May 15, 2016 (Vol. 36, No. 10)
If you’re Serious about Investing in Expression System Success, You Need a Balanced Portfolio
Few balancing acts are as delicate as the optimization of a cell expression system, especially if the expression system is to yield recombinant proteins destined for therapeutic use.
Every expression system—every scheme for optimizing nutrients and cell culture conditions—presents advantages, disadvantages, and tradeoffs with respect to costs, yields, quality metrics, contaminants, availability of system components, time constraints, and so on.
Another important consideration is the degree to which an expression system may be chemically defined. For example, with animal cell cultures, the trend is away from serum and toward defined media, that is, media uncomplicated by the presence of animal-origin components. This trend, well advanced for basal media, is also progressing for media supplements.
Yet attempts to break from serum have a “can’t live with it, can’t live without it” quality. Use of serum to culture cells dates back nearly 70 years and continues to this day, albeit with less devotion.
“Back in the 1950’s, cells from Henrietta Lacks (now called HeLa cells) were isolated and cultivated in horse serum,” recalls Adam Elhofy, Ph.D., CSO, Essential Pharma. “Subsequently, the field moved more to using cow serum. Today, it is often fetal calf serum. However, there is a push to utilize supplements designed to eliminate serum completely. Unfortunately, serum-free supplements still have significant challenges.”
Contemporary bioprocessing has largely eliminated serum. But according to Dr. Elhofy, disposing of serum can be a little like throwing out the baby with the bath water. “The problem is that serum contains more than 300 factors and provides beneficial factors such as fats, cholesterol, small molecules, and hormones,” he says. “Quite frankly, people have not completely solved this problem.”
Lifting the Serum “Curse”
According to Dr. Elhofy, Essential Pharma’s goal was to develop a means to deliver fats and cholesterol in a system that would increase protein production, yet be chemically defined and free of animal-origin components. “Most products on the market have serious shortcomings,” he points out. “In one case, lipids are solubilized with alcohol, but this causes oxidation and destabilization of free fatty acids. Others use dextran, but it adheres to culture vessel walls and subsequently removes cholesterol from the media creating a sink.”
Essential Pharma focused on filling in the hole in this cell culture void. “We developed Cell-Ess®, a novel lipid supplement produced under GMP conditions that serves as a serum-replacement, Dr. Elhofy explains. “We initially targeted its use for the production of monoclonal antibodies (mAbs) because this represents the biggest biotherapeutic sector. We demonstrated that Cell-Ess increased titers 20–40% in bioreactors. Further, there were no significant changes in metabolites or increases in biomass such as cell debris and host proteins.”
Some optimization is required for each specific use. But the process can reliably increase productivity when scaled up.
“Scalability is a problem in many manufacturing platforms,” notes Dr. Elhofy. “We found that we could boost protein biomanufacturing of mAbs by adding the supplement even to a previously optimized serum-free media in a scalable bioreactor run representative of 10,000 L bioreactor runs.”
The company is currently evaluating the serum-free supplement on other aspects of bioprocessing, including single-use bioreactors, protein quality, and cell metabolism.
Optimizing Expression of Biosimilars
When follow-on manufacturers produce a follow-on biologic, or biosimilar, the idea is to emulate an originator company’s “innovator” product. The biosimilar should have a structure that is highly analogous to the innovator product. Also, the biosimilar and the innovator product should have comparable clinical safety, efficacy, and immunogenicity.
Typically, biological medicines are made by or derived from a biological source such as a bacterium or yeast. Cost-efficient production requires the careful optimization of expression and purification.
However, the more complex a biosimilar is, the more difficult it is to produce, according to David Vikström, Ph.D., chief technology officer, Xbrane Bioscience. “The combination of high yield and good quality of the produced target is normally a challenge, but with complex biosimilars it is exceptionally difficult,” he declares.
“For example, our lead biosimilar consists of two different polypeptide chains (in principle to different proteins),” Dr. Vikström continues. “These need to be assembled in a very specific way to get the correct three-dimensional structure, which is essential for the activity of the product. If we were only able to produce high yields of the target and it had the wrong three-dimensional structure, then the product could not be used at all.”
Dr. Vikström adds that optimizing the expression of biosimilars requires that scientists pay attention to the entire production chain. “It is important to consider all processes from design of the genes to be expressed to the strain used, as well as the parameters employed for expression and purification,” he insists. “If one link in the chain is not working optimally, it does not matter if you, for instance, have the best strain or not.”
Another aspect of optimization is scalability of the developed process. “It doesn’t matter if you can reliably produce your produce in a 50 mL shaker,” cautions Dr. Vikström, “if you are unable to upscale it with a stable process.”
Xbrane not only produces its own biosimilars, it also produces those of its clients, acting as a service provider. “Our company,” notes Dr. Vikström, “has developed an Escherichia coli–based protein production technology that has increased production yield by an average of eightfold across multiple different therapeutic proteins.” Xbrane is currently developing a pipeline of biosimilars as well as controlled release drugs.
Boosting Transient Systems
Mammalian expression systems also require rigorous optimization to enhance expression and production. Gavin Barnard, Ph.D., group leader, Eli Lilly and Co., describes his company’s approach to producing high yields of therapeutic proteins as follows: “We need to produce several grams of clinical-grade protein to enable testing for attributes such as safety, efficacy, and stability. Multiply that by the dozens of other different projects, as well as by the many variants per project, and you end up with the need to produce kilogram quantities of high-quality protein.”
Biologic drugs are typically produced in mammalian cell lines such as stable Chinese hamster ovary (CHO) cells. “These cell lines are handy as they produce monoclonal antibodies and other biologic drugs at concentrations greater than 1 g/L,” Dr. Barnard asserts. “Therefore, a single laboratory-scale bioreactor can generate gram quantities of protein.”
“However, it takes 3–9 months to establish a stable CHO cell line,” Dr. Barnard points out. “In research, project teams always need protein yesterday.”
To overcome this challenge, Dr. Barnard’s team developed a transient transfection system capable of maintaining high cell densities. “In a nutshell,” says Dr. Barnard, “we set out to develop a transient CHO expression system that approximated the final drug manufacturing process.
“We used the same cell line, same media, same plasmid DNA, and a production process similar to one that we intend to use to manufacture our drugs a decade from now. We rigorously optimized cell density, DNA, and transfection reagent (polyethyleneimine) concentrations followed by process development strategies.”
The team also improved the system by ensuring that it could co-express XBP1S, a global transcription factor. “XBP1S influences genes involved in the endoplasmic reticulum stress response that triggers the unfolded protein response,” explains Dr. Barnard. “Previously published literature has shown that XBP1S overexpression increases recombinant protein expression in CHO cells. We have not observed any problems using XBP1S.”
The scientists continue to tweak the CHO platform to improve titers further. This will help project teams that often need larger amounts (that is amounts in the 1–5 g range). “For larger amounts, we first make stable CHO pools and then use bioreactors,” notes Dr. Barnard. “We’ve made some improvements in this area, too, and plan to disclose these in the near future.”
In recent years, technologies for cell-line development have improved significantly, remarks Holger Laux, Ph.D., fellow, Novartis. He notes that emerging cell-line engineering tools such as Cas9 (CRISPR-associated endonuclease), ZFN (zinc finger nuclease), and TALEN (transcription activator-like effector nuclease) systems enable researchers to efficiently improve production cell lines with reasonable effort.
“CRISPR/Cas9 technology, with its high efficiency and low cost, can signi?cantly decrease the heavy workload involved in generating, for example, knockout cell lines,” Dr. Laux continues. “Additionally, these engineering tools are benefitting from the recently available Chinese hamster genomes.”
Indeed, the characterization of the CHO genome may revolutionize many aspects of optimization. “The Chinese hamster genome information enables high-throughput technologies, such as next-generation sequencing (NGS) transcriptome analysis to identify new engineering targets,” Dr. Laux states. “Furthermore, the genome sequence data provides key information for gene targeting, opening new opportunities for cell-line engineering.
“Additionally different ‘targeted integration’ technologies have emerged. These technologies allow a plasmid encoding the gene of interest (such as a gene for an antibody) to be integrated into a specific site of the genome. This reduces variation in expression and subsequent screening of high number of clones to select clones suitable for high and stable expression of recombinant proteins.”
Overall, Dr. Laux reports, “researchers can nowadays make use of a variety of excellent tools to better understand the expression of therapeutic proteins as well as improve productivity and quality parameters—capabilities that were missing just a few years ago.”
Scientists also can harness the power of proteomics to help optimize cell culture bioprocess development, according to Deniz Baycin Hizal, Ph.D., scientist II, antibody discovery and protein engineering, MedImmune. “Proteomics,” declares Dr. Baycin Hizal, “can serve an important role in this effort by identifying those factors that enhance the cell’s capacity to produce high yields of protein therapeutics as well as addressing the bottlenecks in low production cell lines.”
Employing a quantitative proteomics approach can also help in characterizing thousands of CHO cellular proteins. “Quantitative proteomics can be used to identify differentially expressed proteins associated with key cellular properties including protein production, cell growth, reduced apoptosis, favorable glycosylation, and optimized medium formulations,” explains Dr. Baycin Hizal. “In quantitative proteomics, we use labeling techniques such as Tandem Mass Tags™ from Thermo Fisher Scientific.”
After data acquisition, bioinformatics tools are used to map the raw data to the in silico digested genome to identify and quantify the differentially expressed proteins between different cell lines. “Applying statistical analysis can reveal the identity of proteins with significantly altered expression patterns between cells lines, notes Dr. Baycin Hizal. “This information can be further interrogated with cellular pathway analysis tools to identify potential bottlenecks and causes of that contribute to low production in cell lines. It is our goal to use cell-line or media engineering approaches to derive solutions that can be used to enhance recombinant protein production in our manufacturing cells.”
Dr. Baycin Hizal has some advice for researchers working on optimizing protein production: “Use the wealth of information and techniques available to study CHO physiology. With the availability of the genome sequence and establishment of informatics tools to study it, new opportunities have arisen to utilize the information to enhance our knowledge of these cellular production factories. As a result, the CHO systems biology era is underway.”
For the future, Dr. Baycin Hizal sees that critical “omics” datasets—including proteomics, transcriptomics, metabolomics, fluxomics, and glycomics—will continue to emerge. She expects that these datasets will allow the elucidation of the molecular basis of CHO cell physiology.
“Integrate these datasets into mathematical models to describe CHO phenotypes,” advises Dr. Baycin Hizal. Doing so, she says, will provide crucial insights related to protein production, cell health and growth, favorable glycosylation, and optimized medium formulations. “In short,” she continues, “use the wide variety of advancements made in biotechnology techniques as tools to devise strategies to enhance the production of novel biotherapeutic drugs.”