3 
spread in nature unless it can compete with the organisms that have been 
selected in the past for their adaptation to the environment. 
2) The novelty of the proposed recombinants was also exaggerated. We 
know that foreign DNA can enter bacteria, and we can expect that such DNA will 
occasionally be integrated into the host chromosome. Recently Chang and Cohen 
have demonstrated that such integration of mammalian DNA occurs in Eh_ coli 
cells, and at the same sites at which the in vitro experiments would cause 
their integration. This is hardly surprising, since the bacteria contain 
precisely the same enzymes that are used in the work in vitro . It now seems 
virtually impossible to doubt that mammalian DNA has been entering bacteria - - 
to be sure, at a low level of efficiency, but on such a scale in nature that 
the events must be quite frequent. 
3) The spread of DNA in the microbial world, by several known mechanisms 
of genetic exchange, has been underestimated. It is increasingly apparent 
that plasmids and restriction enzymes have a major evolutionary role in the 
production in nature of hybrid organisms, with no sharp species boundaries 
between various bacterial strains. 
A) The most serious source of misunderstanding was the failure to recognize 
the relevance of epidemiological principles for a problem that is fundamentally 
one in epidemiology and not in molecular biology. It is true that for precise 
quantitation of the hazards of specific novel organisms we would need data on 
those organisms. However, for a general assessment of the hazards knowledge of 
the class of organisms is sufficient. E. coli plus 0.1% foreign DNA will still 
be coli in its habitat and in its mechanism of spread. Hence knowledge of 
the mechanism of spread of E. coli and related enteric pathogens, and knowledge of 
the control of that spread by sanitation, is thoroughly pertinent. Even though 
[Appendix A — 168] 
