Thursday, October 7, 2010

The Business of Recombinant Protein Production

Since the development of recombinant technology in the late 1970s, the use of recombinant proteins in therapeutics has become an attractive strategy for altering the biology of disease progression. In the early 1980s, recombinant human insulin from Escherichia colibecame the first recombinant pharmaceutical to enter the market. Development of several growth hormones produced in bacteria followed, as did production of monoclonal antibodies such as infliximab and ritxumiab from mammalian cell lines. By January of 2009, the number of protein-based recombinant pharmaceuticals licensed by the U.S. Food and Drug Administration and European Medicines Agency reached 151.1
 Although recombinant protein technology provides a wealth of commercial opportunities, bringing a recombinant protein to the marketplace requires a substantial investment of time and resources. Because the synthesis and purification processes are technically complex and vary with each protein, estimating resources and timelines can be challenging at best. Yet for small companies, meeting these challenges is vital, says Vladimir Kostyukovsky, senior technical manager at Kymanox, Inc. “With start ups, you need to hit your milestones by a certain date, or you’re out of business.”
The most common expression systems for recombinant pharmaceuticals are derived from bacteria (usually E. coli), yeast (usually Saccharomyces cerevisiae), or mammalian cells. Key properties of individual expression systems affect convenience and quality of production. Characteristics inherent to a target protein also affect production capabilities. Making the transition from producing a laboratory stock of protein to commercial quantities also requires strategic planning and flexibility.
Protein synthesis and purification Microbial systems versus mammalian systems
Recombinant protein production in microbial systems (mainly bacteria or yeast) tends to be faster and cheaper than in mammalian systems. The robust cellular structure of bacteria and yeast make them amenable to culturing, but mammalian cells—derived from multicellular organisms—are not adapted to survive outside the body and are thus sensitive to external conditions like shear stress and osmotic shock. Bacteria and yeast replicate every 20 minutes to 2 hours, while mammalian cells do so every 24 hours to 2 days. Mammalian cells require complex—often proprietary—media with amino acids. Bacterial cell medium consists of simple carbon and nitrogen sources and inorganic salts. The cost for bacterial cell media is 90% lower than for mammalian Chinese hamster ovary (CHO) media. Fermentation or bioreactor runs take 24 to 72 hours in bacteria or yeast versus 14 days to 3 weeks in CHO cells.
Yeast and mammalian cell lines have inherent mechanisms for secreting properly folded, active proteins in culture medium. In E. coli, synthesized protein accumulates internally, and cell lysis is required to isolate the target protein. The process of lysis releases intracellular contents such as proteases or endotoxins, which can decrease yield and complicate purification. Recently, a strain of the gram-positive Corynebacterium glutamicum, which does not have endotoxins, has been engineered to express correctly folded, active recombinant proteins directly into the extracellular fermentation broth.
The following characteristics of target proteins influence their likelihood of being expressed and purified in a functionally active state:
Protein size. Eukaryotic cells, such as mammalian and yeast, have the advantage of being able to express proteins of 200 to 250 kDa, the size of most monoclonal antibodies. In some cases, expression of proteins as large as 400 kDa is possible. In E. coli, it is difficult to obtain proteins larger than 60 kDa in soluble forms.
Protein complexity. The number of protein subunits is inversely related to the likelihood of functional expression. Again, eukaryotic cells have a greater advantage in producing fully assembled, functional, multidomain proteins. Monoclonal antibodies, which consist of 4 subunits, can be secreted by eukaryotic cells as fully assembled complexes. E. coli can be used to express multidomain proteins, but they often lack activity.
Glycosylation. Posttranslational addition of sugar motifs to proteins profoundly affects biological activity, function, clearance from circulation, and antigenicity. Patterns of glycosylation are highly species-specific, and mammalian cell lines have glycosylation repertoires most similar to humans.
Common Challenges
Potential complications lie at every step of the process of recombinant protein expression and purification. The complications vary with each specific protein, but because proteins share common motifs, certain technical pitfalls are encountered frequently.
Protein aggregation. Suboptimal conditions during protein expression, purification, or storage can alter structure and result in aggregation or loss of activity. In bacterial expression systems, intracellular accumulation of protein can lead to formation of inclusion body aggregates.
Aggregation can sometimes be reversed by adjusting media composition, growth temperature, inducer concentration, promoter strength, or plasmid copy number. However, inclusion bodies formed by proteins native to eukaryotes but produced in bacteria tend to be resistant to solubilization. Even proteins that are successfully solubilized can become denatured and thus require potentially complicated in vitro refolding processes.
Refolding. The process of refolding is often preceded by a partial purification to remove host cell proteins that could aggregate with the target protein. Because refolding techniques vary and are often empirical, the online database REFOLD provides refolding protocols that have been successful for a wide range of proteins.2
Proteolytic degration. Host proteases, which can degrade the target protein, are commonly active during cell disruption, as intracellular contents are released. Genetic engineering can be used to remove a host’s more active proteases.
Disulfide bond generation. Disulfide bridges, which stabilize protein structure, are often abundant in secreted proteins. As with refolding, disulfide bond generation and maintenance are achieved through empirical processes. Possible strategies include targeting a specific extracellular excretion pathway or overexpression of chaperones or foldases.
Contamination. Contaminants can originate from host cells or from growth medium. Contaminants can interfere with protein function or cause adverse reactions in human consumers, such as through human pathogens harbored in mammalian expression systems or endotoxins in gram-negative bacteria. The process of removing contaminants can be extensive—often involving several rounds of chromatography—and each additional step diminishes yield.
Making the Transition to Clinical Production
The early steps in developing a recombinant pharmaceutical often require multiple iterations to optimize expression and purification of functionally active protein. As development moves closer to commercialization, schedules, pricing, and other aspects of managing a supple chain become paramount. This transition from focusing on technical concerns to the context of the clinic and broader marketplace can be challenging.
Shailesh Maingi, CEO and president of Kineticos Life Sciences, suggests there are key issues to consider throughout the process of production. First, are you going to make the protein yourself or outsource production? “From a business perspective, this decision is probably the most critical,” Maingi says. “If you want to produce it yourself, there is a huge investment required, not only in money, but in time and resources. If you’re a small, virtual company, you’re almost always going to outsource. If you’re a large company, you’re almost always going to insource because you have the capability and you have the technology.”
Smaller companies have the option of producing a lab supply of protein that can be used in preclinical toxicology studies. Outsourcing for larger-scale production can be done prior to initiating clinical trials.
If you decide to outsource, how do you choose a vendor? At the stages of protein expression and purification, working with a vendor that is flexible in its approach is important, because it is impossible to predict how complex the process will be for a given protein. Any vendor should be able to provide a clear proof of concept and performance history of the technology. Cost, quality, and convenience are important considerations as well. During the expression phase, proximity to the vendor can provide convenience for evaluating efficiency and yield of the process.

In summary, while the path to producing a functionally active and meaningful recombinant pharmaceutical can lead to great commercial opportunity, it can also be plagued with technical pitfalls and failures. Careful consideration of various strategies at every step of the process can optimize use of time and resources.

Source: Joel White, Business Manager, Biotechnology; Ajinomoto AminoScience, Raleigh, North Carolina
Drug Discovery & Development - October 07, 2010

No comments: