Influence of precursors and inhibitor on the production of extracellular 5-aminolevulinic acid and biomass by rhodopseudomonas palustris kG31

Article abstract of DOI:10.1271/bbb.80682

5-Aminolevulinic acid (ALA) and the biomass of photosynthetic bacteria, Rhodopseudomonas palustris KG31, have very high potential for development and exploitation as bioherbicide and biofertilizer respectively. In this work, the effects of two precursors

Full text of DOI:10.1271/bbb.80682

Biosci. Biotechnol. Biochem., 73 (5), 987–992, 2009
Influence of Precursors and Inhibitor on the Production
of Extracellular 5-Aminolevulinic Acid and Biomass
by Rhodopseudomonas palustris KG31
y
Angkana SAIKEUR,1 Wanna CHOORIT,1; Poonsuk PRASERTSAN,2
4
Duangporn KANTACHOTE,3 and Ken SASAKI
1School of Agricultural Technology, Walailak University, Nakhon Si Thammarat, 80161, Thailand
2Faculty of Agro-Industry, Prince of Songkla University, Songkhla, 90112, Thailand
3Faculty of Science, Prince of Songkla University, Songkhla, 90112, Thailand
4Faculty of Engineering, Hiroshima International School, Hiroshima 739-0321, Japan
Received September 30, 2008; Accepted January 20, 2009; Online Publication, May 7, 2009
[doi:10.1271/bbb.80682]
5-Aminolevulinic acid (ALA) and the biomass of
photosynthetic bacteria, Rhodopseudomonas palustris
KG31, have very high potential for development and
exploitation as bioherbicide and biofertilizer respec-
tively. In this work, the effects of two precursors and an
inhibitor of aminolevulinic dehydratase (ALAD) added
to the VFA culture medium on the production of ALA
and biomass were investigated. The experimental runs
were carried out according to a Box-Behnken design.
The precursors were added to the medium at the
beginning of cultivation, while the inhibitor was added
after 24 h. Statistical analysis indicated that levulinic
acid (LA) has a positive effect on ALA production while
glycine has a negative effect on biomass production. In
order to enhance both ALA and biomass products, the
most suitable medium was VFA medium supplemented
with 3.0 mM glycine and 10 mM LA, giving ALA and
biomass of 182.91 ꢀM and 3.1 gDCW/l within 54 h.
ties are suitable for application of the culture broth
containing cells and their products in the agricultural
field. Biomass of photosynthetic bacteria has been used
as a source of protein with high amounts of essential
amino acids e.g., methionine and lysine,5,6) and also as a
feed supplement for producing low-cholesterol chicken
eggs.7) ALA can be applied as a plant growth-regulator
and as a weed growth inhibitor,1) in which a concen-
tration range of 0.06–0.6 mM demonstrated plant
regulating activities8) whereas 4 mM ALA could be
used to control weed (Trifolium repens) growth.9)
Application of ALA to date palm revealed that the fruit
chlorophyll content at the Khalal stage was significantly
increased.10)
Despite intensive study of the optimal conditions for
ALA production by adding precursor and inhibitor to the
medium, most of these works have been carried out by
traditionally optimization process (the one-at-a-time
strategy) and were not focused on biomass formation.
Hence in this study, the Box-Behnken design was
applied, whereby the interactions among the factors
were included. A photosynthetic bacterium strain,
KG31, was selected for this study based on our previous
finding that the strain assimilated volatile fatty acids
(acetic acid, propionic acid, butyric acid, and valeric
acid), and grew at wide pH (4.5–8.0) and temperature
ranges (30–42 ꢀC). Thus, these properties might be
useful for the production of ALA and biomass from
wastewater containing volatile fatty acids.
Key words: 5-aminolevulinic acid; Box-Behnken de-
sign; response surface methodology
ALA is an intermediate product found during the
biosynthesis of tetrapyrroles such as heme, chlorophyll,
and vitamin B12. Generally, two distinct metabolic
pathways of ALA synthesis, the Shemin or C4 and
the Beale or C5 pathways have been described.1) In
prokaryotes, such as purple nonsulfur photosynthetic
bacteria, usually ALA is synthesized via the C4 pathway
in which succinyl-CoA and glycine are used as the
precursors, and aminolevulinic synthase (ALAS) cata-
lyzes these two precursors to ALA. Next, two molecules
of ALA are condensed to porphobilinogen (PBG) by the
action of ALAD.2,3) Thus, in the strategy for enhancing
ALA from the bacteria, the precursors of ALAS and the
competitive inhibitor of ALAD, for example LA and
glucose are added.4)
In this work, the influences of succinate, glycine, and
LA on ALA and biomass products of the photosynthetic
bacterium R. palustris KG31 were evaluated. The
results of the design were analyzed and suitable
conditions for ALA and biomass production were
selected after evaluation of the data by the response
surface method (RSM).
Materials and Methods
Among ALA-producing microorganisms, photosyn-
thetic bacteria have received much attention. Besides the
production of ALA, biomass, and pigments, the bacteria
are also able to fix atmospheric nitrogen. These proper-
Media and culture conditions. R. palustris KG31 was isolated from
soil collected from a paddy field in Nakhon Si Thammarat Province,
Thailand, and identified using morphological, biochemical, and
y
To whom correspondence should be addressed. Tel: +66-75-672355; Fax: +66-75-672302; E-mail: cwanna@wu.ac.th
Abbreviations: ALA, 5-aminolevulinic acid; ALAD, aminolevulinic dehydratase; ALAS, aminolevulinic synthase; ANOVA, analysis of variance;
DCW, dry cell weight; df, degree of freedom; LA, levulinic acid; MS, mean square; PBG, porphobilinogen; RSM, response surface method; SS, sum
of squares

988
A. SAIKEUR et al.
Table 1. Level and Code of Variables Chosen for the Box-Behnken
Design
Table 2. Box-Benkhen Design with the Actual and Predicted Values
of ALA Production by R. palustris KG31
Coded levels (mM)
Design
ALA at 48 h (mM)
Actual Predicted
Variable
Coded
Run no.
ꢂ1
0
1
X1
X2
X3
Succinate
Glycine
LA
X1
X2
X3
0
0
0
5
5
5
10
10
10
1
2
10
10
0
5
0
0
5
9.67
38.32
134.47
64.00
47.30
118.75
10.33
119.54
56.20
38.24
61.61
15.28
10.28
106.20
63.49
0.76
42.18
118.40
112.81
53.99
76.97
6.34
3
5
10
10
5
4
5
10
10
10
0
5
10
0
6
5
physiological properties and 16S rDNA sequencing by Pichit Chodok,
a master student of Biotechnology Program, School of Agricultural
Technology, Walailak University, Thailand.
7
5
0
8
10
5
5
10
5
95.42
59.58
65.16
59.58
23.74
18.15
101.00
59.58
9
5
The GM medium was used to maintain the bacterial cells and to
prepare a starter culture. The medium contained 2.0 g/l of yeast
extract, 2.7 g/l of malate, 3.7 g/l of monosodium glutamate, 0.8 g/l of
10
11
12
13
14
15
0
0
5
5
5
5
0
5
0
.
(NH4)2SO4 7H2O, 0.5 g/l of KH2PO4, 0.5 g/l of K2HPO4, 0.2 g/l
.
5
10
0
0
of MgSO4 7H2O, 0.053 g/l of CaCl2 2H2O, 1:2 ꢁ 10ꢂ3 g/l of
.
5
10
5
MnSO4 7H2O, 1:0 ꢁ 10ꢂ3 g/l of thiamine-HCl, 1:0 ꢁ 10ꢂ3 g/l of
.
5
5
nicotinic acid, and 1:0 ꢁ 10ꢂ5 g/l of biotin.11) The initial pH of this
medium was adjusted to 6.5 with 6 M NaOH.
The data in the table under ‘‘Actual’’ represent means of duplicate
experiments.
The VFA medium, used as a test medium, had a composition
similar to the GM medium, except that the yeast extract was 0.5 g/l,
and 1.8 ml/l acetic acid, 1.0 ml/l butyric acid, and 0.2 ml/l propionic
acid were added instead of malate.12) The initial pH of this medium
was adjusted to 7.0 with 6 M NaOH.
The linear model for predicting the optimal point was expressed
according to the equation:
Y (or Z) ¼ b1X1 þ b2X2 þ b3X3
The quadratic second order polynomial equation was as follows:
ð1Þ
Starter cultures were prepared by cultivating R. palustris KG31 in
375-ml flat bottles containing 350-ml the GM medium, and incubated
under the static condition with light intensity of 5,500 lux at 35 ꢃ 1 ꢀC
for 48 h. Then the cells were harvested by centrifugation at 12,000 rpm
for 15 min (RC 5 plus, Sorvell, CT, USA). To prepare the inoculums,
Y (or Z) ¼ b0 þ b1X1 þ b2X2 þ b3X3 þ b12X1X2 þ b13X1X3
2
2
2
þ b23X2X3 þ b11X1 þ b22X2 þ b33X3
ð2Þ
the cell pellets were washed twice and re-adjusted to
a
cell
The coefficient, b0 is the free or off-set term called the intercept.
The terms b1, b2, and b3 are linear coefficients; b12, b13 and b23 are
cross-product coefficients, and b11, b22, and b33 are the quadratic
coefficients.
concentration (OD660) of 1.0 with the VFA medium. The batch
fermentation was performed by inoculating 10% of a starter culture to
a 375-ml flat bottle containing 350-ml of VFA medium plus succinate,
glycine, and LA, as described below. Then the vessels were incubated
as previously described.
The difference between means was evaluated by Duncan’s Multiple
Range Test and P < 0.05 was considered as significant.
Experimental design. Succinate and glycine are the precursors for
ALA production, and LA is a competitive inhibitor of ALAD.13) These
chemical agents were added to the VFA medium at three levels (low,
medium, and high), and coded (ꢂ1, 0 and +1) (Table 1), according to
the Box-Behnken design.14) In this study, the design consisted of
triplicate at the center point and the points as shown in Table 2. A total
of 15 experimental runs were carried out, the precursors were added to
the VFA medium at time 0 h; then the inhibitor was added after 24 h
the cultivation. The experiments were carried out in duplicate. The
actual results are shown in Table 2.
Analytical methods. Biomass was measured as turbidity using a
spectrophotometer (UV-1601, Schimadzu, Kyoto, Japan) at a wave-
length of 660 nm and correlated with the dry cell weight (DCW)
measured by drying the cells at 105 ꢀC for 12 h. Extracellular ALA was
determined by the colorimetric method,15) with the following mod-
ification: The cells were centrifuged at 10,000 rpm for 20 min at 4 ꢀC.
Then 1-ml of the clear supernatants was mixed with 2-ml of 1 M
sodium acetate buffer, pH 4.7, and 50-ml of acetyl acetone. After the
mixtures were boiled for 15 min and cooled, 3.5-ml of modified
Ehrlich’s reagent was added to each sample. The ALA formed was
measured at a wavelength of 553 nm. The residual volatile fatty acids
were determined with HPLC (Waters, MA, USA) with reflective index
detector (Waters, MA, USA). The HPLC condition was as follows:
column, metaCarb H plus column (Varian, CA, USA); temperature,
68 ꢀC; mobile phase, 8.5 mM H2SO4; and flow rate, 0.4 ml/min. The
concentrations of residual volatile fatty acids were calculated from the
peak areas. The pH value was measured using a pH meter (CyberScan
pH 510, Ayer Rajah Crecent, Singapore).
The optimal concentrations of the succinate, glycine, and LA were
optimized by the response surface method (Design-Expert Software,
Stat-Ease, Minneapolis, USA). The three variables (succinate, glycine,
and LA) selected for the statistical analysis were designed X1, X2, and
X3 respectively. As for the predicted responses, Y stands for ALA yield
whereas Z stands for biomass yield. First, the data were entered into
the Design-Expert Software. This software contains sequential tech-
niques that were used for evaluation of the effect of the variables on
the response. Secondly, selecting an appropriate type of model for
the response was done by testing the sequential F-tests, starting with
a linear model and adding terms (quadratic, and higher if appropriate).
Finally, the F-statistic was calculated for each model, and the highest
order model with significant terms was chosen. Tests for significant
sequential models, model equation, and model terms were performed
by employing analysis of variance (ANOVA). The quality of fit of
the model equation was expressed by the coefficient of determination
(R2) and adjusted R2. The software was used for regression analysis
of the experimental data and also to plot the response surface graphs.
The optimal values of the ALA and biomass yields of the experi-
mental conditions were obtained by analyzing the response surface
contour plots. Then the response surface at in-range was selected to
achieve the maximal values of ALA and biomass. Finally, experi-
mental runs were performed to check the validity of the selected
experiments.
Results and Discussion
ALA production
Biochemical data showed that the bacterium utilized
various types of organic carbon such as acetic acid,
propionic acid, butyric acid, glycine, and succinate (data
not shown). Acetic acid, propionic acid, and butyric acid
were supplied as sources of carbon and electron donor
while glycine and succinate were added as precursors to
increase the ALA yield.
The ALA yields of the 15 experimental runs are given
in Table 2. Cells growing in the medium supplemented
with 5 mM glycine and 10 mM LA (run no. 3) produced

Production of ALA and Biomass by Rhodopseudomonas palustris
989
Table 3. Analysis of Variance (ANOVA) for the Fitted Linear Model
for Optimization of ALA Production
the highest ALA yield, at least 8 to 13 fold higher than
cells growing in the medium without added LA (runs
nos. 1, 7, 12, and 13). This indicates that ALA was
greatly enhanced after the addition of LA. It should be
noted that at 24 h, before the addition of LA, the pH
value of all the experimental runs was 6.2–7.6; after
addition, LA, the pH value dropped about 0.5 units. The
pH value of experimental runs nos. 3, 6, 8, and 14
containing ALA in range of 106.20–134.47 mM was 6.6–
7.6, while the pH value of experimental run nos. 1, 7, 12,
and 13 (9.67–15.28 mM ALA) was 8.1–8.4. Possibly not
only LA but also the pH value caused a positive effect in
enhancing the ALA yield. As reported by Lee et al.,
ALAD was inhibited under slight acidic environments.3)
This study aimed to investigate the combined effects
of the selected variables (succinate, glycine, and LA)
to the responses (ALA and biomass), to find the best
models to represent ALA and biomass production and
the optimal concentrations of the selected variables to
the responses. Hence the response surface method was
used. The RSM is a collection of mathematical and
statistical techniques that are useful for the modeling
and analysis of problems in which a response of interest
is influenced by several variables and the objective is to
optimize this response.16) Box et al. mentioned that the
RSM has been used to answer questions: (1) How is a
particular response affected by a given set of input
variables over some specified region of interest? (2)
Which of the inputs gives a product simultaneously
satisfying the desired specifications. And (3), what
values of the inputs yield a maximum for a specific
response, and what is the response surface like close
to this maximum.17) In order to answer, a suitable
approximation for the true functional relationship
between the response and the set of variables should
be found. Normally, that relationship is explained by the
selected model (linear, quadratic, cubic, or so on).
For ALA production, an ANOVA table for the
sequential model sum of squares showed that the linear
model is the highest order model with significance terms
(Prob > F is 0.0010) whereas the quadratic and cubic
models are the Prob > F of 0.7012 and 0.0130
respectively (data not shown). Thus the linear model
represented by regression Eq. (3) for ALA production
by R. palustris KG31 was given as follows:
Source
Model
Residual
df
SS
MS
P-Value
3
11
9
19256.86
6039.47
6010.82
28.56
6418.95
549.04
667.87
14.32
0.0010
Lack of fit
Pure error
Corrected total
0.0212
2
13
25296.33
R2 ¼ 0:7613, adjusted R2 ¼ 0:6961
Table 4. Coefficient of the Response Function to Predicted ALA
Formation
Variable
Coefficient
Standard error
F-Value
P-Value
b0
b1
b2
b3
59.58
ꢂ11.49
5.91
6.05
8.28
8.28
8.28
11.69
1.92
0.51
0.0010
0.1930
0.4908
0.0001
47.33
32.64
The predicted ALA yield was obtained by solving
regression Eq. (3). The analysis showed that the max-
imum ALA yield was 118.40 mM (run no. 3) (Table 2).
Equation (3) indicated that in order to enhance the ALA
yield, glycine and LA should be supplied to the medium.
This also showed that another precursor, succinate, was
available for incorporation with added glycine, leading
to ALA formation by the action of ALAS.18)
It has long been known that the precursors for ALA
formation via the C4 pathway are succinate and glycine.
This work demonstrated that glycine produced by
R. palustris KG31 was a limiting factor as compared
with succinate. In order to improve the ALA yield,
glycine should be added to the culture medium.
However, in this experiment, the effect of glycine was
not as big as LA. As reported by Nishikawa et al.,
R. sphaeroides CR-286 produced 1.5 mM ALA in the
presence of 50 mM glucose, 60 mM glycine, 15 mM LA,
and 1.0% (w/v) yeast extract.19) Fu et al. reported that
high initial concentrations of glycine and succinate did
not improve the ALA yield produced from recombinant
E. coli containing the R. sphaeroides hemA gene.20) In
batch fermentation, the addition of glucose and glycine
was effective to improve ALA production.18) But Chung
et al. found that the optimal condition for ALA
production of recombinant E. coli harboring hemA from
Bradyrhizobium japanicum was 15 mM glycine and
30 mM succinic acid.4) Qin et al. found that the highest
ALA yield of recombinant E. coli harboring hemA from
Agrobacterium radiobacter was achieved in a medium
containing glycine, succinic acid, and glucose, and by
adding LA at the end of the log phase.2)
Y (mM) ¼ 17:83 ꢂ ð2:298X1Þ þ ð1:181X2Þ þ ð9:466X3Þ
ð3Þ
where the variables take their coded values, and
represent ALA yield (Y) as a function of succinate
(X1), glycine (X2), and LA (X3).
Table 3 shows the ANOVA for response surface
linear model. The ANOVA of linear model of Eq. (3)
clearly demonstrates that the model was significant, as
can be seen from the low probability value (P value ¼
0.0010). The soundness of the model can be checked by
the determination coefficient, R2, and the adjusted R2.
The value of adjusted R2 suggested that the total
variation of 69.61% for the ALA was to be attributed
to the independent variables and about 31.39% of the
total variation could not be explained by the model. The
coefficient estimates of Eq. (3) and the corresponding
P-value are shown in Table 4. The P-value of variable
b3 was 0.0001. This result indicates a big effect of LA.
Biomass production
R. palustris biomass was reported to be used as
sources of biofertilizer and Single Cell Protein.21)
Moreover, the bacteria are able to fix molecular nitro-
gen.22) Accordingly, the influence of added succinate,
glycine, and LA on the biomass yield was evaluated.
Table 5 shows the biomass yield of 15 experimental
runs. The data obtained were analyzed as mentioned
above. The highest biomass was obtained in run no. 14
(2.6 gDCW/l), while runs nos. 4, 5, and 13 gave low
biomass content as compared to the others. The F tests

990
A. SAIKEUR et al.
Table 5. Box-Benkehn Design with the Actual and Predicted Values
of Biomass for R. palustris KG31
Table 7. Coefficient of the Response Function to Predicted Biomass
Formation
Biomass at 48 h
Design
Variable
Coefficient
Standard error
F-Value
P-Value
(gDCW/l)
Run no.
b0
2.26
ꢂ0.03
ꢂ0.71
0.18
0.13
0.08
0.08
0.08
0.12
0.12
0.12
0.12
0.12
0.12
14.94
0.14
76.71
4.76
6.80
1.82
5.92
1.42
5.37
34.42
0.0041
0.7274
0.0003
0.0809
0.0478
0.2357
0.0591
0.2874
0.0683
0.0020
X1
X2
X3
Actual
Predicted
b1
b2
1
2
10
10
0
5
0
0
5
1.2
2.6
1.3
0.4
0.8
1.7
1.6
1.8
2.2
2.3
2.3
1.4
0.5
2.6
2.2
1.1
2.8
1.5
0.5
0.8
1.5
1.5
1.8
2.3
2.3
2.3
1.4
0.7
2.5
2.3
b3
b12
b13
b23
b11
b22
b33
ꢂ0.30
0.16
3
5
10
10
5
4
5
10
10
10
0
ꢂ0.28
ꢂ0.14
ꢂ0.28
ꢂ0.70
5
10
0
6
5
7
5
0
8
10
5
5
10
5
9
5
10
11
12
13
14
15
0
0
5
5
5
5
0
5
0
5
10
0
0
5
10
5
5
5
The data in the table under ‘‘Actual’’ represent means of duplicate
experiments.
Table 6. Analysis of Variance (ANOVA) for the Fitted Quadratic
Model for Optimization of Biomass
Source
Model
Residual
df
SS
MS
P-Value
9
5
7.12
0.26
0.25
0.01
7.38
0.79
0.05
0.08
0.01
0.0041
Lack of fit
Pure error
Corrected total
3
0.0639
2
14
R2 ¼ 0:9641, adjusted R2 ¼ 0:8996
are performed as mentioned above. Thus the following
regression equation for biomass yield (Z) was obtained.
The quadratic Eq. (4) is presented.
Fig. 1. Contour Plot of the Interactions of Succinate and Glycine in
Biomass Yield.
Z (gDCW/l) ¼ ð1:278Þ þ ð0:08X1Þ þ ð0:085X2Þ
þ ð0:341X3Þ ꢂ ð0:012X1X2Þ
The coefficient estimates of Eq. (4) and the corre-
sponding P-value are shown in Table 7. The data imply
that glycine (b2) and an interaction of glycine and
succinate (b12) were significant. The P-value of variable
b2 was 0.0003. The P-value of variable b12 was 0.0478.
These results clearly indicate a big effect of glycine. A
contour curve was drawn by fixing the LA concentration
at 5 mM, so that the interaction between succinate and
glycine on the biomass yield during cultivation of the
bacterium for 48 h was obtained (Fig. 1). Figure 1
implies that if succinate was not added or was added
to 10 mM, supplementation of glycine >5 mM would
result in a decrease in biomass contents. In order to
receive about 2.5 gDCW/l, about 3–6 mM succinate
should be added.
Hence, growth and ALA production did not depend
on the same factor. LA had the highest effect on ALA
formation. In the case of glycine, it also had influence on
the ALA yield. Moreover, the addition of succinate
together with glycine resulted in a decrease in biomass
production. Besides, LA lowered the pH value of the
medium and it can also be used as a carbon source for
the bacterium (data not shown).
þ ð0:006X1X3Þ ꢂ ð0:011X2X3Þ
2
2
ꢂ ð0:006X1 Þ ꢂ ð0:011X2 Þ
2
ꢂ ð0:028X3 Þ
ð4Þ
where Z (biomass, gDCW/l) is the predicted response
variable; X1 ꢂ X3 = the actual values of the independ-
ent variables, viz., succinate, glycine, and LA respec-
tively.
A statistical analysis of the process parameters along
with the biomass yields as response values is shown
in Table 5. The optimal biomass of the predicted value
was 2.8 gDCW/l (run no. 2). Run no. 14 achieved a
predicted value of the biomass of 2.5 gDCW/l. The
statistical significance of Eq. (4) was checked by the
F-test, and the ANOVA for the response surface quad-
ratic model is summarized in Table 6. The ANOVA of
the quadratic model of the equation demonstrated that
the model was significant, as can be seen from the low
probability values (P value ¼ 0:0041). The value of
adjusted R2 suggested that the total variation of 89.96%
for the biomass is to be attributed to the independent
variables and about 10.04% of the total variation cannot
be explained by the model.
An attempt to add these precursors at 12, 24, and 36 h
of cultivation time (supplementation of LA at 24 h)
showed that adding times did not increase the ALA

Production of ALA and Biomass by Rhodopseudomonas palustris
Table 8. Concentrations of Succinate, Glycine, and LA Used in the
991
a
Verification Test
200
175
150
125
100
75
Concentration (mM)
Glycine
Vessel
Succinate
LA
Run A
Run 3
0
0
5
0
3
5
0
0
10
10
10
0
Run 14
Control
50
25
concentration significantly. On the other hand, the
addition of LA at 12, 24, and 36 h (supplementation of
the precursors at 0 h) had an effect on ALA formation
(data not shown). Time 24 h, at which the cells growth
entered an early log phase, was the best for adding LA.
0
0
24
30
36
42
48
54
60
66
72
Incubation time (h)
b
6.0
5.0
4.0
3.0
2.0
1.0
0.0
Verification the models
As mentioned above, after cultivation of the bacte-
rium for 48 h, the data of runs nos. 3 and 14 gave the
highest ALA and biomass yields respectively (Table 2
and Table 5). The application of RSM and statistical
analysis clearly confirmed that different factors influ-
enced ALA and biomass production. In order to enhance
both products, response surface method analysis was
performed. The results showed that the most suitable
medium for ALA and biomass production was VFA
medium supplementation with 3.0 mM glycine and
10 mM LA (Table 8).
0
24
30
36
42
48
54
60
66
72
Incubation time (h)
Fig. 2. Time Profiles of ALA Yields (a) and Biomass Yields (b) of
Run A ( ), Run no. 3 ( ) Run no. 14 ( ) and the Control ( ).
Figure 2a shows the levels of ALA along with the
incubation times. The ALA profiles increased with
increasing incubation time, especially in run no. 3 and
run A. At 48 h, the yield of ALA was 8.06, 164.24,
115.00, and 180.21 mM in the control, runs nos. 3 and 14
and run A respectively. The predicted ALA yield in
runs nos. 3 and 14 and run A was 118.40, 101.00, and
116.30 mM respectively. This experimental finding is
in agreement with the model’s prediction. Duncan’s
Multiple Range Test was used to evaluate these results,
and it was found that the amount of ALA produced in the
control experiment was significantly different (p < 0.05)
from the other runs. At 54 h, the ALA yields of run no. 3
(194.70 mM) and run A (182.96 mM) were significantly
different (p < 0.05) from those of run no. 14 (101.79
mM). Thus, the results of run no. 3 and run A indicated
that succinate was not a limited precursor while glycine
and LA were needed for ALA production by the cells.
Also this indicated a major effect on LA toward ALA
formation. It should be noted that in the course of the
experiments, the ALA contents of run no. 3 and run A
were not significant different. This may be because the
optimization experiment was focused on receiving both
ALA and biomass yields. Figure 2a also shows that the
levels of ALA were drastically decreased after 60 h.
Moat and Foster stated that succinate is formed by
reducing malate to fumarate and then, to the succinate in
the Krebs cycle under anaerobic conditions.23)
The growth of the bacterium under the optimal
conditions is shown in Fig. 2b. In run no. 14, growth
of the bacterial cell rapidly developed as compared with
run no. 3 and run A. At 48 h, the biomass yield in the
control, runs nos. 3 and 14, and run A was 1.7, 2.1, and
2.8 and 2.3 gDCW/l respectively (Fig. 2b). However,
the predicted yield of biomass was 1.5, 2.5, and
1.7 gDCW/l, for runs nos. 3 and 4 and run A respec-
tively. At 54 h, the biomass of runs nos. 3 and 14, and
run A was 2.9, 3.4, and 3.1 gDCW/l respectively. At
60 h, the biomass was in a range of 3.8–4.1 gDCW/l.
The biomass yields obtaining at 48 and 54 h under the
experiments of all runs and control were not signifi-
cantly different (p < 0.05).
The volatile fatty acid profiles of run A rapidly
decreased as compared with runs nos. 3 and 14. At 48 h,
acetic acid was 0.88, 1.23, and 0.81 g/l in runs nos. 3
and 14, and run A respectively; butyric acid was 0.71,
0.72, and 0.46 g/l in runs nos. 3 and 14, and run A
respectively (Fig. 3). After 18 h, propionic acid was
completely exhausted (data not shown). Normally, the
profiles of acetic acid were decreased along the
cultivation time but at the end of cultivation (72 h) the
levels of acetic acid of run no. 14 and run A were
slightly increased (Fig. 3a). This is because acetic acid
was formed during the assimilation of propionic and
butyric acids by the bacterium (data not shown).
Figure 3 shows that the volatile fatty acids contained
in run no. 14 were consumed lower than run no. 3 and
run A. This may be because the bacterium preferred
to utilize succinate as a carbon source instead of
volatile fatty acids. This result agrees with the studies
of Qin et al., who reported that succinate serves as a
carbon source for cell growth of the bacterium.2)
ALA production by this bacterium was relatively low
as compared with those previously reported. However,
the utilization of volatile fatty acids which often create
an odor problem and the utilization of light energy are of
the best points for economic wastewater treatment with
ALA production. It should be emphasized that the ALA
content obtained in this study was enough to be used to
promote plant growth.8)

992
A. SAIKEUR et al.
produced by R. palustris KG31 in VFA medium under
a
illumination conditions. Application of the RSM and
statistical analysis led us to find the optimal ALA and
biomass yields in response to the affecting factors,
glycine and LA. Thus, the ALA of 183.0 mM and the
biomass of 3.1 gDCW/l were achieved within 54 h,
cultivation time in VFA medium supplemented with
3.0 mM glycine and 10 mM LA.
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
Acknowledgment
0
24
30
36
42
48
54
60
66
72
The authors are grateful to Walailak University for
providing financial (WU51101) support (to W. C.) to
carry out this experiment.
Incubation time (h)
b
1.2
1.0
0.8
0.6
0.4
0.2
0.0
References
1) Sasaki K, Watanabe M, Tanaka T, and Tanaka T, Appl.
Microbiol. Biotechnol., 58, 23–29 (2002).
2) Qin G, Lin JP, Liu XX, and Cen PL, J. Biosci. Bioeng., 102,
316–322 (2006).
3) Lee DH, Jun WJ, Kin KM, and Shin DH, Enzyme Microb.
Technol., 32, 27–34 (2003).
4) Chung SY, Seo KH, and Rhee JI, Process Biochem., 40, 385–
394 (2005).
0
24
30
36
42
48
54
60
66
72
5) Banerjee S, Azad SA, Vikineswary S, Selvaraj OS, and
Mukherjee TK, Asian-Aust. J. Anim. Sci., 13, 991–994 (2000).
6) Kim JK and Lee BK, Aquacult. Eng., 23, 281–293 (2000).
7) Salma U, Miah AG, Tareq KMA, Maki T, and Tsujii H, Poult.
Sci., 86, 714–719 (2007).
Incubation time (h)
Fig. 3. Time Profiles of Residual Acetic Acid (a) and Residual
Butyric Acid (b) of Run A ( ), Run no. 3 ( ), Run no. 14 ( ), and
the Control ( ).
8) Wang LJ, Jiang WB, and Huang BJ, Physiol. Plant., 121, 258–
264 (2004).
9) Sasaki K, Tanaka T, Nishizawa Y, and Hayashi M, J. Ferment.
Bioeng., 71, 403–406 (1991).
It should be noted that while the bacterium grew
under the control conditions, both ALA (at 48 h = 8.06
mM) and biomass formations (at 48 h = 1.7 gDCW/l)
were lowest. The profiles of volatile fatty acid con-
sumption were also low. Apart from the reasons
mentioned above, these phenomena may be explained
by the fact that the pH value of the culture broth was
alkaline.
The results obtained from this design experiment led
us to predict the medium composition to be used for
ALA and biomass productions. The advantage of this
statistical technique, RSM, led us to build an appropriate
model, evaluating the effects of relevant variables and
searching for optimum conditions for the production of
products and reducing the number of experiments,
whereas traditionally the optimization process in which
a single factor is varied while all other factors are kept
fixed at a specific set of conditions is laborious, time
consuming, and incapable of reaching the true optimum
due to ignoring the interaction among variables.
10) Al-Khateeb AA, Al-Khateeb SA, Okawara R, and
Al-Abdoulhady IA, J. Biol. Sci., 6, 1118–1121 (2006).
11) Noparatnaraporn N, Watanabe M, Choorit W, and Sasaki K,
Biotechnol. Lett., 22, 1867–1870 (2000).
12) Shi XY and Yu HQ, Process Biochem., 40, 2475–2481 (2005).
13) Kim JN, Yun JS, and Ryu HW, J. Chromatogr. A, 938, 137–143
(2001).
14) Box GEP and Behnken DW, Technometrics, 2, 455–475 (1960).
15) Mauzerall D and Granick S, J. Biol. Chem., 219, 435–446
(1956).
16) Montgomery DC, ‘‘Design and Analysis of Experiments’’
4th ed., John Wiley & Sons, New York (1997).
17) Box GEP, Hunter WG, and Hunter JS, ‘‘Statistics for Experi-
ments: An Introduction to Design, Data Analysis, and Model
Building,’’ John Wiley & Sons, New York (1978).
18) Fu WQ, Lin JP, and Cen PL, Bioresour. Technol., 99, 4864–
4870 (2008).
19) Nishikawa S, Watanabe K, Tanaka T, Miyachi N, Hotta Y, and
Murooka Y, J. Biosci. Bioeng., 87, 798–804 (1999).
20) Fu WQ, Lin JP, and Cen PL, Appl. Microbiol. Biotechnol., 75,
777–782 (2007).
21) Carlozzi P, Pushparaj B, Degl’Innocenti A, and Capperucci A,
Appl. Microbiol. Biotechnol., 73, 789–795 (2006).
22) Raymond J, Siefert JL, Staples CR, and Blankenship RE, Mol.
Biol. Evol., 21, 541–554 (2004).
Conclusion
This experiment demonstrated the effects of succi-
nate, glycine, and LA on ALA and biomass yields to be
23) Moat AG and Foster JW, ‘‘Microbial Physiology,’’ Oxford
Blackwell Scientific, London (1995).