Overview of Affinity Purification

Proteins and other macromolecules of interest can be purified from crude extracts or other complex mixtures by a variety of methods. Selective precipitation is perhaps the simplest method for separating one type of macromolecule from another. 

 Most purification methods, however, involve some form of chromatography whereby molecules in solution (mobile phase) are separated based on differences in chemical or physical interaction with a stationary material (solid phase). Gel filtration (also called size-exclusion chromatography or SEC) uses a porous resin material to separate molecules based on size (i.e., physical exclusion). In ion exchange chromatography, molecules are separated according to the strength of their overall ionic interaction with a solid phase material (i.e., nonspecific interactions).

By contrast, affinity chromatography (also called affinity purification) makes use of specific binding interactions between molecules. A particular ligand is chemically immobilized or “coupled” to a solid support so that when a complex mixture is passed over the column, those molecules having specific binding affinity to the ligand become bound. After other sample components are washed away, the bound molecule is stripped from the support, resulting in its purification from the original sample.

Each specific affinity system requires its own set of conditions and presents its own peculiar challenges for a given research purpose. Other Protein Methods articles describe the factors and conditions associated with particular purification systems (see links in side bar near the end of this page). Nevertheless, the general principles involved are the same for all ligand-target binding systems, and these concepts are the focus of this overview.

Affinity purification generally involves the following steps:

1. Incubate crude sample (e.g., cell lysate, cell culture supernatant, or serum) with the affinity support to allow the target molecule in the sample to bind to the immobilized ligand.
2. Wash away non-bound sample components from the support using appropriate buffers that maintain the binding interaction between target and ligand.
3. Elute (dissociate and recover) the target molecule from the immobilized ligand by altering the buffer conditions so that the binding interaction no longer occurs.

Small scale affinity purification using an antibody immobilized to a solid support. Chromatography has three main components: the mobile phase or solvent containing proteins, the stationary or solid phase also called the medium or resin (which may be agarose or other porous resin) and the chromatography column. Affinity chromatography is very selective and provides high resolution with an intermediate to high sample loading capacity. The protein of interest is tightly bound to the resin under conditions that favor specific binding to the ligand, and unbound contaminants are washed off. The bound protein is then recovered in a highly purified form by changing conditions to favor elution. Elution conditions may be specific, such as a competitive ligand, or nonspecific, such as changing pH, ionic strength, or polarity. The target protein is eluted in a purified and concentrated form.

A single pass of a sample (cell lysate, cell culture supernatant, or serum) through an affinity column can achieve greater than 1000-fold purification of a specific protein so that only a single band is detected after gel electrophoresis (e.g., SDS-PAGE) analysis. 

Most commonly, ligands are immobilized or “coupled” directly to solid support material by formation of covalent chemical bonds between particular functional groups on the ligand (e.g., primary amines, sulfhydryls, carboxylic acids, aldehydes) and reactive groups on the support (see related article on Covalent Immobilization). However, indirect coupling approaches are also possible. For example, a glutathione S-transferase (GST)-tagged fusion protein can be first captured to a glutathione support via the glutathione-GST affinity interaction and then secondarily chemically crosslinked to immobilize it. The immobilized GST-tagged fusion protein can then be used to affinity purify binding partner(s) of the fusion protein. 

Ligands that bind to general classes of proteins (e.g., antibodies) or commonly used fusion protein tags (e.g., Histidine tag, or His-tag) are commercially available in pre-immobilized forms ready to use for affinity purification. Alternatively, more specialized ligands such as specific antibodies or antigens of interest can be immobilized using one of several commercially available activated affinity supports; for example, a peptide antigen can be immobilized to a support and used to purify antibodies that recognize the peptide.

Most affinity purification procedures involving protein:ligand interactions use binding buffers at physiologic pH and ionic strength, such as phosphate buffered saline (PBS). This is especially true when antibody:antigen or native protein:protein interactions are the basis for the affinity purification. Once the binding interaction occurs, the support is washed with additional buffer to remove nonbound components of the sample. Nonspecific (e.g., simple ionic) binding interactions can be minimized by adding low levels of detergent or by moderate adjustments to salt concentration in the binding and/or wash buffer. Finally, elution buffer is added to break the binding interaction and release the target molecule, which is then collected in its purified form.

Elution buffers dissociate binding partners by extremes of pH (low or high), high salt (ionic strength), the use of detergents or chaotropic agents that denature one or both of the molecules, removal of a binding factor or competition with a counter ligand. In most cases, subsequent dialysis or desalting is required to exchange the purified protein from elution buffer into a more suitable buffer for storage or downstream analysis.

The most widely used elution buffer for affinity purification based on protein interactions is 0.1 M glycine•HCl, pH 2.5-3.0. This buffer effectively dissociates most protein:protein and antibody:antigen binding interactions without permanently affecting protein structure. However, some antibodies and proteins are damaged by low pH, so eluted protein fractions are best neutralized immediately by addition of 1/10th volume of alkaline buffer such as 1 M Tris•HCl, pH 8.5. Other elution buffers for affinity purification of proteins are listed in the table below.

These conditions apply primarily to protein-protein binding interactions, such as between an antibody and its peptide antigen. Elution buffers for binding interactions between other kinds of molecules may be quite different.

Affinity purification involves the separation of molecules in solution (mobile phase) based on differences in binding interaction with a ligand that is immobilized to a stationary material (solid phase). A support or matrix in affinity purification is any material to which a biospecific ligand is covalently attached. Typically, the material to be used as an affinity matrix is insoluble in the system in which the target molecule is found. Usually, but not always, the insoluble matrix is a solid. Hundreds of substances have been described and utilized as affinity matrices, including agarose, cellulose, dextran, polyacrylamide, latex and controlled pore glass. Useful affinity supports are those with a high surface-area to volume ratio, chemical groups that are easily modified for covalent attachment of ligands, minimal nonspecific binding properties, good flow characteristics and mechanical and chemical stability.

Porous supports (also called resins or gels) generally provide the most useful properties for affinity purification of proteins. These types of supports are usually sugar- or acrylamide-based polymer resins that are produced in solution (i.e., hydrated) as 50-150 µm diameter beads. The beaded format allows these resins to be supplied as wet slurries that can be easily dispensed to fill and "pack" columns with resin beds of any size. The beads are extremely porous and large enough that biomolecules (proteins, etc.) can flow as freely into and through the beads as they can between and around the surface of the beads. Ligands are covalently attached to the bead polymer (external and internal surfaces) by various means. The result is a loose matrix in which sample molecules can freely flow past a high surface area of immobilized ligand.

By far the most widely used matrix for protein affinity purification techniques is crosslinked beaded agarose, which is typically available in 4% and 6% densities. (This means that a 1 mL resin-bed is more than 90% water by volume.) Beaded agarose is good for routine applications as it crushes easily, making it suitable for gravity-flow, low-speed-centrifugation, and low-pressure procedures. Additional crosslinking and/or chemical hardening of beaded agarose resins can increase its ability to withstand higher pressures but can also result in lowering the binding capacity. 

Resins based on polyacrylamide are also used as supports for column affinity chromatography. One example is Thermo Scientific UltraLink Biosupport, which does not compress as easily as typical beaded agarose. UltraLink Biosupport may be used in medium pressure applications with a peristaltic pump or other liquid chromatography systems. Both agarose and UltraLink supports have low nonspecific binding characteristics; nevertheless, they behave slightly differently in particular applications.

Crosslinked 4% beaded agarose and crosslinked 6% beaded agarose are commonly abbreviated as agarose CL-4B and agarose CL-6B, respectively. Also see table below about choosing supports based on scale of use.

Magnetic particles are a completely different type of affinity support from beaded agarose and other porous resins. They are much smaller (typically 1-4 µm diameter) and solid (non-porous). Their small size provides the sufficient surface area-to-volume ratio needed for effective ligand immobilization and affinity purification. Magnetic beads are produced as superparamagnetic iron oxide particles that are covalently coated with silane derivatives. The coating makes the beads inert (to minimize nonspecific binding) and provides the particular chemical groups needed for attaching ligands of interest.

Affinity purification with magnetic particles is not performed in-column. Instead, a few microliters of beads is mixed with several hundred microliters of sample as a loose slurry. During mixing, the beads remain suspended in the sample solution, allowing affinity interactions to occur with the immobilized ligand. After sufficient time for binding has been given, the beads are collected and separated from the sample using a powerful magnet. Typically, simple bench-top procedures are done in microcentrifuge tubes, and pipetting or decanting is used to remove the sample (or wash solutions, etc.) while the magnetic beads are held in place at the bottom or side of the tube with a suitable magnet.

Advantages of magnetic particles over porous resins:

• Magnetic beads exhibit less non-specific binding than porous supports.
• Magnetic beads can be used for cell separation procedures.
• Magnetic beads are suitable for high-throughput automation.

Magnetic beads have generally replaced agarose resin as the preferred support for assay-scale purification techniques, such as immunoprecipitation (IP) and pull-down. Furthermore, increasingly sophisticated and powerful sample-handling instruments are available for performing assays and purification procedures using magnetic separations.

The intended scale of purification and downstream application are perhaps the most important considerations when considering which type of affinity support to use. Differences in pressure limits (see table above), maximum flow rates, and factors such as cost (e.g., magnetic particles would be too costly to use at large scales) determine which support is appropriate to use in a given chromatography system.

Different classes of affinity targets, as well as different purification goals, require consideration of different priorities (e.g., high purity vs. high yield), technical limitations and buffer conditions for development of a successful procedure. The following sections describe some of the most common affinity purification systems. Links to more detailed articles about specific purification methods are provided in the gray boxes.

Several methods of antibody purification involve affinity purification techniques. Typical laboratory-scale antibody production involves relatively small volumes of serum, ascites fluid or culture supernatant. Depending upon how the antibody will be used for various assay and detection methods, it must be partially or fully purified. Three levels of purification specificity include the following approaches:

• Precipitation with ammonium sulfate. This simple technique provides crude purification of total immunoglobulin from other serum proteins.
• Affinity purification with immobilized Protein A, G, A/G or L. These proteins bind to most species and subclasses of IgG, the most abundant type of immunoglobulin produced by mammals in response to immunogens. Ready-to-use resins and purification kits with these proteins are available in many package sizes and formats.
• Affinity purification with immobilized antigen. Covalently immobilizing purified antigen (i.e., the peptide or hapten used as the immunogen to induce production of antibody by the host animal) to an affinity support allows the specific antibody to be purified from crude samples. Activated resins and complete kits for preparing immobilized antigens via a variety of chemistries are available.
   

Specific antibodies are most frequently used to detect antigens of interest in assays, but they also can be used to purify antigens. Because specific antibodies are costly to produce or obtain commercially, this approach is seldom used for large scale purification of antigen. Instead, its use is confined almost entirely to very small-scales, most significantly for immunoprecipitation assays (seen next section).

Nevertheless, when purified antibody is available, it can be covalently immobilized to beaded agarose or other affinity support by any one of several efficient conjugation chemistries. Covalent immobilization via primary amines, as with Thermo Scientific  AminoLink Plus coupling kits, is an especially simple and effective method for preparing an antibody affinity column.

Immunoprecipitation (IP) refers to the small-scale affinity purification of antigen using a specific antibody. Traditional immunoprecipitation involves capturing an antibody-antigen complex with immobilized Protein A or G agarose resin (Protein A or G binds the antibody, which is bound to its antigen), and then recovering the purified antigen in sample loading buffer for gel electrophoresis.

Co-immunoprecipitation (Co-IP) involves attempting to capture and detect not only the direct antigen but also any proteins in the milieu of the cellular lysate that are interacting (i.e., bound to) the antigen. In the traditional format with Protein A or Protein G, this purification scheme involves no less than three levels of affinity interaction.

By adapting and optimizing other methods of antibody immobilization for the small scale needed for IP and Co-IP, several innovations have been developed that overcome many limitations and complications associated with traditional IP techniques.

Like co-immunoprecipitation, pull-down assays are an affinity approach often used for studying protein-protein interactions. However, unlike an IP or Co-IP, pull-down does not involve using an antibody specific to the target protein being studied. The minimal requirement for a pull-down assay is the availability of a purified and tagged protein (the bait) that is used to ‘pull-down’ a protein-binding partner (the prey). The bait protein is created through cloning and expression of a fusion protein or as a covalent modification, such as the addition of a biotin tag (see next two topics). The tagged (e.g., biotinylated) bait protein can be immobilized on a tag-specific affinity support (e.g., streptavidin). Immobilized bait protein is then incubated with a protein solution expressing protein(s) (the prey) that may bind to the bait. These bait-prey protein complexes can then be identified. Alternatively, activated supports can be used to directly immobilize almost any bait molecule.

When proteins are expressed recombinantly, additional amino acids, a functional domain or a whole protein is often appended to aid in the purification and manipulation. These additions to a recombinant protein are known as fusion tags and are added to the DNA that encodes the native protein sequence. One of the most common fusion tags is a short string of six to nine histidine residues (known as the 6xHis or polyHis tag), which will bind to metal ions such as nickel or cobalt. Another fusion tag is glutathione S-transferase (GST), which binds tightly to reduced glutathione.

Other fusion tags include HA, Myc, FLAG (Sigma-Aldrich Co.), MBP, SUMO, and Protein A. Unlike His and GST tags, most of these other tags are called epitope tags because they require specific antibodies (e.g., immobilized anti-HA antibody) for purification. Epitope tags are seldom used for large-scale purification because antibody-based affinity resins are relatively costly compared to simple ligand media such as nickel or glutathione agarose. Instead, epitope tags are more often used for small-scale immunoprecipitation (IP) or Co-IP. 

The properties of fusion tags allow tagged proteins to be manipulated easily in the laboratory. Most significantly, the well-characterized tag-ligand chemistry enables single-step affinity purification of tagged molecules using immobilized versions of their corresponding ligands. Antibodies to fusion tags are also widely available for use in downstream detection and assay methods, eliminating the need to obtain or develop a probe for each specific recombinant protein.

Biotin, also known as vitamin H, is a small molecule (MW 244.3) that is present in tiny amounts in all living cells. The valeric acid side chain of the biotin molecule can be derivatized to incorporate various reactive groups that are used to attach biotin to other molecules. Once biotin is attached to a molecule, the molecule can be captured for detection, immobilization or affinity purification using conjugates or supports based on avidin or streptavidin proteins, which bind strongly and specifically to the biotin group.

Native and recombinant derivatives of avidin and streptavidin proteins are readily available in a wide variety of modified, labeled and immobilized forms. The "avidin-biotin system" (a generic title for all biotin-affinity methods) has been adapted for use in many kinds of research applications for detection or purification.

Because the avidin-biotin affinity interaction is so strong, it is usually impractical to elute biotinylated targets that have been captured to an immobilized avidin or streptavidin support. However, modified versions of biotin labeling reagents have been developed, such as cleavable biotin, iminobiotin and desthiobiotin; these provide readily reversible interactions with streptavidin, making them useful tools for soft-release applications.

In addition to affinity supports and ligands that allow purification of very specific targets (e.g., particular antigens or engineered tags), certain kinds of ligands enable general enrichment or isolation of certain classes of biological molecules. Protein A and Protein G, discussed above, can be thought of as example of this type of affinity system, as they bind to general classes of immunoglobulins. Generally, any unique chemical property or functional group shared by all members of a target set of molecules can become the basis for an enrichment or isolation scheme if a suitable affinity ligand can be identified.

Post-translational modifications (PTM) are good examples of such functional groups that define otherwise unrelated set of molecules. Whether it be phosphorylation, glycosylation or ubiquitination, the PTM has chemical properties that are only subtly distinguishable from other chemical groups by most know chemical ligands. Thus any affinity system can, at best, only enrich for the target class of compounds.

In some cases, the goal of affinity purification is to remove a particular class of undesirable sample components rather than to purify one target molecule. In this sense, the only difference between contaminant removal and traditional affinity purification is that one wishes to keep the flow-through sample and to throw away the bound molecule. In such as scenario, it is important that the binding buffer suitable for sample recovery.

General removal of small molecular weight compounds from protein or other macromolecular sample is usually accomplished by gel filtration (see next section) rather than affinity chromatography. However, where undesired contaminates cannot be differentiated by size and affinity ligands are known that can specifically bind to them in a sample, affinity purification is useful.

The removal of contaminants by affinity is usually performed at the end of a procedure.  For example, detergents that are necessary for cell lysis and protein solubilization can interfere with downstream applications and assays. Several detergent-binding resins are available to process samples in these situations.

Another scenario in which it desirable to remove specific components is for proteomics analysis of serum samples. Often the focus of analysis is on proteins that are much less abundant in the serum or plasma sample than albumin and IgG. Affinity resins, based on specific anti-albumin and anti-IgG antibodies or based on other ligands (Cibacron dye to bind albumin and Protein A/G to bind IgG), are especially useful in these cases. 

Gel filtration (also called size-exclusion chromatography or SEC) uses a porous resin to separate molecules based on size. Small molecules enter the pores of the resin, taking a circuitous route through the column; by contrast large molecules are excluded from the pores, bypassing the internal spaces of the beads and migrating through the column more quickly than the small molecules.

Ion exchange chromatography (IEX or IEC) separates proteins according to the strength of their overall ionic interaction with either negatively of positively charged groups on a resin. By manipulating buffer conditions (e.g., ionic strength and pH), molecules of greater or lesser ionic character can be bound to or dissociated from the solid phase material. IEX supports may either be positively charged (for anion binding) or negatively charged (for cation binding). Furthermore, supports for anion and cation IEX may be characterized for either strong or weak interactions; this does not depict the strength of binding but rather the variation of ionization with pH. Binding with strong exchangers varies minimally with changes in  pH while binding with weak exchangers varies greatly with changes in pH.

Hydrophobic interaction chromatography (HIC) separates proteins based on interactions of external hydrophobic amino acid residues of the protein with hydrophobic groups on a resin.

Multimodal chromatography (MMC) separates proteins similarly to IEX chromatography. MMC uses charged groups on a resin but the groups are modified with a second group which will give a second interaction by which the protein can be purified.

The most common methods of protein purification are all chromatography based. These strategies can be classified into the four application types, which are based upon the properties of the protein to be purified, purity level, and ligand/chemistry. Please note that HIC, SEC, and IEX are useful when isolating a novel protein or when affinity chromatography (AC) is not available. The purity obtained by these methods is protein dependent. Ion exchange and affinity chromatography are two commonly used chromatographic strategies for partial or 1-step purification.

Walker JM. 2009. The Protein Protocols Handbook. Third Edition. Springer-Verlag New York, LLC