July 1, 2010 (Vol. 30, No. 13)
C. Patrick McAtee, Ph.D.
Displacement Chromatography for Biopharmaceutical Process Development
The generally accepted method of purification of process streams for monoclonal antibodies includes capture of the product target with protein A, elution, and acidification to inactivate potential mammalian viruses, followed by cation-exchange chromatography and finally anion-exchange chromatography. There are some variations by manufacturers, which may include an intermediate hydrophobic interaction chromatography step.
As monoclonal antibodies are relatively similar from a structural perspective, purification processes tend to be somewhat consistent from manufacturer to manufacturer. However, with each processing step, the potential for heterogeneity is nonetheless introduced.
A recent report by the CMC Biotech Working Group, a consortium of companies that has prepared a case study on the development of a monoclonal antibody, outlined the steps involved in producing the monoclonal and the implementation of quality by design (QbD) concepts to optimize these steps and simultaneously control or eliminate process variation. So how can one identify and control these process variations—particularly structural variants—and simultaneously deliver a high-quality product with excellent yields?
Displacement chromatography (DC) is a method in which the components are resolved into consecutive rectangular zones of highly concentrated pure substances rather than solvent-separated peaks. The molecules are forced to migrate down the column by an advancing wave of a displacer molecule that has a higher affinity for the stationary phase than does the feed solute. Because of this forced migration, higher product concentrations and purities may be obtained compared to other modes of chromatography. There are distinct differences between displacement and elution chromatography.
DC has advantages over elution chromatography because the process takes advantage of the nonlinearity of the isotherms. A higher mass loading can be separated on a given column and the purified components recovered at significantly higher concentrations.
DC is no more complicated than elution chromatography. However, operating parameters are different, particularly in respect to optimization where displacer concentrations, flow rate, and protein load, are critical.
DC requires high loadings in order to set up a good displacement train. One typically uses loadings of 50% to 80% of the maximum resin capacity. If one does not have enough starting material to load at this level, then it is advisable to switch to a narrower column. Shorter columns will work in DC but recoveries will be lower.
Owing to the convergence of high loading, high recovery, and high purity of DC methods, one can obtain higher throughput per cycle, higher purity, and increased concentrations from each use of the column. This is useful if one needs to purify larger amounts of material and has the added advantage of achieving this with bench-scale analytical columns. However, the other advantage of DC is the ability to isolate trace components and concentrate them while simultaneously removing the main components.
Typically when producing antibodies with mammalian cell culture, structural and charge variants are encountered, which could potentially impact stability and activity of antibody. Many of these alterations are problematic to separate from the intended product and heroic methods are often utilized to obtain enough material to evaluate these discrepancies in pharmacodynamic studies. The thermodynamic properties of DC allow for enrichment of low concentration components and are applicable to isolating and purifying these charge variants.
DC in Action
Sachem recently carried out a study to demonstrate the benefits of DC. In Figure 1 a monoclonal antibody process stream is captured utilizing strong cation exchange. Material is initially pH adjusted following the common protein A or G fractionation step and loaded onto a cation exchange column at pH 4.2–5.2.
In this instance, we chose to run our material on a column with a greater than 50:1 aspect ratio. Flow rate was kept low (0.2 mL/min), which minimized back pressure. Laboratory instrumentation can be a typical analytical HPLC system.
Important components in our set up include a Model 3082-S conductivity detector (Amber Science) and a low-volume flow cell (Knauer) after the UV detector. Sample was loaded onto an equilibrated Resource S 15 column (GE Life Sciences) followed by 1 column volume of loading buffer as a wash.
Sachem Expell SP-1 displacer (in loading buffer) was then loaded at 5 mM and the column was monitored at 280 nm (with additional monitoring at 250 nm for tracking the displacer). The displacer is then removed with 4–18 column volumes of Sachem Regenerate. The column is re-equilibrated for the next run by washing with 6–12 CV of loading buffer with 2.0 M NaCl. The column is stored in a regeneration buffer until next use.
The major peak displaced is typically the desired monoclonal antibody with the variants running immediately prior to and after the main peak. The minor variants are in effect “squeezed” into sharp peaks, which can be further isolated at high recovery for subsequent animal studies.
Two-dimensional chromatography represents the most thorough and rigorous approach to evaluation of the proteome. Variants can readily be identified and lot comparisons made by combining second- dimension chromatography with DC as the first dimension. In our case, the displacement fractions are run neat or diluted in DI water. The reverse-phase solvents are by convention installed on the HPLC channels A and B. The A solvent by convention is the aqueous solvent (water) and the B solvent by convention is the organic solvent (acetonitrile, methanol, propanol). We used a XBridge C18 column (Waters) 250 x 4.6 mm at 1 mL/min flow rate with a gradient from 10% B to 30% B over 40 min followed by a gradient from 30–100% B over 20 minutes.
Figure 2 shows the two dimensional results from the DC run in Figure 1 that have been chromatographed in a second dimension using subsequent elution separation as the chromatography mode. In this 2-D chromatogram, the location of the acidic variant is noted as well as the main antibody peak and basic variants. These peaks may be further isolated and evaluated by or, in the case of secondary reversed phase elution chromatography, may be run directly on-line to a mass spectrometer.
Product heterogeneity is common to monoclonal antibodies and other recombinant biological production and is introduced either upstream during expression or downstream during manufacturing. These variations in the target product are known to affect protein stability and hence, product efficacy. DC approaches enable the investigator to identify and characterize these often previously unseen variants in quantities that are suitable for subsequent preclinical evaluation regimens such as animal pK studies.
Knowledge gained during the preclinical development phase is critical for enhanced understanding of product quality and provides a basis for risk management and increased regulatory flexibility. The QbD initiative attempts to provide guidance on development and to facilitate design of products and processes that maximize efficacy and safety profile while enhancing product manufacturability.
DC can therefore become an enabling technology to compare process variation from one manufacturing step to the next. It allows for the preparative separation and testing of variants that could be present issues in the clinical setting. This information would be essential in development of an adequate QbD program.