ARTICLE IN PRESS
Effect of preculturing conditions on growth of
Lactobacillus rhamnosus on medium containing
glucose and citrate
B.D. Jyotia, A.K. Suresha,b, K.V. Venkatesha,b,*
a
Department of Chemical Engineering, Indian Institute of Technology, Bombay, Powai, Mumbai-400076, India
School of Biosciences and Bioengineering, Indian Institute of Technology, Bombay, Powai, Mumbai-400076, India
b
KEYWORDS
Flux analysis;
Lactobacillus;
Multiple substrates;
Preculturing;
Microaerophilic
Abstract
Lactobacillus rhamnosus can metabolize citrate through a citrate inducible
transport system. The growth curves of L. rhamnosus on medium containing
glucose and citrate was found to be highly dependent on preculturing conditions. It
exhibited diauxic growth when precultured on glucose, but demonstrated
simultaneous consumption when cultured on citrate. The maximum specific
growth rate for cells growing on glucose þ citrate was 0.38 h1, which was higher
than the growth rate on individual substrates (0.28 h1). Simultaneous consumption
also yielded higher net flavour compounds, diacetyl and acetoin. Flux analysis
indicated that L. rhamnosus requires oxygen for balancing excess NADH through
NADH oxidase. The flux analysis provided insights into the metabolic network
of L. rhamnosus.
Introduction
Lactobacillus rhamnosus is a heterolactic acid
bacterium and is extensively used in food industry.
Several studies indicate that this organism is a
possible candidate for industrially synthesizing
flavour compounds such as diacetyl and acetoin
(Lee, 1991; Anuradha et al., 1999). The use of
multiple substrates in the medium, especially
citrate or pyruvate with glucose, has been reported
to increase the yield and productivity of diacetyl
and acetoin (Jyoti et al., 2003).
Various lactic acid bacteria utilize citrate, either
directly or by co-metabolism, for the production of
aroma compounds (Figueroa et al., 2000). Montville
et al. (1987) have suggested that the citrate
transport system was inducible by citrate and the
level of induction depends on the strain. The use of
ARTICLE IN PRESS
36
citrate and lactose in the medium increased the
specific growth rate of Leuconostoc mesenteroides
subspecies cremoris (Schmitt and Divies, 1990).
The diacetyl and acetoin were produced as soon
as the citrate consumption began resulting in a
better productivity of diacetyl and acetoin as
compared to glucose alone in the medium. The
co-metabolism of glucose and citrate by Lactococcus lactis subsp. lactis biovar diacetylactis
indicated that 65% of the carbon source was
converted to flavour producing compound (Goupry
et al., 2000). The co-metabolism of xylose and
citrate gave high yields of diacetyl and acetoin,
11.5% in batch culture and 17.4% in chemostat
culture by Leuconostoc mesenteroides subsp.
mesenteroides (Schmitt et al., 1997). It was
found that in the case of L. rhamnosus, the
presence of lactose does not allow the utilization
of citrate at high concentrations, while the
presence of glucose does not allow the utilization of citrate at low concentrations (Benito de
Cardenas et al., 1992).
It has been seen that the preculturing in
different media affects the fermentation greatly.
The cells utilize the different substrates either
simultaneously or sequentially depending on the
preculturing conditions. When E. coli was grown in
a medium containing glucose þ lactose, due to
catabolite repression glucose was always the
preferred substrate without being affected by the
preculturing conditions (Bailey and Ollis, 1986;
Narang et al., 1997). Kompala et al. (1986)
reported a diauxic growth on glucose and xylose
with xylose as less preferred substrate with
preinoculum being grown on glucose. Doshi and
Venkatesh (1998) observed that when the cells of E.
coli were pre cultured on glucose and grown in a
medium containing glucose, acetate and lactate,
a diauxic growth was observed with lactate as
less preferred substrate, while glucose and
acetate were consumed simultaneously in the first
phase. Figueroa et al. (2000) have shown that L.
rhamnosus grows on citrate only when cultured on
citrate and glucose þ citrate. Thus, it is clear that
preculturing conditions does have an effect on the
growth of L. rhamnosus.
In L. rhamnosus, if citrate permease is an
inducible enzyme, what is the effect of preculturing on this induction? Here, a comparative study
was carried out to study the effect of preculturing
on the growth of L. rhamnosus on citrate. The
analysis was based on growth curves and flux
analysis of the data obtained form experiments
with different culturing conditions. The flux analysis provided insights into the metabolic network of
L. rhamnosus.
Materials and methods
Fermentation and product analysis: The methodology for fermentation and analysis of substrates,
biomass and various products followed was same as
reported by Jyoti et al. (2003). To study the effects
of culturing, the inoculum was grown in presence of
glucose, citrate and glucose þ citrate along with all
other ingredient mentioned in the above reference.
The cell O.D of inoculum was maintained at 0.5 in
all cases at the time of inoculation to the main
fermentation medium. The initial concentration of
glucose and citrate in the main fermentation
medium to study the effect of culturing medium
were as follows: (1) for the inoculum grown in
glucose alone: glucose, 5.6 g/l and citrate 2.2 g/l;
(2) for the inoculum grown in citrate: glucose,
5.3 g/l and citrate 1.8 g/l; and (3) for the inoculum
grown in glucose þ citrate alone: glucose, 6.3 g/l
and citrate 2.2 g/l. The substrates (glucose and
citrate) and the other metabolites (lactate, diacetyl, acetoin and acetate) were measured by high
performance liquid chromatography using a Shimadzu HPLC system and Biorad Aminex HPX-87 H
column with guard columns in series. In this paper,
the data has been presented as absolute accumulation rates (mole/l/h) at 12–14 h of fermentation
(that is for a metabolite M; as dM=dt; slope of the
formation of metabolite at 12–14 h).
Results
Fluxes in L: rhamnosus
Stoichiometric balance method is a well-established technique used to evaluate intracellular
reaction rates. Metabolic balancing is carried out
based on steady-state approximation with respect
to intracellular metabolites. The methodology
requires the knowledge of metabolic network and
accumulation rates of the key extracellular metabolites (Varma and Palsson, 1994). The balances
yield a set of algebraic equations for a network,
A X ¼ R;
ð1Þ
where, A and R are the matrix and vector,
respectively, represents the biochemistry of
the reaction network and accumulation rates of
the extracellular metabolites. X is the vector
containing the flux value for each reaction in
the metabolic network. By knowing A and R, X
can be evaluated by solving the set of algebraic
equations (Venkatesh, 1997).
ARTICLE IN PRESS
37
Table 1. Biochemical reaction network used for flux analysis in L. rhamnosus (Xi represents the flux associated with the
respective reaction)
Glycolysis
GLC þ 2 NAD þ þ 2ADP þ 2Pi-2PYR þ 2H2O þ 2NADH þ 2ATP
X1
Biomass formation
PYR þ 0.3ATP-BIOMASS þ 0.3ADP
X2
Pyruvate breakdown
PYR"PYRE
PYR þ NADH"LACT þ NAD þ
PYR"ACEN þ CO2
PYR þ ACCOA þ NAD þ -DIA þ NADH þ CO2 þ COA
DIA þ NADH-ACEN þ NAD þ
PYR þ COA þ NAD þ -ACCOA þ CO2 þ NADH
X3
X4
X5
X6
X7
X8
Acetate formation
ACCOA þ ADP þ Pi"ACET þ COA þ ATP
X9
TCA Cycle
ACCOA þ OAA þ H2O-CIT þ COA
CIT þ H2O þ 4NAD þ þ ADP þ Pi-OAA þ 2CO2 þ 4NADH þ ATP
X10
X11
Citrate uptake
CITE"CIT
X12
Citrate breakdown
CIT-ACET þ OAA
X13
Decarboxylation reaction
OAA"PYR þ CO2
X14
Oxidation reaction
4NADH þ O2-4NAD þ þ 2H2O
X15
The biochemical reaction network of L. rhamnosus and balance equations are given in Tables 1
and 2, respectively. Glycolysis is represented by a
single reaction to yield 2 mole of pyruvate. The
conversion of citrate to oxaloacetate (as a single
reaction) and the combination of oxaloacetate with
acetyl CoA to citrate represents the TCA cycle.
Citrate is transported into the cell by citrate
permease and later converted to OAA and acetate
(Lee, 1991). Further, OAA undergoes decarboxylation reaction to yield pyruvate. Lactate dehydrogenase converts pyruvate to lactate and also
converts lactate back to pyruvate in the absence
of glucose. It was assumed that the biomass with
46% of carbon was entirely balanced by pyruvate.
Also, 0.3 mole of ATP was assumed to be required
for the formation of biomass (Benthin et al., 1994).
Acetate was formed from acetyl CoA and was a
source of ATP. At the pyruvate node, the flux was
channelled to various metabolites. Pyruvate is
converted to lactic acid for recycling of NAD þ .
Pyruvate is enzymatically converted to acetoin by
decarboxylation. Also, pyruvate stoichiometrically
combines with ACCOA to yield diacetyl. It is clear
from the table that fifteen unknown fluxes have to
be determined to quantify the metabolic network
of L. rhamnosus.
Table 2 shows the balance equations for the
metabolites based on the stoichiometry of the
reactions in the network. The accumulation rates
(components of vector R in Eq. (1)) for GLC, CIT,
DIA, ACEN, LACT, ACET and BIO were determined
using experimental data for media containing
different carbon sources. PYR, ACCOA, OAA and
NADH were assumed to be at pseudo-steady state.
The oxygen uptake rate (X15 ) was determined based
on NADH balance. The ATP balance (Eq. (T14) in
Table 2) was used to verify the feasibility of the flux
distribution and a positive accumulation rate was
ensured. It should be noted that Eqs. (T14) and
(T15) were not used for evaluating fluxes. From
Table 2, it is clear that there are 13 equations
ARTICLE IN PRESS
38
Table 2. Stoichiometric balance equations for the network of L. rhamnosus (see Table 1 for description of X)
q GLU
¼ X1
qt
(T1)
q PYR
¼ 2X1 þ X14 X2 X3 X4 X5 X6 X8
qt
(T2)
q BIO
¼ X2
qt
(T3)
q PYRE
¼ X3
qt
(T4)
q LACT
¼ X4
qt
(T5)
q ACEN
¼ X5 þ X7
qt
(T6)
q DIA
¼ X6 X7
qt
(T7)
q ACCOA
¼ X8 X6 X9 X10
qt
(T8)
q ACET
¼ X19 þ X13
qt
(T9)
q CITE
¼ X12
qt
(T10)
q CIT
¼ X10 þ X12 X11 X13
qt
(T11)
q OAA
¼ X11 þ X13 X10 X14
qt
(T12)
q NADH
¼ 2X1 þ X6 þ X8 þ 4X11 X4 X7 4X15
qt
(T13)
q ATP
¼ 2X1 þ X9 þ X11 0:3X2
qt
(T14)
q O2
¼ X15
qt
(T15)
with 15 unknowns. Two more equations were
required to evaluate the 15 fluxes. When the
medium contained citrate, it was assumed that
the acetate was formed from citrate alone (i.e.,
X9 ¼ 0) and pyruvate was not formed (i.e., X3 ¼ 0).
Also when the growth was on glucose alone, it was
assumed that acetate was formed from acetyl-CoA
(i.e., X13 ¼ 0) and the decarboxylation of OAA did
not occur (X14 ¼ 0). The above set of equations was
solved with these conditions to give a feasible
solution.
Growth on glucose and citrate
Experiments were carried out to study the effect of
culturing conditions on the growth of L. rhamnosus
on medium containing glucose þ citrate. Cells were
previously precultured on medium containing glucose alone, citrate alone or glucose þ citrate. Fig. 1
shows the biomass at different time points when
the cells were cultured on glucose alone. The
profile is a typical diauxic curve without a long lag
phase in between the phases of growth, with
glucose as the preferred substrate. This may
indicate that glucose might regulate citrate permease. Fig. 2 shows the biomass profile with
respect to time for cells precultured on citrate
alone and glucose þ citrate. The curves indicate
that glucose and citrate were simultaneously
consumed. This data signifies that glucose does
not regulate citrate permease contrary to the
result indicated by culturing on glucose alone.
Table 3 shows the maximum specific growth rate
for the three preculturing condition. The specific
growth (of 0.38 h1) was almost same for preculturing on citrate alone and citrate þ glucose.
The resultant high specific growth rate was
due to simultaneous consumption of glucose and
citrate. The growth rate was lower in case of
culturing on glucose alone due to no citrate uptake
in this phase. The specific growth rate was greater
in case of growth on citrate or glucose þ citrate
than growth on just glucose in the medium
(0.28 h1) and just on medium containing citrate
(0.29 h1).
The above growth data with measurements
of glucose, lactate, citrate, diacetyl and acetoin
were obtained for different preculturing conditions. The slope of the curve at 12–14 h of
fermentation was used to evaluate the accumulation rates of different metabolites. Table 4 lists
the absolute and normalized (with respect to
glucose uptake) accumulation rates for different
metabolites. Flux analysis was carried out based on
these accumulation rates obtained at 12–14 h of
fermentation.
Fig. 3 shows the flux distribution for the
growth of L. rhamnosus on medium containing
glucose þ citrate with different preculturing conditions. The numbers shown without brackets represent the flux distribution for the case when the
cells were precultured on glucose as the sole
carbon source. In this case, the incoming flux was
due to consumption of glucose only because citrate
was not consumed. Flux distribution for growth on
glucose demonstrates that at pyruvate node, 59% of
the flux coming from glycolysis was distributed
between lactate and biomass synthesis. The
ARTICLE IN PRESS
39
1. 8
1. 6
Biomass (g/L)
1. 4
1. 2
1.0
0. 8
0.6
0. 4
0. 2
0. 0
0. 0
10.0
20.0
30.0
40.0
50.0
Time (Hours)
Figure 1. Growth of L. rhamnosus on glucose þ citrate for different preculturing conditions (o, represents preculturing
on glucose).
1.8
1.6
Biomass (g/L)
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
Time (Hours)
Figure 2. Growth of L. rhamnosus on glucose þ citrate for different preculturing conditions (&, represents
preculturing on citrate; and, n, represents preculturing on glucose þ citrate).
Table 3. Maximum Specific Growth Rate for L. rhanmnosus (h1) on medium containing different combination of
carbon sources
Precultured on
mmax a
mmax b
Type of growth
Glucose
Citrate
Glucose þ citrtate
0.29
0.37
0.38
0.01
F
F
Diauxic
Simultaneous
Simultaneous
a
Maximum specific growth rate for the first phase of growth.
Maximum specific growth rate for the second phase of growth.
b
remaining 41% of the pyruvate was distributed
among acetoin, diacetyl and acetate with a small
flux toward TCA cycle.
Out of a total flux of 344 for NADH synthesis (200,
14 and 130 from glycolysis, diacetyl synthesis and
TCA recycling), 71 was recycled by lactate formation, 7 was balanced by the conversion of diacetyl
to acetoin and the remaining 264 was recycled by
NADH oxidase. This was balanced by an oxygen
consumption rate of 66 for oxygen. ATP was
synthesized through glycolysis, acetate formation
and a small fraction through TCA.
The numbers shown inside parentheses represent
the flux distribution for the growth of L. rhamnosus
on a medium containing glucose þ citrate with the
cells precultured in a medium containing citrate as
the sole carbon source. Both the substrates were
ARTICLE IN PRESS
40
Table 4. Accumulation Rates and normalized values of
fluxes for extracellular metabolites (normalized with
respect to the consumption rate of glucose)
Preculturing
condition
M
A
F
Glucose
Glucose
Citrate
Lactate
Diacetyl
Acetoin
Acetate
Biomass
4.2 103
0
2.4 103
3 104
6 104
7 104
2.5 103
100
0
57
7
14
17
60
Citrate
Glucose
Citrate
Lactate
Diacetyl
Acetoin
Acetate
Biomass
2.5 103
4.5 104
1.69 103
2.7 104
6.3 104
6.8 104
2.04 103
100
18
67
11
25
27
81
Glucose þ citrate
Glucose
Citrate
Lactate
Diacetyl
Acetoin
Acetate
Biomass
4.8 103
5.3 104
3.4 103
6 104
1.1 103
8 104
3.5 103
100
11
71
13
23
17
73
Nomenclature: M; extracllular metabolite; A; accumulation rates
in moles/l/h, F; normalized flux with respect to glucose uptake.
utilized simultaneously to direct a flux of 218
towards pyruvate. Out of a total flux of 218, the
flux was distributed between lactate (67) and
biomass (60). The excess pyruvate available due
to citrate increased the yields of other metabolites. The yields of 0.09 g acetate/g glucose, 0.05 g
diacetyl/g glucose and 0.12 g acetoin/g glucose
were obtained. The ATP requirement for the growth
of cells was fulfilled mainly by glycolysis and TCA
cycle. Glycolysis (72%), diacetyl formation (6%),
pyruvate conversion to ACCOA (12%) and TCA cycle
(10%) contributed to NADH formation. While lactate formation (24%), diacetyl conversion to acetoin (3%) and oxygen uptake flux of 52 (73%)
contributed to the recycling of NADH.
The numbers shown inside square brackets
present the flux distribution for the growth of L.
rhamnosus on a medium containing glucose
þ citrate with the cells precultured on glucose þ citrate (Fig. 3). The flux distribution in this
case was almost similar to that of preculturing on
citrate alone (indicated by single brackets). Both
the substrates were utilized simultaneously to
direct a flux of 211 towards pyruvate. 68% of the
above flux was distributed between lactate (71)
and biomass (73). The yields of acetate, diacetyl
and acetoin were 0.06 g acetate/g glucose, 0.06 g
diacetyl/g glucose and 0.11 g acetoin/g glucose
respectively. Glycolysis and TCA cycle fulfilled the
ATP requirement for the growth of cells. In this
case, glycolysis (73%), diacetyl formation (7%),
pyruvate conversion to ACCOA (11%) and TCA cycle
(9%) contributed to NADH formation. While lactate
formation (27%), diacetyl conversion to acetoin
(2%) and oxygen uptake flux of 50 (71%) contributed
to the recycling of NADH.
Table 5 shows a comparison between the yields of
metabolites for different preculturing conditions.
The table shows that the yields of biomass and
various metabolites are higher when the cells are
precultured on citrate or glucose þ citrate as
compared to the case when the cells were cultured
on glucose alone. The highest yields of biomass
(0.35 g/g of glucose) and acetate (0.09 g/g of
glucose) were obtained when the cells were
cultured on citrate alone. The yields of diacetyl
and acetoin were found to increase when the cells
were cultured either on citrate or glucose
þ citrate. Table 5 also shows that the normalized uptake rate of oxygen required by L. rhamnosus to balance oxidation state was in the range of
50–70 irrespective of the preculturing conditions
(YO2/S).
Discussion
The stoichiometric balance method was used to
analyse the effect of multiple substrates on the
network of L. rhamnosus. The flux distribution in
the network was evaluated with experimentally
measured extracellular metabolites for medium
containing various substrates and their combinations. Unlike in a homolactic organism, L. rhamnosus cannot convert the pyruvate to lactate at a fast
rate to match the rate of glycolysis. This resulted in
the accumulation of pyruvate inside the cell, which
is toxic to the cell. Therefore, pyruvate gets
channelled to various other metabolites, such as
acetate, diacetyl and acetoin. Since lactate alone
cannot balance the oxidation state, thus oxygen
was also essential for L. rhamnosus to balance the
remaining NADH. The constraints placed by NADH
recycling, ATP and carbon availability, decided the
flux distribution towards different products. NADH
was partly recycled by lactate formation and thus
L. rhamnosus required oxygen for balancing the
oxidation state. This flexibility in L. rhamnosus to
balance the oxidation state was the reason for
simultaneous utilization of glucose and citrate.
ARTICLE IN PRESS
41
GLU
BIO
200 (200) [200]
60 (81) [ 73]
LACT
ACEN
7 (18) [ 17]
7 (7) [ 6 ]
71 (67) [ 71]
14 (18) [ 19]
DIA
PYR
0 (18) [ 11]
48 (34) [ 31]
17 (0) [ 0 ]
ACET
ACCOA
17 (16) [ 11]
0 (18) [ 11]
0 (27) [ 17]
OAA
CIT
CITE
TCA
17 (7) [ 6 ]
Figure 3. Flux distributions for growth of L. rhamnosus on glucose þ citrate for different preculturing conditions. Unbracketed numbers, numbers in parentheses and numbers in square brackets represent the flux distribution for
preculturing on glucose alone, on citrate alone and on glucose þ citrate, respectively.
Table 5. Comparison of metabolite yields grown on medium containing glucose þ citrate, for different preculturing
conditions
Precultured on
Glucose
Citrate
Glucose þ citrate
Yield (g/g of glucose)
YX=S
YLT=S
YAT=S
YD=S
YAN=S
YO2=S
0.26
0.35
0.32
0.28
0.33
0.35
0.05
0.09
0.06
0.03
0.05
0.06
0.07
0.12
0.11
0.66
0.52
0.50
YX=S ¼ biomass yield; YLT=S ¼ lactic acid yield; YD=S ¼ diacetyl yield; YAT=S ¼ acetic acid yield; YAN=S ¼ acetoin yield; and YO2=S represent
the oxygen consumption with respect to glucose consumption.
The yield of diacetyl increased for the medium
which was inoculated by those cells which were
pre-inoculated in a medium containing citrate
(0.05 g diacetyl/g of glucose) or glucose þ citrtate
(0.06 g diacetyl/g of glucose) as compared to
glucose (0.03 g diacetyl/g of glucose).
ARTICLE IN PRESS
42
The above analysis shows that the culturing
conditions affect the growth as well as the
formation of metabolites. Since citrate permease
is inducible, it requires a lag time for its synthesis.
It can be postulated that if this lag time is
completed in the preculturing stage, then the cells
consume both glucose and citrate. Thus, the
diauxic growth when cells were precultured on
glucose alone, was just a manifestation of synthesis
time required for citrate permease. The maximum
flux towards diacetyl þ acetoin was found to be
higher than that for simultaneous consumption of
glucose þ citrate in comparison to the sequential
utilization of glucose and citrate.
The flux analysis demonstrated a low oxygen
uptake rates for the growth of L. rhamnosus
indicating that the organism is microaerophilic. A
typical aerobic organism requires about 300–400
normalized oxygen uptake rates (equivalent yield,
YO2/S, of 3–4) for glucose uptake (Vallino 1991). L.
rhamnosus, a microaerophilic organism, demonstrated an oxygen uptake rate of only 10–15% of
that of an aerobic organism. The metabolic state in
the fermenter was thus affected by the culturing
status of L. rhamnosus as demonstrated by the flux
balance.
The above discussion indicates that to further
evaluate the network of hetero-lactic organisms,
perturbation experiments sholuld be planned
around the pyruvate node. The experimental
analysis reported here demonstrated the effect of
preculturing effects on the grwoth of L. rhamnosus
and also the feasibility of metabolic engineering of
L. rhamnosus towards optimization of diacetyl.
References
Anuradha, R., Suresh, A.K., Venkatesh, K.V., 1999.
Simultaneous saccharification and fermentation of
starch to lactic acid. Proc. Biochem. 35, 367–375.
Bailey, J.E., Ollis, D.V., 1986. Biochemical Engineering
Fundamentals. Mc-Graw Hill, Singapore.
Benito de Cardenas, I.L., Medina, R., Oliver, G., 1992.
Effect of glucose and lactose on the utilization of
citrate by Lactobacillus casei subsp. rhamnosus ATCC
7469. Rev. Argentina de Microbiol. 24, 136–144.
Benthin, S., Schulze, U., Nielsen, J., Villadsen, J., 1994.
Growth energetics of Lactcoccus cremoris FD1 during
energy-, carbon- and nitrogen-limitation in steady
state and transient cultures. Chem. Eng. Sci. 49,
589–609.
Doshi, P., Venkatesh, K.V., 1998. An optimal model for
microbial growth in multiple substrate environment:
simultaneous and sequential utilization. Pro. Biochem. 33, 663–670.
Figueroa, M.R., Alvarez, F., Pesce de Ruiz Holgado, A.,
Oliver, G., Sesma, F., 2000. Citrate utilization by
homo- and hetero-fermentative lactobacilli. Microbiol. Res. 154, 313–320.
Goupry, S., Croguennec, T., Gentil, E., Robins, R.J.,
2000. Metabolic flux in glucose/citrate co-fermentation by lactic acid bacteria as measured by isotopic
ratio analysis. FEMS Microbiol. Lett. 182, 207–211.
Jyoti, B.D., Suresh, A.K., Venkatesh, K.V., 2003. Diacetyl
production and growth of lactobacillus rhamnosus on
multiple substrates. World J. Microbiol. Biotechnol.
19, 509–514.
Kompala, D.S., Ramakrishna, D., Jensen, N.B., Tsao,
G.T., 1986. Investigation of bacterial growth on
multiple substrates: experimental evaluation of cybernatic models. Biotechnol. Bioeng. 28, 1044–1055.
Lee, W., 1991. Production of diacetyl (2,3, butanediol)
by continuous fermentation with simultaneous
product separation. Ph.D. Thesis, Purdue University,
USA
Montville, T.J., Hsu, A.H., Meyer, M.E., 1987. Influence of
carbon sources on lactic acid, cell mass and diacetylactoin production in Lactobacillus planatarum. Appl.
Environ. Microbiol. 53, 1798–1802.
Narang, A., Konopka, A.E., Ramkrishna, D., 1997. New
patterns of mixed-substrate growth in batch cultures
of E. coli K12. Biotechnol. Bioeng. 55, 747–757.
Schmitt, P., Divies, C., 1990. Co-metabolism of citrate
and lactose by Leuconostoc mesenteroides subsp.
cremoris. J. Ferment. Bioeng. 71, 72–74.
Schmitt, P., Vasseur, C., Phalip, V., Huang, D.Q., Divies,
C., Prevost, H., 1997. Diacetyl and acetoin production
from the co-metabolism of citrate and xylose by
Leuconostoc mesenteroides subsp. mesenteroides.
Appl. Microbiol. Biotechnol. 47, 715–718.
Vallino, J.J., 1991. Microbial metabolism through metabolic flux analysis and local network perturbations.
Ph.D. thesis, MIT, USA.
Varma, A., Palsson, B.O., 1994. Stoichiometric flux
balance models quantitatively predict growth and
metabolic by-product secretion in wild-type Escherichia coli W3110. Appl. Environmen. Microbiol. 60,
3724–3731.
Venkatesh, K.V., 1997. Metabolic flux analysis of lactic
acid fermentation: effects of pH and lactate ion
concentration. Process Biochem. 32, 651–655.