Backroom Boys
Page 22
Of course, Venter’s database wouldn’t exert its hold for ever. No information stays proprietary in the long run. But his company might succeed at gripping tight for twenty, thirty, forty years, at what ought to have been a time of ferment, of amazing discovery. Sometimes, over the months that followed, you’d hear defenders of Venter arguing that this wouldn’t matter. There had been a tremendous controversy about the patenting of individual genes, they pointed out – and that didn’t seem to have halted work on those genes, did it? The gene BRCA1, for example, implicated in the breast cancers of women who had a faulty form of it on chromosome 17, had been patented by Myriad Genetics, the biotech company which identified it. Yet hundreds of scientific papers had since been published about it. Why should the genome be any different? Because the genome represented information a whole order of magnitude more fundamental, a whole step further upstream towards the most basic recipe for life. Asserting a proprietary right over the whole genome was not like holding a patent for, say, a particular model of car; it was like saying you had a patent on the very idea of a car in general, a patent that covered every conceivable self-propelled personal transportation device there ever had been – and ever might be in future – while your patent lasted. People kept saying, in 1998, that the twenty-first century was going to be the age of the life sciences, the century when applied genetics banished a thousand diseases, abolished a thousand sources of suffering, creating the same kind of step change in human mastery over the terms of human existence that mass mobility had done in the twentieth century thanks to the internal combustion engine. Well, in 1898, tiny manufacturers all over Europe and North America had been experimenting with different ways to fit a petrol motor together with four wheels and a transmission. Only a few designs, from a few firms, prospered; but it was the multiple experiments in multiple directions that allowed the few successful designs to emerge and to form the foundation of the car industry. No one had rights over the whole class of designs. If they had – if Mr Rolls and Mr Royce had been able to prevent Mr Ford, Mr Renault, Mr Morris, Mr Buick, Mr Daimler and Mr Benz from trying out ideas – then the twentieth century might not have been the era of the automobile after all. If Craig Venter acquired equivalent blocking powers over the stuff of life, at the turn of the twenty-first century, he might spoil the promise of this new age. There seemed a realistic chance that he might bugger up a renaissance. He might so restrict and Balkanise genome research that, even when the reign of his database ended, scientists might remain locked in a vicious circle of mutual distrust, and open access to the whole thing might never quite arrive.
It was to avoid this kind of danger that the Wellcome Trust had called a conference two years earlier, in Bermuda. Then, the pressing problem had been the tendency of individual research groups to cling to their own particular stretches of the genome, as if they had to treat them as property or risk losing the right to work on them. The biggest sequencing centres, like the Sanger, were just preparing for the first attempts at mass production, and they needed to know that they’d have a clear run through the chromosomes – while the smaller labs needed to know that if they let the big centres do that and perform the sequencing in bulk at a few big facilities, their research topics wouldn’t be whisked out of their grasp. They needed to know that the data was not going to be claimed; that it would flow back to them, and they’d still be acknowledged as the ones with the standing to analyse it and publish on it, in accordance with the ancient scientific commandment of Thou Shalt Not Muscle In. Wellcome rented an entire out-of-season hotel, the Hamilton Princess, and filled it with interested parties from both sides of the Atlantic, Venter included. Because the hotel rules specified that gentlemen must wear a jacket, the male part of the world of genome research arrived looking more formal than many of its members were entirely comfortable with. The administrators and the grant-givers wore neat managerial suits, the scattering of TV-friendly biologist media stars wore Armani, dapper geneticists of a certain age wore blue blazers with brass buttons in honour of the seaside, and the mutinous remainder turned up in plaid things, striped things and floppy green things that looked as if they had been hanging for a long, long time in garden sheds. Among them, cajoling, moved Michael Morgan.
What was required was a means to remove everyone’s anxieties about the data being grabbed by ensuring that no one could grab it. With John Sulston standing at the front, scribbling a record on an overhead-projector transparency, they thrashed out a set of rules, later to be known as the Bermuda Principles. The big sequencing centres, said the rules, would automate the release of sequence data. Whenever their production line had generated more than two thousand bases of continuous sequence, out it would go onto the internet, in its raw state, where anyone could look at it, and copy it, and annotate it, and make surmises about it. No discretion involved, no nod required from a self-interested human, no holding back bits because they looked interesting. Out it would all go, day after day, out beyond recall into the medium that was a byword for its uncontrollability. Then came the statement of intent. Aim to have all sequence freely available, wrote Sulston on the overhead, and in the public domain for both research and development, in order to maximise its benefit to society. And astonishingly, considering what was going to happen in 1998, everyone agreed. Everyone, including Craig Venter. There was some grumbling, there was some unwillingness among those whose thoughts ran yachtwards and who could see perfectly well that the Principles were incompatible with most genome-related business plans they might come up with. But no one had a business plan, yet; large-scale sequencing was a new territory, and for now the initiative lay with those who were ready to go ahead and had a strong idea of how to proceed. ‘It did at the end of the day require some … well, bullying would be the wrong way to put it, but I was accused by Craig Venter of what he called “social engineering”,’ Michael Morgan explained, unrepentantly. ‘It seemed to me that the funding organisations had to take the lead and say, “You do it this way, or you won’t receive funds.” That was the final threat that was hanging. We never had to use it …’ Conviction and the evident need for a functioning social contract for genome science did the rest, helped maybe by the universal human desire not to be left out of a crowd that everybody else seems to be joining. And so for two years the Bermuda solution succeeded. The big sequencers – the Sanger Centre, Bob Waterston’s lab in St Louis, the Whitehead Institute at MIT, Baylor College of Medicine in Texas and the Joint Genome Institute run by the US Department of Energy in Los Alamos – posted their data instantly on the net. The whole global community of interested scientists contributed their pennyworths to the collective task of understanding the data. The whole gigantic, organic, intrinsically interwoven endeavour rolled forward undivided. A virtuous circle of mutual trust had been created.
And Venter and ABI had just violated it. It seemed to John Sulston now that almost everyone had no alternative but to resist, because ‘free release’ had benefited almost everyone. The Wellcome Trust was in favour of it: ‘As a charity, it’s our duty to get the data into the public domain,’ Michael Morgan said to me. It had been in the interest of the scientific consumers of the information, who’d neither been locked out by a subscription fee they couldn’t pay nor had to divert any of their precious funding. It had been in the interest of the potential industrial consumers of the data, from the giant pharmaceutical companies intrigued by the thought of tailoring medicines to fit individual patients, to the biotech start-ups who wanted to do interesting things involving RNA and antibodies, just downstream of the genome. Why should all their separate business ideas suddenly have become dependent on Venter?
So it was clear to Sulston what had to happen. The Sanger Centre, and the other public sequencing labs, would just have to accelerate to match Venter’s speed. They would have to buy the same number of ABI’s new capillary machines as Venter was being given (ensuring that ABI did very nicely out of the situation whatever happened). They would have to pump out the data freely
onto the internet and destroy the rationale for anyone to subscribe to the proprietary database. You didn’t have to believe in the gift economy of science to think this was the only possible move. Impeccably orthodox economic logic said so too. This particular bid for monopoly could not be dealt with by the classic remedy of replacing monopoly with multiple competing businesses. Supplying genome information was a business that could only exist as a business if it were a monopoly. No monopoly, no business. Therefore, the only viable alternative to monopoly was to treat the supply of genome information exactly as the Bermuda Principles had been treating it, as one of those desirable things the market cannot furnish at all: as a public good, like a library, like a lighthouse. You didn’t have to believe in the scientific gift economy. But it pleased John Sulston all the same that the way to defeat Craig Venter was to give away the human genome as fast as he tried to sell it.
If, that is, the public project survived the week.
Sunday 10 May 1998
Craig Venter had been talking to a friendly journalist in the hope of hurrying the news agenda along a little in the direction the new company needed it to go. This morning, a million brunching New Yorkers saw the result. Nicholas Wade’s piece on the front page of the Sunday New York Times reported the surrender of the public genome project as if it were a dead certainty, almost as if it had already happened. The suggestions that Venter and Hunkapiller had made to Francis Collins and Harold Varmus on Friday – rejoice, co-operate, do the mouse genome instead – appeared to have solidified magically into statements of policy by Collins and Varmus. Wade made it sound as if they were in a state of acquiescent excitement at their rescue from the slow old toils and coils of their own genome effort. ‘Both said that the plan, if successful, would enable them to reach a desired goal sooner. Dr Collins said he planned to integrate his program with the new company’s initiative … by focusing on the many projects that are needed to interpret the human DNA sequence, such as sequencing the genomes of mice and other animals.’ The public project was referred as ‘the Federal human genome project’, or just as ‘the Government’. Oh, those Federal types! With their inefficient Federal ways, and their long Federal lunches, and their creeping, cautious, unwieldy idea of science! Why wouldn’t they rejoice when a dynamic entrepreneur burst into their stuffy halls and set their research free? Those of the million brunching New Yorkers who traded stocks with E-Trade or Schwab Online reminded themselves to look out for the new company’s shares when its IPO came. As they were intended to. Nowhere in Wade’s piece was there a single hint that any non-American might be involved in the issue in any way; that the human genome project, now invigorated by market forces, was anything but a 100 per cent guaranteed all-American enterprise.
*
The new company had no name as yet. It would be months before the branding consultants came up with ‘Celera’. Meanwhile, for ease, it had to be known as something. The disrespectful onlookers in the scientific world came up with a nickname. They called it The Ventapiller. It appeared to be very hungry.
Monday 11 May 1998
Hasty discussions over the weekend in Francis Collins’ office had led to the conclusion that the only politically feasible move was to offer the Ventapiller a cautious welcome. It was impossible to say anything directly critical. In President Clinton’s Washington, public bodies didn’t go up against private initiatives. The public sector justified its existence by demonstrating how entirely compatible it was with the interests of the private sector. The officers of the public genome project in the US would have to provide a spectacle of public-private harmony, while avoiding explicit collision with the Bayh-Dole Act and hoping that co-operation between the two efforts could be made to mean something substantial, despite the signs to the contrary. Collins’ deputy sent a letter to the heads of the major genome centres, putting the best face possible on the matter. ‘This is a very exciting development, providing a major infusion of resources and new technology at a critical juncture in the progress of the genome project.’ The public project was not yet lost, but the situation was balanced on a fine edge and trembling like a see-saw with equal weights at both ends. It could tilt either way. The best they seemed to be able to do right now was to avoid tilting it the Ventapiller’s way.
So that morning in Washington, in a display of harmony verging on humiliation, Harold Varmus and Francis Collins, together with their colleague Ari Patrinos from the Department of Energy, appeared at a joint press conference with Craig Venter, Mike Hunkapiller and the chairman of Perkin-Elmer, Tony White. Yes, said Harold Varmus politely, the advent of the Ventapiller would certainly ‘move things along more quickly’. Great news! Glad tidings! beamed Craig Venter, again. He had a new idea to share with everybody: the company, he pledged, would release its human-sequence data free of charge every three months, only charging a subscription fee to those who required day-by-day access to the database. (If this had ever actually happened, it would have represented an attack of true dot-com madness by Venter’s backers. It would have meant that Perkin-Elmer believed they could make back their $150–200 million investment out of the marginal advantage subscribers got by looking at genome data three months early. Of course, it was 1998. In 1998, people really did found companies dedicated to the proposition that if you gave stuff away, profits would somehow follow. But it never happened in this case. The company never instituted the three-months idea. Instead, over the years that followed, the company allowed researchers to download a free ration of 500,000 bases of human sequence per week. It sounded generous, until you worked out that, at that rate, it was going to take 120 years to view the whole genome. To get the ration, you had to negotiate a deliberately un-user-friendly website and then click on a button accepting the company’s rights over the data.) Well, that certainly sounded interesting and desirable, replied Francis Collins, carefully. Nonetheless, he thought it would be ‘vastly premature’ for the public project to stop its own sequencing work just yet. Venter disagreed. Mice! They were the creatures the public project should be looking at. Sequencing the mouse genome, it should be stressed, was a respectable idea, high on the scientific agenda anyway because of the mammalian comparison it would allow with the human sequence. Venter was not being unequivocally insulting – but the comical decline in dignity from human to mouse was not an accident either.
Still nobody in either camp remembered the existence of the Brits.
*
Meanwhile, across the Atlantic, on the water meadows of the River Granta it was an ordinary working Monday, and the Sanger Centre was sequencing away. It was, after all, not just a competition of ideas that the public project was getting into. The knowledge that constituted the human genome was not the kind you find in a brilliant equation or an elegant observation. It was knowledge in bulk, three billion digits long; industrial knowledge, to be obtained by industrial means. If the Sanger Centre was going to race Venter, it would be a contest of factory management. The guardian angel of British technology, had there been such a person, might at this point have laid her head in her hands, stuffed her hair into her mouth and wept.
But she would have been wrong.
*
Sequencing is not a direct process. You cannot just read the order of the chemical bases straight off the DNA molecule. The double spiral of deoxyribonucleic acid is several orders of magnitude too small for that. It exists down in the nano-realm, where objects are assemblages of individual atoms and distances are measured in nanometres, or one-billionths of a metre. In other words, far below the threshold of scale at which the human power to manipulate matter presently ends. In every one of the cells of the human body (except for the egg cells in women and the sperm cells in men), the full set of twenty-three pairs of chromosomes, one lot from each parent, floats around higgledy-piggledy inside the cell nucleus like a scattering of crimped Twiglets. These you can examine under a microscope, although at this magnification, the tip of even the finest surgical needle already looks like the prow of an appr
oaching supertanker. Each chromosome is actually composed of one incredibly long, incredibly thin molecule of DNA wrapped around a core of protein and coiled and re-coiled so densely on the core that sheer volume makes it visible. But the filament itself is only a few nanometres wide; along it, the chemical bases that hold together the two sides of the twisted ladder of atoms come only 0.34 nanometres apart. You cannot act on something this size. You cannot move it around, you cannot move a pointer along it, you cannot squeeze it through a device. You cannot even look at it. (Needless to say, if you could, you’d find that the four different bases were not handily coded in four different colours like the little metal plates and rods in the model Crick and Watson built when they were working out the twisted-ladder shape of the molecule, in 1953.) The transmission system is missing, the almost unimaginable system of ever-smaller linkages you’d need to transmit an action down from our macro domain to the nano-realm and then to transmit the result of it back up again. You can’t get at the sequence that way. It is not locked. It has evolved to be read, and it is read, every minute of every day in every organism on the planet from oak trees to earwigs. But it has evolved to be read by an interface operating on the same scale as itself. In a living cell, the information in the DNA is accessed by a piece of organic machinery called an RNA polymerase. The RNA polymerase is a molecular contraption that works its way along the double helix, very much like the read head of a tape recorder. It registers whether the base it has reached is adenine (A), guanine (G), cytosine (C) or thymine (T). Then it finds an appropriate counterpart chemical from the free-floating soup of substances in the nucleus and adds it to the string of RNA it is building. The RNA string thus contains the same data as the DNA, but unlike the DNA, it can leave the cell nucleus and pass through another reading device, the ribosome, to instruct the creation of whatever protein it is which that cell in that earwig or oak tree requires just then. Maybe some day, if nanotechnology bears out its promise, it will be possible to build something like an artificial polymerase which crawls along the double helix shouting out, ‘G! T! T! A! G! C!’ to its giant handlers. But for now, although human cells can read DNA directly, human minds have to do it by a roundabout route. You have to resort to science’s ancient manoeuvre: coaxing the answer to the question you can ask to give you the answer to the question you wanted to ask.