23 
counterparts. Though this approach is sinipie and attractive in principle^ 
this step has many pitfalls and unknowns, and these shall certainly have to 
be examined very carefully before any such therapy could be considered. 
I have spoken about the benefits briefly. There are many others, and 
we could spend hours discussing them. I want to turn now to the question 
of risks. 
From the beginning, some people expressed reservations about the risks 
attending the construction and propagation of hybrid or recombinant DNA 
molecules. In fact, I had been the target of that concern, nearly 4 years 
ago, when we had contemplated putting tumor virus genes into J£. coli. But 
a more concerted action occurred when scientists attending the 1973 Gordon 
Research Conference on Nucleic Acids requested that the National Academy of 
Sciences and the Institute of Medicine consider the question of whether 
this research created potential hazards to man and his environment. 
Subsequently, at the request of the Academy, I convened a group of 
about a dozen scientists to meet during April of 1974 at the Massachusetts 
Institute of Technology to consider this question. 
Despite our own strong commitment to this work, we felt it necessary 
to call attention to the potential — and here I would like to stress the 
word potential, because none of the risks that I have talked about can be 
documented as having occurred or being very likely to occur. They are 
potential risks. 
Nevertheless, we felt it necessary to call attention to these poten- 
tial risks of several lines of work with biochemically constructed hybrid 
DNAs. We suggested that scientists working in this field throughout the 
world should join us, the signators to the letter or report which was pub- 
lished in Science and Nature and the Proceedings of the National Academy 
of Sciences , should join us in deferring, and I stress the word deferring, 
voluntarily, at least two kinds of experiments. This pause was to be a 
temporary measure to secure time until an international meeting of consul- 
tants with greater expertise and broader concerns could evaluate the nature 
and the magnitude of the potential risks. 
Though each individual assesses the risks involved with the construc- 
tion and propagation of recombinant DNAs differently, there are few, if 
any, that I know of who believe that this methodology is free of any risks. 
What then, are the possible risks? 
First, there is the possibility of genetically altering bacteria so 
that they have a greater capacity to cause disease or to resist treatment. 
Genetic determinants promoting pathogenicity are known to occur on bacte- 
rial plasmids that are being used to introduce and transfer genes into bac- 
teria. Particularly risky is the possibility that certain infections 
[164] 
n 
i 
i 
