(620an) Development of an Automated in-Line Fraction Collection Device for Single-Sample, Multi-Peak 2D-HPLC Protein Characterization

Authors: 
Lewis, A., Johns Hopkins University
Lute, S., FDA/CDER
Brorson, K. A., FDA CDER OPQ OBP
Agarabi, C. D., FDA/CDER/OPQ

The formation of aggregates during the manufacturing process of monoclonal antibodies is an important issue that can lead to problems downstream affecting the safety and efficacy of the drug product.  Thus, analytical tools for monitoring protein aggregation during manufacturing would be a highly desirable component for implementing process analytical technology (PAT) and quality by design (QbD) in the biopharmaceutical industry.  We have established an automated, off-line method for the two-dimensional (2D) chromatographic characterization of the multiple protein peaks that can result due to aggregation.   This method utilizes a unique fraction collection device, herein termed Medusa, which was developed to enable the collection of up to twelve fractions between column one and two.  Medusa consists of two electronically-controlled 12-position/13-port valve systems connected with 100 µl loops, and was placed in line with an Agilent 1200 series HPLC system configured for 2D chromatography.  In our system, column one is Protein A and separation by size exclusion is column two.  This method was challenged using an aggregated model IgG1 protein injected onto the first dimension column.  Samples from multiple protein peaks of interest following purification from column one are fractionated by automated switching of the HPLC valve module to direct flow to the Medusa, and coordinated opening and closing of the Medusa’s valves allowing for the isolation of these time-based fractions in the connected loops.  Each sample can then be sequentially loaded onto the second chromatographic column by sequentially opening the Medusa valves.  By placing the Medusa in-line with the 2D-HPLC in this way, we facilitate the timely analysis of multiple protein peaks from column one in a rapid, high-throughput manner.