In august 2002, Jeronimo Cello, Aniko Paul, and Eckard Wimmer published a now famous paper in Science reporting that they had artificially synthesized a live poliovirus. They did so by following the sequenced polio RNA genome, which is published online, and by connecting corresponding strands of DNA, some of which they purchased via mail order. The synthesized DNA then was transcribed to form viral RNA and added to a cell-free extract–a fluid made up of the contents of punctured cells–resulting in the production of a virus that paralyzed mice. The scientists conducted their work in a lab at Stony Brook University headed by Wimmer, who played a leading role in the genetic sequencing of polio.
Their study, titled “Chemical Synthesis of Poliovirus cDNA: Generation of Infectious Virus in the Absence of Natural Template,” proved the accuracy of existing genetic sequencing methods and held out the possibility that this technique could be useful in vaccine creation. Yet when it was published–less than a year after 9/11 and the 2001 anthrax attacks–much of the reaction was negative. Cello, Paul, and Wimmer's colleagues scornfully labelled their work irresponsible, and the study was the focus of intense media attention due to its dual-use implications–the potential that it could be used to synthesize other deadly viruses such as smallpox.
Since that time, the Cello, Paul, and Wimmer study has become one of the most cited examples in dual-use research and synthetic biology policy debates. To investigate the security, safety, and regulatory implications of such work, Michael Selgelid, a bioethicist at the Australian National University in Canberra, and Lorna Weir, a sociologist at York University in Toronto, conducted interviews with these three researchers in 2008 as part of a larger collaborative research project on the burgeoning field of synthetic biology. Selgelid and Weir's aim was to allow the three scientists to look back on the impact of their study and discuss the developments in both science and policy since its publication. Although Cello, Paul, and Wimmer were interviewed separately, many of the same questions were asked of each, and so their responses appear side by side.
SELGELID AND WEIR: When did the idea for the synthetic polio study first occur to you?
WIMMER: In 1991. But the biggest question at the time was who would pay for the research. We didn't have any grant money. Also, it was difficult work because we needed very large numbers of oligonucleotides [short fragments of DNA or RNA]. Back then, it was a major and esoteric undertaking to synthesize even one gene–let alone the entire poliovirus genome.
SELGELID AND WEIR: How was the research financed?
WIMMER: By the U.S. Defense Department. You couldn't really go to the National Institutes of Health (NIH) and say, “We want to synthesize poliovirus.” During the Clinton administration, I had been recruited to work with the Defense Advanced Research Projects Agency (DARPA) to evaluate grant applications for projects that might help protect the United States against bioterrorism. This was long before 9/11. We met about three times a year to review these applications, and the idea was that DARPA might look favorably at proposals that NIH wouldn't fund.
After about two years of reading successful grant applications, I decided I had a sufficiently DARPA-esque project–that is, to make synthetic poliovirus from scratch. If I succeeded, it would be a wake-up call because then anyone would be able to do the same, including synthesizing other viruses such as smallpox. On the basis of that argument, DARPA gave our lab about $300,000.
PAUL: In 1998, Wimmer and I wrote the DARPA grant proposal. We received the grant, which I believe was for three years.
Cello came from Sweden in 1999 as a postdoc to work on the project. I trained him in genetic engineering and worked alongside him, but he did almost all of the benchwork. The poliovirus genome was put together from three large fragments. Cello was involved in making that first fragment from oligonucleotides–a very time-consuming exercise. Since we had shown it was possible to synthesize the polio genome from this work, we purchased the other two fragments Cello then did work with the mice and the tissue cultures, and we completed the virus synthesis in cell extracts that were derived from human cancer cells.
“By the time we went full blast into DNA synthesis, it was very evident–more so than any experiment that I have ever done–that we would succeed.”
SELGELID AND WEIR: How did synthesizing poliovirus fit into your overall research?
WIMMER: I'm trained as an organic chemist. I chose to work on viruses because, from the very beginning of my career, they seemed to me to be chemicals. Yet they had the intriguing property of behaving like they were living if put in the right environment. In my first 13 years in polio virology we examined the structure of the virus. This culminated in the determination of the sequence of its genome, which was published in Nature in 1981.
The next step was to ask how the poliovirus replicates when it enters the cell nucleus. Eventually, we were able to make a cell extract with no living cells in it. If we took this extract and added viral RNA, the genome would begin to grow. Not only that, it would replicate and make viruses. We published that finding in a 1991 Science article “Cell-Free, De Novo Synthesis of Poliovirus.”
SELGELID AND WEIR: Did anyone raise security concerns after that first 1991 study was published?
WIMMER: No. At the time, bioterrorism wasn't much of an issue, although DARPA and certain military agencies were worried about security. Every time I went to a review committee meeting, one of the officers showed up to give a talk on bioterrorism. This was in the early to mid-1990s. Nothing much happened until the anthrax attacks after 9/11. When we initially published that you can make poliovirus in a cell-free juice [from natural viral RNA], no one came to us and said, “That's scary!” But there were people who were worried about the ethics of creating life in a test tube.
SELGELID AND WEIR: Were there particular scientific challenges in synthesizing the polio genome?
WIMMER: Very few. The scientific challenge was to convert the synthetic RNA into a live virus outside living cells, and we had already devised the cell extract system to do this. By the time we went full blast into synthesis, it was very evident–more so than any experiment that I have ever done–that we would succeed.
SELGELID AND WEIR: Some people have claimed that your 2002 article had a predictable result and therefore was scientifically uninteresting. How do you reply to such claims?
WIMMER: It's difficult to respond to that. If you claim in experimental science that something can be done, you must do the experiment. Talking about it doesn't count. Of course, in our paper we had said that once you actually succeeded in putting together a genome for viruses chemically, you could change the sequence of the genome to a large extent, which could be very beneficial. So for example, in humans, a specific set of nucleotides are used for encoding amino acids–these are known as codons. Because there's a redundancy of codons, there are some that are used very rarely, while others are used often. Now certain pairs of codons don't like each other–that is, one codon placed next to another codon may cause friction, whereas other codon pairs may love being next to each other. This is called “codon pair bias.” Changing codon pairing by using strictly synonymous codons [codons that encode the same amino acid] doesn't change the encoded amino acid or protein, but it may modify the amount that's synthesized. We have shown that if you introduce less frequently occurring pairings into the poliovirus genome–in other words, pairs that hate each other–the virus is dead or, depending on the changes, weakened. This was a big surprise, and we published a paper on this, “Virus Attenuation by Genome-Scale Changes in Codon Pair Bias,” in 2008 in Science suggesting that this could be a novel strategy to generate vaccines.
CELLO: Our experiment showed that the synthesis of an infectious agent was feasible. We also proved that the sequence of poliovirus–and therefore, the methodology for sequencing–was correct. We showed that viruses are chemicals. It's a simple idea, but you have to prove it. It's not enough to make a hypothesis. When we assume things without proof and then we repeat that assumption, it's not science.
Also, when the 2002 article was published, people didn't focus on the finding that the synthetic virus was much weaker than the wild version in mice. They were mainly interested in the synthesis of polio itself. When we synthesized sequences, we put markers in regions that we supposed wouldn't affect the virus. One of those markers ended up actually attenuating the virus. Later work showed where the attenuation point was located.
Since then, we've synthesized viruses with this attenuated region and are working to stabilize this attenuation for use in vaccines. Many vaccines, particularly the Sabin vaccine [an orally administered polio vaccine], are attenuated but can revert easily to non-attenuated forms. With a synthesized virus, we hypothesized that we can change the codons without changing the expression of the protein or the amount of protein produced. In other words, such a virus could be attenuated but very stable. If it's stable, it can be used in a vaccine that may be useful also in the treatment of cancer.
“The day the paper was published we became outcasts. Even our colleagues didn't talk to us. Most of them think we caused an unnecessary uproar. They talk to us now, but many still think it was a bad experiment.”
SELGELID AND WEIR: Do we understand you correctly that from the time the synthesis of poliovirus was first proposed to DARPA you had at least three explicit intentions: (1) demonstrate proof of principle, (2) send a message that would-be bioterrorists might be able to synthesize viruses, which also has implications for eradication, and (3) use codon pair bias to weaken the virus?
WIMMER: There actually was a fourth as well: proving the idea that one can look at viruses as pure chemicals. If that's the case, you can synthesize any virus.
SELGELID AND WEIR: Do you think we are already there now with smallpox?
WIMMER: Absolutely. In fact, the progress is much faster than I imagined given Craig Venter [a biotech entrepreneur and biologist who is best known for competing against the U.S. government to sequence the human genome] and his 2008 synthesis of the genome of Mycoplasma genitalium.
SELGELID AND WEIR: Were any security concerns raised in the editorial or peer review process with Science?
CELLO: No. The main purpose of the study was to demonstrate the principle that viruses are chemicals and that, as chemicals, they can be synthesized. That was in 1998. When the paper was eventually published in 2002–after 9/11–its significance was completely different, and so a debate about whether it should have been published or not ensued. Potential uses of our research findings by rogue interests were discussed in the draft paper we submitted to Science, but the journal edited them out.
SELGELID AND WEIR: What kind of publicity had you anticipated before the article was published?
WIMMER: I didn't anticipate the hostility. I mean, the day the paper was published we became outcasts. Even our colleagues didn't talk to us. Most of them think we caused an unnecessary uproar. They talk to us now, but many still think it was a bad experiment.
SELGELID AND WEIR: How did the press find out about the study?
CELLO: What happened was that Science put an embargo on the paper but submitted it to a large number of journalists. We couldn't talk about it in public, which we let Stony Brook University know about. All of a sudden the day the embargo ended, Stony Brook realized that our article was huge; so it called a press conference. The university media office came up with the strategy that Wimmer would be the only person to talk to the press. Sixty to seventy networks broadcast from the university, and journalists called us from all over the world. The telephone lines crashed. It was amazing.
PAUL: Wimmer was on the phone continuously. Cello and I were present, but we were literally hiding because the press was so hostile. We didn't want people to know we were involved in the work. This was definitely considered bad publicity for the university. Then there was the overflow that came to Cello and me. This went on for months. It's still going on today, but to a much lesser extent.
SELGELID AND WEIR: Was there ever any serious consideration of not publishing the paper for security reasons?
PAUL: No, there was absolutely nothing dangerous with our work. Everybody is vaccinated against polio. It was totally harmless from that point of view. For terrorists, it would have made no sense to start putting this virus together because they already had an ample supply of poliovirus available. Polio is present in old medical samples that are stored in freezers, and it can still be bought from suppliers. It's not like smallpox, which isn't as easily accessible.
WIMMER: Nobody is really scared that someone will want to make poliovirus. It became more of an issue in 2002. The World Health Organization wanted to eradicate polio–and then along comes Wimmer who says no virus will ever be eradicated because the genetic sequences for many viruses are available on the internet. And if they're online, they can be synthesized.
CELLO: The government knew everything, and they never put any restrictions on our publishing. In the end you can do two things: You can give this information to the government to keep secret, or you can make it public. The drawback of the second option is that rogue minds could use this information against society. But is the problem really rogue minds or government secrecy? Look at past uses of biological weapons. Very few have been used by rogue groups; most were produced and used by state-run programs.
SELGELID AND WEIR: If putting the sequences for dangerous viruses online means that eradication isn't possible, is that a reason for not publishing them?
WIMMER: That's an interesting question. Should one restrict the dissemination of scientific information? I personally don't think information should be restricted. But there is a line. Craig Venter sequenced smallpox, and it's been published. At the time he published it, nobody ever thought you could synthesize viruses. It's typical of the rapid progress in biotechnology and medical research. That, of course, brings up the question of whether the circulation of information should be slowed down or restricted. I think research is always ahead of people's thinking. Regardless, if you try to prevent publication, you can't stop it everywhere. You might be able to stop it in the United States, but then it could be published internationally.
“I'm amazed that some people think we can now, with synthetic biology, magically do whatever we want. It isn't easy to create something that's much more pathogenic than what nature has created.”
SELGELID AND WEIR: Can you imagine a paper that shouldn't be published?
WIMMER: I've been asked that before. If somebody found out how monkeypox could be changed genetically to infect and spread among humans, I guess that would be something you wouldn't want to publicize. But why would somebody bother manipulating monkeypox when they could just make smallpox? As the molecular biologist Joshua Lederberg said, “There's no technical solution to the problem of biological weapons. It needs an ethical, human, and moral solution if it's going to happen at all.”
SELGELID AND WEIR: Does synthetic genomics create new biosafety or biosecurity problems?
CELLO: Biosafety issues predate synthetic biology. People have been manipulating germs for 100 years and have been using restriction enzymes since the 1970s. I'm amazed that some people think we can now, with synthetic biology, magically do whatever we want. It isn't easy to create something that's much more pathogenic than what nature has created. But I do agree that events during the last decade have brought more sense of responsibility regarding biosafety issues. There has to be regulation. You can't just do anything. Synthetic biology does pose more of a threat than the life sciences previously had, so we have to be more careful regarding ethical issues in synthetic biology.
PAUL: I think recombinant DNA created a series of biosafety problems, and synthetic biology has increased them. With genetic engineering you could make smaller changes to genomes, but now, with synthetic biology, you can synthesize a new organism and make much larger changes. The way I see it, the danger is that if some terrorist group wants to make an organism that's harmful, they'll be able to equip a lab to do so for only a few hundred thousand dollars.
SELGELID AND WEIR: It's often said that the scientific community needs to better educate and govern itself. What role could, or should, education of scientists and codes of conduct play?
WIMMER: This brings up the question of licensing, where researchers take an examination like a medical doctor in order to do research and have a code of ethics that says they cannot produce anything harmful. The problem is that, in research, you don't really know what the outcome of an experiment will be. You don't really know if within five years there could be a seemingly innocuous, but ultimately horrible, result, like the 2001 Australian mousepox virus experiment. The people who did that had no inkling that their genetic engineering would create a virus that kills vaccinated animals. We're living in an era where we should be aware of the dangers, but we shouldn't drive ourselves nuts. I'm worried that governments could block scientific investigation.
SELGELID AND WEIR: What do you think about the way that the policy debate has unfolded and the direction in which policy making has gone since the publication of your article?
CELLO: The debates alerted scientists that they need to be less naive and to think more about the implications of their work. Also, a new review process for scientific articles was created due to reaction to our paper. But things will happen, such as the mousepox experiment in Australia, that one cannot foresee. Law enforcement also is checking the background of everybody. I'm in charge of a Biosafety Level 3 lab where we work with select agents, and so the FBI has checked my background. My only concern is that having so many rules may deter scientists from actually doing groundbreaking science. My opinion is that policies and oversight mechanisms should be developed and that we scientists should think about bioethical questions and the impact of our research on the world.
