Succinic acid production from sugarcane bagasse hemicellulose hydrolysate by Actinobacillus succinogenes Free

Abstract

This article is based on a presentation at the 32nd Symposium on Biotechnology for Fuels and Chemicals.

Electronic supplementary material

The online version of this article (doi:10.1007/s10295-010-0874-7) contains supplementary material, which is available to authorized users.

Introduction

Succinic acid is currently produced from crude oil by catalytic hydrogenation of maleic anhydride to succinic anhydride and subsequent hydration or by the direct catalytic hydrogenation of maleic acid [29]. Population growth and the associated demand for energy and goods coupled with more restrictive environmental regulations and growing concern about carbon emissions from the burning of fossil fuels have intensified the search for renewable energy feedstocks to substitute or complement fossil fuel sources [21, 23]. Sugarcane bagasse, a lignocellulosic material composed of 50% cellulose, 25% hemicellulose and 25% lignin, is one of the most important agricultural residues produced in Brazil, amounting to approximately 250 kg per ton sugarcane [28]. A diluted acid pretreatment of sugarcane bagasse generates a liquid phase (hemicellulose hydrolysate) composed mainly of xylose, which can, in turn, be used as a substrate for biotechnological and chemical processes [7, 11]. The use of a renewable agricultural feedstock for the production of succinic acid reduces the need to draw upon limited oil reserves, thus diversifying a biorefinery’s potential product portfolio [17]. The utilization of such feedstocks within the concept of biorefinery has been successfully investigated for the fermentative production of a range of products, including lactic acid, ethanol, polyhydroxybutyrate (PHB) and succinic acid [6].

Succinic acid, also known as amber acid or butanedioic acid, is a dicarboxylic acid with the molecular formula C₄H₆O₄. It can be used as a precursor for the production of many chemicals for use in the agricultural, food processing and pharmaceutical industries [for example, as surfactants, detergents, adipic acid, 1,4 butanediol, tetrahydrofuran, N-methyl pyrrolidinone, 2-pyrrolidinone, succinate salts, gamma-butyrolactone, various green solvents, biodegradable polymers, such as polybutyrate succinate (PBS) and ingredients to stimulate animal and plant growth] [22, 29]. Several research teams are working on the development of industrial-level fermentation processes for succinic acid production using strains of Anaerobiospirillum succiniciproducens, Actinobacillus succinogenes, Mannheimia succiniciproducens [14, 24, 30] and Corynebacterium glutamicum and recombinant Escherichia coli strains [3, 19]. A. succinogenes is able to produce a relatively large amount of succinic acid from a broad range of carbon sources, such as arabinose, cellobiose, fructose, galactose, glucose, lactose, maltose, mannitol, mannose, sorbitol, sucrose, xylose or salicin under anaerobic conditions [8]. This microbial species is a facultatively anaerobic, Gram-negative bacterium that naturally produces high concentrations of succinate as a fermentation end product in addition to formate, acetate and ethanol [8, 26]. CO₂ functions as an electron acceptor and alters the flux of phosphoenolpyruvate (PEP), which induces the metabolism of the substrate to pyruvate and lactate/ethanol at low CO₂ levels and succinate at high CO₂ levels [26].

The composition of culture medium significantly affects the yield of any fermentation product. Physiological and nutritional factors, such as the initial sugar concentration, complex nitrogen sources, inoculum size and cell acclimatization, carbonate ion concentrations, magnesium, pH and the temperature of the growth medium, are reported to be the most critical factors affecting both cell growth and succinic acid production [1, 14].

Objective

The ultimate aim of the study reported here was to develop a cheap and suitable medium for succinic acid production from the hemicellulose hydrolysate of sugarcane bagasse in a batchwise culture of A. succinogenes. To this end, we studied the influence of the medium components (NaHCO₃, Yeast Extract, K₂HPO₄, MgSO₄) and their interaction on succinic acid production using an experimental design strategy.

Materials and methods

Microorganism and inoculum preparation

The strain of A. succinogenes CIP 106512 was obtained from the Pasteur Institute. The culture stock was maintained on Trypticase Soy Agar (TSA) slants at 4°C. The inoculum was grown in a liquid medium composed of xylose, 20 g l⁻¹; yeast extract, 2.0 g l⁻¹; MgSO₄, 1.0 g l⁻¹; NaHCO₃, 12.0 g l⁻¹; K₂HPO₄, 1.5 g l⁻¹. The cultures were shaken at 150 rpm and 37°C in an orbital shaking incubator for 24 h, which was the time required for the microorganisms to enter the exponential growth phase. After growth, the cells needed for each fermentation assay were separated from the media by centrifugation at 8,000 rpm and 10°C for 20 min.

Fermentation assays in flasks

The fermentation was initially carried out in 500-ml flasks containing 200 ml of a synthetic medium containing the same nutrients as those in the medium for the inoculum preparation. The medium was inoculated with 10% (v/v) of inoculum and the fermentation under anaerobic conditions was carried out as described elsewhere [12] at 150 rpm and 37°C in a shaker for 24 h. The system was set to create anaerobic conditions by introducing the desired CO₂ [12, 24]. The injection of CO₂ into the flasks was achieved using a gas cylinder, with the flow rate monitored with a gas manometer. Tygon tubes were used to introduce the gas into the interior of the flasks via connections in which membranes were present to purify the gas as it entered. Disposable syringes attached to the end of the tubes penetrated the cork on the surface of the flasks, thus promoting the flow of gas into the region between the fermentation environment and the Erlenmeyer surface (headspace). The pH was adjusted to 7.0 through the addition of sodium hydroxide (NaOH). A separately autoclaved solution of carbohydrate was added aseptically to the medium after autoclaving (Fig. 1).

Fig. 1

Schematic representation of CO₂ distribution in shake flasks for medium optimization in succinic acid fermentation

Open in new tabDownload slide

The results were analyzed by experimental design to define the optimal conditions for succinic acid production, and the predicted values were validated experimentally in batch fermentation.

Fermentation assays in bioreactor using synthetic medium

All batch fermentations were carried out at 37°C in a 2-l bioreactor (BIOFLO III; New Brunswick Scientific, Edison, NJ) with a working volume of 1.5 l. The initial xylose concentration in the medium was 20 g l⁻¹, and the temperature and agitation speed were controlled at 37°C and 150 rpm, respectively. The bioreactor was continuously supplied with CO₂ from the gas cylinder at a flow rate of 0.05 vvm during the whole fermentation period. A second bioreactor was operated without a supply of external CO₂ to compare how essential the feed of this gas was in the succinic acid production. The pH was monitored using a sterilizable pH electrode and maintained at 7.0 through the addition of 1 M NaOH. The kinetics of the fermentation process were evaluated under the optimal conditions established in the shake flask experiments. In the second part of this study, the fermentation was performed using the hydrolysate obtained from sugarcane bagasse with a high xylose concentration (52 g l⁻¹).

Sugarcane diluted acid pretreatment/hemicellulose hydrolysis

The sugarcane bagasse (Saccharum spp.) was provided by Costa Pinto Distillery (SP, Brazil). An acid pretreatment was developed to disrupt the lignocellulosic matrix of the sugarcane bagasse and hydrolyze the hemicellulosic component [H₂SO₄, 1% (v/v); solid:liquid ratio, 1:2; 121°C; 40-min pretreatment] [4]. The hydrolysate obtained was separated from the solid phase by pressure filtration and neutralized until pH 6.0 with Ca(OH)₂; it was then filtered again to remove the generated precipitate.

Fermentation assays in bioreactor using sugarcane hemicellulose hydrolysate

The cells were initially activated in the synthetic medium described above and then cultured in two acclimatization stages on a hydrolysate-containing medium; first, with 25% (v/v) medium and second, with 50% (v/v) medium. The cultures were grown in a rotatory shaker at 37°C and 150 rpm. After acclimatization, the cells required for each fermentation assay were separated from the media by centrifugation at 8000 rpm and 10°C for 20 min and inoculated into the bioreactor with all of the same nutrients used in the inoculum preparation medium, with the exception of xylose, which was substituted by the hemicellulose hydrolysate (xylose concentration of 52 g l⁻¹) derived from sugarcane bagasse pretreatment. The kinetics of the fermentation process operating in the bioreactor were evaluated under the optimal conditions established in the shake flask experiments using the statistical software package Design-Expert ver. 6.0 (Stat-Easem Minneapolis, MN).

Optimization of conditions for succinic acid production

The data collected on succinic acid production were subjected to analysis of variance (ANOVA) in order to evaluate statistical significance. The mathematical relationship between the independent and response variables was calculated. The strategy of sequential experimental design was adopted for optimization of succinic acid production. The first experimental 2⁴⁻¹ fractional factorial design with four variables (NaHCO₃, yeast extract, K₂HPO₄, MgSO₄) was carried out on two levels, namely, minimum and maximum, coded as “−1” e “+1”, respectively, with two replicates of the center point (Table 1). A second experimental central composite rotational design (CCRD) (Table 2) was conducted to study the influence of the medium components on succinic acid production.

The response surface methodology (RSM) and the Pareto chart were also used to investigate the aggregate effect of multiple variables and to seek the optimal conditions for this multivariable system [5].

Analytical methods

Cell concentration was determined by measuring the optical density of a diluted sample at 600 nm (Spectrum Lab 22 PC; S) using a standard curve of absorbance against dry cell mass [16, 30]. Xylose concentration in the samples was determined by high-performance liquid chromatography (HPLC) using the Waters chromatographic system (Waters, Milford, MA) consisting of a HPX-87p column (Bio-Rad, Hercules, CA), a Waters 510 pump, a refractive index detector (Waters model 410) and HP 3390A integrator. The succinic acid concentration produced also was estimated on high performance liquid chromatography (HPLC). Samples (20 μl) were injected into a column (C18, 250 length × 4.6 mm internal diameter, 9 μm; StrodsII Peek) at a flow rate of 0.9 ml min⁻¹. Degassed and filtered sulfuric acid (0.01 N) was used as the mobile phase [8]. Succinic acid yield was defined as the amount of succinic acid produced from 1 g of xylose (expressed as a percentage).

Results and discussion

Fractional factorial design and analysis

The results of the experimental 2⁴⁻¹ fractional factorial design with four variables are presented in Table 3. A set of ten trials was generated (8 independent and 3 repetitions of the central point). Statistical significance of the respective model equations was checked using ANOVA. The final succinic acid concentration ranged from 4.74 to 9.02 g l⁻¹. The highest succinic acid concentrations (experiments 3, 5, 8, 9 and 10) were coincident with the presence of NaHCO₃ and MgSO₄, and the maximum succinic acid concentration of 9.02 g l⁻¹ was achieved in 24 h of incubation at 37°C and an agitation speed of 150 rpm, with the following medium composition: xylose, 20 g l⁻¹; K₂HPO₄, 10.0 g l⁻¹; MgSO₄, 2.0 g l⁻¹; yeast extract, 5.0 g l⁻¹; NaHCO₃, 10.0 g l⁻¹.

According to the variance analysis, the best model for succinic acid concentration was the Reduced Quadratic Model, and the values of R² (0.952) and pure error (0.064) show that the model is in agreement with the experimental data. The model remained significant, with a Fisher’s value of 29.85 and a statistically nonsignificant lack of fit (0.120) (P > 0.1), as shown in Table 4.

These results enabled plotting the histogram of the influence of the process variable (Fig. 2a), which provided important data on the statistical relevance of the factors and their interactions. Pareto diagram analysis (Fig. 2b) revealed that the variables D (NaHCO₃) and B (MgSO₄) had significant positive linear effects on succinic acid production. NaHCO₃ (parameter D) had the most significant factorial design effect, followed in decreasing order of effect by MgSO₄ (parameter B), yeast extract (parameter C) and K₂HPO₄ (parameter C); as independent variables, these had a lesser influence on succinic acid concentration.

Fig. 2

a Histogram of the influence of variables and their interactions, b Pareto chart for minimal confidence range of 90%, according to the sum of squares in 2⁴⁻¹ fractional factorial design. A K₂HPO₄, B MgSO₄, C Yeast extract, D NaHCO₃

Open in new tabDownload slide

Figure 2 reveals that NaHCO₃ had a significant effect on achieving a high succinic acid concentration, indicating that in the next trial the concentration of this parameter should be increased. In contrast, K₂HPO₄ showed no significant effect (0.1159), suggesting that K₂HPO₄ could be eliminated from the medium. However, the interaction between K₂HPO₄ and yeast extract (AC) had a significant influence in this analysis, with values of P > F of less than 0.1 (0.0952). MgSO₄ and K₂HPO₄, with their components magnesium and phosphorus, play an important role in cell physiology and morphology [8].

According to the results of the first experimental design, the greatest efficiency was obtained in experiment no. 8. ANOVA analysis showed that the maximum succinic acid concentration occurred when the highest levels of parameters (B) and ©) were combined with a lower concentration of K₂HPO₄, indicating that there is a maximum. However, the values of the central points are close to the maximum obtained in the investigated range, indicating the need to add axial points to a subsequent trial in order to optimize succinic acid production (in factorial design only linear models can be evaluated). Thus, the next step was to evaluate succinic acid production using the CCRD technique, since this approach enables the response surface to be predicted with curvature plots and the maximum values of the response variables to be visualized. The maximum succinic acid concentration (9.02 g l⁻¹) can be enhanced, as the center points are still lower than the factorial points, showing an increasing tendency in the direction of the maximum point. All four parameters evaluated were significant (high F value) and kept in the model. The analysis of effects revealed that the levels of NaHCO₃ and MgSO₄ could be increased. However, any increase in K₂HPO₄ concentration may reduce succinic acid production since impurities, such as other organic acids (acetic and formic acids) and salts (HPO₄²⁻), can cause a decrease in the extraction yield of succinic acid from an aqueous phase due to their ionic strength. Thus, the concentration of K₂HPO₄ should be minimized in the fermentation process to optimize the downstream processing of succinic acid separation [9].

Central composite rotational design and analysis

The results of the CCRD are presented in Table 5. A set of 30 trials was generated (27 independent and 3 repetitions of the central point). According to the results of the second experimental design, an increase in succinic acid concentration was achieved, increasing from a maximum of 9.05 g l⁻¹ (first experimental design) to 12.8 g l⁻¹ (second experimental design—CCRD). The final succinic acid concentration varied between 6.04 and 12.8 g l⁻¹. A maximum succinic acid concentration was obtained in experiment no. 2, with the following conditions: xylose concentration, 20 g l⁻¹; K₂HPO₄, 5.0 g l⁻¹; MgSO₄, 2.0 g l⁻¹; yeast extract, 3.0 g l⁻¹; NaHCO₃; 10 g l⁻¹.

The data presented in Table 6 corroborate the results of the experimental 2⁴⁻¹ fractional factorial design in showing that the best model for succinic acid concentration was the Reduced Quadratic Model, based on the highest value for R² (0.995), indicating 99.5% of the response variability. This model remained significant, with a Fisher value of 34.81, thus denoting a curvilinear plane, as can be observed in the ANOVA (Table 4). Moreover, the residue was low and statistically nonsignificant, which did not invalidate the model for predictive purposes because the equation had a high R² value. The lack of fit was not significant (P > 0.05) which, combined with the F values of the parameters and the determination coefficient (R²), showed that the model adequately adjusts to the experimental points, representing the confidence of the results [2].

The histogram of variable influence and interactions in the CCRD experimental design are presented in Fig. 3. Parameter D (NaHCO₃) was the most influential, followed by parameter B (MgSO₄) and parameter C (Yeast extract) and their interactions. The interaction between AB, followed by BC, AD and CD presented more significant effects than those of parameter A (K₂HPO₄).

Fig. 3

Histogram of the influence of the process variables and their interactions, according to the sum of squares in the experimental central composite rotational design (CCRD). A K₂HPO₄, B MgSO₄, C yeast extract, D NaHCO₃

Open in new tabDownload slide

Figure 4 shows the three-dimensional (3-D) response contour and surface plots, which represent the regression equations. Analysis of Fig. 4a and b shows that increasing concentrations of MgSO₄ and yeast extract can result in increases in the final succinic acid concentration. Figure 4c shows that the increase in K₂HPO₄ concentration contributes to a reduced succinic acid production. In contrast, Fig. 4b shows that the interaction between K₂HPO₄ and yeast extract (AC) has a significant effect. The interaction between these four variables is also shown with high Fisher values (Table 6), indicating the large degree of synergism between the nutrients being fermented by A. succinogenes, producing larger amounts of succinic acid. Figure 4d shows that the highest interaction was between MgSO₄ and NaHCO₃, which confirms the statistical analysis. In all experiments, the highest levels of these nutrients promote an increase in succinic acid concentration.

Fig. 4

Response surface plots showing the effect of parameters and their combined effects on succinic acid concentration. Interaction between the parameters AB (a), AC (b), AD (c), BD (d). A K₂HPO₄, B MgSO₄, C Yeast extract, D NaHCO₃

Open in new tabDownload slide

Succinic acid production was increased by A. succinogenes when an optimal combination of nutrients was used. Analysis of the response measured (concentration of succinic acid) emphasizes experimental condition number 2 since it is the most efficient. The effect of NaHCO₃, keeping MgSO₄ to a minimum level of (2.0 g l⁻¹) and yeast extract at the center point (3.0 g l⁻¹), indicates that the optimal point for succinic acid production is around the maximum level of NaHCO₃ (10.0 g l⁻¹). Therefore, when the fermentation process occurs in the presence of high levels of K₂HPO₄, succinic acid concentration is affected. All of the nutrients were significant and essential to the metabolism of the bacteria. The concentrations of amino nitrogen (yeast extract) in the fermentation broth usually reflect the concentrations of proteins and amino acids [21].

Liu et al. [16] demonstrated that there was no significant inhibition of cell growth or succinic acid production by magnesium ions using A. succinogenes strain CGMCC1593. Nitrogen sources generally play a significant role because these nutrients are directly linked with cell proliferation and metabolite biosynthesis. It has been shown that yeast extract, which contains protein, lipid, vitamins and others, supplies important growth factors for succinic acid production, but its cost may be a limitation for its application in industrial processes [25]. Recent studies on similar types of fermentations have demonstrated that the nitrogen deficiency stimulated the biosynthesis of other metabolites, such as acetic and formic acids [20]. Variance analysis showed the adequacy of the model, which was validated experimentally in the bioreactor.

Thus, the medium obtained after optimization by the strategy of sequential experimental design had the following composition: xylose, 20 g l⁻¹; yeast extract, 5.0 g l⁻¹, MgSO₄, 3.0 g l⁻¹; NaHCO₃, 10 g l⁻¹. The validation of the experimental design was carried out in the bioreactor (Figs. 5, 6).

Fig. 5

Concentration of cells (filled diamond), xylose (filled square) and succinic acid (filled triangle) during the fermentation process carried out under controlled conditions (37°C, pH 7.0, 150 rpm) using synthetic medium with 22.0 g/l of xylose. Medium with low (a) and high (b) CO₂ supply

Open in new tabDownload slide

 

Fig. 6

Concentration of cells (filled diamond), xylose (filled square) and succinic acid (filled triangle) during the fermentation of sugarcane bagasse hemicellulose hydrolysate containing 52.0 g  l⁻¹ xylose, carried out under controlled conditions (37°C, pH 7.0, 150 rpm). Medium with low (a) and high (b) CO₂ supply

Open in new tabDownload slide

Succinic acid fermentation

The best conditions for the fermentation process indicated by CCRD (Table 5) were used to prepare the fermentation medium for succinic acid production in the bioreactor experiment. Under response surface methodology (RSM)-optimized conditions, a concentration of 10.04 g l⁻¹ succinic acid, a volumetric productivity (Q  _P) of 0.478 g l⁻¹ h⁻¹ and an efficiency (E  _f) of 58.4% were obtained in 24 h (Fig. 5a). An increase in succinic acid production was subsequently obtained in a second bioreactor operated with a supply of external CO₂, resulting in a final succinic acid concentration of 14.22 g l⁻¹, a volumetric productivity of 0.677 g l⁻¹ h⁻¹ and an efficiency (E  _f) of 82.8% (Fig. 5b). A similar trend was reported by Lee et al. [13], who observed that 13.5 g l⁻¹ of succinic acid was produced when 100% CO₂ was supplied to the bacterium M. succiniciproducens. These outcomes were likely achieved because of the supply of carbon dioxide. Similarly, van der Werf et al. [26] suggested that the production of succinate requires CO₂ fixation and also that CO₂ concentration regulates the level of the key enzymes of the PEP carboxykinase pathway in A. succinogenes. These researchers further confirmed that high levels of CO₂ stimulated PEP carboxykinase levels, whereas the levels of alcohol dehydrogenase and lactate dehydrogenases were significantly decreased.

Succinic acid fermentation of sugarcane bagasse hemicellulose hydrolysate

The final verification of the statistical model for succinic acid production was also carried in a 2-l bioreactor. Figure 6 shows the behavior of the hydrolysate during the fermentation process under the conditions predicted in the statistical analysis. The hydrolysate was inoculated with a cell concentration of 1.02 g l⁻¹; the fermentation was run at 37°C for 24 h, pH 7.0, with an agitation of 150 rpm, and the initial xylose concentration was 52 g l⁻¹.

The succinic acid concentration reached its maximum value (19.0 g l⁻¹) after 24 h of fermentation, corresponding to a yield of product per substrate consumed of 0.465 g g⁻¹, a volumetric productivity of 0.903 g l⁻¹ h⁻¹ and a fermentation efficiency of 45.6% (Fig. 6a).

The effect of 10 g l⁻¹ of NaHCO₃ was similar to that of MgCO₃ used by Vemuri et al. [27], who obtained a final succinic acid concentration of 10.53 g l⁻¹; nonetheless, concentrations of NaHCO₃ higher than 10.0 g l⁻¹ did not significantly increase the succinic acid concentration. The results of this analysis also suggest strategies for improving CO₂ utilization. For example, at intermediate CO₂ concentrations, overexpression of the carbonic anhydrase enzyme, which catalyzes the hydration of CO₂ to HCO₃⁻, may be effective. If pure CO₂ were to be used, further improvement in succinate production should result from increased enzyme expression, as has been previously observed with pyruvate carboxylase [27].

Similarly, Fig. 6b shows the fermentation of hemicellulose hydrolysate in the presence of carbon dioxide, which was continuously introduced into the medium at 0.05 vvm throughout the process. The final succinic acid concentration, volumetric productivity and efficiency were 22.5 g l⁻¹, 1.014 g l⁻¹ h⁻¹ and 55.4%, respectively, with a fermentation time of approximately 24 h, at a temperature of 37°C, agitation of 150 rpm and pH 7.0 (Table 7).

Table 8 shows the results previously reported in the literature using various bacterial strains in succinic acid fermentation processes as well as the main results obtained in our study. A review of the relevant literature reveals that many published studies used other materials, such as wood hydrolysate and corn steep liquor [1, 14], and other microorganisms, such as naturally occurring M. succiniciproducens [9, 10, 13], A. succiniciproducens [12, 18] and genetically engineered E. coli [15, 19]. Other authors reported that A. succinogenes could use a wide range of carbohydrates as carbon sources and ferment straw hydrolysate or cane molasses, making this species a promising agent for succinic acid production [16].

Conclusions

The sequential experimental design strategy used to study the addition of nutrients in the fermentation medium was an important tool for maximizing final succinic acid concentration through RSM.

The results of our study on succinic acid production from sugarcane bagasse hemicellulose hydrolysate by means of A. succinogenes fermentation indicate that it is a promising procedure provided that an optimal combination of nutrients is used. Increased levels of NaHCO₃ and MgSO₄ were primarily responsible for the increased succinic acid concentration, followed by yeast extract and KH₂PO₄, which were less significant as independent variables, but revealed themselves to be important when they were involved in the interactions.

The conversion yield of succinic acid from sugarcane bagasse was relatively high in the batch cultivation of A. succinogenes. Under the best conversion conditions, dictated by the results of our statistical analysis and validated experimentally, a succinic acid concentration of 22.5 g l⁻¹ was achieved in hemicellulose hydrolysate fermentation.

Succinic acid will likely be one of the future platform chemical derivatives from renewable resources. Microbial production of succinic acid, in particular, may prove a novel “green technology,” since CO₂ fixation is integrated to carboxylation and decarboxylation in both catabolism and anabolism. Based on recent research results, cost reduction and scale-up are the main objectives of new process development in this area.

Acknowledgments

The authors are grateful to the Brazilian Council for Research (CNPq); the Rio de Janeiro Foundation for Science and Technology (FAPERJ) and the Brazilian Oil Company (PETROBRAS) for financial support of this research.

References