3. The response surface of gibberellic acid production (mg l − 1) to changes in pH and C:N ratios. The equation mentioned gives the relationship between gibberellic acid production and the C:N-to-pH ratio. Production of gibberellic acid can be calculated in function of C:N ratio and pH for the experimental conditions mentioned with a significance at the 0.05 probability level.
Fig. 4. The response surface of gibberellic acid production (mg l − 1) to changes in pH and rice flour concentration (g l − 1). The equation mentioned gives the relationship between gibberellic acid production and the rice flour-to-pH ratio. Production of gibberellic acid can be calculated in function of rice flour concentration and pH for the experimental conditions mentioned with a significance at the 0.05 probability level.
E.M. Escamilla S. et al. / Journal of Biotechnology 76 (2000) 147–155
153
Fig. 5. The response surface of gibberellic acid production (mg l − 1) to changes in C:N ratio and rice flour concentration (g l − 1). The equation mentioned gives the relationship between gibberellic acid production and the rice flour-to-C:N ratio. Production of gibberellic acid can be calculated in function of rice flour concentration and C:N for the experimental conditions mentioned with a significance at the 0.05 probability level.
nitrogen-limited and balanced medium, i.e. initial optimal C:N ratio of 36.8, but the growth rate in the production phase is low in an unbalanced medium. Production of gibberellins starts when N is depleted and continues when sufficient C substrate is available. At the onset of the stationary phase, a correlation can be found between depletion of NH+ 4 N and initiation of gibberellic acid production. The production of gibberellic acid at a concentration of 2 g l − 1 rice flour was greater when the C:N ratios were 36.8 and 122 than for values in between (Fig. 5). The effect of temperature on the production of gibberellic acid was statistical significant (Table 2). The response surface of gibberellic acid production to changes in pH and temperature showed that greatest production was found at 30°C; increasing with increased pH (Fig. 6). The response surface of gibberellic acid production showed that the greatest production was found at 30°C and rice flour concentration of 2 g l − 1 (Fig. 7). The largest production of gibberellic acid could theoretically be obtained with a C:N ratio of 36.8, a rice flour concentration of 2 g l − 1, pH 5 and at 30°C. As this combination was not tested for
(Table 1), the gibberellic acid production was calculated by a predictive equation using the values that would give a maximum production (Wu and Hobbs, 1987): Y =Yt + (A −Yt) +(B+Yt) +(C− Yt) +(D–Yt)
(1)
where Y is the gibberellic acid production estimated, Yt is the experimental mean of gibberellic acid for all experiments, and A (temperature), B (pH), C (C:N ratio), and D (rice flour concentration) are the gibberellic acid production for the concentration of each factor that gave the largest gibberellic acid production, i.e. C:N ratio of 36.8, a rice flour concentration of 2 g l − 1, pH 5 and at 30°C. The equation predicted a gibberellic acid concentration of 3.26 g l − 1. An experiment was done in duplicate to test the prediction of the model and a concentration of 3.90 g l − 1 was obtained, more than 3-times greater then previously reported (e.g. Holme and Zacharias, 1965). Apart from a greater production of gibberellic acid, the pellets with the immobilized mycelia of G. fujikuroi can be re-used up to seven times, each run taking 192 h, before the pellets disintegrates; largely reducing the costs of production.
154
E.M. Escamilla S. et al. / Journal of Biotechnology 76 (2000) 147–155
4. Conclusion The response surface method allowed to find the relationship between pH, C:N ratio, rice flour concentration and temperature that gave the largest production of gibberellic acid and the factor that most affected its produc-
tion, i.e. pH. It was found that the gibberellic acid production in a fluidized bioreactor with immobilized mycelia of G. fujikuroi in Ca-polygalacturonate could reach 3.9 g l − 1, more than 3-times greater than previously reported with the possibility to expand the production to eight cycles.
Fig. 6. The response surface of gibberellic acid production (mg l − 1) to changes in temperature (°C) and pH. The equation mentioned gives the relationship between gibberellic acid production and pH-to-temperature ratio. Production of gibberellic acid can be calculated in function of pH and temperature for the experimental conditions mentioned with a significance at the 0.05 probability level.
Fig. 7. The response surface of gibberellic acid production (mg l − 1) to changes in temperature (°C) and rice flour concentration (g l − 1). The equation mentioned gives the relationship between gibberellic acid production and rice flour-to-temperature ratio. Production of gibberellic acid can be calculated in function of rice flour concentration and temperature for the experimental conditions mentioned with a significance at the 0.05 probability level.
E.M. Escamilla S. et al. / Journal of Biotechnology 76 (2000) 147–155
Acknowledgements The research was funded by Consejo Nacional de Ciencia y Tecnologı´a (CONACyT, grant 0739B9506) and the Consejo del Sistema Nacional de Educacio´n Tecnolo´gica (COSNET 493.96-P). We are gratefully to the CRODE group for technical assistance and Drs F. Thalasso and R. Farrera for comments on an early draft of this paper.
References A.O.A.C., 1990. Official Methods of Analysis. 15th Edition of the Association of Official Chemists, Washington, DC. Balan, J., Fuska, J., Kuhr, Y., Kuhrova´, V., 1970. Bikaverin, an antibiotic from Gibberella fujikuroi against Leishmania brasiliensis. Folia Microbiol. 15, 479–484. Borrow, A., Brown, S., Jefferys, E.G., Kessell, R.J., Lloyd, E.C., Lloyd, P.B., Rothwell, B., Swait, J.C., 1964. The kinetics of metabolism of Gibberella fujikuroi in stirred culture. Can. J. Microbiol. 10, 407–444. Bru¨ckner, B., Blechsmidt, D., 1991. The gibberellins fermentation. Critical Rev. Biotechnol. 11, 161–192. Bu’Lock, J.D., Detroy, R.W., Hostalek, Z., Munim-alShakarchi, A., 1974. Regulation of secondary biosynthesis in Gibberella fujikuroi. Trans. Br. Mycol. Soc. 62, 377–389. Cheetham, P.S.J., 1980. Developments in the immobilization of microbial cells and their applications. In: Wiseman, A. (Ed.), Topics in Enzyme and Fermentation Technology, vol. 4. Ellis Horwood, Chichester, pp. 189–238. Coolbaugh, R.C., 1983. Early stages of gibberellin bisoysnthesis. In: Crozier, A. (Ed.), The Biochemistry and Physiology of Gibberellins. Praeger, New York, pp. 53–98. Fredickson, A.G., Magee, R.D., Tsuchiya, H.M., 1979. Mathematical models for fermentation processes. Adv. Appl. Microbiol. 13, 419. Gutke, R., 1980. Theoretical basis of kinetics growth, metabolism and product formation of micro-organisms. UNEP/UNESCO/ICRO training course, Jena, Science Academy of East Germany, Central Institute for Microbiology and Experimental Therapy (ZIMET), vol. 1, pp. 30– 112.
.
155
Hollmann, D., Switalski, J., Geipel, S., Onken, U., 1995. Extractive fermentation of gibberellic acid by Gibberella fujikuroi. J. Ferment. Bioeng. 79, 594 – 600. Holme, T., Zacharias, B., 1965. Gibberellic acid formation in continuous culture. Biotechnol. Bioeng. 7, 405 – 415. Jefferys, E.G., 1970. The gibberellin fermentation. Adv. Appl. Microbiol. 13, 283 – 316. Karel, S.F., Libicki, S.B., Robertson, C.R., 1985. The immobilization of whole cells: engineering principles. Chem. Engin. Sci. 40, 1321 – 1354. Knorr, D., 1987. Immobilization of microbial and cultured plant cells. In: Schwartzberg, H.G. (Ed.), Food Technology and Biotechnology. Dekker, USA, pp. 287 – 319. Kossen, N.W.F., Oosterhuis, N.M.G., 1985. Modelling and scaling-up of bioreactors in biotechnology. In: Rehm, H.J., Reed, G. (Eds.), Biotechnology Processes, vol. 2. VCH, Wenheim, pp. 571 – 605. Kumar, P.K.R., Lonsane, B.K., 1987. Gibberellic acid by solid state fermentation: consistent and improved yields. Biotechnol. Bioeng. 30, 267 – 271. Kumar, P.K.R., Lonsane, B.K., 1988. Immobilized growing cells of Gibberella fujikuroi P-3 for production of gibberellic acid and pigment in batch and semi-continuous cultures. Appl. Microbiol. Biotechnol. 28, 537 – 542. Lu, Z.X., Xie, Z.C., Kumakura, M., 1995. Production of gibberellic acid in Gibberella fujikuroi adhered onto polymeric fibrous carriers. Process Biochem. 30, 661 – 665. Miller, G.L., 1959. Use of dinitrosalicylic acid reagent for determination of reducing sugars. Anal. Chem. 31, 426 – 428. Montes, M.C., Magan˜a, P.I., 1991. Dehydrogenation of steroids by Arthrobacter simplex immobilized in calcium polygalacturonate beads. J. Ind. Microbiol. 8, 259 – 264. Nielsen, J., Villadsen, J., 1992. Modelling of microbial kinetics. Chem. Eng. Sci. 47, 4225 – 4270. Peace, G.S., 1991. Taguchi Methods: A Hands-On Approach. Addison-Wesley, USA, New York. Pitel, D.W., Vining, L.C., Arsenault, G.P., 1971. Biosynthesis of gibberellins in Gibberella fujikuroi. The sequence after gibberellin A4. Can. J. Biochem. 49, 194 – 200. Rehm, H.J., 1980. Industrielle Mikrobiologie. Springer Verlag, New York. Wu, Y., Hobbs, M.W., 1987. Quality Engineering: Product & Process Design Optimization. American Supplier Institute Inc, Derbon, Michigan, USA. (Ok).
