Abstract
Hyaluronic acid (HA) dispersion obtained from the bacteria Streptococcus equi was concentrated by electrofiltration. In the conventional downstream processing of HA, extraction and precipitation lead to increase in environmental issues, structural changes, and time and energy related costs. Using electrofiltration as an alternative technology delivers solutions to these limitations. Experiments were conducted in order to test the applicability of electrofiltration to downstream processing of the negatively charged HA. The structural changes and molecular weight distributions, often a consequence of the employed separation method, were tested by analysis of the initial dispersions and final products. In comparison to the conventional filtration, concentration factors were increased up to almost four times without any detectable structural change in the final product.
Keywords: Concentration, Electrofiltration, Filtration kinetics, Hyaluronic acid, Product quality
Abbreviations
- HA
hyaluronic acid
- RI
refractive index
1. Introduction
Hyaluronic acid (HA) is an anionic linear polysaccharide composed of alternating glucuronic acid and N‐acetylglucosamine units with a molecular weight in the range of 104–107 Da depending on its origin and production technology 1, 2, 3. HA is present in the extracellular matrix of vertebrate epithelial, neural, and connective tissues. Physicochemical properties such as biodegradability, hydrophilicity, and biocompatibility define HA as an important biopolymer with a variety of biomedical, pharmaceutical, and cosmetic applications 4, 5, 6. Depending on applications, the value of HA products varies from US$ 2000 to 60,000 per kilogram 7.
Commercially, HA is either extracted from rooster combs or produced by fermentation of certain strains of Streptococcus, which synthesize the biopolymer as part of their outer capsule 8, 9, 10. Extraction of HA from rooster combs has some disadvantages such as probability of inflammatory reactions. Additionally, HA obtained from animal sources forms complexes with proteoglycans, which complicates the purification process and increases production costs 11. Besides these disadvantages, optimization potential of product yield and quality through control of cultural conditions and genetic engineering justify the recent increased interest in the microbial production of HA 7.
Common separation and purification of HA mainly comprises precipitation with organic solvents such as ethanol, acetone, isopropanol, etc. Nevertheless, the use of organic solvents is a time‐consuming process, not environmentally friendly and also can lead to change in the quality of HA 12, 13. Molecular weight distribution being a key parameter for the quality of the product is highly affected by the applied purification strategy 14. In biotechnology industry, the current downstream processing trends are directed toward integrated, faster, and more effective processes, incorporating mild operational conditions. Alternative approaches including membrane filtration can be introduced to solve the technological challenges. However, pressure‐driven membrane processes lead to membrane fouling and subsequent permeate flux decline 15.
On the other hand, most biopolymers are charged and application of an electrical field leads to the migration of molecules in electric field 16. The electric charge of HA enables its purification by electrofiltration, which is a separation technique combining membrane filtration and electrophoresis in a dead‐end process. The aim of this work was to investigate the applicability and potential of electrofiltration as an alternative downstream strategy in biotechnological production of HA. By application of both pressure and electric field, an improvement of the filtration kinetics was expected. Structural changes in HA and molecular weight determinations are the critical factors affecting the applicability.
2. Materials and methods
2.1. Materials and experimental conditions
HA was purchased from Sigma‐Aldrich (Seelze, Germany) with product number 53747, which was produced by Streptococcus equi. Investigations were performed in a lab‐scale electrofiltration cell designed with flushing electrodes (Fig. 1) and the detailed experimental approach was previously described 17. Online control of filtrate mass, pH, and conductivity was performed in order to ensure the proper process realization. The use of flushing chamber system enables pH and temperature control, prevents the contact of the biopolymer with electrodes, and rinses out the electrolysis products formed at the electrodes. The various buffer solutions, which were prepared according to the conductivity and pH of the initial HA dispersions, contained potassium dihydrogen phosphate–dipotassium hydrogen phosphate (buffer I), potassium hydrogen phthalate–NaOH (buffer II), and acetic acid–sodium acetate (buffer III). HA was dispersed in demineralized water with a concentration of 2 g/L and pH of 5.6. In order to obtain reproducible results, all experiments were repeated several times and new membranes with a surface area of 0.001 m2 (polyethersulfone, 0.1 μm; Pall, USA) were used for each experiment.
Figure 1.
Lab‐scale electrofiltration cell. Product chamber (in the middle) and flushing chambers (on the left and right sides).
2.2. Characterization methods
2.2.1. Zeta potential measurements
Zeta potential is an important characteristic defining the stability and electrophoretic mobility of the products. In order to evaluate the electrofiltration capacity, zeta potential measurements of HA dispersions were performed using Malvern Zetasizer 5000 with PCS software V1.52, Rev1 (Malvern Ltd, UK). Each sample was measured five times and a purchased standard solution (Malvern Ltd.) was used for calibration.
2.2.2. Molecular weight distribution
As a key parameter for the quality of the product, the molecular weight of HA was determined by using SEC combined with multiangle laser light scattering (DAWN DSP; Wyatt Technology Corporation, Santa Barbara, CA, USA) and refractive index (RI) detection. The system was composed of an Agilent 1100 Series Chromatography system with an RI detector using a TSKgel G6000PWxl column (Tosoh Bioscience GmbH, Stuttgart, Germany) of 7.8 mm inner diameter and 30 cm length with a guard column. The multiangle laser light scattering detector was calibrated with Toluene (VWR, Germany) and normalized with BSA (5 mg/mL) in 30 mM ammonium acetate buffer. RI calibration constants were generated with anhydrous sodium chloride. Analyses were performed at room temperature using ammonium acetate buffer as the mobile phase at a flow rate of 0.8 mL/min. A sample volume of 100 μL was injected and the average of three replicates was used for evaluation.
2.2.3. FT‐Raman spectroscopy
The influence of the purification technology upon the structure of the product was estimated by applying FT‐Raman spectroscopy. The used FT‐spectrometer Vertex 80 with attached Raman module RAM II (Bruker Optik GmbH, Germany) employed a 1064 nm NdYAG excitation laser, an interferometer, and a high sensitivity near infrared Ge‐detector. For Raman analysis, an adequate amount of each sample was lyophilized for at least 20 h. The applied laser power was maintained at 400 mW in order to improve data comparison. The received spectra were evaluated by OPUS 6.5 software (Bruker Optik GmbH).
3. Results and discussion
3.1. Filtration kinetics is influenced by process parameters
The effects of applied electric field strength and pressure were investigated by varying both parameters. The t f/V f (filtration time/filtrate volume) versus V f (filtrate volume) graphs interpreted the filtration characteristics based on the slope alteration of the curves, describing the specific filter cake resistances. The standard progress of electrofiltration experiments has three distinct sections divided according to the changes in specific filter cake resistances 18. In the section I due to the higher effect of hydrodynamic resistance force in comparison to the electrophoretic force, the specific filter cake resistance is high. After the equilibrium of the forces in section II, no further filter cake is formed on the cathode side for a negatively charged product. In the section III, as a result of the constant accumulation of product on the anode side membrane, the formed filter cake reaches the cathode side filter film, increasing the specific filter cake resistance.
In order to determine the influence of specific buffer solutions on the filtration kinetics of HA dispersions, several experiments with various buffer compositions were conducted. The t f/V f versus V f graph of the applied buffer solutions is shown in Fig. 2.
Figure 2.
t f/V f versus V f diagrams of HA dispersions under conditions of 2 V/mm and 1 bar with various buffer solutions (pH 5.6, temperature: 15°C) used. Buffer I: dihydrogen phosphate–dipotassium hydrogen phosphate, Buffer II: potassium hydrogen phthalate–NaOH, and Buffer III: acetic acid–sodium acetate.
All curves in Fig. 2 representing specific filter cake resistances exhibit almost the same tendency in the first two sections of the curves. However, the experiments using buffer II and buffer III delivered, in the third section, a higher specific filter cake resistance in comparison to the experiment with the buffer I. The result could be explained by the conductivity of the membranes, which depends on the variety of ions. Concerning the lower specific filter cake resistance, the use of buffer I led to higher permeate mass for the same experimental time and therefore was used for further experiments. The experiments demonstrated that different buffer solutions have different influences on filtration kinetics and could be further investigated for a detailed consideration.
The effect of various process parameters upon the efficiency of electrofiltration of HA dispersions was investigated by conducting different experiments with 1, 2, and 4 bars overpressure with and without applied electric field. The effect of applied pressure on filtration kinetics of HA dispersions is demonstrated in Fig. 3 based on the gradient of filtrate mass per membrane surface area.
Figure 3.
Filtrate mass/membrane area versa filtration time diagrams of HA dispersions at 0 V/mm with different applied overpressures (pH 5.6, temperature: 15°C). Gradient of filtrate mass by the increase in pressure presents the compressibility of the product.
According to Darcy's law, application of higher pressure results in more mass of filtrate. However, compressibility of the biopolymers hinders the proportional increase in filtrate mass with pressure. According to the initial HA concentration (2 g/L), concentration factors were 2.8, 4.5, and 5 for 1, 2, and 4 bars conditions, respectively, and the obtained corresponding average filtrate fluxes were determined to be 15, 23, and 25 L/m2h. The differences between the permeate masses obtained with 2 and 4 bars are insignificant within the experimental frame. Figure 3 demonstrates that application of higher pressure compresses the HA filter cakes, increasing filter cake resistances. Without application of electric field, the hydrodynamic resistance force is the only force that has effect on the filter chamber and the filter cakes are formed on anode and cathode side membranes. The compression results in the blocking of the membranes on both sides, preventing the filtration capacity of HA dispersions.
The effect of applied electric field is demonstrated in Fig. 4 in terms of t f/V f versus V f and compared with conventional filtration. HA contains carboxylic groups designating its negative charge 19. The zeta potential of HA dispersion was determined to be –37 mV at pH 6. Based on the high migration capability of the molecules, experiments revealed promising results considering the filtration technology of HA dispersions. The charge facilitates the separation of the biopolymer by means of electric field. The application of electric field leads to a decrease in specific filter cake resistances as demonstrated by the slopes of the curves. The experiment without electric field is characterized by a steeper curve compared to the ones derived after application of electric field of 2 and 4 V/mm. The results present the high positive effect of electric field on the filtration kinetics of HA. This result is based on the electrophoretic migration of negatively charged HA molecules toward anode side membrane. Cathode side membrane remains almost free of filter cake allowing an undisturbed perfusion of the filtrate. The experiments under electric field follow the already described theoretical division of the electrofiltration process into sections. However, only the experiment with 4 V/mm reached the third section where the specific filter cake increases significantly. The result was caused by the rapid filtration of the dispersion with higher applied electric field strength. The concentration factors at 4 bars were 5, 15, and 18 for 0, 2, and 4 V/mm conditions, respectively, and the corresponding average filtrate fluxes were determined to be 22, 60, and 80 L/m2h for the elapsed time. In comparison to conventional filtration, the concentration factor based on the permeate mass of samples was increased up to almost four times for the same experimental time for HA dispersions, when 4 V/mm electric field strength under 4 bars was applied. For a defined filtrate mass (49 g), the filtration duration was 3 h without application of electric field, whereas for the same amount of filtrate mass the duration was only 14 min representing a time saving of approximately 92%. The composition of filter cake after application of 4 bars and 4 V/mm is presented in Fig. 5. Furthermore, the energy need was 8100 J/100 g filtrate, whereas energy requirement for a conventional method such as extraction is higher 20. In case of higher salt contents as may be present in real fermentation broths, electric current increases with a negative impact on energy balance 21. Ongoing research is targeted to overcome the limitations by in situ desalination during the filtration process.
Figure 4.
t f/V f versus V f diagrams of HA dispersions at 4 bars with different applied electric field strengths (pH 5.6, temperature: 15°C). The slope of the curves is characteristic for the specific filter cake resistances.
Figure 5.
Filter cake composition of HA at 4 bars and 4 V/mm applied. The product has high viscosity with a transparent appearance.
3.2. Product characteristics are not significantly influenced by electrofiltration parameters
The alteration of the chemical structure of HA is another criterion to evaluate the applicability of electrofiltration. In order to detect possible changes in the structure due to the applied electric field, samples which were taken before and after electrofiltration were analyzed using FT‐Raman spectroscopy. The spectra of the HA samples derived from experiments under maximum applied pressure (4 bars) with 0, 2, 4 V/mm are presented in Fig. 6 together with the sample of unprocessed HA dispersion.
Figure 6.
Raman spectra of HA samples before and after electrofiltration at a pressure of 4 bars and applied electric field strenghts of 0, 2, and 4 V/mm. No structural deformation was detected by the increase in electric field strength.
The main bands are composed of N‐H stretching (1), C‐H stretching(2), Amid I and C=C stretching (3), C–N stretching and C–H deformation (4), C–H bend (5), Amid III (6), and C–C and C–O stretchings (7) 22. The results reveal no significant difference between the spectra before and after electrofiltration and, respectively, the different experimental conditions. Therefore, the application of electrofiltration expresses no negative effect on the structure of HA molecules. Slight differences between the intensity of the HA spectra were caused by diminutive differences in molecule size. However, in general there was no detectable oxidative or reductive change in any functional group induced by electrofiltration.
Another important product characteristic of HA influencing its properties is the molecular weight. With (4 V/mm) and without applied electric field (0 V/mm), the average molecular weight remained in the range of 451–462 kDa (Table 1). Together with the Raman spectra, no structural change is revealed with respect to molecular weight. The evaluation of possible structural changes and characteristics arising from the purification method demonstrated that electrofiltration is a relevant and promising technique in downstream processing of HA delivering required unchanged specifics.
Table 1.
The weight average molecular weights (M w) and the number average molecular weights (M n) of samples obtained after application of 0, 2, and 4 V/mm
| Electric field strength (V/mm) | M w in Da ± SD | M n in Da ± SD | Polydispersity |
|---|---|---|---|
| 0 | 4.51 E5 ± 1.3% | 6.29 E5 ± 1.3% | 1.4 |
| 2 | 4.51 E5 ± 3.1% | 6.44 E5 ± 6.8% | 1.4 |
| 4 | 4.62 E5 ± 0.7% | 6.37 E5 ± 1.0% | 1.4 |
4. Concluding remarks
The present work suggests electrofiltration as a novel downstream processing method in the production of HA. In the state of the art of HA production processes, filtration is not possible, as presented in the experiments without electrical field. Increasing HA concentration in technical scale is only possible by precipitation with organic solvents. The inherent disadvantage of using organic solvents can be overcome by electrofiltration making HA production potentially more sustainable. The results confirmed that the application of electric field in the filtration technology of HA dispersions reduces processing time and increases concentration factors compared with no electric field applied.
It has been demonstrated that the average molecular weight and structural composition of HA remained the same after introduction of electric field during processing. In addition, the influence of different buffer solutions upon filtration kinetics was explored. The obtained results established electrofiltration as an alternative technology to common downstream processes. Depending on production procedure, salt content of fermentation broth varies, and consequently influences filtration behavior. As a result of increased ionic strength, electric current increases leading to an increase in energy consumption. In order to reduce overall energy requirement, integration of an in situ desalination system into the electrofiltration set up can be a valuable contribution. For the optimization of the purification process and its successful implementation in industrial‐scale technologies, further experiments including separation of HA after biotechnological production are required.
Practical application
Hyaluronic acid is a high molecular weight polysaccharide present in the extracellular matrix of mammalian connective tissues as well as in some bacteria. Separation of hyaluronic acid is usually performed by application of organic solvents, which leads to high costs and change in product quality. Owing to surface charge of hyaluronic acid caused by carboxylic groups, electrofiltration can be applied as an alternative technology. Instead of conventional methods in the downstream processing of the biopolymer, application of electrofiltration results in a significant increase in concentration and the lack of unwanted structural changes in HA molecules. Therefore, electrofiltration opens new possibilities for HA production meeting the current downstream processing trends.
The authors have declared no conflict of interest.
Acknowledgement
The present work was prepared in the course of a prior study at Karlsruhe Institute of Technology in Karlsruhe, Germany. The authors express their gratitude to BMBF (Federal Ministry of Education and Research) project number 31P5568 for financial support and to Thomas Hivert for the illustration of electrofiltration cell.
5 References
- 1. Armstrong, D. C. , Johns, M. R. , Culture conditions affect the molecular weight properties of hyaluronic acid produced by Streptococcus zooepidemicus . Appl. Environ. Microbiol. 1997, 63, 2759–2764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Chong, B. F. , Blank, L. M. , Mclaughlin, R. , Nielsen, L. K. , Microbial hyaluronic acid production. Appl. Microbiol. Biotechnol. 2005, 66, 341–351. [DOI] [PubMed] [Google Scholar]
- 3. Rangaswamy, V. , Jain, D. , An efficient process for production and purification of hyaluronic acid from Streptococcus equi subsp. zooepidemicus . Biotechnol. Lett. 2008, 30, 493–496. [DOI] [PubMed] [Google Scholar]
- 4. Lapcik, L. , Lapcik, L. , De Smedt, S. , Demeester, J. et al., Hyaluronan: Preparation, structure, properties, and applications. Chem. Rev. 1998, 98, 2663–2684. [DOI] [PubMed] [Google Scholar]
- 5. Akdamar, H. A. , Yılmaz Sarıözlü, N. , Atılır Özcan, A. , Ersöz A. et al., Separation and purification of hyaluronic acid by glucuronic acid imprinted microbeads. Mat. Sci. Eng. C 2009, 29, 1404–1408. [Google Scholar]
- 6. Sze, J. H. , Brownlie, J. C. , Love, C. A. , Biotecnological production of hyaluronic acid: A mini review. 3 Biotech 2016, 6, 67–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Pires A. M. B., Macedo, A. C. , Eguchi, S. Y. , Santana, M. H. A. , Microbial production of hyaluronic acid from agricultural resource derivatives. Bioresour. Technol. 2010, 101, 6506–6509. [DOI] [PubMed] [Google Scholar]
- 8. Kim, C. W. , Kim, S. J. , Park, S. Y. , A novel approach to the production of hyaluronic acid by Streptococcus zooepidemicus . J. Microbiol. Biotechnol. 2006, 16, 1849–1855. [Google Scholar]
- 9. Chien, L. J. , Lee, C. K. , Enhanced hyaluronic acid production in Bacillus subtilis by coexpressing bacterial hemoglobin. Biotechnol. Prog. 2007, 23, 1017–1022. [DOI] [PubMed] [Google Scholar]
- 10. Chen, S. J. , Chen, J. L. , Huang, W. C. , Chen, H. L. , Fermentation process development for hyaluronic acid production by Streptococcus zooepidemicus ATCC 39920. Korean J. Chem. Eng. 2009, 26, 428–432. [Google Scholar]
- 11. O`regan, M. , Martini, I. , Crescenzi, F. , Deluca C. et al., Molecular mechanisms and genetics of hyaluronan biosynthesis. Int. J. Biol. Macromol. 1994, 16, 283–286. [DOI] [PubMed] [Google Scholar]
- 12. Zhou, H. D. , Ni, J. R. , Huang, W. , Zhang, J. D. , Separation of hyaluronic acid from fermentation broth by tangential flow microfiltration and ultrafiltration. Sep. Purif. Technol. 2006, 52, 29–38. [Google Scholar]
- 13. Murado, M. A. , Montemayor M. I., Cabo M. L., Vazquez, J. A. et al., Optimization of extraction and purification process of hyaluronic acid from fish eyeball. Food Bioprod. Process. 2012, 90, 491–498. [Google Scholar]
- 14. Sousa, A. S. , Guimaraes, A. P. , Goncalves, C. V. , Silva, I. J. et al., Purification and characterization of microbial hyaluronic acid by solvent precipitation and size‐exclusion chromatography. Sep. Sci. Technol. 2009, 44, 906–923. [Google Scholar]
- 15. Chen, J. C. , Li, Q. L. , Elimelech, M. , In situ monitoring techniques for concentration polarization and fouling phenomena in membrane filtration. Adv. Colloid Interface Sci. 2004, 107, 83–108. [DOI] [PubMed] [Google Scholar]
- 16. Posten, C. , Principles of Mechanical Bioseparation, Shaker Verlag, Aachen: 2007. [Google Scholar]
- 17. Käppler T., Posten C., Fractionation of proteins with two‐sided electro‐ultrafiltration. J. Biotechnol. 2007, 128, 895–907. [DOI] [PubMed] [Google Scholar]
- 18. Hofmann, R. , Posten, C. , Improvement of dead‐end filtration of biopolymers with pressure electrofiltration. Chem. Eng. Sci. 2003, 58, 3847–3858. [Google Scholar]
- 19. Liu, X. , Sun, C. , Zang, H. , Wang, W. et al., Capillary electrophoresis for simultaneous analysis of heparin, chondroitin sulfate and hyaluronic acid and its application in preperations and synovial fluid. J. Chromatogr. Sci. 2012, 50, 373–379. [DOI] [PubMed] [Google Scholar]
- 20. Gözke, G. , Prechtl, C. , Kirschhöfer, F. , Mothes, G. et al., Electrofiltration as a purification strategy for microbial poly‐(3‐hydroxybutyrate). Bioresour. Technol. 2012, 123, 272–278. [DOI] [PubMed] [Google Scholar]
- 21. Gözke, G. , Kirschhöfer, F. , Heissler, S. , Trutnau, M. et al., Filtration kinetics of chitosan separation by electrofiltration. Biotechnol. J. 2011, 7, 262–274. [DOI] [PubMed] [Google Scholar]
- 22. Alkrad, J. A. , Mrestani, Y. , Stroehl, D. , Wartewig, S. et al., Characterization of enzymatically digested hyaluronic acid using NMR, Raman, IR, and UV‐Vis spectroscopies. J. Pharm. Biomed. Anal. 2003, 31, 545–550. [DOI] [PubMed] [Google Scholar]