2016 Chromatography Buyer’s Guide

When it comes to separating proteins out of a solution, liquid chromatography is one of the most widely used techniques. The strength of the methodology comes from the ability to exploit physiochemical properties such as size, chirality, hydrophobicity, charge or binding to a specific ligand, to specifically sort, concentrate, desalt or purify proteins from a heterogenous solution. These molecules—often contained in solutions such as cell supernatant or bacterial lysate—are passed over columns typically containing beads (called resins or media and often referred to as “solid phase”) with which they interact in varying but generally predictable ways.

Lab-scale protein preparation was once largely performed in gravity-fed glass columns or homebrewed low-pressure pump-driven systems. Recently, researchers have embraced commercial low- to medium-pressure fast protein liquid chromatography (FPLC) instrumentation, which provides the ability to push the solution (“liquid phase”) through a column at greater, but controlled, pressure (and therefore higher speed), perhaps taking advantage of automation and modular components such as a variety of detectors.

The biocolumn market is fairly fragmented, with certain vendors being far more dominant in the high-pressure liquid chromatography (HPLC) and ultra-high-pressure liquid chromatography (UHPLC/UPLC) markets than in FPLC, and vice versa. Systems have to be designed to work optimally with the choice of columns. For example, FPLC/LPLC (low-pressure liquid chromatography) columns work best on FPLC/LPLC systems. Vendors such as Agilent, Bio-Rad, Cube Biotech GmbH, GE Healthcare Life Sciences, MilliporeSigma, QIAGEN, Thermo Fisher, Waters and many other tool providers offer a diverse array of products to address researchers’ needs.

The features of the columns used to perform chromatography are widely diverse. This guide focuses on affinity-based media, ion-exchange media and other media as well as the columns used in the workflow to prepare protein for lab-scale research. Many of the insights that inform these choices are useful in deciding which reagents best suit specific projects and can also be translated to scaled-up protein-purification workflow processes.

Workflow

Chromatography can be used to analyze, quantitate and understand the structure of small amounts of protein, and for this HPLC or UPLC is the system of choice. For purification, the protein is not so much the subject of study as a tool. In a typical bench-scale protein purification, the researcher may be looking to prepare tens or hundreds of milligrams of product to use for an assay, as a reagent or for some other research need. It’s more about the ease and speed of the purification than the absolute purity or overall yield: How fast can I take a crude sample such as bacteria and cell culture and pull out the protein I’m interested in, with the amount I need? Bench scale is obviously different from pilot or process scale, in which larger quantities of a product may be produced as a therapeutic, for example.

Scientists should begin thinking about how to purify their protein long before they have a crude cell lysate or supernatant in hand. If it’s a recombinant protein, this is the time to engineer in an affinity tag or fusion to facilitate pulling out the protein. This is also the stage, with amino acid sequence in hand—even absent a tag—when researchers can look for properties to help determine the most efficacious media and strategy. The protein’s isoelectric point (pl) compared with the pH of the buffer, for example, can help decide whether an anion-exchange or a cation-exchange column might be helpful. Similarly, finding evidence of hydrophobic patches in the sequence may suggest the utility of hydrophobic interaction chromatography (HIC).

A typical purification begins with some sort of affinity step, such as running supernatant over Protein G to pull out antibodies, or grabbing histidine-tagged proteins using a metal chelate column. For most native proteins, such affinity purification is not an option, though, and in those cases researchers would likely look at more traditional techniques, such as ion-exchange (IEX) chromatography and HIC, perhaps followed by size-exclusion chromatography (SEC, aka gel filtration).

"It is important to perform these separation steps in the right order, using the right column with the appropriate amount of media."

Such steps are often performed sequentially (and are sometimes combined) in a purification workflow, comprising the so-called capture, intermediate and polishing steps. Beginning with bulk separations, and refining the purification as the process progresses, affinity resins and IEX (both anion and cation exchangers), and to a lesser extent HIC, do most of the heavy lifting. SEC is, in most instances, reserved for polishing.

Affinity-based media

An affinity FPLC resin is generally composed of a base bead—agarose and its derivatives are the most common—to which is attached a functional group with a particular specificity for the protein of interest.

Beads with a variety of ligands are available commercially. For example, GE’s “Affinity chromatography columns and media selection guide” boasts pre-packed columns and bulk resins that can bind proteins from albumin to fibronectin. (Bio-Rad  also provides an online guide to assist researchers with column and resin selections).

Many of these affinity resins are used for what is called group-specific purification, meaning they bind a host of related proteins, such as certain kinases, proteases, or glycoproteins. These media often are used to pull out proteins that have been tagged with biotin, polyhistidine, glutathione S-transferase (GST) or maltose binding protein (MBP), for example.

Some media specifically recognize classes, subclasses or fragments of immunoglobulins; for example, modern resins with a modified Protein A can yield greater than 95% purity in a single affinity step. But researchers can also attach their own ligands to media with activated functional groups such as epoxides, hydrazines and amines. A bound antibody of choice can be used to capture its corresponding antigen on the column (analogously to immunoprecipitation), for example, or an immobilized drug inhibitor can capture its binding partner. The interaction is often highly specific, but elution may occur in harsh conditions that need to be tested and optimized to avoid denaturing the protein.

Bind-and-elute chromatography often can be scaled up without many changes to the base media—an important consideration for projects that may transition to production scale. This provides peace of mind, as the entire protocol would not need to be re-engineered in case, say, the researchers isolated a novel therapeutic or reagent. It’s also important to consider whether such a scale-up is commercially feasible: Is the resin prohibitively expensive, or is the attached antibody difficult to obtain? Depending on the answers to these questions, researchers may want to consider more scalable alternatives early in the process.

Because the protein is being selectively bound, bead size is less of a concern for affinity chromatography than it is for doing separation based on protein size—you obtain the purity very easily, and don’t need to focus on the resolution.

Capture resins tend to be larger, in part because the sample being loaded is typically a clarified lysate, chock full of viscous protein, which would likely clog a column made from (say) 10-mm beads.

"One of the advantages of affinity chromatography is that there is virtually no limit to the volume of sample that can be loaded onto the column, as long as the resin has sufficient capacity to adsorb the protein in the sample—the rest just flows through."

(The binding capacity is generally provided by the vendor in terms of milligrams of protein per milliliter of chromatography medium.)

Many affinity columns can be run by gravity, and some—like the metal chelate columns used to pull down polyhistidine-tagged proteins—are quick and easy to use for certain applications. Gravity columns may not provide quite the level of reproducibility as a pressure-driven column, nor will they provide a chromatogram, but often these are not a concern for bespoke purifications.

In fact, affinity chromatography doesn’t even need to be performed in a typical column. Small volumes often can be run in spin columns. And magnetic beads may be a good option for automating large numbers of purifications.

Ion-exchange media

"As efficient as affinity columns are, there are almost always some contaminants that come along for the ride."

So although it’s best to minimize the number of purification steps—every additional step has a cost, in terms of product yield—affinity purification often is followed by some sort of IEX as a polishing step. And if an affinity resin is not available, ion exchange can function as an initial step in the purification. IEX tends to be lower cost, offer a higher capacity—which is especially important for a capture step—and be easier than trying to separate on the basis of size.

For practical purposes, IEX media can be divided into two categories: cation exchangers, which attract positively charged molecules, and anion exchangers, which bind negatively charged molecules. Each category includes both strong and weak binders. Buffer conditions such as salt concentration and pH are used to further modify the binding properties of these resins.

The functional groups responsible for strong anion exchange or weak cation exchange do not tend to differ substantially among manufacturers: Cation exchangers typically contain a sulphonyl or carboxyl group, for example, while anion exchangers sport a quaternary ammonium ion or tertiary or secondary amine. The base resins themselves often contribute to the binding profile, as well.

Researchers may start with a run through a column using larger beads, so as not to foul the column, followed by a run through the column with smaller beads—or a second IEX step using an oppositely charged resin, or even HIC—as a polishing step.

HIC is essentially the flip side of IEX. In HIC, hydrophobic ligands are used to bind the hydrophobic patches on the surface of proteins under high salt concentration.

Under the right conditions, smaller beads can provide higher resolution. But most people use affinity, IEX or HIC resins not for that reason but rather as a grab column.

Mixed-mode or multimodal media

Mixed-mode or multimodal resins, which comprise a more recent category, are generating a lot of interest. These combine multiple selection properties, such as IEX and HIC functionalities, on the same bead—usually on the same ligand. Researchers cannot predict a priori exactly how their protein will interact with the hydrophilic and charged moieties of a mixed-mode resin, so using these resins may require more optimization than with a single-mode resin. But it may be worth the effort, because adjusting the conditions can make both process contaminants (other proteins found in the sample) and product impurities (such as truncations and small charge variants) amenable to fine separation.

Unlike other media described here, mixed-mode ligands tend to differ substantially among vendors, with one using an aromatic while another uses a branched hydrocarbon as the hydrophobic element, for example. There are also a host of other specialized chromatography resins that can be used in the FPLC (and gravity) arenas. To take just one example, Bio-Rad’s ceramic hydroxyapatite—in which the CaPO4 bead doubles as the functional group—is a combined anion and cation exchanger, the action of which can be modulated by pH and salt concentration.

Because of the many options available from various tool providers, it’s advisable to screen as many resins as possible when considering multimodal media. This will help you determine which gives the best separation of your particular protein.

Size-exclusion media

As the media name applies, in SEC, proteins in a solution flow in and around the resin as they make their way through the column. Smaller molecules take longer to make it through the column, because they enter and travel through the bead’s pores to a greater extent than larger molecules. SEC is a good way to separate groups of proteins that have large differences in molecular weight, such as monomers from dimers and aggregates. It’s also useful in cleaning up, desalting and exchanging buffers in a sample. For downstream processes such as mass spectrometry or crystallography, in which a high level of purity may be needed, SEC often is used as a polishing step.

In the case of SEC media offerings, many vendors provide a diverse array of products, specifically differing in bead material and pore sizes. For example, GE’s new Superdex™ and Superose™ Increase line of SEC resins’ small bead sizes—about 10-mm —and very narrow particle size distribution can act as a bridge between FPLC and HPLC, enabling the researcher to separate proteins with nearly analytical-level resolution using low-pressure chromatography systems.

Wrap it up

To prepare small amounts of protein for downstream applications ranging from crystallography to ELISAs to inhibitor studies, researchers most often rely on chromatography using low- to medium-pressure-driven FPLC instruments.

A variety of media use different physiochemical properties to effect separations. Affinity chromatography is generally the preferred first line, as it provides the most selective separations and thus the greatest purity—if an appropriate ligand to bind the protein of interest is available. When a charged or hydrophobic protein is to be purified, IEX, HIC and mixed-mode chromatography are the obvious next choices. Yet because different molecules may share the same charge or hydrophobicity, several iterations—often in “opposite directions”—may be necessary. SEC, when used, is typically reserved for final, polishing steps. Oftentimes multiple columns are used in series to achieve the desired purity.

We acknowledge the following for useful discussions in preparing this article: Gurmil Gendeh (Agilent), Maja Petkovic (AMSBIO), Jeff Habel (Bio-Rad), John Daicic (GE Healthcare and Life Science), Paul Lynch (Thermo Fisher Scientific), Michael Miley (Univ. North Carolina at Chapel Hill)

Image: Shutter Stock