- Email: [email protected]
Extractive bioconversions in aqueous two-phase systems
171
Extractive bioconversions in aqueous two-phase systems Gerben M Zijlstra*, Cornelis D de Gooijert and Johannes Trampert Although extractive bioconversions in aqueous two-phase systems (ATPSs) have been studied for over a decade, this has not yet resulted in widespread industrial application. The main reasons are the cost of the phase-forming polymers and the complexity of ATPS behavior. A number of recent developments may give a new impetus to this technology. First of all, the use of extractive bioconversions in ATPSs has recently been extended to high-value protein products, while in the meantime the development of low-cost ATPSs is ongoing. Furthermore, novel chromatographic methods enable the analysis of polymer and metabolite compositions in complex ATPS mixtures, and recently employed statistical experimental designs provide a tool for efficient data gathering, while they also reveal synergistic effects between process parameters. Together, these developments open the way towards improved modeling of partitioning behavior in ATPSs.
Addresses *Gist-brocades/Bio-lntermediair BV, PO Box 454, 9700 AL Groningen, The Netherlands; e-mail: [email protected] tWageningen Agricultural University, Department of Food Science, Food and Bioprocess Engineering Group, PO Box 8129, 6700 EV Wageningen, The Netherlands Correspondence: Gerben M Zijlstra Current Opinion in Biotechnology 1998, 9:171 - 176 http://biornednet.corn/elecref/0958166900900171
© Current Biology Ltd ISSN 0958-1669 Abbreviations ATPSs aqueoustwo-phase systems PEG polyethyleneglycol
and productivity of existing processes. T h e y are obligatory, however, when highly toxic or highly unstable products are made [1,7,8]. An additional advantage of extractive bioconversions is that they have the possibility to be scaled up to industrial size application [3,9]. Yield improvements may be obtained by reduction of product inhibition, through immediate product removal, or by prevention of product degradation, through minimization of the product residence time in the vicinity of the biocatalyst. Furthermore, in situ product recovery techniques offer the opportunity for continuous processing with biocatalyst retention, which may greatly improve the overall volumetric productivity. Finally, through integrating part of the downstream processing with biocatalysis, the number of subsequent downstream processing steps may be reduced [1]. The major disadvantages that have hampered widespread industrial application of extractive bioconversions with ATPSs are the costs of the polymers [3,9,10,11°,12"], and the difficulties in predicting the biocatalyst and product partitioning, as well as the biocatalyst activity in ATPSs, due to the lack of adequate mathematical models [1,3,13,14]. Key issues in designing technically and economically feasible processes are the partitioning of biocatalysts and their products in ATPSs, the activity and stability of biocatalysts in ATPSs, the mode of operation, equipment design, process control, downstream processing and, what is most important, process economics. In this paper, recent developments that may improve the technical and economical feasibility of extractive bioconversions with ATPSs are reviewed.
Introduction
Extractive bioconversion is one of the techniques for in situ product recovery, the separation of products from their biocatalyst immediately after they are formed [1]. In extractive bioconversions two immiscible liquid phases are employed, the objective being to run a bioconversion in one liquid phase, while the product is extracted into the other phase. Aqueous two-phase systems are formed by mixing two aqueous polymer solutions (or a polymer and a salt solution). Above certain concentrations two immiscible aqueous phases will form, each enriched in one of the polymers [2]. Aqueous two-phase systems (ATPSs) are especially useful in extractive bioconversions, because they constitute a relatively mild environment for biocatalysts [2,3] and they are suitable for extraction of hydrophilic products, such as proteins [4-6]. The benefits of in situ product recovery techniques are to be found mainly in their potential for improving yield
Partitioning
The first objective in designing ATPSs for extractive bioconversions is to separate the product and its biocatalyst. For efficient processing the biocatalyst has to partition completcly to one of the liquid phases, while the product concentrates in the other phase. In this manner the biocatalyst can be retained and reused by recycling the biocatalyst-containing phase, the product concentration in the vicinity of the biocatalyst is low, and a biocatalyst-free product phase is available for further downstream processing. Cells, like all large particles, mainly partition between one of the liquid phases and the interface. The cell concentration in the other phase is typically several orders of magnitude less, so in essence it is free of cells [2,3,11°,15-17,18°°]. During continuous processing some leakage of cells into the extracting phase is less
172
Biochemicalengineering
of a problem, because it is compensated for by cell growth; however, prior to further downstream processing an additional cell separation step will be required. Enzymes, like all macromolecular sized particles, are distributed far more evenly between the two liquid phases than cells. During continuous processing, therefore, enzymes will be extracted rapidly with the product extracting phase and enzyme retention in the bioreactor will be poor [2,3]. In contrast to cells, this leakage is not compensated for and therefore may be problematic. To achieve a one-sided partitioning of enzymes, whole-cell enzyme preparations [19], immobilized enzymes [20] and enzymes chemically coupled to polyethylene glycol (PEG) [21] have been used. Typical products of extractive bioconversions in ATPSs so far have been, on one hand, low molecular weight inhibitory products, such as lactic acid [11°,12°,22,23°°], sugars from particulate substrates, such as starch [24-27], cellulose [28] or chitin [17,29], and easily degradable products, such as antibiotics [30,31]. On the other hand, high-molecular weight products, such as extracellular enzymes [16,32,33] or even high-value (recombinant) proteins [18°°,34,35°°,36°°], have been produced in ATPSs. Low molecular weight products generally partition evenly between the phases, high molecular weight products tend to concentrate in one of the liquid phases. Despite numerous theoretical models [37], to date no practical models exist to predict the partitioning of biocatalysts and their products in a given ATPS, based on their physical and chemical properties. Furthermore, the partitioning of biocatalysts, products, and also polymers may change as bioconversions progress, because biomass and metabolites can alter the composition of the aqueous phases [16,23°°]. Two recently employed approaches [18°°,23°°,38,39°,40 ° ] may greatly assist in generating dynamic models that describe the partitioning in ATPS media for extractive bioconversions more accurately, also taking biomass and metabolite effects into account. In the first place, the use of analytical tools to directly determine the composition of each phase in complex mixtures [23 °°] constitutes a significant improvement compared to the traditional situation where phase diagrams in water were used to simulate those in complex mixtures. Secondly, the use of statistical experimental design [18°°,38,39°,40 °] may provide empirical partitioning models in which more relevant factors, such as metabolite and biomass concentrations, can be included, while the required number of experiments to generate data for these models is reduced to a minimum. Because selecting ATPSs with the proper biocatalyst partitioning is the starting point in ATPS design, the opportunities for maximizing product partitioning to the opposite phase are sometimes limited. Essentially, two
approaches can be used for improving product partitioning in a given system. One is to add a specific extractant to the phase system, such as an affinity ligand coupled to a phase-forming polymer [40°,41-43]; however, this may affect the partitioning of other components, for instance cells, as well [40°,44]. Furthermore, these additives may be expensive. T h e other approach is to alter the product in such a way that its partitioning improves. With recombinant proteins tagging is a common procedure to improve further downstream processing. This procedure has also been used in ATPSs [45,46]. Here the consequences with respect to protein folding and removal of the tag further downstream in the process have to be considered.
Activity and stability of biocatalysts T h e second major objective in ATPS design is to create an environment that allows biocatalysis, that is, cell growth and production of enzyme activity, at similar specific activities and stabilities compared to culture media or buffers without polymers. Extractive bioconversions with enzymes have been executed so far in polymer-polymer ATPSs. T h e specific enzyme activities in ATPSs are generally not affected by the phase-forming polymers and are similar to those in reference media [3,19,21,24,25]. Reports on the effects of ATPSs on enzyme stability vary from a slight increase [24,25,28] to a slight decrease [29]. Cell growth has been reported in ATPSs with neutral polymers, with charged polymers and even with polymers and salts. Bacterial and fungal cell growth appears to be hardly affected in ATPSs with moderate concentrations of neutral polymers [3,12°,16,17,32,33,47], unless oxygen transfer due to the increased viscosity becomes rate limiting; however, changes in morphology, such as cell aggregation and also changes in cell-surface hydrophobicity, have been reported [3]. Charged polymers can offer the advantage of improved product partitioning as well as ease of polymer recovery. Recently, the cultivation of lactic acid producing bacteria in ATPSs containing a charged polymer was described [11",22]. T h e inhibitory effect of the charged polymer was largely overcome by adaptation of the cells to the ATPS culture medium, although cell aggregation was still observed. T h e production rates were similar to those of the control cultures. A major breakthrough is the recently reported cultivation of several bacterial species in polymer-salt ATPSs [10,34]. Polymer-sah ATPSs can be cost effective for protein extractions [6,10,48]. Furthermore, their suitability for large-scale extractions of products with a wide range of hydrophobicities has long been recognized [6,48]. Cell growth was found to be largely unaffected by PEG when the molecular weight was 4000kD or higher;
Extractive bioconversionsin aqueous two-phase systems Zijlstra,de Gooijerand Tramper
however, it was inhibited exponentially with increasing salt concentrations [10]. Another major breakthrough is the recently reported cultivation of animal cell lines excreting high value proteins, such as monoclonal antibodies, in ATPSs [36°°]. In ATPSs of PEG and dextran, hybridoma cell growth was shown over a period of months to be similar to growth rates of control cultures. Finally, successful plant cell cultivation in ATPSs has been reported in ATPSs with neutral polymers [49,50]; however, product excretion, required for successful in situ product removal, remains an issue to be improved. To date, no models are available to describe the effect of the phase-forming constituents on cell growth and production. Again, like the partitioning behavior, empirical correlations need to be developed in order to produce guidelines for designing better ATPSs. M o d e of o p e r a t i o n T h e mode of operation of extractive bioconversions can vary from simple batch processes to full continuous processes with biocatalyst recycling or even polymer recycling. Batch extractive bioconversions have been reported most frequently in recent literature [10,11°,17,18°',20,22,26,27, 29,36°°,51], their main advantage being simplicity. Once an ATPS has been designed that provides adequate separation between the biocatalyst and its product, and supports adequate activity and stability of the biocatalyst in small-scale experiments, the process can easily be scaled up in conventional stirred tanks. T h e product can be harvested by removing the product-containing phase by settling or centrifugation. T h e time required for batch process development, therefore, is minimal. Disadvantages in the case of low molecular weight products, or other products that distribute evenly between the phases, are that the product-biocatalyst interaction is substantial and the final product yield in the extracting phase is low. Furthermore, since the residence time of the product in the reaction mixture is equal to the batch cycle time, the yield of unstable or easily degradable products may be compromised. Finally, especially for bioconversions with cells, the volumetric productivity of batch processes is low compared to fed-batch processes or continuous processes with cell retention. Continuous processing with biocatalyst and polymer recycling may compensate for most of the disadvantages of batch processes. This highly integrated approach, however, inherently leads to a higher level of complexity and it poses stringent requirements on both equipment design and process control. Many reports so far, therefore, only describe repeated extractive batch processes, where the product extracting-phase is periodically replaced with fresh extracting phase [12°,23"°,25,34,35"°]. Also,
173
continuous processes in which the extracting phase is continuously replaced and the biocatalyst-containing phase is recycled by means of an in-line settler are described [16,19,28,47]. In-line recycling of polymers from the product-containing phase, after removal of the product, has been reported rarely, because of the difficulties to achieve this on a small scale [24]. Equipment design There are important issues to take into account during the equipment design for extractive bioconversions: firstly, the mass transfer of metabolites and product within each phase; secondly, the mass transfer of metabolites and product between the phases; thirdly, the phase separation; fourthly, the separation of the product from the extracting phase; and finally, the recycling of polymers. Stirred tanks, operated at moderate stirring speeds, generally have been found to provide adequate mass transfer for extractive bioconversions in ATPSs. Because of the low interfacial tensions in ATPSs, emulsions with high interfacial areas, and thus with high mass transfer rates, will already occur at low power input [3,9,52,53]. To separate the ATPS emulsions, however, long settling times are required. In batch extractions, phase separation can be obtained by settling in the stirred tank itself. For continuous extractive bioconversions with biocatalyst recycling, however, dedicated settlers are required. Several settler designs have been described in the literature, although none of them seem to be very effective at high flow rates [16,19,24,47]. Furthermore, the residence time in these settlers may cause oxygen depletion when growing cells are used [16,47], and when cells partition to the interface they can accumulate in the settler and compromise the effectiveness of the phase separation [47]. Complete cell partitioning to one of the liquid phases, therefore, is of paramount importance. Although for large-scale ATPS extractions the use of industrial continuous (disk-stack) centrifuges has been reported [6], and meanwhile specific continuous centrifuges for cell recycling have become available on the market, so far their use in extractive bioconversions with ATPSs has not been reported. To avoid phase-separation difficulties, and moreover to obtain multistage extraction of the product, the opportunities for counter-current extraction using spray and other types of column extractors have been investigated [52]. For polymer-polymer ATPSs, however, the mass-transfer coefficients were extremely low. Furthermore, to avoid flooding, the residence time in these extractors has to be long, which practically excludes their use for extractive bioconversions with cells. Process control In batch processes only some basic process parameters in the stirred tank need to be controlled. Typically the
1 74
Biochemical engineering
temperature, pH, stirrer speed, and in case of cells the dissolved oxygen concentration are important. In continuous processes, beside the process parameters of the stirred tank the parameters of the phase separator, as well as the parameters of incoming and outgoing fluxes, also need to be controlled. For effective continuous extractive bioconversions the aim is to obtain and maintain a steady-state situation, under optimal process conditions, usually after an initial dynamic phase where, for instance, the biomass increases. Recent studies have shown, however, that shifts in the polymer concentrations of each phase can occur during the dynamic phase of extractive bioconversions. These shifts can result in changes in cell, metabolite and product partitioning and, ultimately, in system breakdown [23"°,47]. Controlling the ATPS composition, therefore, is vital in continuous extractive bioconversions. Novel on-line analytical methods may be used for this purpose [23°°]. Furthermore, process models consisting of mass balances of all phase-system constituents, models describing the partitioning, as well as models describing the bioconversion in ATPSs, seem to be necessary for process control, especially during the dynamic phase [54,55]. Downstream
processing
Depending on the product type and the required purity, one or a combination of conventional downstream processing techniques, such as extraction, evaporation, filtration, ultrafiltration, chromatography and so on, can be used for further purification. T h e first purification step generally includes separation of the polymers from the product. Recently, new polymer types have been employed that form a precipitate [56] or an immiscible liquid phase highly concentrated in the polymer [11°,12",57,58] after a shift in pH or temperature. T h e polymers can be recovered or even recycled after phase separation, while the polymer-lean product phase is available for further downstream processing. Process
explored [10,34] and polymer recycling schemes are being developed [ 11 °, 12",24,57,58]; and finally, the opportunities for using extractive bioconversions in high-value-product manufacturing are being explored [18°°,35°',36°',39°,40"]. Also, scale appears to be crucial for the economically effective implementation of extractions [13]. Once the knowledge of extractive bioconversions has increased to a level where rational process design can take place and in the meantime low-cost ATPSs have become available, extractive bioconversions with ATPSs become an important tool for the large-scale manufacturing of biologicals. Conclusions
Extractive bioconvcrsions with ATPSs offer a number of unique features which make them exceptionally useful for large-scale manufacturing of biologicals. T h e current knowledge of this technology, however, is not sufficient for rational process design. Furthermore, the costs associated with implementing extractive bioconversions are still too high. T h e continuing developments in the field of ATPSs, in combination with important recent findings, may greatly improve the practical applicability of extractive bioconversions with ATPSs in the near future. Acknowledgements
This workwas supported by Gist-brocades/Bio-lntermediairBV,Groningen, The Netherlands and by a grant under the Dutch PBTS-Biotechnology program, Ministry of EconomicAffairsNr. BIO89041. References and recommended reading Papers of particular interest, published within the annual period of review, have been highlighted as: * ..
Freeman A, Woodley JM, Lilly MD: In situ product removal as a tool for bioprocessing. Bio/Techno/ogy 1993, 11:1007-1012. AIbertsson PA: Partitioning of Cel/ Partic/es and Macromo/ecu/es, edn 3. New York: John Wiley & Sons; 1986. Andersson E, Hahn-H~igerdal B: Bioconversions in aqueous two-phase systems. Enzyme Microb Techno11990, 12:242-253.
economics
Obviously, for the breakthrough of extractive bioconversions with ATPSs it is essential that the costs of implementation are favorable compared to the revenues of the product, and that the extractive bioconversion is more cost effective than conventional processing. Previous economical evaluations have shown that the polymer costs largely determine the process economics. In particular, fractionated dextran is too expensive for industrial application in extractive bioconversions [3,9,53,59]. To tackle this problem, several approaches have been employed; firstly, low-cost alternatives to dextran are being developed, mainly based on starch derivatives [11°,60,61]; secondly, the possibilities for using low cost PEG-salt systems for extractive bioconversions are being
of special interest of outstanding interest
Kroner K, Kula MR: Extraction of enzymes in aqueous twophase systems. Process Biochem 1978, 13:7-10. Kula MR, Kroner K, Hustedt H: Purification of enzymes by liquidliquid extraction. Adv Biotechno/Eng 1982, 24:73-118.
7. 8. 9. lO.
Hustedt H, Kronor KH, Papamichael N: Continuous cross-current aqueous two-phase extraction of enzymes from biomass: automated recovery in production scale. Process Biochem 1988, 23:129-137. Daugulis AJ: Integrated reaction and product recovery in bioreactor systems. Biotechno/Prog 1988, 4:113-122. Daugulis AJ: Integrated fermentation and recovery processes. Curr Opin Biotechno/1991, 5 :192-195. Kaul R, Mattiasson B: Extractive bioconversions in aqueous two-phase systems. Bioprocess Techno/t 991, 11:173-188. Kuboi R, Maruki T, Tanaka H, Komasawa I: Fermentation of Bacillus subtilis ATCC 6633 and production of subtilin in polyethylene glycol/phosphate aqueous two-phase systems. J Ferm Bioeng 1994, 78:431-436.
Extractive bioconversions in aqueous two-phase systems Zijlstra, de Gooijer and Tramper
11.
KwonYJ, Kaul R, Mattiasson B: Extractive lactic acid fermentation in poly(ethyleneimine)-based aqueous two-phase system. Biotechno/Bioeng 1996, 50:280-290. This study shows the usefulness of charged polymers in combination with low-cost polymers in extractive fermentations with bacteria. Charged polymers can be separated from the product extracting phase by phase separation using multivalent ions.
29.
Chen JP, Lee MS: Enhanced production of Serratia marcescens chitinase in PEG/dextran aqueous two-phase systems. Enz Microb Techno/1995, 17:1021-1027.
30.
Chang HN, Lee YH, Lee CY: Continuous production of 6 - A P A in an aqueous two-phase system. Ann NY Acad Sci 1992, 672:643-648.
31.
Paquet V, Myint M, Roque C, Soucialle P: Partitioning of pristinamycins in aqueous two-phase systems: a first step toward the development of antibiotic production by extractive fermentation. Biotechno/Bioeng 1994, 44:445-451.
32.
PerssonI, Staalbrand H, Tjerneld F, Hahn-h~gerdal B: Semicontinuous production of cellulolytic enzymes with Trichoderma reesei Rutgers C30 in an aqueous two-phase system. App/ Biochern Biotechno/1992, 27:27-36.
33.
PerssonI, Tjerneld F, Hahn-haegerdal B: Influence of cultivation conditions on the production of cellulolytic enzymes with Trichoderma reesei Rutgers C30 in aqueous two-phase systems. App/ Biochem Biotechno/1992, 27:9-25.
34.
Kuboi R, Umakoshi H, Komasawa h Extractive cultivation of Escherichia coil using poly(ethylene glycol)/phosphate aqueous two-phase systems to produce intracellular ]3galactosidase. Biotechno/Prog 1995, 11:202-207.
•
12. •
PlanasJ, Radstrom P, Tjerneld F, Hahn-Hagerdal B: Enhanced production of lactic acid through the use of a novel aqueous two-phase system as an extractive fermentation system. App/ Microbiol Biotechnol 1996, 45:737-743. This study shows the usefulness of EO-PO polymers in extractive fermentations with bacteria. The EO-PO polymers can be separated from the product extracting phase by a temperature shift. 13.
Asenjo JA: Industrial prospects of aqueous two-phase processes. J Chem Technol Biotechnol 1994, 59:109.
14.
MistrySL, Kaul A, Merchuk JC, Asenjo JA: Mathematical modeling and computer simulation of aqueous two-phase continuous protein extraction J Chromatogr 1996, 741:151-163.
15.
Walter H: Separation and subfractionation of blood cell populations based on their surface properties by partitioning in two-polymer aqueous phase systems. In Ceil Separation: Methods and Se/ected Appfications. Edited by Pretlow TG II, Pretlow ThP II. New York: Academic Press; 1982, 1:261-299.
16.
Park KM, Wang NS: c~-amylase fermentation with Bacillus amylofiquefaciens in an aqueous two-phase system. Biotechno/ Prog 1991,7:439-444.
17.
Chen JP, Lee MS: Simultaneous production and partition of chitinase during growth of Serratia marcescens in an aqueous two-phase system. Biotechnol Tech 1994, 8:?83-?88.
18. •o
Zijlstra GM, Michielsen M, De Gooijer CD, Van der Pol LA, TramperJ: Hybridoma and CHO cell partitioning in aqueous two-phase systems. Biotechnol Prog 1996, 12:363-370. This study uses a statistical experimental design approach to obtain an empirical model for the partitioning of cells. These models are of paramount importance for improved process control of extractive fermentations.
1 ?5
35. o.
UmakoshiH, Yano K, Kuboi R, Komasawa h Extractive cultivation of recombinant Escherichia coil using aqueous two-phase systems for production and separation of intracellular heat shock proteins. Biotechno/Prog 1996, 12:51-56. This study demonstrates the feasibility of producing high-value recombinant proteins from E. coil by means of extractive fermentation in ATPSs. A semicontinuous production scheme is proposed, to allow periodical cell disruption and product release. 36.
Zijlstra GM, de Gooijer CD, Van der Pol LA, Tramper J: Design of aqueous two-phase systems supporting animal cell growth: a first step towards extractive bioconversions. Enz Microb Techno/1996, 19:2-8, This study demonstrates that the principle of extractive fermentation in ATPSs can also be applied to animal cells. The benefits of extractive fermentations, continuous operation with scalable cell retention, in situ product removal and integration of downstream processing with fermentation, now become available for animal cell cultivation. o.
19.
TomaskaM, Stredansky M, Tomaskova A, Sturdik E: Lactose hydrolysis in aqueous two-phase system by whole-cell betagalactosidase of Kluyveromyces marxianus. Bioproc Eng 1995, 12:17-20.
20.
Kondo A, Urabe T, Higashitani K: Bioconversions in an aqueous two-phase system using enzymes immobilized on ultrafine silica particles. J Ferrn Bioeng 1994, 6:700-703.
37.
Abbott N, Blankschtein D, Hatton A: On protein partitioning in two-phase aqueous polymer systems. Bioseparation 1990, 1:191-225.
21.
MukatakaS, Haynes CA, Prausnitz JM, Blanch HW: Extractive bioconversions in aqueous two-phase systems: enzymatic hydrolysis of casein proteins. Biotechno/Bioeng 1992, 40:195206.
38.
Hart RA, Ogez JR, Builder SE: Use of multifactorial analysis to develop aqueous two-phase systems for isolation of nonnative IgF1. Bioseparation 1995, 5:113-121.
22.
Dissing U, Mattiasson B: Cultivation of Lactococcus lactis in a polyelectrolyte-neutral polymer aqueous two-phase system. Biotechno/Lett 1994, 16:333-338.
23.
PlanasJ, Lefebvre D, Tjerneld F, Hahn-H&gerdahl B: Analysis of phase composition in aqueous two-phase systems using a two-column chromatographic method: application to lactic acid production by extractive fermentation. Biotechno/Bioeng 1997, 54:303-311. This study unambiguously shows that drifts in phase compositions occur during extractive fermentations. In order to maintain a steady state during continuous processing, sophisticated process control is necessary. The on line measurement method described is an important tool in process control. e•
39.
Zijlstra GM, Michielsen MJF, de Gooijer CD, Van der Pol LA, Tramper J: Separation of hybridoma cells from their I g G product using aqueous two-phase systems. Bioseparation 1996, 6:201-210. This study uses a statistical experimental design to generate data for the modeling of product partitioning. •
40. •
Zijlstra GM, Michielsen MJF, de Gooijer CD, Tramper J: Selection criteria for effective ligands in extractive fermentations with animal cells. Bioseparations 1998, in press. This study characterizes the effects of ligands coupled to PEG on both product and cell partitioning. 41.
Koppersl&ger G, Birkenmeier G: Affinity partitioning and extraction of proteins. Bioseparation 1990, 1:235-254.
42.
BirkenmeierG, Vijayalakshmi MA, Stigbrand T, Kopperschl&ger G: Immobilized metal ion affinity partitioning, a method combining metal-protein interaction and partitioning of proteins in aqueous two-phase systems. J Chromatogr 1991, 539:26?-277.
24.
LarssonM, Arasaratnam V, Mattiasson B: Integration of bioconversion and downstream processing: starch hydrolysis in an aqueous two-phase system. Biotechnol Bioeng 1989, 33:758-766.
25.
Tjerneld F, Persson I, Lee JM: Enzymatic cellulose hydrolysis in an attrition bioreactor combined with an aqueous two-phase system. Biotechnol Bioeng 1991, 37:876-882.
43.
26.
Liakopoulou-Kyriakides M, Karakatsanis A, Stamatoudis M: Enzymic hydrolysis of starch in agitated PEG-dextran aqueous two-phase system. Starch 1996, 48:291-294.
Garg N, Galaev IY, Mattiasson B: Dye-affinity techniques for bioprocessing: recent developments. J Mo/Recognit 1996, 9:259-274.
44.
2"7.
Min KH, Chang WJ, Koo YM: Increased production of betacyclodextrin using an aqueous two-phase system. Biotechno/ Tech 1996, 10:395-400.
Hamamoto R, Kamihira M, lijima S: Specific separation of animal cells using aqueous two-phase systems. J Ferm Bioeng 1996, 82:73-76.
45.
Taguchi F, Yamada K, Hasegawa K, Taki-Saito T, Hara K: Continuous hydrogen production by Clostridium sp. strain no. 2 from cellulose hydrolysate in an aqueous two-phase system. J Ferm Bioeng 1996, 82:80-83.
K6hler K, Ljungquist C, Kondo A, Veide A, Nilsson B: Engineering proteins to enhance their partition coefficients in aqueous two-phase systems. Bio/Techno/ogy 1991, 9:642-646.
46.
Carlsson M, Berggren K, Linse P, Veide A, Tjerneld F: Effects of fused tryptophan-rich peptides to a recombinant protein A domain on the partitioning in polyethylene glycol/dextran
28.
176
Biochemical engineering
aqueous two-phase systems. J Chromatogr A 1996, 756:107117 4?.
KatzbauerB, Cesi V, Narodoslawsky M, Moser A: Extractive lactic acid fermentation using aqueous two-phase systems. Chem Biochem Eng 19951 9:79-87.
48.
Kula MR: Trends and future prospects of aqueous two-phase extraction. Bioseparation 1990, 1:181-189.
49.
Buitelaar RM, Leenen EJTM, Tramper J: Growth and secondary metabolite production by hairy roots of tagetes patula in aqueous two-phase systems. Biocatalysis 1992, 6:73-80.
50.
IlievaMP, Bakalova A, Mihneva M, Pavlov A, Dolapchiev L: Production of phosphomonoesterases by Nicotiana tabacum 1507 in an aqueous two-phase system. Biotechnol Bioeng 1996, 51:488-493.
51.
52.
53.
54.
RamadasM, Hoist O, Mattiasson B: Production of amyloglucosidase by Aspergillus niger under different cultivation regimens. World J Microbiol Biotechno11996, 12:267-271. JoshiJB, Sawant SB, Rao KSMSR, Patil TA, Rostami KM, Sikdar SK: Continuous counter-current tw0-phase aqueous extraction. Bioseparation 1990, 1:311-324. Huddleston JG, Lyddiatt A: Aqueous two-phase systems in biochemical recovery. Appl Biochem Biotechno11990, 26:249280. JarzebskiAB, Malinowski JJ: Analysis of continuous fermentation processes in aqueous two-phase systems. Bioprocess Eng 1992, 7:315-31"7.
55.
BaughmanDR, Liu YA: An expert network for predictive modeling and optimal design of extractive bioseparations in aqueous two-phase systems. Ind Eng Chem Res 1994, 33:2668-2687.
56.
KamiharaM, Kaul R, Mattiasson B: Purification of recombinant protein A by aqueous two-phase extraction integrated with affinity precipitation. Biotechno/Bioeng 1992, 40:1381-1387.
67
FarkasT, Stalbrand H, Tjerneld F: Partitioning of betamannanase and alpha-galactosidase from Aspergillus Niger in Ucon/Reppal aqueous two-phase systems and using temperature induced phase separation. Bioseparation 1996, 6:147-157.
58.
Lu M, Albertsson PA, Johansson G, Tjerneld F: Ucon-benzoyl dextran aqueous two-phase systems: protein purification with phase component recycling. J Chromatogr B Biomed App11996, 680:65-?0.
59.
Datar R: Economics of primary separation steps in relation to fermentation and genetic engineering. Proc Biochem 1986, Feb:l 9-26.
60.
TjerneldF, Berner S, Cajarville A, Johansson G: New aqueous two-phase systems based on hydroxypropyl starch useful in enzyme extraction. Enz Microbiol Technol 1986, 8:41 ?-423.
61.
Sz[agDC, Guiliano KA, Snyder SM: A low cost aqueous two-phase system for affinity extraction. In Bioseparations. Downstream Processing in Biotechnology. Edited by Hamel JF, Hunter JB, Sikdar SK. Washington DC: ACS Symposium series; 1990, 419:71-86.
Extractive bioconversions in aqueous two-phase systems Gerben M Zijlstra*, Cornelis D de Gooijert and Johannes Trampert Although extractive bioconversions in aqueous two-phase systems (ATPSs) have been studied for over a decade, this has not yet resulted in widespread industrial application. The main reasons are the cost of the phase-forming polymers and the complexity of ATPS behavior. A number of recent developments may give a new impetus to this technology. First of all, the use of extractive bioconversions in ATPSs has recently been extended to high-value protein products, while in the meantime the development of low-cost ATPSs is ongoing. Furthermore, novel chromatographic methods enable the analysis of polymer and metabolite compositions in complex ATPS mixtures, and recently employed statistical experimental designs provide a tool for efficient data gathering, while they also reveal synergistic effects between process parameters. Together, these developments open the way towards improved modeling of partitioning behavior in ATPSs.
Addresses *Gist-brocades/Bio-lntermediair BV, PO Box 454, 9700 AL Groningen, The Netherlands; e-mail: [email protected] tWageningen Agricultural University, Department of Food Science, Food and Bioprocess Engineering Group, PO Box 8129, 6700 EV Wageningen, The Netherlands Correspondence: Gerben M Zijlstra Current Opinion in Biotechnology 1998, 9:171 - 176 http://biornednet.corn/elecref/0958166900900171
© Current Biology Ltd ISSN 0958-1669 Abbreviations ATPSs aqueoustwo-phase systems PEG polyethyleneglycol
and productivity of existing processes. T h e y are obligatory, however, when highly toxic or highly unstable products are made [1,7,8]. An additional advantage of extractive bioconversions is that they have the possibility to be scaled up to industrial size application [3,9]. Yield improvements may be obtained by reduction of product inhibition, through immediate product removal, or by prevention of product degradation, through minimization of the product residence time in the vicinity of the biocatalyst. Furthermore, in situ product recovery techniques offer the opportunity for continuous processing with biocatalyst retention, which may greatly improve the overall volumetric productivity. Finally, through integrating part of the downstream processing with biocatalysis, the number of subsequent downstream processing steps may be reduced [1]. The major disadvantages that have hampered widespread industrial application of extractive bioconversions with ATPSs are the costs of the polymers [3,9,10,11°,12"], and the difficulties in predicting the biocatalyst and product partitioning, as well as the biocatalyst activity in ATPSs, due to the lack of adequate mathematical models [1,3,13,14]. Key issues in designing technically and economically feasible processes are the partitioning of biocatalysts and their products in ATPSs, the activity and stability of biocatalysts in ATPSs, the mode of operation, equipment design, process control, downstream processing and, what is most important, process economics. In this paper, recent developments that may improve the technical and economical feasibility of extractive bioconversions with ATPSs are reviewed.
Introduction
Extractive bioconversion is one of the techniques for in situ product recovery, the separation of products from their biocatalyst immediately after they are formed [1]. In extractive bioconversions two immiscible liquid phases are employed, the objective being to run a bioconversion in one liquid phase, while the product is extracted into the other phase. Aqueous two-phase systems are formed by mixing two aqueous polymer solutions (or a polymer and a salt solution). Above certain concentrations two immiscible aqueous phases will form, each enriched in one of the polymers [2]. Aqueous two-phase systems (ATPSs) are especially useful in extractive bioconversions, because they constitute a relatively mild environment for biocatalysts [2,3] and they are suitable for extraction of hydrophilic products, such as proteins [4-6]. The benefits of in situ product recovery techniques are to be found mainly in their potential for improving yield
Partitioning
The first objective in designing ATPSs for extractive bioconversions is to separate the product and its biocatalyst. For efficient processing the biocatalyst has to partition completcly to one of the liquid phases, while the product concentrates in the other phase. In this manner the biocatalyst can be retained and reused by recycling the biocatalyst-containing phase, the product concentration in the vicinity of the biocatalyst is low, and a biocatalyst-free product phase is available for further downstream processing. Cells, like all large particles, mainly partition between one of the liquid phases and the interface. The cell concentration in the other phase is typically several orders of magnitude less, so in essence it is free of cells [2,3,11°,15-17,18°°]. During continuous processing some leakage of cells into the extracting phase is less
172
Biochemicalengineering
of a problem, because it is compensated for by cell growth; however, prior to further downstream processing an additional cell separation step will be required. Enzymes, like all macromolecular sized particles, are distributed far more evenly between the two liquid phases than cells. During continuous processing, therefore, enzymes will be extracted rapidly with the product extracting phase and enzyme retention in the bioreactor will be poor [2,3]. In contrast to cells, this leakage is not compensated for and therefore may be problematic. To achieve a one-sided partitioning of enzymes, whole-cell enzyme preparations [19], immobilized enzymes [20] and enzymes chemically coupled to polyethylene glycol (PEG) [21] have been used. Typical products of extractive bioconversions in ATPSs so far have been, on one hand, low molecular weight inhibitory products, such as lactic acid [11°,12°,22,23°°], sugars from particulate substrates, such as starch [24-27], cellulose [28] or chitin [17,29], and easily degradable products, such as antibiotics [30,31]. On the other hand, high-molecular weight products, such as extracellular enzymes [16,32,33] or even high-value (recombinant) proteins [18°°,34,35°°,36°°], have been produced in ATPSs. Low molecular weight products generally partition evenly between the phases, high molecular weight products tend to concentrate in one of the liquid phases. Despite numerous theoretical models [37], to date no practical models exist to predict the partitioning of biocatalysts and their products in a given ATPS, based on their physical and chemical properties. Furthermore, the partitioning of biocatalysts, products, and also polymers may change as bioconversions progress, because biomass and metabolites can alter the composition of the aqueous phases [16,23°°]. Two recently employed approaches [18°°,23°°,38,39°,40 ° ] may greatly assist in generating dynamic models that describe the partitioning in ATPS media for extractive bioconversions more accurately, also taking biomass and metabolite effects into account. In the first place, the use of analytical tools to directly determine the composition of each phase in complex mixtures [23 °°] constitutes a significant improvement compared to the traditional situation where phase diagrams in water were used to simulate those in complex mixtures. Secondly, the use of statistical experimental design [18°°,38,39°,40 °] may provide empirical partitioning models in which more relevant factors, such as metabolite and biomass concentrations, can be included, while the required number of experiments to generate data for these models is reduced to a minimum. Because selecting ATPSs with the proper biocatalyst partitioning is the starting point in ATPS design, the opportunities for maximizing product partitioning to the opposite phase are sometimes limited. Essentially, two
approaches can be used for improving product partitioning in a given system. One is to add a specific extractant to the phase system, such as an affinity ligand coupled to a phase-forming polymer [40°,41-43]; however, this may affect the partitioning of other components, for instance cells, as well [40°,44]. Furthermore, these additives may be expensive. T h e other approach is to alter the product in such a way that its partitioning improves. With recombinant proteins tagging is a common procedure to improve further downstream processing. This procedure has also been used in ATPSs [45,46]. Here the consequences with respect to protein folding and removal of the tag further downstream in the process have to be considered.
Activity and stability of biocatalysts T h e second major objective in ATPS design is to create an environment that allows biocatalysis, that is, cell growth and production of enzyme activity, at similar specific activities and stabilities compared to culture media or buffers without polymers. Extractive bioconversions with enzymes have been executed so far in polymer-polymer ATPSs. T h e specific enzyme activities in ATPSs are generally not affected by the phase-forming polymers and are similar to those in reference media [3,19,21,24,25]. Reports on the effects of ATPSs on enzyme stability vary from a slight increase [24,25,28] to a slight decrease [29]. Cell growth has been reported in ATPSs with neutral polymers, with charged polymers and even with polymers and salts. Bacterial and fungal cell growth appears to be hardly affected in ATPSs with moderate concentrations of neutral polymers [3,12°,16,17,32,33,47], unless oxygen transfer due to the increased viscosity becomes rate limiting; however, changes in morphology, such as cell aggregation and also changes in cell-surface hydrophobicity, have been reported [3]. Charged polymers can offer the advantage of improved product partitioning as well as ease of polymer recovery. Recently, the cultivation of lactic acid producing bacteria in ATPSs containing a charged polymer was described [11",22]. T h e inhibitory effect of the charged polymer was largely overcome by adaptation of the cells to the ATPS culture medium, although cell aggregation was still observed. T h e production rates were similar to those of the control cultures. A major breakthrough is the recently reported cultivation of several bacterial species in polymer-salt ATPSs [10,34]. Polymer-sah ATPSs can be cost effective for protein extractions [6,10,48]. Furthermore, their suitability for large-scale extractions of products with a wide range of hydrophobicities has long been recognized [6,48]. Cell growth was found to be largely unaffected by PEG when the molecular weight was 4000kD or higher;
Extractive bioconversionsin aqueous two-phase systems Zijlstra,de Gooijerand Tramper
however, it was inhibited exponentially with increasing salt concentrations [10]. Another major breakthrough is the recently reported cultivation of animal cell lines excreting high value proteins, such as monoclonal antibodies, in ATPSs [36°°]. In ATPSs of PEG and dextran, hybridoma cell growth was shown over a period of months to be similar to growth rates of control cultures. Finally, successful plant cell cultivation in ATPSs has been reported in ATPSs with neutral polymers [49,50]; however, product excretion, required for successful in situ product removal, remains an issue to be improved. To date, no models are available to describe the effect of the phase-forming constituents on cell growth and production. Again, like the partitioning behavior, empirical correlations need to be developed in order to produce guidelines for designing better ATPSs. M o d e of o p e r a t i o n T h e mode of operation of extractive bioconversions can vary from simple batch processes to full continuous processes with biocatalyst recycling or even polymer recycling. Batch extractive bioconversions have been reported most frequently in recent literature [10,11°,17,18°',20,22,26,27, 29,36°°,51], their main advantage being simplicity. Once an ATPS has been designed that provides adequate separation between the biocatalyst and its product, and supports adequate activity and stability of the biocatalyst in small-scale experiments, the process can easily be scaled up in conventional stirred tanks. T h e product can be harvested by removing the product-containing phase by settling or centrifugation. T h e time required for batch process development, therefore, is minimal. Disadvantages in the case of low molecular weight products, or other products that distribute evenly between the phases, are that the product-biocatalyst interaction is substantial and the final product yield in the extracting phase is low. Furthermore, since the residence time of the product in the reaction mixture is equal to the batch cycle time, the yield of unstable or easily degradable products may be compromised. Finally, especially for bioconversions with cells, the volumetric productivity of batch processes is low compared to fed-batch processes or continuous processes with cell retention. Continuous processing with biocatalyst and polymer recycling may compensate for most of the disadvantages of batch processes. This highly integrated approach, however, inherently leads to a higher level of complexity and it poses stringent requirements on both equipment design and process control. Many reports so far, therefore, only describe repeated extractive batch processes, where the product extracting-phase is periodically replaced with fresh extracting phase [12°,23"°,25,34,35"°]. Also,
173
continuous processes in which the extracting phase is continuously replaced and the biocatalyst-containing phase is recycled by means of an in-line settler are described [16,19,28,47]. In-line recycling of polymers from the product-containing phase, after removal of the product, has been reported rarely, because of the difficulties to achieve this on a small scale [24]. Equipment design There are important issues to take into account during the equipment design for extractive bioconversions: firstly, the mass transfer of metabolites and product within each phase; secondly, the mass transfer of metabolites and product between the phases; thirdly, the phase separation; fourthly, the separation of the product from the extracting phase; and finally, the recycling of polymers. Stirred tanks, operated at moderate stirring speeds, generally have been found to provide adequate mass transfer for extractive bioconversions in ATPSs. Because of the low interfacial tensions in ATPSs, emulsions with high interfacial areas, and thus with high mass transfer rates, will already occur at low power input [3,9,52,53]. To separate the ATPS emulsions, however, long settling times are required. In batch extractions, phase separation can be obtained by settling in the stirred tank itself. For continuous extractive bioconversions with biocatalyst recycling, however, dedicated settlers are required. Several settler designs have been described in the literature, although none of them seem to be very effective at high flow rates [16,19,24,47]. Furthermore, the residence time in these settlers may cause oxygen depletion when growing cells are used [16,47], and when cells partition to the interface they can accumulate in the settler and compromise the effectiveness of the phase separation [47]. Complete cell partitioning to one of the liquid phases, therefore, is of paramount importance. Although for large-scale ATPS extractions the use of industrial continuous (disk-stack) centrifuges has been reported [6], and meanwhile specific continuous centrifuges for cell recycling have become available on the market, so far their use in extractive bioconversions with ATPSs has not been reported. To avoid phase-separation difficulties, and moreover to obtain multistage extraction of the product, the opportunities for counter-current extraction using spray and other types of column extractors have been investigated [52]. For polymer-polymer ATPSs, however, the mass-transfer coefficients were extremely low. Furthermore, to avoid flooding, the residence time in these extractors has to be long, which practically excludes their use for extractive bioconversions with cells. Process control In batch processes only some basic process parameters in the stirred tank need to be controlled. Typically the
1 74
Biochemical engineering
temperature, pH, stirrer speed, and in case of cells the dissolved oxygen concentration are important. In continuous processes, beside the process parameters of the stirred tank the parameters of the phase separator, as well as the parameters of incoming and outgoing fluxes, also need to be controlled. For effective continuous extractive bioconversions the aim is to obtain and maintain a steady-state situation, under optimal process conditions, usually after an initial dynamic phase where, for instance, the biomass increases. Recent studies have shown, however, that shifts in the polymer concentrations of each phase can occur during the dynamic phase of extractive bioconversions. These shifts can result in changes in cell, metabolite and product partitioning and, ultimately, in system breakdown [23"°,47]. Controlling the ATPS composition, therefore, is vital in continuous extractive bioconversions. Novel on-line analytical methods may be used for this purpose [23°°]. Furthermore, process models consisting of mass balances of all phase-system constituents, models describing the partitioning, as well as models describing the bioconversion in ATPSs, seem to be necessary for process control, especially during the dynamic phase [54,55]. Downstream
processing
Depending on the product type and the required purity, one or a combination of conventional downstream processing techniques, such as extraction, evaporation, filtration, ultrafiltration, chromatography and so on, can be used for further purification. T h e first purification step generally includes separation of the polymers from the product. Recently, new polymer types have been employed that form a precipitate [56] or an immiscible liquid phase highly concentrated in the polymer [11°,12",57,58] after a shift in pH or temperature. T h e polymers can be recovered or even recycled after phase separation, while the polymer-lean product phase is available for further downstream processing. Process
explored [10,34] and polymer recycling schemes are being developed [ 11 °, 12",24,57,58]; and finally, the opportunities for using extractive bioconversions in high-value-product manufacturing are being explored [18°°,35°',36°',39°,40"]. Also, scale appears to be crucial for the economically effective implementation of extractions [13]. Once the knowledge of extractive bioconversions has increased to a level where rational process design can take place and in the meantime low-cost ATPSs have become available, extractive bioconversions with ATPSs become an important tool for the large-scale manufacturing of biologicals. Conclusions
Extractive bioconvcrsions with ATPSs offer a number of unique features which make them exceptionally useful for large-scale manufacturing of biologicals. T h e current knowledge of this technology, however, is not sufficient for rational process design. Furthermore, the costs associated with implementing extractive bioconversions are still too high. T h e continuing developments in the field of ATPSs, in combination with important recent findings, may greatly improve the practical applicability of extractive bioconversions with ATPSs in the near future. Acknowledgements
This workwas supported by Gist-brocades/Bio-lntermediairBV,Groningen, The Netherlands and by a grant under the Dutch PBTS-Biotechnology program, Ministry of EconomicAffairsNr. BIO89041. References and recommended reading Papers of particular interest, published within the annual period of review, have been highlighted as: * ..
Freeman A, Woodley JM, Lilly MD: In situ product removal as a tool for bioprocessing. Bio/Techno/ogy 1993, 11:1007-1012. AIbertsson PA: Partitioning of Cel/ Partic/es and Macromo/ecu/es, edn 3. New York: John Wiley & Sons; 1986. Andersson E, Hahn-H~igerdal B: Bioconversions in aqueous two-phase systems. Enzyme Microb Techno11990, 12:242-253.
economics
Obviously, for the breakthrough of extractive bioconversions with ATPSs it is essential that the costs of implementation are favorable compared to the revenues of the product, and that the extractive bioconversion is more cost effective than conventional processing. Previous economical evaluations have shown that the polymer costs largely determine the process economics. In particular, fractionated dextran is too expensive for industrial application in extractive bioconversions [3,9,53,59]. To tackle this problem, several approaches have been employed; firstly, low-cost alternatives to dextran are being developed, mainly based on starch derivatives [11°,60,61]; secondly, the possibilities for using low cost PEG-salt systems for extractive bioconversions are being
of special interest of outstanding interest
Kroner K, Kula MR: Extraction of enzymes in aqueous twophase systems. Process Biochem 1978, 13:7-10. Kula MR, Kroner K, Hustedt H: Purification of enzymes by liquidliquid extraction. Adv Biotechno/Eng 1982, 24:73-118.
7. 8. 9. lO.
Hustedt H, Kronor KH, Papamichael N: Continuous cross-current aqueous two-phase extraction of enzymes from biomass: automated recovery in production scale. Process Biochem 1988, 23:129-137. Daugulis AJ: Integrated reaction and product recovery in bioreactor systems. Biotechno/Prog 1988, 4:113-122. Daugulis AJ: Integrated fermentation and recovery processes. Curr Opin Biotechno/1991, 5 :192-195. Kaul R, Mattiasson B: Extractive bioconversions in aqueous two-phase systems. Bioprocess Techno/t 991, 11:173-188. Kuboi R, Maruki T, Tanaka H, Komasawa I: Fermentation of Bacillus subtilis ATCC 6633 and production of subtilin in polyethylene glycol/phosphate aqueous two-phase systems. J Ferm Bioeng 1994, 78:431-436.
Extractive bioconversions in aqueous two-phase systems Zijlstra, de Gooijer and Tramper
11.
KwonYJ, Kaul R, Mattiasson B: Extractive lactic acid fermentation in poly(ethyleneimine)-based aqueous two-phase system. Biotechno/Bioeng 1996, 50:280-290. This study shows the usefulness of charged polymers in combination with low-cost polymers in extractive fermentations with bacteria. Charged polymers can be separated from the product extracting phase by phase separation using multivalent ions.
29.
Chen JP, Lee MS: Enhanced production of Serratia marcescens chitinase in PEG/dextran aqueous two-phase systems. Enz Microb Techno/1995, 17:1021-1027.
30.
Chang HN, Lee YH, Lee CY: Continuous production of 6 - A P A in an aqueous two-phase system. Ann NY Acad Sci 1992, 672:643-648.
31.
Paquet V, Myint M, Roque C, Soucialle P: Partitioning of pristinamycins in aqueous two-phase systems: a first step toward the development of antibiotic production by extractive fermentation. Biotechno/Bioeng 1994, 44:445-451.
32.
PerssonI, Staalbrand H, Tjerneld F, Hahn-h~gerdal B: Semicontinuous production of cellulolytic enzymes with Trichoderma reesei Rutgers C30 in an aqueous two-phase system. App/ Biochern Biotechno/1992, 27:27-36.
33.
PerssonI, Tjerneld F, Hahn-haegerdal B: Influence of cultivation conditions on the production of cellulolytic enzymes with Trichoderma reesei Rutgers C30 in aqueous two-phase systems. App/ Biochem Biotechno/1992, 27:9-25.
34.
Kuboi R, Umakoshi H, Komasawa h Extractive cultivation of Escherichia coil using poly(ethylene glycol)/phosphate aqueous two-phase systems to produce intracellular ]3galactosidase. Biotechno/Prog 1995, 11:202-207.
•
12. •
PlanasJ, Radstrom P, Tjerneld F, Hahn-Hagerdal B: Enhanced production of lactic acid through the use of a novel aqueous two-phase system as an extractive fermentation system. App/ Microbiol Biotechnol 1996, 45:737-743. This study shows the usefulness of EO-PO polymers in extractive fermentations with bacteria. The EO-PO polymers can be separated from the product extracting phase by a temperature shift. 13.
Asenjo JA: Industrial prospects of aqueous two-phase processes. J Chem Technol Biotechnol 1994, 59:109.
14.
MistrySL, Kaul A, Merchuk JC, Asenjo JA: Mathematical modeling and computer simulation of aqueous two-phase continuous protein extraction J Chromatogr 1996, 741:151-163.
15.
Walter H: Separation and subfractionation of blood cell populations based on their surface properties by partitioning in two-polymer aqueous phase systems. In Ceil Separation: Methods and Se/ected Appfications. Edited by Pretlow TG II, Pretlow ThP II. New York: Academic Press; 1982, 1:261-299.
16.
Park KM, Wang NS: c~-amylase fermentation with Bacillus amylofiquefaciens in an aqueous two-phase system. Biotechno/ Prog 1991,7:439-444.
17.
Chen JP, Lee MS: Simultaneous production and partition of chitinase during growth of Serratia marcescens in an aqueous two-phase system. Biotechnol Tech 1994, 8:?83-?88.
18. •o
Zijlstra GM, Michielsen M, De Gooijer CD, Van der Pol LA, TramperJ: Hybridoma and CHO cell partitioning in aqueous two-phase systems. Biotechnol Prog 1996, 12:363-370. This study uses a statistical experimental design approach to obtain an empirical model for the partitioning of cells. These models are of paramount importance for improved process control of extractive fermentations.
1 ?5
35. o.
UmakoshiH, Yano K, Kuboi R, Komasawa h Extractive cultivation of recombinant Escherichia coil using aqueous two-phase systems for production and separation of intracellular heat shock proteins. Biotechno/Prog 1996, 12:51-56. This study demonstrates the feasibility of producing high-value recombinant proteins from E. coil by means of extractive fermentation in ATPSs. A semicontinuous production scheme is proposed, to allow periodical cell disruption and product release. 36.
Zijlstra GM, de Gooijer CD, Van der Pol LA, Tramper J: Design of aqueous two-phase systems supporting animal cell growth: a first step towards extractive bioconversions. Enz Microb Techno/1996, 19:2-8, This study demonstrates that the principle of extractive fermentation in ATPSs can also be applied to animal cells. The benefits of extractive fermentations, continuous operation with scalable cell retention, in situ product removal and integration of downstream processing with fermentation, now become available for animal cell cultivation. o.
19.
TomaskaM, Stredansky M, Tomaskova A, Sturdik E: Lactose hydrolysis in aqueous two-phase system by whole-cell betagalactosidase of Kluyveromyces marxianus. Bioproc Eng 1995, 12:17-20.
20.
Kondo A, Urabe T, Higashitani K: Bioconversions in an aqueous two-phase system using enzymes immobilized on ultrafine silica particles. J Ferrn Bioeng 1994, 6:700-703.
37.
Abbott N, Blankschtein D, Hatton A: On protein partitioning in two-phase aqueous polymer systems. Bioseparation 1990, 1:191-225.
21.
MukatakaS, Haynes CA, Prausnitz JM, Blanch HW: Extractive bioconversions in aqueous two-phase systems: enzymatic hydrolysis of casein proteins. Biotechno/Bioeng 1992, 40:195206.
38.
Hart RA, Ogez JR, Builder SE: Use of multifactorial analysis to develop aqueous two-phase systems for isolation of nonnative IgF1. Bioseparation 1995, 5:113-121.
22.
Dissing U, Mattiasson B: Cultivation of Lactococcus lactis in a polyelectrolyte-neutral polymer aqueous two-phase system. Biotechno/Lett 1994, 16:333-338.
23.
PlanasJ, Lefebvre D, Tjerneld F, Hahn-H&gerdahl B: Analysis of phase composition in aqueous two-phase systems using a two-column chromatographic method: application to lactic acid production by extractive fermentation. Biotechno/Bioeng 1997, 54:303-311. This study unambiguously shows that drifts in phase compositions occur during extractive fermentations. In order to maintain a steady state during continuous processing, sophisticated process control is necessary. The on line measurement method described is an important tool in process control. e•
39.
Zijlstra GM, Michielsen MJF, de Gooijer CD, Van der Pol LA, Tramper J: Separation of hybridoma cells from their I g G product using aqueous two-phase systems. Bioseparation 1996, 6:201-210. This study uses a statistical experimental design to generate data for the modeling of product partitioning. •
40. •
Zijlstra GM, Michielsen MJF, de Gooijer CD, Tramper J: Selection criteria for effective ligands in extractive fermentations with animal cells. Bioseparations 1998, in press. This study characterizes the effects of ligands coupled to PEG on both product and cell partitioning. 41.
Koppersl&ger G, Birkenmeier G: Affinity partitioning and extraction of proteins. Bioseparation 1990, 1:235-254.
42.
BirkenmeierG, Vijayalakshmi MA, Stigbrand T, Kopperschl&ger G: Immobilized metal ion affinity partitioning, a method combining metal-protein interaction and partitioning of proteins in aqueous two-phase systems. J Chromatogr 1991, 539:26?-277.
24.
LarssonM, Arasaratnam V, Mattiasson B: Integration of bioconversion and downstream processing: starch hydrolysis in an aqueous two-phase system. Biotechnol Bioeng 1989, 33:758-766.
25.
Tjerneld F, Persson I, Lee JM: Enzymatic cellulose hydrolysis in an attrition bioreactor combined with an aqueous two-phase system. Biotechnol Bioeng 1991, 37:876-882.
43.
26.
Liakopoulou-Kyriakides M, Karakatsanis A, Stamatoudis M: Enzymic hydrolysis of starch in agitated PEG-dextran aqueous two-phase system. Starch 1996, 48:291-294.
Garg N, Galaev IY, Mattiasson B: Dye-affinity techniques for bioprocessing: recent developments. J Mo/Recognit 1996, 9:259-274.
44.
2"7.
Min KH, Chang WJ, Koo YM: Increased production of betacyclodextrin using an aqueous two-phase system. Biotechno/ Tech 1996, 10:395-400.
Hamamoto R, Kamihira M, lijima S: Specific separation of animal cells using aqueous two-phase systems. J Ferm Bioeng 1996, 82:73-76.
45.
Taguchi F, Yamada K, Hasegawa K, Taki-Saito T, Hara K: Continuous hydrogen production by Clostridium sp. strain no. 2 from cellulose hydrolysate in an aqueous two-phase system. J Ferm Bioeng 1996, 82:80-83.
K6hler K, Ljungquist C, Kondo A, Veide A, Nilsson B: Engineering proteins to enhance their partition coefficients in aqueous two-phase systems. Bio/Techno/ogy 1991, 9:642-646.
46.
Carlsson M, Berggren K, Linse P, Veide A, Tjerneld F: Effects of fused tryptophan-rich peptides to a recombinant protein A domain on the partitioning in polyethylene glycol/dextran
28.
176
Biochemical engineering
aqueous two-phase systems. J Chromatogr A 1996, 756:107117 4?.
KatzbauerB, Cesi V, Narodoslawsky M, Moser A: Extractive lactic acid fermentation using aqueous two-phase systems. Chem Biochem Eng 19951 9:79-87.
48.
Kula MR: Trends and future prospects of aqueous two-phase extraction. Bioseparation 1990, 1:181-189.
49.
Buitelaar RM, Leenen EJTM, Tramper J: Growth and secondary metabolite production by hairy roots of tagetes patula in aqueous two-phase systems. Biocatalysis 1992, 6:73-80.
50.
IlievaMP, Bakalova A, Mihneva M, Pavlov A, Dolapchiev L: Production of phosphomonoesterases by Nicotiana tabacum 1507 in an aqueous two-phase system. Biotechnol Bioeng 1996, 51:488-493.
51.
52.
53.
54.
RamadasM, Hoist O, Mattiasson B: Production of amyloglucosidase by Aspergillus niger under different cultivation regimens. World J Microbiol Biotechno11996, 12:267-271. JoshiJB, Sawant SB, Rao KSMSR, Patil TA, Rostami KM, Sikdar SK: Continuous counter-current tw0-phase aqueous extraction. Bioseparation 1990, 1:311-324. Huddleston JG, Lyddiatt A: Aqueous two-phase systems in biochemical recovery. Appl Biochem Biotechno11990, 26:249280. JarzebskiAB, Malinowski JJ: Analysis of continuous fermentation processes in aqueous two-phase systems. Bioprocess Eng 1992, 7:315-31"7.
55.
BaughmanDR, Liu YA: An expert network for predictive modeling and optimal design of extractive bioseparations in aqueous two-phase systems. Ind Eng Chem Res 1994, 33:2668-2687.
56.
KamiharaM, Kaul R, Mattiasson B: Purification of recombinant protein A by aqueous two-phase extraction integrated with affinity precipitation. Biotechno/Bioeng 1992, 40:1381-1387.
67
FarkasT, Stalbrand H, Tjerneld F: Partitioning of betamannanase and alpha-galactosidase from Aspergillus Niger in Ucon/Reppal aqueous two-phase systems and using temperature induced phase separation. Bioseparation 1996, 6:147-157.
58.
Lu M, Albertsson PA, Johansson G, Tjerneld F: Ucon-benzoyl dextran aqueous two-phase systems: protein purification with phase component recycling. J Chromatogr B Biomed App11996, 680:65-?0.
59.
Datar R: Economics of primary separation steps in relation to fermentation and genetic engineering. Proc Biochem 1986, Feb:l 9-26.
60.
TjerneldF, Berner S, Cajarville A, Johansson G: New aqueous two-phase systems based on hydroxypropyl starch useful in enzyme extraction. Enz Microbiol Technol 1986, 8:41 ?-423.
61.
Sz[agDC, Guiliano KA, Snyder SM: A low cost aqueous two-phase system for affinity extraction. In Bioseparations. Downstream Processing in Biotechnology. Edited by Hamel JF, Hunter JB, Sikdar SK. Washington DC: ACS Symposium series; 1990, 419:71-86.