Building the Ecological Immune System

Fred Hapgood

Policy groupies, like generals, always prefer to refight the last war. A case in point is how the battles over biotech are being forced into the ruts dug during the hostilities over nuclear power twenty- five years ago. The rhetoric about the powers of big business, the fecklessness of captured regulators, the importance of stopping the development of the technology, and increasing the weight of regulation are all lifted straight from that playbook, as are, in many cases, the guerilla tactics.

Unfortunately for both sides and the middle, these two technologies could not be more different. One way of assessing the social meaning of a technology is to plumb the pool of potential applications. Nuclear power plants made power, period. They had a small and stable universe of partners. Genetically engineered plants are doing everything. They are cleaning air and water and concentrating minerals for the mining industries. They are being used as sensors, luminescing when some change is detected in the environment. They are being developed to make specialty fabrics (such as industrially useful amounts of spider silk), materials (biodegradable plastics, specialty woods), oils, and lubricants. The list of gardening applications - - fluorescing flowers, better habit control, longer flowering seasons, low-light adaptations -- spreads over every gardener's wish list. All in addition to the food applications that generally fill the papers and which also run on beyond any counting: modifying plants to resist pests or increase their tolerance for poor soils or harsh climates, crowding tolerance, resistance to spoilage, improved photosynthesis, added nutrients and vaccines, and subtracted allergens, toxins, or carcinogens. I won't even begin with the possibilities for "bioartists" who want to "raise issues".

Further, as impressive as the current state of the art is, the fundamental capacities of the technology are themselves on a rapidly rising curve. The biotech one reads about today rests on the use, or reuse, of already evolved genes, genes that now exist in nature. As handsome as that universe is, there are a number of techniques under development that contemplate expanding it with novel genes, genes with no natural representatives. One filters the results of a process that generates very large numbers of mutations randomly; another is to combine parts of genes from several species into a single portmanteau gene (as opposed to simply lifting a gene intact from one species). The next step, programming genes directly, the way we do computers, is years off but clearly in sight.

Nor is programmability the end- state of the technology by any means. There are several labs around the world working on reengineering cells themselves to support a much wider range of reactions and materials. One common theme in this research (there are at least two research groups in my own city working specifically on this) is to find ways of getting cells to build biological computers that would be maintained inside the cell and yet be programmable from outside. (Ref for fact checker: Construction of a genetic toggle switch in Escherichia coli. Nature 403(6767):339- 342.) Such onboard cellular computers would give us a very flexible control over the nature and operation of the synthetic machinery of the cell. You do not have to go very far down this route before you begin to arrive at a general purpose self- replicating molecular assembler, an artifact that is to atoms what a computer's processor is to bits, and that works for wages that can be measured out in tablespoons of bouillon. In short, this technology is like programming: it speaks to just about every industrial and cultural sector.

Biotech differs just as radically from nuclear power in a second respect: the price of setting up a business in that sector. This useful number, what economists call 'entry costs,' tells you at a glance much about who is going to be using a technology, why, where it will fit in a civilization, and how and by whom it will be regulated. In nuclear power these costs were in the billions of dollars. The access costs of biotechnology, measured by the prices of the materials and equipment required, are close to zero.

This point can be missed because we often read about biotech in the context of the development of the science. Maintaining a fully staffed research lab is indeed expensive, but the cost of engineering products based on the results of that research goes down every year. Instruments get cheaper and more powerful, more laboratory procedures get codified into paint- by- number textbook protocols, and more genes get sequenced and published. You can get a feeling for the trajectory of the costs by paging to the back of Science magazine and looking at the ads of the service bureaus competing to make genetic sequences to your specification. .75 a base. .60 a base. .49 a base. This is an intensely competitive business and that intensity reflects the low cost of access.

Or you might check out a catalog from Edvotek, a biotechnology education company. I went on a shopping spree on their online site, clicking on everything I thought I might need to start doing microbial biotech in my kitchen. I came up with a total bill of $10,000, which is about middle of the range as hobby startup costs go. When I was through I called Edvotek's President, Jack Chirikjian, to ask him what else I needed. "$10,000!" he said, obviously exasperated at my Yuppie self- indulgence. "You don't need to spend a penny more than $3500!" (A plant biotech lab might be a little more -- Edvotek is going to announce this line at about the time you will be reading this article -- but not by much.) And if $3500 is too much you might consult a series of 1998 articles in Scientific American's wonderful Amateur Scientist column which tell you exactly how to build many of the basic tools yourself.

There are information access costs, but information gets onto the internet and from there it goes everywhere. The intrinsic interrelatedness of living things (apples and oranges, the two organisms that symbolize the very idea of mutual difference, share more than 90% of their genome) makes the effect of this information distribution especially powerful. Discoveries coming out of work on, say, the growth rate of potatoes, might well be applicable to onions or grains or apples. Once a gene is transplanted into a single plant variety, classical breeding techniques can be used to spread it to all the varieties in that species. And if the development technology is cheap, the distribution technology - - self- replicating seeds -- is beneath cheap.

What these low costs say, among other points, is that biotech, far from being a conspiracy of rootless multinationals, is likely to be their undoing. Large companies need to be able to control their market; their bets are too big to risk. In low access-cost technologies, like writing or fashion design or crafts of any kind, corporate strategies are a risk, because new talents or ideas can pop up at any time and do something unpredictable to the market.

Biotech is doubly inhospitable to large companies in that the product is self-replicating, which raises extremely knotty intellectual property protection issues. Monsanto tried to put its finger in that dike by funding development of a so-called "terminator" gene, a gene that would prevent the genetically modifyed organism from replicating, but as the software industry well knows, it always costs more to put copy protection in than to strip it out. I don't know how long biotech- savvy anticorporate activists would need to figure out how to inhibit any possible form of a terminator gene, but almost certainly they would need less time than it took to design it. (Indeed, the best hope for bigness in biotech is if huge upfront testing costs are imposed on the sector. The irony of this common interest between big capital and the activists is well- known inside the industry, but isn't much talked about.)

Both the critics and the sector, by which I mean both the academic and industrial sides, seem wedded to the nuclear power model. The critics are reluctant to admit that a general ban is impossible, while the sector is reluctant to admit that biotech is as uncontrolled as it is, because it feels that would unsettle the public. I recently posted a query about 'garage biotech' to a professional mailing list and got the following reprimand from an executive.

Articles (that)... suggest that much is easily do-able (and it is) for little money (and it is) and then profile what can be done, can't help but fire up the full spectrum of imaginations (the good, the bad, and the ugly). ... Taking this activity below the public radar and into garages is the last thing we ought to be doing if we want to continue to gain greater public acceptance. That it is being done, I have no doubt, but to directly or indirectly "promote" it through articles would in my view be the height of irresponsibility, at least for the foreseeable future.

The problem with denying the truth is not just that the strategy is bound to explode in your lap sooner or later; time is being lost that could be used to organize real solutions to problems that are not going to be chased away by denial. For instance, the concern that genetically engineered organisms will escape into the wild seems more than plausible; it seems stone inevitable, sooner or later. Not very many are likely to survive, being handicapped by whatever assignment we have bred into them, but a handful might. If there is no chance of preventing these releases with the top-down model of regulation, the only alternative is to start developing the science needed to forge biotech into a tool for bottom-up biological control -- so that we know how to send genetically modified organisms out after other GMOs.

The great advantage of this research is that it could be readily adapted to fight in a much more critical battle now being waged in every region of the globe against species invading from other ecologies. The list of real disasters spreading from these invasions, of real habitats destroyed, real jobs lost, real species driven over the brink, makes the worst speculations about the ecological effects of biotech gone wrong pale by comparison. Conventional defenses against these invasions have had hardly any effect, and the few victories that have been posted are likely to be reversed at any time.

The only defense that can be guaranteed to work, though at present only in theory, is engineering microbial organisms custom- built to infect and debilitate specific aliens. Yet so far as I know, no person of stature in the biotech sector has called for such a program, either as a tool for resisting the spread of exotic species or a defense against the concern of accidental release. Instead the sector has contented itself with throwing sand on the studies that seem to show that releases might have happened, or could happen.

This defensiveness and general lack of leadership might be traced to the assumptions of the debate, which is that biotech can be controlled. The fear seems to be that if argument goes against it, then that control might be used to throttle the industry back and even shut it down, the way the nuclear power industry was. To argue for the design and release of ecological antibodies is to move the discussion into what seems to be a forbidden zone: the admission that this technology can not be stopped.

Unfortunately we live in this zone. This is our address. If both sides to this debate could rise above their reluctance to seem impotent, we could probably do a pretty good job of cleaning up after the technology if we needed to. To do that, we are going to need to start thinking about building an ecological immune system, preferably sooner rather than later.