Size-exclusion chromatographic protein refolding: fundamentals, modeling and operation

Size-exclusion chromatography (SEC) has proven its capability to refold a variety of proteins using a range of gel filtration column materials, demonstrated in the growing body of experimental evidence. However, little effort has been allocated to the development of mechanistic models describing size-exclusion chromatographic refolding reactors (SECRR). Mechanistic models are important since they provide a link between process variables like denatured and reduced protein feed concentration (C_f,D&;R), flow rate, column length, etc., and performance indicators like refolding yield (Y_N), thereby opening the possibility for in silico design of SECRRs. A critical step, in the formulation of such models, is the selection of an adequate reaction mechanism, which provides the direct link between the separation and the refolding yield. Therefore, in this work we present a methodology using a SEC refolding reactor model, supported by a library of reaction mechanisms, to estimate a suitable reaction scheme using experimental SEC refolding data. SEC refolding data is used since it provides information about the mass distribution of monomers and aggregates after refolding, information not readily available from batch dilution refolding data alone. Additionally, this work presents (1) a systematic analysis of the reaction mechanisms considered using characteristic time analysis and Damköhler maps, revealing (a) the direct effect of a given reaction mechanism on the shape of the SEC refolding chromatogram (number of peaks and resolution) and (b) the effect that the competition between convection, refolding and aggregation is likely to have on the SEC refolding yield; (2) a comparison between the SECR reactor and the batch dilution refolding reactor based on mechanistic modeling, quantitatively showing the advantages of the former over the latter; and (3) the successful application of the modeling based strategy to study the SEC refolding data of an industrially relevant protein. In principle, the presented modeling strategy can be applied to any protein refolded using any gel filtration material, providing the proper mass balances and activity measurements are available.     

Dispersive mixing and intraparticle partitioning of protein in size-exclusion chromatographic refolding

Refolding enables bioprocesses predicated on proteins expressed as inclusion bodies in Escherichia coli. Optimization of size-exclusion chromatography (SEC) refolding is a significant challenge because a wide range of factors, including the choice of gel media, the column dimensions and configuration, affect the final yield in a protein-specific manner. In this study, we investigated these factors by relating them to dispersive mixing and partitioning of refolding molecules within the SEC pore structure. Lysozyme was refolded using SEC resins giving different column dispersion and chromatography resolution. Despite a low separation resolution, the desalting SEC resin Sephadex G-25 resulted in a refolding yield that was 12-30% higher than those obtained with Superdex 75 and Superdex 200. This finding supported the notion that SEC refolding was enhanced by dispersive mixing, which was increased by a wide particle size distribution of the Sephadex G-25 used. Column dispersion was further improved by strategically placing an inlet gap before the packed resin beds, leading to a 20% increase in refolding yield. Refolding yield in Superdex 75 was 20% higher than that in Superdex 200 under conditions giving similar dispersive mixing. This yield enhancement is expected to be protein-specific since Superdex 75 was chosen to specifically maximize partitioning of lysozyme molecules within the resin particles, reducing the likelihood of aggregation during refolding. The highest refolding yield (65%) was achieved using a Sephadex G-25 column with a 15 mm inlet gap, suggesting that desalting systems optimized for dispersive mixing might be an economical and generic alternative for preparative SEC protein refolding.     

Continuous matrix-assisted refolding of proteins

A refolding reactor was developed for continuous matrix-assisted refolding of proteins. The reactor was composed of an annular chromatography system and an ultrafiltration system to recycle aggregated proteins produced during the refolding reaction. The feed solution containing the denatured protein was continuously fed to the rotating bed perfused with buffer promoting folding of the protein. As the protein passed through the column, it was separated from chaotropic and reducing agents and the refolding process took place. Native proteins and aggregates could be continuously separated due to different molecular size. The exit stream containing aggregates was collected, concentrated by ultrafiltration and recycled to the feed solution. The high concentrations of chaotropic and reducing agents in the feed solution enabled dissociation of the recycled aggregates and consequently were fed again to the refolding reactor. When the initial feed mixture of denatured protein is used up, only buffer-containing chaotropic agents and recycled aggregates are fully converted to native protein. This process resulted in a stoichiometric conversion from the denatured protein to its correctly folded native state. The system was tested with bovine alpha-lactalbumin as model protein. Superdex 75 PrepGrade was used as size-exclusion medium. The yield of 30% active monomer in the batch process was improved to 41% at a recycling rate of 65%. Assuming that the aggregates can be redissolved and recycled into the feed stream in a quantitative manner, a refolding yield close to 100% is possible. The method can be also applied to other chromatographic principles suited for the separation of aggregates.     

Molecular simulation of surfactant-assisted protein refolding

Protein refolding to its native state in vitro is a challenging problem in biotechnology, i.e., in the biomedical, pharmaceutical, and food industry. Protein aggregation and misfolding usually inhibit the recovery of proteins with their native states. These problems can be partially solved by adding a surfactant into a suitable solution environment. However, the process of this surfactant-assisted protein refolding is not well understood. In this paper, we wish to report on the first-ever simulations of surfactant-assisted protein refolding. For these studies, we defined a simple model for the protein and the surfactant and investigated how a surfactant affected the folding behavior of a two-dimensional lattice protein molecule. The model protein and model surfactant were chosen such that we could capture the important features of the folding process and the interaction between the protein and the surfactant, namely, the hydrophobic interaction. It was shown that, in the absence of surfactants, a protein in an "energy trap" conformation, i.e., a local energy minima, could not fold into the native form, which was characterized by a global energy minimum. The addition of surfactants created folding pathways via the formation of protein-surfactant complexes and thus enabled the conformations that fell into energy trap states to escape from these traps and to form the native proteins. The simulation results also showed that it was necessary to match the hydrophobicity of surfactant to the concentration of denaturant, which was added to control the folding or unfolding of a protein. The surfactants with different hydrophobicity had their own concentration range on assisting protein refolding. All of these simulations agreed well with experimental results reported elsewhere, indicating both the validity of the simulations presented here and the potential application of the simulations for the design of a surfactant on assisting protein refolding.     

Molecular dynamics for surfactant-assisted protein refolding

Surfactants are widely used to refold recombinant proteins that are produced as inclusion bodies in E. Coli. However, the microscopic details of the surfactant-assisted protein refolding processes are yet to be uncovered. In the present work, the authors aim to provide insights into the effect of hydrophobic interactions of a denatured protein with surfactant molecules on the refolding kinetics and equilibrium by using the Langevin dynamics for coarse-grained models. The authors have investigated the folding behavior of a beta-barrel protein in the presence of surfactants of different hydrophobicities and concentrations. It is shown that the protein folding process follows a "collapse-rearrangement" mechanism, i.e., the denatured protein first falls into a collapsed state before acquiring the native conformation. In comparison with the protein folding without surfactants, the protein-surfactant hydrophobic interactions promote the collapse of a denatured protein and, consequently, the formation of a hydrophobic core. However, the surfactants must be released from the hydrophobic core during the rearrangement step, in which the native conformation is formed. The simulation results can be qualitatively reproduced by experiments.     

Molecular simulation of polymer assisted protein refolding

Protein refolding in vitro, the formation of the tertiary structure that enables the protein to display its biological function, can be significantly enhanced by adding a polymer of an appropriate hydrophobicity and concentration into the refolding buffer. A molecular simulation of the refolding of a two-dimensional simple lattice protein was presented. A protein folding map recording the occurrence frequency of specified conformations was derived, from which the refolding thermodynamics and kinetics were interpreted. It is shown that, in the absence of polymer, the protein falls into the "energy trapped" conformations characterized by a high intramolecular hydrophobic interaction, denoted as HH contact, and a high magnitude of the structure overlap function, chi. This makes it difficult for the protein to fold to the native state. The polymer with a suitable chain length, concentration, and hydrophobicity has formed complex with partially folded protein and created diversified intermediates with low chi. This gives more pathways for the protein to fold to the native state. At a given hydrophobicity, the short chain polymer has a broader concentration range where it assists protein folding than those of long chains. The above simulation agrees well with the experimental results reported elsewhere [Cleland et al., J. Biol. Chem. 267, 13327 (1992); ibid., Bio/Technology 10, 1013 (1992); Chen et al., Enzyme Microb. Technol. 32, 120 (2003); Lu et al., Biochem. Eng. J. 24, 55 (2005); ibid., J. Chem. Phys. 122, 134902 (2005); ibid., Biochem. Eng. J. (to be published)] and is of fundamental importance for the design and application of polymers for protein refolding.     

Continuous full filling capillary electrochromatography-electrospraying chromatographic nanoparticles

The influence of instrumental parameters affecting the ionization in continuous full filling capillary electrochromatography/electrospray ionization mass spectrometry (CFF‐CEC/ESI‐MS) was investigated. The investigated parameters were the BGE and sheath liquid ion strength and organic modifier content, the nebulizer gas pressure, and the concentration of nanoparticles in the BGE. It was found that the nebulizer pressure had the largest influence on the separation efficiency and apparent retention. It was shown that even the lowest pressure investigated was sufficient to guide the nanoparticle flow away from the mass spectrometer inlet. A nebulizer pressure of 5 psi was found to be optimal; increasing the pressure significantly decreased the separation efficiency due to the generation of a hydrodynamic flow. Generally, the ion strength of both the BGE and the sheath liquid were found to have very moderate effects on the separation of a homologous series of dialkyl phthalates, whereas the ionization efficiency was found to be unaffected by the nanoparticles and the separation efficiency was found to increase with increasing concentrations up to 3.8 mg/mL, whereafter it was observed to drop. The optimized method was linear over a wide concentration range and presented LOD and LOQ more than threefold lower than those previously reported using CFF‐CEC/ESI‐MS.     

Controlled oxidative protein refolding using an ion-exchange column

Column-based refolding of complex and highly disulfide-bonded proteins simplifies protein renaturation at both preparative and process scale by integrating and automating a number of operations commonly used in dilution refolding. Bovine serum albumin (BSA) was used as a model protein for refolding and oxido-shuffling on an ion-exchange column to give a refolding yield of 55% after 40 h incubation. Successful on-column refolding was conducted at protein concentrations of up to 10 mg/ml and refolded protein, purified from misfolded forms, was eluted directly from the column at a concentration of 3 mg/ml. This technique integrates the dithiothreitol removal, refolding, concentration and purification steps, achieving a high level of process simplification and automation, and a significant saving in reagent costs when scaled. Importantly, the current result suggests that it is possible to controllably refold disulfide-bonded proteins using common and inexpensive matrices, and that it is not always necessary to control protein-surface interactions using affinity tags and expensive chromatographic matrices. Moreover, it is possible to strictly control the oxidative refolding environment once denatured protein is bound to the ion-exchange column, thus allowing precisely controlled oxido-shuffling.     

Continuous stationary phase gradients for planar chromatographic media

A simple, elegant method for the formation of a continuous stationary phase gradient for use in chromatographic separations is described. Its applicability to separation science is demonstrated using thin-layer chromatography as a test case. Gradient stationary phases were formed on activated High Performance Thin-Layer Chromatography (HP-TLC) plates using a newly developed methodology termed "controlled rate infusion". Specifically, the SiOH groups on the activated HP-TLC plates were reacted with 3-aminopropyltriethoxysilane (APTEOS) in a time dependent fashion by using a programmable syringe pump to control the rate of APTEOS infusion into the deposition reservoir. The shape (profile) of the gradient was controlled by the rate of infusion and imaged by taking advantage of the concentration-dependent color formation reaction between amine groups and ninhydrin. The advantages of such gradients in optimizing the retention and separation of various components in different mixtures were illustrated using mixtures of (1) four weak acids and bases and (2) three widely used over-the-counter drugs. The separation of the individual components on the gradient stationary phase was clearly improved relative to those on either traditional normal-phase TLC plates or uniformly amine-modified TLC plates. Precise control over component retention and separation was also demonstrated by strategically modifying the steepness of the gradient.     

Mechanism of simultaneously refolding and purification of proteins by hydrophobic interaction chromatographic unit and applications

The hydrophobic amino acid residues of a denatured protein molecule tend to react with the particles of the stationary phase of hydrophobic interaction chromatography (STHIC). These hydrophobic interactions prevent the denatured protein molecules from aggregating with each other. The STHIC can provide high enough energy to a denatured protein molecule to make it dehydration and to refold it into its native or various intermediate states. The outcome not only depends on the specific interactions between amino acids, the structure of STHIC, but also depends on the association between the STHIC and mobile phase. The mechanism of protein refolding and the principle of its quality control by HPHIC were also presented. By appropriate selection of the chromatographic condition, several denatured proteins can be refolded and separated simultaneously in a single chromatographic run. A specially designed unit, with diameter much larger than its length, was designed and employed for both laboratory and preparative     

A complete picture of protein unfolding and refolding in surfactants

Interactions between proteins and surfactants are of relevance in many applications including food, washing powder formulations, and drug formulation. The anionic surfactant sodium dodecyl sulfate (SDS) is known to unfold globular proteins, while the non-ionic surfactant octaethyleneglycol monododecyl ether (C₁₂E₈) can be used to refold proteins from their SDS-denatured state. While unfolding have been studied in detail at the protein level, a complete picture of the interplay between protein and surfactant in these processes is lacking. This gap in our knowledge is addressed in the current work, using the β-sheet-rich globular protein β-lactoglobulin (bLG). We combined stopped-flow time-resolved SAXS, fluorescence, and circular dichroism, respectively, to provide an unprecedented in-depth picture of the different steps involved in both protein unfolding and refolding in the presence of SDS and C₁₂E₈. During unfolding, core–shell bLG-SDS complexes were formed within ∼10 ms. This involved an initial rapid process where protein and SDS formed aggregates, followed by two slower processes, where the complexes first disaggregated into single protein structures situated asymmetrically on the SDS micelles, followed by isotropic redistribution of the protein. Refolding kinetics (>100 s) were slower than unfolding (<30 s), and involved rearrangements within the mixing deadtime (∼5 ms) and transient accumulation of unfolded monomeric protein, differing in structure from the original bLG-SDS structure. Refolding of bLG involved two steps: extraction of most of the SDS from the complexes followed by protein refolding. These results reveal that surfactant-mediated unfolding and refolding of proteins are complex processes with rearrangements occurring on time scales from sub-milliseconds to minutes.

The time-resolved study reveals several transition states during SDS-induced unfolding of the protein, as well as under refolding of the protein by the nonionic surfactant C₁₂E₈.     

Continuous melting and ion chromatographic analyses of ice cores

A new method for determining concentrations of organic and inorganic ions in ice cores by continuous melting and contemporaneous ion chromatographic analyses was developed. A subcore is melted on a melting device and the meltwater produced is collected in two parallel sample loops and then analyzed simultaneously by two ion chromatographs, one for anions and one for cations. For most of the analyzed species, lower or equal blank values were achieved with the continuous melting and analysis technique compared to the conventional analysis. Comparison of the continuous melting and ion chromatographic analysis with the conventional analysis of a real ice core segment showed good agreement in concentration profiles and total amounts of ionic species. Thus, the newly developed method is well suited for ice core analysis and has the advantages of lower ice consumption, less time-consuming sample preparation and lower risk of contamination.     

Continuous chromatographic separation of glucose-fructose mixtures using anion-exchange resins

Summary A semi-continuous, counter-current chromatographic refiner (SCCR) was packed with an anion resin in the bisulphite form and used to separate mixtures of sugars from different feedstocks. Products with purities of more than 99% were achieved and solid concentrations of the fructose products of 12% w/v were possible. The deterioration of the anion resin was controlled by simple precautions and thus its useful life was substantially prolonged.     

Considerations of sample application and elution during size-exclusion chromatography-based protein refolding

A mechanism for size-exclusion chromatography-based protein refolding is described. The model considers the steps of loading the denatured protein onto a gel filtration column, and protein elution. The model predictions are compared with results of refolding lysozyme (10 and 20 mg/ml) using Superdex 75 HR. The main collapse in protein structure occurred immediately after loading, where the partition coefficient of unfolded lysozyme increased from 0.1 to 0.48 for the partially folded molecule. Use of a refolding buffer as the mobile phase resulted in complete refolding of lysozyme; this eluted at an elution volume of 15.6 ml with a final partition coefficient of 0.54. The model predicted the elution volume of refolded lysozyme at 19.3 ml.     

Continuous full filling capillary electrochromatography: chromatographic performance and reproducibility

Continuous full filling capillary electrochromatography with nanoparticles as pseudostationary phase interfaced with electrospray ionisation mass spectrometric detection was used for reversed phase separations with very high separation efficiency. Several batches of nanoparticles were synthesised and their electrochromatographic performance were evaluated. Different parameters, such as repeatability, reproducibility, limit of detection, and peak asymmetry, were investigated yielding excellent results. The stability of the system over wide pH ranges and over time was found to be excellent. Very high separation efficiencies with over 1.1 million theoretical plates per metre were obtained. The limit of detection for the investigated dialkyl phthalates was approximately 1.0 micromol L(-1), corresponding to 3-5 fmol injected. After preparation, nanoparticle suspensions could be used without further treatment for at least an entire working day with maintained chromatographic qualities.     

Adsorptive refolding of a highly disulfide-bonded inclusion body protein using anion-exchange chromatography

α-Fetoprotein (AFP) is a prospective biopharmaceutical candidate currently undergoing advanced-stage clinical trials for autoimmune indications. The high AFP expression yields in the form of inclusion bodies in Escherichia coli renders the inclusion body route potentially advantageous for process scale commercial manufacture, if high-throughput refolding can be achieved. This study reports the successful development of an ‘anion-exchange chromatography’-based refolding process for recombinant human AFP (rhAFP), which carries the challenges of contaminant spectrum and molecule complexity. rhAFP was readily refolded on-column at rhAFP concentrations unachievable with dilution refolding due to viscosity and solubility constraints. DEAE-FF functioned as a refolding enhancer to achieve rhAFP refolding yield of 28% and product purity of 95% in 3 h, at 1 mg/ml protein refolding concentration. Optimization of both refolding and chromatography column operation parameters (i.e. resin chemistry, column geometry, redox potential and feed conditioning) significantly improved rhAFP refolding efficiency. Compared to dilution refolding, on-column rhAFP refolding productivity was 9-fold higher, while that of off-column refolding was more than an order of magnitude higher. Successful demonstration that a simple anion-exchange column can, in a single step, readily refold and purify semi-crude rhAFP comprising 16 disulfide bonds, will certainly extend the application of column refolding to a myriad of complex industrial inclusion body proteins.     

Toward chromatographic analysis of interacting protein networks

Protein complexes, collectively referred to as the cellular interactome, appear to play a major role in cellular regulation. At present it is thought that the interactome could be composed of hundreds of protein assemblies. The objective of the work described here was to examine the prospect that chromatographic methods widely used in the preparative isolation of native proteins could be incorporated into global proteomics methods in such a way that the primary structure of protein complexes of sufficient stability to survive chromatography could be recognized along with their participation in protein complexes. Because wide differences in sizes are a unique feature of protein complexes, size-exclusion chromatography (SEC) was incorporated into all the fractionation strategies examined. Anion-exchange chromatography (AEC) and hydrophobic-interaction chromatography (HIC) were also examined because of the broad utility that these methods have shown in the preparation of proteins with native structure. Slightly more than a third of all proteins identified in yeast lysates were found to elute from SEC, AEC, and HIC columns with an apparent molecular weight much higher than that predicted from their parent gene. These results were interpreted to mean that these proteins were migrating through columns as components of protein complexes. Based on studies with multidimensional SEC-->RPLC (reversed-phase liquid chromatography), AEC-->SEC, and HIC-->SEC systems, it was concluded that recognition of proteins in complexes could be easily incorporated into multidimensional chromatographic methods for global proteomics when at least one of the fractionation dimensions included SEC of native proteins.     

Continuous matrix assisted refolding of alpha-lactalbumin by ion exchange chromatography with recycling of aggregates combined with ultradiafiltration

Continuous matrix assisted refolding (MAR) can be achieved on a solid support by using a continuous chromatographic system. Recycling the aggregate fraction, simultaneously formed during a refolding reaction, can further increase the refolding yield. Due to the nature of this reaction, aggregates are the main reason for a refolding yield below stoichiometric conversion. A preparative continuous annular chromatographic system (P-CAC) equipped with an ion exchange resin was used to continuously refold the model protein alpha-lactalbumin. For this purpose, this protein was denatured, reduced and adsorbed on the ion exchange resin. Elution was performed with or without redox reagents in the buffer system permitting fast formation of the native disulfide bonds. In the case redox reagents were present, the protein refolds then during its residence time on the matrix. However, aggregate formation is also increased and refolding yields are lower. Tightly bound aggregates were removed from the column by 2M guanidinium hydrochloride. In order to increase the system yield, this aggregate fraction was recycled after lowering the conductivity by ultradiafiltration and adjustment of the protein concentration by dilution. For on-column refolding, recycling of aggregates at a recycling rate of 0.17 increased the system yield from 25% to 30%. An algorithm was developed to show interdependencies of the single influencing parameters. The operability of the system was demonstrated but limitations due to instability of the P-CAC, especially inhomogeneous flow and peak wobbling, have to be considered.