Use and Optimization of a Dual-Flowrate Loading Strategy To Maximize Throughput in Protein-A Affinity Chromatography Sanchayita Ghose, † Deepak Nagrath, ‡ Brian Hubbard, † Clayton Brooks, † and Steven M. Cramer* ,‡ Purification Process Development, Amgen Inc., Seattle, Washington, and Department of Chemical & Biological Engineering, Rensselaer Polytechnic Institute, Troy, New York 12180 The effect of an alternate strategy employing two different flowrates during loading was explored as a means of increasing system productivity in Protein-A chromatog- raphy. The effect of such a loading strategy was evaluated using a chromatographic model that was able to accurately predict experimental breakthrough curves for this Protein-A system. A gradient-based optimization routine is carried out to establish the optimal loading conditions (initial and final flowrates and switching time). The two-step loading strategy (using a higher flowrate during the initial stages followed by a lower flowrate) was evaluated for an Fc-fusion protein and was found to result in significant improvements in process throughput. In an extension of this optimization routine, dynamic loading capacity and productivity were simultaneously optimized using a weighted objective function, and this result was compared to that obtained with the single flowrate. Again, the dual-flowrate strategy was found to be superior. Introduction Monoclonal antibodies have emerged as an important and rapidly expanding class of drugs for the treatment of several human diseases. Approximately 30% of bio- pharmaceuticals currently in clinical trials are antibodies and Fc-fusion proteins (1). Given the large scale at which many of these antibodies are produced (hundreds of kilograms/year), it is no surprise that the commercial success of these biomolecules hinges on the development of efficient and economic downstream processes. Protein-A affinity chromatography has come to be used routinely in industry for the direct capture and purifica- tion of monoclonal antibodies and Fc-fusion proteins from complex cell culture media (2-4). The large mass re- quirements of antibodies coupled with the high cost of Protein-A media have made binding capacity and through- put on Protein-A primary concerns because they govern process economics in commercial scale purification pro- cesses. The development of a Protein-A affinity chromatogra- phy process typically involves consideration of several factors. Many commercial Protein-A media are available that vary with respect to their backbone matrix, bead and pore size and the source of the Protein-A ligand. Typically, choice of the Protein-A resin depends on the best compromise between loading capacity, eluate purity and flow characteristics of the stationary phase. For a particular stationary phase, efforts are often made to maximize dynamic capacity in order to allow efficient and maximal utilization of the expensive Protein-A media. Typically, Protein-A media have a finite lifetime defined in terms of the maximum number of column reuses that are possible. Given the finite lifetime, increasing the dynamic loading capacity on the column reduces the resin cost per unit quantity of antibody purified. Another important consideration in bioprocessing is production rate or throughput (5, 6). Production rate can be especially important for Protein-A chromatography. Because the media is very expensive, rather than using a large column to process a batch of antibody in a single cycle, typical bioprocess applications run a smaller column for several cycles to purify a single batch. This reduces the risk of capital cost if the column is compro- mised during operation and also brings the column diameter into a practical range. Cycling increases the total purification time and thereby decreases the produc- tion rate. Thus, processing time can be an important factor in Protein-A step development. Fahrner et al. (7) discuss the importance of considering the optimal flow- rate on Protein-A and suggest that higher flowrates will reduce process time without significantly affecting pro- cess capacity. Processing time has been mentioned to be critical to process development for three reasons (8). First, if purification is the limiting factor in a production facility, then a direct improvement in process time will increase throughput. Often, Protein-A is the first capture step and is often the rate-limiting step as a result of the large volume loads and low protein concentrations as- sociated with the process. Second, the product stability in the harvested cell culture fluid can limit allowable hold and processing times. Third, cell culture fluid is a rich medium that can promote an increase in bioburden. Minimizing processing time can help to decrease biobur- den contamination. The authors (8) even suggest using a resin with a slightly lower binding capacity but better flow characteristics to enable a decrease in the overall processing time. Fahrner et al. (6) have compared the performance of several Protein-A resins using both volumetric (amount of protein purified per unit time per unit column volume) * To whom all correspondence should be addressed. Phone: (518) 276-6198. Email: crames@rpi.edu. † Amgen Inc. ‡ Rensselaer Polytechnic Institute. 830 Biotechnol. Prog. 2004, 20, 830-840 10.1021/bp0342654 CCC: $27.50 © 2004 American Chemical Society and American Institute of Chemical Engineers Published on Web 02/14/2004