A Comparative Study of Models to Predict Protein Adsorption

Activated carbons and carbon-containing poly(vinyl alcohol) cryogels: characterization, protein adsorption and possibility of myoglobin clearance

Adsorption of myoglobin (Mb), bovine serum albumin (BSA) and γ-globulin (GG) onto activated carbons (ACs) with different pore size distributions, and poly(vinyl alcohol) (PVA) monolithic cryogels containing AC particles was studied. The highest initial rate of Mb adsorption was observed for AC having the largest specific surface area (1939 m(2) g(-1)) and pore volume (1.82 cm(3) g(-1)). The adsorption kinetics of proteins was characterized by a bimodal shape of the distribution f(D) function of an effective diffusion coefficient. Adsorption isotherms of Mb and GG were of Freundlich type within the studied range of equilibrium concentrations (10-150 μg mL(-1)). The distributions of free energy of protein adsorption were bimodal and reflected both interactions with carbon surfaces and self-association of proteins. Adsorbed amounts of Mb were the highest among the proteins studied (up to 700 mg g(-1) carbon), which was attributed to the higher fraction of pores accessible for Mb. Incorporation of carbon particles into PVA-based cryogel resulted in macroporous monolithic composite materials (AC-PVA) exhibiting good flow-through properties. Scanning electron microscopy of the composites showed macroporous aggregates of carbon particles held together by films and bridges of PVA. The rates of adsorption and adsorbed amounts of proteins on AC-PVA were reduced compared to the pristine carbon and depended on the carbon content in the composites. Nevertheless, adsorption of Mb on AC-PVA took place even in the presence of 500-fold higher concentration of BSA. This indicated a possibility of Mb clearance from blood plasma using the PVA-carbon monoliths.

Protein diffusion in charged polyacrylamide gels. Visualization and analysis

Protein diffusion in anionic, cross-linked polyacrylamide-based gels supported in fused-silica capillaries was characterized by a direct visualization method. Microphotography was used to obtain transient protein concentration profiles in these gels using cytochrome c as a probe molecule. Gels based on acrylamido-methylpropane sulfonic acid with 2.5-10% N,N'-methylene-bisacrylamide as a cross-linker and with a total polymer concentration of 0.21 g/cm3 yielded diffuse protein concentration profiles which were quantitatively consistent with a Fickian diffusion model. An analytical method was developed to calculate the diffusivity as a function of protein concentration in the gel from the experimental profiles. The diffusivity was found to assume values in the range 2.5-5.5x10(-8) cm2/s and varied somewhat with the protein concentration in the gel. The effects of some of the polymer properties, such as cross-link density, polymer concentration and charge, were also investigated for a limited range of conditions to derive qualitative trends. Results showed that the transport rates increased with a decrease in the cross-link density, were extremely reduced when the polymer concentration was doubled, and were slightly increased when the charge density was decreased by half by polymerizing a 1:1 mixture of acrylamide and acrylamido-methylpropane sulfonic acid monomers.

Two-component protein adsorption to the cation exchanger S Sepharose FF

A model system, consisting of bovine serum albumin, lysozyme and the cation exchanger S Sepharose FF, was used to investigate multicomponent protein adsorption to ion exchangers. Two models, one based on a complete absence of competition between adsorbing molecules and the other a competitive model, based on the assumption that all adsorption sites are available to both proteins, have been compared to experimental results. Evidence for competitive adsorption was seen in experiments in which breakthrough curves and the profiles of adsorbed proteins in packed beds were determined. However, although the results for packed-bed experiments were more closely predicted by the fully competitive model, some discrepancies were found suggesting that when considering multicomponent protein adsorption to ion exchangers it may also be necessary to take account of factors such as the molecular size of the adsorbing proteins and any potential inter-protein interactions, factors which may hinder the development of a general model of multicomponent protein adsorption to ion exchangers.