Available via license: CC BY-NC 4.0
Content may be subject to copyright.
ISSN 1330-9862 scientific note
(FTB-2313)
Improvement of Folate Biosynthesis by Lactic Acid Bacteria
Using Response Surface Methodology
Norfarina Muhamad Nor1, Rosfarizan Mohamad1,2*, Hooi Ling Foo1,2
and Raha Abdul Rahim2,3
1Department of Bioprocess Technology, Faculty of Biotechnology and Biomolecular Sciences,
University of Putra Malaysia, 43400 UPM Serdang, Selangor Darul Ehsan, Malaysia
2Laboratory of Industrial Biotechnology, Institute of Bioscience, University of Putra Malaysia,
43400 UPM Serdang, Selangor Darul Ehsan, Malaysia
3Department of Cell and Molecular Biology, Faculty of Biotechnology and Biomolecular Sciences,
University of Putra Malaysia, 43400 UPM Serdang, Selangor Darul Ehsan, Malaysia
Received: July 8, 2009
Accepted: October 6, 2009
Summary
Lactic acid bacteria (Lactococcus lactis NZ9000, Lactococcus lactis MG1363, Lactobacillus
plantarum I-UL4 and Lactobacillus johnsonii DSM 20553) have been screened for their ability
to produce folate intracellularly and/or extracellularly. L. plantarum I-UL4 was shown to
be superior producer of folate compared to other strains. Statistically based experimental
designs were used to optimize the medium formulation for the growth of L. plantarum
I-UL4 and folate biosynthesis. The optimal values of important factors were determined
by response surface methodology (RSM). The effects of carbon sources, nitrogen sources
and para-aminobenzoic acid (PABA) concentrations on folate biosynthesis were determined
prior to RSM study. The biosynthesis of folate by L. plantarum I-UL4 increased from 36.36
to 60.39 mg/L using the optimized medium formulation compared to the selective Man de
Rogosa Sharpe (MRS) medium. Conditions for the optimal growth of L. plantarum I-UL4
and folate biosynthesis as suggested by RSM were as follows: lactose 20 g/L, meat extract
16.57 g/L and PABA 10 mM.
Key words: folate, lactic acid bacteria, Lactobacillus plantarum I-UL4, response surface method-
ology
Introduction
Folate plays an important role in human life as one
of the most important essential components for the syn-
thesis of purine, guanine, adenine, pyrimidine and thy-
mine. Sufficient daily doses of folate may prevent di-
seases such as colon cancer, growth retardation, anaemia
and neural tube defects in newborn. The human body
needs 200 to 400 mg of folate per day. However, preg-
nant women are advised to take double doses (1). Folate
is mostly found in green, leafy vegetables, legumes,
beans, citrus, and dairy products, but spinach and beans
have been found to contain the highest amount of folate
243
N. MUHAMAD NOR et al.: Folate Biosynthesis by Lactic Acid Bacteria, Food Technol. Biotechnol. 48 (2) 243–250 (2010)
*Corresponding author; Phone: ++603 8946 7518; Fax: ++603 8946 7510; E-mail: farizan@biotech.upm.edu.my
(2). Folic acid (also known as vitamin B9 or folacin),
folate (the naturally occurring form), pteroyl-L-glutamic
acid and pteroyl-L-glutamate are forms of the water-
-soluble vitamin B9. Folic acid is not biologically active
but after the conversion of tetrahydrofolate and other
derivatives into dihydrofolic acid in the liver, it becomes
biologically important. Folic acid found in fortified foods
such as vitamin supplements is the synthetic form of
water-soluble vitamin B9 (3). Based on the bioavailabil-
ity, folic acid is more stable than the natural form of
folate because folic acid is more readily and quickly ab-
sorbed into the human body.
On the other hand, a good selection of microbial spe-
cies such as lactic acid bacteria (LAB) and cultural con-
ditions could enhance the level of folate in the ferment-
ed milk and dairy products (4). It has been reported that
the folate levels in most of the products are low (5,6).
Therefore, several researchers have focused on the bio-
synthesis of folate using LAB strains (7). Large differ-
ences in folate biosynthesis for different LAB strains and
different growth conditions have been reported in liter-
ature. Lactococcus lactis, as claimed by Sybesma et al. (8),
has a great potential to produce high amount of folate
and its metabolism was further explored to maximize
folate biosynthesis. Other researchers have also reported
that many strains of bifidobacterium were able to pro-
duce folate (5). Among the Lactobacillus species, only L.
plantarum can produce folate, whereas other species were
shown to consume folate during growth (7).
The introduction of response surface methodology
(RSM) by Box and Wilson (9) gives an alternative tech-
nique to analyze the optimized culture conditions, thus
enabling the researcher to design the experiment, build
the blocks, and evaluate the effects of different growth
factors and responses. This approach allows the re-
searchers to design experiments based on the RSM re-
sults and produce maximum yield of desired products
(10). Furthermore, one of the major constraints involved
in designing new cultivation media is the high number
of experiments involved. RSM, a combination of good
experimental design, regression modelling techniques
and optimization, is a useful tool for process improve-
ment. Therefore, RSM is a valuable tool used prior to
industrial level production (11). Most of the researches
on the optimization of medium formulation and cultural
conditions for biosynthesis of products by LAB were
focused on the conventional method rather than RSM
approach (5,7,12,13).
The present work has been conducted to identify
folate producers amongst our laboratory collection of
LAB using microbiological assay. Folate in a biological
extract is usually determined by a microbiological assay
and this method is highly sensitive and ideal for routine
assaying. The optimization of medium formulation for
the growth and folate biosynthesis by the selected strains
was then investigated through a conventional method
and statistical approach of RSM.
Materials and Methods
Bacterial strains and maintenance
LAB strains were maintained in 5 % (by volume)
glycerol at –80 °C. L. lactis NZ9000 and L. lactis MG1363
were kind gifts of Kees Leenhouts, the Netherlands. L.
plantarum I-UL4 was isolated from local fermented food
(14,15) and L. johnsonii DSM 20553 was purchased from
DSMZ (German Collection of Microorganisms and Cell
Cultures), Braunschweig, Germany.
Growth of LAB cultures
L. lactis NZ9000 and MG1363 were cultivated in
folate-free M17 medium (16) supplemented with 0.5 %
(m/V) glucose. L. plantarum I-UL4 and L. johnsonii DSM
20553 were cultivated in de Man-Rogosa-Sharpe (MRS)
medium (17). All cultivations were carried out in a 250-mL
Erlenmeyer flask containing 100 mL of medium at pH=7.
The LAB strains were incubated at 30 °C at the agitation
speed of 100 rpm for 24 h. The strains were cultivated in
modified MRS medium containing (in g/L): glucose 10,
peptone 10, yeast extract 5, meat extract 5, potassium
hydrogen phosphate 2, sodium acetate 5, triammonium
citrate 2, magnesium sulphate 0.2, manganese sulphate
0.2 and Tween 80 1 mL/L with the addition of para-ami-
nobenzoic acid (PABA, 0.01 mM). All cultivations were
seeded with 5 % (by volume) inoculum. Inocula were
prepared by inoculating a colony of the strains grown
on an M17 or MRS agar plate into 5 mL of M17 or MRS
broth in a 10-mL test tube with continuous shaking (100
rpm) in a water bath at 30 oC for 12 h.
Central composite design
Three variables and five levels were used in this
study. The three variables used were lactose, meat ex-
tract and PABA concentration. The experimental design
for central composite design (CCD) is shown in Table 1.
The results of the CCD were statistically evaluated
and the data were analyzed by Design Expert v. 6.0.6
(Stat-Ease Inc., MN, USA). Each variable was investigat-
ed as the independent and dependent variable. Each in-
dependent variable indicates high level (+1) and low level
(–1). An axial distance ±a was chosen to be 1.4 to make
the design rotatable, which means that the value is at
the same distance from the design centre. The centre
point, or typically known as dependent variable, was in-
dicated as 0 and maintained at constant value, which
provided an unbiased estimate of the process error vari-
ance. This centre point was set at middle point value,
and six centres of this experiment were included to
avoid the missing of a non-linear relationship (18). The
response of folate concentration was subjected to quad-
ratic regression model and expressed by the second-
-order polynomial:
Y=bo+Sbjxj+Sbjjxj2+Sbjkxjxk/1/
where Yis the folate concentration, and bj,bjj and bjk are
linear, quadratic and interactive coefficient, respectively.
The F-value was considered to be significant. The lack of
fit should be non-significant and produce a good multi-
ple correlation coefficient (R2).
Analytical determination
Cell and substrate concentrations
A volume of 6 mL of the sample was withdrawn at
1-hour intervals for the analysis of folate, cell concen-
tration and reducing sugar concentration. A volume of 5
mL of the sample was used for the analysis of intra- and
extracellular folate and 1-mL sample was used for sub-
244 N. MUHAMAD NOR et al.: Folate Biosynthesis by Lactic Acid Bacteria, Food Technol. Biotechnol. 48 (2) 243–250 (2010)
Table 1. Coded and real values of variables selected for CCD
Variable Unit Values of CCD variables
–a–1 0 +1 +a
A lactose g/L 1.89 5 12.5 20 23.11
B meat extract g/L 1.89 5 12.5 20 23.11
C PABA mM1.56 3 6.5 10 11.45
strate (19) and cell concentration analysis. The 1-mL
samples were centrifuged (10 000´g, 10 min, 4 oC) to
separate the cell pellet from the supernatant. The super-
natants were collected for substrate determination and
the absorbance was measured at 540 nm, while the cell
pellets were used for cell concentration determination.
The cell pellets were transferred to a pre-weighed dry
filter paper and dried for 24 h at 100 oC. Dry cell mass
(DCM) values were correlated with absorbance measure-
ments at 600 nm to obtain a calibration standard of A600 nm
vs. DCM. A600 nm readings were subsequently used to deter-
mine DCM.
Folate concentration
Folate concentration was quantified using a Lacto-
bacillus casei microbiological assay. Cells and supernat-
ants recovered from a cell culture were used to measure
both intra- and extracellular folate concentrations as de-
scribed previously (20). L. casei ATCC 7469 was used for
folate bioassay and stored at –80 °C in MRS medium sup-
plemented with 15 % (by volume) of glycerol. In the assay,
L. casei ATCC 7469 was pre-grown in Folic Acid Casei
Medium (Difco Laboratories, Surrey, UK) supplemented
with folate 0.3 mg/L and the culture was incubated for
18 h at 37 °C. Volumes of 1.5 mL of aliquots of the cul-
ture were then stored in sterile tubes at –80 °C until use
for folate determination. For measuring the intracellular
and extracellular folate, a volume of 5 mL of cultivation
broth was used and centrifuged (10 000´g, 10 min, at
20 °C) to separate the cell and the supernatant. The cells
were washed with 0.1 M sodium acetate buffer (pH=4.8)
and1%(byvolume)ofascorbicacidandresuspended
with the same buffer. The samples were then incubated
at 100 °C for 5 min to release the folate from the cells.
The supernatant was diluted 1:1 with 0.1 M sodium ace-
tate buffer (pH=4.8) and 1 % (by volume) of ascorbic acid.
The analysis of total folate concentration, including poly-
glutamyl folate, was conducted after enzymatic deconju-
gation of the folate samples with human plasma (Sigma-
-Aldrich, Malaysia) and incubated for4hat37°C.The
microbiological assay was determined in 96-well micro-
titer plates. The wells were filled by adding 8 mL of work-
ing buffer containing 0.1 M potassium phosphate buffer
with sodium acetate buffer (pH=4.8) and1%(byvol
-
ume) of ascorbic acid to 122 mL of samples or 60 mLof
reference sample, and the volume was increased by ster-
ile distilled H2Ountilitreached150mL and prior to
filling with 150 mL of Folic Acid Casei Medium. The assay
plate was covered and incubated at 37 °C for 18 h. For
reference samples, folate was dissolved in the same buf-
fer at a concentration ranging from 0 to 0.30 mg/L. The
growth of LAB strains in the 96-well microtiter plates
was determined by measuring the absorbance at 620 nm
using the Zenyth Microplate Reader (Biochrom Ltd, Cam-
bridge, UK).
Results and Discussion
Selection of folate producer
The results obtained in this study showed that the
growth performance of folate producer was strain-specific.
All the strains grew well and produced comparable intra-
and extracellular folate. About 7.67 g/L of maximum cell
concentration (Xmax) was achieved in the cultivation of
L. plantarum I-UL4 with a specific growth rate (m)of0.39
h–1 after 18 h of cultivation using MRS medium. The
highest folate level was detected in L. plantarum I-UL4
(36.36 mg/L). Lactobacillus spp. was able to grow well and
produce substantially high folate using chemically de-
fined or complex media, compared to other strains (7).
However, 27.63 mg/L of folate were obtained in the cul-
tivation of L. johnsonii DSM 20553 with the maximum
cell concentration of 7.73 g/L, which was the lowest, com-
pared to the other strains.
Higher growth yield, Yx/s (1.69 g/g) and slightly lower
folate production (30.61 mg/L) were obtained for L. lactis
NZ9000 compared to L. plantarum I-UL4. These results
suggest that the substrate was mainly consumed for the
biomass build-up rather than the biosynthesis of folate.
Furthermore, high cell concentration increased the visco-
sity of the culture, which could reduce the mixing effi-
ciency and hence limit the substrate availability for fo-
late biosynthesis (21,22).
The highest efficiency of cells to produce folate (Yp/x)
was observed in the cultivation of L. lactis MG1363 (6.31
mg/g). It was shown that this species has maximum effi-
ciency in producing folate compared to other strains. From
the results, it can be suggested that the biosynthesis of
folate was non-growth associated process since good
growth was unable to promote higher folate biosynthe-
sis as observed in the cultivation of L. lactis NZ9000 and
L. johnsonii DSM 20553. The highest productivity was ob-
served in the cultivation of L. plantarum I-UL4 (3.71 mg/
(L·h)). Hence, L. plantarum I-UL4 was selected as the su-
perior folate producer and used for further optimization
studies on medium composition via conventional and
RSM approaches.
Effect of carbon sources on folate production
Preliminary optimization study on medium compo-
sition was conducted to determine important factors
that could enhance the biosynthesis of folate. Three dif-
ferent carbon sources (lactose, maltose and glucose) were
selected in this study and L. plantarum I-UL4 grew well
on all of them. The highest biosynthesis of folate (36.19
mg/L) was obtained when lactose was used as the car-
bon source, compared to maltose and glucose. The strain
also showed high efficiency when using lactose as a car-
bon source to produce folate (data not shown). The high-
est productivity of folate was obtained in the cultivation
process using lactose (1.97 mg/(L·h)) compared to glu-
cose (1.53 mg/(L·h)) and maltose (1.13 mg/(L·h)). Lactose
is normally found in milk and it is the preferred carbon
source for the growth of LAB (23).
Effect of nitrogen sources on folate production
The maximum cell concentration (Xmax) was obtained
when yeast extract was used as a nitrogen source (data
not shown). Yeast extract, which consists of nitrogenous
compounds and growth factors, stimulates cell growth
(24,25). However, the best nitrogen source for the high-
est biosynthesis of folate (47.01 mg/L) was obtained when
using meat extract. Slightly lower folate concentration was
obtained when using yeast extract (42.83 mg/L) and pep-
245
N. MUHAMAD NOR et al.: Folate Biosynthesis by Lactic Acid Bacteria, Food Technol. Biotechnol. 48 (2) 243–250 (2010)
tone (43.71 mg/L) as nitrogen sources (data not shown).
Even though the cell efficiency of folate biosynthesis (Yp/x)
was the highest on peptone compared to meat extract
and yeast extract, the highest productivity of folate (1.56
mg/(L·h)) was found on meat extract (data not shown).
Therefore, meat extract is the most suitable nitrogen source
for folate biosynthesis although it is quite an expensive
source of nitrogen, especially for industrial application.
Hence, a precise concentration of meat extract needed in
a medium for higher folate biosynthesis should be fur-
ther investigated.
Optimization of lactose concentration for folate
production
Generally, a higher concentration of lactose in the
medium exhibits an increased cell concentration. In this
study, approx. 7.07 g/L of maximum cell concentration
(Xmax) was achieved using 20 g/L of lactose (Table 2).
However, slightly lower folate concentration was obtained
when more lactose was added to the medium. This in-
dicated that the carbon flux in cells favoured more the
build-up of cell biomass or biosynthesis of by-products
rather than the folate biosynthesis. Excessive carbon
source in the culture broth reduces biomass yield and
growth efficiency due to metabolism overflow. Lactose
concentration of 10 g/L was optimal for the strain to pro-
duce the highest folate concentration (35.94 mg/L) (Table
2). As shown earlier, the biosynthesis of folate was a non-
-growth associated process, therefore the balanced car-
bon flux in cells is required through the folate pathway
and other metabolite pathways (5,7,12).
Optimization of meat extract concentration for folate
production
The growth performance and the kinetic parameter
values of L. plantarum I-UL4 cultivation using meat ex-
tract as a nitrogen source are shown in Table 3. When 15
g/L of meat extract were used, maximum cell concen-
tration (Xmax) of 4.75 g/L and the highest folate biosyn-
thesis (52.82 mg/L) were obtained. The growth of L. plan-
tarum I-UL4 and folate biosynthesis were reduced with
meat extract concentration above 15 g/L in the medium.
Further increment of nitrogen source in the medium
showed an inhibitive effect on the growth performance
of the strain and folate biosynthesis. This finding is not
in agreement with previous studies (7,8) where batch
cultures of L. lactis MG1363 supplemented with growth-
-inhibiting substances increased the folate biosynthesis.
Such discrepancies may be due to the use of meat ex-
tract in the medium, which consists of nitrogenous com-
pounds and growth factor.
Optimization of PABA concentration for folate
production
The addition of PABA into the medium promoted
good growth of L. plantarum I-UL4 and folate biosyn-
thesis. PABA concentration of 0.1 to 10 mM significantly
increased the biosynthesis of folate (Table 4). However,
an addition of above 15 mM of PABA caused cell inhibi-
tion and decreased biosynthesis of folate. At the concen-
tration of 50 to 100 mM of PABA, the growth was signifi-
cantly inhibited and the substrate was not fully utilized
by L. plantarum I-UL4. The strain efficiently produced
high amount of folate (55.02 mg/L) at 10 mM of PABA.
This result suggested that the biosynthesis of folate by L.
plantarum I-UL4 was enhanced by the addition of PABA
to the medium. In the absence of PABA, low folate lev-
els were found to be produced by L. lactis strain (7) and
PABA is an important precursor of folate biosynthesis
(5).
Optimization of lactose, meat extract, and PABA by
central composite design
A second experiment was conducted as a result of
the aforementioned findings, in an attempt to further
optimize the independent levels of lactose, meat extract
and PABA with a face-centred central composite design
(CCD) using Design Expert software. The results of 20
experiments to evaluate the effect of three factors of me-
dium composition that influenced folate biosynthesis are
shown in Table 5. Centre points with a coded value (0)
were repeated six times in order to estimate pure error
for the lack of fit test. Models with a significant lack of
fit should not be used for predictions. The insignificant
lack of fit is most desirable at p>0.1.
246 N. MUHAMAD NOR et al.: Folate Biosynthesis by Lactic Acid Bacteria, Food Technol. Biotechnol. 48 (2) 243–250 (2010)
Table 2. The performance and kinetic parameter values of folate
biosynthesis by L. plantarum I-UL4 using different concentra-
tions of lactose
Kinetic parameter g(lactose)/(g/L)
5 101520
t/h 18181212
Xmax/(g cell/L) 5.75 5.92 7.00 7.07
Pmax/(mg folate/L) 34.22 35.94 31.10 30.47
Si–Sf/(g substrate consumed/L) 3.89 6.84 12.11 15.24
m/h–1 0.37 0.39 0.39 0.34
Yx/s/(g cell/g substrate) 1.48 0.87 0.58 0.46
Yp/s/(mg folate/g substrate) 8.80 5.25 2.57 2.00
Yp/x/(mg folate/g cell) 5.95 6.07 4.44 4.31
Pr/(mg folate/(L·h)) 0.90 1.01 1.12 0.92
Table 3. The performance and kinetic parameter values of folate
biosynthesis by L. plantarum I-UL4 using different concentra-
tions of meat extract
Kinetic parameter g(meat extract)/(g/L)
5 101520
t/h88128
Xmax/(g cell/L) 3.1 4.06 4.75 3.93
Pmax/(mg folate/L) 44.14 49.62 52.82 48.71
Si–Sf/(g substrate consumed/L) 7.89 7.67 7.62 7.52
m/h–1 0.31 0.4 0.32 0.38
Yx/s/(g cell/g substrate) 0.39 0.53 0.62 0.52
Yp/s/(mg folate/g substrate) 5.59 6.47 6.93 6.48
Yp/x/(mg folate/g cell) 14.24 12.22 11.12 12.39
Pr/(mg folate/(L·h)) 2.16 1.25 1.47 1.07
The results from the CCD showed that the optimal
concentrations of lactose, meat extract and PABA for fo-
late biosynthesis were 20 g/L, 20 g/L and 10 mM, re-
spectively. The Design Expert software uses an optimi-
zation method that allows the criteria for all variables
and responses to be set. This optimization method takes
into consideration a combination of criteria in the cal-
culation of the optimum points. Therefore, based on the
setting for criteria of lactose (in range), meat extract (in
range), PABA (in range) and response (maximize), the
optimum point of lactose (20 g/L), meat extract (16.57
g/L) and PABA (10 mM) was suggested.
The maximum response predicted from the model
was 58.51 mg/L. Repeated experiments were performed
to verify the predicted optimum value. A maximum fo-
late concentration (60.39 mg/L) was obtained from run
no. 8. Although the actual experimental response value
at the optimum point was slightly higher than the pre-
dicted value, statistically, there was no significant differ-
ence. The maximum folate concentration obtained from
the optimized medium composition was compared with
the MRS standard medium, and it was apparent that the
optimized medium formulation significantly improved
folate biosynthesis by about twofold. It is interesting to
note that L. plantarum has an ability to produce about 45
mg/L of folate (4) as well as other beneficial metabolites
including bacteriocin (26). In this study, even higher pro-
duction of folate by L. plantarum I-UL4 was obtained.
Regression analysis was performed to fit the response
function with the experimental data. The data obtained
(Table 6) were fitted to a quadratic polynomial model,
and the obtained full actual model is shown in Eq. 2:
Y=52.42+3.93A+2.74B+2.19C–0.22A2–2.78B2+
+0.36C2+0.73AB–0.43AC–0.42BC /2/
where Y represents the predicted responses; A, B and C
are coded values of lactose, meat extract and PABA con-
centration, respectively.
247
N. MUHAMAD NOR et al.: Folate Biosynthesis by Lactic Acid Bacteria, Food Technol. Biotechnol. 48 (2) 243–250 (2010)
Table 4. The performance and kinetic parameter values of folate biosynthesis by L. plantarum I-UL4 using different concentrations of
PAB A
Kinetic parameter c(PABA)/mM
0.1 0.3 1.0 3.0 5.0 7.0 10 15 20 50 100
t/h 121818121812121212 8 6
Xmax/(g cell/L) 3.36 4.42 4.30 4.76 4.09 3.76 4.36 2.74 1.86 1.08 0.95
Pmax/(mg folate/L) 39.89 44.16 45.86 45.90 45.38 49.27 55.02 41.79 39.80 26.82 24.93
Si–Sf/(g substrate consumed/L) 7.36 7.51 7.22 7.15 6.07 6.59 6.62 5.50 5.76 6.13 5.20
m/h–1 0.23 0.19 0.21 0.22 0.21 0.20 0.28 0.21 0.21 0.13 0.15
Yx/s/(g cell/g substrate) 0.46 0.59 0.60 0.67 0.67 0.57 0.66 0.50 0.32 0.18 0.18
Yp/s/(mg folate/g substrate) 5.42 5.88 6.35 6.42 7.48 7.48 8.34 7.60 6.91 4.38 4.79
Yp/x/(mg folate/g cell) 11.87 9.99 10.67 9.64 11.10 13.10 12.66 15.25 21.40 24.83 26.24
Pr/(mg folate/(L·h)) 0.99 1.85 1.38 1.00 0.43 0.95 2.11 0.48 0.34 1.05 1.27
Table 5. Central composite design with a real value and response
to folate concentration (actual and predicted values)
Run g(A)/(g/L) g(B)/(g/L) c(C)/mM
g(folate)/(mg/L)
Obtained
value Predicted
value
1 5 5 3 38.70 40.80
2 20 5 3 50.11 48.06
3 5 20 3 47.94 45.66
4 20 20 3 54.86 55.84
5 5 5 10 47.64 46.88
6 20 5 10 49.91 52.41
7 5 20 10 47.80 50.07
8 20 20 10 60.39 58.51
9 12.5 12.5 6.5 53.00 52.43
10 12.5 12.5 6.5 52.73 52.43
11 12.5 12.5 6.5 51.09 52.41
12 12.5 12.5 6.5 52.73 52.41
13 1.90 12.5 6.5 47.27 46.42
14 23.11 12.5 6.5 57.11 57.52
15 12.5 1.90 6.5 44.14 42.97
16 12.5 23.11 6.5 49.99 50.72
17 12.5 12.5 1.55 49.05 50.03
18 12.5 12.5 11.45 57.64 56.22
19 12.5 12.5 6.5 50.97 52.41
20 12.5 12.5 6.5 53.56 52.41
A–lactose,B–meatextract,C–PABA
Table 6. Regression coefficient and the significance for the re-
sponse of folate biosynthesis
Factor Degree of
freedom
Coeffi-
cient
estimate
Standard
error
Computed
t-distri-
bution p>|t|
Intercept 1 52.42 0.93 56.37 0.0012
A 1 3.93 0.62 6.34 0.0011
B 1 2.74 0.62 4.42 0.0017
C 1 2.19 0.62 3.53 0.0065
A21 –0.22 0.76 –0.29 0.7775
B21 –2.78 0.76 –3.66 0.0052
C21 0.36 0.76 0.47 0.6511
AB 1 0.73 0.76 0.96 0.3632
AC 1 –0.43 0.76 –0.57 0.5826
BC 1 –0.42 0.76 –0.55 0.5955
The student's t-distribution and the corresponding
p-values along with the second order coefficient are
shown in Table 6. p-Value is used as a tool to determine
the significance of each coefficient and higher p-value
indicated higher significance of corresponding coeffi-
cient (10). The parameter was estimated and the corre-
sponding p-values showed that A (lactose), B (meat extract)
and C (PABA) had significant effect on the biosynthesis
of folate. Positive coefficient for A, B and C was indicat-
ed as a linear effect to the response. Table 6 shows some
of the model terms of response of folate biosynthesis (A2,
C2, AB, AC and BC), which had p-value>0.05. Therefore,
they were eliminated from the model and simplified
quadratic model equation most suitably described the
folate biosynthesis as follows:
Y=52.42+3.93A+2.74B+2.19C–2.78B2/3/
The independent variables were fitted to the second
order. Table 7 shows the results of the analysis of vari-
ance (ANOVA) to indicate the adequacy of the fitted
model. The low p-value (p<0.001) indicated that the ob-
tained equation was appropriate and suitable after mod-
el reduction. The determination of R2coefficient, cor-
relation and model significance (F-value) were used to
analyze the adequacy of the model. The quality of fit of
the equation was expressed by the determination coeffi-
cient, R2. Common study showed that a good R2should
be at least 80 %. The value of coefficient R2=0.9063 in-
dicated that the model could explain about 91 % of the
variability and it was attributed to the independent vari-
ables. Model significance (F-value) is a measure of vari-
ation of the data around the mean. Also, the F-value
indicates that the present model can serve as a good
prediction of the experimental results. Central composite
rotatable design (CCRD) analysis was unable to support
a full cubic model; the results from the model were ac-
companied by aliased statement, which indicated that not
all parameters could be uniquely estimated.
Fig. 1 depicts the three-dimensional plot showing the
effect of lactose (A) and meat extract (B) as responses to
folate biosynthesis. Based on the p-value derived from
the CCD analysis (Table 6), it shows no interactions be-
tween these two variables. Biosynthesis of folate was in-
creased when high concentration of meat extract was
added to the medium, while keeping lactose at low con-
centration.
Fig. 2 shows the combined effects of meat extract (B)
and PABA (C) as response to folate biosynthesis. From
the analysis, the independent variables affect the biosyn-
thesis, but there were no interactions between meat ex-
tract and PABA, based on p-value results. This indicates
that all independent variables cannot be interacting due
248 N. MUHAMAD NOR et al.: Folate Biosynthesis by Lactic Acid Bacteria, Food Technol. Biotechnol. 48 (2) 243–250 (2010)
Table 7. Analysis of variance (ANOVA) for the quadratic model of folate biosynthesis
Source Sum of squares Degree of freedom Mean F-value Prob>F
Linear 332.68 3.00 14.82 <0.0001
Cross product 7.16 3.00 0.27 0.8442
Quadratic 63.42 3.00 4.56 0.0331
Cubic 35.64 4.00 7.36 0.0252
Total 52127.26 13.00 2606.36
Pure error 4.86 4.00 1.22
R2=0.9063
43.476
46.9084
50.3409
53.7733
57.2057
Fola
t
e
5.00
8.75
12.50
16.25
20.00
5.00
8.75
12.50
16.25
20.00
A: lactose
B:m
eate
xtract
Fig. 1. Response surface plot of folate biosynthesis as a function
of lactose and meat extract concentrations
45.51
48.2216
50.9332
53.6448
56.3564
Fola
t
e
5.00
8.75
12.50
16.25
20.0
0
3.00
4.75
6.50
8.25
10.00
B:m
eate
xtract
C: PABA
Fig. 2. Response surface plot of folate biosynthesis as a function
of meat extract and PABA concentrations
to the insignificant model of p-value test. Low concen-
tration of meat extract and the addition of high con-
centration of PABA to the medium induced high biosyn-
thesis of folate. The addition of PABA stimulated the
biosynthesis of folate. In L. plantarum I-UL4, folate bio-
synthesis was shown to be dependent on the concentra-
tion of PABA in the medium. Generally, with the addi-
tion of PABA, the biosynthesis of folate was increased
by about twofold compared to the standard MRS me-
dium.
Conclusion
Biosynthesis of folate was strain-dependent and all
investigated LAB strains were able to grow well and pro-
duce folate. The highest folate biosynthesis was obtained
in L. plantarum I-UL4 culture using standard MRS me-
dium. A better understanding of the relationship among
lactose, meat extract and PABA was obtained by RSM,
which was used as a statistical tool to improve the folate
biosynthesis of L. plantarum I-UL4. RSM analyses demon-
strated that optimum biosynthesis of folate can be success-
fully predicted and the combination of lactose (20 g/L),
meat extract (16.57 g/L) and PABA (10 mM) significantly
improved folate biosynthesis. The optimized medium
formulation could be used for the cultivation process of
L. plantarum I-UL4 for folate biosynthesis in a bioreactor
system associated with a process control strategy.
Nomenclature
t: time at maximum cell concentration
Xmax: maximum cell concentration (g cell/L)
Pmax: maximum folate concentration (mg folate/L)
Si–Sf: substrate consumed (g substrate/L)
Si: initial substrate concentration (g substrate/L)
Sf: final substrate concentration (g substrate/L)
m: specific growth rate (h–1)
Yx/s: growth yield coefficient (g cell/g substrate)
Yp/s: folate yield based on substrate utilized
(mg folate/g substrate)
Yp/x: folate biosynthesis per cell (mg folate/g cell)
Pr: folate productivity (mg folate/(L·h))
Acknowledgements
We would like to acknowledge the financial support
from Science Fund, the Ministry of Science, Technology
and Innovation of Malaysia, Project reference: 02-01-04-
SF0330 and Graduate Research Scheme Fellowship (GRF)
from the University of Putra Malaysia for Norfarina
Muhamad Nor.
References
1. K.M. Forssén, M.I. Jägerstad, K. Wigertz, C.M. Witthöft,
Folates and dairy products: A critical update, J.Am.Coll.
Nutr. (Suppl. 2), 19 (2000) 100–110.
2. R.R. Eitenmiller, W.O. Landen: Vitamin Analysis for the
Health and Food Sciences, CRC Press LLC, New York, USA
(1999).
3. R. Póo-Prieto, D.B. Haytowitz, J.M. Holden, G. Rogers, S.F.
Choumenkovitch, P.F. Jacques, J. Selhub, Use of the affi-
nity/HPLC method for quantitative estimation of folic
acid in enriched cereal-grain products, J.Nutr.136 (2006)
3079– 3083.
4. J.G. Le Blanc, G. Savoy de Giori, E.J. Smid, J. Hugenholtz,
F. Sesma: Folate Production by Lactic Acid Bacteria and
Other Food-Grade Microorganisms. In: Communicating
Current Research and Educational Topics and Trends in Applied
Microbiology, A. Méndez-Vilas (Ed.), Formatex, Badajoz,
Spain (2007) pp. 239–339.
5. A. Pompei, L. Cordisco, A. Amaretti, S. Zanoni, D. Matte-
uzzi, M. Rossi, Folate production by bifidobacteria as a po-
tential probiotic property, Appl.Environ.Microbiol.73 (2007)
179–185.
6. R.G. Crittenden, N.Z. Martinez, M.J. Playne, Synthesis and
utilization of folate by yogurt starter cultures and probiot-
ic bacteria, Int.J.Food Microbiol.80 (2003) 217–222.
7. W. Sybesma, M. Starrenburg, L. Tijsseling, M.H.N. Hoefna-
gal, J. Hugenholtz, Effects of cultivation conditions on fo-
late production by lactic acid bacteria, Appl.Environ.Mi-
crobiol.69 (2003) 4542–4548.
8. W. Sybesma, M. Starrenburg, M. Kleerebezem, I. Mierau,
W.M. de Vos, J. Hugenholtz, Increased production of folate
by metabolic engineering of Lactococcus lactis,Appl.Envi-
ron.Microbiol.69 (2003) 3069–3076.
9. G.E.P. Box, K.B. Wilson, On the experimental attainment of
optimum conditions, J.Roy. Stat.Soc.B,1(1951) 1–45.
10. D.C. Montgomery, R.H. Myers: Response Surface Methodo-
logy: Process and Product Optimization Using Designed Expe-
riments, John Wiley & Sons, Inc, New York, USA (1997) pp.
427–510.
11. N.N. Nawani, B.P. Kapadnis, Optimization of chitinase pro-
duction using statistics based experimental designs, Pro-
cess Biochem.40 (2005) 651–660.
12. M.Y. Lin, C.M. Young, Folate levels in cultures of lactic acid
bacteria, Int.Dairy J.10 (2000) 409–413.
13. S. Sirisansaneeyakul, T. Luangipat, W. Vanichsriratana, T.
Srinophakun, H.H.H. Chen, Y. Chisti, Optimization of lac-
tic acid production by immobilized Lactococcus lactis IO-1,
J.Ind.Microbiol.Biotechnol.34 (2007) 381–391.
14. H.L. Foo, T.C. Loh, F.L. Law, Y.Z. Lim, C.N. Kufli, G. Ru-
sul, Effects of feeding Lactobacillus plantarum I-UL4 isolat-
ed from Malaysian tempeh on growth performance, faecal
flora and lactic acid bacteria and plasma cholesterol con-
centrations in post-weaning rats, Food Sci.Biotechnol.12 (2003)
403–408.
15. H.L. Foo, T.C. Loh, P.W. Lai, Y.Z. Lim, C.N. Kufli, G. Ru-
sul, Effects of adding Lactobacillus plantarum I-UL4 meta-
bolites in drinking water of rats, Pak. J.Nutr.2(2003) 283–
288.
16. B.E. Terzaghi, W.E. Sandine, Improved medium for lactic
streptococci and their bacteriophages, Appl.Environ. Micro-
biol.29 (1975) 807–813.
17. J.C. De Man, M. Rogosa, M.E. Sharpe, Cultivation medium
for lactobacilli, J.Appl.Biotechnol.21 (1977) 130–135.
18. L. Rodrigues, J. Teixeira, R. Oliveira, H.C. van der Mei,
Response surface optimization of the medium components
for the production of biosurfactants by probiotic bacteria,
Process Biochem.41 (2006) 1–10.
19. G.L. Miller, Use of dinitrosalycilic acid reagent for deter-
mination of reducing sugar, Anal.Chem.31 (1959) 426–428.
20. D.W. Horne, D. Patterson, Lactobacillus casei microbiologi-
cal assay of folic acid derivatives in 96-well microtiter pla-
tes, Clin.Chem.34 (1988) 2357–2359.
21. A.R. Berry, C.M.M. Franco, W. Zhang, A.P.J. Middelberg,
Growth and lactic acid production in batch culture of Lac-
tobacillus rhamnosus in a defined medium, Biotechnol.Lett.
21 (1999) 163–167.
249
N. MUHAMAD NOR et al.: Folate Biosynthesis by Lactic Acid Bacteria, Food Technol. Biotechnol. 48 (2) 243–250 (2010)
22. M. Rosfarizan, A.B. Ariff, M.A. Hassan, M.I.A. Karim, Kojic
acid production by Aspergillus flavus using gelatinized and
hydrolyzed sago starch as carbon sources, Folia Microbiol.
43 (1998) 459–464.
23. B. Bibal, G. Goma, Y. Vayssier, A. Pareilleux, Influence of
pH, lactose and lactic acid on the growth of Streptococcus
cremoris: A kinetic study, Appl.Microbiol.Biotechnol.28 (1988)
340–344.
24. M.S. Kalil, H.S. Alshiyab, W.M.W. Yusoff, Effect of nitro-
gen source and carbon to nitrogen ratio on hydrogen pro-
duction using C. acetobutylicum,Am.J.Biochem.Biotechnol.
4(2008) 393–401.
25. V.M.G. Lima, N. Krieger, M.I.M. Sarquis, D.A. Mitchell, L.P.
Ramos, J.D. Fontana, Effect of nitrogen and carbon sources
on lipase production by Penicillium aurantiogriseum, Food
Technol.Biotechnol.41 (2003) 105–110.
26. S.D. Todorov, L.M.T. Dicks, Effect of growth medium on
bacteriocin production by Lactobacillus plantarum ST194BZ,
a strain isolated from boza, Food Technol.Biotechnol.43 (2005)
165–173.
250 N. MUHAMAD NOR et al.: Folate Biosynthesis by Lactic Acid Bacteria, Food Technol. Biotechnol. 48 (2) 243–250 (2010)
