Search

In recent years, confocal laser scanning microscopy has been developed into a non-invasive tool to probe intra-particle profiles of protein in chromatographic adsorbents. A necessary prerequisite when using this technique lies in the labeling of proteins with fluorescent probes. The quality of the obtained results is thus strongly dependent on the probes used, its sensitivity on experimental parameters and the change of protein characteristics upon binding. In this review, the fundamental issues when using fluorescent probes are described, before giving a critical evaluation on published literature in the field of confocal laser scanning microscopy for the analysis of chromatographic principles.

In the current study we developed a process for the capture of pDNA exploiting the ability of aqueous two-phase systems to differentiate between different forms of DNA. In these systems scpDNA exhibits a near quantitative partitioning in the salt-rich bottom phase. The successive recovery from the salt rich bottom phase is accomplished by a novel membrane step. The polish operation to meet final purity demands is again based on a system exploiting a combination of the denaturation of the nucleic acids present, specific renaturation of scpDNA, and an ATP system able to differentiate between the renatured scpDNA and the denatured contaminants such as ocpDNA and genomic host DNA. This polish step thus allows a rapid and efficient separation of scpDNA from contaminating nucleic acids which up to date otherwise only can be accomplished with much more cumbersome chromatographic methods. In a benchmark comparison, it could be shown that the newly developed process exhibits a comparable yield to an industrial standard process while at the same time showing superior performance in terms of purity and process time. Additionally it could be shown that the developed polish procedure can be applied as a standalone module to support already existing processes.

Expanded bed adsorption (EBA) is an integrative step in downstream processing allowing the direct capture of target proteins from cell-containing feedstocks. Extensive co-adsorption of biomass, however, may hamper the application of this technique. The latter is especially observed at anion exchange processes as cells or cell debris are negatively charged under common anion exchange conditions. The restrictions observed under these conditions are, however, directly related to processing steps prior to fluidised bed application. In this study, it could be shown that the effective surface charge of cell debris obtained during homogenisation is closely related to the debris size and thus to the homogenisation method and conditions. The amount and thus effect of cells binding to the adsorbent could be significantly decreased when optimising the homogenisation step not only towards optimal product release but towards a reduction of debris size and charge. The lower size and charge of the debris results not only in a reduced retention probability but also, in a lower collision probability between debris and adsorbent. The applicability was shown in an example where the homogenisation conditions of E. coli were optimised towards EBA applications. In a previous report (Reichert et al., 2001) studying the suitability of EBA for the capture of formate dehydrogenate from E. coli homogenate the pseudo affinity resin Streamline Red was identified as the only suitable adsorbent. The new approach, however, led to a system where anion exchange as capture step became possible, however, to the cost of binding capacity., (2006 Wiley Periodicals, Inc.)

In this study, the interplay of two linked equilibria is examined, one concerning an aqueous two-phase system (ATPS) composed of poly(ethylene glycol) (PEG) and salt employed to partition plasmid DNA (pDNA), and the other a potential structural transition of pDNA depending on PEG and salt concentration and other system parameters. The boundary conditions for pDNA partitioning are set by PEG and salt concentrations, PEG molecular weight, pH, and temperature. While investigating these parameters, it was found that a small increase/decrease of the respective values led to a drastic and significant change in pDNA behavior. This behavior could be attributed to a coil-globule transition of the pDNA triggered by the respective phase conditions. The combination of this structural change, aggregation effects linked to the transition process, and the electrostatic potential difference found in PEG-salt systems thus offers a sensitive way to separate nucleic acid forms on the basis of their unique property to undergo coil-globule transitions under distinct system properties.

Expanded bed adsorption (EBA) was evaluated for the isolation of the human insulin precursor MI3, expressed and secreted by the yeast Saccharomyces cerevisiae. The isoelectric point of the insulin precursor (pH 5.3) makes cation exchange a prime candidate for direct adsorption. In order to find a suitable window of operation for the process the adsorption equilibrium was analysed in a wide range of operating conditions (pH and conductivity) and for three different stationary phases. The same array of operating conditions was examined with regard to stable fluidisation of the adsorbents in S. cerevisiae suspensions. Interactions of the yeast with the fluidised stationary phase were investigated by a pulse response technique and the hydrodynamics of the fluidised bed under process conditions by residence time distribution analysis. The case study demonstrates that by parallel examination of product binding and fluidisation quality a window of operation can be found. Analysis of the binding kinetics by breakthrough experiments and modelling led to the definition of a set of operating conditions, which yield a compromise between optimal use of the equilibrium capacity provided by the adsorbent and high throughput required for an industrial separation. After initial experiments on the bench scale the protocol was transferred successfully to pilot scale demonstrating the design of a reliable operation.

Future developments in gene therapy and DNA vaccination depend on cost-effective large-scale production of pharmaceutical-grade pDNA (plasmid DNA). Given the large amount of impurities present in the feedstock, purification processes that have high specificity and capacity at a moderate cost are required. In the present study, we describe a non-chromatographic procedure based on aqueous two-phase extraction allowing a fast and simply scalable capture step. PEG [poly(ethylene glycol)] in combination with potassium citrate or potassium phosphate was tested as phase component for extraction. By increasing either PEG or salt concentration, the partitioning of nucleic acids changed from bottom to top phase. Phase systems with a composition of 15% PEG 800 and 20% potassium phosphate at pH 7.0 showed a strong partitioning of pDNA to the bottom phase, linked to a clear decrease in open circular pDNA, while proteins, genomic DNA and RNA remain at the top or at the interphase. A great advantage of the current process is that the complete procedure of lysis, precipitation, clarification and extraction can be performed in a single vessel. The number of denatured and sheared genomic DNAs in a spiking experiment was found to be depleted by more than 99%.

The gene for the newly described D-amidase from Variovorax paradoxus (Krieg et al. 2002) was cloned and functionally expressed in Escherichia coli. Since native enzyme was available in minute amounts only, we determined the N-terminal sequence of the enzyme and utilized the Universal GenomeWalker Approach to make use of the common internal sequence of the amidase signature family. The high GC content of the gene made it necessary to employ an appropriate DNA polymerase in the amplification reactions. Thus, the sequence of the complete gene and the flanking regions was established. In independent experiments, the gene was then amplified from genomic DNA of V. paradoxus, expressed in E. coli, and characterized. The recombinant enzyme has a specific activity of 1.7 units/mg with racemic tert-leucine amide as substrate and is a homodimer of 49.6-kDa monomers.

The economic feasibility of a biotechnological production process is directly linked to the number and efficiency of the processes required during purification in order to reach the requested product specifications. Expanded bed adsorption is an integrated process that combines solid-liquid separation and product recovery into a single unit operation. This approach suggests an increased overall yield, lower requirements for capital investments and consumables and most importantly a reduced process time. The success of an integrative process is, however, closely linked to a detailed understanding of the biochemical principles involved and the constraints arising from feedstock complexity. This paper reviews the hydrodynamic principles that expanded bed adsorption is based on, discusses current developments in resin and column design, and finally presents a methodology for targeted process development in expanded bed adsorption.

Expanded bed adsorption is an innovative chromatographic technology that allows the introduction of particle-containing feedstock without the risk of blocking the bed. Provided a perfectly classified fluidized bed (termed expanded bed) is formed in the crude feedstock and the biomass is not influencing protein transport towards the adsorbent surface, a sorption performance comparable to packed beds is found. The influence of biomass on the hydrodynamic stability of expanded beds is essential and was investigated systematically in this article. Residence-time distribution analyses were performed using model systems and a yeast suspension under various fluid-phase conditions. It is demonstrated that three factors (biomass/adsorbent interactions, biomass concentration, and flow rate) play an interdependent role disturbing the classified fluidization of an expanded bed. A clear correlation between the degree of aggregative fluidization--obtained by PDE modeling of RTD data--and the expansion behavior of the fluidized bed has been found. Thus, combining three analytical methods, namely cell transmission index analysis, expansion analysis, and RTD analysis provides a solid base for understanding and control of the fluidization behavior and thus further process design during the initial phase of process development.

An experimental study on the interplay of sorbent structure and fluid phase conditions (pH) has been carried out examining adsorption and transport of bovine serum albumin (BSA) and a monoclonal antibody (IgG 2a) on SP Sepharose Fast Flow and SP Sepharose XL. SP Sepharose Fast Flow is characterised by a relatively open pore network, while SP Sepharose XL is a composite structure with ligand-carrying dextran chains filling the pore space. Both adsorbents have similar ionic capacity. Protein transport and adsorption profiles were evaluated using confocal laser scanning microscopy. Under all investigated conditions, BSA uptake could be adequately explained by a pore diffusion mechanism. The adsorption profiles obtained for IgG 2a, however, indicated that changes in fluid phase conditions as well as a change in the solid phase structure could result in a more complex uptake mechanism as compared to pore diffusion alone. This mechanism results in a fast transport of proteins into the adsorbent, followed by an overshoot of protein in the center of the sorbent and a setback towards a homogeneous adsorption profile.

The impact of different transport mechanism on chromatographic performance was studied by confocal laser scanning microscopy (CLSM) for solutions containing bovine serum albumin (BSA) and monoclonal IgG 2a under different solid- and fluid-phase conditions. During this investigation, a clear influence of the uptake mechanism on the affinity of the respective proteins for the different adsorbents and thus separation performance of the chromatographic process could be observed. For the system SP Sepharose Fast Flow at pH 4.5 pore diffusion could be ascribed to be the dominant transport mechanism for both proteins and the adsorption profiles resembled a pattern similar to that described by the 'shrinking core' model. Under these conditions a significantly higher affinity towards the adsorbent was found for BSA when compared to IgG 2a. With changing fluid- and solid-phase conditions, however, a change of the transport mode for IgG 2a could be detected. While the exact mechanism is still unresolved it could be concluded that both occurrence and magnitude of the now governing transport mechanism depended on protein properties and interaction with the adsorbent surface. For the system SP Sepharose XL at pH 5.0 both parameters leading to the change in IgG 2a uptake were combined resulting in a clear change of the system affinity towards the IgG 2a molecule, while BSA adsorption was restricted to the most outer shell of the sorbent.

Expanded bed adsorption is an integrated technology that allows the introduction of particle-containing feedstock without the risk of blocking the bed. The biomass particles contained in the feedstock have to be treated as an integral part of the process and potential interactions between suspended biomass and the adsorbent must be excluded during process design. Because the electrostatic forces dominate the interactions between the biomass and adsorbent, the zeta potential has been studied as a tool to characterize biomass/adsorbent electrostatic interactions. The zeta potentials of four types of biomass (yeast intact cells, yeast homogenate, Escherichia coli intact cells, and E. coli homogenate) and two types of ion exchanger were measured systematically at varying process conditions. Using the cell transmission index from biomass pulse-response experiments as a parameter, the relations between zeta potential and the biomass/adsorbent interaction were evaluated. Combining the influences from zeta potential of adsorbent (zeta(a)), zeta potential of biomass (zeta(b)), and biomass size (d), parameter (-zeta(a)zeta(b)d) was found to be an appropriate indicator of the biomass/adsorbent interactions in expanded beds under various liquid-phase conditions for different types of biomass. The threshold value of parameter (-zeta(a)zeta(b)d) can be defined as 120 mV(2) microm for cell transmission of >90%, which means that systems with (-zeta(a)zeta(b)d) < 120 may have a considerable probability of forming stable expanded beds in a biomass suspension under the particular experimental conditions., (Copyright 2003 Wiley Periodicals, Inc. Biotechnol Bioeng 83: 149-157, 2003.)

Expanded bed adsorption is an integrated technology that allows the introduction of a particle containing feedstock without the risk of blocking the bed. Provided a perfectly classified fluidized bed (termed expanded bed) is formed in the crude feed, a sorption performance comparable to packed beds is found. During the application of biomass containing samples to stable expanded beds an increase in bed expansion due to the higher density and viscosity of the feed is encountered. In this article it is investigated whether the expanded bed condition is also fulfilled during the transition in bed expansion from lower to higher density (i.e., from an equilibration buffer to a biomass containing feedstock). Residence time distribution analyses were performed by using model systems and a yeast suspension during this transition phase. It is shown that in systems in which the biomass does not interact with the fluidized stationary phase, the perfectly classified fluidization is maintained also during this transition phase regardless of the type of feedstock. Additional bed expansion takes place in an "ordered" manner without compromising bed stability. In case of biomass/adsorbent interactions, a deterioration in bed stability is found directly when the crude feed is loaded., (Copyright 2002 Wiley Periodicals, Inc.)

A new experimental set-up for on-line visualization of the intra-particle uptake kinetics during packed bed chromatography has been designed and tested. Confocal laser scanning microscopy was used to analyze the dynamics of protein adsorption to porous stationary phases. In combination with this, a flow cell was developed that could be packed with chromatography media and operated as a fully functional mini-scale chromatography column. Adsorption profiles of single- and two-component mixtures containing BSA and IgG 2a during packed bed cation-exchange chromatography were investigated. The two proteins appear to exhibit different transport characteristics. For BSA a classical "shrinking core" behavior could be detected. The profiles obtained during IgG 2a adsorption point toward a different transport mode, which deviates from the classical pore-diffusion picture. For the two-component system, a superposition of the single-component profiles combined with a classical displacement of the weaker bound protein species was found. The results indicate that depending on the adsorbed protein the uptake can vary tremendously, even for adsorption to the same chromatographic support. It is clearly shown that the new microcolumn allows in situ quantitative investigations of protein adsorption dynamics within a single particle, which adds a new tool to the available methods for characterizing and optimizing protein adsorption chromatography., (Copyright 2002 Wiley Periodicals, Inc.)

The peptide amidase from Stenotrophomonas maltophilia catalyses predominantly the hydrolysis of the C-terminal amide bond in peptide amides. Peptide bonds or amide functions in amino acid side-chains are not hydrolysed. This specificity makes peptide amidase (Pam) interesting for different biotechnological applications. Pam belongs to the amidase signature (AS) family. It is the first protein within this family whose tertiary structure has been solved. The structure of the native Pam has been determined with a resolution of 1.4A and in complex with the competitive inhibitor chymostatin at a resolution of 1.8A. Chymostatin, which forms acyl adducts with many serine proteases, binds non-covalently to this enzyme.Pam folds as a very compact single-domain protein. The AS sequence represents a core domain that is covered by alpha-helices. This AS domain contains the catalytic residues. It is topologically homologous to the phosphoinositol phosphatase domain. The structural data do not support the recently proposed Ser-Lys catalytic dyad mechanism for AS enzymes. Our results are in agreement with the role of Ser226 as the primary nucleophile but differ concerning the roles of Ser202 and Lys123: Ser202, with direct contact both to the substrate molecule and to Ser226, presumably serves as an acid/bases catalyst. Lys123, with direct contact to Ser202 but no contact to Ser226 or the substrate molecule, most likely acts as an acid catalyst.

Benzoylformate decarboxylase (BFD) from Pseudomonas putida was subjected to directed molecular evolution to generate mutants with increased carboligase activity which is a side reaction of the enzyme. After a single round of random mutagenesis mutants were isolated which exhibited a 5-fold increased carboligase activity in aqueous buffer compared to the wild-type enzyme with a high enantiomeric excess of the product (S)-2-hydroxy-1-phenyl-propanone. From the same library, mutants with enhanced carboligase activity in water-miscible organic solvents have been isolated. The selected mutants have been characterized by sequencing, revealing that all mutants carry a mutation at Leu476, which is close to the active site but does not directly interact with the active center. BFD-L476Q has a 5-fold higher carboligase activity than the wild-type enzyme. L476 was subjected to saturation mutagenesis yielding eight different mutants with up to 5-fold increased carboligase activity. Surprisingly, all L476 mutants catalyze the formation of 2-hydroxy-1-phenyl-propanone with significantly higher enantioselectivity than the wild-type enzyme although enantioselectivity was not a selection parameter. Leu476 potentially plays the role of a gatekeeper of the active site of BFD, possibly by controlling the release of the product. The biocatalyst could be significantly improved for its side reaction, the C-C bond formation and for application under conditions that are not optimized in nature.

Here we present data to demonstrate how partitioning of a hydrophilic enzyme can be directed to the hydrophobic detergent-enriched phase of an aqueous two-phase system by addition of short stretches of amino acid residues to the protein molecule. The target enzyme was the industrially important endoglucanase I, EGI (endo-1,4-beta-D-glucan-4-glucanohydrolase, EC 3.2.1.4, Cel7B) of Trichoderma reesei. We investigated the partitioning of three EGI variants containing various C-terminal peptide extensions including Trp-Pro motifs of different lengths and localizations. Additionally, a recently developed system composed of the thermoseparating copolymer HM-EOPO was utilized to study the effects of fusion tags. The addition of peptides containing tryptohan residues enhanced the partitioning of EGI to the HM-EOPO-rich phase. The system composed of a nonionic detergent (Agrimul NRE1205) resulted in the highest partition coefficient (K = 31) and yield (90%) with the construct EGI(core-P5)(WP)(4) containing (Trp-Pro)(4) after a short linker stretch. A recombinant strain of T. reesei Rut-C30 for large-scale production was constructed in which the fusion protein EGI(core-P5)(WP)(4) was expressed from the strong promoter of the cellulase gene cbh1. The fusion protein was successfully expressed and secreted from the fungus during shake-flask cultivations. Cultivation in a 28-L bioreactor however, revealed that the fusion protein is sensitive to proteases. Consequently, only low production levels were obtained in large-scale production trials., (Copyright 2002 Wiley Periodicals, Inc.)

The peptide amidase (Pam) from the gram-negative bacterium Stenotrophomonas maltophilia catalyzes predominantly the hydrolysis of the C-terminal amide bond in peptide amides. Its gene ( pam) was isolated by Southern hybridization using a DNA probe derived from the known N-terminal amino acid sequence. Pam is a member of the amidase signature family and was identified as a periplasmic protein by an N-terminal signal peptide found in the gene. The processed protein consists of 503 amino acids with a molecular mass of 53.5 kDa. The recombinant enzyme with a C-terminal His(6) tag has a monomeric structure and its isoelectric point is 6.3. The dipeptide amide L-Ala- L-Phe-NH(2) is hydrolyzed in the absence of cofactors to L-Ala- L-Phe-OH and ammonia with V(max)=194 U/mg and K(m) <0.5 mM. The natural function of Pam remains unclear. Chymostatin ( K(i)<0.3 microM) and Pefabloc SC ( K(i) not determined) were identified as inhibitors. When the gene was expressed in Escherichia coli on a 12-l scale, the specific activity in the crude extract was 60 U/mg, compared to 0.24 U/mg in S. maltophilia. In the expression system, Pam made up about 31% of the total soluble cell protein. From 75 g wet cells, 2.1 g of >95% pure enzyme was obtained, which corresponds to a total activity of 416,000 units.

The peptide amidase from Stenotrophomonas maltophilia selectively hydrolyses the C-terminal amide bond in peptide amides. Crystals have been obtained by sitting-drop vapour diffusion from solution containing polyethylene glycol (PEG) 6000, HEPES pH 7.5, glycerine and sodium azide (NaN(3)). The crystals belong to the monoclinic space group P2(1), with unit-cell parameters a = 74.18, b = 62.60, c = 101.91 A, beta = 90 degrees. X-ray data from these crystals diffracted at the European Synchrotron Radiation Facility (ESRF, France) ID14-1 beamline to 1.4 A.

The suitability of ion-selective electrodes (ISE) for the determination of residence time distribution (RTD) in turbid, cell-containing fluids was examined. The electrodes were found to give reproducible signals in biomass-containing feedstock with up to 20% wet weight of solids. The enhanced feedstock compatibility of IES, when compared to other tracer sensing devices, allows the study of expanded bed system hydrodynamics under relevant operating conditions. Within the linear range of the corresponding ISE-tracer pair, both examined ISE (Li(+)- or Br(-)-selective) showed to be insensitive against the range of flow rate and pH normally employed during expanded bed adsorption (EBA) of proteins. Analyzing the RTD obtained after a perfect ion tracer pulse in terms of the PDE model (PDE, axially dispersed plug-flow exchanging mass with stagnant zones) gave a quantitative description of the underlying hydrodynamic situation during EBA processing. These data provided a powerful tool to make predictions on the adsorptive global process performance with a defined feedstock type and composition. The link between the hydrodynamic events during feedstock application and the actual process performance was shown when applying intact yeast cell suspensions at different biomass content (up to 7.5% wet weight) and buffer conductivity (5-12 mS) onto an EBA column filled with the adsorbent Streamline Q XL as fluidized phase. On the basis of our experimental results, a guideline for the successful application of the ISE/RTD method to EBA process design is presented.

Formate dehydrogenase (FDH) is an enzyme of industrial interest, which is recombinantly expressed as an intracellular protein in Escherichia coli. In order to establish an efficient and reliable purification protocol, an expanded bed adsorption (EBA) process was developed, starting from the crude bacterial homogenate. EBA process design was performed with the goal of finding operating conditions which, on one hand, allow efficient adsorption of the target protein and which, on the other hand, support the formation of a perfectly classified fluidised bed (expanded bed) in the crude feed solution. A pseudo-affinity ligand (Procion Red HE3B) was used to bind the FDH with high selectivity and reasonable capacity (maximum equilibrium capacity of 30 U/ml). Additionally, a simplified modelling approach, involving small packed beds for generation of process parameters, was employed for defining the operating conditions during sample application. In combination with extended elution studies, a process was set up, which could be scaled up to 7.5 l of adsorbent volume yielding a total amount of 100,000 U of 94% pure FDH per run. On this scale, 19 l of a benzonase-treated E. coli homogenate of 15% wet-weight (pH 7.5, 9 mS/cm conductivity) were loaded to the pseudo-affinity adsorbent (0.25 m sed. bed height, 5 x 10(-4) m/s fluid velocity). After a series of two wash steps, a particle-free eluate pool was obtained with 85% yield of FDH. This excellently demonstrates the suitability of expanded bed adsorption for efficient isolation of proteins by combining solid-liquid separation with adsorptive purification in a single unit operation.

Peptide amidase-catalysed amidations of the C- terminal carboxylic group of peptides were studied using model substrates of a large series of N(alpha)-protected di-, tri-, tetra- and penta-peptides in the presence of NH4HCO3 as the ammonium source. The maximal yields of amide syntheses were achieved in a medium consisting of acetonitrile with 20-25 vol% of dimethylformamide and 3 vol% of water. Under these conditions, the substrate specificity of the enzyme was more restricted in the synthetic reaction than was found for the amide hydrolysis. Elongation of the peptide chain had a negative effect on enzymic amidation. Thus the direct amidation of N(alpha)-t-butoxycarbonyl-protected Leu-enkephalin resulted in a low yield of protected enkephalin amide.

Fungal hydrophobins are a group of surface active, self-assembling proteins. The filamentous fungus Trichoderma reesei produces two (class II) hydrophobins, HFBI and HFBII. We have studied how these water-soluble hydrophobins behave in two-phase systems using a series of nonionic surfactants with different characteristics. It was found that both hydrophobins, but especially HFBI, had a very high affinity for the surfactants. The highest partitioning coefficient, over 2500, was observed for HFBI with C(11)EO(2). Reducing the disulfides in the protein resulted in a complete loss of affinity for the surfactant, which demonstrates that the interaction is dependent on the disulfide-stabilized conformation. The hydrophobins could be efficiently extracted back from the surfactant phase by addition of alcohols such as isobutanol. Effects of the type of surfactant, temperature, pH, and ionic strength were investigated. The use of this method for purifying the proteins from crude fungal culture supernatants is demonstrated and implications of the protein-polymer interaction are discussed.

The purification of human chymotrypsinogen B (hCTRB) after expression and secretion by the yeast Pichia pastoris is described based on two different approaches using integrated initial recovery. Extraction employing aqueous two-phase systems (ATPS) from poly(ethylene glycol) and sodium sulfate allows direct processing of cell containing yeast suspensions of 50% wet weight. The target protein is obtained partially purified in the top phase while cells and cell debris are partitioned to the bottom phase of the system. hCTRB is further purified by adsorption from the top phase to the cation exchanger SP Sepharose Big Beads and elution in a salt step. The single step isolation of hCTRB is possible by expanded bed adsorption (EBA) using a fluidized cation exchanger (Streamline SP XL). A design strategy is shown taking both target protein binding and stable fluidization of the stationary phase in cell containing suspensions into consideration. For the example of hCTRB isolation from cell containing P. pastoris suspensions, a successful use of this strategy is demonstrated. Both initial recovery strategies deliver a product that can be further purified and formulated by ultrafiltration/diafiltration followed by lyophilization, resulting in a homogeneous product. Scale-up to 30-90 L of culture suspension was shown for both methods, resulting in a product of similar quality. Comparing both strategies reveals that the two-step ATPS route is better suited for high cell density cultures, while the single step EBA method is preferred for cultures of moderate cell density. This is due to the fact that application of EBA is restricted to suspensions of 10-12.5% wet weight cell concentration, thus necessitating dilution of the original broth prior to sample application. The data presented show that integrated recovery operations are a valuable alternative to traditional processing for systems that are problematic during initial solid-liquid separation.

Thirty-one different actinomycete strains were used in a genetic screening using PCR and Southern hybridization methods to detect N-acetylamino acid racemases (AAR) in order to obtain enzymes with different properties. Cloning and sequencing of a 2.5 kb EcoRI DNA fragment from Amycolatopsis orientalis subsp. lurida revealed the coding gene of an N-acetylamino acid racemase, which had identities to the aar gene of Amycolatopsis sp. TS-1-60 [Tokuyama and Hatano (1995) Appl Microbiol Biotechnol 42:884-889] of 86% at the level of DNA, and 90% at the level of amino acids. The heterologous overexpression in Escherichia coli resulted in a specific activity of about 0.2 U/mg of this racemase. A two-step purification with heat treatment followed by anion-exchange chromatography led to almost homogeneous enzyme. The optimum pH of the enzyme was 8.0 and it was stable at 50 degrees C for 30 min. The relative molecular mass of the native enzyme and the subunit was calculated to be 300 kDa and 40 kDa by gel filtration and SDS-PAGE, respectively. The isoelectric point (pI) of the AAR was 4.4. It catalyzed the racemization of optically active N-acetylamino acids such as N-acetyl-L- or -D-methionine and N-acetyl-L-phenylalanine. Further characterization of the racemase demonstrated a requirement for divalent metal ions (Co2+, Mn2+, Mg2+) for activity and inhibition by EDTA and p-hydroxymercuribenzoic acid. AAR is sensitive to substrate inhibition at concentrations exceeding 200 mM.

The physical behavior of the binary phase systems of the non-ionic polyoxyethylene detergent Agrimul NRE 1205 and water was investigated. This technical detergent can be used for the large-scale recovery of biomolecules in detergent based aqueous two-phase systems. The phase diagram was determined. It shows significant and unexpected differences to highly purified detergents. Very similar to neat detergents the phase diagram can be influenced by auxiliary chemicals thus shifting the entire phase diagram in general to lower temperatures. This was demonstrated by lowering the cloud-point by various additions. The concentration factor, as an important parameter of a first capture step in purification was investigated and modeled. Auxiliary chemicals, temperature change and change in detergent concentration also influence the viscosity and density of the phases. These experimental data are shown. They can help to explain the separation behavior of proteins. In large-scale separations aqueous two-phase systems are separated using disc-stack centrifuges. It is demonstrated that this is not a feasible method for detergent-based aqueous two-phase extraction and the physical reason is presented.

Expanded bed adsorption (EBA) is an integrated technology for the primary recovery of proteins from crude feedstock. Interactions between solid matter in the feed suspension and fluidised adsorbent particles influence bed stability and therefore have a significant impact on protein adsorption in expanded beds. In order to design efficient and reliable EBA processes a strategy is needed, which allows to find operating conditions, where these adverse events do not take place. In this paper a methodological approach is presented, which allows systematic characterisation and minimisation of cell/adsorbent interactions with as little experimental effort as possible. Adsorption of BSA to the anion exchanger Streamline Q XL from a suspension containing S. cerevisiae cells was chosen as a model system with a strong affinity of the biomass towards the stationary phase. Finite bath biomass adsorption experiments were developed as an initial screening method to estimate a potential interference. The adhesiveness of S. cerevisiae to the anion exchanger could be reduced significantly by increasing the conductivity of the feedstock. A biomass pulse response method was used to find optimal operation conditions showing no cell/adsorbent interactions. A good correlation was found between the finite bath test and the pulse experiment for a variety of suspensions (intact yeast cells, E. coli homogenate and hybridoma cells) and adsorbents (Streamline Q XL, DEAE and SP), which allows to predict cell/adsorbent interactions in expanded beds just from finite bath adsorption tests. Under the optimised operating conditions obtained using the prior methods, the stability of the expanded bed was investigated during fluidisation in biomass containing feedstock (up to 15% yeast on wet weight basis) employing residence time distribution analysis and evaluation by an advanced model. Based on these studies threshold values were defined for the individual experiments, which have to be achieved in order to obtain an efficient EBA process. Breakthrough experiments were conducted to characterise the efficiency of BSA adsorption from S. cerevisiae suspensions in EBA mode under varying operating conditions. This allowed to correlate the stability of the expanded bed with its sorption efficiency and therefore could be used to verify the threshold values defined. The approach presented in this work provides a fast and simple way to minimise cell/adsorbent interactions and to define a window of operation for protein purification using EBA.

The parameters important for an optimisation of cloud point extraction in technical scale were investigated using a genetically engineered fusion protein derived from endoglucanase I expressed in Trichoderma reesei and the nonionic polyoxyethylene Agrimul NRE 1205. The key parameters are temperature, detergent concentration, and additional salts. These parameters are interdependent, thus there is an optimum in the partition coefficient with respect to detergent concentration and a maximum for the partition coefficient and the yield with respect to temperature. These results were confirmed for the detergent C12E5 to demonstrate that these optima are due to the nature of polyoxyethylenes. Cloud point extraction was found to be only slightly affected by pH. In the case studied extraction of whole broth is favourable for a high yield and partition coefficient, since fusion protein adhering to the cells can be solubilized. However some loss of detergent which remains in the fungal biomass was observed.

Extrachromosomal elements were found in a strain of X. dendrorhous, and were characterized as linear DNA forming two well defined groups, pPh1 with 3 high-copy-number molecules, pPh11 (6.9 kb), pPh12 (5.7), pPh13 (4.7), and pPh2 with 2 low-copy-number molecules, pPh21 (3.6 kb), pPh22 (3.0). A 4077 bp fragment from pPh13 was cloned in pUC18 (pDK1) and sequenced (accession no. AJ 278,424). Seven putative ORF and some possible regulator sequences were defined.

A pH-auxostatic fed-batch process was developed for the secretory production of a fusion protein consisting of the pro-part of Staphylococcus hyicus lipase and two synthetic human calcitonin (hCT) precursor repeats under the control of a xylose-inducible promotor from Staphylococcus xylosus. Using glycerol as the energy source and pH-controlled addition of yeast extract resulted in the production of 2000 mg 1(-1) of the fusion protein (420 mg 1(-1) of the recombinant hCT precursor) within 14 h, reaching 45 g 1(-1) cell dry mass with Staphylococcus carnosus in a stirred-tank reactor. Product titer and space-time yield (30 mg calcitonin precursor 1(-1) h(-1)) were thus improved by a factor of 2, and 4.5, respectively, compared to Escherichia coli expression-secretion systems for the production of calcitonin precursors. Two hundred grams of the fusion protein was secreted by the recombinant S. carnosus on a 150-1 scale (scale-up factor of 50) with a minimum use of technical-grade yeast extract (40 mg fusion protein g(-1) yeast extract).

During recombinant E. coli fermentation with high-expression levels inclusion bodies are often formed. Aqueous two-phase systems have been successfully used in the presence of urea for the initial recovery step of inclusion bodies from E. coli. Basic studies of the complex interactions are lacking. For a systematic study of protein partitioning in the presence of urea we selected T4-lysozyme mutants with different thermal stability as a model. The stabilization of these variants by phase components was investigated measuring the fluorescence emission of tryptophan residues in the protein. Protein structure was stabilized at pH 7 in the order of S0(4)(2-) >> PEG = Dextran > H(2)O. The conformation of proteins was shown to have a strong influence on the partitioning in aqueous two-phase systems. Tryptophan and its homologuous di- and tripeptdides were partitioned in similar phase systems to normalize for contribution from hydrophobic interactions., (Copyright 2000 John Wiley & Sons, Inc.)

Mathematical strategies were applied to optimise the extraction of recombinant leucine dehydrogenase from E. coli homogenates and endoglucanase 1 from culture filtrates of Trichoderma reesei in polyethylene glycol-phosphate systems. The goal was to test mathematical tools which could facilitate the optimisation procedure in aqueous two-phase systems. A modified simplex approach, the method of steepest ascent and a genetic algorithm were successfully applied and compared. The methods differ in the height of the optimum found, the number of experiments and the time required. The genetic algorithm proved to be an optimisation procedure which can be used well in aqueous two-phase systems. The simplex procedure has to be further improved.

A method for the production of recombinant L-leucine dehydrogenase from Bacillus cereus in pilot scale is described employing the temperature induced runaway replication vector pIET98 and the Escherichia coli host strain BL21. Fed-batch cultivation using a semi-synthetic high-cell densitiy medium was adjusted in 5-L scale to yield a constant growth rate of 0,17 h(-1) and a final cell concentration of 27 g dry weight/L by exponentially increasing the nutrient supply. Runaway replication and thus, LeuDH expression was induced during the feeding phase by increasing the cultivation temperature to 41 degrees C yielding a specific enzyme activity of 110 U/mg, which corresponds to 30% of the soluble cell protein. The cultivation was terminated when the dissolved oxygen content fell below 10% saturation. The final volume activity was 600,000 U/L cultivation. No change in growth, cell density, or expression activity was observed scaling up the cultivation volume to 200 L. Thus, 120,000,000 units L-leucine dehydrogenase were obtained from one cultivation. The purification of L-leucine dehydrogenase to homogeneity was carried out by heat denaturation, liquid-liquid extraction, gel filtration, and anion-exchange chromatography to give pure enzyme in 65% yield. The integrity of the recombinant enzyme was tested measuring the molecular weight and determining the N-terminal amino acid sequence., (Copyright 2000 John Wiley & Sons, Inc.)

The established Escherichia coli expression vectors ptrc99a, pKK223-3, pPLlambda, pAsk75, pRA95, and pRA96, which differ in copy number, mode of induction, selection marker, and use of par sequences for stabilization, were investigated for the stable expression of recombinant L-leucine dehydrogenase from Bacillus cereus with a view to large-scale production. Best results were achieved with pIET98, a runaway-replication system derived from pRA96. Expression of L-leucine dehydrogenase was controlled by its constitutive B. cereus promoter and depended on host strain, cultivation temperature, induction time, and media composition. After cell cultivation at 30 degrees C and shifting to 41 degrees C to induce plasmid replication, E. coli BL21[pIET98] yielded 200 U LeuDH/mg protein, which corresponds to >50% of the soluble cell protein. Continuous cultivation in a semisynthetic high-cell-density medium verified structural and segregational stability over 100 generations in the absence of a selection pressure.

Expanded bed adsorption (EBA) is a primary recovery operation allowing the adsorption of proteins directly from unclarified feedstock, e.g. culture suspensions, homogenates or crude extracts. Thus solid-liquid separation is combined with adsorptive purification in a single step. The concept of integration requires that the solid components of the feed solution are regarded as a part of the process, which influences stability, reproducibility, and overall performance. This aspect is investigated here at the example of the influence of presence and concentration of intact yeast cells (S. cerevisiae) on the adsorption of model proteins (hen egg white lysozyme and bovine serum albumin) to various stationary phases (cation and anion-exchange, hydrophobic interaction, immobilised metal affinity). The interaction of the cells with the adsorbents is determined qualitatively and quantitatively by a pulse response method as well as by a finite bath technique under different operating conditions. The consequence of these interactions for the stability of expanded beds in suspensions of varying cell concentration is measured by residence time distributions (RTDs) after tracer pulse injection (NaBr, LiCl). Analysis of the measured RTD by the PDE model allows the calculation of the fraction of perfectly fluidised bed (phi), a parameter which may be regarded as a critical quantity for the estimation of the quality of fluidisation of adsorbents in cell containing suspensions. The correlation between bed stability and performance is made by analysing the breakthrough of model proteins during adsorption from unclarified yeast culture broth. A clear relationship is found between the degree of cell/adsorbent interaction, bed stability in terms of the phi parameter, and the sorption efficiency. Only beds characterised by a phi value larger than 0.8 in the presence of cells will show a conserved performance compared to adsorption from cell free solutions. A drop in phi, which is due to interactions of the fluidised adsorbent particles with cells from the feed, will directly result in a reduced breakthrough efficiency. The data presented highlight the importance of including the potential interaction of solid feedstock components and the expanded adsorbents into the design of EBA processes, as the interrelation found here is a key factor for the overall performance of EBA as a truly integrated operation.

Multienzyme reaction systems with simultaneous coenzyme regeneration have been investigated in a continuously operated membrane reactor at bench scale. NAD(H) covalently bound to polyethylene glycol with a molecular weight of 10(4) [PEG-10,000-NAD(H)] was used as coenzyme. It could be retained in the membrane reactor together with the enzymes. L-leucine dehydrogenase (LEUDH) was used as catalyst for the reductive amination of alpha-ketoisocaproate (2-oxo-4-methylpentanoic acid) to L-leucine. Formate dehydrogenase (FDH) was used for the regeneration of NADH. Kinetic experiments were carried out to obtain data which could be used in a kinetic model in order to predict the performance of an enzyme membrane reactor for the continuous production of L-leucine. The kinetic constants V(max) and k(m) of the enzymes are all in the same range regardless of whether native NAD(H) or PEG-10,000-NAD(H) is used as coenzyme. L-leucine was produced continuously out of alpha-ketoisocaproate for 48 days; a maximal conversion of 99.7% was reached. The space-time yield was 324 mmol/L day (or 42.5 g/L day).

The gene of the NAD-dependent formate dehydrogenase (FDH) from the yeast Candida boidinii was cloned by PCR using genomic DNA as a template. Expression of the gene in Escherichia coli yielded functional FDH with about 20% of the soluble cell protein. To confirm the hypothesis of a thiol-coupled inactivation process, both cysteine residues in the primary structure of the enzyme have been exchanged by site-directed mutagenesis using a homology model based on the 3D structure of FDH from Pseudomonas sp. 101 and from related dehydrogenases. Compared to the wt enzyme, most of the mutants were significantly more stable towards oxidative stress in the presence of Cu(II) ions, whereas the temperature optima and kinetic constants of the enzymatic reaction are not significantly altered by the mutations. Determination of the Tm values revealed that the stability at temperatures above 50 degrees C is optimal for the native and the recombinant wt enzyme (Tm 57 degrees C), whereas the Tm values of the mutant enzymes vary in the range 44-52 degrees C. Best results in initial tests concerning the application of the enzyme for regeneration of NADH in biotransformation of trimethyl pyruvate to Ltert leucine were obtained with two mutants, FDHC23S and FDHC23S/C262A, which are significantly more stable than the wt enzyme.

Detergent-based aqueous two-phase systems have the advantage to require only one auxiliary chemical to induce phase separation above the cloud point. In a systematic study the efficiency of tryptophan-rich peptide tags was investigated to enhance the partitioning of an enzyme to the detergent-rich phase using cutinase as an example. Up to 90% enzyme activity could be extracted in a single step from whole broth of recombinant Saccharomyces cerevisiae expressing cutinase variants carrying a (WP)4 tag. In contrast, the extraction yield of wild type cutinase was 2-3% only. The detergent concentration and the temperature are the main parameters to optimize the extraction yield. Considering availability, extraction yields, and price the detergent Agrimul NRE 1205 served best for enzyme recovery.

The use of confocal scanning laser microscopy (CSLM) has recently been described for the visualization of intraparticle protein profiles during single-protein finite bath uptake experiments. By coupling of fluorescent molecules to proteins the penetration of porous media by labeled macromolecules could be detected by scanning single adsorbent particles for fluorescence emission after laser excitation. Thus the internal protein distribution profile, which is a central element in modeling of protein transport in porous adsorbents, became experimentally accessible. Results from the simultaneous visualization of two proteins by this technology are shown here. The use of two different fluorescent dyes for protein labeling and two independent detectors in the CSLM allowed for the first time ever the direct observation of a two-component diffusion process within a porous stationary phase. The finite bath uptake of human immunoglobulin G (hIgG) and bovine serum albumin (BSA) to two different ion exchange adsorbents (SP Sepharose Fast Flow and Source 30S) and to an affinity adsorbent (Protein A Sepharose) was measured using Cy5 and Oregon Green as labels. Single adsorbent particles were scanned for intensity distribution of fluorescence emission from the two fluorophors. The intraparticle profiles obtained from the confocal images were translated into a relative protein concentration thus allowing the calculation of protein uptake kinetics from direct measurement in the stationary phase. The confocal technique may prove to be a very powerful means of data generation for modeling of multi-component mass transfer phenomena in protein adsorption., (Copyright 1999 John Wiley & Sons, Inc.)