Comparing Mammalian Expression Systems

Almost every pharmaceutical and biopharmaceutical company in the world depends on the use of recombinant stable cell lines to enable drug discovery, development, and often manufacturing of biologics. It normally falls on multidisciplinary upstream development teams to attain this goal, requiring a wide variety of technologies and skill sets such as laboratory robotics, optical analyzers, molecular biology, and data processing. The large capital investment required to procure the equipment and expertise necessary to develop biologics can be cost prohibitive, which has spurred growth of a service sector to provide timely, cost-effective protein expression solutions. The outsourcing industry has grown consistently, paralleling the development of therapeutic biologics and leading to a wide array of protein expression technologies.

The impact protein expression level has on production costs cannot be overemphasized. High expression levels translate directly to lower cost of goods because of lowered capitalization costs in facilities and facility qualification, reduced operating costs, fewer production runs, and savings on consumables such as media. Equally important are stable cell lines for greater process consistency and control of final product purity. Expression levels and cell density are two critical attributes, with expression levels ranging 20–60 pcd (pg/cell/day) for commercially viable processes and cell densities of 10–150 million cells/mL (1). For an average process and cell density, 15–18 pcd equates to ∼1 g/L. Well-established systems involve extensive host cell line development for high cell densities in concert with specific vector systems. An example is the well-known Lonza GS system, which produces titers ranging 0.8–8 gm/L (2). The upper levels of titers demonstrate what can occur in test production lines that have been optimized over a decade, although in our experience titers of <4 g/L are normally seen for most standard processes.

Market needs can range from small quantities for vaccines to hundreds of kilograms for antibodies, and high titers with rapid delivery will always constitute a successful business model for the service industry. Many factors other than titer, however, can be equally important in decision making. Among the first considerations are host cell type and vector system. Both must match the expression requirements of a given product, especially regarding posttranslational processing features such as modified amino acids, glycosylation, and proteolytic cleavage, which sometimes will require the ability to coexpress processing enzymes. The greatest challenge is optimization of production specifics to reduce cost and meet regulatory standards. Areas of attention include gene amplification, animal-derived components such as fetal calf serum for cell expansion, selection components such as methotrexate or antibiotics, the required number of clones screened, specific productivity, and overall system robustness for various types of growth and scale-up related to suspension or adherent cell lines. So an ideal expression system would have the following features:

Successful commercialization of a diverse range of expression technologies has produced a variety of choices for customers, from start-up biotechnology ventures to established companies with limited available production capacity, as well as burgeoning opportunities for contract manufacturing, from single-to full-service providers. Table 1234 contain information to help companies prioritize their choices based on the services provided or business model used. They have been categorized into fee-for-service; license fees, royalties, and outlicensing options; and contract manufacturing organizations. The underlying commonality in all these platforms is the potential for creating stable clonal cell lines.

Table 1.Fee-for-service companies

Fee for Service Companies

Fee-for-service companies (Table 1) provide their customers with enhanced throughput technology to help them develop high-quality production cell clones (as in the case of Invitrogen and SAFC) or can provide a complete production model from DNA to CGMP manufacturing (as with ProBioGen).

Invitrogen (www.invitrogen.com) has invested significantly in its PD-Direct suite of services, which encompasses cell-line construction, screening, media optimization, and even selected mutagenesis using the Revolution human antibody system licensed from Morphotek (www.morphotek.com). Invitrogen has broad experience in vector development and can customize its vectors to preferred production cell types including Chinese hamster ovary (CHO) and NS0 cells. The company’s Free-Style protein production services can include transient protein expression as a guide to selection of an appropriate and stable production platform. Novel technology includes Morphotek’s evolutionary process to select desirable characteristics (e.g., high titer and certain cell traits) by allowing natural mutations to persist through reversible inhibition of host cells’ DNA repair machinery. That can be useful during early stage development to increase the host cells’ ability to secrete protein or improve the functionality of that product.

ProBioGen (www.probiogen.de) is primarily a know-how-based company, which makes it difficult to outlicense and positions it primarily as a fee-for-service company. It focuses on designing cell lines to improve productivity, achieve serum independence (e.g., CHO and NS0), grow in suspension, or derive posttranslational modifications through metabolic engineering, for example. Additionally, the company is developing avian and human cell lines within a proprietary framework, presumably to enter into the outlicensing and royalties arena. Fully documented development takes six to seven months after starting with cell lines selected for growth in animal-component-free media.

As with Invitrogen, SAFC (www.sigmaaldrich.com) has invested significantly in cell-line screening services as an extension of its media business by partnering with Cyntellect (www.cyntellect.com) to exploit its Cell Xpress clone-picking system, which uses laser-enabled analysis and processing (LEAP) technology for cell imaging and manipulation. The Cell Xpress system enables selection of individual clones based on growth rates, protein secretion, clonal size, or other environmental parameters at a rate of 30,000 cells/minute. Generic expression vectors are often acceptable for this type of system because clone selection is based not on enhanced expression, but rather the indiscriminate approach of screening thousands of clones for the rare high producer. Both Invitrogen and SAFC are primarily interested in developing their media business, so they are very approachable regarding services.

Licensing/Royalty Companies

A number of companies have developed cell lines, expression vectors, or a combination thereof to license for fees or royalties (Tables 2 and 3). Companies focusing on human or human-like glycosylation patterns are generating novel production cell lines, some that grow to very high cell densities. For example, Vivalis (www.vivalis.com) has developed a serumfree avian embryonic stem cell line that produces antibodies with a glycosylation pattern similar to those of humans, with reduced fucose and improved antibody-dependent cell cytotoxicity (ADCC), which is important for cancer therapy. Cevec Pharmaceuticals (www.cevecpharmaceuticals.com) has developed an immortalized human amniocyte cell line with favorable glycosylation patterns that can grow to high cell densities in suspension culture. Crucell (www.crucell.com) has a technology that is focused on a better production host cell, the PER.C6 human cell line, which is discussed further below. All three companies make their cell lines available through outlicensing, and they may also generate production cell lines for customers using these cell lines.

Table 2.Companies with cell-based technologies

Companies developing robust new expression-vector-based technologies have used site-directed recombination to foster “optimized” chromosomal integration or novel genetic elements to modulate chromatin surrounding the site of integration. Other technologies include the use of miniature chromosomes and retroviral delivery systems. These technologies (Table 3) provide customers with multiple options for discovery and preclinical protein production.

Table 3.Companies with vector-based technology

Cellectis (www.cellectis.com) recently entered the bioproduction market to provide access to its proprietary Meganuclease Recombination System (MRS) technology through partnership agreements. MRS uses rational genome engineering technologies that involve DNA cleavage and recombination enzymes to target sitespecific integration of expression constructs into a stable site in CHO or other cell types. Recombination of vector DNA into desirable chromosomal loci removes the possibility of random integrations into nonexpressing heterochromatic DNA sites and thus increases the proportion of “good” expressers in a pool of transfectants. Similar technology is also marketed by Invitrogen with its Flp-In DNA recombination system that uses Flp recombinase to target vector DNA into engineered host chromosomal sites. A number of cell lines (including CHO) have been developed with stable recombination acceptor sites already present.

Chromos Molecular Systems (www.chromos.com) has developed and patented its Artificial Chromosome Expression (ACE) miniature chromosome system designed to introduce multiple genes into a single host cell. The technology allows for coupled selection of multiple gene cassettes that could be useful for expressing multisubunit proteins and engineering complex metabolic pathways. The Sure Cell Line Development platform offered by Selexis (www.selexis.com) takes advantage of proprietary genomic scaffold/matrix attachment regions (S/MARs) to expeditiously develop stable, productive cell lines.

Millipore’s (www.millipore.com) Ubiquitous Chromatin Opening Element (UCOE) technology falls into a similar category of chromatinstabilizing DNA sequences. UCOE cell lines can have good expression from single copy-number cells, even if the DNA integrates into silent, heterochromatic regions of their chromosomes. Such versatility helps foster the rapid creation of stable, expressing cells. The company outlicenses this technology for fees and royalties, part of a move into the research tool market.

Companies that offer proprietary technology coupled with production capabilities include Gala Biotech (www.gala.com) and Crucell, the latter having partnered with wellknown contract manufacturing organization DSM (www.dsm.com) to create a fee-for-service production entity known as Percivia (the PER.C6 Development Center) in Cambridge, MA. Crucell is an established vaccine development company that developed its well-documented, immortalized, serum-free production cell line known for a strong safety profile, good productivity, and human glycosylation patterns. Additionally, Percivia has developed the XD process, which supports growth of PER.C6 cells to very high densities (150,000,000 cells/mL) with concomitant high productivity (5). Production licensing is based on geographic location, so to avoid confusion, both companies should be contacted regarding licensing. Crucell also offers STAR and STAR-Select systems for expression of proteins in CHO cells. STAR elements are reported to be highly species-conserved DNA sequences that flank expression constructs and improve both recombinant protein productivity and cell line stability.

Another player in this arena is Gala Biotech in collaboration with Catalent Pharma Solutions (www.catalent.com). As a stand-alone service, Gala supports cell-line development from DNA to CGMP production for Phase 1–2 clinical materials. The company’s proprietary Gene Product Expression (GPEx) technology uses a retroviral delivery and expression system to engineer mammalian cell lines (e.g., CHO) with multicopy, stable DNA insertions. Transfection efficiencies are reported to be high, reducing the need to screen large numbers of clones and thus reducing cell-line development timelines.

Contract Manufacturing Organizations

With respect to market share, CMOs dominate the upstream process development service sector, so it is natural for them to also supply cell-line construction services. In terms of overall production capacity, the main players are Boehringer Ingelheim (www.boehringeringelheim.com) and Lonza (www.lonza.com), which currently operate an impressive 13% and 6% of total mammalian cell culture capacity worldwide respectively. All the remaining CMOs capture about 8% of total capacity (6). The CMO mammalian cell culture market has doubled in size over the past four years, and that trend is expected to continue as small and midsize biotechnology companies continue to outsource their production of clinical trial materials and cell line development. With this growth will come demand for more rapid development timelines and increased productivities, which will drive technology development.

Companies listed in Table 4 are primarily CGMP contract manufacturing businesses that offer cell-line construction as part of their overall services. Boehringer Ingelheim, Lonza, and CMC Biologics (www.cmcbio.com) all offer clinical and commercial manufacturing, whereas Cobra Biomanufacturing (www.cobrabio.com) provides only clinical materials at present. Licensing fees and royalties related to cell-line development are expected to be negotiable with all CMOs, depending to what extent a company also uses their manufacturing services. In some cases, part of the licensing is tied directly to manufacturing. That is not always the case, however. For example, the CHEF1 system provided by CMC Biologics has been licensed to more than a dozen companies. The technology platforms are all similar in that they have streamlined vector, cell line, and process development to rapidly deliver GMP-purified protein with complete analytical, formulation, and regulatory support.

Table 4.Contract manufacturers (license fees, royalties, and fee for service)

The Lonza GS system (GS for glutamine synthetase) has currently set an industry standard for antibody production of 5 g/L in chemically defined media. Historically, high titers have required GS DNA amplification, and there is a wealth of literature citations on the company’s website covering titer and stability findings associated with this amplification process. Cell lines and media conditions probably contribute most to the high levels of production seen by Lonza because of the very high cell densities achieved on scale-up. Recent work with the company’s proprietary chemically defined, animal-component-free (CDACF) medium and CDACF-preadapted CHO cells shows a rapid development of pools producing up to 2 g/L in disposable bioreactors. Of course, the time-consuming process of clonal cell selection would need to be implemented for any CGMP process.

Both CMC Biologics and Cobra Biomanufacturing offer short CGMP development timelines, in part because they can create stable clonal CHO cell lines without amplification. For instance, a subsidiary of CMC Biologics in Bothell, WA, CMC ICOS has demonstrated the ability to go from DNA to CGMP-grade clinical material for antibodies in as little as 12 months using the proprietary CHEF1 vector. It makes use of the Chinese hamster EF1 (elongation factor 1) promoter-enhancer to drive high-level transcription from low-copy-number, stable host chromosome integrations (7). Furthermore, unlike the GS system, this has multiple selectable markers supporting supertransfection of multiple expression constructs to enable detailed metabolic engineering of a production cell line for expression of very complex modified glycoproteins. Millipore’s UCOE technology has been out-licensed to Cobra Biomanufacturing for CGMP production, and it allows for rapid generation of stable cell lines through its ability to integrate in low copy numbers and drive high-level expression.

Boehringer Ingelheim has recently added the BI-HEX (Boehringer Ingelheim High Expression) system to its own enormous production capacity. This system is proprietary and claims fast-track production of “high-quality, high-titer processes” in CHO cells. The BI-HEX platform includes vectors with novel genetic elements, serum-free transfection of suspension-adapted CHO cells, chemically defined media, and high-throughput clone selection.

A Driving Force

Recent advances in expression technologies have greatly increased productivity, thus becoming an enabling technology in the commercial success of biotherapeutics that may require high doses, such as antibodies. For expression services customers, high-quality protein production is a requisite that depends heavily on the downstream quality and analytical strengths of the manufacturer. The technologies covered here focus on variables such as titer and timeliness and in some cases product quality (e.g., with respect to metabolic engineering and control of posttranslational modifications), which of course should be discussed in more detail with vendors before entering into the process of developing a drug. Generating higher expression titers will be a driving force for the industry in the foreseeable future, as long as product quality remains uncompromised. Unfortunately, higher titers are sometimes generated at the expense of quality, so the support of good analytical tools is as important as increasing product yield. Timelines for process development depend on many factors — e.g., the type of protein produced, quantities needed, and product quality specifications — making it difficult to directly compare technologies with seemingly contrasting endpoints.

Clearly we are in a period of dramatic change in biologics development. Over the past two decades, with the acceptance of mammalian cell culture as a standard production tool for complex proteins, product titers have increased from milligram-to gram-per-liter concentrations. This is reminiscent of the dramatic increases in antibiotic production during the latter half of the 20th century. And as in the case of penicillin V and G, the titer increase in biologicals has come about by incremental improvements in expression systems, production technologies, metabolic engineering, and media improvements that allow for increasing cell densities. Future technologies will most certainly push expression levels to new heights, providing even more opportunities and competition for production services providers and leading to even more choices for their customers. However, in the foreseeable future, generation of a CGMP-grade production clone will be the first bottleneck encountered in the long development cycle of a modern protein-based therapeutic.

REFERENCES

1.) Coco-Martin, JM, and MM. Harmsen. 2008. A Review of Therapeutic Protein Expression By Mammalian Cells. BioProcess Int. 6:28-33.

2.) Birch, J. 2007..

3.) Weidle, UH, P Buckel, and J. Wienberg. 1988. Amplified Expression Constructs for Human Tissue-Type Plasminogen Activator in Chinese Hamster Ovary Cells: Instability in the Absence of Selective Pressure. Gene 66:193-203.

4.) Fann, CH. 2000. Limitations to the Amplification and Stability of Human Tissue-Type Plasminogen Activator Expression By Chinese Hamster Ovary Cells. Biotechnol. Bioeng. 69:204-212.

5.) Lee, G. 2008.. Cell Line Development Using a PERC6 Host Cell Line: A Highly Efficient Expression Platform.

6.) 2006.The Protein Therapeutics Market: The Science and Business of a Growing Sector, Kalorama Information, Rockville.

7.) Running Deer, J, and DS. Allison. 2004. High-Level Expression of Proteins in Mammalian Cells Using Transcription Regulatory Sequences from the Chinese Hamster EF-1 Gene. Biotechnol. Prog. 20:880-889.

8.) 2007. Boehringer Ingelheim: From Mind to Market — Our Services in Biopharmaceutical Contract Development and Manufacturing. BioProcess Int. 5:38.

9.) Carrier, T. 2008. Team Up for Today’s Cell Culture Challenge: PD-Direct BioProcess Services. BioProcess Int. 5:146.

10.) Kayser, K. 2006. Cell Line Engineering Methods for Improving Productivity. BioProcess Int. 4:S6-S13.

11.) Wynne, J. 2007. UCOE Technology Maximizes Protein Expression. BioProcess Int. 5:140.