The use of affinity-based tools has become invaluable as a platform for basic research and in the development of drugs and diagnostics. Applications include affinity chromatography and affinity tag fusions for efficient purification of proteins as well as methods to probe the protein network interactions on a whole-proteome level. A variety of selection systems has been described for in vitro evolution of affinity reagents using combinatorial libraries, which make it possible to create high-affinity reagents to virtually all biomolecules, as exemplified by generation of therapeutic antibodies and new protein scaffold binders. The strategies for high-throughput generation of affinity reagents have also opened up the possibility of generating specific protein probes on a whole-proteome level. Recently, such affinity proteomics have allowed the detailed analysis of human protein expression in a comprehensive manner both in normal and disease tissue using tissue microarrays and confocal microscopy.
Life is based on specific interactions between biomolecules. The underlying affinities form the basis for molecular recognition events that make up the complex machinery of all living organisms, including man. In fact, the genome project (1,2) has taught us that the number of protein-coding genes is probably as few as between 21,000 (3) and 23,000 (4) and this reinforces the notion that cellular processes are built up by complex networks of specific interactions. The affinities between various molecules in biological systems range from low-affinity interactions to very high-affinity interactions in the picomolar range. The interactions can be transient, such as the molecules in signal pathways, or very stable, such as heterodimer-forming protein complexes or multicomponent organelles, such as ribosomes or proteosomes.
In life science, affinity has been used as a tool to study cellular processes in normal and disease tissues, but it has also been used to develop products for diagnostics and therapeutics. In fact, most pharmaceutical and diagnostic assays are based on affinity between a product and a biomolecular target. In this review, some of the development and use of affinity reagents during the last 25 years will be discussed, although it is important to note that the review is by no means meant to be comprehensive, but rather will show some examples of affinity applications in the field of “biotechniques.”Affinity Chromatography
The use of affinity for purification of proteins through chromatography was first described in the late 60s (5). The method relies on the use of an affinity ligand coupled to a matrix to allow specific capture of the product from a complex mixture. In this way, an essentially pure product can be obtained with a single operation. The most frequent use of affinity chromatography during the last decade has been the purification of antibodies using recombinant protein A (6) or protein G (7). Most monoclonal antibodies used for research and diagnostics and essentially all therapeutic antibodies used to treat patients have been purified using affinity chromatography (8). Recently, protein engineering and design have been used to create new affinity reagents more suitable for affinity chromatography, as exemplified by the protein A derivative engineered to be stable during industrial “cleaning-in-place” procedures involving 0.1 M NaOH (9).
Another application of affinity capture is the technique most commonly called immunoprecipitation (10), which is based on the use of a specific antibody coupled to a solid matrix to capture the protein targets, often in a complex with its interaction partners. This technique has become very popular, with applications ranging from molecular profiling of protein modifications to pathway mapping and network analysis (11). Affinity purification has also been used to facilitate analysis of plasma and serum samples based on affinity capture to remove the most abundant proteins in sera, such as albumin and IgG (12) or transferrin (13). This affinity procedure allows, for some cases, a more sensitive analysis in proteomics efforts aimed at discovering biomarkers useful for distinguishing patients with a particular disease.Affinity Tags
With recombinant DNA methods it is possible to create fusion proteins consisting of a protein to be studied and various tags used for detection and purification (14). Such affinity tags have been used for the generation of purified fusion proteins in a multitude of applications, including structural genomics, antibody generation, and interaction analysis. The first affinity tag was described in 1983 (15) and during the last 25 years many alternative systems have been described ((Table 1)), all having advantages and disadvantages depending on the application and the requirement for specificity, solubility, and the binding and elution conditions. The most often used affinity tag today is probably the His-tag, consisting of a short peptide of histidine residues, which allows a convenient affinity chromatography step using metal-chelating chromatography (16).Table 1. Examples of Affinity Tags
Some examples of affinity tags used for purification of fusion proteins ordered by the year of first publication. The size refers to the minimal size, although the tag is in some cases larger. For details and references, see References 14, and 56,.
An important application of affinity tags is the study of protein networks using tandem affinity protein (TAP) fusions (17). With the aid of two affinity tags, it is possible to capture proteins interacting with the target molecule and by careful elution yield a mixture of proteins corresponding to interactive partners that can subsequently be identified with mass spectrometry. Recently, this procedure has been used to create a proteome-wide network map of yeast (18) and it can be envisioned that the procedure could be used to study in a proteome-wide manner the interactions in humans and model organisms, such as in rodents. In this context, a complementary affinity-based technology for protein interaction analysis is the two-hybrid system (19) based on a selection system that senses the interaction between two affinity partners, one of which is fused to a DNA-binding domain (“bait”) and the other to a transcriptional activation domain (“prey”). In such a way, whole proteomes can be probed for interactions to a particular protein and this can be extended to create proteome-wide network maps (19).