(326c) Use of Expanded-Bed Chromatography (EBA) as An Alternative to Clarification and Capture Steps in Downstream Processing for Fc-Fusion Proteins and Antibodies

Authors: 
Xu, X. - Presenter, Bristol-Myers Squibb
Epting, K. L. - Presenter, Bristol-Myers Squibb
Hickey, J. L. - Presenter, Bristol-Myers Squibb
Ghose, S. - Presenter, Bristol-Myers Squibb


Preparative EBA utilizes a fluidized bed in a bottom-up flow direction to process crude culture broths and combines clarification and capture into one unit operation. Elimination of the clarification steps which normally include centrifugation and depth filtration for protein purification can improve yield and productivity, and reduce the overall downstream cost. As some of its well-known shortcomings are being properly addressed by recent progresses in resin and hardware designs, EBA is drawing increasingly more attention in downstream processing for biotherapeutics. The work presented here first characterizes packed-bed performance of two EBA resins which are tungsten carbide densified agarose resins developed using unique synthesis method to achieve suitable particle density and structure. Using an Fc-fusion protein and two antibodies, resin capacity, selectivity and other properties are discussed, as compared to MabSelect Protein A. Then, the Protein A resin was investigated using clarified and unclarified harvest materials in two EBA columns which differ only in column configurations. Results show that the chromatographic performance of the resin is very similar in packed bed and EBA columns in terms of impurity clearance and leached Protein A level. The similarity seen between the two modes allows most of the condition screening for EBA to be performed quite efficiently in packed bed columns. In addition, the breakthrough profile in EBA is sensitive to settled bed height (SBH) and indirectly associated with residence time (RT). By using an industry scale-down EBA column (1 cm i.d. × 50 cm SBH), preparative EBA runs reproducibly demonstrated a dynamic binding capacity comparable to that of a packed-bed column, with slightly lower CHOP reduction due mainly to longer RT in the EBA runs. The high level of consistency of different runs using various load materials (e.g., with and without biomass) indicates very specific binding of target proteins on the resin. Cell clearance capability in EBA is satisfactory at properly controlled bed expansion. The implications of EBA to downstream cost and efficiency will also be discussed using an example.