Accounting for host cell protein behavior in anion-exchange chromatography

Area of Science:

• Biotechnology
• Chemical Engineering
• Protein Chemistry

Background:

• Host cell proteins (HCPs) are critical impurities in recombinant biologic production.
• HCPs complicate downstream processing (DSP) due to similar separation behaviors to target proteins.

Purpose of the Study:

• Characterize the separation behavior of HCP mixtures during anion-exchange chromatography (AEX).
• Evaluate a statistical methodology for predicting protein separation behavior in DSP.

Main Methods:

• Aqueous two-phase partitioning and 2D electrophoresis to determine HCP physicochemical properties (pI, molecular weight, hydrophobicity).
• Multivariate random forest (MVRF) model applied to characterize separation data from corn germ and E. coli extracts.
• Root mean squared error used to validate the predictive accuracy of the MVRF method for AEX.

Main Results:

• Physicochemical properties of HCPs from corn germ and E. coli were characterized.
• An MVRF model was developed to predict protein separation behavior during AEX.
• The MVRF method demonstrated high accuracy, with a root mean squared error not exceeding 0.045 for AEX.

Conclusions:

• Characterizing HCP separation behavior is crucial for optimizing downstream processing.
• The developed MVRF statistical methodology effectively predicts protein separation in AEX.
• This predictive tool can enhance the production of high-purity recombinant biologics.