An efficient and large-scale preparation process for polysialic acid by Escherichia coli CCTCC M208088

Biochemical Engineering Journal 53 (2010) 97–103

Contents lists available at ScienceDirect

Biochemical Engineering Journal journal homepage: www.elsevier.com/locate/bej

An efficient and large-scale preparation process for polysialic acid by Escherichia coli CCTCC M208088 Jin-Long Liu, Xiao-Bei Zhan ∗ , Jian-Rong Wu, Chi-Chung Lin, Dan-Feng Yu The Key Laboratory of Industrial Biotechnology of Ministry of Education, School of Biotechnology, Jiangnan University, 1800 Lihu Avenue, Wuxi, Jiangsu Province 214122, PR China

a r t i c l e

i n f o

Article history: Received 13 May 2010 Received in revised form 22 August 2010 Accepted 21 September 2010

Keywords: Escherichia coli Polysialic acid Fermentation Purification

a b s t r a c t Polysialic acid (PSA) is a novel pharmaceutical material used in control release for protein drugs or in biomedical applications as scaffold. An efficient pilot production process for bacterial PSA was developed. Our PSA fermentation process by Escherichia coli CCTCC M208088 was optimized in a 500 L fermenter using a novel strategy by controlling pH with ammonia water feeding coupled with sorbitol supplementation. The resulting PSA level increased to 5500 mg/L as compared with the 1500 mg/L of the control. Furthermore, the process for the PSA purification from the fermentation broth was also established. PSA was isolated from the broth by ethanol precipitation, filtration with perlite as filter aid, followed by cetyl pyridinium chloride (CPC) precipitation and lyophilization. The final PSA product obtained had 98.1 ± 1.6% purity at 56.1 ± 1.7% recovery rate. Infrared spectroscopy and NMR spectroscopy analysis indicated that the structure of resulting PSA was identical to the published ␣-2,8 linked polysialic acid. © 2010 Elsevier B.V. All rights reserved.

1. Introduction Polysialic acid (PSA) is a polymer of sialic acid with the degree of polymerization usually between 8 and 200 residues linked by ␣2,8- and/or ␣-2,9-glycosidic (ketosidic) bonds which exist mostly in the terminal location of the glycoconjugates or the cell membrane surface of mammals and some bacteria [1,2]. As a result of its exterior surface location in certain biomolecules, PSA plays important roles in a variety of vital biological processes such as embryogenesis, neural cell growth, differentiation, cell–cell mediation and membrane transport [3]. PSA can be mainly used in the control release of drugs and as scaffold material in biomedical applications. PSA is a poor immunogen in humans and other mammals. It does not trigger the formation of antibodies required for phagocytic removal of the invasive organism [4]. On that basis, polysialylation of proteinbased drug leads to significant improvement of biological activity and prolongs the residence time of the conjugate in the blood circulation [5]. Furthermore, based on the fact that the biosynthesis of PSA in the adult brain induces and supports neuronal regeneration, combined with its trait of self-aggregates to form solids, PSA is suggested to be an ideal scaffold material for brain operation [6,7]. Due to limited availability, the market price of PSA is as high as US$200 per gram [8]. Bacteria such as Neisseria meningitides, Escherichia coli, Haemophilus ducreyi and Pasteurella haemolytica have been found to produce capsular PSA in their culture broth

∗ Corresponding author. Tel.: +86 510 85918299; fax: +86 510 85918299. E-mail address: [email protected] (X.-B. Zhan). 1369-703X/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.bej.2010.09.015

[9,10]. Several groups have been working on the production of PSA with different bacterial strains. However, their yields were rather low and could not meet commercial demand [11–13]. RodriguezAparicio et al. optimized the physical and chemical conditions on the production of PSA by E. coli K235 and their PSA production was 1350 mg/L [11]. Rode et al. optimized the culture medium for the production of PSA by E. coli K1 and the PSA production was higher than 1500 mg/L [12]. Recently, Kapre and Shaligram developed a process for bacterial PSA production in 30 L fermenter, but their resulting PSA production only reached 3000 mg/L [13]. During PSA production by E. coli, the fermentation medium pH has a significant effect on the biosynthesis of the PSA. RodriguezAparicio et al. found the environmental pH had a significant influence on the activity of key enzymes regulating the biosynthesis of PSA in E. coli [11]. The optimal pH for the PSA synthesis was determined in our previous work [14]. In addition to the medium pH, the nutritional composition of the medium also influences the PSA biosynthesis significantly, particularly the carbon and nitrogen sources. Sorbitol is the most commonly used carbon source in the PSA fermentation with E. coli. Previously, we demonstrated that the biosynthesis of the PSA was inhibited as sorbitol level higher than 40 g/L [15]. Furthermore, it was observed that ammonia concentration could severely affect the cell growth and PSA production with E. coli, and a flow injection analysis (FIA) system was employed by Honda et al. to monitor and control the ammonia in an optimal level of 0.3 g/L, which resulted in a maximum PSA production of 1900 mg/L [16]. Hitherto, few reports on the isolation of PSA from the fermentation broth have been published impacting the industrial production of PSA. To the best of our knowledge, there were only two reports

98

J.-L. Liu et al. / Biochemical Engineering Journal 53 (2010) 97–103

on the PSA purification from the fermentation broth. Kapre and Shaligram developed an entire process for the PSA purification, but their process needed a series of intricate and expensive column chromatographic processing [13]. Rode et al. also developed a process for PSA purification, but it worked in a small scale with only a 20% recovery yield [12]. In our previous report, we obtained a high-yield PSA strain E. coli by conventional mutation. Subsequently, higher PSA production was achieved in the lab level through optimization of the physical and chemical condition for the PSA fermentation [14,15,17]. In the present work, a complete pilot production process for bacterial PSA was developed. PSA fermentation process was optimized and verified in a 500 L fermenter. The process for the purification of bacterial PSA in pilot scale was also established. Our new, efficient and complete process has been shown to be capable of large scale production of high-purity PSA.

2.5 L/min. Ultrafiltration (MWCO 10 kDa) was subsequently conducted to reduce the liquid volume. For the purification of the PSA, a detailed treatment protocol (see Section 3) was developed. 2.4. Analytical methods A 0.5 ml fermentation broth was sampled, followed by centrifugation at 10,000 × g for 10 min. The amount of PSA in the supernatant of the culture was determined by using the resorcinol method [18]. The amount of sorbitol was tested using the periodic acid–chromotropic acid colorimetric method [19]. The level of ammonium chloride was assayed by the method of Nesseler [20]. Biomass was determined by drying the pellet at 80 ◦ C to a constant weight. Protein concentration was analyzed according to the Bradford method [21] and bovine serum albumin (BSA) (0.1 mg/ml) was used as standard. All results were the average of three independent replicate assays.

2. Materials and methods 2.5. IR spectroscopy 2.1. Microorganism and culture media The parent strain E. coli K235 is an original clinical isolate from a patient with urinary tract infection. In order to increase the PSA production, the parent strain E. coli K235 was mutated by various approach successively. It was firstly mutated by ultraviolet and then 60 Co radiation and subsequently underwent N-methylN-nitro-N-nitrosoguanidine (NTG) treatment as well as nitrogen ion implantation. During the mutation, the agar plate with 1% Bromothymol Blue indicator solution was used as the primary screening for the high-yield PSA strain. The fast-growing colonies with relatively large-size yellow zones were selected for further confirmation in the flask culture. Finally, a mutant strain, E. coli CCTCC M208088 was isolated. It demonstrated higher PSA productivity in shake flask cultivation, which was registered by the China Center for Type Culture Collection. Slant culture medium consisted of (g/L): NaCl 5, peptone 10, beef extract 3.0, agar 20 at pH 7.0. The seed medium-1 contained (g/L): NaCl 5, peptone 10, and beef extract 3.0 at pH 7.0. The seed medium-2 contained (g/L): sorbitol 20, ammonium chloride 0.5, K2 HPO4 2.5, MgSO4 0.9 and peptone 1.5 at pH 7.8. The medium for batch fermentation contained (g/L): sorbitol 40, ammonium chloride 5, K2 HPO4 2.5, MgSO4 0.9 and peptone 1.5 at pH 7.8. 2.2. PSA production with E. coli CCTCC M208088 in 500 L fermenter The production of PSA by E. coli CCTCC M208088 was carried out in a 500 L fermenter (Lehui Industry Company Ltd., Ningbo, China) with a working volume 300 L. One loop of cells from a fresh slant were transferred to 250 mL flask containing 30 mL seed medium-1 and grew on a rotary shaker for 8 h. The harvested seed culture was inoculated into a 3000 mL flask with 1000 mL seed medium-2 and incubated for 12 h. Then the prepared seed culture was pumped into fermenter with 4% (v/v) inoculation size. The basic fermentation parameters in 500 L fermenter were: temperature 37 ◦ C, aeration 1.5 vvm (volume of air per volume of medium per min) and agitation 300 rpm. Variable parameters and fermentation nutrient supplementation employed are listed in Section 3. 2.3. Purification of PSA from the fermentation broth After harvest, the fermentation broth was pasteurized at 80 ◦ C for 30 min. The following separation of biomass was achieved by continuous centrifugation at 12,000 × g with a flow rate of

The purified PSA was extruded to slice with potassium bromide, and then scanned by the infrared spectrometer IR-440 (Shimadzu, Japan) in the range of 4000–400 cm−1 . 2.6. NMR spectroscopy The structure of the purified PSA was determined with a Bruker BioSpin Corporation (Billerica, Massachusetts, USA) DPX 400 NMR at 35 ◦ C. The PSA was dissolved in D2 O at 80 mg/ml. The 13 C 1D spectra were recorded at 100 MHz with 13 C carrier at 90 ppm (sw 250 ppm), acquiring 64k complex points (acquisition time 1.3 s), and referenced to an internal TMS standard at 0 ppm. 3. Results and discussion 3.1. Optimization of fermentation process to enhance PSA production Microbial fermentation for PSA production with E. coli CCTCC M208088 was carried out in a 500 L pilot fermenter. As shown in Fig. 1A, sorbitol and ammonium chloride (NH4 + ) were rapidly consumed for bacterial growth and PSA production. Ammonium chloride was exhausted at 16 h of the fermentation process. pH level sharply decreased from initial pH 7.2 to the low value of 4.0 due to the accumulated acidic metabolites and subsequently flatten up to a plateau level. The maximum biomass of 7.6 g/L and PSA production of 1535 mg/L were achieved by the end of the fermentation. It was noteworthy that a steep decline of the bacterial growth and PSA synthesis occurred at the moment when ammonium chloride was exhausted as well as pH decreased to its minimum level. Thus, we inferred that the exhaustion of the nitrogen source and the lower pH at the early stage of fermentation may negatively influence the bacterial growth and PSA synthesis. Environmental pH has a significant influence on the PSA synthesis and the optimal pH level (pH 6.4) for the PSA fermentation by E. coli have been reported in our previous study [11,14]. In addition, constant-speed feeding of NH4 Cl solution during the PSA fermentation employed in our previous study had shown a positive effect on stimulating cell growth, but an excess accumulation of NH4 Cl was usually observed at the stationary phase, which hindered the further synthesis of PSA and caused some problems during the subsequent purification of PSA, particularly in the steps of ethanol precipitation and CPC treatment. Therefore, in order to supply adequate nitrogen source as well as control pH in an optimal level for the PSA fermentation, a novel strategy of feeding ammonia water for control pH at 6.4 was adopted. Using this method, nitrogen

J.-L. Liu et al. / Biochemical Engineering Journal 53 (2010) 97–103

99

Table 1 Results of PSA fermentation by various methods. Parameters

Fermentation time (h) Biomass (g/L) PSA production (mg/L) PSA productivity (mg/(L h)) PSA yield per gram cell (mg/g)

PSA fermentation Aa

Bb

Cc

56 7.64 ± 0.19 1535 ± 73 27.4 ± 1.3 201 ± 4.5

40 11.86 ± 0.28 2566 ± 130 64.2 ± 3.3 216 ± 5.8

28 16.45 ± 0.34 5505 ± 232 197.0 ± 8.3 334 ± 6.7

a

PSA fermentation A: PSA fermentation without pH control. PSA fermentation B: PSA fermentation with feeding of ammonia for pH control. c PSA fermentation C: PSA fermentation with feeding of ammonia for pH control and sorbitol supplementation. b

Fig. 1. PSA fermentation by various methods in 500 L fermenter: biomass (), PSA (), sorbitol (), NH4 + (), pH (–). (A) PSA fermentation without pH control. (B) PSA fermentation with feeding of ammonia for pH control. (C) PSA fermentation with feeding of ammonia for pH control and sorbitol supplementation.

source – NH4 + could be maintained in a proper level at around 0.06 mol/L (Fig. 1 B), which could further stimulated the cell growth and PSA synthesis. As a result, a maximum biomass of 11.86 g/L and PSA production of 2566 mg/L were attained, which were 55.2% and 67.2% higher than those of the method without pH control, respectively. However, a new problem derived from this method was observed – the carbon source sorbitol was consumed so rapidly for bacterial growth stimulated by the NH4 + that it was used up at the early stage of the fermentation (Fig. 1 B). Thus, it was necessary to increase the total sorbitol level in the fermenter as pH was controlled with feeding ammonia water during the PSA fermentation. Since high initial substrate level usually inhibits bacterial growth and the biosynthesis of the PSA would be inhibited as sorbitol level ≥40 g/L [15,22]. A 6000 g sorbitol, equal to 1/2 of the initial amount of sorbitol, was pumped into the fermenter at 16 h of the process. As a result, the maximal level of 16.45 g/L biomass and 5505 mg/L PSA were attained (Fig. 1 C), respectively, which were the highest

level reported up to now. In addition, the fermentation time was decreased to 28 h as compared to the 56 h of the method without pH control. As shown in Table 1, substantial enhancement of 2.5 folds in PSA production and 6.2 folds in PSA productivity were achieved through our process optimization, respectively. Sialic acid lyase and sialyltransferase play important roles in regulating the PSA synthesis in E. coli, and their activity can be drastically influenced by the environmental pH [11,23]. Aisaka et al. reported that sialic acid lyase was most active at pH 6.5–7.0, which was in agreement with our experimental results [23]. Various carbon sources such as xylose, glucose and sorbitol were used in the PSA fermentation with E. coli in different reports, but sorbitol showed the highest capability in supporting the PSA biosynthesis [11–13]. Furthermore, sorbitol level had a significant effect on the PSA synthesis during the PSA fermentation, and the optimal sorbitol level – 20–40 g/L for the PSA synthesis was observed in our previous study [15]. Vimr and Troy reported that sialic acid lyase from E. coli is a sialic acid-inducible and glucose-repressive enzyme [24]. It may be the reason why the conventional carbon source glucose is not competitive in supporting PSA synthesis as compared to sorbitol. Sorbitol is a carbohydrate that can induce osmotic pressure for the microorganisms. Conceivably, the osmotic stress imposed by sorbitol may simulate the PSA synthesis and the bacteria which are generally isolates from infected urinary tract could survive by enhancing bacterial virulence with the PSA polymer. Moreover, the ammonia level had a remarkable effect on the PSA synthesis during the fermentation, and the suitable ammonia level – about 0.3 g/L was reported by Honda et al. in which casamino acid feeding was controlled by a flow injection analysis (FIA) system to keep the ammonia at a optimal level [16]. This was similar with our experimental results that maintaining the ammonia at around 0.06 mol/L by feeding of ammonia water. Consequently, a suitable level of environmental pH, carbon and nitrogen source were critical in the PSA fermentation with E. coli. In the present study, by applying the developed strategy of controlling pH with ammonia water feeding coupled with sorbitol supplementation, medium pH, NH4 + and sorbitol could be maintained in a proper level for PSA fermentation, resulting in the significant improvement of the PSA production. 3.2. Purification of PSA from the fermentation broth The impurities in the E. coli CCTCC M208088 fermentation broth include biomass, culture medium residuals and proteins. After harvest, the fermentation broth was processed as described in Section 2. To recover PSA from the fermentation broth, various physical and chemical treatments were tested as follows. In a typical microbial polysaccharide purification process, precipitation of the polysaccharide with ionic surfactant from crude extract is the most common primary recovery method [25,26]. In the present work, one of typical cationic surfactants: cetyl pyridinium chloride (CPC) was utilized to separate the PSA from the

100

J.-L. Liu et al. / Biochemical Engineering Journal 53 (2010) 97–103

Fig. 2. Effect of CPC and NaCl levels, temperature as well as pH on PSA recovery.

fermentation broth. Factors such as temperature, pH of aqueous solution, NaCl and CPC level, were investigated in order to optimize the precipitation process (PSA recovery was calculated based on the distribution of PSA in the precipitate and supernatant). As shown in Fig. 2, that the amount of CPC added was a crucial factor leading to PSA precipitation. The experimental results indicated that 3 g CPC/g PSA was the optimal level for the precipitation process. PSA recovery decreased as the CPC level higher or lower than this critical point. As to the effect of NaCl, PSA recovery decreased from 95 to 60% with the addition of NaCl, as NaCl concentration increased from 0 to 0.2 mol/L. Furthermore, with the increase of the pH, PSA recovery experienced an initial rise, then underwent a slight decrease and maintained more or less at the same level thereafter. The highest PSA recovery was obtained in the range of pH 6–7. With increasing temperature, PSA recovery experienced an initial flat plateau and subsequently a sharp decrease at 40 ◦ C. Based on our experimental results, the optimal condition for the CPC precipitation recovery of PSA was optimized as follows: addition of 3 g CPC/g PSA, pH of 6–7, NaCl of 0.1 mol/L or less and temperature lower than 40 ◦ C. It is well known that interactions between ionic surfactants and oppositely charged polysaccharide are quite strong in aqueous solution resulting in the precipitation of the polysaccharide [27,28]. Several types of molecular interactions may occur in the formation of surfactant–polysaccharide precipitate, including electrostatic force, hydrophobic and hydrogen bonding [29,30]. Bacterial PSA is negatively charged due to the high content of carboxyl groups which could be completely neutralized by the positively charged CPC at a certain mass ratio. The optimal CPC level determined for PSA recovery was 3 g CPC/g PSA. This determination was similar to the results reported by Goddard and Hannan in which the polymer–surfactant complexes can be dissolved by excessive level of surfactant and the maximum precipitate appeared to correspond to a single layer of surfactant adsorbed by the polymer [30]. As for the redissolution by NaCl, the PSA recovery declined at higher NaCl level as the electrostatic attraction of PSA and CPC was reduced by the screening effect which was caused by the

high salt concentration used [29]. In addition, the environmental pH had an obvious influence on both the electrostatic force and hydrogen bonding of the surfactant–polysaccharide complex [29]. Temperature is another vital factor for polysaccharide precipitation. Usually, the redissolution of the polysaccharide precipitate is energy-consuming. Thus it was observed that the CPC–PSA precipitate was generated at a lower temperature and dissolved at a higher temperature (≥40 ◦ C). In addition to CPC, PSA can also be precipitated by ethanol. Ethanol can lower the dielectric constant of the aqueous solution, decreasing the polysaccharide–solvent affinity, promoting the insolubility of the polysaccharide leading to the precipitation of polysaccharide [31]. We found that ethanol was not only capable of separation the PSA from the broth and it also can wash out impurities such as coloured components, extraneous salts as well as the remaining cells. However, neither the ethanol nor the CPC precipitation can effectively separate the PSA and contaminative proteins since most of the PSA and proteins would be precipitated simultaneously as the fermentation broth was treated with ethanol or CPC (Table 2). Proteins were the main impurities existed in the resulting PSA product. For the evaluation of the separation of proteins from the PSA, the PSA to proteins ratios in the supernatant and precipitate were analyzed, as shown in Table 2. Both of the treatment with CPC and ethanol resulted in a complete precipitation of the PSA. During the CPC treatment, protein content of the precipitate reduced by 50% while the application of ethanol just led to a removal of 10% of the proteins. To achieve complete depletion of the contaminating proteins, another effective treatment method must be adopted. It was observed that proteins in the PSA fermentation broth could be effectively absorbed by perlite. Thus, filtration with perlite as the filter aid was demonstrated to be an effective method for the removal of proteins in the PSA fermentation broth. As the efficiency of absorption between perlite and protein was dose and pH-dependent, the efficiency of protein removal was mainly determined by the level of perlite and the pH of the aqueous solution. As shown in Fig. 3, protein removal capacity was enhanced with a

J.-L. Liu et al. / Biochemical Engineering Journal 53 (2010) 97–103

101

Table 2 Distribution of PSA and proteins in the precipitation and supernatant in CPC as well as ethanol treatment. Methods

CPC treatment Ethanol treatment

PSA

Protein

In supernatant

In precipitation

In supernatant

In precipitation

1.42 ± 0.33% 2.33 ± 0.29%

96.14 ± 0.79% 95.23 ± 2.34%

50.32 ± 0.93% 10.32 ± 1.03%

47.35 ± 0.47% 87.32 ± 2.12%

higher perlite use level. However, the recovery of PSA decreased. Similarly, an obvious improvement in the removal of proteins with a distinct decrease in PSA recovery yield was observed with increasing pH. Higher perlite use level and pH are favorable for the adsorption of proteins by the perlite. Thus, during the filtration with perlite as filter aid, removal of the proteins can be improved by increasing the level of perlite and pH of the aqueous solution. However, higher level of perlite and pH may be problematic during the filtration process, leading to the polymer retaining in the filter

cake layer and notable reduction of the PSA recovery. Consequently, the ideal condition for filtration had to be a tradeoff between the recovery yield and protein removal. As a result, pH 10 and 8% (m/v) of perlite was determined to be the optimum condition for the filtration. To achieve a thorough purification of the product at a low cost, the above steps were combined together in a defined order: “ethanol precipitation → filtration → CPC precipitation → ethanol precipitation”. Ethanol precipitation was followed by filtration with

Fig. 3. Effect of perlite use level and pH on the filtration purification of PSA. Removal of protein (), recovery of PSA ().

Fig. 4. Flow chart of pilot production process for PSA.

102

J.-L. Liu et al. / Biochemical Engineering Journal 53 (2010) 97–103

4. Conclusion An efficient and large-scale production process for PSA by E. coli CCTCC M208088 has been developed. In order to improve the PSA production process, a novel strategy of controlling pH with ammonia feeding coupled with carbon source supplementation was implemented. Using this strategy, significant enhancement of PSA production was achieved. Improved processes for the recovery of PSA from the fermentation broth, effective purification methods including ethanol precipitation, filtration and CPC precipitation, were introduced and optimized. Using our new improved processes for PSA purification, better than 98% high-purity PSA at over 50% total recovery was achieved. Based on our current findings, a commercially feasible production process for PSA would be attainable. Acknowledgements Fig. 5. Infrared spectrum of the prepared PSA.

perlite as filter aid. The following CPC precipitate led to further isolation of the PSA from the impurities. Subsequently, the resulting precipitate of PSA–CPC was dissolved in 0.8 mol/L NaCl solution followed by another ethanol precipitation treatment. Finally, PSA product was obtained after the precipitate was dried in vacuum to constant weight. The whole purification process was summarized in the flow chart shown (Fig. 4). Using this purification protocol, the final PSA product obtained had 98.1 ± 1.6% purity at 56.1 ± 1.7% recovery rate. The remaining trace impurities in the resulting PSA product contain endotoxin, protein, nucleic acid, etc. The method for removing them is currently under development in our laboratory. During the purification, loss of PSA mainly occurred at filtration. As filtration was conducted in a plate-frame filter with perlite as filter aid and nylon filter cloth, seeking more effective material to substitute perlite and nylon filter cloth may further improve PSA recovery yield. Current work in our laboratory has been undertaken in this direction.

3.3. Characterization of the resulting polysialic acid As shown in Figs. 5 and 6, the purified PSA product was analyzed by infrared spectroscopy and 13 C NMR. The infrared spectra were compared with the published spectra of the pure PSA and the characteristic absorption peaks of our purified PSA were similar to that from the published report [32]. The NMR spectra were also comparable to those of the published spectra of ␣-2,8 linked PSA and all peaks can be definitely identified as well as allocated accordingly [12,33].

Fig. 6.

13

C NMR spectra of the prepared PSA.

Funding for this project was provided by the National High Technology Research and Development Program of China (2006AA02Z207) and by Research Fund for the Doctoral Program of Higher Education of China (20077434). References [1] T. Angata, A. Varki, Chemical diversity in the sialic acids and related alpha-keto acids: an evolutionary perspective, Chem. Rev. 102 (2002) 439–469. [2] F.A. Troy, Sialobiology and the Polysialic Acid Glycotope: Occurrence, Structure, Function, Synthesis, and Glycopathology, Plenum Press, New York, 1996, pp. 95–144. [3] E.R Vimr, K.A. Kalivoda, E.L. Deszo, S.M. Steenbergen, Diversity of microbial sialic acid metabolism, Microbiol. Mol. Biol. Rev. 68 (2004) 132–153. [4] D.E. Wunder, W. Aaronson, S.F. Hayes, J. Bliss, R.P. Silver, Nucleotide sequence and mutational analysis of the gene encoding KpsD, a periplasmic protein involved in transport of polysialic acid in Escherichia coli K1, J. Bacteriol. 176 (1994) 4025–4033. [5] G. Gregoriadis, S. Jain, I. Papaioannous, P. Liang, Improving the therapeutic efficacy of peptides and proteins: a role for polysialic acids, Int. J. Pharm. 300 (2005) 125–130. [6] S. Bruns, Y. Stark, M. Wieland, F. Stahl, C. Kasper, T. Scheper, Fast and efficient screening system for new biomaterials in tissue engineering: a model for peripheral nerve regeneration, J. Biomed. Mater. Res. 81A (2007) 736–747. [7] Y. Stark, S. Bruns, F. Stahl, C. Kasper, M. Wesemann, C. Grothe, T. Scheper, A study on polysialic acid as a biomaterial for cell culture applications, J. Biomed. Mater. Res. 85 (2007) 1–13. [8] Nacalai USA Inc., 2008, http://www.nacalaiusa.com/product.php?id=33, accessed 01/06/09. [9] M.R. Mizanu, N.L. Pohl, Bacterial CMP-sialic acid synthetases: production, properties, and applications, Appl. Microbiol. Biotechnol. 80 (2008) 757–765. [10] M. Ringenberg, C. Lichtensteiger, E. Vimr, Redirection of sialic acid metabolism in genetically engineered Escherichia coli, Glycobiology 11 (2001) 533–539. [11] L.B. Rodriguez-Aparicio, A. Reglero, A.I. Ortiz, J.M. Luenge, Effect of physical and chemical conditions on the production of colominic acid by Escherichia coli in a defined medium, Appl. Microbiol. Biotechnol. 27 (1988) 474–483. [12] B. Rode, C. Endres, C. Ran, F. Stahl, S. Beutel, C. Kasper, S. Galuska, R. Geyer, M. Muhlenhoff, R. Gerardy-Schahn, T. Scheper, Large-scale production and homogenous purification of long chain polysialic acids from E. coli K1, J. Biotechnol. 135 (2008) 202–209. [13] S.V. Kapre, U. Shaligram, Process for the preparation of highly pure polysialic acid of high molecular weights, PCT Int. Patent Application WO 2008035373 (2008). [14] X.B. Zhan, L. Zhu, J.R. Wu, Z.Y. Zheng, W. Jia, Production of polysialic acid from fed-batch fermentation with pH control, Biochem. Eng. J. 11 (2002) 201–204. [15] J.L. Liu, J.R. Wu, F.D. Shen, D.F. Yu, Q. Zhang, X.B. Zhan, A new strategy to enhance polysialic acid production by controlling sorbitol concentration in cultivation of Escherichia coli K235, Afr. J. Biotechnol. 9 (2010) 2422–2426. [16] H. Honda, T. Nakazeko, K. Ogiso, Y. Kawase, N. Aoki, M. Kawase, T. Kobayashi, Colominic acid production from Escherichia coli in a fed-batch culture under the control of ammonium ions using an FIA system, J. Ferment. Bioeng. 83 (1997) 59–63. [17] X.B. Zhan, Z.Y. Zheng, W. Jia, L. Zhu, The breeding of Escherichia coli 60 Co mutagensis strain and the condition on the production of polysialic acid, J. Food Sci. Biotechnol. (Chinese) 19 (2000) 315–318. [18] L. Svennerholm, Quantitative estimation of sialic acids, Biochim. Biophys. Acta 24 (1957) 604–611. [19] M. Pesez, J. Bartos, Colorimetric and Fluorometric Analysis of Organic Compounds and Drugs, Marce Dekker, New York, 1974, pp. 291–328. [20] J.E. Harwood, D.J. Huysen, Automated analysis of ammonia in water, Water Res. 4 (1970) 695–704.

J.-L. Liu et al. / Biochemical Engineering Journal 53 (2010) 97–103 [21] M.M. Bradford, A. Rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding, Anal. Biochem. 72 (1976) 248–254. [22] H. Huang, D. Ridgway, T. Gu, M. Moo-Young, Enhanced amylase production by Bacillus subtilis using a dual exponential feeding strategy, Bioprocess Biosyst. Eng. 27 (2004) 63–69. [23] K. Aisaka, A. Igarashi, K. Yamaguchi, T. Uwajima, Purification, crystallization and characterization of N-acetylneuraminate lyase from Escherichia coli, Biochem. J. 276 (1991) 541–546. [24] E.R. Vimr, F.A. Troy, Identification of an inducible catabolic system for sialic acids (nan) in Escherichia coli, J. Bacteriol. 164 (1985) 845–853. [25] A. Nimrod, B. Greenman, D. Kanner, Method of producing high molecular weight sodium hyaluronate by fermentation of Streptococcus, US Patent 4780414, 1988. [26] R. Kurane, H. Matsuyama, Production of Bioflocculant by mixed culture, Biosci. Biotechnol. Biochem. 58 (1994) 1589–1594.

103

[27] K. Shirahama, The Nature of Polymer–Surfactant Interactions, Marcel Dekker, New York, 1998, pp. 143–191. [28] M. Ishiguro, W.F. Tan, L.K. Koopal, Binding of cationic surfactants to humic substances, Colloids Surf. A: Physicochem. Eng. Aspects 306 (2007) 29–39. [29] F.P. Yang, M.C. Chin, K.C. Lee, Interaction of hyaluronic acid (HA) with organosilicon (Si-QAC) modified magnetite for HA recovery, Sep. Sci. Technol. 42 (2007) 1747–1760. [30] E.D. Goddard, R.B. Hannan, Polymer/surfactant interactions, J. Am. Oil Chem. Soc. 54 (1977) 561–566. [31] I.H. Smith, G.W. Pace, Recovery of microbial polysaccharide, J. Chem. Technol. Biotechnol. 32 (1982) 119–129. [32] K. Hotta, M. Kurokawa, Isolation and identification of two sialic acids from the jelly coat of sea urchin eggs, J. Biol. Chem. 245 (1970) 6307–6311. [33] W. Egan, T.Y. Liu, Structural studies on the sialic acid polysaccharide antigen of E. coli strain Bos-12, Biochemistry 16 (1977) 3687–3692.