9
Food Sci. Biotechnol. 23(6): 1911-1919 (2014) DOI 10.1007/s10068-014-0261-7 Optimization of Succinic Acid Production from Cane Molasses by Actinobacillus succinogenes GXAS137 Using Response Surface Methodology (RSM) Naikun Shen, Qingyan Wang, Yan Qin, Jin Zhu, Qixi Zhu, Huizhi Mi, Yutuo Wei, and Ribo Huang Received November 4, 2013; revised May 27, 2014; accepted May 27, 2014; published online December 31, 2014 © KoSFoST and Springer 2014 Abstract A method combining a Plackett-Burman design (PBD), the steepest ascent method (SA), and a Box- Behnken design (BBD) was developed to optimize succinic acid production from cane molasses by Actinobacillus succinogenes GXAS137. The important parameters were (g/L): total sugars of cane molasses (85 g/L), yeast extract (8.84 g/L), and MgCO 3 (63.1 g/L). Verification experiments indicated that the maximal succinic acid production reached 57.43±0.86 g/L, which agreed with the predicted value (57.12 g/L). In addition, batch and fed-batch fermentations were carried out in a 1.3 L stirred bioreactor. Compared with a batch fermentation that produced 57.96 g/L of succinic acid at 60 h, a fed-batch fermentation, performed to minimize the inhibition effect of the substrate, produced 64.34 g/L of succinic acid at 60 h. The combined method is powerful for selection of optimized conditions for succinic acid production from cane molasses. Keywords: succinic acid, cane molasses, Plackett-Burman design, response surface methodology, fermentation Introduction Succinic acid is a member of the 4-carbon dicarboxylic acid family and has a wide range of applications in the fields of surfactants, green solvents, and pharmaceutical intermediates as an important 4-carbon intermediate compound (1). Succinic acid has been conventionally produced from fossil raw materials for more than a century (2). Due to depletion of fossil resources and a strong demand for environmentally friendly energy sources, biological production of succinic acid has attracted great interest (3). Succinic acid can be produced via mircrobial fermentation using Actinobacillus succinogenes (4,5), Mannheimaia succiniciproducens (6), Anaerobiospirillum succiniciproducens (7), and recombinant Escherichia coli (8). A. succinogenes, a member of the Pasteurellaceae family, is one of the most promising succinic acid producers due to an ability to produce high concentrations of succinic acid naturally from a broad range of carbon sources (4,9). The fermentation cost of bio-based succinic acid is, however, a key factor when competing with petroleum-based succinic acid production. To obtain a high production of succinic acid, synthetic fermentation media compositions are often complicated by addition of a high concentration of glucose and 10 kinds of vitamins (10,11), leading to increased expense and a lack of facility for economic bulk succinic acid production. Therefore, use of cheap carbon sources rather than glucose is important for economical production of succinic acid. Use of the natural substrates of starch (12) and cellulose (13) is economically unfavorable because of a requirement for pretreatment in order to release fermentable sugars. Furthermore, A. succinogenes is a fastidious organism (14) and high concentrations of complex growth supplements, mainly yeast extract, are required for use of renewable Naikun Shen, Qingyan Wang, Yan Qin, Yutuo Wei, Ribo Huang () Guangxi Key Laboratory of Subtropical Bio-resource Conservation and Utilization, College of Life Science and Technology, Guangxi University, Nanning, Guangxi 530005, China Tel: +86-0771-2503902; Fax: +86-0771-2503908 E-mail: [email protected] Naikun Shen, Qingyan Wang, Yan Qin, Jin Zhu, Qixi Zhu, Huizhi Mi, Ribo Huang Guangxi Academy of Sciences, State Key Laboratory of Non-food Biomass and Enzyme Technology, National Engineering Research Center for Non-food Biorefinery, Guangxi Biomass Industrialization Engineering Institute, Guangxi Key Laboratory of Biorefinery, Nanning, Guangxi 530007, China RESEARCH ARTICLE

Optimization of succinic acid production from cane molasses by Actinobacillus succinogenes GXAS137 using response surface methodology (RSM)

  • Upload
    ribo

  • View
    218

  • Download
    3

Embed Size (px)

Citation preview

Page 1: Optimization of succinic acid production from cane molasses by Actinobacillus succinogenes GXAS137 using response surface methodology (RSM)

Food Sci. Biotechnol. 23(6): 1911-1919 (2014)

DOI 10.1007/s10068-014-0261-7

Optimization of Succinic Acid Production from Cane Molasses by

Actinobacillus succinogenes GXAS137 Using Response Surface

Methodology (RSM)

Naikun Shen, Qingyan Wang, Yan Qin, Jin Zhu, Qixi Zhu, Huizhi Mi, Yutuo Wei, and Ribo Huang

Received November 4, 2013; revised May 27, 2014; accepted May 27, 2014; published online December 31, 2014

© KoSFoST and Springer 2014

Abstract A method combining a Plackett-Burman design

(PBD), the steepest ascent method (SA), and a Box-

Behnken design (BBD) was developed to optimize succinic

acid production from cane molasses by Actinobacillus

succinogenes GXAS137. The important parameters were

(g/L): total sugars of cane molasses (85 g/L), yeast extract

(8.84 g/L), and MgCO3 (63.1 g/L). Verification experiments

indicated that the maximal succinic acid production

reached 57.43±0.86 g/L, which agreed with the predicted

value (57.12 g/L). In addition, batch and fed-batch

fermentations were carried out in a 1.3 L stirred bioreactor.

Compared with a batch fermentation that produced 57.96

g/L of succinic acid at 60 h, a fed-batch fermentation,

performed to minimize the inhibition effect of the substrate,

produced 64.34 g/L of succinic acid at 60 h. The combined

method is powerful for selection of optimized conditions

for succinic acid production from cane molasses.

Keywords: succinic acid, cane molasses, Plackett-Burman

design, response surface methodology, fermentation

Introduction

Succinic acid is a member of the 4-carbon dicarboxylic

acid family and has a wide range of applications in the

fields of surfactants, green solvents, and pharmaceutical

intermediates as an important 4-carbon intermediate

compound (1). Succinic acid has been conventionally

produced from fossil raw materials for more than a century

(2). Due to depletion of fossil resources and a strong

demand for environmentally friendly energy sources,

biological production of succinic acid has attracted great

interest (3).

Succinic acid can be produced via mircrobial fermentation

using Actinobacillus succinogenes (4,5), Mannheimaia

succiniciproducens (6), Anaerobiospirillum succiniciproducens

(7), and recombinant Escherichia coli (8). A. succinogenes,

a member of the Pasteurellaceae family, is one of the most

promising succinic acid producers due to an ability to

produce high concentrations of succinic acid naturally from

a broad range of carbon sources (4,9). The fermentation

cost of bio-based succinic acid is, however, a key factor

when competing with petroleum-based succinic acid

production. To obtain a high production of succinic acid,

synthetic fermentation media compositions are often

complicated by addition of a high concentration of glucose

and 10 kinds of vitamins (10,11), leading to increased

expense and a lack of facility for economic bulk succinic

acid production. Therefore, use of cheap carbon sources

rather than glucose is important for economical production

of succinic acid. Use of the natural substrates of starch (12)

and cellulose (13) is economically unfavorable because of

a requirement for pretreatment in order to release fermentable

sugars. Furthermore, A. succinogenes is a fastidious organism

(14) and high concentrations of complex growth supplements,

mainly yeast extract, are required for use of renewable

Naikun Shen, Qingyan Wang, Yan Qin, Yutuo Wei, Ribo Huang (�)Guangxi Key Laboratory of Subtropical Bio-resource Conservation andUtilization, College of Life Science and Technology, Guangxi University,Nanning, Guangxi 530005, ChinaTel: +86-0771-2503902; Fax: +86-0771-2503908E-mail: [email protected]

Naikun Shen, Qingyan Wang, Yan Qin, Jin Zhu, Qixi Zhu, Huizhi Mi,Ribo HuangGuangxi Academy of Sciences, State Key Laboratory of Non-foodBiomass and Enzyme Technology, National Engineering Research Centerfor Non-food Biorefinery, Guangxi Biomass Industrialization EngineeringInstitute, Guangxi Key Laboratory of Biorefinery, Nanning, Guangxi530007, China

RESEARCH ARTICLE

Page 2: Optimization of succinic acid production from cane molasses by Actinobacillus succinogenes GXAS137 using response surface methodology (RSM)

1912 Shen et al.

resources.

Molasses is a byproduct of the sugar industry that is

readily available at a relatively low cost. Molasses contains

water, sugars, nitrogen compounds, organic acids, amino

acids, and heavy metals (15). According to previous reports,

the highest concentration of succinic acid was only 46.4 g/

L when cane molasses was used as a carbon source (16).

Cane molasses has been used for production of lactic acid

(17), ethanol (18), and citric acid (19). Therefore, it is

desirable to develop an effective method for succinic acid

production using cane molasses.

There are no reports on optimization of succinic acid

production from cane molasses using Actinobacillus

succinogenes. Response surface methodology (RSM) has

been successfully applied for optimization of bioprocess

(20). Compared with single variable methods, RSM is a

powerful mathematical model based on statistical techniques

in which interactions between multiple process variables

can be identified using fewer experimental trials. In the

present work, an optimization method is proposed including

steps of (a) use of a one-factor-at-a-time approach to

determine the critical components of a medium, (b) use of

a 2-level Plackeet-Burman design (PBD) to identify important

parameters and experimental levels for further optimization,

(c) application of the path of steepest ascent (SA) approach

to the biggest region of succinic acid production, and (d)

use of a Box-Behnken design (BBD) to develop mathematical

models for estimation of relationships between responses

obtained using the optimum values of parameters.

Materials and Methods

Chemicals and materials All chemicals used were

obtained from either Sinochem (Shanghai, China) or OXOID

(Hampshire, UK), except where otherwise specified.

Microorganism and growth conditions The GXAS137

strain of Actinobacillus succinogenes isolated from the

bovine rumen was obtained from the China Center for

Type Culture Collection (No. CCTCC M 2011399) and

was maintained in 20% glycerol at −80oC. Cells were

grown in 100 mL sealed anaerobic bottles containing 50

mL of medium. The medium for inoculum cultures was

composed of glucose (20 g/L), yeast extract (5.0 g/L), corn

steep liquid (dry power) (2.5 g/L), NaH2PO4·H2O (9.6 g/

L), K2HPO4 (15.5 g/L), and NaHCO3 (10 g/L). All liquid

cultures were incubated in a rotary shaker (New Brunswick

Scientific Co., Edison, NJ, USA) at 37oC and 100 rpm.

Treatment of cane molasses The molasses used in this

study was obtained from a local beet sugar refinery in

Guangxi, China. Molasses contained sucrose (370 g/L),

water (250 g/L), converted sugars (glucose and fructose)

(100 g/L), other carbohydrates (26 g/L), crude proteins

(41 g/L), crude fats (0.8 g/L), ash (90 g/L), salt (49 g/L),

and metal ions (83 g/L). Crude molasses was diluted using

distilled water to obtain a total sugar concentration of 250

g/L. For sulfuric acid treatment, the molasses solution was

adjusted to pH 3.5 using 5 M H2SO4, and heated at 60oC

for 2 h. After centrifugation at 8,000×g in an Avanti J-26

XP centrifuge (Beckman Coulter Inc., Fullerton, CA, USA)

for 15 min, the supernatant was adjusted to pH 7.0 using 10

M NaOH (16).

Fermentation in anaerobic bottles For anaerobic bottle

fermentation, exponentially growing cells were inoculated

into 250 mL sealed anaerobic bottles with 100 mL of

medium containing yeast extract (2.5 g/L) and a salt

mixture including K2HPO4 (3.0 g/L), NaH2PO4·H2O

(2.0 g/L), MgCl2·6H2O (0.2 g/L), and NaCl (1.0 g/L). The

pH of the medium was maintained by addition of 40 g of

MgCO3. Separately autoclaved cane molasses was added

aseptically to the medium to make up the desired sugar

concentration (70 g/L). The sterile medium was sparged

with CO2, and Na2S·9H2O (a final concentration of 0.2 g/L)

was added before inoculation to ensure strictly anaerobic

conditions. For fermentation, the medium was inoculated

with a 5% seed culture.

Fermentation in stirred bioreactors Batch fermentation

was carried out in a 1.3 L fermenter (Eppendorf BioFlo/

CelliGen 115, Hamburg, Germany) with an initial broth

volume of 0.8 L. All fermentations were performed at 37oC

at an agitation speed of 200 rpm and a CO2 flow rate of 0.3

L/min. The fermentation medium was the same as the

medium in the anaerobic bottles and the pH was controlled

using MgCO3 during the fermentation process. Samples

were taken every 4 h during the entire fermentation cycle,

which was terminated after 60 h. Fed-batch fermentation

was carried out under the same conditions and the same

total sugar concentrations as for batch fermentations. When

the concentration of total sugars was lower than 15 g/L, a

concentrated molasses solution containing 300 g/L of total

sugar was fed into the stirred bioreactor using a peristaltic

pump to maintain the sugar concentration within 10-15 g/

L during the course of fermentation.

Experimental design

One-factor-at-a-time design: Before RSM was applied,

the approximate medium composition for cultivation of

A.succinogenes was determined by varying one factor at a

time while keeping all other factors constant. The medium

components (carbon, nitrogen, and metal ions) were selected

for maximum succinic acid production while the temperature,

pH, and fermentation time were examined as conditions for

Page 3: Optimization of succinic acid production from cane molasses by Actinobacillus succinogenes GXAS137 using response surface methodology (RSM)

Succinic Acid Production 1913

the fermentation. The nutritional factors were optimized by

changing one factor at a time and keeping other variables

constant.

Plackett-Burman design (PBD): The methodology of

PBD is a powerful and useful tool for rapid searching for

key factors in a multivariable system. This design does not

consider the interaction effects among the variables. PBD

does not determine exact quantities, but can provide

important information about each factor based on relatively

few experiments (21). The design is based on the first order

model:

Y=β0+βixi (1)

where Y is the predicted response, β0 and βi are constant

coefficients, and xi is a coded independent factor. The

purpose of using PBD was to identify the key medium

components for succinic acid production.

Steepest ascent (SA) design: This method was used to

move rapidly towards the neighborhood of the optimum

response (11). Experiments were designed to determine a

suitable direction by increasing or decreasing the

concentrations of variables based on PBD results. The

optimum point in the optimal range was used as the center

point for optimization using a central composite design.

Box-Behnken design (BBD): BBDs are response surface

designs especially constructed to require only 3 levels,

coded as 1, 0, and +1. The levels are formed by combining

2-level factorial designs with incomplete block designs

(22). The response surface analysis is based on multiple

linear regressions that take into account the main,

quadratic, and interaction effects in accordance with the

following equation:

(2)

; i, j =1, 2, 3...,

where Y is the predicted response, β0 is an offset term, βi

is the liner effect, βii is the quadratic effect, βij is the

interaction effect, and ε is experimental error. The variables

xi and xj represent the independent variables (medium

components) in the form of coded values as follows:

i=1, 2, 3, (3)

where xi and Xi are the dimensionless and the actual values

of the independent variable i, X is the actual value of the

independent variable i at the central point, and ∆Xi is the

step change of Xi corresponding to a unit variation of the

dimensionless value.

Methods of analysis Bacterial growth was determined

using optical density (OD) measurements at 660 nm (DU

800 UV/VIS Spectrophotometer; Beckman, CA, USA).

Cultures were diluted using 0.2 M HCl, then centrifuged at

10,000×g for 10 min in order to ensure that none of the

MgCO3 remained undissolved (23). Cell pellets were

washed 3 times using distilled water to remove all medium

components.

Concentrations of organic acids and sugars were

determined using HPLC. The culture broth was centrifuged

at 10,000×g for 10 min, filtered, and 10 µL of each test

sample was subjected to HPLC (Ultimate 3000; Dionex

Co., Sunnyvale, CA USA) equipped with a tunable UV

detector set at 210 nm. An Aminex HPX-87H ion-

exchange column (Rezex ROA-Organic Acid H+, 300×7.8

mm; Phenomene, Torrence, CA, USA) was eluted using

0.005 N H2SO4 as a mobile phase at a flow rate of 0.6 mL/

min. The column temperature was maintained at 45ºC and

a refraction index (RI) detector was used. The temperature

of the RI detector was 55ºC. Amounts of residual total

sugars (glucose, fructose, and sucrose) were determined as

an amount of sucrose based on the DNS method (24) after

acid hydrolysis (adjusted to pH 1.0 using HCl and heated

to 100ºC for 30 min) with sucrose as standard.

Statistical analysis All experiments were repeated 3

times. Data were presented as mean±standard deviation

(SD). All statistical analyses except regression analysis and

surface layers, were performed using one-way analysis of

variance (ANOVA) in IBM-SPSS software (version 19.0;

IBM, Chicago, IL, USA). Treatments were compared

using ANOVA followed by Tukey test. Student’s t-tests

were performed to confirm the comparisons between

groups. All statistical tests were performed at a 5%

signicance level (p<0.05). Design Expert software (version

7.1.3; Stat-Ease Inc., Minneapolis, MN, USA) was used for

regression analysis (R2, standard error, SS, MS, and F

value) and to estimate the coefficients of the regression

equation. Surface layers were also obtained using Design

Expert software (version 7.1.3; Stat-Ease Inc.). The

succinic acid yield was defined as the amount of succinic

acid obtained from 1 g of sugar consumed, expressed as a

percentage.

Results and Discussion

Medium components screened for succinic acid

production with a one-at-a-time method Key parameters

for succinic acid production were identified using a one-at-

a-time method. Key parameters were 40.00 g/L of initial

total sugars for cane molasses, 4.00 g/L of yeast extract,

Y β0

= βixi∑ βiixi2

∑ βijxixji 1=

n

∑ ε+ + + +

I j≠

xiXi Xi–

∆Xi

-------------=

Page 4: Optimization of succinic acid production from cane molasses by Actinobacillus succinogenes GXAS137 using response surface methodology (RSM)

1914 Shen et al.

1.00 g/L of K2HPO4, 1.50 g/L of KH2PO4, 1.00 g/L of NaCl,

0.60 g/L of MnCl2, 0.30 g/L of CaCl2, and 40.00 g/L of

MgCO3 at pH 7.00. The yield of succinic acid reached

30.60 g/L in 60 h at 37oC (Table 1).

Results obtained from a PBD Based on results of the

one-at-a-time analysis, a PBD for 12 trials with 2

concentration levels of was undertaken to evaluate the

significance of 9 medium components (Table 2). Each

variable was represented as 2 levels with a high level

denoted by (+1) and a low level denoted by (−1). All

experiments were performed in triplicate and the average

value of succinic acid concentrations (g/L) after 60 h was

reported as the response (Y). Variables with confidence

levels greater than 95% were considered to significantly

(p<0.05) influence the production of succinic acid.

Based on above results, 9 variables were analyzed using

the PBD, and the maximum and minimum effect of each

variable on the concentration of succinic acid was determined

using Student’s t-test (Table 3). According to the absolute

value of t, the effects of these 9 variables were in the order

of X2>X4>X3>X5>X8>X6>X7>X9>X1. The variables X2, X3,

X4, X5, and X7 had positive effects on succinic acid

production whereas X1, X6, X8, and X9 showed negative

effects on succinic acid production. The variables X2

(initial total sugars of cane molasses), X3 (yeast extract),

and X4 (MgCO3) had significant (p<0.05) effects on

production of succinic acid with confidence levels greater

than 95%. The remaining variables had confidence levels

of <95% and were considered not to be significant

(p>0.05). These insignificant factors were not included in

subsequent optimization experiments, but were used in all

trials at the (−1) level for a negative contribution and a (+1)

level for a positive contribution. A first order model

equation was derived to represent succinic acid production

as a function of the independent variables:

Table 1. Succinic acid fermentation optimized based on a one-at-a-time method1)

Total sugars (g/L) Succinic acid (g/L) Acetic acid (g/L) Formic acid (g/L)Cell concentration

(OD660)Succinic acid yield

(%)

40.00 30.60±0.4 3.41±0.2 2.81±0.5 6.32±0.8 76.50±0.4

1)Each value is a mean of 3 parallel replicates and is reported as mean±SD.

Table 2. The Plackett-Burman experimental design and corresponding results

Run X11)

X2 X3 X4 X5 X6 X7 X8 X9 Y2)

1 1 -1 1 -1 1 -1 1 -1 1 030.84±0.433)

2 -1 1 1 1 -1 -1 1 -1 1 31.69±0.35

3 1 1 -1 1 1 1 -1 -1 1 33.12±0.50

4 1 1 -1 -1 -1 -1 -1 1 1 31.64±0.29

5 -1 -1 -1 -1 1 1 1 1 1 34.01±0.34

6 1 -1 1 1 1 -1 -1 1 -1 32.40±0.42

7 1 -1 -1 1 -1 1 1 -1 -1 33.69±0.28

8 -1 1 1 -1 1 1 -1 -1 -1 32.70±0.41

9 -1 -1 -1 -1 -1 -1 -1 -1 -1 30.70±0.29

10 -1 -1 1 1 -1 1 -1 1 1 30.17±0.38

11 -1 1 -1 1 1 -1 1 1 -1 31.91±0.37

12 1 1 1 -1 -1 1 1 1 -1 30.44±0.39

1)Each variable was tested at a low (−1) and at a high level (+1) level. The 2 levels of each variable were: X1 (NaCl), 1.0 and 1.5 g/L; X2 (totalsugars of cane molasses), 40 and 60 g/L; X3 (yeast extract), 4 and 6 g/L; X4 (MgCO3), 40 and 50 g/L; X5 (MnCl2), 0.6 and 0.8 g/L; X6

(KH2PO4), 1.5 and 3.0 g/L; X7 (K2HPO4), 1.0 and 1.5 g/L; X8 (MgCl2), 1.0 and 1.5 g/L; and X9 (CaCl2), 0.3 and 0.6 g/L2)Y is the succinic acid concentration (g/L).3)Each value is a mean of 3 parallel replicates and is represented as mean±SD.

Table 3. Succinic acid production based on results of thePlackett-Burman design (PBD)

VariablesCoefficient

estimateStandard

errort Prob>| t | Order

Intercept 31.94 0.20 9.67 0.0049 0

X1

1) -0.11 0. 20 -1.16 0.3645 9

X2 0.86 0. 20 18.50 0.0026 1*

X3 0.46 0. 20 5.18 0.0525 3*

X4 0.46 0. 20 5.34 0.0497 2*

X5 0.29 0. 20 1.71 0.1379 4

X6 -0.25 0. 20 -1.66 0.1568 6

X7 0.22 0. 20 1.74 0.1564 7

X8 -0.27 0. 20 -1.33 0.2263 5

X9 -0.19 0. 20 -1.98 0.1419 8

1)The symbols are the same as in Table 1; *statistically significant atp<0.05

Page 5: Optimization of succinic acid production from cane molasses by Actinobacillus succinogenes GXAS137 using response surface methodology (RSM)

Succinic Acid Production 1915

Y=31.94−0.11X1+0.86X2+0.46X3+0.46X4+0.29X5

−0.25X6+0.22X7−0.27X8−0.19X9 (4)

Results obtained from SA analysis The coefficients of

X2, X3, and X4 were positive, indicating that the path of

steepest ascent should increase their concentration in order

to enhance succinic acid production. The center point of

the PBD was considered to be the origin of the path. The

experimental design and responses for the steepest ascent

path analysis are shown in Table 4. The succinic acid yield

showed a maximum at run 4. The highest production

response was 52.33 g/L when the concentrations of initial

total sugars of cane molasses (X2), yeast extract (X3), and

MgCO3 (X4) were 75, 9, and 65 g/L, respectively,

indicating that this point was near the region of maximum

production response.

Optimization of screening culture conditions using

response surface methodology (BBD) Based on the

information above, the point that represented 75.0 g/L of

glucose, 9.0 g/L of yeast extract, and 65.0 g/L of MgCO3

served as the central BBD level for 3 variables. The coded

and real levels are listed in Table 5, and results for an

analysis of variance (ANOVA) for the BBD are shown in

Table 6. The experimental BBD results were fit to a

second-order polynomial equation as follows:

Y=55.10+6.30X1−1.23X2−2.32X3−2.82X12−3.45X2

2

−4.20X32+0.065X1X2−0.43X1X3+0.11X2X3 (5)

The efficiency of fit of the model was checked using the

determination coefficient (R2). The value of the coefficient

of determination (R2=0.9883) indicated that less than 2%

of the total variation was not explained by the model (Table

5). The values of “Probability>F” (0.03) less than 0.05

indicated that model terms were significant (p<0.05).

There was only a 3% chance could occur due to noise.

Changes in the parameter modeled as the 2 factors moved

Table 4. Steepest ascent experimental design and correspondingresults

RunITSCM1)

(g/L)YE2)

(g/L)MgCO3

(g/L)Succinic acid

(g/L)

1 60.00 6.00 50.00 33.61±0.28

2 65.00 7.00 55.00 39.46±0.35

3 70.00 8.00 60.00 46.66±0.40

4 75.00 9.00 65.00 52.33±0.37

5 80.00 10.00 70.00 50.15±0.43

6 85.00 11.00 75.00 48.24±0.37

1)Initial total sugar of cane molasses.2)Yeast extract

Table 5. The Box-Behnken design and experimental responsesof the dependent variable Y (succinic acid concentration, g/L)

Run

Coded values andreal values

Y (succinic acid g/L)

X11)

X22)

X33) Experimental Predicted

1 70 (-1) 8 (-1) 65 (0) 44.14±0.58 43.82

2 70 (-1) 10 (1) 65 (0) 41.98±0.46 41.23

3 80 (1) 8 (-1) 65 (0) 55.54±0.52 56.29

4 80 (1) 10 (1) 65 (0) 53.64±049 53.96

5 75 (0) 8 (-1) 60 (-1) 51.91±0.68 51.10

6 75 (0) 8 (-1) 70 (1) 45.86±0.57 46.25

7 75 (0) 10 (1) 60 (-1) 48.82±0.41 48.43

8 75 (0) 10 (1) 70 (1) 43.20±0.37 44.01

9 70 (-1) 9 (0) 60 (-1) 42.54±0.46 43.67

10 80 (1) 9 (0) 60 (-1) 57.07±0.50 57.13

11 70 (-1) 9 (0) 70 (1) 39.96±0.82 39.90

12 80 (1) 9 (0) 70 (1) 52.76±0.49 51.63

13 75 (0) 9 (0) 65 (0) 55.15±0.71 55.10

14 75 (0) 9 (0) 65 (0) 55.48±0.53 55.10

15 75 (0) 9 (0) 65 (0) 54.67±0.47 55.10

1)X1 is the initial total sugar of cane molasses (g/L).2)X2 is the yeast extract (g/L).3)X3 is MgCO3 (g/L).

Table 6. Analysis of variance (ANOVA) and coefficients estimated for succinic acid production

Term Coefficient SS1) MS F value Pr>F

model 55.1 494.080 54.90 47.13 0.0003*

X1 6.30 317.390 317.390 272.49 0.0001*

X2 -1.23 12.030 12.030 10.33 0.0236*

X3 -2.32 43.060 43.060 36.97 0.0017*

X1*X1 -2.82 29.360 29.360 25.21 0.0040*

X1*X2 0.065 0.017 0.017 0.015 0.9088

X1*X3 -0.43 0.750 0.750 0.640 0.4592

X2*X2 -3.45 44.080 44.080 37.84 0.0017*

X2*X3 0.11 0.046 0.046 0.040 0.8499

X3*X3 -4.20 65.050 65.050 55.85 0.0007*

Lack of fit 5.490 1.830 11.04 0.0842*

1)SS, sum of squares; MS, mean square; *significant at p<0.05, R2=0.9883, adjusted R2= 0.9674

Page 6: Optimization of succinic acid production from cane molasses by Actinobacillus succinogenes GXAS137 using response surface methodology (RSM)

1916 Shen et al.

along those levels, while the other factor were held

constant at the central point, are shown in Fig. 1.

The independent variable (X1) had a significant (p<0.05)

effect with a positive coefficient (Table 6), according to

which an increase in the concentration of X1 led to an

increase in the succinic acid yield. The negative signs of

the independent variables X2 and X3, and the squared

variables X12, X2

2 , and X32 revealed a reduction in succinic

acid production when the concentrations were increased in

the system. The same phenomenon was observed with the

interaction term of X1X3.

A second-order polynomial model was used to calculate

the values of the initial total sugar content of cane

molasses, yeast extract, and MgCO3, in order to determine

the maximum succinic acid concentration that corresponded

to the optimum levels of these variables. The experimental

data were fitted using Equation (5), which indicated that

the concentrations of initial total sugar of cane molasses

(X1=85 g/L), yeast extract (X2=8.84 g/L), and MgCO3

(X3=63.1 g/L) resulted in a maximum succinic acid

concentration (57.12 g/L). Therefore, optimization of succinic

acid production from cane molasses using A. succinogenes

GXAS137 was achieved.

Succinic acid is an end-product of the sugar substrate, so

the initial sugar concentration influenced cell growth and

metabolites production. Urbance et al. (25) reported that

A. succinogenes could tolerate up to 160 g/L of an initial

glucose concentration in batch fermentation. However,

significant decreases in biomass, succinic acid production,

and sugar use were observed when the initial sugar

concentration was over 65 g/L. In this study, a maximum

value of succinic acid was obtained in a culture with an

initial total sugar concentration of 85 g/L, a higher value

than reported (65 g/L) by both Liu et al. (16) and Lin et al.

(26). This might be due to differences in strains used and

experimental conditions.

It has been reported that A. succinogenes is a fastidious

organism that requires complex nutrients, such as amino

acids and vitamins, for cell growth. Therefore, the yeast

extract was determined to be a key factor for production of

succinic acid because it affected cell growth directly as a

nutrient. The yeast extract contained many trace substances,

such as folic acid, pantothenic acid, biotin, and vitamins

B1, B2, B6, and B12. This may be an important reason why

many vitamins could be omitted while maintaining efficient

succinic acid accumulation in this study. The maximum

value of succinic acid was obtained in a culture with a

yeast extract concentration 8.84 g/L, which was lower than

a previously reported yeast extract concentration of glucose

as carbon source (11). Thus, molasses contain some nitrogen

source that meets the demands of cell growth. Therefore,

the cost of providing a nitrogen source for cane molasses

fermentation should be reduced. Replacement of yeast

extract using other protein sources was unsuccessful,

causing low product concentrations (data not shown). An

in-depth study to identify an inexpensive nitrogen source

for succinic acid production from cane molasses is in

progress.

MgCO3 was also identified as a key factor because it

was used as a neutralizing agent and as a CO2 donor. In

bio-based succinic acid production, the culture pH is

known to be a key factor for both cell growth and succinic

Fig. 1. Response surface curve for succinic acid production byActinobacillus succinogenes GXAS-137 showing the combinationeffects of (A) total sugars of cane molasses and yeast extract,(B), and (C) yeast extract and MgCO3. Other factors wereconstant at 0 levels. YE, yeast extract; TSCM, total sugars of canemolasses

Page 7: Optimization of succinic acid production from cane molasses by Actinobacillus succinogenes GXAS137 using response surface methodology (RSM)

Succinic Acid Production 1917

acid production. The culture pH can influence cellular

metabolism by changing the chemical environment and by

affecting enzyme activity (27). A near-neutral pH culture

environment is suitable for A. succinogenes. However, the

culture pH quickly decreased during fermentation due to

rapid accumulation of succinic acid and acetic acid. MgCO3

was used to maintain a constant pH value in the fermentation

broth in this study. The final pH of a medium with an

initial pH of 7.5 was close to neutral (approximately 6.5)

during fermentation, and a neutral environment was helpful

for accumulation of succinic acid. Therefore, the effect of

medium optimization using MgCO3 may be partly related

to the influence of the medium pH.

The level of dissolved CO2 and the ionization equilibrium

between HCO3− and CO3

2− are direct factors that influence

production of succinic acid (28). As an important CO2 donor

in A. succinogenes fermentation, MgCO3 reacted with

organic acids in the fermentation broth and caused an

increase in the dissolved concentrations of HCO3, CO32,

and CO2. Recent studies have also reported that magnesium

ions play an important role in maintaining cellular metabolism

because these ions are a cofactor for phosphoenolpyruvate

(PEP) carboxykinase, which is a key enzyme for succinic

acid production (29). These properties make MgCO3 a key

factor for improved succinic acid production.

Validation of the optimization fermentation medium

The combination levels of the 3 key factors (initial total

sugars of cane molasses, yeast extract, and MgCO3) were

predicted based on the BBD polynomial model. The

applicability of the model and the accuracy of the

prediction were checked based on verification experiments

performed in triplicate using the optimized conditions

representing the maximum point of the concentration of

succinic acid to verify the modelling results. The predicted

average concentration of succinic acid was 57.12 g/L, and

the average concentration determined by experiment was

57.43±0.86 g/L. This is an improvement in succinic acid

Fig. 2. Time course of cell growth and production of organic acids in batch (a, b) and fed-batch (c, d) fermentations withpretreated cane molasses. Cells were grown in a 1.3 L stirred bioreactor with a cane molasses total sugar content of 85 g/L, yeast extractof 8.84 g/L, and MgCO3 of 63.1 g/L. In fed-batch fermentation, the final total sugar concentration was the same as the batch fermentation

and the initial total sugar concentration of 35 g/L was maintained at 10-15 g/L during the fermentation process. Symbols are sucrose (■),glucose (▲), fructose (▼), total sugar (△), succinic acid (□), acetic acid (▽), and OD660 (◇).

Page 8: Optimization of succinic acid production from cane molasses by Actinobacillus succinogenes GXAS137 using response surface methodology (RSM)

1918 Shen et al.

concentration by approximately 87% relative to the amount

obtained by optimizing a single variable of the culture

medium, where the concentration of succinic acid was only

30.60 g/L. The good correlation between the predicted and

experimental values after optimization justified the validity

of the response model and the existence of an optimum

point.

Fermentation in stirred bioreactors The fermentation

performance of the fed-batch process was much better than

the batch process (Fig. 2). Substrate inhibition was a clear

limitation to succinic acid production by A. succinogenes

in batch fermentation. When the sugar concentration was

maintained at a low level during the fed-batch fermentation

process, cells grew faster and the steady state phase was

prolonged due to elimination of substrate inhibition. In fed-

batch fermentation, the succinic acid concentration using

cane molasses reached 64.34 g/L, higher than previously

reported values (16,27) and succinic acid productivity and

yield were 1.07 g/L/h and 75.69%, which were higher than

corresponding batch fermentation values of 57.96 g/L, 0.97

g/L/h and 67.87%, respectively. Thus, the fed-batch process

was more efficient for cultivation than the batch process for

succinic acid production by A. succinogenes.

Acknowledgments This work was supported by the

National Natural Science Foundation of China (No.

31160023), the Natural Science Foundation of Guangxi

Province (No. 2013GXNSFBA019102), the Science

foundation of Guangxi Academy of Science (No. 13YJ22SW),

and the BaGui Scholars Program Foundation of Guangxi

Province, China.

Disclosure The authors declare no conflict of interest.

References

1. Song H, Lee SY. Production of succinic acid by bacterialfermentation. Enzyme Microb. Technol. 39: 352-361 (2006)

2. Wilke D. Chemicals from biotechnology: molecular plant geneticswill challenge the chemical and the fermentation industry. Appl.Microbiol. Biotechnol. 52: 135-145 (1999)

3. Xi YL, Chang KQ, Zhang JH, Bai XF, Jiang M, Wei P. Effect ofbiotin and a similar compound on succinic acid fermentation byActinobacillus succinogenes in a chemically defined medium.Biochem. Eng. J. 69: 87-92 (2012)

4. Guettler MV, Rumler D, Jain MK. Actinobacillus succinogenes sp.nov., a novel succinic-acid-producing strain from the bovine rumen.Int. J. Syst. Bacteriol. 49: 207-216 (1999)

5. Wang CX, Lee YB, Shahbazi A, Xiu SN. Succinic acid productionfrom cheese whey using Actinobacillus succinogenes 130 Z. Appl.Biochem. Biotechnol. 145: 111-119 (2008)

6. Oh IJ, Lee HW, Park CH, Lee SY, Lee J. Succinic acid productionby continuous fermentation process using Mannheimaia succinici-producens LPK7. J. Microbiol. Biotechnol. 18: 908-912 (2008)

7. Lee PC, Lee SY, Hong SH, Chang HN, Park SC. Biological

conversion of wood hydrolysate to succinic acid by Anaero-biospirillum succiniciproducens. Biotechnol. Lett. 25: 111-114(2003)

8. Jiang M, Liu SW, Chen JF, Yu KQ, Li Y, Yue FF, Xu B, Wei P.Effect of growth phase feeding strategies on succinate production bymetabolically engineered Escherichia coli. Appl. Environ. Microb.76: 1298-1300 (2010)

9. Mckinlay J, Vieille C, Zeikus JG. Prospects for a bio-basedsuccinate industry. Appl. Microbiol. Biotechnol. 76: 727-740 (2007)

10. Mckinlay J, Zeikus JG, Vieille C. Insights into Actinobacillussuccinogenes fermentative metabolism in a chemically definedgrowth medium. Appl. Environ. Microbiol. 71: 6651-6656 (2005)

11. Zhu LW, Wang CC, Liu RS, Li HM, Wang DJ, Tang YJ.Actinobacillus succinogenes ATCC 55618 fermentation mediumoptimization for the production of succinic acid by response surfacemethodology. J. Biomed. Biotechnol. 2012: 626137 (2012)

12. Dorado MP, Lin SKC, Koutinas A, Du CY, Wang RH, Webb C.Cereal-based biorefinery development: Utilisation of wheat millingby-products for the production of succinic acid. J. Biotechnol. 143:

51-59 (2009)13. Xi YL, Dai WY, Xu R, Zhang JH, Chen KQ, Jiang M, Wei P,

Ouyang PK. Ultrasonic pretreatment and acid hydrolysis ofsugarcane bagasse for succinic acid production using Actinobacillussuccinogenes. Bioproc. Biosyst. Eng. 36: 1779-1785 (2013)

14. Mckinlay J, Zeikus JG, Vieille C. Insights into Actinobacillussuccinogenes fermentative metabolism in a chemically definedgrowth medium. Appl. Environ. Microbiol. 71: 6651-6656 (2005)

15. Kotzamanidis C, Roukas T, Skaracis G. Optimization of lactic acidproduction from beet molasses by Lactobacillus delbrueckiiNCIMB 8130. World J. Microbiol. Biotechnol. 18: 441-448 (2002)

16. Liu YP, Zhang P, Sun ZH, Ni Y, Dong JJ, Zhu LL. Economicalsuccinic acid production from cane molasses by Actinobacillussuccinogene. Bioresource Technol. 99: 1736-1742 (2008)

17. Dumbrepatil A, Adsul M, Chaudhari S, Khire J, Gokhale D.Utilization of molasses sugar for lactic acid production byLactobacillus delbrueckii subsp. delbrueckii Mutant Uc-3 in BatchFermentation. Appl. Environ. Microb. 74: 333-335 (2008)

18. Ghorbani F, Younesi H, Esmaeili SB, Najafpour G. Cane molassesfermentation for continuous ethanol production in an immobilizedcells reactor by Saccharomyces cerevisiae. Renew. Energ. 36: 503-509 (2011)

19. Ikamul H, ALI S, Qadeer M, Iqbal J. Citric acid production byselected mutants of Aspergillus niger from cane molasses.Bioresource Technol. 93: 125-130 (2004)

20. Azma M, Mohamed MS, Mohamed R, Rahim RA, Ariff, AB.Improvement of medium composition for heterotrophic cultivationof green microalgae, Tetraselmis suecica, using response surfacemethodology. Biochem. Eng. J. 53: 187-195 (2011)

21. Roukas T, Niavi P, Kotzekidou P. A new medium for spore productionof Blakeslea trispora using response surface methodology. World J.Microbiol. Biotechnol. 27: 307-317 (2011)

22. Febe F, Abdulhameed S, Madhavan NK, Sumitra R, Sanjoy G,George S, Ashok P. Use of response surface methodology foroptimizing process parameters for the production of α-amylaseby Aspergillus oryzae. Biochem. Eng. J. 15: 107-115 (2003)

23. Song H, Huh YS, Lee SY, Hong WH, Hong YK. Recovery ofsuccinic acid produced by fermentation of a metabolically engineeredMannheimia succiniciproducens strain. J. Biotechnol. 132: 445-452(2007)

24. Miller GL. Use of dinitrosalicylic acid reagent for determination ofreducing sugar. Anal. Chem. 31: 426-428 (1959)

25. Urbance SE, Pometto AL 3rd, Dispirito AA, Denli Y. Evaluation ofsuccinic acid continuous and repeat-batch biofilm fermentation byActinobacillus succinogenes using plastic composite supportbioreactors. Appl. Microbiol. Biotechnol. 65: 664-70 (2004)

26. Lin SKC, Du C, Koutinas A, Wang R, Webb C. Substrate andproduct inhibition kinetics in succinic acid production byActinobacillus succinogenes. Biochem. Eng. J. 41: 128-135 (2008)

27. Podkovyrov SM, Zeikus JG. Purification and characterization ofphosphoenolpyruvate carboxykinase, a catabolic CO2-fixing enzyme,

Page 9: Optimization of succinic acid production from cane molasses by Actinobacillus succinogenes GXAS137 using response surface methodology (RSM)

Succinic Acid Production 1919

from Anaerobiospirillum succiniciproducens. J. Gen. Microbiol.139: 223-228 (1993)

28. McKinlay JB, Vieille C. 13C-metabolic flux analysis of Actinobacillussuccinogenes fermentative metabolism at different NaHCO3 and H2

concentrations. Metab. Eng. 10: 55-68 (2008)

29. Bazaes S, Toncio M, Laivenieks M, Zeikus JG, Cardemil E.Comparative kinetic effects of Mn(II), Mg(II), and the ATP/ADPratio on phosphoenolpyruvate carboxykinases from Anaerobiospirillumsucciniciproducens and Saccharomyces cerevisiae. Protein J. 26:265-269 (2007)