YouTube Facebook LinkedIn Google+ Twitter Xingrss  

High-Throughput Biologics Production

By Robert M. Frederickson

Sept 15, 2005 | High-throughput science has generally been the domain of drug screening and any of a variety of discovery-based “omic” endeavors. However, miniaturization and massively parallel analysis are finding a separate niche at the other end of the drug development pipeline — in upstream process development and production.Part of this change has to do with the increase in number of protein-based drugs — such as recombinant proteins and monoclonal antibodies. Although there are perhaps only 10 to 12 protein drugs on the market today, there are a few hundred in various stages of clinical development.

Years ago, a drug was a small molecule that could be chemically synthesized in a test tube. If the drug was a success and was to be marketed, you simply scaled-up the synthesis from test tube to reaction tank. Unfortunately, the production of proteins is much more complicated and tends to take place in cultured animal cells that have been engineered to produce the particular drug of interest. The genetic background of the cell is key to producing reasonable quantities of the protein drug with the appropriate posttranslational modifications.

The problem is that there is a myriad of factors that need to be optimized for the production of pharmaceutical-grade proteins produced in cultured cells — from the pH to the temperature to the composition of the medium used. Testing of these conditions in sufficient numbers to obtain reliable data can quickly become tens of thousands of experiments, involving tens of thousands of culture flasks, which is clearly impracticable from both the standpoint of time and money. Moreover, once the appropriate conditions have been identified, transferring them from a culture flask to a 5,000-gallon reaction tank is not necessarily straightforward or even possible.

Massachusetts-based BioProcessors has addressed this issue with the introduction of the SimCell MicroBioreactor — essentially a microarray of miniaturized bioreactors in a 96-well plate format. Made of a proprietary plastic material, the device contains embedded sensors for pH, CO2, etc., and molded microfluidic channels that allow real-time sampling of the culture medium. Two hundred and ten of these microbioreactors can be used simultaneously in the BioProcessors SimCell Automated management system, which can inoculate cultures, perform maintenance feeds of reagents, and monitor metabolites and biomass in an automated fashion. This translates into 1,260 benchtop bioreactor experiments in two weeks — a roughly 100-fold increase in throughput. Moreover, each microbioreactor has a capacity of 200-800 μL, which minimizes material use.

Importantly, users can design and execute experiments with a software interface and monitor cell counts and culture conditions in real time. Data management is also automated and integrated. While the system is available for purchase and integration into existing drug development facilities, it can also be accessed via contract research services provided by BioProcessors, which may be attractive for smaller enterprises with more limited needs. Both Amgen and Novo Nordisk are currently using the SimCell system, with another medium-sized pharmaceutical producer expected to join their ranks.

SimCell cuts the cost of a bioreactor or shake flask data point dramatically. Cost estimates for running benchtop bioreactors or shake flasks range from $2,000 to $4,000 or $200 to 400 each, respectively, where the equivalent cost for a microbioreactor data point is in the order of $150.

The SimWare system has been developed such that is can easily archive massive amounts of data. It is also possible to import data from “offline” systems such as bioanalyzers, production reactors, benchtop systems, and even shake flasks. With the availability of all this information about a single expression system in one informatics platform, the researcher can make cross-system, cross-facility, and cross-time comparisons, download data back to statistical analysis software packages, and conduct detailed data mining in an effort to extract all available knowledge. The potential benefit is to advance in general an organization’s general know-how of cell culture, help facilitate remote facility collaborations, and ultimately support the filing of regulatory approval documents, the implementation of process analytical technology to cell culture processes, and the ultimate implementation of truly “excellent manufacturing” processes.

E-mail Robert M. Frederickson:


View Next Related Story
Click here to login and leave a comment.  


Add Comment

Text Only 2000 character limit

Page 1 of 1

For reprints and/or copyright permission, please contact  Terry Manning, 781.972.1349 ,