The use of quantitative structure-activity relationship models to develop optimized processes for the removal of tobacco host cell proteins during biopharmaceutical production.

The production of recombinant pharmaceutical proteins in plants benefits from the low cost of upstream production and the greater scalability of plants compared to fermenter-based systems. Now that manufacturing processes that comply with current good manufacturing practices have been developed, plants can compete with established platforms on equal terms. However, the costs of downstream processing remain high, in part because of the dedicated process steps required to remove plant-specific process-related impurities. We therefore investigated whether the ideal strategy for the chromatographic removal of tobacco host cell proteins can be predicted by quantitative structure-activity relationship (QSAR) modeling to reduce the process development time and overall costs. We identified more than 100 tobacco proteins by mass spectrometry and their structures were reconstructed from X-ray crystallography, nuclear magnetic resonance spectroscopy and/or homology modeling data. The resulting three-dimensional models were used to calculate protein descriptors, and significant descriptors were selected based on recently-published retention data for model proteins to develop QSAR models for protein retention on anion, cation and mixed-mode resins. The predicted protein retention profiles were compared with experimental results using crude tobacco protein extracts. Because of the generic nature of the method, it can easily be transferred to other expression systems such as mammalian cells. The quality of the models and potential improvements are discussed.

The effect of multi-component adsorption on selectivity in ion exchange displacement systems.

This paper examines chemically selective displacement chromatography using affinity ranking plots, batch displacer screening experiments, column displacements, multi-component adsorption isotherms and spectroscopy. The affinity ranking plot indicated that the displacers, sucrose octasulfate (SOS) and tatrazine, should possess sufficient affinity to displace the proteins amyloglucosidase and apoferritin over a wide range of operating conditions. In addition, the plots indicated that the separation of these proteins by displacement chromatography would be extremely difficult. Further, the two proteins were shown to have very similar retention times under shallow linear gradient conditions. When batch displacement experiments were carried out, both tartrazine and SOS exhibited significant selectivity differences with respect to their ability to displace these two proteins, in contrast to the affinity ranking plot results. Column displacement experiments carried out with sucrose octasulfate agreed with the predictions of the affinity ranking plots, with both proteins being displaced but poorly resolved under several column displacement conditions. On the other hand, column displacement with tartrazine as the displacer resulted in the selective displacement and partial purification of apoferritin. Single- and multi-component isotherms of the proteins with or without the presence of displacers were determined and were used to help explain the selectivity reversals observed in the column and batch displacement experiments. In addition, fluorescence and CD spectra suggested that the displacers did not induce any structural changes to either of the proteins. The results in this paper indicate that multi-component adsorption behavior can be exploited for creating chemically selective displacement separations.

A hybrid model framework for the optimization of preparative chromatographic processes.

An optimization framework based on the use of hybrid models is presented for preparative chromatographic processes. The first step in the hybrid model strategy involves the experimental determination of the parameters of the physical model, which consists of the full general rate model coupled with the kinetic form of the steric mass action isotherm. These parameters are then used to carry out a set of simulations with the physical model to obtain data on the functional relationship between various objective functions and decision variables. The resulting data is then used to estimate the parameters for neural-network-based empirical models. These empirical models are developed in order to enable the exploration of a wide variety of different design scenarios without any additional computational requirements. The resulting empirical models are then used with a sequential quadratic programming optimization algorithm to maximize the objective function, production rate times yield (in the presence of solubility and purity constraints), for binary and tertiary model protein systems. The use of hybrid empirical models to represent complex preparative chromatographic systems significantly reduces the computational time required for simulation and optimization. In addition, it allows both multivariable optimization and rapid exploration of different scenarios for optimal design.

Comparison of linear gradient and displacement separations in ion-exchange systems.

The linear gradient mode of chromatography is the most widely employed mode of operation in ion-exchange chromatographic separations. However, in recent years, the displacement mode has received considerable attention because of its promise of high throughput and high resolution. To enable a comparison of these two modes of chromatography, it is essential to identify the optimum operating conditions for each. We employed an iterative algorithm to carry out the necessary optimization. The Steric Mass Action model of ion-exchange chromatography is used in concert with the solid-film linear-driving force model to describe the chromatographic behavior of the solutes in these systems. The performances of displacement and gradient modes of chromatography are compared for different types of separation problems. It turns out that for "easy" separations, both the modes are equally effective. However, for challenging separations, the displacement mode is superior to the gradient mode. Our results shed significant light on the performance of gradient and displacement modes in protein ion-exchange systems.

Purification of recombinant brain derived neurotrophic factor using reversed phase displacement chromatography.

This work investigates the utility of RPLC displacement chromatography for the purification of recombinant brain derived neurotrophic factor (rHu-BDNF) from its variants and E. coli. protein (ECP) impurities. The closely associated variants (six in total) differ by one amino acid from the native BDNF and thus pose a challenging separation problem. Several operational parameters were investigated to study their effects on the yield of the displacement process. The results indicated that the concentration of trifluoroacetic acid (TFA) in the buffer was a key factor in achieving the desired purification. Displacement chromatography on an analytical scale column resulted in extremely high purity and yield in a single chromatographic step. The process was successfully scaled-up with respect to particle and column diameter. The production rate of a pilot scale RPLC displacement process was shown to be 23 times higher than the combined production rates of the current preparative ion exchange and hydrophobic interaction gradient elution steps that are used to remove variant and ECP impurities, respectively.

Characterization and modeling of monolithic stationary phases: application to preparative chromatography.

A methodology for characterizing and modeling preparative separations on monolithic ion-exchange stationary phases is presented. A dimensionless group analysis was carried out to determine the relative importance of mass transfer and kinetic resistances on this stationary phase. In contrast to conventional beaded morphologies, the continuous bed stationary phase was found to possess enhanced mass transport properties resulting in kinetic resistance as the dominant non-ideality. Accordingly, a reaction-dispersive steric-mass action formalism was successfully utilized for simulating preparative displacement chromatography on this resin. Since kinetics were found to be important on this column morphology, mobile phase salt concentration was found to be an important variable during displacement chromatography on this stationary phase. An increase in the mobile phase salt concentration was found to significantly improve the displacement separation of a model protein mixture. The formalism presented in this paper provides a better understanding of preparative chromatography in monolithic resin systems and a means of simulating separations on this class of chromatographic stationary phases.

Purification of an oligonucleotide at high column loading by high affinity, low-molecular-mass displacers.

The development of efficient techniques for large-scale oligonucleotide purification is of great interest due to the increased demand for antisense oligonucleotides as therapeutics as well as their use for target validation and gene functionalization. This paper describes the use of anion-exchange displacement chromatography for the purification of 20-mer phosphorothioate oligonucleotide from its closely related impurities using low-molecular-mass amaranth as the displacer. Experiments were carried out to examine the effect of the feed load on the performance of the displacement chromatography. In contrast to prior work, displacement chromatography was successfully scaled-up to high column loadings while maintaining high purity and yields. Experiments carried out on a Source 15Q column indicated that crude oligonucleotide loading as high as 39.2 mg/ml of column were readily processed, resulting in product recovery of 86% and purity of 92%. These results demonstrate that anion-exchange displacement chromatography can indeed be employed for large-scale oligonucleotide separations at high column loading.

Prediction of protein retention in ion-exchange systems using molecular descriptors obtained from crystal structure.

In this paper, a novel approach is described for the a priori prediction of protein retention in ion exchange systems. Quantitative structure retention relationship (QSRR) models based on a genetic algorithm/partial least squares approach were developed using experimental chromatographic data in concert with molecular descriptors computed using protein crystal structures. The resulting QSRR models were well-correlated, with cross-validated r2 values of 0.938 and 0.907, and the predictive power of these models was demonstrated using proteins not included in the derivation of the models. Importantly, these models were able to predict selectivity reversals observed with two different stationary phase materials. To our knowledge, this is the first published example of predictive QSRR models of protein retention based on crystal structure data.

Purification of oligonucleotides by high affinity, low molecular weight displacers.

High affinity, low molecular weight anionic displacers were successfully employed for the purification of antisense oligonucleotides. Several important structural characteristics were identified that contribute to the affinity of low molecular weight displacers to a hydrophilized polystyrene divinyl benzene anion exchanger. Sulfonic acid groups were found to possess higher affinity than carboxylic acid and phosphate functionalities, and nonspecific interactions (particularly hydrophobic interactions) were shown to play a major role in the retention process on this stationary phase material. Using this information, two high affinity, low molecular weight displacers were identified. These molecules are relatively inexpensive organic dyes that possess multiple sulfonic acid moieties, as well as aromatic functionalities, which increase nonspecific interactions with the stationary phase. These high affinity displacers, which can be readily detected, were then employed to displace several strongly retained antisense oligonucleotides that could not be displaced by previously established low molecular weight displacers. The displacement process resulted in very high purities of the antisense oligonucleotides. The results presented in this paper are significant in that they demonstrate that low molecular weight displacers for ion-exchange chromatography can possess equal to or greater affinities than their higher molecular weight counterparts, when nonspecific interactions with the stationary phase are exploited. In addition, the results illustrate the high resolutions possible with displacement chromatography and demonstrate an attractive technology for the process scale purification of oligonucleotides.

Optimization of ion-exchange displacement separations. I. Validation of an iterative scheme and its use as a methods development tool.

Displacement chromatography has been demonstrated to be a powerful, high-resolution preparative tool. The performance of displacement systems can be affected by a variety of factors such as the feed load, flow-rate, initial salt concentration and the displacer partition ratio. Thus, the optimization of displacement separations is a uniquely challenging problem. In this manuscript, an iterative optimization scheme has been presented whereby one can identify the optimum operating conditions for displacement separations at a given level of loading on a given resin material. The solid film linear driving force model has been employed in concert with the Steric Mass Action formalism of ion-exchange chromatography to describe the chromatographic behavior in these systems. Simple pulse techniques have been employed to estimate the transport parameters. The iterative scheme has been validated using a rigorous Feasible Sequential Quadratic Programming algorithm. Finally, the utility of the iterative optimization scheme as a methods development tool for displacement separations has been demonstrated for a difficult separation. The results indicate that the use of the optimization scheme leads to significantly better performance than standard rules of thumb.

Optimization of ion-exchange displacement separations. II. Comparison of displacement separations on various ion-exchange resins.

A variety of stationary-phase materials are currently available for the chromatographic purification of biomolecules. However, the effect of various resin characteristics on the performance of displacement chromatography has not been studied in depth. In Part I, a novel iterative scheme was presented for the rapid optimization of displacement separations in ion-exchange systems. In this article, the optimization scheme is employed to identify the optimum operating conditions for displacement separations on various ion-exchange resin materials. In addition, the effect of different classes of separation problems (e.g., diverging, converging or parallel affinity lines) on the performance of displacement separations is also presented. The solid film linear driving force model is employed in concert with the Steric Mass Action isotherm to describe the chromatographic behavior in these systems. The results presented in this article provide insight into the effects of resin capacity and efficiency as well as the type of separation problem on the performance of various ion-exchange displacement systems.

Hydrophobic displacement chromatography of proteins.

Displacement chromatography of proteins was successfully carried out in both hydrophobic interaction and reversed-phase chromatographic systems using low-molecular weight displacers. The displacers employed for hydrophobic displacement chromatography were water soluble, charged molecules containing several short alkyl and/or aryl groups. Spectroscopy was employed to verify the absence of structural changes to the proteins displaced on these hydrophobic supports. Displacement chromatography on a reversed-phase material was employed to purify a growth factor protein from its closely related variants, demonstrating the high resolutions that can be achieved by hydrophobic displacement chromatography. This process combines the high-resolution/high-throughput characteristics of displacement chromatography with the unique selectivity of these hydrophobic supports and offers the chromatographic engineer a powerful tool for the preparative purification of proteins.

Characterization of recombinant human brain-derived neurotrophic factor variants.

The purpose of this work was the chemical characterization of variants of the recombinant human brain derived neurotrophic factor (rHu-BDNF), expressed in Escherichia coli. This paper also addresses the question of the in vitro activity of these variants. Chemical characterization of the variants employed peptide mapping using Glu-C protease and cyanogen bromide digestion on reduced and alkylated variants followed by the analysis of the digested peptides using mass spectrometry and Edman sequencing. The BDNF variants in this work have been designated by the order of their elution as observed from the high temperature RPLC assay. It was determined that Peaks 1 and 2, which eluted just before the predominant BDNF peak, had methionine sulfoxide instead of methionine at positions 31 and 61, respectively. Peak 4, which is chromatographically a single peak, contained three variants. Two of these variants had norleucine instead of methionine, at positions 61 and 92, respectively, while the third had methionine sulfoxide instead of methionine at position 92. Peak 5 had norleucine at position 31 instead of methionine. All of these variants showed in vitro biological activity consistent with the BDNF standard, suggesting the preservation of the trkB receptor-ligand binding domain of the variants.

A holistic approach to protein secondary structure characterization using amide I band Raman spectroscopy.

We have developed a holistic protein structure estimation technique using amide I band Raman spectroscopy. This technique combines the superposition of reference spectra for pure secondary structure elements with simultaneous aromatic, fluorescence, and solvent background subtraction, and is applicable to solution, suspension, and solid protein samples. A key component of this technique was the calculation of the reference spectra for ordered helix, unordered helix, and sheet, turns, and unordered structures from a series of well-characterized reference proteins. We accurately account for the overlap between the amide I and non-amide I regions and allow for different scattering efficiencies for different secondary structures. For hydrated samples, we allowed for the possibility that bound water spectra differ from the bulk water spectra. Our computed reference spectra compare well with previous experimental and theoretical results in the literature. We have demonstrated the use of these reference spectra for the estimation of secondary structures of proteins in solution, suspension, and dry solid forms. The agreement between our structure estimates and the corresponding determinations from X-ray crystallography is good.

Cation-exchange displacement chromatography for the purification of recombinant protein therapeutics from variants.

Removal of low level impurities that are closely related to the bioproduct is a commonly encountered challenge in the purification of biopharmaceuticals. These separations are typically carried out by using shallow linear salt gradients at relatively low column loadings, significantly limiting the throughput of the purification process. In this manuscript we examine the utility of displacement chromatography for the purification of recombinant human brain-derived neurotrophic factor, rHuBDNF. The utility of displacement chromatography is compared to gradient elution for the removal of variants of the rHuBDNF. The results demonstrate that displacement chromatography is capable of achieving high yields and purity at high column loadings. Displacements developed on 20 microns and 50 microns cation-exchange resins are shown to provide 8-fold and 4.5-fold increases in production rates, respectively, when compared to an existing linear gradient elution operation. These results demonstrate the efficacy of displacement chromatography for the purification of therapeutic proteins from complex feed streams.

Structural characteristics of low-molecular-mass displacers for cation-exchange chromatography.

The relative efficacy of a variety of low-molecular-mass displacers was examined using a displacer ranking plot. This method enables an evaluation of the dynamic affinity of a variety of displacers over a range of operating conditions. Several homologous series of molecules were evaluated to provide insight into the effects of various structural features on displacer efficacy. The results indicate that linear flexible geometries may have advantages over branched or cyclic structures. Data also indicate that the spreading out of charges may increase affinity. The incorporation of aromatic moieties in these displacers, particularly near the surface of the molecules, appears to result in a dramatic increase in displacer affinity. The ability of several high-affinity low-molecular-mass displacers a very strongly bound cationic protein is also examined. The results confirm the predictions of the theory and indicate that it is indeed possible to displace highly bound macromolecules with low-molecular-mass dispatchers. The work presented in this paper indicates that non-specific interactions can be exploited for producing high-affinity low-molecular-mass displacers.

Purification of an antigenic vaccine protein by selective displacement chromatography.

A recent advance in the state of the art of displacement chromatography has been the development of selective displacement chromatography. In this process, the bioproduct of interest is selectively displaced while impurities with lower retention are eluted in the induced salt gradient and higher retained impurities are desorbed after the breakthrough of the displacer front. In this manuscript, selective displacement chromatography is employed to purify an antigenic vaccine protein (AVP) from an industrial process stream. Displacers were screened and an operating regime plot was employed to establish appropriate conditions for selective displacement. The selective displacement process was successful and resulted in AVP that was equivalent in purity to product obtained at commercial production scale after conventional step gradient chromatography. Methods used to characterize the purified protein include size-exclusion chromatography, SDS-PAGE, isoelectric focusing, N-terminal amino acid sequence analysis, and amino acid composition analysis. This is the first report of the purification of a commercially and pharmaceutically significant protein using selective displacement chromatography and thereby sets the stage for the implementation of selective displacement chromatography for the downstream processing of biologicals.

Structural characteristics of low-molecular-mass displacers for cation-exchange chromatography. II. Role of the stationary phase.

The relative efficacy of a variety of low-molecular-mass displacers was examined on three different stationary phase materials. Several homologous series of displacer molecules were evaluated on these ion-exchange resins using a displacer ranking plot based on the steric mass action model. The results demonstrate that while aromaticity and hydrophobicity can play a significant role in the affinity of displacer molecules on polymethacrylate based and hydrophilized polystyrene-divinylbenzene based materials, this effect is much less pronounced on an agarose based resin. The work presented in this paper demonstrates that different structural features of low-molecular-mass displacers can dominate their affinity on various stationary phase materials employed and provides rules of thumb for the design of high affinity, low-molecular-mass displacers for a variety of commercial cation-exchange materials.

Low-molecular-weight displacers for high-resolution protein separations.

The resolving power of displacement chromatography using low-molecular-weight displacers was investigated using a model mixture containing bovine and horse heart cytochrome c. The linear and nonlinear adsorption behavior of these two proteins was examined in cation-exchange chromatography and shown to be quite similar. Furthermore, an analysis of the dynamic affinity of these proteins indicated extremely similar affinities under displacement conditions. Despite the extreme similarities in the adsorption behavior, displacement chromatography using a protected amino acid displacer resulted in excellent separation of the proteins with both high yields and purity. These results indicate that displacement chromatography may be efficacious for a wide variety of difficult protein separation problems.

Selective displacement chromatography of proteins.

In contrast to high molecular weight polyelectrolyte displacers, the efficacy of low molecular weight displacers are dependent on both mobile phase salt and displacer concentration. This sensitivity to the operating conditions opens up the possibility of carrying out selective displacement where the product(s) of interest can be selectively displaced while the low affinity impurities can be desorbed in the induced salt gradient ahead of the displacement train, and the high affinity impurities either retained or desorbed in the displacer zone. This type of displacement combines the operational advantages of step gradient and the high resolution inherent in a true displacement process, in a single operation. Theoretical expressions are presented for establishing selective displacement operating conditions (initial salt concentration, displacer concentration) based on the Steric Mass Action parameters of the displacer and the linear Steric Mass Action parameters of the feed proteins. Experimental results are presented to elucidate the concept of selective displacement in both cation and anion exchange systems. A mixture of alpha-lactalbumin and beta-lactoglobulin A and B has been used for anion-exchange systems; a four-protein mixture consisting of ribonuclease B, bovine and horse heart cytochrome c, and lysozyme has been employed in cation exchange systems. This article also demonstrates that on-line monitoring can be readily employed for the selective displacement process, thus facilitating the scale-up and control of the process. This work sets the stage for the development of robust large scale high resolution separations using selective displacement chromatography. (c) 1997 John Wiley & Sons, Inc. Biotechnol Bioeng 56: 119-129, 1997.