Scale-up of affinity membrane modules: comparison between lumped and physical models.

Membrane chromatography represents one of the emerging technologies for downstream processing in the biotechnology industry. This process is currently used in polishing steps for antibody manufacturing, while its application is still under development for the capture step. To promote its employment in large-scale processes, it is crucial to develop a simple, yet reliable, simulation tool able to describe the process performance in a predictive way at all scales. In this work, the physical model for the description of protein purification with affinity membrane chromatography has been used to predict the performance of scaled-up systems and compared with the lumped model, frequently used for its deceptive simplicity. Two commonly used binding kinetics have been implemented in the models, namely the Langmuir and the bi-Langmuir equations. The two models describe equally well experimental data obtained in a lab-scale apparatus, while, on the contrary, important differences are observed in scaled-up systems even at the early stages of breakthrough, which are particularly relevant in industrial-scale operations. It is seen that for both kinetics, the physical model is more appropriate and safer to use for scale-up purposes.

Use of the Dispersion Coefficient as the Sole Structural Parameter to Model Membrane Chromatography.

The characterization and modelling of membrane chromatography processes require the axial dispersion coefficient as a relevant and effective intrinsic property of porous media, instead of arbitrary assumptions on pore size distribution. The dispersion coefficient can be easily measured by experiments completely independent of chromatographic tests. The paper presents the prediction of experimentally obtained breakthrough curves using B14-TRZ-Epoxy2 membranes as a test case; the mathematical model implemented is based on the use of the experimentally measured axial dispersion coefficient as an input parameter. Application of the model and its comparison with the data demonstrate that alternative ways of explaining the shape of breakthrough curves, based on unverified assumptions about the membrane pore size distribution, are not feasible and not effectively supported by experimental evidence. In contrast, the axial dispersion coefficient is the only measurable parameter that accounts for all the different contributions to the dispersion phenomenon that occurs in the membrane chromatography process, including the effects due to porous structure and pore size distribution. Therefore, mathematical models that rely on the mere assumption of pore size distribution, regardless of the role of the axial dispersion coefficient, are in fact arbitrary and ultimately misleading.

Gas Transport in Glassy Polymers.

This Special Issue of Membranes provides an updated and comprehensive overview of the state of fundamental knowledge on the fluid sorption and transport in glassy polymers, combining original experimental and modeling works, as well as reviews, prepared by renowned experts [...].

Affinity Membranes and Monoliths for Protein Purification.

Affinity capture represents an important step in downstream processing of proteins and it is conventionally performed through a chromatographic process. The performance of this step highly depends on the type of matrix employed. In particular, resin beads and convective materials, such as membranes and monoliths, are the commonly available supports. The present work deals with non-competitive binding of bovine serum albumin (BSA) on different chromatographic media functionalized with Cibacron Blue F3GA (CB). The aim is to set up the development of the purification process starting from the lab-scale characterization of a commercially available CB resin, regenerated cellulose membranes and polymeric monoliths, functionalized with CB to identify the best option. The performance of the three different chromatographic media is evaluated in terms of BSA binding capacity and productivity. The experimental investigation shows promising results for regenerated cellulose membranes and monoliths, whose performance are comparable with those of the packed column tested. It was demonstrated that the capacity of convective stationary phases does not depend on flow rate, in the range investigated, and that the productivity that can be achieved with membranes is 10 to 20 times higher depending on the initial BSA concentration value, and with monoliths it is approximately twice that of beads, at the same superficial velocity.

Gas Transport in Glassy Polymers: Prediction of Diffusional Time Lag.

The transport of gases in glassy polymeric membranes has been analyzed by means of a fundamental approach based on the nonequilibrium thermodynamic model for glassy polymers (NET-GP) that considers the penetrant chemical potential gradient as the actual driving force of the diffusional process. The diffusivity of a penetrant is thus described as the product of a purely kinetic quantity, the penetrant mobility, and a thermodynamic factor, accounting for the chemical potential dependence on its concentration in the polymer. The NET-GP approach, and the nonequilibrium lattice fluid (NELF) model in particular, describes the thermodynamic behavior of penetrant/polymer mixtures in the glassy state, at each pressure or composition. Moreover, the mobility is considered to follow a simple exponential dependence on penetrant concentration, as typically observed experimentally, using only two adjustable parameters, the infinite dilution penetrant mobility L10 and the plasticization factor ß, both determined from the analysis of the dependence of steady state permeability on upstream pressure. The available literature data of diffusional time lag as a function of penetrant upstream pressure has been reviewed and compared with model predictions, obtained after the values of the two model parameters (L10 and ß), have been conveniently determined from steady state permeability data. The model is shown to be able to describe very accurately the experimental time lag behaviors for all penetrant/polymer pairs inspected, including those presenting an increasing permeability with increasing upstream pressure. The model is thus more appropriate than the one based on Dual Mode Sorption, which usually provides an unsatisfactory description of time lag and required an ad hoc modification.

Modelling and simulation of affinity membrane adsorption.

A mathematical model for the adsorption of biomolecules on affinity membranes is presented. The model considers convection, diffusion and adsorption kinetics on the membrane module as well as the influence of dead end volumes and lag times; an analysis of flow distribution on the whole system is also included. The parameters used in the simulations were obtained from equilibrium and dynamic experimental data measured for the adsorption of human IgG on A2P-Sartoepoxy affinity membranes. The identification of a bi-Langmuir kinetic mechanisms for the experimental system investigated was paramount for a correct process description and the simulated breakthrough curves were in good agreement with the experimental data. The proposed model provides a new insight into the phenomena involved in the adsorption on affinity membranes and it is a valuable tool to assess the use of membrane adsorbers in large scale processes.

Experimental characterization of the transport phenomena, adsorption, and elution in a protein A affinity monolithic medium.

A commercially available convective interaction media (CIM) Protein A monolithic column was fully characterized in view of its application for the affinity capture of IgG in monoclonal antibody production processes. By means of moment analysis, the interstitial porosity and axial dispersion coefficient were determined. The frontal analysis method of characteristic points was employed, for the first time with monolithic media, to determine the dynamic binding capacity. The effects of the flow rate and pH on the total recovery of polyclonal IgG and elution profile were evaluated. A comparison with literature data for Protein A chromatography beads demonstrate the superior bed utilization of monolithic media, which gave better performance at lower residence times.

Separation of MBP fusion proteins through affinity membranes.

Cellulose microporous membranes have been modified in order to obtain a stationary phase specific for the recovery of a class of fusion proteins containing the maltose binding protein domain, through affinity chromatography separations. The feasibility of a single step separation process for the recovery of large amounts of the desired product has been considered. To that purpose, a preparative scale module has been realized, suitable for flat sheet membranes. The affinity matrix used proved to be highly selective toward the fusion proteins examined. The binding capacity determined is comparable with the nominal binding capacity of commercially available supports. The influence of the relevant working parameters, such as flow rate, on the performances of the recovery process has been studied.

Solubility of gases and liquids in glassy polymers.

This review discusses a macroscopic thermodynamic procedure to calculate the solubility of gases, vapors, and liquids in glassy polymers that is based on the general procedure provided by the nonequilibrium thermodynamics for glassy polymers (NET-GP) method. Several examples are presented using various nonequilibrium (NE) models including lattice fluid (NELF), statistical associating fluid theory (NE-SAFT), and perturbed hard sphere chain (NE-PHSC). Particular applications illustrate the calculation of infinite-dilution solubility coefficients in different glassy polymers and the prediction of solubility isotherms for different gases and vapors in pure polymers as well as in polymer blends. The determination of model parameters is discussed, and the predictive abilities of the models are illustrated. Attention is also given to the solubility of gas mixtures and solubility isotherms in nanocomposite mixed matrices. The fractional free volume determined from solubility data can be used to correlate solute diffusivities in mixed matrices.

Influence of protein adsorption kinetics on breakthrough broadening in membrane affinity chromatography.

Existing mathematical models developed to describe membrane affinity chromatography are unable to match the complete breakthrough curve when a single Langmuir adsorption isotherm is used, because important deviations from the observed behavior are systematically encountered in the simulation of breakthrough broadening near saturation. The relevant information required to overcome that limitation has been obtained by considering simultaneously both loading and washing curves, thus evaluating the adsorption data at equilibrium and recognizing what are the appropriate adsorption mechanisms affecting the observed behavior. The analysis indicates that a bi-Langmuir binding kinetics is essential for a correct process description up to the saturation of the stationary phase, together with the use of the relevant transport phenomena already identified for the experimental system investigated. The input parameters used to generate the resulting simulations are evaluated from separate experiments, independent from the chromatographic process. Model calibration and validation is accomplished comparing model simulations with experimental data measured by feeding pure human immunoglobulin G (IgG) solutions as well as a cell culture supernatant containing human monoclonal IgG(1) to B14-TRZ-Epoxy2 bio-mimetic affinity membranes. The simulations obtained are in good agreement with the experimental data over the entire adsorption and washing stages, and breakthrough tailing appears to be associated to the reversible binding sites of the bi-Langmuir mechanism. Remarkably, the model proposed is able to predict with good accuracy the purification of IgG from a complex mixture simply on the basis of the results obtained from pure IgG solutions.

A validated model for the simulation of protein purification through affinity membrane chromatography.

A mathematical model is proposed for the description of protein purification through membrane affinity chromatography. The model describes all the three stages of the chromatographic cycle and takes into account convection, axial dispersion and binding reaction kinetics in the porous membrane matrix, while boundary layer mass transfer resistance is shown to be negligible. All the model parameters have a precise physical meaning which enables their evaluation through separate experimental measurements, independent of the chromatographic cycle. Model testing and validation has been performed with experimental chromatographic cycles carried out with pure IgG solutions as well as with complex mixtures containing IgG(1), using new affinity membranes. The comparison between model calculations and experimental data showed good agreement for all stages of the affinity cycle. In particular, for loading and washing steps binding kinetics was found so fast that adsorption equilibrium was sufficient to describe the observed behavior; as a result, the model simulations are entirely predictive for the adsorption and washing phases. On the contrary, in the elution step the reaction rate is comparable to that of the other simultaneous transport phenomena. The model is able to predict the performance of chromatographic purification of IgG from complex mixtures simply on the basis of the parameter values obtained from pure IgG solutions.

Influence of different spacer arms on Mimetic Ligand™ A2P and B14 membranes for human IgG purification.

Microporous membranes are an attractive alternative to circumvent the typical drawbacks associated to bead-based chromatography. In particular, the present work intends to evaluate different affinity membranes for antibody capture, to be used as an alternative to Protein A resins. To this aim, two Mimetic Ligands™ A2P and B14, were coupled onto different epoxide and azide group activated membrane supports using different spacer arms and immobilization chemistries. The spacer chemistries investigated were 1,2-diaminoethane (2LP), 3,6-dioxa-1,8-octanedithiol (DES) and [1,2,3] triazole (TRZ). These new mimetic membrane materials were investigated by static and by dynamic binding capacity studies, using pure polyclonal human immunoglobulin G (IgG) solutions as well as a real cell culture supernatant containing monoclonal IgG(1). The best results were obtained by combining the new B14 ligand with a TRZ-spacer and an improved Epoxy 2 membrane support material. The new B14-TRZ-Epoxy 2 membrane adsorbent provided binding capacities of approximately 3.1mg/mL, besides (i) a good selectivity towards IgG, (ii) high IgG recoveries of above 90%, (iii) a high Pluronic-F68 tolerance and (iv) no B14-ligand leakage under harsh cleaning-in-place conditions (0.6M sodium hydroxide). Furthermore, foreseeable improvements in binding capacity will promote the implementation of membrane adsorbers in antibody manufacturing.

Understanding ligand-protein interactions in affinity membrane chromatography for antibody purification.

Affinity chromatography with Protein A beads has become the conventional unit operation for the primary capture of monoclonal antibodies. However, Protein A activated supports are expensive and ligand leakage is an issue to be considered. In addition, the limited production capabilities of the chromatographic process drive the research towards feasible alternatives. The use of synthetic ligands as Protein A substitutes has been considered in this work. Synthetic ligands, that mimic the interaction between Protein A and the constant fragment (Fc) of immunoglobulins, have been immobilized on cellulosic membrane supports. The resulting affinity membranes have been experimentally characterized with pure immunoglobulin G (IgG). The effects of the membrane support and of the spacer arm on the ligand-ligate interaction have been studied in detail. Experimental data have been compared with molecular dynamic simulations with the aim of better understanding the interaction mechanisms. Molecular dynamic simulations were performed in explicit water, modelling the membrane as a matrix of overlapped glucopyranose units. Electrostatic charges of the ligand and spacer were calculated through ab initio methods to complete the force field used to model the membrane. The simulations enabled to elucidate how the interactions of surface, spacer and ligand with IgG, contribute to the formation of the bond between protein and affinity membrane.

Performance of a new protein A affinity membrane for the primary recovery of antibodies.

Recovery of antibodies with Protein A affinity chromatography columns has become the standard for the biotechnology industry. Membrane affinity chromatography has not yet experienced extensive application due to the lower capacity of membrane supports compared to chromatographic beads. In this work, new affinity membranes endowed with an interesting binding capacity for human IgG are studied in view of their application in the capturing step of a monoclonal antibody production process. The membranes have been extensively tested with pure IgG solutions and with a cell culture supernatant containing IgG1. The effects of feed flow rate and IgG concentration on the separation performances have been studied in detail, considering in particular binding capacity, selectivity and recovery. These new high capacity affinity membranes appear good candidates to avoid the throughput limitations and other well-known drawbacks of traditional bead-based chromatographic columns.

Preparation and characterization of polysulfone affinity membranes bearing a synthetic peptide ligand for the separation of murine immunoglobulins.

Affinity membranes have been prepared by immobilizing D-PAM, a synthetic ligand that exhibits affinity for the Fc portion of antibodies, onto poliethersulfone microporous membranes. The ligand density has been measured and the ligand utilization was evaluated and compared with literature data available for chromatographic beads. The resulting new affinity membranes have been experimentally characterized and tested by using pure murine IgG solutions and mouse serum. Equilibrium and kinetic parameters have been obtained in batch experiments using pure protein solutions. The highest binding capacity measured for murine IgG was 45 microg/cm(2) obtained at 1.2 mg/mL protein concentration at equilibrium, while the maximum static binding capacity calculated with the Langmuir model was 81 microg/cm(2). The adsorption of murine IgG on the affinity membranes was described using different isotherms: Freundlich and Temkin models have been considered and critically compared with the Langmuir adsorption model. A dynamic binding capacity of 21 microg/cm(2) was obtained by feeding a solution of 0.3 mg/mL of murine IgG, confirming the results obtained in batch experiments at the same concentration. The affinity membranes considered are endowed with good binding capacity for murine IgG and good selectivity for immunoglobulins and can be considered for the capturing step of an antibody production process.