Bacterial fermentation of recombinant major wasp allergen Antigen 5 using oxygen limiting growth conditions improves yield and quality of inclusion bodies

A process for bacterial expression and purification of the recombinant major wasp allergen Antigen 5 (Ves v 5) was developed to produce protein for diagnostic and therapeutic applications for type 1 allergic diseases. Special attention was focused on medium selection, fermentation conditions, and efficient refolding procedures. A soy based medium was used for fermentation to avoid peptone from animal origin. Animal-derived peptone required the use of isopropyl-beta-D-thiogalactopyranoside (IPTG) for the induction of expression. In the case of soy peptone, a constitutive expression was observed, suggesting the presence of a component that mimics IPTG. Batch cultivation at reduced stirrer speed caused a reduced biomass due to oxygen limitation. However, subsequent purification and processing of inclusion bodies yielded significantly higher amount of product. Furthermore, the protein composition of the inclusion bodies differed. Inclusion bodies were denatured and subjected to diafiltration. Detailed monitoring of diafiltration enabled the determination of the transition point. Final purification was conducted using cation-exchange and size-exclusion chromatography. Purified recombinant Ves v 5 was analyzed by RP-HPLC, CD-spectroscopy, SDS-PAGE, and quantification ELISA. Up to 15 mg highly purified Ves v 5 per litre bioreactor volume were obtained, with endotoxin concentrations less than 20 EU mg(-1) protein and high comparability to the natural counterpart. Analytical results confirm the suitability of the recombinant protein for diagnostic and clinical applications. The results clearly demonstrate that not only biomass, but especially growth conditions play a key role in the production of recombinant Ves v 5. This has an influence on inclusion body formation, which in turn influences the renaturation rate and absolute product yield. This might also be true for other recombinant proteins that accumulate as inclusion bodies in Escherichia coli.

CFD-aided design of a dynamic filter for mammalian cell separation

In the present work, a rotating disk filter was designed for mammalian cell separation with the aim of avoiding both cell damage and membrane fouling. Different geometric and operational variables of the rotating disk filter were studied using computational fluid dynamics (CFD) by varying rotor radius, rotor angle, membrane-rotor distance, and angular velocity. The combinations of these variables followed a statistical design, so that an analysis of the CFD results provided correlations describing the average shear stress on the membrane surface and the maximum shear stress in the whole module as a function of the variables studied. Based on these correlations, and on the shear resistance levels of Chinese hamster ovary (CHO) and baby hamster kidney (BHK) cell lines, which were investigated using a cone-and-plate viscosimeter, it was possible to determine the geometry and angular velocity that would minimize both cell damage and membrane fouling. After construction, the filter was tested in filtration experiments at increasing permeate fluxes. Cell viability remained >90% for the duration of the experiments (2.5 h), and no indication of fouling was observed. It was shown that the designed dynamic filter is able to effectively avoid both cell damage and membrane fouling, and thus can be used for mammalian cell harvesting and perfusion.

On the molecular interaction between lactoferrin and the dye Red HE-3b. A novel approach for docking a charged and highly flexible molecule to protein surfaces

Lactoferrin (Lf) is a non-heme, iron binding protein present in many physiological fluids of vertebrates where its main role is the microbicidal activity. It has been isolated by different methods, including dye-affinity chromatography. Red HE-3B is one of the most common triazinic dyes applied in protein purification, but scant knowledge is available on structural details and on the energetics of its interaction with proteins. In this work we present a computational approach useful for identifying possible binding sites for Red HE-3B in apo and holo forms of Lfs from human and bovine source. A new geometrical description of Red HE-3B is introduced which greatly simplifies the conformational analysis. This approach proved to be of particular advantage for addressing conformational ensembles of highly flexible molecules. Predictions from this analysis were correlated with experimentally observed dye-binding sites, as mapped by protection from proteolysis in Red HE-3B/Lf complexes. This method could bear relevance for the screening of possible dye-binding sites in proteins whose structure is known and as a potential tool for the design of engineered protein variants which could be purified by dye-affinity chromatography.

An integrated process for mammalian cell perfusion cultivation and product purification using a dynamic filter

In the present work, a dynamic filter was employed to develop an integrated perfusion/purification process. A recombinant CHO cell line producing a human anti-HIV IgG was employed in the experiments. In the first part of this work, the dynamic filter was fitted with conventional microfiltration membranes and tested as a new external cell retention device for perfusion cultivations. The filter was connected to a running perfusion bioreactor and operated for approximately 400 h at an average cell concentration of 10 million cells mL(-)(1), whereby cell viability remained above 90% and no problems of sterility were experienced. In the second part of this work, the dynamic filter was employed to simultaneously carry out cell separation and product purification, using membrane adsorbers containing Protein A affinity ligands. An automated system was built, which integrated the features of an automated perfusion bioreactor and of a liquid chromatography system. The IgG was continuously adsorbed onto the affinity membranes and was periodically recovered through elution cycles. After connection of the filter, the system was operated for approximately 300 h, whereby three elution cycles were carried out. No progressive increase in transmembrane pressure was observed, indicating no membrane fouling problems, and the IgG was recovered practically free of contaminants in a 14-fold concentrated form, indicating that the integrated, one-step perfusion/purification process developed during this work is a promising alternative for the production of biologicals derived from mammalian cells.

Endotoxin removal by affinity sorbents

Affinity sorbents and detoxification strategies are described to remove different amounts of endotoxin. Advantages and disadvantages of the employed ligands are discussed and it is shown that both electrostatic and hydrophobic interactions contribute to the association of ligands and endotoxins. Furthermore, the flexibility of the ligand is more important than an exact structural match between ligand and ligate. Owing to the formation of endotoxin micelles and vesicles, microfiltration membrane adsorbers are particularly effective since mass transfer restrictions are almost absent in the flow-through pores.

Selective adsorption of endotoxin inside a polycationic network of flat-sheet microfiltration membranes

For the removal of remaining amounts of endotoxin, sorbents with high selectivity for endotoxin are required. Typically, particulate sorbents with positively charged ligands, such as histidine, polymyxin B poly-L-lysine and poly(ethyleneimine) (PEI), display moderate to high removal efficiencies in an environment of low ionic strength. It was found that polycationic ligands are most suitable to meet an endotoxin concentration which is below the threshold level required for parenteralia. Furthermore, protein recoveries close to 100% are obtained if the decontamination is performed at a pH close to the pI of acidic proteins. The high selectivity is probably caused by complexation of the polycationic ligand with the polyanionic endotoxin, leading to interactions with KD < 10(-9) M using PEI and assuming M(r) = 10 kDa for monomeric endotoxin; with BSA the same ligand reveals only KD = 4 x 10(-6) M. Using polymer-coated microfiltration membranes, immobilization of positively charged ligands leads to membrane adsorbers which are generally superior to chromatographic adsorbers and allow faster processing. Since immobilization takes place at polymer chains, low-molecular-weight ligands mainly add positive charges to the hydrophilic polymer. Consequently, membrane adsorbers with low-molecular-weight ligands, even DEAE, demonstrate similar selectivity to PEI or poly-L-lysine.

Endotoxin removal from protein solutions

Endotoxins liberated by gram-negative bacteria are frequent contaminations of protein solutions derived from bioprocesses. Because of their high toxicity in vivo and in vitro, their removal is essential for a safe parenteral administration. A general method for the removal of endotoxins from protein solutions is not available. Methods used for decontamination of water, such as ultrafiltration, have little effect on endotoxin levels in protein solutions. Various techniques described in the patent literature are not broadly applicable, as they are tailored to meet specific product requirements. Besides ion-exchangers and two-phase extraction, affinity techniques are applied with varying success. Also, taylor-made endotoxin-selective adsorber matrices for the prevention of endotoxin contamination and endotoxin removal are discussed for this purpose. After giving an overview of the properties of endotoxins and the significance of endotoxin contamination, this review intends to provide an overall picture of the various methods employed for their removal. Avenues are pointed out how to optimise a method with regard to the specific properties of endotoxins in aqueous solution.

Preparative two-step purification of recombinant human basic fibroblast growth factor from high-cell-density cultivation of Escherichia coli

Aggregation and precipitation are major pitfalls during bioprocessing and purification of recombinant human basic fibroblast growth factor (rh-bFGF). In order to gain high yields of the soluble protein monomer with high biological activity, an efficient downstream process was developed, focussing on the combination of expanded bed adsorption (EBA) and heparin chromatography. After expression in E. coli TG1:plambdaFGFB, cells were harvested and washed; then the rh-bFGF was released via high pressure homogenization. The high viscosity of the feedstock of about 40 mPa s, showing non-newtonian behaviour, was reduced to 2 mPa s by the addition of DNase. The homogenate (5.6 l) was loaded directly on an expanded bed column (C-50) packed with the strong cation-exchanger Streamline SP. In the eluates, histone-like (HU) protein was identified as the main protein contaminant by sequence analysis. The thermodynamics and kinetics of rh-bFGF adsorption from the whole broth protein mixture were determined in view of competition and displacement effects with host-derived proteins. Optimal binding and elution conditions were developed with knowledge of the dependence of rh-bFGF adsorption isotherms on the salt concentration to allow direct application of eluates onto Heparin HyperD. This affinity support maintained selectivity and efficiency under CIP and over a wide range of flow-rates; both is advantageous for the flexibility of the purification protocol in view of a scalable process. Remaining DNA and HU protein were separated by Heparin HyperD. The endotoxin level decreased from approximately 1,000,000 EU/ml in the whole broth to 10 EU in 3 mg bFGF per ml. The final purification protocol yields >99% pure rh-bFGF as judged from SDS-PAGE and MALDI-TOF mass spectrometry with high mitogenic activity (ED50=1-1.5 ng/ml) of the lyophilized sample. In comparison to the conventional process, the overall protein recovery rose by 15% to 65% with saving time and costs.

Expanded-bed chromatography in primary protein purification

Chromatography in stable expanded beds enables proteins to be recovered directly from cultivations of microorganisms or cells and preparations of disrupted cells, without the need for prior removal of suspended solids. The general performance of an expanded bed is comparable to a packed bed owing to reduced mixing of the adsorbent particles in the column. However, optimal operating conditions are more restricted than in a packed bed due to the dependence of bed expansion on the size and density of the adsorbent particles as well as the viscosity and density of the feedstock. The feedstock composition may become the most limiting restriction owing to interactions of adsorbent particles with cell surfaces, DNA and other substances, leading to their aggregation and consequently to bed instabilities and channeling. Despite these difficulties, expanded-bed chromatography has found widespread applications in the large scale purification of proteins from mammalian cell and microbial feedstocks in industrial bioprocessing. The basics and implementation of expanded-bed chromatography, its advantages as well as problems encountered in the use of this technique for the direct extraction of proteins from unclarified feedstocks are addressed.

Continuous-bed chromatography for the analysis and purification of recombinant human basic fibroblast growth factor

The chromatographic properties of the commercial cation exchanger UNO-S1 (35x7 mm) was investigated using lysozyme from hen egg white as model protein and recombinant human basic fibroblast growth factor (rh-bFGF) from a high cell density cultivation of E. coli. The dynamic capacity for lysozyme (c(o) = 1 mg/ml) in 100 mM acetate buffer, pH 5 was 27 mg per ml sorbent. It was found independent of the flow-rate from 78 to 935 cm/h owing to the absence of mass transfer restrictions with this column concept. Regarding the selectivity for rh-bFGF and the capacity for lysozyme, no changes were apparent after cleaning-in-place (CIP) procedures with 0.5 M NaOH. Clogging of the column by a clarified crude cell homogenate of E. coli was not critical as precipitates were removed by reversal of the flow during CIP. Rh-bFGF elutes in three consequent peaks from the UNO-S1 column, which could be attributed to soluble rh-bFGF aggregates of different size. The dynamics of rh-bFGF aggregation and reaggregation in the crude feedstock was monitored by fast gradient elution chromatography.

Proteinase K digestion of proteins improves detection of bacterial endotoxins by the Limulus amebocyte lysate assay: application for endotoxin removal from cationic proteins

Cationic proteins, such as lysozyme, ribonuclease A, and human IgG, impaired the detection of endotoxins with the Limulus amebocyte lysate assay (LAL assay) through formation of endotoxin-protein complexes, demonstrating pronounced masking of endotoxins. Methods, such as phenol extraction, dilution heating, and perchloric acid treatment failed to demask the endotoxins. Also, digestion with trypsin, chymotrypsin, or pronase recovered only 10 to 20% of the applied endotoxins. However, endotoxin recoveries up to 100% were obtained with proteinase K digestion of the samples prior to the LAL assay. This method was then applied to examine the impact of endotoxin masking on endotoxin removal from protein solutions by selective adsorption on membrane adsorbers. It was found that poly-L-lysine and poly(ethyleneimine) as endotoxin-selective ligands were able to pull endotoxins off the proteins studied, thereby guaranteeing successful decontamination.

Endotoxin removal with poly(ethyleneimine)-immobilized adsorbers: Sepharose 4B versus flat sheet and hollow fibre membranes

Poly(ethyleneimine) was immobilized on poly(vinyl alcohol)-coated nylon flat sheet membranes, poly(vinyl alcohol) and poly(ethylenevinyl alcohol) hollow fibre membranes as well as Sepharose 4B. The resulting poly(ethyleneimine)-immobilized adsorbers were used for removal of E. coli derived endotoxin from buffers and bovine serum albumin solutions. The efficiency of poly(ethyleneimine) proved to be constant over a wide pH range, including phosphate buffered saline. The performance depended upon the matrix type employed: endotoxin clearance factors varied from 100 to 120,000 in protein-free solutions and 40 to 33,000 in solutions of bovine serum albumin using 6000 EU/ml as feed concentration. The best adsorber was the flat sheet membrane-immobilized poly(ethyleneimine), followed by the hollow fibre-immobilized poly(ethyleneimine) and poly(ethyleneimine)-Sepharose. The factors influencing endotoxin clearance were the mass transport (convective systems were superior to the diffusive system), the chemical composition and the surface structure of the underlying matrix.

Nylon-Based Affinity Membranes: Impacts of Surface Modification on Protein Adsorption

Nylon microfiltration membranes were activated with bisoxirane and formaldehyde at terminal amino groups and amide groups of the nylon polymer, respectively. Dextrans were covalently immobilized on these activated membranes to yield dextran-coated membrane matrices. Both procedures led to a significant reduction of hemoglobin adsorption; however, bisoxirane activation required additional cross-linking of dextran and a second dextran layer to yield comparable quality of dextran-coated membranes than formaldehyde activation. Formaldehyde activation was easiest and cheapest and resulted in membranes with highest dextran density and relatively lowest nonspecific hemoglobin adsorption. Dextrans of &Mmacr;w >/= 40,000 were required for bisoxirane-activated membranes, whereas dextrans of &Mmacr;w = 6000 were sufficient for formaldehyde-activated membranes. Both activation methods resulted in stable coatings at low and high pH; however, formaldehyde-activated membranes were unstable under strongly acidic conditions at pH < 3. Dextran coils were found responsible for the reduction of the hydraulic permeability but also for the high ligand densities obtained after immobilization of Cibacron Blue F3G-A (360 nmol/cm2) and iminodiacetic acid (400 nmol/cm2). The thermodynamics of protein adsorption on dye ligand affinity (DLA) membranes corresponded with chromatographic sorbents and dye ligand conjugates, with the dextran coating demonstrating similar structure than dextrans in solution. Protein adsorption took place in the extended coil structure of dextrans with binding capacities up to 730 µg/cm2 lysozyme on DLA membranes and 470 µg/cm2 concanavalin A on metal chelate affinity membranes. Copyright 1997 Academic Press.

Membrane adsorbers for selective removal of bacterial endotoxin

Surface-modified flat-sheet microfiltration membranes were functionalised with poly-L-lysine, polymyxin B, poly(ethyleneimine), L-histidine, histamine, alpha-amylase and DEAE as well as deoxycholate. Their suitability to remove endotoxin from both buffers and protein solutions was examined using bovine serum albumin, murine IgG1 and lysozyme as model proteins. In protein-free solutions reduction from 6000 EU/ml to <0.1 EU/ml was achieved with all applied ligands; only alpha-amylase as well as L-histidine and histamine, when immobilized via the non-ionic spacer bisoxirane, exhibited low clearance factors at neutral pH. The adsorption of endotoxin is mainly ruled by electrostatic interaction forces. Thus in multi-component systems, such as endotoxin-contaminated protein solutions, competing interactions take place: acidic proteins compete with endotoxin for binding sites at the membrane adsorbers, basic proteins compete with the ligands for endotoxin and act as endotoxin carriers. With properly chosen conditions the membrane adsorbers presented here show exceptional effectiveness also in the presence of proteins. They are generally superior to functionalised Sepharose chromatographic sorbents and allow fast processing. They may contribute to reduce the risks in the application of parenterals and diagnostics.

Purification of (His)6EcoRV [recombinant restriction endonuclease EcoRV fused to a (His)6 affinity domain] by metal-chelate affinity chromatography

The chromatographic purification of (His)6EcoRV, a fusion protein consisting of a hexahistidine affinity domain and restriction endonuclease EcoRV produced from recombinant Escherichia coli, led to high product concentrations (> or = 1 mg/ml) in the preparative mode. Increasing the amount of applied crude cell homogenate caused competition with host-specific proteins was achieved by pre-adsorption on to a DEAE anion-exchange sorbent. This, in combination with 0.5-1 M NaCl in the adsorption buffer, assured a purity > 95% and a total protein recovery of approximately 34% in the preparative mode. Contamination of the product with about 2 mol of Ni(11)/mol of (His)6EcoRV was found due to metal-ion transfer to the N-terminal high-affinity binding site at (His)6. Tris(carboxymethyl)ethylenediamine (TED)-Sepharose was employed as an Ni(11) adsorber. One passage of Ni(11)-contaminated protein solutions through the TED-Sepharose column resulted in a decrease in the Ni(11) content in the (His)6EcoRV fractions below the detection limit (approximately 0.02 mg/l) of the atomic-adsorption spectrophotometer.

A study of combined filtration and adsorption on nylon-based dye-affinity membranes: separation of recombinant L-alanine dehydrogenase from crude fermentation broth

Dextran, hydroxyethylcellulose (HEC), and poly(vinyl alcohol) PVA were covalently linked to bisoxirane-activated nylon membranes. Cibacron Blue F3G-A was immobilized on to these membranes to yield a dye-affinity membrane. The hydrodynamic permeability of affinity membranes was reduced to approximately 50% of that of the original Nylon membrane due to extension of polymer coils into flow-through pores. Adsorption of pre-purified human serum albumin (HSA) and malate dehydrogenase (MDH) displayed highest maximum binding capacities on HEC-coated dye-ligand-affinity membranes, ranging from (163 micrograms/cm2 for HSA to 316 micrograms/cm2 for MDH. The protein recovery of HSA was 100% on dextran-coated membranes compared with 70% on PVA-coated membranes, whereas almost 100% recovery was found for MDH, independent of the polymer. Application of crude supernatant from recombinant Escherichia coli yielded purification factors of 7.4, 8.9 and 11.2 for recombinant alanine dehydrogenase from Mycobacterium tuberculosis for HEC-, dextran- and PVA-coated membranes respectively. Dynamic capacities decreased remarkably to approximately 3 micrograms/cm2 due to co-adsorption of host proteins. The presence of cell debris caused only a slight decrease of purification factors, but a dramatic decrease of the permeability of affinity membranes due to development of a particle layer in front of the membranes. Although enzyme recoveries were up to 90% using cell-free supernatant, more than 50% of the product was lost due to polarization, concentration and rejection at particle layers when using crude homogenates. In order to further improve this integrated downstream process, sophisticated membrane techniques are required by which the formation of a filter cake is circumvented. Further refinement of polymer-coated membranes would not help one to avoid this problem.

Strategies for the depyrogenation of contaminated immunoglobulin G solutions by histidine-immobilized hollow fiber membrane

The depyrogenation of different IgG solutions using the histidine-linked hollow fiber membrane developed in our laboratory is presented here. Three strategies for endotoxin (ET) removal were investigated according to the immobilized histidine's ability to bind different immunoglobulins: (1) ET removal from 1 mg/ml non histidine-binding mouse monoclonal IgG1 (MabCD4) solution was achieved in the presence of acetate buffer (pH 5.0) without any protein loss. (2) For contaminated human IgG, combined adsorption of ET and IgG in the presence of MOPS of Tris buffer was tested, followed by differential elution using increasing salt concentrations. This attempt was not successful since ET were quantitatively found in the IgG elution fraction. (3) Alternatively, it was proposed to adsorb selectively ET in the presence of acetate buffer (pH 5.0) under non binding conditions for human IgG. Human IgG could then be purified if necessary with the same membrane in the presence of MOPS buffer (pH 6.5). With a 1 m2 histidine-PEVA module under these operating conditions, it is estimated that the depyrogenation of 3 l of 1 mg/ml IgG (human or murine) solution containing 80 EU/ml of ET should be possible.

Purification of murine IgG1 on group specific affinity sorbents

Murine IgG1 a CD4-antibody, was produced from the cell line MAX 16 H 5. DEAE-Sepharose CL-6B and group specific sorbents based on different ligands, such as protein A-Sepharose, AvidChrom, Thiophilic Resin (T-Resin) Fractogel EMD TA 650S, Sepharose 4B-immobilized metal chelate and histidine sorbents, were compared in terms of their suitability for the large-scale purification of this IgG. Evaluation criteria were selectivity for IgG1 protein capacity and recovery, stability under cleaning-in-place (CIP) conditions and the rate of IgG1 adsorption. Both thiophilic gels performed best under these conditions, with Fractogel EMD TA 650S demonstrating slightly higher contamination with other proteins compared to the T-Resin based on 6% beaded Agarose. Higher purities were obtained with these gels compared to conventional purification techniques employing a combination of ion exchange and hydrophobic interaction chromatography. Favorable adsorption kinetics were found on DEAE-Sepharose with the effective diffusivity in the pores of the sorbent being five-times accelerated compared to the diffusion coefficient of IgG1 in solution. Multilayer adsorption of the IgG1 was experienced from adsorption isotherms on DEAE-Sepharose and Fractogel EMD TA 650S. This was linked to an early breakthrough of this protein during frontal chromatography. Co-purification of pyrogens between two-(Thiophilic Resin) and five-times (DEAE-Sepharose) of the original level was significant with all sorbents except protein A-Sepharose and AvidChrom. Thus an additional chromatographic step is required in case of pyrogen contamination of the cell culture homogenate.

Purification of recombinant human basic fibroblast growth factor: stability of selective sorbents under cleaning in place conditions

Human basic fibroblast growth factor (bFGF) was produced from recombinant Escherichia coli by high-cell-density cultivation. In order to develop a purification strategy for large-scale purification, chromatographic sorbents with different anionic functional groups were compared in terms of selectivity for bFGF and stability under cleaning in place (CIP) conditions. Heparin-Sepharose CL-6B, Fractogel EMD-SO3- 650 (S) and SP-Sepharose (high performance) were found suitable for this purpose with decreasing selectivity in that order. Each sorbent was treated eight times under CIP conditions employing both 0.2 and 1.0 M NaOH, in order to study modifications of these sorbents. Heparin-Sepharose displayed more than 50% loss of capacity after the first CIP treatment and decreasing selectivity with each cycle. Both cation exchangers displayed almost constant results regarding selectivity and capacity. The Fractogel EMD-SO3- exhibited only slightly lower selectivity for bFGF than Heparin-Sepharose and the highest capacity of all sorbents tested. Agglomeration of bFGF at low salt concentrations was a serious problem. By direct application of pooled fractions from Fractogel EMD-SO3- onto Heparin-Sepharose a highly pure product was obtained; however, the recovery after Heparin-Sepharose was only 30%.

Removal of endotoxins by affinity sorbents

Histidine, histamine and polymyxin B affinity sorbents were employed for the removal of Escherichia coli-derived endotoxins. Their effectiveness was compared with those of poly-L-lysine-Sepharose and DEAE-Sepharose. All sorbents reduced the concentration of endotoxins from an E. coli culture filtrate to tolerable levels. However, their effectiveness was not higher than that of the anion exchanger, which displayed clearance rates of up to 15,000. Endotoxin removal from protein solutions depended on the net charge of the desired protein. Lysozyme as a model for positively charged proteins enhanced endotoxin removal. In contrast, only low initial contamination levels (< 34 EU/ml) were reduced to tolerable levels from bovine serum albumin (BSA) as the negatively charged protein model owing to competition of BSA and endotoxins for adsorption sites. Hence also a low BSA recovery was observed after the treatment whereas the lysozyme recovery was almost 100%. At pH values below the isoelectric point of BSA, endotoxin removal was also more effective. The best conditions for the decontamination were found at neutral pH and low ionic strength (< or = 20 mM phosphate). Ionic forces between ligands and endotoxins are dominant at this ionic strength; hydrophobic interactions are not very effective. Hence the selectivities of all sorbents towards endotoxins are not exceptionally high. DEAE-anion exchangers are the most suitable sorbents for the removal of endotoxins from solutions accommodating positively charged proteins owing to their low cost and high capacity. Poly-L-lysine-Sepharose was most effective for the removal of small amounts of endotoxins from solutions of negatively charged proteins. The "affinity ligands" histamine, histidine and polymyxin B were effective for the removal of endotoxins from E. coli filtrate; however, their effectiveness decreased dramatically in the presence of BSA and it was lower than for poly-L-lysine- and DEAE-Sepharose in the presence of lysozyme.

Immobilized-metal-chelate regenerable carriers: (I). Adsorption and stability of penicillin G amidohydrolase from Escherichia coli

Penicillin G amidohydrolase (PGA) was immobilized on Cu(II)-chelate regenerable sorbents. A long spacer was essential for binding, such as bisoxirane in the case of Sepharose 4B or glycidoxypropyltrimethoxysilane in the case of silica-based carriers. The stability of the PGA-carrier was determined both by the interaction forces between PGA and the metal-chelate sorbent and the presence of penicillin G (Pen G). The force of interaction between the enzyme and the metal-chelate sorbent was low, and Pen G competed for binding sites at high concentrations. The carrier with a small pore size demonstrated diffusion restrictions during immobilization of PGA, resulting in low activities of the immobilized enzyme. This carrier could not be completely regenerated. Carriers with an average pore size of 55 nm or larger displayed fewer diffusion restrictions. The corresponding Cu(II)-chelate sorbents were regenerated several times.

High-performance liquid chromatography of amino acids, peptides and proteins. XCV. Thermodynamic and kinetic investigations on rigid and soft affinity gels with varying particle and pore sizes: comparison of thermodynamic parameters and the adsorption behaviour of proteins evaluated from bath and frontal analysis experiments

The thermodynamic constants, associated with the interaction of three proteins with triazine dye affinity sorbents, have been derived from bath and frontal analysis experiments. In cases where mass-transfer restrictions are very high, calculation of the thermodynamic constants directly from frontal analysis experiments could not be achieved. In such cases, a portion of the adsorbate was always present in the effluent, a situation which has its effect as the split peak phenomenon. With Fractogel-based triazine dye affinity sorbents none of the test proteins applied in frontal analysis were adsorbed. A similar behaviour was observed for a Cellufine sorbent during the adsorption of human serum albumin and the Blue Sepharose CL6B sorbent during the adsorption of alcohol dehydrogenase, which displayed much slower apparent adsorption kinetics than observed in the bath experiments. These phenomena were shown to be associated with changes in the gel structure, caused in part by the column packing procedure. Silica-based sorbents performed better in the adsorption of lysozyme in the column mode than soft-gel affinity sorbents, as was evident in the higher capacities and steeper breakthrough curves. At high protein concentrations (feedstock concentration greater than 0.2 mg/ml) breakthrough curves obtained with small- and large-particle-size sorbents, but of constant pore size, were found to be identical. This finding demonstrates that the use of small-particle-size sorbents (e.g. particle diameter, dp less than or equal to 5 microns) for the preparative isolation of proteins may not be justified when operating in the overload mode. With other higher-molecular-weight proteins and the silica-based sorbent systems examined, the small-particle-size sorbents (dp = 5 microns) displayed less symmetrical shapes of their breakthrough curves than the larger-particle-size and soft-gel sorbents. This behaviour was further exacerbated when non-porous glass or silica-based sorbents were utilized. These non-porous affinity sorbents displayed nearly rectangular breakthrough shapes at the onset of the adsorption process, but comparatively slow adsorption kinetics became evident as saturation was approached. This phenomenon has been attributed to surface rearrangement and/or reorientation of the adsorbed proteins, particularly with sorbents of high ligand densities.

High-performance liquid chromatography of amino acids, peptides and proteins. XCII Thermodynamic and kinetic investigations on rigid and soft affinity gels with varying particle and pore sizes

In these investigations Cibacron Blue F3GA was immobilized on soft gels, porous silicas, and non-porous glass beads. Hen egg white lysozyme, human serum albumin and yeast alcohol dehydrogenase were used as adsorbates with the dye-affinity sorbents. Batch experiments with continuous monitoring of protein concentration were employed to evaluate thermodynamic and kinetic behaviour of these proteins in finite bath systems. The observed adsorption kinetic rates of interaction of the above proteins with each of the dye-affinity sorbents were found to decrease with increasing protein molecular weight. Equilibration times, in the batch experimental mode, of the adsorption of lysozyme on the dye-affinity sorbents varied from 20 s for the non-porous glass beads with a size range of 20-40 microns to more than 60 min in the case of a porous sorbent with a particle diameter of 100-300 microns and 60 nm pore size. Furthermore, equilibration times, which represent the overall adsorption rates incorporating all the non-equilibrium effects, increased with all affinity systems when adsorption took place in the non-linear portion of the isotherm. The most dramatic increase was observed when sorbents with relatively high protein size to pore size ratios, lambda, were employed.

High-performance liquid affinity chromatography with phenylboronic acid, benzamidine, tri-L-alanine, and concanavalin A immobilized on 3-isothiocyanatopropyltriethoxysilane-activated nonporous monodisperse silicas

Nonporous, microparticulate, monodisperse silicas with particle diameters between 0.7 and 2.1 microns are introduced as stationary phases in high-performance affinity chromatography. The immobilization of m-aminophenylboronic acid, p-aminobenzamidine, tri-L-alanine, and concanavalin A onto these silicas was successfully achieved using 3-isothiocyanatopropyl-triethoxysilane as an activation reagent. Immobilized phenylboronic acid was applied to the isolation of nucleosides, nucleotides, and glycoprotein hormones such as bovine follicotropin and human chorionic gonadotropin, while immobilized benzamidine was employed for the isolation of the serine proteases thrombin and trypsin, immobilized tri-L-alanine for the separation of pig pancreatic elastase and human leukocyte elastase, and immobilized concanavalin A for the isolation of horseradish peroxidase. In all affinity chromatographic systems studied, the nonporous monodisperse silicas showed improved chromatographic performance compared to results obtained with porous silica supports using identical activation and immobilization procedures. Furthermore, frontal analysis was used as a method to evaluate the influence of experimental parameters on biological activity and accessible ligand densities. Only minor changes in bioactivity were found with the nonporous affinity supports, where accessibilities were typically higher than ca. 60%. The immobilization of affinity ligands onto porous supports as used in this and associated papers thus represents a successful general procedure for the preparation of stable matrices with fast kinetics for use in high-performance affinity chromatography.