While it’s too early to predict that CHO’s days are numbered, the biopharmaceutical industry is clearly looking for a next-generation expression platform to address sky-high production costs. In a July speech at the Brookings Institution, U.S. FDA Commissioner Scott Gottlieb, M.D., criticized biopharma on several points: “Less than two percent of Americans use biologics, but they [the medicines] account for 40 percent of total spending on prescription drugs. They also represent 70 percent of the growth in drug spending from 2010 to 2015 and are expected to be the fastest growing segment of drug spending.”
Regulators are not the only ones dissatisfied. Last year, Biogen’s vice president of international manufacturing, Eliana Clark, explained that her company was looking for alternatives to CHO. “How do we really lower the cost of manufacturing to a place where we can actually make affordable medicines for people? We did a lot of modeling with MIT, and we realized the CHO cell line won’t get us there.”
Meanwhile, Biopharma Reporter’s 2017 survey suggests that nearly half of biomanufacturing experts believe that the industry is “too reliant on Chinese hamster ovary expression systems.”
Alternative expression hosts
Over the years numerous protein expression systems have cropped up, each promising untold efficiencies, cost- and time-savings, improved patient safety, and access. Among the more interesting is the C1 system from Dyadic, based on the fungus Myceliophthora thermophila.
C1 is robust, versatile, and easily manipulated for protein expression, and—little known fact—fungi are more closely related to animals than either plants or bacteria. C1 does not require proof of concept with regard to large-scale production of commercially significant proteins. Dyadic has applied C1 to the manufacture of high-purity enzymes at scales of up to 500,000 liters, but the commercialization of industrial chemicals is vastly less complex than for biopharmaceuticals.
The company is involved in several collaborations to bring C1 to mainstream biopharmaceutical production. One project, with the European Union, has demonstrated a high level of production of an antigen protein. Another, with Mitsubishi Tanabe Pharma, is aiming to overcome gene expression challenges for two important therapeutic compounds. Dyadic is also working with the Israel Institute for Biological Research to advance C1 for producing recombinant vaccines and neutralizing agents, and investigating the potential of C1 for mainstream therapeutic enzymes and proteins.
Alongside these efforts, Dyadic has tackled protease generation, which can delay development of conventional mammalian expression systems for months, sometimes permanently. Expression systems generate proteases in an attempt to eliminate self-generating but foreign proteins. Dyadic has identified protease-coding genes in M. thermophila and is on the way to reducing them by a factor of nine. In September, Dyadic entered a collaboration with Sanofi-Aventis to demonstrate proof-of-concept for C1 to produce “multiple” biologic vaccines and pharmaceuticals.
Much has been written about post-translational modifications of therapeutic glycoproteins, particularly on efforts to assure human-like glycosylation. Compared with yeast or bacteria, C1 attaches glycans in patterns that resemble human glycosylation: O-glycosylation (problematic) to date appears absent. Dyadic has undertaken efforts to engineer C1’s glycosylation to resemble more common mammalian forms such as G0, G0F, G2, and G2F as well.
Image: Dyadic's C1 gene expression platform is based on a patented and proprietary genetically modified strain of the fungus Myceliophthora thermophile.
Working with what you know
CHO has been described as the "workhorse" expression system for monoclonal antibodies. Numerous established commercial processes are based on large-scale CHO expression, to the point where vendors and biomanufacturers talk of “platform,” or plug-and-play processes based on that cell line. Has this occurred by design, by accident, or by default?
Dr. Jamie Freeman, product manager at Horizon Discovery, has another theory: that biotech’s reliance on CHO came about through trial and error. “CHO was one of many options available at the time. While CHO isn’t perfect, and even as other systems show potential today, given what was available 30 years ago and limitations in our ability to modify different systems, developers at the time struck a good balance between expression and safety considerations.”
The biggest process improvement, in terms of improving CHO, has been in media and feed development. Experts like Dr. Florian Wurm, at the Swiss Federal Polytechnic Institute, Lausanne, have gone on record to say that media development has been the single most significant improvement responsible for improved CHO productivity. Wurm has also expressed dismay over CHO’s genomic instability—a factor that can only detract from prospects of improving the cell line through genetic engineering.
According to Freeman, the emergence of chemically defined media has helped to reduce batch-to-batch variability, while improving safety by eliminating a potential source of animal-derived pathogens. “Serum-free media has enabled the adoption of design of experiment for process development. Additionally the introduction of improved selection systems, particularly the GS knockout system, significantly shortens cell line development times while substantially increasing average titres.”
Unsurprisingly, Freeman is less sanguine about wide-scale adoption of CHO alternatives. “Any new entrant to the market would initially need to demonstrate that it can generate safe therapeutics that are acceptable to regulators. They must also demonstrate that any benefit is significant enough to warrant changing their manufacturing platform.”
Part of this exercise requires that novel expression systems be compatible with existing upstream and downstream equipment, operations, infrastructure, and workflows. Significant incompatibility with, say, existing harvest or capture unit operations would likely sink any “new CHO.”
A good place to begin, Freeman says, is for products for which CHO is unsuitable, which would allow time and create the imperative for processes and regulators to adjust. “But this will take time. Many systems have been proposed over the years, from insect cells to modified yeasts, but the benefits have never proven significant enough to displace CHO, which is still the first port of call for producing complex biotherapeutics.”
At the same time Freeman likes the potential for further improvements in CHO: “Modern biotech tools such as transposase-mediated integration, or small-scale genome editing (for example the GS system), have improved the performance of CHO-mediated expression, but the potential impact is far greater than this. Combining the new improved sequence based on Horizon Discovery’s GS knockout CHO K1 line with CRISPR screening will allow the identification of genetic targets that improve a number of attributes, from titre through to product quality and even genomic instability.”
These targets can be modified further to improve CHO, and even combined with synthetic biology approaches for improved processability. “Extensive genome engineering could meet or exceed any improvements promised by novel expression systems, while retaining the assurance that comes with our extensive experience with CHO cells,” Freeman says.