Kinetics of Gibberella fujikuroi and GA3 production by

solid state fermentation in aired columns bioreactor

Cristina M. M. Machado1, Carlos R. Soccol2 e Bruno O. Oishi2
1

Embrapa Hortalias BR 060 km 09. 70359-970 Braslia DF

Laboratrio de Processos Biotecnolgicos, Departamento de Engenharia Qumica,
Universidade Federal do Paran (UFPR), 81531-970 Curitiba PR

ABSTRACT
In this work the kinetics of G. fujikuroi growth and GA3 production was studied, using an
aired column bioreactor connected to a gas chromatograph, to make the exit gas analysis.
With the respirometric data, a logarithmic correlation between accumulated CO 2 and biomass
production was determined. The solid medium consisted in coffee husk pre-treated with alkali
solution, mixed with cassava bagasse (7:3. dry wt basis), with pH 5.2, moisture of 77%.
Cultivation was carried out in inoculated columns, with forced aeration of 0.24 l air.h-1.(g
substrate)-1. in the first three days, and of 0.72 l air.h-1.(g substrate)-1 in the remaining period.
The m obtained was 0.052 h-1 (between 24 and 48 hours of fermentation). A production of
0.925 g GA3.(kg substrate)-1 was achieved after 6 days of fermentation.
INTRODUCTION
Gibberellins (GAs) are a group of diterpenoid acids that function as plant growth regulators
influencing a range of developmental processes in higher plants. GA 3 is a high-valued plant
growth regulator with various applications in agriculture. Its high price, however, have limited
its use to high-premium crops. The industrial process currently used for the production of GA3
is based on submerged fermentation (SmF) techniques, using Gibberella fujikuroi or
Fusarium moniliforme. In spite of the use of the best process technology, the yield of GA 3 is
low, and it demands a very extensive downstream processing. Thus, it is essential to look
beyond the conventional submerged technique and explore other alternatives to achieve a
more economical process. Lower production costs could lead to a more extensive application
of GA3 in agriculture e the consequent benefits (Jefferys, 1970; Kumar & Lonsane, 1989;
1987).
Solid state fermentation (SSF) has been investigated to increase the yields of GA 3 and also
to decrease production costs (Kumar & Lonsane, 1989, 1987; Bandelier et al., 1997; Tomasini
et al., 1997; Pastrana, et al., 1995; Gelmi et al., 2000). Indeed, recently the authors have made
an Industrial Patent, demonstrating that a mixed substrate of important Brazilian
agroindustrial wastes (coffee husk and cassava bagasse) is an practicable alternative for the
production of GA3 by SSF (Soccol, Machado & Oliveira, 2000).
The SSF technique has shown a number of economic advantages over SmF process in the
production of microbial biomass and metabolites and the valorization of agroindustrial byproducts (Pandey, 1992; Soccol, 1995). SSF systems permit the growth of the microorganisms
on solid material in absence or near absence of free water. Part of the water is tightly absorbed
within the solid matrix (Pandey et al., 2001). General and microbiological aspects of SSF,
operation system conditions and scaling-up strategies have been reviewed in pertinent
literature (Raimbault, 1997; Pandey et al., 2001; Hesseltine, 1972).

The impossibility of separation of biomass from substrate and the heterogeneous

characteristics of SSF processes are the principal difficulties found when kinetics
accomplishment is attained. These facts impede the acquirement of representative samples
and are particularly acute in the case of fungal growth and mycelial production. Undoubtedly,
kinetic process is the procedure that will describe the development of the process. Even with
the difficulties that are encountered in SSF process, kinetic procedure can not be substituted
by goodwill, subjectivism or even by the simple and overall process description. (Pandey et
al., 2001). Although that, there is a lack of works studying the kinetics of gibberellic acid
production by solid state fermentation.
Automatic on-line analysis of CO 2 and O2 in the exit gases coming from SSF reactors allows
real time information on the physiological state of the cultures to be obtained and correlation
with other factors as biomass to be established. It is also useful for the monitoring of diverse
biotransformations and has possible applications to the scale-up of processes. Furthermore,
gas measurements in aerobic cultures allow for the calculation of the respirator activity rate
(r) from the natural logarithm of total production of CO 2. Since this parameter is obtained
from a larger number of data points, a more accurate value of then that calculated from the
diverse methods of biomass estimation may be obtained (Pandey et al., 2001; Pintado et al.,
1998; Saucedo-Castaeda et al., 1994).
The present work was undertaken to study the kinetics of growth and production of GA 3 by
G. fujikuroi, using aerated columns bioreactor connected to a gas chromatograph, to make the
exit gas analysis. With the respirometric data, a logarithmic correlation between accumulated
CO2 and biomass production was determined and the mathematical model were compared
with experimental data, heating to a better knowledge about the behavior of the system for
future works in scaling up, and comparison with submerged fermentation.
MATERIALS AND METHODS
Microorganisms and SSF technique
The strain of Gibberella fujikuroi LPB-06, maintained in potato dextrose agar (PDA) was
inoculated in Czapek-Dox medium in rotatory shaker at 30 C for 4 days for developing the
inoculum (Bandelier et al., 1997).
The solid medium consists in coffee husk pre-treated with alkali solution (KOH 0.25%)
autoclaved in 100 C for 45 minutes, mixed with cassava bagasse (7:3. dry wt basis), with pH
5.2, moisture of 77%, adjusted by adding a mineral salts-acid solution containing (%) 0.03
FeSO4 and 0.01 (NH4)2SO4. sterilized at 121oC for 20 min (Machado et al., 2002).
Cultivation and sampling
Cultivation was carried out in 14 inoculated columns, with forced aeration as described by
Raimbault & Alazard (1980). The water bath was maintained the temperature at 29 C. The air
flow was set at 0.24 l air.(h.g DM) -1 in the first 72 h, and 0.72 l air.(h.g DM) -1 after the 3rd day
of fermentation. Every 24 hours two columns were taken of the system, for analysis.
On-line gas monitoring system measured the flowrate and gas composition of the outlet air
stream from four columns. This columns were connected to a gas chromatograph linked to a
personal computer, as described by Saucedo-Castaeda et al, 1994. The respirometric data
were processed en Excel , in witch instantaneous curves of CO 2 and O2 calculations were
determined using the equations developed by Pandey et al., 2001
Analytical Techniques
The material from each column was separated in two parts. The first one was extracted with
phosphate buffer pH 8.0, and analyzed for GA3 with spectrophotometric method of Holbrook
et al., 1961.

In the other part were determined moisture content and loss of weight in the substrate by
gravimetric method. Reducing sugars and total sugars (after hydrolysis with H 2SO4) were
determined by the Somogyi-Nelson spectrophotometric method (Nelson, 1944; Somogyi,
1945). Non-reducing sugars were estimated by difference. Proteic nitrogen was precipitated
by Stutzer method (Vervack, 1973) followed by Kjedahl determination, assuming that protein
content was equal to 6.25 times the N content (IAL, 1985). Biomass was calculated as twice
the protein, assuming that 50% of fungal biomass is protein. In the biomass calculation,
adjustment was made for protein in the substrate.
Correlation between CO2 and biomass produced
To determine the correlation between biomass and CO2, the logistic law (Ebune et al., 1995)
was applied to describe the biomass production. This model (equations 1 and 2) describes the
mycelial growth in logarithmic and stationary phases in solid-culture of substrate.

dX
X

X 1
dt
X m

(1)

Xm
Xm
1

(2)
mt
X0
1 e
Where X is the biomass (g.(kg DM)-1) at time t (h); Xm the maximum biomass concentration,
m the maximum specific growth rate (h -1) and X0 is the initial biomass concentration. The
quality of the mathematical model proposed were availed by the regression coefficient (R 2),
and by the comparison between the values of Xm, X0 e m, obtained experimentally and by the
model.
Then, using this model, a theoretical value of biomass was determined for every time that
had a CO2 analysis. Finally, a regression of the curve accumulated CO 2 vs. biomass was made,
leading to a logarithmic correlation between these two variables:
X yLn(CO2 acum ) z
(3)
X

Kinetic calculations
The kinetic parameters were determined according the following equations:
a) specific growth rate ():

X2

X1

dX

t2

t1

dt

(4)

Where X is the biomass (g.(kg DM)-1) at time t (h); and the specific growth rate (h-1). The
maximum specific growth rate (m) is calculated by the same equation (4) considering the
exponential growth phase. The global specific growth rate ( ) is the medium of daily
specific growth rates.
b) Rate of product formation (rP)
P P1
rP 2
(5)
t 2 t1
Where P is the GA3 concentration (g.(kg DM)-1) at time t (h), and rP the rate of GA3 production
(g.(kg DM.h)-1). The maximum rate of product formation (rPmax) is calculated by the same
equation (5), in linear period of bigger GA 3 production and global metabolite production rP
is given by the medium of daily rP values.

RESULTS AND DISCUSSION

Experimental data obtained
A tendency of increasing of moisture and water activity was observed during the
fermentation. This behavior is explained by the saturated air aeration made during the process,
heading a significant mass transfer while the experiment. In erlenmeyer flasks without
aeration, with the same substrate (Machado, 2001), the moisture decreased gradually. A small
change in pH happened, oscillating between 5.2 and 5.6. Probably this variation is due to the
difficulties in acquiring representative samples for this analysis.
The loss of weight in the substrate is an important factor in SSF, once its variation can be
considered an indirect measure of microbial growth. In this experiment, loss of weight
reached 9% at 7th day of fermentation and followed the Monod model: a lag phase of 48 h,
exponential phase 48-120 h and stabilization until 144 h.
Table 1 was made considering data for kinetic study and correlation of CO 2 and biomass.
Although it is a complex substrate, observing sugars behavior, the main carbon source for this
fermentation was considered the non-reducing sugars, which became mainly from cassava
bagasse starch. Therefore, in Table 1 these were the values considered in the column S.
TABLE 1: Kinetic parameters in SSF for gibberellic acid production
Time
X
S
P
rp

(h)
g.(kg DM)-1 g.(kg DM)-1 g.(kg DM)-1 (h-1) g.(kg DM.h)-1
0
1.122
76.665
0.000
24
1.677
64.226
0.153
0.0167
0.006
48
5.872
38.742
0.206
0.0522
0.002
72
10.791
37.504
0.315
0.0254
0.005
96
15.764
36.886
0.417
0.0158
0.004
120
19.276
36.443
0.711
0.0084
0.012
144
21.743
35.519
0.925
0.0050
0.009
168
24.185
28.664
0.740
0.0044
-0.008
Where: DM dry matter (dry fermented substrate); X biomass; S substrate (non-reducing sugars);
P product (gibberellic acid); especific growth rate; rp GA3 production per hour

Correlation of biomass CO2 production

In SSF processes the amount of biomass produced in one moment of fermentation can be
related to the O2 consume pattern, if the fraction employed in metabolites synthesis is known
or irrelevant. As the GA3 synthesis has several oxidative stages, and the substrate used in this
work is a mixture of carbohydrates, the mathematical modeling of the process is extremely
complex. Otherwise, once CO2 is only produced by microbial respiration, this parameter
seemed to be more directly related to biomass.
Hence, according to previously described, the Logistical law of microbial growth (2) was
applied to experimental data for biomass, and the following constants were determined: Xm =
24.38; X0 = 1.171; m = 0.037. The high regression coefficient (0.996) obtained and the good
correspondence with the experimental values in Table 1 (Xm = 24.18 g.(kg MS)-1; X0 = 1.122
g.(kg MS)-1; max = 0.0522 h-1) proves the adequacy of the model. Then, a logarithmic
correlation between accumulated CO2 and biomass were determined, leading to the following
equation (R2 = 0.879):
X estim 3,9872 ln(CO2 acum ) 23,824

(6)

Kinetics of fermentation
The kinetic parameters of the microorganism growth were established with the values of
experimental and estimated biomass. As shown in Figure 1, the two growth curves are very
similar, and follow a Monod model having the same lag phase (0 to 24 hours), and
exponential growth phase (24 to 48 hours). The desacelerated growth phase started in both
cases at 48 hours but the stationary phase appeared in experimental data after 120 hours and
wasnt found in estimated data.
The maximum specific growth rates (max) were determinate according to equation 4,
between 24 and 48 hours (exponential growth phase), and where almost the same for
experimental and estimated biomass (0.0522 h-1 and 0.0520 h-1 respectively). For the global
specific growth rate ( ), calculated between 0 and 144 h, the same value of 0.0206 h -1 was
found.
3,0
2,5

ln X

2,0
1,5
1,0
0,5

experimental

estimated

0,0
0

24

48

72

96

120

144

Time(h)

FIGURE 1: Phases of experimental and estimated biomass evolution

The main product of this fermentation is gibberellic acid. After 6 days a concentration of
0.925 g GA3.(kg DM)-1 was achieved. In Table 1 can be observed that the better production of
the metabolite was between 96 and 120 h of fermentation, just after the exponential growth
phase. This behavior is characteristic for secondary metabolites kinetics as gibberellic acid.
The decaying of GA3 in substrate after some time of fermentation, was observed by other
authors working with GA3 production in SSF using different substrates and bioreactors
(Tomasini et al., 1997; Bandelier et al., 1997; Machado et al., 2001). This phenomenon is
probably due for GA3 decomposition in productive system, once the metabolite is extremely
unstable in aqueous solutions (Perrez et al., 1996; Gelmi et al., 2002).
The kinetic parameters related with metabolite production, where calculated according to
equation 6. The maximum GA3 production rate was determined between 96 and 120 h: rPmax =
0.012 g.(kg MS.h)-1 and rP = 0.0063 g.(kg MS.h)-1 .
Another parameter that can be used for defining the studied fermentation is the yield based
on product formation (YP/X), defined as the amount of formed product per unity of biomass
(equation 7). By this criteria, between 0 and 144 h, YP/X = 44.8 mg GA3.(g biomass)-1.
YP
X

P
X

(7)

In air forced columns bioreator, Gelmi et al. (2000) reached 0.760 g GA3.(kg MS)-1 using an
inert support of Amberlite with nutrient solution, after 5 days of fermentation. The m was

0.046 h-1 between 0 and 48 h. The YP/X was 65 mg GA3.(g biomass)-1 , value 40% bigger than
obtained in this work. This confirms the information given by the authors that the strain used
(G. fujikuroi ATCC 12616) is hyperproducer of GA 3. Otherwise, there were a low biomass
production and as a result of product. Tomasini et al. (1997) obtained 0.24 g GA3.(kg DM)-1
after 3 days of fermentation using wheat flour as substrate. The m obtained was of 0.033 h-1
between 0 and 48 h., with a YP/X of 50 mg GA3.(g biomass)-1 revealing a strain with an
analogous productivity then LPB-06, but a less efficient system then the studied in this work.
CONCLUSIONS
The aim of this work was to determine some important kinetic parameters of this kind of
fermentation, taking it as a model for future works in the GA 3 production by solid state
fermentation. Moreover it was proved the efficiency of the substrates studied using a aerated
columns system, confirming the results obtained by the authors in erlenmeyer flasks (Soccol
et al., 2001; Machado et al., 2001). Indeed, considering the same system, the concentration
achieved is the best found, having the advantage of the utilization of agroindustrial residues as
substrates.
ACKNOWLEDGEMENTS
We thank European Union (project INCO DC: NCT970185), PN&D-CAF -coordinator
EMBRAPA, Brazil (project n o 07.199.079.01 and 07.199.079.02) for the financial assistance.
C. R. Soccol thanks CNPq for Scientific Productivity scholarship, and C. M. M. Machado
thanks CAPES for PhD scholarship.
REFERENCES
Bandelier, S., Renaud, R. & Durand, A. (1997), Production of gibberellic acid by fed-batch solid state
fermentation in as aseptic pilot-scale reactor. Process Biochemistry, v. 32 n.2, p.141-145.
Ebune, A.; Al-Asheh, S. & Duvnjak, Z. (1995), Effects on phosphate, surfactants and glucose on phytase
prodction and hydrolysis of phytic acid in canola meal by Aspergillus ficuum during solid-satate fermention.
Bioresource Technology, v. 54 , p. 241-247.
Gelmi, C.; Perez-Correa, R. & Agosin, E. (2002), Modeling Gibberella fujikuroi growth and GA3 production in
solid-state fermentation. Process Biochemistry, v. 37, p.1033-1040.
Gelmi, C.; Perez-Correa, R.; Gonzalez, M. & Agosin, E. (2000), Solid substrate cultivation of Gibberella
fujikuroi on an inert support. Process Biochemistry, v. 35, p. 1227-1233.
Hesseltine, C. W. Solid State Fermentation (1972), Biotechnology and Bioengineering, v. 14, p. 517-532.
Holbrook, A. A.; Edge, W. J. W.; Baily , F. (1961), Spectrophotometric method for determination of gibberellic
acid. Advances in Chemical Series, v. 28, p. 159-167.
IAL Institulo Adolfo Lutz (1985), Normas Analticas do Instituto Adolfo Lutz, So Paulo, Brasil.
Jefferys, E. G. (1970), The gibberellin fermentation. Advances in Applied Microbiology, v.13, p.283-316.
Kumar, P. K. R. & Lonsane, B. K. (1987), Gibberellic acid by solid state fermentation: consistent and
improvement yields. Biotechnology and Bioengineering, v.30, p.267-271.
Kumar, P. K. R. & Lonsane, B. K. (1989), Microbial Production of Gibberellins: state of the art. Advances in
Applied Microbiology, v.34, p.29-139.

Machado, C. M. M.; Oliveira, B. H.; Pandey, A.; Soccol, C. R., (2001), Coffee husk as substrate for the
production of gibberellic acid by fermentation. Chapter 37 In: Sera, T., Soccol, C. R., Pandey, A., Roussos, S.
Book on Coffee Biotechnology and Quality, Kluwer Academic Publishers.
Machado, C. M. M.; Soccol, C. R.; Pandey, A. (2002), Gibberellic acid production by solid sate fermentation in
coffee husk. Applied Biochemistry and Biotechnology v.102, n.103, p.179-192.
Nelson, N. A. (1944) A photometric adaptation of the Somogyi method for determination of glucose. Journal of
Biological Chemistry, v.153, p.375-380.
Pandey, A., (1992), Recent developments in solid-state fermentation. Process Biochemistry, v.27, p.109-117.
Pandey, A.; Soccol, C. R.; Rodriguez-Leon, J. A; Nigham, P. (2001) Solid State Fermentation in Biotechnology
Fundamentals and Applications, Asistech Publishers Inc, England.
Pastrana, L. M.; Gonzlez, M. P.; Torrado, A. & Murado, M. A. (1995) Interactions affecting gibberellic acid
production in solid-state culture: a factorial study. Enzyme and Microbial Technology, v. 17, p. 784-790.
Perrez, F. J.; Vechiolla, A.; Pinto, M. & Agosin, E. (1996), Gibberellic acid decomposition and its loss of
biological activity in aqueous solutions. Phytochemistry, v. 41, n.3, p.675-679.
Pintado, J.; Lonsane, B. K.; Gaime-Perraud, I & Roussos, S. (1998), On-line monitoring of citric acid production
in solid-state culture by respirometry. Process Biochemistry, v.33, n.5, p.513-518.
Raimbault, M. (1997) General and microbiological aspects of solid-substrate fermentation.
Training Course in Solid State Fermentation, Curitiba, Brasil.

International

Raimbault, M.; Alazard, D. (1980), Culture Method to Study Fungal Growth in Solid Fermentation, European
Journal of Applied Microbiology and Biotechnology, v. 9, p. 199-209.
Saucedo-Castaeda, G. Trejo-Hernndez, M. R.; Lonsane, B. K.; Navarro, J. M.; Roussos, S.; Dufour, D. &
Raimbault, M. (1994), On-line automated monitoring and control systems for CO 2 and O2 in aerobic and
anaerobic solid-state fermentations. Process Biochemistry, v.29, n.1, p.13-24.
Soccol, C. R. (1995), Aplicaes da fermentao no estado slido na valorizao de resduos agroindustriais.
Flach Agricultura, v. 4 n. 2, p. 3-4.
Soccol, C. R., Machado, C. M. M., Oliveira, . H. (2000), Produo de cido giberlico por fermentao no
estado slido em substrato misto. Patente Industrial n 525-8 , Brasil.
Somogyi, M. (1945), A new reagent for the determination of sugars. Journal of Biological Chemistry, v. 160, p.
61-68.
Tomasini, A.; Fajardo, C. & Barrios-Gonzles, J. (1997), Gibberellic acid production using different solid-state
fermentation systems. Word Journal of Microbiology and Biotechnology v.13, p.203-206.
Vervack, W. (1973), Analyse des Aliments methodes courants danalyse. Louvain la Neuve: Laboratoire de
Biochemie de la Nutricion, UCL, France.