Model-based rational methodology for protein purification process synthesis

Abstract

A model-based rational protein purification process synthesis methodology that addresses the challenge of selecting the most optimal process scheme from several possible alternatives is presented in this study. The main rationale behind the methodology is to keep the number of purification units in the final process to the minimum through a systematic cycle of flowsheet synthesis, optimization, evaluation and the rational elimination of the least feasible process options at each purification step, taking into account the specific needs of the purification step. Process evaluation is based on techno-economic performance obtained by model-based optimization of the integrated processes using validated column models. The methodology was illustrated by synthesizing a process for the purification of monoclonal antibody from crude hybridoma cell culture supernatant using four non-affinity chromatographic methods (AEX, CEX, SEC and HIC). The results showed that four out of thirteen evaluated process options satisfied all pre-defined product specifications, with overall yields ≥90%. In order of increasing product cost, these were CEX–AEX<AEX–CEX<AEX–HIC<HIC–AEX, taking into account the need for inter-stage conditioning. The main advantages of the methodology include savings in experimentation, computational time and effort; unbiased evaluation of purification schemes by using their ideal operating conditions; and the consideration of the specific needs of each purification step during process evaluation. Finally, the methodology is generic for cell-derived proteins, irrespective of the host organism or application.