Alternative media for lactic acid production by lactobacillus delbrueckii NRRL b 445

0960–3085/03/$23.50+0.00
# Institution of Chemical Engineers
www.ingentaselect.com
=
titles
=
09603085.htm Trans IChemE, Vol 81, Part C, September 2003
ALTERNATIVE MEDIA FOR LACTIC ACID PRODUCTION
BY LACTOBACILLUS DELBRUECKII NRRL B-445
S. J. TE
´
LLEZ-LUIS
1,2
, A. B. MOLDES
2
, M. VA
´
ZQUEZ
1 ,3
and J. L. ALONSO
2
1
Department of Food Science and Technology, UAM Reynosa-Aztla´n, Universidad Autonoma de Tamaulipas, Me
´
xico
2
Department of Chemical Engineering, University of Vigo (Facultad de Ourense), Vigo, Spain
3
A
´
rea de Tecnolog‡´a de los Alimentos, Escuela Polite´cnica Superior, Departamento de Qu‡´mica Anal‡´tica, Universidad

de Santiago de Compostela, Lugo, Spain
L
actic acid bacteria are generally recognized as nutritionally fastidious. The complexity
of the media increases the cost of lactic acid production. In this study a low-cost
nutrient medium based on corn steep liq uor (CSL) was developed for lactic acid pro-
duction by
Lactobacillus delbrueckii
NRRL B-445. Starting from a medium containing
90 g l
¡
1
glucose and 20 g l
¡
1
CSL as a sole nutrient source, 70.7 g l
¡
1
lactic acid was obtained
with an economic ef ciency of 98 g lactic acid per
„
nutrient. Other media, made with CSL
and each individual component (yeast extract, peptone, sodiu m acetate, sodium citrate,
K
2
HPO
4
, MgSO
4
¢
7H

2
O, MnSO
4
¢
H
2
O or FeSO
4
¢
7H
2
O) of a general (Mercier) lactobacilli
medium were also assayed. The highest economical ef ciency (134 g lactic acid per
„
nutrient)
was obtained supplementing 10 g l
¡
1
CSL with 0.05 g l
¡
1
FeSO
4
¢
7H
2
O. Additionally, lactic
acid production and glucose consumption were mathematically modelled and the regression
parameters obtained were correlated with CSL concentration by linear or exponential
equations.

Keywords: Lactobacillus delbrueckii; lactic acid, corn steep liquor, n utrients study;
mathematical modelling.
INTRODUCTION
Many applications in dairy, beverage, confectionery, meat
and poultry industries have been found for lactic acid
(2-hydroxy propionic acid) and its derivates. Lactic acid is
employed in food industry as acid ulant,  avour and preser-
vative due to its mild taste that does not hide the weaker
aromatic  avours of some foods. Additionally, it has a
technological application during cheese and yoghurt produc-
tion, producing the coagulation of the casein fraction.
An other important appli cation of this compound is the
production of polylactic acid (PLA)-based degradable plas-
tics (Chahal, 1991; Ozen and Ozilgen, 1992).
Lactic acid can be obtained by chemical synthesis from
petroleum-based products or by microbial fermentation.
Many lactobaci lli strains a nd some fungus like
Rhizopus
oryzae
can bioconvert glucose and other sugars to lactic acid
(Zhou
et al.
, 1999; Hofvendahl and Hahn-Hagerdal, 2000).
Owing to its asymmetric c arbon, lactic acid can occur in two
optically active enantiomers, L and D (Vick-Roy, 1985), but
an important advantage of microbial fermentation over
chemical synthesis is that it is possible to produce exclu-
sively one of the isomers.
Lactic acid bacteria u se sugars via different pathways
resulting in homo-, hetero- o r mixed acid fermentations.

Homofermentation gives only lactic ac id as the end product
of glucose metabolism by the Embden–Meyerhof–Parnas
pathway. In heterofermentation, equimolar amounts of lactic
acid, carbon dioxide and ethanol or acetate are formed from
glucose via the phosphoke tolase pathway (Chahal, 1991).
Several factors that affect lactic acid production by micro-
organisms are medium composition (carbohydrate source,
sugar concentration and growth factors), temperature,
presence of oxygen, pH and product concentration
(Burgos-Rubio
et al.
, 2000).
Lactobacillus delbrueckii
NRRL B-445, also named as
Lactobacillus rhamnosus
ATCC 10863 (Hofvendahl and Hahn-Hagerdal, 2000), is a
homofermentative lactic acid b acterium that produces
mainly L-lactic acid.
In spite of the advantages, fermentations must be cost
competitive with chemical synthesis. Fermentation medium
can represent almost 30% of the cost for a microbial
fermentation (Miller and Churchill, 1986). Lactic acid
bacteria have limited capacity to synthesize B-vitamins
and amino acids (Hofvendahl and Hahn-Hagerdal, 2000).
Yeast extract is used as the main source of nitrogen and
vitamins for lactic acid production by microorganisms, but it
is too expensive for larg e-scale fermentations. Complex
media commonly employed for growth of lactic acid
bacteria are not economically attractive due to their high
amount of expensive nutrients such as yeast extract, peptone

and salts (Mercier
et al.
, 1992). Various nitrogen sources
were tested for lactic acid production by bacteria but none
of these gave lactic acid concent rations as high as that
obtained with yeast extract (Nancib
et al.
, 2001; Te´llez-
Luis
et al.
, 2003). However, new low-cost media for lactic
acid fermentation are desirable in order to decrease the
production cost.
250
Corn steep liquor (CSL) is a low-cost nutritional medium
employed successfully in the production of ethan ol by
Zymomonas mobilis
(Kadam and Newman, 1997; Silveira
et al.
, 2001), succinic acid by
Anaerobiospirillum succini-
ciproducens
(Lee
et al.
, 2000) or arabinanase by
Fusarium
oxysporum
Cheilas
et al.
, 2000). It could replace some of

the expensive nutrients in the complex medium employed to
grow
L. delbrueckii
.
The aim of t his study was to develop a low-cost nutrient
medium based on CSL for lactic acid production by
L. delbrueckii
NRRL B-445. Additionally, lactic acid produc-
tion and glucose consumption were modelled and the regres-
sion parameters obtained were correlated with CSL
concentration.
MATERIALS AND METHODS
Strains and Culture Conditions
L. delbrueckii
NRRL B-445 was obtain ed from the
United States Department of Agriculture Northern National
Research Laboratory in Peoria, IL. The strain was grown on
plates using the complete media proposed by Mercier
et al.
(1992), which contains 20 g l
¡
1
glucose, 5 g l
¡
1
yeast
extract, 10 g l
¡
1
peptone, 5 g l

¡
1
sodium acetate, 2 g l
¡
1
sodium citrate, 2 g l
¡
1
K
2
HPO
4
, 0.58 g l
¡
1
MgSO
4
¢
7H
2
O,
0.12 g l
¡
1
MnSO
4
¢
H
2
O, 0.05 g l

¡
1
FeSO
4
¢
7H
2
O and 10 g l
¡
1
agar at 37
¯
C for 24 h. Inocula were prepared by washing
cells from plates with 5 ml sterile water. Biomass in the
inocula was measured by optical density at 600 nm and
adjusted to equivalent values by dilution with water to
obtain 6 g l
¡
1
dry cells. Inocula were 5 ml. The experiments
were carried out in 250 ml Erlenmeyer  asks with a  nal
volume of 100 ml usi ng different media. The content of
nitrogen in the yeast extract, peptone and CSL used was
11.9, 12 and 13% in dry basis, respectively. The content of
water in the yeast extract, peptone and CSL used was 3.5, 4
and 50%, respectively.
Calcium carbonate (10 g) was added previously to inocu-
lation. After inoculation, fermentations were carried out in
orbital shakers at 41.5
¯

C and 200 rpm for 96–98 h. The pH
was kept constan t around 6 due to the lactic acid formed was
neutralized by the present of calcium carbonate. Samples
(2 ml) were taken at random time in tervals and centrifuged
at 16,000
g
for 3 min. The supernatants were used immedi-
ately for various analyses.
Analytical Methods
Glucose, lactic acid and ac etic acid were determined by
high performance liquid chromatography (HPLC) using a
Transgenomic ION-300 column and an isocratic elution
with a  ow rate of 0.4 ml min
¡
1
. The mobile phase was
0.0025 M H
2
SO
4
. The oven temperature was 65
¯
C and a
refractive index (RI) detector was used.
Statistical Analysis
All experimental data were obtained in triplicate and
mean values are g iven. Linear and non-linear regression
analyses of experimental data were performed usi ng
commercial software (Microsoft Excel 2000, Microsoft
Corporation, Redmond, WA, USA).

RESULTS AND DISCUSSION
The cost of nutrients is an important aspect in the
fermentation of glucose to lactic acid by
L. delbrueckii
.
General lactobacilli media such as Mercier medium and
MRS medium are very complex with many expensive
nutrients. Table 1 shows the components of the Mercier
medium as well as the cost of each nutrient. The price of
CSL is also included in the same table. As it can be noted,
the price of CSL is two to three times lower than the price of
yeast extract and peptone, respectively.
Comparative Study Between Mercier Medium
and CSL-based Media
Media containing different concentrations of CSL (1, 3, 5,
10 or 20 g l
¡
1
) were tested in the fermentation of 90 g l
¡
1
glucose to lactic acid. For comparative purposes a batch
fermentation using the Mercier medium with 90 g l
¡
1
glucose was also performed.
Figure 1 shows the results of the batch experiments
for lactic acid and glucose concentrations. The highest
lactic acid concentration was obtained using the Mercier
medium (76.2 g l

¡
1
). However, a similar pattern was shown
by the fermentation with 20 g l
¡
1
CSL, obtaining 70.7 g l
¡
1
lactic acid at the end of the fermentation. Using lower
concentrations of CSL, lower lactic acid concentrations
and volumetric productivities were obtained. These facts
suggest that CSL at concentration lower than 20 g l
¡
1
did
not supply the require d nutrients for the metabolism of
L. delbrueckii
. Additionally, acetic acid was quanti ed
and negligible concentrations were obtained (data not
shown). This was important because it demonstrated that
L. delbrueckii
maintains the homofermentative pathway in
the presence of CSL.
In the experiment using the Mercier medium, the glucose
concentration was completely consumed at the end of the
fermentation (Figure 1b). However, a  nal glucose concen-
tration of 12.0 g l
¡
1

was observed in the experiments carried
out with medium con taining 20 g l
¡
1
CSL. Using lower
concentrations of CSL, higher concentrations of residual
glucose were obtained. This suggested that CSL is limited in
some nutrients.
Table 2 shows numerical values of lactic acid concentra-
tion, product yield (
Y
p
=
s
), product ef ciency (
E
p
=
s
) and
economic ef ciency (
E
p
=
„
) after 98 h of fermentation.
Product yield was de ned as grams lactic acid produced
per gram glucose consumed, product ef ciency as g rams
lactic acid produced per gram initial glucose and economic
Table 1. Prices of nu trients used in

experiments.
Nutrient Cost (
„
=
kg)
Corn steep liquor 36.06
Yeast extract 76.74
Peptone 112.27
Sodium acetate 13.94
Sodium citrate 20.73
K
2
HPO
4
30.29
MgSO
4
10.58
MnSO
4
15.03
FeSO
4
11.54
ALTERNATIVE MEDIA FOR LACTIC ACID PRODUCTION 251
Trans IChemE, Vol 81, Part C, September 2003
ef ciency as grams lactic acid produced per cost unit of
nutrients (
„
). As it can be observed,

E
p
=
s
decreased with the
decrease of CSL concentration. However,
Y
p
=
s
was main-
tained around 0.9 g g
¡
1
using the Mercier medium or media
containing 20 or 10 g l
¡
1
CSL.
It i s interesting to select the cheapest media that allow the
highest lactic acid concentra tion to be obtained. The
E
p
=
„
is
an adequate parameter to compare media from an econom-
ical poin t of view. The parameter
E
p

=
„
showed that it was
more pro t able to use a medium with 20 g l
¡
1
CSL than the
Mercier medium because using 20 g l
¡
1
CSL, 98 g of lactic
acid were produced per euro of n utrients while only 50 g
lactic acid were obtained per euro when the Mercier medium
was employed.
Mathematical Modelling of Fermentation with
CSL as a Sole Nutritional Source
Lactic acid production and glucose consumption were
mathematically modelled and the regression parameters
obtained were correlated with CSL concentration by linear
or exponential equati ons. A mathematical model was
adopted from another study to describe the fermentative
production of lactic acid (Mercier
et al.
, 1992):
d
P
d
t
ˆ P
r

P
1
¡
P
P
m
 ´
(1)
where
t
is time,
P
is lactic acid concentration,
P
m
is
maximum concentration of lactic acid, and
P
r
is the ratio
between the initial volumetric rate of produ ct formation (
r
p
)
and the initial product concentration
P
0
. Equation (1) can be
directly solved to give the following expression:
P ˆ

P
0
P
m
e
P
r
t
P
m
¡ P
0
‡ P
0
e
P
r
t
(2)
From the series of experimental data for lactic acid
concentration during fermentation, the model parameters
P
0
,
P
m
and
P
r
can b e calculated for each fermentation

medium by non-linear regression using the least-squares
method. Table 3 shows the kinetic and statistical parameters.
Figure 1a shows the experimental and predicted data for
these batches. The coef cient
r
2
showed a good agreement
between experimental and predicted data. The value of the
F
-test probability showed that the model for 1 g l
¡
1
CSL
medium is the least accurate due to the low value of lactic
acid concentration obtained for this medium.
The models predict maximum lactic acid concentrations of
74.9 g l
¡
1
for the Mercier medium, 66.6 g l
¡
1
for 20 g CSL
l
¡
1
medium and 25.9 g l
¡
1
for 10 g CSL l

¡
1
medium at 98 h.
The regression parameters obtained for each experiment were
Figure 1. Experimental and calculated dependence of lactic acid and
glucose concentrations on the fermentation time correspon ding to fermen-
tations of 90 g l
¡
1
glucose with different concentrations of corn steep liquor
and the Mercier medium.
Table 2. Results for the lactic acid pro duction by Lactobacillus delbrueckii
using different concentrations of CSL and the Mercier medium. All
fermentations were with 90 g l
¡
1
glucose.
Medium
Lactic acid
concentration
(g l
¡
1
)
Y
p
=
s
(g g
¡

1
)
E
p
=
s
(g g
¡
1
)
E
p
=
„
(g
„
¡
1
)
Mercier medium 76.20 0.85 0.85 50
20 g l
¡
1
CSL 70.73 0.91 0.79 98
10 g l
¡
1
CSL 27.13 0.90 0.30 75
5 g l
¡

1
CSL 15.59 0.66 0.17 86
3 g l
¡
1
CSL 8.55 0.42 0.10 79
1 g l
¡
1
CSL 2.57 0.20 0.03 71
Table 3. Results obtained by regression analysis of lactic acid p roduction and glucose consumption by Lactobacillus delbrueckii u sing different media (all
media include 90 g l
¡
1
glucose).
Lactic acid production Glucose consumption
Medium P
0
(g l
¡
1
) P
m
(g l
¡
1
) P
r
(h
¡

1
) r
2
F-test of
probability Y
p
=
s
(g g
¡
1
) r
2
F-test of
probability
Mercier medium 2.81 74.97 0.181 0.9982 0.9759 0.91 0.9966 0.9931
20 g l
¡
1
CSL 4.82 66.66 0.091 0.9783 0.9041 0.91 0.9935 0.9933
10 g l
¡
1
CSL 1.69 25.91 0.137 0.9851 0.9311 0.86 0.9834 0.9767
5 g l
¡
1
CSL 0.99 15.38 0.166 0.9941 0.9559 0.63 0.9883 0.9782
3 g l
¡

1
CSL 0.67 8.88 0.199 0.9892 0.9432 0.46 0.9808 0.9648
1 g l
¡
1
CSL 0.28 2.73 0.412 0.9856 0.8505 0.21 0.9536 0.9775
Trans IChemE, Vol 81, Part C, September 2003
252 TE
´
LLEZ-LUIS
et al.
correlated with the CSL concentration by mean of empirical
equations.
P
0
,
P
m
and
P
r
were related to the CSL concentra-
tion given by equations (3 )–(5), respectively:
P
0
ˆ
0:2371
C
csl
¡

0:1555 (3)
P
m
ˆ
3:3189
C
csl
¡
1:9735 (4)
P
r
ˆ
0:3788
¢ C
¡
0:4733
csl
(5)
The coef cient
r
2
(0.9710) for
P
0
, (0.9858) for
P
m
and
(0.9749) for
P

r
con rmed that the empirical equations  t the
data well. By combining equations (3)–(5) with the model of
equation (2), it is possible to predict the lactic acid concen-
tration at any time for CSL concentrations and time in the
range studied (0–98 h and 1–20 g l
¡
1
CSL). Figure 2 shows
how the generalized model predicts the dependence of lactic
acid concentration on different CSL concentra tions and time
using the model p arameters. This kind of surface response
allows the selection o f different conditions in order to
achieve the same results.
The consumptionof glucose by
L. delb rueckii
can be given
by the following equation (obtained from the
Y
p
=
s
de nition):
S ˆ S
0
¡
1
Y
p
=

s
(
P ¡ P
0
) (6)
where
Y
p
=
s
,
P
and
P
0
were de ned above,
S
is the glucose
concentration (g l
¡
1
) and
S
0
is the initial glucose concentra-
tion. Using the series of experimental data concernin g
glucose concentration
=
time and the regression parameters
of equation (2), the model parameter

Y
s
=
p
can b e calculated
for each fermentation medium by non-linear regression
using the least-squares method. Table 3 lists the numerical
values of
Y
p
=
s
and statistical parameters obtained for the
glucose consumption and Figure 1b shows the experimental
and predicted data for these fermentations. The parameter
Y
p
=
s
varied with
C
csl
according to the following equation:
1
Y
p
=
s
ˆ
4:163

¢ C
0:5072
csl
(7)
The statistical parameter
r
2
for the empirical equation (7)
was signi cant (0.9645). Combining equation (5) with
equation (6), a generalized model for predicting glucose
consumption in CSL media was also developed. Figure 3
shows the prediction of the generalized model for the
dependence of glucose concentration with the CSL concen-
tration and time. The model predicts that more than 45 g l
¡
1
glucose remaine d in the medium when less than 10 g l
¡
1
CSL
is used. Both models would be very useful for op timization.
Fermentation of Supplemented CSL Media
In order to increase the lactic acid production and the
economic ef ciency, experiments were conducted using
10 g l
¡
1
CSL supplemented with one component of the
Figure 2. Prediction of the generalized model for the dependence of lactic
acid concentration on the corn steep liquor concentration and time.

Figure 3. Prediction of the generalized model for the dependence of glucose
concentration on the corn steep liquo r concentration and time.
Table 4. Results for the lactic acid production by Lactobacillus delbrueckii using CSL supplemented with other nutrients from the
Mercier medium.
Medium
Lactic acid
concentration (g l
¡
1
) Y
p
=
s
(g g
¡
1
) E
p
=
s
(g g
¡
1
)
E
p
=
„
(g lactic acid
per

„
nutrients)
10 g l
¡
1
CSL 27.13 0.90 0.30 75
10 g l
¡
1
CSL ‡ 10 g l
¡
1
peptone 76.71 0.85 0.85 52
10 g l
¡
1
CSL ‡ 5 g l
¡
1
yeast extract 77.64 0.89 0.86 104
10 g l
¡
1
CSL ‡ 2 g l
¡
1
citrate 41.30 0.72 0.46 103
10 g l
¡
1

CSL ‡ 5 g l
¡
1
acetate 40.04 0.77 0.44 111
10 g l
¡
1
CSL ‡ 2 g l
¡
1
K
2
HPO
4
29.10 0.66 0.32 69
10 g l
¡
1
CSL ‡ 0.58g l
¡
1
MgSO
4
25.49 0.84 0.27 67
10 g l
¡
1
CSL ‡ 0.12g l
¡
1

MnSO
4
37.35 0.94 0.41 103
10 g l
¡
1
CSL ‡ 0.05g l
¡
1
FeSO
4
48.53 0.85 0.54 134
Trans IChemE, Vol 81, Part C, September 2003
ALTERNATIVE MEDIA FOR LACTIC ACID PRODUCTION 253
Mercier medium at a time, in the same concentrations.
Table 4 lists the medium used and results for lactic acid
concentration obtained and
Y
p
=
s
,
E
p
=
s
and
E
p
=

„
calculated at
98 h of fermentation for media of supplemented CSL.
Figure 4 shows the experimental and calculated results
from the fermentation of 10 g l
¡
1
CSL alone as control an d
10 g l
¡
1
CSL supplemented with peptone or yeast extract
(the main nutritional components of the Mercier medium).
Both exhibited a signi cant effect. The highes t lactic acid
concentration (77.64 g l
¡
1
) was obtained by supplementing
10 g l
¡
1
CSL with 5 g l
¡
1
yeast extract. Similar lactic acid
concentration (76.71 g l
¡
1
) was also obtained by supple-
menting with 10 g l

¡
1
peptone. These values compare very
well with those achieved with th e Mercier medium and the
20 g l
¡
1
CSL medium. The
E
p
=
s
was also higher in the above
two cases. Although the glucose consumed was different,
the
Y
p
=
s
was similar or slight lower than that without
supplementation. The values of
Y
p
=
s
compare well with
those reported using other microorganisms like
Rhizopus
oryzae
(Zhou

et al.
, 1999). The
E
p
=
„
was decreased by
30.66% using the medium supplemented with peptone and
increased by 38.66% when using the medium supplemented
with yeast extract (Table 4). The
E
p
=
„
values showed that it
is most economically interesting to supplement CSL with
yeast extract than with peptone. The importan ce of yeast
extract in the preculture media is stressed (Amrane and
Prigent, 1994). They proposed that the main contributors of
yeast extract are the purine and pyridine bases and B group
Figure 4. Experimental and calculated dependence of lactic acid and
glucose concentrations on the fermentation time correspon ding to fermen-
tations of 90 g l
¡
1
glucose with different concentrations of corn steep liquor
alone and supplemented with peptone or yeast extract.
Figure 5. Experimental and calculated dependence of lactic acid and
glucose concentration on the fermentation time corresponding to fermenta-
tions of 90 g l

¡
1
glucose with different concentrations of corn steep liquor
supplemented with citrate, acetate or phosphate.
Figure 6. Experimental and calculated dependence of lactic acid and
glucose concentration on the fermentation time corresponding to fermenta-
tions of 90 g l
¡
1
glucose with different concentrations of corn steep liquor
supplemented with MgSO
4
, MnSO
4
or FeSO
4
.
Trans IChemE, Vol 81, Part C, September 2003
254 TE
´
LLEZ-LUIS
et al.
vitamins. The importance of yeast extract to
Lactobacilli
has
been reported (Hujanem and Linko, 1996).
Figure 5 shows experimental results for the fermentat ion
of 10 g l
¡
1

CSL supplemented with the carboxylic salts and
mineral acids (sodium citrate, sodium acetate and sodium
phosphate) of the Mercier medium. Citrate, acetate and phos-
phate decreased signi cantly the value of
Y
p
=
s
but the
E
p
=
s
was increased after supplementing with citrate or acetate.
Using 10 g l
¡
1
CSL alone, lactic acid concentration was
27.13 g l
¡
1
. This value slightly increased with supplementa-
tion with acetate or citrate but enhancement was not observed
with phosphate.The involvement of citrate and acetate in the
metabolism cycles could be the explanation.The supplemen-
tation of CSL with citrate and acetate enhanced the economy
of the lactic acid fermentation.
E
p
=

„
was 111 g lactic acid per
„
nutrient for the fermentation of the 10 g l
¡
1
CSL medium
supplemented with 5 g l
¡
1
acetate. The
E
p
=
„
of med ia with
citrate was the same as that for yeast extract (Table 4).
Figure 6 shows the experimental results from the fermen-
tation of 10 g l
¡
1
CSL supplemented with MgSO
4
, MnSO
4
and FeSO
4
, the mineral sources of the Mercier medium. The
lactic acid concentration obtained by supplementing CSL
with MgSO

4
remained the same as that of medium without
supplementation at 98 h . It is reported that magnesium is
a key element in lactic acid fermentation (Thomas and
Ingledew, 1990). In our study, addition of MgSO
4
had no
effect on la ctic acid production. This must have occurred
because CSL contains 1.5% Mg
2
‡
on a dry basis (Zabriskie
et al.
, 1980). Using MgSO
4
as a component of the CSL
medium decreased the
E
p
=
„
of the process.
Better results were obtained by supplementing with
MnSO
4
or FeSO
4
(Table 4). Although the highest lactic
acid concentration was obtained by supplementing with
yeast extract, the

E
p
=
„
showed that the better supplement
is 0.05 g l
¡
1
FeSO
4
because 134 g lactic acid per
„
of
nutrients was obtained. This value is 75% higher than the
E
p
=
„
obtained with 10 g l
¡
1
CSL, 36% higher than to
20 g l
¡
1
CSL medium and 168% higher than the
E
p
=
„

value when the Mercier medium was used.
Mathematical Modelling of Fermentation with CSL
Supplemented with Nutritional Source
The experimental l actic acid production and glucose
concentration d ata were examined using equation (2) and
equation (6). The kinetic parameters of
P
0
,
P
m
and
P
r
were
calculated for each fermentation medium by non-linear
regression. The results are shown in Table 5 together with
the statistical p arameters. The coef cient
r
2
showed that all
the equations obtained were well  tted and Fig ures 4–6
con rm the good agreement between experimental and
predicted data. The value of
F
-test probability also showed
that the model was accurate. The values of
P
0
obtained are

higher than the values reported by others (Para jo´
et al.
,
1996). This was because the CSL contains a low concentra-
tion of lactic acid (Hull
et al.
, 1996).
Table 5 also shows the parameter
Y
p
=
s
and statistical
parameters for the consumption of glucose. The coef cient
showed a good agreement between experimental and
predicted data. Figures 4–6 also display comparison
between experimental and predicted data. The values of
Y
p
=
s
are in agreement with those reported in the literature
(Parajo´
et al.
, 1996).
CONCLUSIONS
Alternative media based on CSL were evaluated in order
to improve the eco nomic ef ciency of the lactic acid
production by
Lactobacillus delbrueckii

NRRL B-445. CSL
is a cheaper nutrient source than other complex media like
that proposed by Mercier. In this work, it was demonstrated
that a medium containing 1 0 g l
¡
1
corn steep liquor is more
economically ef cient than the Mercier medium but it is not
a balanced nutritional medium for
Lactobacillus delbrueckii
.
It can be improved by adding other supplements such as
yeast extract or mineral salts. A medium containing CSL
(10 g l
¡
1
) with 0.05 g l
¡
1
FeSO
4
is an economically ef cient
medium for lactic acid production by
Lactobacillus
delbrueckii
NRRL B-445.
REFERENCES
Amrane, A. and Prigent, Y., 1994, Mathematical model for lactic acid
production from lactose in batch culture: model development and simula-
tion, J Chem Technol Biotechnol, 60: 241–246.

Burgos-Rubio, C.N., Okos, M.R. and Wankat, P.C., 2000, Kinetic studies of
the conversion of differen t substrates to lactic acid using lactobacillus
bulgaricus, Biotech nol Prog, 16: 305–314.
Chahal, S.P., 1991, Lactic acid, in Ullmann’s Encyclopedia of Industrial
Chemistry (VCH, Berlin, Germany), Vol A15, pp 97–105.
Table 5. Results obtained by regression analysis of lactic acid production and glucose consumption by Lactobacillus delbrueckii using CSL supplemented
with other components of Mercier medium (all media include 90 g l
¡
1
glucose).
Lactic acid production Glucose consumption
Medium
a
P
0
(g l
¡
1
) P
m
(g l
¡
1
) P
r
(h
¡
1
) r
2

F-test of
probability Y
p
=
s
(g g
¡
1
) r
2
F-test of
probability
CSL 1.51 26.09 0.141 0.9967 0.9606 0.80 0.9987 0.9602
CSL ‡ peptone 3.80 72.64 0.150 0.9898 0.9309 0.81 0.9858 0.9956
CSL ‡ yeast extract 3.46 73.95 0.169 0.9890 0.9350 0.84 0.9942 0.9578
CSL ‡ citrate 3.72 45.59 0.048 0.9859 0.9298 0.62 0.9978 0.9129
CSL ‡ acetate 4.47 42.89 0.048 0.9720 0.9081 0.76 0.9930 0.9573
CSL ‡ K
2
HPO
4
2.02 27.89 0.111 0.9977 0.9377 0.71 0.9926 0.9137
CSL ‡ MgSO
4
2.47 22.83 0.078 0.9859 0.9291 0.77 0.9952 0.9492
CSL ‡ MnSO
4
4.11 38.14 0.051 0.9773 0.9036 0.67 0.9956 0.9675
CSL ‡ FeSO
4

4.78 60.38 0.041 0.9771 0.9232 0.74 0.9925 0.9176
a
The CSL concentration was 10 g l
¡
1
in all the experiments and the concentrations of the other nutrients were the same as used in Mercier medium.
Trans IChemE, Vol 81, Part C, September 2003
ALTERNATIVE MEDIA FOR LACTIC ACID PRODUCTION 255
Cheilas, T., Stoupis, T., Christakopoulos, P., Katapodis, P., Mamma, D.,
Hatzinikolaou, D.G., Kekos, D. and Macris, B.J., 2000, Hemicellulolytic
activity of Fusarium oxysporum grown on sugar beet pulp. Production
of extracellular arabinase, Proc Biochem, 35: 557–561.
Hofvendahl, K. and Hahn-Hagerdal, B., 2000, Factors affecting the fermen-
tative lactic acid production from renewable resources, Enzyme Microb
Technol, 26: 87–107.
Hujanen, M. and Linko, Y.Y., 1996, Effect of temperature and various
nitrogen sources on L ‡ lactic acid production by Lactobacillus casei,
Appl Microbiol Biotechnol, 45: 307–313.
Hull, S.R., Peters, E., Cox, C., Montgomery, R., 1996, Composition of
corn steep water during experimental steeping, J Agric Food Chem, 44:
3521–3527.
Kadam, K.L. and Newman, M.N., 1997, Development of a low-cost
fermentation medium for ethanol production from biomass, Appl
Microbiol Biotechnol, 47: 625–629.
Lee, P.C., Lee, W.G., Lee, S.Y., Chang, H.N. and Chang, Y.K., 2000,
Fermentative production of succinic acid from glucose and corn steep
liquor by Anaerobiospirillum succiniciproducens, Biotechnol Bioprocess
Eng, 5: 379–381.
Mercier, P., Yerushalmi, L., Rouleau, D. and Dochain, D., 1992, Kinetics
of lactic acid fermentation on glucose and corn by Lactobacillus

amylophilus, J Chem Tech Biotechnol, 55: 111–121.
Miller, T.L. and Churchill, B.W., 1986, Substrates for large-scale fermenta-
tions, in Manual of Industrial Microbiology and Biotechnology,
Demain, A.L. and Solomon, L.A. (eds) (American Society for Micro-
biology, Washington, DC, USA).
Nancib, N., Nancib, A., Boudjelal, A., Benslimane, C., Blanchard, F.
and Boudrant, J., 2001, The effect of supplementation by
different nitrogen sources on the production of lactic acid from
data juice by Lactobacillus casei subsp. rhamnosus, Biores Technol,
78(2): 149–153.
Ozen, S. and Ozilg en, M., 1992, Effects of substrate concentration on
growth an d lactic acid production by mixed cultures of Lactobacillus
bulgaricus and Streptococcus thermophillus, J C hem Technol Biotechnol,
54: 57–61.
Parajo´, J.C., Alonso, J.L. and Santos, V., 1996, Lactic acid from wood, Proc
Biochem, 31: 271–280.
Silveira, M.N., Wisbeck, E., Hoch, I. and Jonas, R., 2001, Production of
glucose-fructo se ox idoreductase and ethanol by Zimomonas mobilis
ATCC29191 in medium containing corn steep liquour as a source of
vitamins, Appl Microbiol Biotechnol, 55: 442–445.
Te´llez-Luis, S.J., Moldes, A.B., Alonso, J.L. and Va
´
zquez, M., 2003,
Optimization of lactic acid productio n by Lactobacillus delbrueckii
through response surface methodology, J Food Sci, 37: 267–274.
Thomas, K.C. and Ingledew, W.N., 1990, Fuel ethanol production: effects of
free amino nitrogen on fermentation of very-high-gravity wheat mashes,
Appl Environ Micro biol, 56: 2046–2050.
Vick-Roy, T.B., 1985, Comprehensive Bio technology, Moo Young, M. (ed.)
(Pergamon Press, Oxford, UK).

Zabriskie, D.W., Armiger, W.B., Phillips, D.H. and Albano, P.A., 1980,
Traders’ Guide to Fermentation Med ia Formulation (Traders Protein,
Memphis, TN, USA).
Zhou, Y., Dom‡
´
nguez, J.M., Cao, N.J., Gong, C.S. and Tsao, G.T., 1999,
Optimization of L-lactic acid production from glucose by Rhizopus
oryzae ATCC 52 311, Appl Biochem Biotechnol, 77–79: 401–407.
ACKNOWLEDGEMENTS
The Authors are grateful to Xunta de Galicia for the  nancial support of
this work (Project XUGA PGIDT00PXI38301PR). A grant from the
PROMEP program of the Secretar‡
´
a de Educacio´n Pu
´
blica (Me´xico) to
author Te´llez-Luis is gratefully acknowledged.
ADDRESS
Correspondence concerning this paper should be addressed to
Dr M. Va´zquez, A
´
rea de Tecnolog‡´a de los Alimentos, Escuela Polite´cnica
Superior, Departamento de Qu‡
´
mica Anal‡
´
tica, Universidad de Santiago
de Compostela, Campus de Lug o, 27002 Lugo, Spain.
E-mail:
The manuscript was received 7 May 2002 and accepted for publication

after revision 30 April 2003.
Trans IChemE, Vol 81, Part C, September 2003
256 TE
´
LLEZ-LUIS
et al.