Editor's Note: Since this article was written, European regulators have recommended the use, and paved the way for approval, of GTC Biotherapeutics' ATryn, which would become the first transgenics-derived product on the market. Click here for more information.
To say that therapeutic protein production in transgenic plants and animals has been long on promise, short on products, is an understatement. Nineteen years after demonstrating proof of principle, in transgenic mice, we still await the first transgenic protein product approval.
Transgenics — put simply, modifying the genetic material of one organism with that of another — should be a paradigm-buster, with greater potential for lowering cost of goods for biologicals than all the buzzwords of the last decade combined. Compared with cell culture and fermentation, transgenics offers numerous process advantages:
- Lower capital and operating costs upstream, especially during expression;
- Depending on the source, easier downstream separations;
- The manufacture of proteins that are poorly expressed, if at all, in mammalian cell culture, yeast, or bacteria;
- Elimination of risky, unreliable sourcing for some biologicals from animal and human tissue;
- Long-term stabilization of proteins in seeds or plant tissues;
- Proteins in physical forms that require less downstream manipulation.
Although the science and engineering have been validated elegantly and repeatedly, transgenics has encountered an entrenched fermentation culture, technical difficulties, and regulatory hurdles that at times appear insurmountable. Biotech’s primary expression model, mammalian cell culture, has spawned dozens of products, billions of dollars of revenue, and a warm, cozy environment in which regulators, manufacturers and vendors can operate. Transgenics represent an abrupt departure — a new set of manufacturing and regulatory challenges, and untold potential.
Biotechnology has previously faced paradigm shifts of similar magnitude. Experts recoiled at the prospect of manufacturing injectible drugs in yeast, pathogenic bacteria, and especially in cancer cells. Eventually, good science rendered these objections moot. Ironically, one of the battlegrounds in transgenics today is the source of transgenically produced proteins — food plants and animals that humans have ingested safely since time immemorial, but which present difficulties within the predominant GMP culture.
Protein expression in cultured cells is an uphill battle in every sense. Because cells exist out of their biological context, processors must pour nutrients and energy into bioreactors to keep cells alive and growing, and to fuel their protein-making apparatus. Cultured cells were not meant to express foreign proteins, even less so at production-worthy titers. Moreover, mammalian cells prefer to attach to the sides of vessels, much as they stick to other cells in tissues, rather than grow in suspension as bioprocessors prefer.
Animals and plants naturally make high-abundance proteins in specific tissues or fluids — blood, milk, eggs, leaves, seeds. As complex organisms, they are able to sequester therapeutic proteins and still go about their normal existence. Contrast this with fermentation, where exposure to foreign proteins by the entire organism or cell often spells trouble.
Most obvious benefits upstream
With transgenics, upstream operations are greatly simplified and reduced in terms of time, cost and complexity. At the development stage, when cell culture engineers are specifying stainless steel vessels, bioreactors, control equipment, piping, clean-in-place rigs, and a hundred-million-dollar facility, transgenics is concerned with agricultural issues — acquiring land, veterinary and husbandry expertise in the case of animals or plant specialization for food crops, and simple processing equipment for grinding grains or separating milk.
“Transgenic expression systems can be so much more attractive, financially, than CHO [Chinese hamster ovary] cells,” says Tom Newberry, VP at GTC Biotherapeutics (Framingham, Mass.). Economic benefits include flexibility of capital and operations, profitability even at low volumes, and lower cost of goods. Scaleup involves breeding more animals or growing more crops, compared with designing, building and validating a larger facility. Without the facility and equipment overhead, early-stage companies enjoy the freedom to allocate resources optimally.
GTC, formerly Genzyme Transgenics, concentrates on proteins that express poorly in cells, for example blood proteins, which are isolated from animal or human blood serum. GTC’s lead product, ATryn (recombinant human anti-thrombin), a plasma protein with anti-coagulant and anti-inflammatory properties, is currently under development in Europe, the Middle East and Canada with Leo Pharma (Ballerup, Denmark).
Even as FDA urges pharm/biotech to embrace and manage risk, companies have seemingly become more risk-averse. Nowhere is this more evident than in transgenics, where “nobody wants to be first,” says Newberry. In February 2006, the European Medicines Agency (EMEA) turned down GTC’s marketing application for ATryn, citing an insufficient number of patients in clinical studies and a minor discrepancy between the submitted purification methodology and the one used for the Phase III study. GTC and Leo Pharma have re-submitted their application. GTC has invested $200 million in ATryn and plans to pursue FDA approval, regardless of EMEA’s final decision.
Interestingly, GTC positions itself, reasonably successfully, as a contract manufacturer using transgenic cows and goats as an alternative production platform. Although no such products have hit the market, these “backup” agreements provide the company with funds that have helped the firm persevere in a business where cash burns quickly.
Expression and productivity
Expression rates in bioreactors are critical, since they determine the eventual production capacity or volumetric productivity. No analogous relationship exists for transgenics. “If you need more material to meet a production quota you simply breed more animals or grow more crops,” Newberry says. “This factor ultimately allows greater freedom in process design, to the point where capturing every last molecule becomes less critical.”
Traditionally, capital costs are driven by bioreactor capacity and facility. Although upstream capital allocation influences cost of goods and the price that customers pay to the same extent as purification and fill/finish, capital outlays strongly influence the number and types of products a company is capable of developing. Transgenics’ ultimate value may therefore be in allowing companies to work on multiple products in parallel, thereby reducing the risk associated with any particular product.
Newberry cites a Biogen cell culture facility that recently sold for $500 million. Multi-hundred-million dollar facilities are common in biotech, to the point where many development-stage companies cannot afford to build facilities. For transgenics, the same capacity may be had for $40-100 million, a cost that may be spread over the product’s development timeline instead of up front. “A traditional biotech company with $500 million in its pocket might have to bet on one or at most two products,” notes GTC’s Newberry. “Using a transgenic platform, you can spread that risk over maybe ten products.”
Although upstream facility costs will be low for transgenics, the expertise required to rear animals or grow transgenic crops is not easily found among traditional bio-entrepreneurs. That, and the potentially tricky hybrid (FDA/USDA) regulatory climate, will cause transgenics firms to seek out contract manufacturing and development relationships wherever possible. Companies should seek out university veterinary or agricultural facilities for the upstream part, believes Charles Clerecuzio, senior director at engineering firm IPS (Philadelphia). Initial purification might even be possible before the product is officially moved into GMP facilities. Just how the regulatory oversight for this idea plays out, given FDA’s obligation to regulate drug products, is unknown.
Clerecuzio sees tomorrow’s transgenic startups retaining direct oversight over only those operations they deem proprietary, for example the molecular biology, embryology or downstream operations, and contracting out everything else.
Transgenic plants have been the source of controversy, mostly due to genetically modified crops. Biolex (Pittsboro, N.C.) takes a different approach. Instead of cultivating crops outdoors, the company’s LEX System uses permanently transfected lemna (duckweed) plants grown in an indoor controlled environment in scaled containers. Lemna plants double in mass every 36 hours and produce perfect clones of themselves without setting out seeds or spores. Biolex recently doubled its manufacturing capacity through an expansion that took about half a year and minimal capital investment. This capacity should serve Biolex through Phase III manufacturing, says Glen Williams, senior VP of operations.
Upstream capacity expansion is as simple as constructing a building with controlled temperature and light. Since a sterile environment is unnecessary, the huge investment in cleaning, steam-in-place, and associated validation simply goes away. “There’s no chance of a failed batch because of a sterility breach caused by turning a valve incorrectly,” Williams says. “We’re basically growing plants in a GMP facility.” LEX System simplifies downstream operations as well, since there is no need for viral clearance. “Every time you introduce a viral clearance step you create a room full of paper, and a new expense category,” he says.
Streamlined operations cut facility construction time as well, from five to seven years for a cell culture facility to three years or less for the LEX System. The platform can therefore postpone capital expenditure decisions until they are fairly certain a product will enter Phase III.
Biolex focuses on developing proteins that are difficult to make in cell culture, a strategy that has attracted attention from mainstream biomanufacturers. A recently expanded agreement with Medarex (Princeton, N.J.) is in addition to one with Centocor (Horsham, Pa.), that combined will cover up to ten protein therapeutics. Locteron, Biolex’s lead controlled-release alpha interferon product, is under development with delivery specialty firm OctoPlus (Leiden, The Netherlands). Biolex expects to begin Phase II testing by the end of 2006. Additional products include another human interferon in early-stage testing, an anti-thrombolytic (clot-busting) drug, BLX 155, and three additional protein therapeutic candidates.
Halfway between traditional cell culture and transgenic plants is Greenovation’s (Freiburg, Germany) moss bioreactor, which uses glass instead of stainless steel vessels to grow transiently transfected moss. The company claims that this organism churns out proteins with human-like glycosylation patterns, a benefit both for manufacturing and clinical development. The company generates stable production strains in four to six months, which is comparable to cell culture, but scaleup is easier and less expensive. Processing is simplified because moss secretes product protein into the medium.
Transgenics provide less obvious benefits, compared with cell culture, further downstream. As with CHO processing, animal-based transgenics will require demonstration of viral clearance but plant-based feedstock would not since no plant viruses have ever been known to infect humans.
Developers are content to get product in a form that can be fed directly into capture or ion exchange chromatography columns. How that state is achieved depends on the raw form of the product. Much as in the food industry, eggs are separated, milk is concentrated, and green plants are ground up and extracted. GTC, for example, uses a tangential flow filtration step to reduce fluids and remove non-protein small molecules from milk, yielding a product that feeds directly into the chromatography column.
Raw feed material tends to be quite concentrated for transgenic production compared with cell culture stock. Base expression levels in goat milk ranges from 2-8 g/L of protein, which doubles or triples after a filtration step. CHO systems generate between 1 and 3 g/L reliably. With CHO cells, manufacturers must often capture every last target molecule to generate product economically. The higher concentration for transgenic feedstock means manufacturers can focus on overall process economics rather than strictly on recovery. In addition, transgenic processing is continuous, which optimizes downstream utilization compared with fermentation.
Proteins expressed in milk also are more likely, compared with CHO, to emerge in the “proper” chemical and physical forms, with the right folds and post-translational modifications (such as phosphorylation and glycosylation) already in place. In a best-case scenario, such products would require less manipulation downstream and attract less regulatory scrutiny.
Although no products are currently made at the ton scale through transgenics, today’s smallish production facilities offer a glimpse into what the future holds: a chimera, if you will, of agricultural and pharmaceutical processing. Upstream, it would be difficult to tell the difference between cow-, goat-, or plant-based protein production systems from their food counterparts. Both are regulated by the U.S. Department of Agriculture and follow good agricultural practices. Once the product enters the processing plant, GMPs take over. “At this point you’re dealing with a traditional bioprocessing facility,” says Charles Clerecuzio.
With contamination always a concern for manufacturers and regulators, the challenge is managing two radically different facility types, both operationally and from a regulatory perspective: cows, for example, at one end, and a Class 10,000 purification suite at the other. Biomanufacturers have instituted best practices for dealing with endotoxins from bacteria, or viruses from mammalian cells, but not for possible infectious carry-over from live animals or insects living in food plants. The contamination issue boils down to the differences in regulatory perception between small molecule drugs and biologicals.
“FDA is usually satisfied with a 99.9% pure small molecule drug, but if it’s a biological they want to know the identity of that 0.1%,” Clerecuzio notes. Companies will pay special attention, he says, to lot identification, segregation of product and equipment, chain of custody and cold-chaining potentially large volumes of raw product.
These issues are ancient history for traditional biotech. Many bioproducts are or have been harvested directly from animals or humans. Some of these proteins are now produced through recombinant technology but one large-volume product, flu vaccine, is still cultured in chicken eggs and will be for years to come. Nevertheless, the newness of transgenics assures that these issues will be re-visited, and afforded the special welcome that always greets new technologies.
Biomanufacturers’ obsession with cleanliness upstream may be based more on fetish than fact. Fermentations and cell cultures run in classified environments present this paradox: While the outside of the reactor is kept sterile, the product is swimming in a broth of living and dead organisms, waste products and all. Chemical engineer-turned-consultant Jim Agalloco of Agallaco & Associates (Belle Mead, N.J.) has long held that employing clean areas for fermentation and cell culture is a waste of time. “You don’t need a classified environment for the reactor,” he says.
Similarly, concerns about the sterility of the source, be it animal or plant, for transgenic proteins are probably misplaced. Aseptic fill/finish is biopharm’s only hard and fast sterility standard, but manufacturers mistakenly believe they must apply it everywhere in the process, Agalloco says. It would be unfortunate for transgenic technology to be held back because of concerns about sterility, either at the agricultural source or in transfer from USDA- to FDA-regulated areas.
IT'S A TOUGH BUSINESS
Although no data suggests that transgenics companies go out of business more frequently than other biotechs, that is the prevailing opinion. Scientifically speaking, transgenic protein production should be a slam-dunk. Unfortunately, great technology alone does assure success. Large Scale Biology Corp. (LSBC; Vacaville, Calif.) had the breathtaking science and business plan to succeed in transgenics. What it lacked was time and a well-greased development pathway for its pipeline products.
The company used an abrasive spray technique to transfect nicotinia plants — a species related to tobacco — to produce therapeutic proteins for the duration of the life of the plant. On paper the process was everything that could be expected of transgenics: high protein expression yields (about 1g/kg of biomass), transient transfection (so plants did not pass the new trait onto offspring), and the need for proximal contact between the spray and the plant, eliminating the possibility of affecting other flora or engendering the wrath of environmental activists. Since it used plants, there was no chance of retaining pathogenic organisms or toxins. On the strategic side, LSBC was targeting proteins that were difficult to make in CHO cells, for example aprotinin, which is currently harvested from the lungs of cattle in slaughterhouses. Scaleup was as simple as planting more nicotinia. Larry Grill, Ph.D., an LSBC founder and now a consultant, estimates that 15 acres of nicotinia would “conservatively” yield 100 kg of therapeutic protein.
Upstream processing could easily have been confused with a food production process. After harvesting, the biomass was chopped into small pieces, treated with bleach to kill any organisms growing on the leaves, then ground, pressed to obtain protein-containing juice, centrifuged, and extracted. No seed train, stainless steel tanks or piping, or clean-in-place equipment. LSBC built a processing plant in Kentucky, for about $12 million, with a capacity of 6,000 pounds of biomass (or 6 kg of target protein) per day — by any standards, manufacturing scale.
LSBC entered Chapter 11 on March 2 and began auctioning off its physical assets on May 18, 2006.