No abstract

A straightforward epitope imprinting procedure is used to prepare IgG affinity membranes directly applicable to mAb capture under near process realistic conditions.

Key life science disciplines (e.g. diagnostics, proteomics, protein purification) rely on selective protein binders that serve the purpose of specifically capturing a protein in a complex matrix for either analytical or preparative use. [1,2] These binders are commonly of biological origin, for example, antibodies for protein fractionation or detection in proteomics or antibody-binding proteins (e.g. Protein A) for antibody capture in downstream processing. Common to most biologically derived protein binders are their lability, high cost, denaturation tendency, and intrinsically low binding capacity.Robust artificial protein binders in the form of molecularly imprinted polymers (MIPs) could potentially overcome these limitations, thus offering a step change in the above disciplines. [3] In spite of numerous reports that describe protein-imprinted hydrogels, advances towards generic and robust imprinting techniques have been slow. One reason is the need for employing a low cross-linking level in order to provide a mesh size of the network large enough for the protein to penetrate. The memory effects of these gels are thus easily erased, thereby preventing repeated use of the gels. Various forms of surface-imprinting techniques have been used with promising results to address this problem. [4] However, robust imprinting techniques that afford materials, which can compete with established bioaffinity media in terms of both affinity and capacity, are still lacking.By combining our previously developed hierarchical imprinting technique [5] for small molecules with protein imprinting we herein demonstrate a general technique ( Figure 1) to obtain protein-imprinted separation media that could potentially serve this purpose. As model proteins in this proof of concept study, the human blood plasma proteins, human serum albumin (HSA; 68 kDa, pI = 4.7), and immunoglobulin G (IgG; 150 kDa, pI % 9 for the monoclonal antibody in this study) [6,7] were chosen. As a first step, the proteins were physically immobilized on wide-pore silica (d p = 50 nm) at or close to their isoelectric points. This procedure is known to result in the formation of a strongly adsorbed monolayer with only minor loss of the native protein structure. [6] HSA was nearly quantitatively adsorbed under these conditions, whereas IgG showed a somewhat lower affinity ( Figure S1 in the Supporting Information). Saturation of HSA adsorption was reached at approximately 100 mg g À1 , which corresponds to a protein surface density of approximately 2 mg m À2 , consistent with a monolayer coverage. HSA-and IgG-modified silica templates were then prepared by adjusting the protein surface coverage of approximately 20 %-the submonolayer coverage being chosen in order to reduce surface protein-protein interactions.In order to benchmark the hierarchically imprinted polymers in comparison to reported bulk materials, we chose the common redox-initiated polymerization of methylenebisacrylamide (MBA) as cross-linking monomer (10 % w/w) and acrylamide (AAm) as functional mon...

No abstract

The gelatinases are a subgroup of the matrix metalloproteinase family. The interaction of their C-terminal hemopexin-like domain with a tissue inhibitor of metalloproteinases (TIMP) is a major part of the regulatory mechanisms of gelatinases. To investigate the interaction of the hemopexin-like domain of gelatinase B (92-Pex) and TIMP-1, we expressed the individual domain in Pichia pastoris. The active refolded domain was purified by ion exchange chromatography and gel filtration. We investigated the formation of the 92-Pex/TIMP-1 complex by surface plasmon resonance (SPR). The dissociation constant Kd was calculated to be 0.86 nM. Analogous to the complex of the hemopexin-like domain of gelatinase A and TIMP-2 (Olson, M. W. et al., 1997), the binding curves of the 92-Pex/TIMP-1 complex were best fitted with a monophasic model.