| Making recombinant proteins in chicken eggs is a good idea but an elusive reality. As the field of avian transgenics moves haltingly forward, it embodies the best and worst of biotech: tremendous profit potential and prospects for human benefit, along with hype, litigiousness, duplication of effort, and a culture dominated by suspicion and secrecy. |
The dirt road leading into the Michigan chicken farm is unmarked except for a street address, scrawled in magic marker on a piece of cardboard stapled to a wooden stake. No sign, no name. “We’re hard to find, we’re a very clandestine organization,” admits Bill MacArthur, founder of GeneWorks Inc., which owns the farm. “For a lot of reasons.” Chief among them is fear of industrial espionage. GeneWorks makes transgenic chickens, and it doesn’t want competitors spying on its activities.
Secrecy and suspicion pervade the emerging field of avian transgenics. GeneWorks, based in Ann Arbor, is actually one of the more open companies. MacArthur happily discloses, in general terms, his approach to transforming chickens; many of his rivals will not even do that. “It’s a very competitive industry, obviously, because the market need is so significant,” explains AviGenics Inc.'s VP of corporate development Tony Cruz. “And so it is, in fact, a race. And the winner of the race is going to get rewarded. That’s why I think the various companies are being silent.” AviGenics is a good example. Cruz refers to two “proprietary methods” for transforming chickens which he claims are “progressing quite well,” but will not say what they are.
Much, however, can be inferred about the potential and the pitfalls of manufacturing therapeutic proteins in chicken eggs. If the technology works, it will offer a vastly cheaper and more efficient way of making proteins than current overburdened bioreactor-based systems -- and one superior to transgenic goats and plants. The payoff would be immense. But none of the eight avian transgenics companies has yet made the system work, or has apparently even come close.
Their four main approaches to transforming the chicken -- retroviral vectors, stem cells, sperm-mediated transgenesis and cloning -- show promise, but all face major technical obstacles. Each company is frantically trying to solve the problems in order to be first to market.
The chicken, in theory, should be a fantastic production vessel for recombinant proteins, because hens are nature’s most efficient bioreactors. The typical hen lays 250 eggs a year, with the egg white alone containing four grams of protein. If a mere ten milligrams of that -- just 0.25 percent -- is recombinant human protein, it’ll be enough for efficient scale-up. Because chickens are an “off-the-shelf” technology whose production and maintenance have been thoroughly rationalized by the poultry industry, GeneWorks CEO Steve Sensoli estimates that an avian production system could generate proteins for $10 per gram. This is two full orders of magnitude cheaper than a state-of-the-art bioreactor growing engineered Escherichia coli or Chinese hamster ovary (CHO) cells. And the hen only takes six months to reach sexual maturity, compared to 18 months or more for livestock -- a huge advantage for drug companies rushing products to market. “It could take years to grow up a herd of these [large] animals,” says Mike West, CEO of Advanced Cell Technology Inc. (ACT). “Whereas chickens [can] fill the solar system in a few years.”
But chickens are wickedly difficult to work with. Livestock are much easier. Although they've now switched to nuclear transfer technology, GTC Biotherapeutics Inc. (formerly Genzyme Transgenics Corp.) and PPL Therapeutics plc originally created transgenic livestock by micro-injecting DNA into the immature egg, the oocyte, following in vitro fertilization. In chickens, that’s almost impossible to do, because the tiny egg is buried in the yolk. “It’s like finding a pea in a swimming pool full of tar,” says GeneWorks' MacArthur. And, so far, chicken egg yolks can’t be fertilized in vitro. Going after sperm is even more futile, since as many as thirty sperm may end up fertilizing the yolk, and there’s no way to know which one will end up contributing to the embryo.
Every possible solution brings its own problems. One option is to wait until the egg is laid, then use retroviral vectors to deliver the human gene of interest through a temporary window in the egg shell. This is the only proven method for generating transgenic chickens. “We know that it works,” says MacArthur. Unfortunately, gene integration is random and thus susceptible to gene silencing. The expressed gene is often lethal to the developing embryo. Finally, germline transmission is rare.
The overall result is a stultifying inefficiency. AviGenics, in early research that was finally published in the April 2002 issue of Nature Biotechnology (Nature Biotechnology 20: 396-399 [2002]), had to raise over a thousand chickens to get one that expressed even microgram quantities of a marker protein in its eggs. The marker gene was inserted using a retroviral vector based on avian leukosis virus. Cruz argues that one bird is enough -- “One transgenic founder can give rise to thousands of offspring in a matter of months,” he says --but the depressing truth is that each therapeutic protein project will probably require the genetic engineering of thousands of birds to find one stably expressing high levels of protein without ill effects on the bird. That is, if it can be done at all.
Cruz claims AviGenics has proprietary nonviral methods improving these odds, but won’t reveal them. MacArthur thinks the Athens, GA-based company is removing the egg from the chicken before laying, while still in the oviduct, the hen’s reproductive tract -- a very labor-intensive process. (Cruz won’t deny or confirm.) “What AviGenics is doing, they take it very high up [in the oviduct], after it’s fertilized, they take it and manipulate it,” he says. “They take that yolk, put it back into the chicken, and it lays an egg.” This allows prescreening for transgene insertion and thus greater efficiency. Unfortunately, most of the time the inserted human DNA remains episomal and doesn’t integrate, or kills the embryo, says MacArthur, who doesn’t think AviGenics can do any better than one transgenic bird out of 5,000 injected embryos. “Take 5,000 hens and maybe hit or miss?” he says dismissively. “Let somebody else do that.”
MacArthur is sticking to retroviral transfection of the laid egg, despite the perils of gene silencing. (Chicken embryos shut down viral elements in their genomes.) He’s trying various tricks and stunts to jack up protein expression and to fool the chicken into ignoring the vector so that the gene can do its stuff. It’s mostly trial and error. “In the end, it’s a numbers game,” he says. “The more transgenic birds you make [and] the more you breed ‘em out, the more likely it is you’re going to get one with decent expression levels.” So far, no luck.
MacArthur founded GeneWorks, the first avian transgenics company, in 1996. A molecular biologist with a veterinary degree, he has raised over $18 million in venture capital and hopes to succeed within two years. If he does, GeneWorks will still have to contend with the FDA, which is critical of retroviral methods because of the theoretical possibility of recombination or complementation with endogenous viral elements in the chicken. That’s more of a poultry health issue than a human one, because GeneWorks’ vector, reticuloendotheliosis virus (REV), does not infect humans, but regulatory issues are a major hurdle for any company working with retroviruses.
Stem cells are cleaner, but have their own problems. In theory, it should be possible to achieve targeted gene transfer in chicken embryonic stem cells, without viral vectors, as is routinely done in mice. The chicken, in fact, is the only animal besides the mouse whose stem cells are capable of creating a “chimera,” or genetic mixture, after implantation in another individual. Thus, germline transmission and at least some high-performing birds should be achievable. The stem cell approach “allows you to stick in transgenes of any size, that give you high levels of expression and tissue specificity, and you can do knockouts,” says Rob Etches, VP of research for Burlingame, CA-based Origen Therapeutics Inc. (Etches was the first to make chimeric chickens.) Knocking out chicken immunoglobulin loci with human ones allows one to make human polyclonal antibodies in the bird, a key element of Origen’s business strategy.
But stem cells have two (so far) insurmountable problems. “For a quirky reason they tend not to become reproductive cells very easily,” says ACT's West. “So when the chickens with the particular genetic modification breed, they don’t pass on the gene.” The second, more intractable problem is the inability to keep stem cells in culture long enough to select for the ones that have incorporated the human genes. “Maybe one in ten thousand integrates,” says MacArthur. “Only those that have maintained an integrated copy will grow as colonies. It takes about a month.” So far, says MacArthur, no one has kept transformed stem cells intact, without differentiating, longer than a week. The right culture medium must first be found.
Origen is counting on stem cells to make its chickens. Etches won’t comment on culture methods, nor will he reveal how long he can keep transformed avian stem cells in culture. “That sort of technology should be peer reviewed before I make statements about where we are in our technical progress,” he says. MacArthur doesn’t think Etches is anywhere close. Of course, neither is MacArthur, who also has a stem cell project going. Can it be done at all? “We’re asking basically an impossible task,” he says. “What’s the probability of success? Low. Will it happen over the next fifteen years? Yep. Is it going to cost millions of dollars to do it? You bet. We’re looking at … a $50 million project.”
Advanced Cell Technology, the Worcester, MA company that vaulted to fame by cloning the first human embryo last year, also has an avian project. ACT president Mike West co-founded Origen in 1997, but left a year later. Instead of the stem cell, he has opted for the slightly more mature “embryonic germ cell,” which is destined to become sperm or egg. This solves the germline transmission problem, but these cells are just as hard to culture and transform as stem cells. “We’ve had a hard time getting DNA into these cells, certainly to get a targeted modification,” says West, one of the few people, along with MacArthur, working in the industry who will publicly admit to specific problems. Although West says the system “is very close to being there,” ACT has committed only two scientists to chicken work.
No one knows how many people Viragen Inc. has working on chickens, because the company won’t say. The publicly traded Plantation, FL firm is the most tight-lipped company in this secretive industry. “As to any of the technology, I’m afraid you’re not going to find us very cooperative,” says Viragen CEO Gerald Smith. “We don’t put highly technical information in the public domain, nor are we required to.” But Viragen puts out no meaningful information, technical or otherwise, about transgenic chickens. It will not reveal its general approach to genetic transformation, much less results. “It is not an appropriate time to talk about our programme in avian transgenics just now,” writes Viragen collaborator Helen Sang, a senior scientist at Scotland’s Roslin Institute, in an email.
Despite the information blackout, some things can be inferred about Viragen’s approach. Sang has worked for more than a decade on transforming the early chicken embryo. “She’s the acknowledged expert on early avian embryogenesis,” says MacArthur. “You can’t teach her anything. She’s developed it all.” So Sang and Viragen may be removing the newly fertilized egg from the oviduct, transforming it in vitro, then either returning it to the bird or to a surrogate egg shell to hatch. This method is labor-intensive, technically demanding and (using micro-injected plasmid DNA) inefficient. It’s possible that Sang is using transposable elements -- “jumping genes” -- as vectors, because they integrate efficiently. (She has published on this.) But transposable elements may be unstable, making reliable germline transmission a problem.
Avian Transgenics Companies
|
Company
|
Main approach to
transformation of chicken
|
|
Advanced Cell Technology
Worcester, MA
|
Avian embryonic germ cell
|
|
AviGenics
Athens, GA
|
Retroviral-mediated; other methods confidential [See story]
|
|
BioAgri
City of Industry, CA
|
Sperm-mediated transgenesis
|
|
GeneWorks
Ann Arbor, MI
|
Retroviral mediated; some stem cell work
|
|
Origen Therapeutics
Burlingame, CA
|
Avian embryonic stem cell
|
|
TranXenoGen
Shrewsbury, MA
|
Direct egg transfection; other methods
|
|
Viragen
Plantation, FL
|
Confidential [See story]
|
|
Vivalis
Nantes, France
|
Avian embryonic stem cell
|
The Roslin Institute (along with Viragen, presumably) is also trying to clone chickens. So are AviGenics and TranXenoGen Inc. Cloning, or nuclear transfer, would obviate the need to culture stem cells for long periods, because it’s easy to precisely transform adult cells before inserting them into the chicken egg. But no one has yet managed to clone a chicken. “There are a number of technical hurdles [to] overcome that the mammalian folks don’t face,” says Origen's Etches. “The egg is a much more difficult structure to work with, with its large, yolky mass.”
“You would predict that [cloning chickens] would be far more inefficient than mammals,” agrees ACT's West. “But probably doable somehow.” MacArthur thinks cloning chickens is a real longshot, at least in the near term. “We don’t think that is anywhere near even as probable as the stem cell,” he says.
TranXenoGen has a more sophisticated disclosure strategy than Viragen. Instead of keeping its methods under wraps, the Shrewsbury, MA firm says its scientists are trying the full gamut of techniques. “Direct-egg transfection,” “gene-testes transfection,” chicken embryonic cells and cloning -- TranXenoGen does it all. On May 30, 2002, the company announced that it had achieved expression of two monoclonal antibodies in the egg whites of chimeric chickens, using the direct-egg method. “We believe that this is the first reported expression of monoclonal antibodies in transgenic chicken eggs,” said CEO George Uveges in a press release.
By the industry’s standards of openness, this amounted to a Magna Carta. However, the expression level achieved—“up to 1.5ng/ml”—is minuscule. “Generally you need about a mig per mil, a gram per liter… to get in the [commercial] arena,” says MacArthur. More importantly, TranXenoGen hasn’t demonstrated germline transmission of these genes. Difficult as it is, many companies and labs have been able to engineer chimeric birds to express foreign proteins in their eggs. But only high expression levels, together with germline transmission and reproducibility, constitute true proof of principle. No one, including TranXenoGen, has yet demonstrated this.
Uveges is not any more forthcoming than his competitors about the details behind the hype. Is TranXenoGen using plasmid DNA to transfect eggs? “You’re getting into the science part,” he protests. How is TranXenoGen getting genes past the tough defense mechanisms of the testes and sperm? “I’m not going to go into details.” How efficient is direct-egg transfection in generating chimeric birds? “We’re not commenting on that at this point.”
“They won’t share their data,” complains MacArthur. “Nor do they have to. Coca-Cola doesn’t share its recipes either. [But] it’s impossible to get an objective view of what’s going on there.”
Nonetheless, Uveges expects TranXenoGen to achieve “repeatable” protein expression by mid-year -- his version of “proof of principle” -- and commercial levels of expression by the end of 2002. If the company succeeds, strategic partnerships and more cash should flow in, perhaps enabling a new stock offering. The need is there. Two years ago, TranXenoGen raised $19 million in an IPO on London’s Alternative Investment Market, and almost half the money is now gone.
One possible sleeper company is BioAgri Corp. Founded four years ago by Taiwan native and CalTech graduate Ken Wang, the tiny City of Industry, CA firm last year reported making 10 transgenic chickens (in 88 attempts) using sperm-mediated transgenesis, a notoriously difficult technique. The goal is to insert the human gene into rooster sperm before fertilization, automatically achieving germline transmission and skipping the chimeric generation -- thus shortening product development by six months over methods that involve transforming the multi-cell embryo and generating chimeric birds.
“Sperm-mediated transgenesis is another Holy Grail,” says MacArthur. “You can’t do it, it doesn’t even work in mammals. It’s never been faithfully reproduced -- never.” It’s almost impossible to integrate foreign genes in sperm, because sperm are quiescent -- they don’t divide -- and their DNA is condensed.
But BioAgri uses a trick: It links its foreign gene to a monoclonal antibody reactive to an antigen on sperm cells. This causes the DNA to bind to the sperm, and also protects the DNA as it penetrates the sperm, believes BioAgri chief scientist Jin Qian.
The company has a long, long way to go. It must now reproduce these very preliminary data, demonstrate germline transmission, and show high, tissue-specific protein expression. This is unlikely to happen soon: The company is still trying to get local officials to approve its plan for a chicken farm, and will need many millions of dollars to build the necessary feeding and breeding facilities, as well as add to its meager U.S. staff of six.
Money is a worry for all these companies. They took advantage of the biotech finance boom of 2000 to fill their coffers, but cash is starting to run low. Pilot project agreements with drug companies bring in little. What’s needed are full-blown production contracts, and those won’t happen until proof of principle is demonstrated. “There are too many players, and [the field is] too young,” says MacArthur. What drug company executives are saying, he adds, is, “’There’s five of you out there, and nobody’s got a damn thing to show yet. So we’ll just wait. You guys spend your venture capital ‘til you’ve got a result, then come talk to us.’”
Progress, in the meantime, has been agonizingly slow. “Not a damn thing’s happened in the last 10 years,” says MacArthur. “It’s still the same old technology. Some improvements on expression vectors, some things like that, but the general mechanics are unchanged. And it’s not through lack of effort. I would say clearly $50 million has gone into this, counting all the companies working on it, and [we’re] pretty much right where we were 10 years ago. We’ve ruled out a lot of things, so we’ve tried a lot and failed a lot, but as far as the basic technology -- no.”
One barrier to progress is lack of cooperation. “Everybody shuts up,” says MacArthur. “There’s almost no transfer of data from one company to another. So I’m sure there’s redundant work going on.”
Cooperation seems to be the last thing on these companies' minds. Instead, a series of bitter lawsuits have poisoned relationships within the industry. Two years ago, AviGenics, Viragen and the Roslin Institute were talking about a possible collaboration. (The Roslin Institute’s Helen Sang was then a member of AviGenics’ scientific advisory board.) Then things turned ugly. When Viragen and the Roslin Institute announced a chicken partnership in December 2000, AviGenics charged Sang with violating her consulting agreement and with sharing AviGenics' trade secrets with Viragen. Sang, Viragen and the Roslin Institute filed a pre-emptive lawsuit in California against AviGenics and denied the charges.
Since then, Avigenics has sued Sang and Viragen; Viragen and the Roslin Institute sued right back, this time for patent infringement; and Avigenics countersued. According to the Athens (Georgia) Banner-Herald, the two companies have trashed each other in Clarke County Superior Court. The litigation continues.
The strife is not helping the field progress. Data sharing is almost nonexistent. MacArthur says professional meetings are uninformative. “Unless you get the whole picture, it’s a worthless talk,” he says. “I’ve been to so many talks where people say, ‘Yeah, we have migs per mil.’ ‘How did you do it?’ ‘Well, we can’t say.’ Well then… I don’t want to hear it. And they don’t want to hear it from me.”
ACT's West doesn’t think there’s any more secrecy in avian transgenics than in other biotech fields. “Cutting-edge type areas are always highly competitive,” he says. “And I haven’t seen any more [secrecy] here in the avian sector than anywhere else. In fact, maybe less.” But putting out vague press releases celebrating minor advances is probably counterproductive. “A lot of people are out there flapping their lip,” says MacArthur. “And the avian industry as a whole will look bad. If they keep promising and not delivering, eventually everyone is going to turn their back on it as a nonviable technology.”
Maybe there’s no solution to this dilemma. The profit imperative works against public disclosure and the sharing of information and resources between companies. “They are constrained by commercial pressure to keep the details of their work under wraps,” noted an editorial in Lancet Oncology. “In doing so they risk wasted time and duplication of effort by other researchers seeking the same goals… this seems a great shame. Both the chicken and the egg seem to have taken second place to the buck.”
Ironically, the field of avian transgenics is hardly on a confident march towards success and riches. It's more of a blindfold race towards an ever-receding finish line.
-- Ken Garber (kengarber@prodigy.net) is a freelance science writer based in Ann Arbor, MI. |