With the advent of new non-platform biomolecules, designing the downstream process (DSP) has become increasingly difficult. Knowledge based techniques and tools, including mechanistic and statistical models, have shown a strong potential for reducing the difficulty in designing the purification steps for these molecules. Over the last year, we have created such tools, allowing for the in depth portrayal of the different steps involved in DSP. These tools are operated within the context of work packages allowing, in a first instance, to easily and quickly model each step individually. The different work packages are then combined to create full process models. In this talk, we discuss these different tools and techniques in detail. First, the chemistry of the different solutions involved in the downstream process is explained. This allows for the description of the charge of all proteins and of the pH, conductivity, ionic strength of all solutions, streams and unit operations in the process. This information, in conjunction with mechanistic models, is then utilized to provide in depth descriptions of the several chromatographic, buffer exchange and conditioning steps involved. Further, all the developed techniques and tools are presented alongside design and validation experiments. Using the ability to describe every unit operation and stream, the entire process is then optimized, as a whole. This optimization can be further enhanced by considering individual step performance constraints (i.e. product pool conductivity following a polishing step). Further, the robustness of the designed process and its sensitivity to potential changes in process parameters is analyzed. Of course, in early phase design, the finality of such tools is not a perfect understanding of the process or perfectly fitting mechanistic models, but achieving an optimally designed process easily, quickly and with minimum risk. In fact, all design experiments are meant to be useful even without the proposed approaches. This allows us to minimize the time and risk involved in process design while maximizing the process knowledge and understanding achieved.

Using knowledge for downstream process design

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.