October 15, 2014 (Vol. 34, No. 18)

Rivals Need to Demonstrate Better Cost and Mechanical Robustness Capabilities

Protein remains at the top of methodologies for capturing or otherwise isolating monoclonal antibodies from bioprocess streams. Although resins have much higher capacities and lower sensitivity to cleaning agents than a decade ago, issues of cost, ligand leaching, and mechanical robustness remain.

The view of GE Healthcare, other top resin manufacturers, and customers is that Protein A-based affinity chromatography is currently the most economically practical capture methodology for commercial purification of monoclonal antibodies.

“We have investigated and even patented alternative and complementary technologies such as aqueous two-phase separation, and have supported customers who prefer to purify mAbs without Protein A,” says Jonathan Royce, bioprocess senior product manager, antibody affinity media, GE Healthcare Life Sciences.

“But we do not broadly recommend these practices due to process economy and time- to-market considerations. At the same time, in addition to looking at ways to improve Protein A technology, we are investing in the development of newer affinity resins, for example those based on Protein L.”

According to Royce numerous researchers have and continue to investigate alternatives to Protein A capture, including expanded bed adsorption, crystallization, and precipitation. However, in terms of yield, capacity, and ease of scaleup, operation, and process development, Protein A offers significant advantages leading to overall better process economy and shorter development times for new molecules.

“Alternatives such as Protein A synthetic resins show lower selectivity, affinity, as well as nonspecific adsorption and potential differences in selectivity between two antibodies,” Royce observes.

“In addition, adsorbents such as hydrophobic charge interaction chromatography might be more difficult to apply to mAb purification.”

GE Healthcare Life Sciences’ ÄKTA avant is a preparative chromatography system designed for fast and secure development of scalable methods and processes. ÄKTA avant is available in two versions—ÄKTA avant 25, which is optimized for media screening and method optimization using small columns, and ÄKTA avant 150, which is designed for scaling up to larger columns, as well as fine tuning and robustness testing of the optimized process.

Promising Methodology

Nevertheless, alternatives to Protein A might be a promising methodology for the purification of a next-generation of biotherapeutics, including antibody fragments. Alternative technologies will have to offer similar advantages in terms of high selectivity, long lifetime, and ease of developing platform processes.

Royce notes that not all IgG types bind to Protein A, and in these cases alternative purification techniques may be required. Furthermore, new classes of antibody-based therapies like antibody fragments and bispecific antibodies may lack the Fc region for which Protein A has an affinity. “In these cases, alternatives such as Protein L or recombinant affinity ligands may be required,” Royce adds.

As with any process-development activities, investigations into alternatives to Protein A are best conducted early in a product’s lifecycle. Otherwise the risk of expensive changes, including process revalidation may easily erase any savings gained from developing alternative processes.


Alternatives to Protein A capture include precipitation, crystallization, cation exchange, and mixed-mode chromatography. One type of mixed-mode separation—hydrophobic charge induction chromatography—employs charge and hydrophobicity in the stationary phase. Despite promises of equal or better affinity to monoclonal antibodies (and non-mAb therapeutic proteins as well), and of course lower cost, Protein A resins remain the gold standard.

“Protein A’s highly specific interaction with antibodies’ Fc region is difficult to reproduce,” says Melissa Holstein, Ph.D., applications engineer at EMD Millipore. “While many alternative capture technologies are being investigated, I am not sure if there are any with serious potential to disrupt the industry.”

From the perspective of a likely entry point within a product’s lifecycle, alternative capture techniques are in a very difficult spot. For established processes, technical inertia and aversion to regulatory risk ensure Protein A’s pre-eminent status for the product’s entire lifecycle.

Sponsors of early-stage molecules are equally unlikely to adopt alternatives because they know Protein A works; companies do not enjoy the luxury of working on science projects when time-to-market is everything. The “platforming” of downstream operations has streamlined development activities and would in fact not be possible without confidence in Protein A capture.

“Even if alternatives are a fraction of the cost of Protein A, it is unlikely that the same level of overall productivity and product quality can be achieved with them, time after time,” adds Dr. Holstein.

That resin manufacturers continue to improve Protein A is another factor working against alternatives. The next-generation Protein A products, like EMD Millipore’s Eshmuno® A and GE’s MabSelect Sure, are alkaline-resistant and claim up to 200 cleaning cycles, a significant improvement over first-gen resins, with very low ligand leaching.

Another Perspective

Vendors of traditional capture resins make excellent points. Scientists may thrive on innovation, but in biopharmaceuticals they prefer novelties that bring new molecules to the clinic faster, and with less regulatory uncertainty, than on new separation methodologies.

Yet the potential for “disruption” in large-scale separations continues to intrigue. Affilogic, for example, has had a remarkable string of successes with its Nanofitin® purification platform. Nanofitins are recombinant proteins that serve as proprietary affinity ligands to a host of biomolecules. Their high affinity makes Nanofitins suitable for targeting, detection, and, more apropos to this article, affinity chromatography.

Because they are derived from ligands on extremophile bacteria, Nanofitins have remarkable pH (1–13) and temperature (up to 85° C) stability.

“A hundred percent success rate is of course impossible, but Nanofitins have worked on each of the 35 projects we’ve worked on,” says Olivier Kitten, Ph.D., CEO.

Affinity to IgGs of types 1, 2, and 4 are very strong. IgG3 binding is somewhat weaker, “but it might suffice for purification,” adds Dr. Kitten. Moreover, because they do not rely on conventional binding to the Fc region, they can differentiate between isoforms, for example phosphorylated and nonphosphorylated proteins.

Affilogic designs Nanofitins for specific molecules by screening the compound against a library of affinity molecules, a process that Dr. Kitten describes as “brute force.”

The ability to tailor Nanofitins for specific binding sites allows design of affinity ligands to broad molecular classes, e.g., the Fc region on mAbs and broad antibody capture capability. “But we have designed other Nanofitins for affinity to specific regions on customers’ molecules, which eliminates binding to other antibodies,” Dr. Kitten tells GEN. “The effect is tunable.”

He points out that Nanofitins are expected to “cost less” than Protein A: “We could not imagine them costing more.” But cost, as he notes, is a tricky concept for complex biomanufacturing campaigns. “The principal benefit,” he says, “will be specificity,” which translates to higher quality and perhaps yields as well.

Nanofitins will also find application for antibody fragments that might be deficient in Fc binding capability.