Search

Controlled shear affinity filtration (CSAF) is an integrated bioprocess that positions a contoured rotor above a membrane affinity chromatography column to permit the capture and purification of a secreted protein product directly from cell culture. Here, computational fluid dynamics (CFD) simulations previously used on a laboratory-scale unit (Francis et al., Biotechnol. Bioeng. 2005, 95, 1207-1217) are extended to study the fluid hydrodynamics and expected filter performance of the CSAF device for rotor sizes up to 140 cm in radius. We show that the fluid hydrodynamics within the rotor chamber of larger-scale CSAF units are complex and include turbulent boundary layers; thus, CFD likely provides the only reliable route to CSAF scale-up. We then model design improvements that will be required for CSAF scale-up to permit processing of industrial feedstock. The result is the in silico design of a preparative CSAF device with an optimized rotor 140 cm in radius. The scaled up device has an effective filtration area of 5.93 m(2), which should allow for complete processing in ca. 2 h of 1000 L of culture harvested from either a perfusion, fed-batch or batch bioreactor. Finally, a novel method for the parallelization of CSAF units is presented for use in bioprocessing operations larger than 1000 L.

A novel two-zone model (TZM) is presented to describe the rate of solute uptake by the stationary phase of a sorption-type chromatography column. The TZM divides the porous stationary-phase particle into an inner protein-free core and an outer protein-containing zone where intraparticle transport is limited by pore diffusion and binding follows Langmuir theory. The TZM and the classic pore-diffusion model (PDM) of chromatography are applied to the prediction of stationary-phase uptake and elution bands within a cellulose-based affinity chromatography column designed to selectively purify proteins genetically labelled with a CBM9 (family 9 cellulose binding module) affinity tag. Under both linear and nonlinear loading conditions, the TZM closely matches rates of protein uptake within the stationary phase particles as measured by confocal laser scanning microscopy, while the PDM deviates from experiment in the linear-binding region. As a result, the TZM is shown to provide improved predictions of product breakthrough, including elution behavior from a bacterial lysate feed.

Entropic interaction chromatography (EIC) provides efficient size-based separation of protein mixtures through the entropy change associated with solute partitioning into a layer of hydrophilic homopolymer that has been end-grafted within the pores of a macroporous chromatography support. In this work, surface-initiated atom-transfer radical polymerization (ATRP) is used to prepare a library of EIC stationary phases covering a wide range of grafted-chain densities and molecular weights. Exhaustive chain cleavage and analysis by saponification and GPC-MALLS, respectively, show that the new ATRP synthesis procedure allows for excellent control over graft molecular weight and polydispersity. The method is used to prepare high-density grafts (up to 0.164 +/- 0.005 chains/nm(2)) that extend the range of EIC applications to include efficient buffer-exchange and desalting of protein preparations. Reducing the graft density allows for greater partitioning of high molecular weight solutes, extending the linear range of the selectivity curve. Increasing graft molecular weight also alters selectivity, but more directly affects column capacity by increasing the volume of the grafted layer. Protein partitioning in high-density EIC columns is found to decrease with mobile-phase velocity (u). Although solute mass transfer resistances leading to an increase in plate height can explain this effect, pressure drop data across the column are indicative of weak convective flow through at least a fraction of the grafted architecture. Modeling of the grafted brush properties in the presence of solvent flow by subjecting a self-consistent-field theory representation of the brush to a viscous shear force predicts that the grafted chains will tilt and elongate in the direction of flow. The shear force may therefore act to reduce the number of conformations available to chains, increasing their rigidity without significantly altering the thickness of the grafted layer. A reduction in protein partitioning is then predicted when the dependence on u of the solute entropy loss is stronger than that of the grafted polymer, a condition met at high graft densities.

Controlled shear affinity filtration (CSAF) is a novel integrated processing technology that positions a rotor directly above an affinity membrane chromatography column to permit protein capture and purification directly from cell culture. The conical rotor is intended to provide a uniform and tunable shear stress at the membrane surface that inhibits membrane fouling and cell cake formation by providing a hydrodynamic force away from and a drag force parallel to the membrane surface. Computational fluid dynamics (CFD) simulations are used to show that the rotor in the original CSAF device (Vogel et al., 2002) does not provide uniform shear stress at the membrane surface. This results in the need to operate the system at unnecessarily high rotor speeds to reach a required shear stress of at least 0.17 Pa at every radial position of the membrane surface, compromising the scale-up of the technology. Results from CFD simulations are compared with particle image velocimetry (PIV) experiments and a numerical solution for low Reynolds number conditions to confirm that our CFD model accurately describes the hydrodynamics in the rotor chamber of the CSAF device over a range of rotor velocities, filtrate fluxes, and (both laminar and turbulent) retentate flows. CFD simulations were then carried out in combination with a root-finding method to optimize the shape of the CSAF rotor. The optimized rotor geometry produces a nearly constant shear stress of 0.17 Pa at a rotational velocity of 250 rpm, 60% lower than the original CSAF design. This permits the optimized CSAF device to be scaled up to a maximum rotor diameter 2.5 times larger than is permissible in the original device, thereby providing more than a sixfold increase in volumetric throughput., (Copyright 2006 Wiley Periodicals, Inc.)

Partitioning of a macromolecule into the interfacial volume occupied by a grafted polymer brush decreases the configurational entropy (DeltaSbrush(c)) of the terminally attached linear polymer chains due to a loss of free volume. Self-consistent field theory (SCF) calculations are used to show that DeltaSbrush(c) is a strong function of both the size (MWp) of the partitioning macromolecule and the depth of penetration into the brush volume. We further demonstrate that the strong dependence of DeltaSbrush(c) on MWp provides a novel and powerful platform, which we call entropic interaction chromatography (EIC), for efficiently separating mixtures of proteins on the basis of size. Two EIC columns, differing primarily in polymer grafting density, were prepared by growing a brush of poly(methoxyethyl acrylamide) chains on the surface of a wide-pore (1,000-A pores, 64-microm diameter rigid beads) resin (Toyopearl AF-650M) bearing surface aldehyde groups. Semipreparative 0.1-L columns packed with either EIC resin provide reduced-plate heights of 2 or less for efficient separation of globular protein mixtures over at least three molecular-weight decades. Protein partitioning within these wide-pore EIC columns is shown to be effectively modeled as a thermodynamically controlled process, allowing partition coefficients (K(P)) and elution chromatograms to be accurately predicted using a column model that combines SCF calculation of K(P) values with an equilibrium-dispersion type model of solute transport through the column. This model is used to explore the dependence of column separation efficiency on brush properties, predicting that optimal separation of proteins over a broad MWp range is achieved at low to moderate grafting densities and intermediate chain lengths., (Copyright (c) 2005 Wiley Periodicals, Inc.)

Dynamic Monte Carlo (DMC) simulations of the adsorption of simple protein-like chains are used to more clearly define the molecular basis for the dependence of adsorption thermodynamics on the stability of the unique lowest-energy "native state" conformation of the chain. Arai and Norde were among the first to show that proteins of low native-state stability strongly denature upon adsorption to weakly attractive sorbent surfaces, while relatively modest changes in conformation are observed in stable proteins under identical adsorption conditions. When the protein has a low native-state stability, favorable adsorption entropies are typically observed in such systems, leading to the general belief that the chain gains conformational entropy during adsorption through a net reduction in intramolecular interactions specific to the native-state structure. Analysis of energy landscapes generated from our DMC simulation results show that a net loss in specific intramolecular interactions can lead to a positive delta(ads)S under certain adsorption conditions. However, the influence of chain conformation on delta(ads)S is found to correlate more directly with the manner in which the unique states of the system are distributed among the energy levels available to the adsorbed chain. Delta(ads)S is found to tend toward a maximum for adsorption processes described by thermally averaged energy landscapes in which the energy levels carrying the highest Boltzmann weights have a high degree of conformational degeneracy. This condition is met when the average interaction energy between the chain and the sorbent equals that between two hydrophobic segments of the chain.

Liquid-liquid extraction in two-phase aqueous complex-fluid systems has been proposed as a scalable, versatile, and cost-effective purification method for the downstream processing of biotechnological products. In the case of two-phase aqueous micellar systems, careful choices of the phase-forming surfactants or surfactant mixtures allow these systems to separate biomolecules based on size, hydrophobicity, charge, or specific affinity. In this article, we investigate the affinity-enhanced partitioning of a model affinity-tagged protei--green fluorescent protein fused to a family 9 carbohydrate-binding module (CBM9-GFP)--in a two-phase aqueous micellar system generated from the nonionic surfactant n-decyl beta-D-glucopyranoside (C10G1), which acts simultaneously as the phase-former and the affinity ligand. In this simple system, CBM9-GFP was extracted preferentially into the micelle-rich phase, despite the opposing tendency of the steric, excluded-volume interactions operating between the protein and the micelles. We obtained more than a sixfold increase (from 0.47 to 3.1) in the protein partition coefficient (Kp), as compared to a control case where the affinity interactions were "turned off" by the addition of a competitive inhibitor (glucose). It was demonstrated conclusively that the observed increase in Kp can be attributed to the specific affinity between the CBM9 domain and the affinity surfactant C10G1, suggesting that the method can be generally applied to any CBM9-tagged protein. To rationalize the observed phenomenon of affinity-enhanced partitioning in two-phase aqueous micellar systems, we formulated a theoretical framework to model the protein partition coefficient. The modeling approach accounts for both the excluded-volume interactions and the affinity interactions between the protein and the surfactants, and considers the contributions from the monomeric and the micellar surfactants separately. The model was shown to be consistent with the experimental data, as well as with our current understanding of the CBM9 domain.

Proteins prefer interfaces, and in aqueous solutions they rapidly adsorb to available solid-liquid interfaces. The adsorption process often involves a change in protein conformation at the surface that can result in functional inactivation of the protein. These changes in protein conformation, which are thought to lead to the formation of an entangled gel-like layer of denatured protein, are responsible for a number of deliterious processes, including biofouling on contact lenses and medical implants. The adsorption process is generally irreversible; dilution of protein in the solution phase does not result in protein desorption from the solid. Presumably, this is due to the effects of the protein denaturation and entanglement process on the rate constant for desorption. Nonspecific protein adsorption to solid-liquid interfaces is, therefore, a kinetically controlled process. Hence, measuring and understanding the kinetics of protein adsorption to solid surfaces, including the kinetics of protein conformational changes, is of considerable interest. We have developed a sensor that responds to protein adsorption kinetics and also, we believe, is sensitive to protein conformational changes during adsorption. The device is operated by monitoring the change in resonant frequency of an elastomeric film (25μm thick), as an aqueous protein solution is exposed to the surface. Since the mass of a monolayer of the protein or other adsorbent is an extremely small fraction of the mass of the film, the observed change in resonant frequency is due almost entirely to changes in the surface tension of the film. Upon exposing the elastomeric film to a protein solution, we observe a continuing change in resonant frequency for more than 24 h, which is well beyond the time it would take for the population of proteins on the surface to equilibrate under diffusion-limited kinetics. This prolonged response is likely due to the surface energy changes of the sensor as the adsorbed... [ABSTRACT FROM AUTHOR]

δ-crystallin is directly related to argininosuccinate lyase (ASL), and catalyzes the reversible hydrolysis of argininosuccinate to arginine and fumarate. Two δ-crystallin isoforms exist in duck lenses, δ1 and δ2, which are 94% identical in amino acid sequence. Although the sequences of duck δ2-crystallin (dδc2) and duck δ1-crystallin (dδc1) are 69 and 71% identical to that of human ASL, respectively, only dδc2 has maintained ASL activity. Domain exchange experiments and comparisons of various δ-crystallin structures have suggested that the amino acid substitutions in the 20's (residues 22-31) and 70's (residues 74-89) loops of dδc1 are responsible for the loss of enzyme activity in this isoform. To test this hypothesis, Å double loop mutant (DLM) of dδc1 was constructed in which all the residues that differ between the two isoforms in the 20's and 70's loops were mutated to those of dδc2. Contrary to expectations, kinetic analysis of the DLM found that it was enzymatically inactive. Furthermore, binding of argininosuccinate by the DLM, as well as the dδc1, could not be detected by isothermal titration calorimetry (ITC). To examine the conformation of the 20's and 70's loops in the DLM, and to understand why the DLM is unable to bind the substrate, its structure was determined to 2.5 Å resolution. Comparison of this structure with both wild-type dδc1 and dδc2 structures reveals that the conformations of the 20's and 70's loops in the DLM mutant are very similar to those of dδc2. This suggests that the five amino acid substitutions in domain 1 which lie outside of the two loop regions and which are different in the DLM, and dδc2, must be important enzymatically. The structure of the DLM in complex with sulfate was also determined to 2.2 Å resolution. This structure demonstrates that the conformational changes of the 280's loop and domain 3, previously observed in dδc1, also occur in the DLM upon sulfate binding, reinforcing the hypothesis that these events may occur in the active dδc2 protein during catalysis. [ABSTRACT FROM AUTHOR]

A novel inexpensive affinity purification technology is described based on recombinant expression in Escherichia coli of the polypeptide or protein target fused through its N-terminus to TmXyn10ACBM9-2 (CBM9), the C-terminal family 9 carbohydrate-binding module of xylanase 10A from Thermotoga maritima. Measured association constants (Ka) for adsorption of CBM9 to insoluble allomorphs of cellulose are between 2×105 and 8×106 M-1. CBM9 also binds a range of soluble sugars, including glucose. As a result, a 1 M glucose solution is effective in eluting CBM9 and CBM9-tagged fusion proteins from a very inexpensive commercially-available cellulose-based capture column. A processing site is encoded at the C-terminus of the tag to facilitate its rapid and quantitative removal by Factor Xa to recover the desired target protein sequence following affinity purification. Fusion of the CBM9 affinity tag to the N-terminus of green fluorescent protein (GFP) from the jellyfish, Aquorin victoria, is shown to yield >200 mg l-1 of expressed soluble fusion protein that can be affinity separated from clarified cell lysate to a purity of >95% at a yield of 86%. [Copyright &y& Elsevier]