Scientists Use Various Methods To Tackle Protein Purification

Tedious, time-consuming, exasperating: Those are perhaps the three words that researchers use most commonly to describe protein purification. Proteins, and their little sisters the peptides, are among the most difficult compounds to purify. Enzyme degradation and chemical decomposition conspire to ruin protein samples throughout the steps of purification. Proteases in the original sample and varying salt concentrations can result in degradation or precipitation. Varying pH and temperature may irreversibly alter the protein's conformation. "It really is a hybrid problem of purification and formulation in order to keep the protein functional. You have to be heads-up," says Pete Gagnon, president and scientific director at Tucson, Ariz.-based Validated Biosystems Inc. Validated Biosystems reverses a common relationship between researcher and custom services: Rather than require the investigator to send material to its facility, the company goes to the research lab and tackles the problem in-house. "We help [researchers] develop processes that ensure the final [product] meets their specifications," states Gagnon. "There are so many nuances about working in a lab that [can't be communicated on paper]. Being there can make the difference between a great prep and winding up with egg on your face." Indeed, proteins have a sinister way of making even experts feel foolish. To tackle the problem, researchers can turn to a vast array of methods that have been developed over the years. Popular chromatographic approaches rely on a protein's interaction with some molecule linked to the matrix of a column (affinity chromatography); the size of the protein (size exclusion); the pH at which it is electrically neutral (chromatofocusing); hydrophobic interactions (reverse-phase); and overall charge (ion exchange). Another approach, embraced by many, is to express the protein as a conjugate to a tag or another protein that is easily lifted from a solution. A string of six histidine residues (6xHis) at one terminus of the protein, for example, will bind tightly to nickel metal bound to a column. Thus anchored, the protein remains on the column until extraneous material is cleared, and then eluted by either reducing the pH or adding imidazole. EXPRESSION SYSTEM: QIAexpress bacterial expression vectors from QIAGEN offer the opportunity to place the 6xHis tag at either the N- or C-terminus. The QIAexpress protein purification and expression system from QIAGEN Inc. of Valencia, Calif., uses nickel nitrilotriacetic acid (Ni-NTA) to bind proteins with a 6xHis tag. The tag and the purification procedure are compatible with any expression system, be it bacteria, mammalian cell, baculovirus, or yeast, the company claims. Most animal species do not mount an immune response to the tag, and it rarely interferes with protein structure or function. Because the histidine-nickel interaction does not depend upon the structure of the tagged protein, the system can be used under denaturing conditions, often simplifying purification. The tags' very high affinity for the nickel ions immobilized on the resin means that, even if the protein expression level is low, the protein can be purified to homogeneity under native or under denaturing conditions, according to the company. The QIAexpress system includes products for the detection and assay of 6xHis-tagged biomolecules: antihistidine tag monoclonal antibodies, Ni-NTA conjugates, and Ni-NTA-coated microtiter plates. Ni-NTA is provided in several different formats bound to agarose (Ni-NTA Agarose or Ni-NTA Superflow), bound to silica (Ni-NTA Silica), or as Ni-NTA Spin Columns for protein minipreps. METAL-BASED SYSTEM: Clontech’s TALON Metal Affinity Resin for the purification of 6xHis-tagged proteins Clontech Laboratories Inc. of Palo Alto, Calif., also offers a metal-based purification system, the TALON Metal Affinity Resin. Like nickel-based systems, TALON can separate proteins under denaturing conditions. However, the nickel in other systems can leach from the column, contaminating the protein sample with the metal. To circumvent the problem, Clontech has developed a tetradentate chelator that coordinates the metal ion at all four positions, anchoring the metal to the column. The metal's two remaining coordination sites are oriented away from the solid support, easily accessible to 6xHis proteins as they pass through the column. There is virtually no loss of metal under the recommended conditions, the company reports. TALONspin spin columns can purify 2 mg to 4 mg of 6xHis protein, and Clontech also offers empty disposable polypropylene columns for use with TALON resin in a gravity flow format. Not all commonly used tags are as unassuming as 6xHis. They may interfere with secretion, protein folding, or some function of the protein. And even if the tag causes no problems, some researchers would prefer an unlabeled protein just to eliminate all potential interference in a proposed experiment. The trouble is, tags are not always easy to remove. "Some tags you have to cleave with a protease, and you are left with it in your preparation . . . and the proteases don't [always] cleave with really high fidelity-they may cleave within your target protein," reports Emma Jean Hess, product manager at New England BioLabs Inc. of Beverly, Mass. One solution to that dilemma is the company's IMPACT I System. The IMPACT I uses a protein-splicing agent, called an intein, that-at low temperatures in the presence of thiols-undergoes a self-cleavage reaction at its N-terminus. The investigator inserts the gene encoding the target protein into a multiple cloning site of a vector to fuse the C-terminus of the target gene to the N-terminus of the intein. The intein, in turn, is fused to DNA encoding a 5 kDa chitin binding domain. Crude extracts of cells from an E. coli expression system can then be passed through a chitin column, binding the fused protein. On the column, the conjugate undergoes an intein-mediated self cleavage after overnight incubation at 4¡C in the presence of 1,4-dithiothreitol or 2-mercaptoethanol. As a result, the target protein is released while the intein-chitin binding domain remains bound to the column. Such self-cleaving tags are a clever trick, but they will never replace tried-and-true chromatography approaches, says Gagnon. "What you don't see publicized is that those techniques also bind a whole set of contaminants, and raise a set of [regulatory] problems all their own . . . by the time you're done with cleavage, I don't think it's that advantageous [over traditional chromatography]," he maintains. But he also concedes that in the battle to purify proteins, any approach is welcome: "If it works for you and you're set up for it, it's a good tool. My biggest concern with kits is people trying to substitute them for a real understanding of their protein and how it behaves." But when the basic research ends, and the protein work enters into the development and production processes, "you're out of kit territory . . . you need people who understand the chemistries. There's a huge variation in how they can be applied to achieve results, and that's where you need experience," he remarks. There are two general approaches to scale-up problems: Outsource purification to a contractor, or purchase chromatography equipment and acquire the expertise internally. At purification labs, "new services are brought in quickly, so investigators can spend their time doing research and not developing new methods every time they face a challenging new protein. It also cuts cost because instrumentation isn't being duplicated," notes Thomas A. Grover, assistant director for service labs at Utah State University's Biotechnology Center. The Biotechnology Center's lab has facilities for most purification methods as well as trained technicians dedicated to operating them, explains Grover. Many similar centers exist at universities around the United States, and although they are usually intended to meet the needs of the university's researchers, "companies will occasionally hire [custom labs] to culture the bacterial system, harvest a protein and do a purification, and then propose a purification scheme. The company can decide if it's worth their while to scale up and build the facilities themselves," says Grover. Conducting clinical trials involves new regulatory hurdles. "Companies often contract protein production out so as not to delay clinical trials. We have Good Manufacturing Practices [GMP] facilities, and a lot of organizations-especially research groups, don't want to commit precious capital resources [to meet those standards]," notes David Fischer, director of customer service at Goodwin Biotechnology Inc., a custom protein production facility in Plantation, Fla. GMP refers to the Food and Drug Administration's regulations for the production of drug products. "GMP is not just an instrument, it's a group of instruments and clean-room facilities . . . the number of air changes per hour required to keep the rooms clean is horrendous. There's a substantial commitment of capital to the brick and mortar of the facility and personnel, training, and documentation," Fischer adds. Fischer also points out that the specifications for the finished protein become more stringent as it progresses to Phase II and Phase III trials. Most of its competitors deliver proteins that meet Phase III specifications, says Fischer, but Goodwin Biotechnology concentrates on providing proteins that meet the less stringent requirements of Phase I and II trials. At Phase III trials, the protein-purification process is usually set in stone in preparation for product approval. But in Phase I and II, "we can be a little more nimble. We can bob and weave to help clients meet their needs," explains Michael Battersby, manager of operations for Goodwin Biotechnology. Doing it yourself, of course, has its own set of unique problems. A central headache for process developers is that the protocol a researcher works out on a small scale is very rarely useful on a larger scale. Conventional agarose "soft gels" are made up to 10- to 20-micron beads that partially absorb compounds as they travel a column. However, at higher flow rates, the increased pressure distorts these beads and reduces separation power. To get around the problem, chemists moving toward large-scale production often go to larger-size beads. "But the majority of the [media-interaction] chemistries change with bead size," points out Edward F. Hawkins, product marketing manager for protein purification at Beckman Instruments Inc. of Fullerton, Calif. As a result, all the painstaking work to produce a few milligrams in vitro is usually no help when it comes to scaling up. "Soft gel" packing gained favor because the beads offered a high carrying capacity, but at increasing pressures they compress, and as the shape of a soft bead distorts under very high pressure, its resolution quickly nose-dives. Rigid particles with open pores-an alternative-operate well at high speeds but have limited capacity. SELECTED SUPPLIERS OF PRODUCTS AND SERVICES FOR PROTEIN PURIFICATION: Beckman Instruments Inc. BioSepra Inc. Clontech Laboratories Inc. Goodwin Biotechnology Inc. New England BioLabs Inc. Qiagen Inc. Utah State University Biotech Laboratory Validated Biosystems Inc. To help get around the problem, Beckman has acquired the rights to market a new column packing-dubbed HyperDiffusion-from BioSepra Inc. of Marlborough, Mass. According to Beckman, HyperDiffusion delivers the highest dynamic capacity (or the "real-world" capacity under elution conditions, as Hawkins puts it) of any chromatographic media. Instead of a malleable bead, HyperDiffusion consists of a ceramic with hydro-gel filled pores. The ceramic coating and hydro-gel combine to offer the high-pressure stability of rigid packing with the higher carrying capacity of a gel. Although the media were originally developed for production scale, BioSepra "scaled them down so that Beckman could put them in prepackaged columns, and researchers in drug discovery and methods development can use them. [The columns] offer consistent chemistry all the way from research to methods development . . . to process development," according to Hawkins. Beckman is developing specialized columns using the media, including the recently introduced Protein A HyperD, designed for purification of monoclonal and polyclonal immunoglobulin G antibodies. Protein A appears on the surface of most species of bacteria, and most human IgG antibodies will recognize it. Fused to the HyperDiffusion media, it offers affinity purification at high flow rates and large capacities: The columns can handle a flow rate up to 2,000 cm per hour with binding capacities in excess of 30 mg per ml over the entire flow range, the company claims. SYSTEMATIC APPROACH: Beckman's ProScale protein-purification program includes workstations, software, and chemistries. To ease the research to methods to process transition, Beckman has created what it calls the "first logical, systematic approach to protein purification." The ProScale incorporates the HyperDiffusion media and involves a Windows-based purification simulation software package to model chromatographic separations. "Say you're considering three different ion exchange resins: After initial runs with your protein, you put those values into the simulation engine, and it generates optimized run conditions for you. You can easily change the size of the column, the flow rate, and other parameters. Then you can take those optimal methods into the scale-up package and estimate how much it would cost to take those schemes into production," says Hawkins. WORKSTATION: Beckman’s BioSys 2000 can help fine-tune and automate a separation. Beckman offers three workstations to go with the package. The BioSys 500 series is designed for the research stage, in which it is important to run an established method routinely while delivering accurate buffer gradients from microliters up to 30 ml per minute. When promising research results merit greater quantities of the protein for preclinical studies, Beckman's BioSys 2000 methods development workstation can help fine-tune and automate a separation. At the process-development stage, the ProSys workstation with simulation and scale-up software speeds identification of an optimized method for production, says Hawkins. Whatever stage you're at, there are no magic bullets to shoot down any protein purification challenge. "The real story is that you have to evaluate a lot of options, do a lot of hard work, and make difficult choices. The unique characteristics of the protein and regulatory requirements define the technique-you can't go the other way around. People try, but the protein always wins," says Gagnon. James Kling, a science writer based in Bellingham, Wash., can be reached online at jkling@nasw.org.

Scientists Use Various Methods To Tackle Protein Purification

Interested in reading more?

Become a Member of