SCOPE 44 - Introduction of Genetically Modified Organisms into the Environment

I discuss here some general and historical aspects of recombinant DNA concerns and my own involvement with them. This may seem inappropriate to this volume, either because it rehashes ostensibly dead issues or because it includes political and regulatory, as well as scientific, consideration. However, I believe the history is important to understanding our present outlook. Labelling an issue 'dead' implies that it once was alive, whereas much of the recombinant DNA debate was based on notions that were never scientifically viable. The role of competition in limiting the spread of engineered organisms, for example, was scarcely mentioned at the 1975 Asilomar Conference on Recombinant DNA, although most of us would agree that it is central, based on concepts that were in place long before 1975. I have also seen enough discussions of the underlying science to realize how often they can be influenced by some political second guessing of the probable impact; so I would rather put that aspect on the table rather than under it.

2.1 RECOMBINANT DNA CONCERNS: HISTORICAL BACKGROUND

Each of our perspectives on this subject is inevitably conditioned by personal experiences. Briefly, my own participation has been as follows: I attended the Asilomar Conference on Recombinant DNA and chaired the first Stanford University institutional Recombinant DNA Committee the following year. I was a member of the NIH Recombinant Advisory Committee (RAC) from 1977 to 1981 and participated in the COGENE Meeting on Recombinant DNA and Genetic Experimentation held at Wye in 1979. Since 1981 my involvement has been slight, including service on one advisory committee to the Environmental Protection Agency and current membership on the Biosafety Committee of a local genetic engineering company.

When I joined RAC, the NIH Guidelines were already in place. As time progressed, I became increasingly conscious that the Guidelines were arbitrary, based more on public relations than on serious science, responsive primarily to fear rather than to danger. The authors of the Guidelines themselves seemed unable to remember or to agree on why or how they had arrived at a particular classification level for a given experiment. Most of my energy was spent working for relaxations of the Guidelines, trying to narrow their focus to any areas where perceptions of danger might have a rational basis. One of my last acts as an RAC member was to sponsor, along with David Baltimore, a proposal that would have removed all but a small fraction of recombinant DNA experiments from RAC jurisdiction.

After I left RAC, our motion came to a vote and was defeated. I was unhappy with the outcome, not so much because my side had lost (as it always had done at times) but because of the reasons for the vote. I read the minutes of that RAC meeting carefully and was dismayed by the virtual irrelevance of most of the discussion to the purposes for which I thought RAC had been created. Scientific and lay members alike ignored the question of `Will this make the world a safer place to live?' in favor of `How will this play in Peoria?' The most influential opposition to our motion came not from traditional advocates of genetic regulation but from industrial and academic sources, such as the chairman of the Harvard Biosafety Committee. Their concern was that, if RAC withdrew regulatory supervision of most recombinant DNA research, political pressures for regulation at other levels would mount. From that time onward RAC became, in my opinion, a dummy committee with a sham function. Ostensibly it protects the public from danger, but in fact it serves more to protect Harvard from the Cambridge City Council and Genentech from the California State Assembly.

2.2 CURRENT CONCERNS

How does all this relate to the issue of deliberate release of engineered microorganisms? Does my position on the NIH Guidelines imply that I favor complete laissez-faire? Not at all. Both deliberate release and large-scale production were among the few areas to which Guidelines would have continued to apply under the most liberal revisions ever proposed. Attempts by others to extend exemptions for small-scale laboratory experiments to these other levels have generally rested on devious logic which I cannot endorse.

For example, along with other RAC members, I labored long and ultimately successfully to exempt most cloning in Escherichia coli K12 from the Guidelines. First, K12 is a highly compromised laboratory strain, unable to compete with wild-type bacteria; second, we saw no reason to suspect that a small segment of foreign DNA would make it harmful. Both considerations apply with some force to the possible escape of a few bacteria from a small laboratory culture., If they escape, they are unlikely to spread; and even if they spread, no damage is likely to result. If we compare a culture of K12 produced by shotgun cloning of random samples of insect DNA to a culture containing a mixture of natural bacteria, such as an enrichment culture inoculated with soil or sewage, the former seems if anything safer than the latter. That would certainly be my decision if someone asked me which of the two cultures I would prefer to drink.

On the other hand, if someone were to isolate a single component strain from either culture intending to propagate and disseminate large amounts of it, I would consider it prudent to review its properties carefully for any unexpected problems it might cause. After the K12 exemption, some industrial representatives argued to RAC that, `If something is safe in small amounts, it's safe in large amounts,' a position that, in context, impressed me as a deliberate misrepresentation. Small amounts of cultures of no known or predicted hazard do not warrant elaborate containment measures. That does not imply total absence of risk, and large-scale production is the stage at which more detailed checking is feasible.

2.3 NATURAL MICROORGANISMS COMPARED TO ENGINEERED ONES

As implied above, my concerns about large-scale propagation and dissemination are not restricted to engineered organisms but extend to natural ones as well. I see no rational scientific basis to distinguish between them. However, at least in the United States, regulation of industrial microorganisms has been spotty. Some areas may be overregulated, but others have been underregulated. For example, enzyme-containing detergents were eventually removed from the market-place, but after, rather than before, significant damage had been inflicted on both workers and consumers by these product of natural bacteria. The legal system makes it simpler to impose regulation on new processes rather than established ones, and complete equity may be attainable only in the very long range.

Much has been said and written about the novelty of engineered organisms, including the hypothesis that all imaginable heterologous gene combinations have already occurred in nature and have failed the test of natural selection. Discussion of this question can become confused by imprecision in the level of specification. It cannot be literally true that all conceivable individual gene combinations have been tried. The history of life on earth has been too short even to allow realization of all combinations that could result from reshuffling the human gene pool, let alone the vastly larger number of potential heterologous combinations that might be generated by interspecific transfer (Campbell, 1979; Regal, 1986).

One may of course reason that, since not all combinations have been tried, some particular untested combination might be selectively advantageous and therefore could spread and perhaps cause problems. The logic is unassailable, but in the context of evaluating engineered organisms, I think it misses the main point. The first thing that requires emphasis, in terms of the expectation of Western society in the 1990s is that, like it or not, we already inhabit an inherently uncertain world. A natural genetic lottery operates on all biological populations, including our own, and may generate disruptive new gene combinations at times and places beyond our power to predict. The uncertainties associated with engineered organisms must be measured against this background. The question is not whether individually novel combinations might be created, but whether molecular biologists are sampling a different (and perhaps more dangerous) pool of possibilities than nature samples on a grander scale. I see no reason to think that they do, but I am less certain of that answer than I am of my analysis of what question should be asked.

The expected impact of genetic novelties is closely related to one's position on evolutionary theory. All neo-Darwinian evolutionists see evolution as the joint result of mutation and selection, but only some imagine that the rate of evolution is limited by the rate of fresh (not necessarily unique) mutation (Nei, 1983).

CONCLUSIONS

Public discussion of the safety issues related to engineered organisms has frequently been clouded by decisions of regulatory bodies that were dictated more by politics than by logic. If political considerations are discounted, there are no convincing scientific grounds for distinguishing engineered organisms from natural ones. Because organisms of either type could pose unforeseen hazards, some safety testing is desirable before large-scale propagation.

REFERENCES

Nei, M. (1983) Genetic polymorphism and the role of mutation in evolution. In: Nei, M. and Koehn, R.K. (Eds.) Evolution of Genes and Proteins, Sinauer Associates, Sunderland, Mass, pp. 165-90.

Regal, P.J. (1986) Models of genetically engineered organisms and their biological impact. In: Mooney, H.A. and Drake, J.A. (Eds.) Ecological Studies, Vol. 58, Springer-Verlag, New York, pp. 111-29.