Expanded-bed adsorption

The development of molecular biology techniques has enabled researchers to generate large quantities of biologically important molecules from alien sources—transgenic bacteria, plants, and animals. Although this ability has revolutionized the production and delivery of pharmaceutical and therapeutic products, one problem has remained: All of the protein in the world won’t do you any good if you can’t purify it.

Anyone who has purified protein from bacteria will recall the hassles of lysing bacterial pellets. Regardless of whether you are performing enzyme- or detergent-based lysis, or physical cell disruption, there is still the problem ofseveral rounds of lysis and solution clarification. Even at this stage, many chromatographic resins are prone to clogging, and sample feedlots may require further rounds of clarification by ultracentrifugation or ultrafiltration. And with each stage of purification, not only is valuable time lost, but also valuable sample, and yields can drop dramatically in these first few steps. What chromatographers need is a method that combines sample preparation with the first stage of chromatography. Welcome to expanded-bed adsorption (EBA) chromatography.

Technology
EBA uses apparatus that is familiar to most users of standard liquid chromatography. The column has a flow adapter that is positioned to suit the specific step of resin preparation or protein purification. And a series of pumps and valves, connected through the adapter and bottom of the column, control the flow rate and direction of the buffer and sample loading (1, 2). Thus, it is feasible to perform preliminary EBA trials with a little ingenuity and standard chromatographic equipment. However, for those wishing to pursue the technology more aggressively, the two main EBA vendors are Amersham Biosciences (Uppsala, Sweden; www.amershambiosciences.com) and UpFront Chromatography A/S (Copenhagen, Denmark; www.upfront-dk.com).

In the more traditional packed-bed methods, where the resin is confined between the bottom of the column and the flow adapter, clogging occurs when particulate matter and cell debris cannot flow around the closely packed resin beads. In contrast, EBA columns are fed from below, and the adapter is held away from the packed resin level, giving the resin room to expand and thus creating spaces between the beads.

The mechanics of EBA
When the resin is packed in the column, the beads sit close together and leave little room for large aggregates and clumps to maneuver (Figure 1-1). Under typical chromatographic conditions, where the flow comes from the top of the column, this would lead to a buildup of particulate matter on the resin surface, eventually clogging the column.

As buffer is injected from below, the resin becomes fluidized and the beads form a stable concentration gradient when their sedimentation velocity equals the upward liquid flow velocity (Figure 1-2). The goal is to achieve a flow rate that exceeds the terminal velocity of the feedlot particulates but not that of the resin beads.To accentuate this difference, many standard resin beads have been modified to include inert quartz or metal alloy cores, depending on the application and source.

Unlike traditional resins, in which the beads tend to be uniform in size,the beads of EBA resins are variable, typically ranging from 50 to 400 µm. Thus, the larger particles populate the lower portion of the fluidized bed while the smaller particles populate the upper portions. If the beads are too small, expansion will occur at velocities comparable to the escape velocity of the particulate contaminants, lowering purification efficiency. Likewise, if the beads are too large, fluidization requires higher flow rates and protein binding is impaired because of improper diffusion among the beads.

As the sample feedlot is injected, the particulates and cell debris move freely around the resin beads and eventually leave through the top of the column. As with any chromatographic step, the resin then undergoes strenuous washing to limit nonspecific interactions between the particulates and resin. Meanwhile, the compounds of interest interact with the beads and are retained on the column (Figure 1-3). The column is then allowed to pack, the flow is reversed, and the compound is eluted from the beads as in traditional methods (Figure 1-4).

It is this last step that separates EBA from both batch-mode and fluidized-bed adsorption chromatography. In the latter methods, the compounds of interest can be eluted from the resin only in a batch- or stepwise manner, so there is little or no resolution of eluted molecules.By contrast, a packed column allows the use of elution gradients and improves elution peak resolution.

“The advantage is a higher recovery,” says Guenter Jagschies, vice president of industrial separations for Amersham Biosciences. “That’s where the money comes out for the customer.” Jagschies says that EBA users often see a 25% increase in their recovery. “That is very important,” he continues, “because we are talking about the first step in the downstream process, and you can never have more product after any later step. You always lose something.”

Process applications
As scientists scale up biopharmaceutical production, however, they are quickly reaching a point at which they will have two options. They must either install new fermentation capacity or increase the productivity of their cell lines. “This is an issue in the monoclonal antibody business,” says Jagschies, “where in 3 or 4 years’ time, we are not going to have enough production capacity in terms of fermenters to meet the demands from the biopharmaceutical industry.”

According to Jagschies, recent advances in increasing the monoclonal antibody (mAb) productivity of Chinese hamster ovary cells from a few hundred micrograms per liter to 1 g/L have been paralleled by increasing problems with cell death and the contamination of mAb solutions with cellular proteases. This decreases the final yield, purity, and homogeneity of the mAbs. “People have found that by using EBA very early in the process,” adds Jagschies, “they can capture these proteases and can reduce these kinds of problems. They actually got purer products as compared to [using] packed-bed columns.”

Beyond mAb purification, Kaoru Kobayashi of Yoshitomi Pharmaceutical Industries recently reported on the purification of pharmaceutical-grade recombinant human serum albumin (rHSA) that had been expressed in the methylotrophic yeast Pichia pastoris (3). HSA is used in the treatment of severe hypoalbuminemia and traumatic shock, but it suffers from the fact that it is purified from whole blood and may thus contain bloodborne contaminants such as HIV. Expressing rHSA in a nonhuman system, however, eliminates these contaminants. Similarly, Jörg Thömmes and colleagues at the Heinrich-Heine University and Research Centre Jülich in Jülich (Germany) and Mucos Pharma GmbH and Co. in Geretsried (Germany) used EBA as part of an integrated fermentation and purification process for the isolation of recombinant human chymotrypsinogen B expressed in P. pastoris (4).

Nonprotein applications
EBA also may have a large impact in the preparation of pharmaceutical-quality DNA for nonviral vaccines and gene therapeutics (5). Owen Thomas and colleagues at the Centre for Process Biotechnology in the Technical University of Denmark (Lyngby; www.ibt.dtu.dk/cpb/) worked closely with developers at UpFront Chromatography to explore the use of EBA in the separation of plasmid DNA from the shear-sensitive solids that result from the lysis of microbial cells. The solids include flocculant material from the cell membranes and organelles as well as host nucleic acids.

The Danish researchers developed a series of polyethylene imine- and diethylaminoethyl-linked agarose resins that bind DNA more tightly than commercial anion exchange resins, although at low ionic strength they have experienced some problems of individual DNA strands binding to more than one resin bead, effectively cross-linking the beads.

Further, the researchers developed a plasmid-specific affinity resin by linking avidin to the resin bead and biotin to an oligonucleotide, the sequence of which matches a portion of the target plasmid. By first forming a triple helix between the biotinylated oligonucleotide and the plasmid, the researchers can draw the plasmid from solution through the binding of biotin by avidin.

Nonprocess applications
Perhaps the most interesting use of EBA is not in the isolation of potential therapeutics but in diagnostics and epidemiology. Isabelle Accoceberry and her colleagues in Bordeaux (France) recently reported on the use of immunoaffinity EBA in the isolation of parasite spores from human stool samples (6). The researchers created a mAb that bound a protein on the surface of the spores of Enterocytozoon bieneusi, an opportunistic protistan parasite that causes diarrhea, malabsorption, and weight loss. Conjugating the mAbs to a Streamline rProtein A adsorbent, the French group isolated spores from numerous human stool samples, detecting the spores by an indirect immunofluorescence antibody test.

“The ability to obtain a pure population of whole mature spores,” wrote Accoceberry, “free of all fecal contaminants, should facilitate biological, biochemical, and immunological studies of the infective stage of E. bieneusi.”

Future developments
At the same time that researchers continue to find new and exciting ways to apply EBA technologies, system engineers are hard at work trying to expand the properties of the EBA hardware. According to Jagschies, Amersham Biosciences hopes to address some of the other issues in the generation of products, such as making EBA more suitable for the loading of samples directly from fermenters and increasing the volumes that can be run through the system. In part, achieving these goals will require increasing the robustness of the system to make it more accessible to process engineers who are more accustomed to filtration.

Amersham Biosciences is also trying to expand the repertoire of resin matrices. The company is looking to create variations on the ion-exchange resins that will allow more specific binding of products and limit the effects of salt concentration. Beyond this, the future seems to lie in an expanded line of affinity chromatography matrices.

While not necessarily the light at the end of the bioprocessing tunnel, EBA certainly looks to become a popular purification method with both the bench scientists who have to clarify a sample and the laboratory managers who have to clarify the bottom line.

Further reading

1. Mattiasson, B.; Nandakumar, M. P. Physicochemical Basis of Expanded-Bed Adsorption for Protein Purification. Handbook of Bioseparations, Vol. 2: Separation Science and Technology; S. Ahuja, Ed.; Academic Press: San Diego, 2000.
2. Shiloach, J.; Kennedy, R. M. Expanded-Bed Adsorption Process for Protein Capture. Handbook of Bioseparations, Vol. 2: Separation Science and Technology; S. Ahuja, Ed.; Academic Press: San Diego, 2000.
3. Kobayashi, K. Downstream 2000, 31, 5.
4. Thömmes, J.; et al. Biotechnol. Prog. 2001, 17, 503–512.
5. Ferriera, G. N. M.; et al. TIBTECH 2000, 18, 380–388.
6. Accoceberry, I.; et al. J. Clin. Microbiol. 2001, 39, 1947–1951.