Affinity chromatography is vital to protein research and development

Because they are derived from living organisms, biological products present a potential source of pathogens. By John M. Curling.

Recombinant DNA and hybridoma technologies have given us more than 100 new biopharmaceutical products. Marketed entities must show very low levels of host-cell proteins, cell culture components, and endotoxins. Every effort must be made to remove potential viral or other infectious agents, such as prions. DNA vaccines and products for gene therapy also require purification from cellular impurities and other contaminants that may enter the strictly controlled processing chain.

Affinity chromatography (the most specific form of separation) is at the core of Prometic BioSciences' portfolio technologies for purifying proteins and DNA. The technique exploits biologically relevant interactions such as those found in antibody-antigen recognition, enzyme-inhibitor binding, cell-surface recognition processes, and cell-cell association. Since its introduction, affinity chromatography has been widely used in both academic laboratories and industry, with an average of three publications a day citing the technique. More than 150 patents using it have been granted in the past two years alone, and over 60 per cent of all purification protocols involve the technique, testifying to its central position in protein purification.

Finding new ligands

Biometric affinity uses synthetic ligands to mimic their natural counterparts. Previously referred to as pseudoaffinity, this technology has its origins in the application of triazinyl dyes. Such compounds are analogues of adenyl-containing cofactors and can be used to purify a wide range of enzymes needing such cofactors. Triazine derivatives containing the blue anthraquinone chromophore exhibit significant binding for albumin, interferon, and certain vitamin K dependent coagulation factors.

Although they are chemically related, the new ligand entities developed today have little in common with the fortuitous discoveries of the 1970s. Arrays of ligands can be made using solid phase chemistry. A simple two-dimensional library of 1000 amine substituents generates a total of 500000 possible affinity ligands. By increasing the possible substitution sites using branched and cyclic triazine scaffolds, the diversity can be increased to about 50 billion in a 5-D library.

Instead of actually synthesising such libraries, avirtual' library spaces are constructed. General-property libraries are screened to identify probable aproperty spaces' in which real sublibraries are developed to generate ligand candidates for a target protein. High-resolution crystal structures are helpful at the early design stage but not mandatory or success. Once identified, ligand candidates can be developed into products useful at both laboratory and industrial processing scales.

Research, not development. Methods using proteinaceous ligands (such as Protein A for monoclonal antibodies) suffer from alkaline instability, a critical issue in manufacturing that affects product safety and cost because of reduced-lifetime chromatography resins. To solve that problem, work has been done to enhance stability by exchanging asparagines and glutamines in polypeptide ligands. Resultant products still require cloning into suitable host organisms so they may demonstrate academic elegance. But they are unlikely to be used at production scales because manufacturers are unwilling to introduce new biological products into their processes, which would increase costs and introduce validation problems.

Similarly, phage display with its inherent diversity is a discovery technique par excellence ­ but it generates ligands that require complex peptide synthesis or development in yeast or similar organisms. Resultant ligands may require modification for attachment to solid supports.

Peptide libraries also provide valuable sources for ligands, but internal rotation and bond flexibility restrict their use. However, 6- to 8-mer peptides generally are large enough to provide high specificity and may serve as excellent starting points for the development of synthetic mimics using a rigid triazine backbone.

Affinity applications

By far the most widespread use of affinity chromatography, both at the laboratory scale and in industrial processing, is the purification of monoclonal antibodies from cell culture of ascites. Whether involving humanised chimeras, fully humanised antibodies, FAb fragments, or antibody fusions, antibody purification is important at any scale. Modelled on the IgG dipeptide binding locus of Protein A, the synthetic alternatives MAbsorbent A1P and A2P provide significant advantages over the protein ligand.

Affinity separations can be performed as two-phase partitioning and precipitation, but chromatography remains the primary technology to which they are applied, with almost all separations dependent on a beaded agarose matrix. That naturally occurring polysaccharide has special gelling properties so it can be manufactured in cross-linked, beaded form for both laboratory and industrial use. ProMetic has developed PuraBead XL, a matrix of narrow bead size distribution and rigidity.

Synthetic matrices are also used, particularly in anegative' steps. Usually, a target protein is temporarily fixed (bound) by its immobilised ligands and impurities are washed away. In such apositive' chromatography, high binding capacity is a desirable product attribute. However, products must be free from exogenous contaminants such as viruses, endotoxins, and prions, for which the risks may be real or theoretical. Removal of such biocontaminants demands lower-capacity resins but high-performance adsorption or destruction formats. Those demands are sometimes met with membranes or matrices (such as the inert Fluorosorb) that are rigid and resistant in very harsh condition. Furthermore, the perfluorocarbon base is resistant to gamma radiation so it can be terminally sterilised to remove excess radiolabels or antibodies ­ before cancer diagnosis and therapy, for example.

Proteomics. Study of the human proteome using plasma or tissue samples frequently requires the removal of abundant proteins, that mask the presence of low-concentration proteins in 2D electrophoresis. When the high-concentration proteins albumin and immunoglobulin are removed from plasma, for example, many trace proteins are revealed. Little or nothing is yet known about most of those proteins. Removal of the bulk proteins using suitable mixtures of Mimetic Blue and MAbsorbent A1P allows the remaining proteins to be concentrated, identified, and analysed.

Purification is critical to all genomic and proteomic research and biopharmaceutical development. Every step from basic research to marketable product requires purified proteins and/or nucleic acids that are free of host-cell proteins, cell culture components, endotoxins, and other contaminants. The marriage of genome-related sciences and protein biochemistry ­ proteomics ­ will succeed or fail on the rapidity and specificity of separation science.

John M. Curling is a senior scientist for Prometic BioSciences, Burtonsville, Maryland, USA. Or visit

Recent Issues