3 . 
ex perimen ts not t o _ be performed : 
1. Introduction of the DNA from mammals or. any warm-blooded 
animals into E. coj i_. (this applies to total DNA and "purified" 
segments, cf. ’’below). The possibilities of introducing tumor virus 
genomes, enzymes converting bacterial metabolites to small molecule 
hormones (e.g. adrenaline) and many others, make such experiments 
extremely hazardous. 
2. Introduction of DNA coding for resistance to antibiotics 
into any bacteria in which this drug resistance does not naturally 
occur. 
3. The use of currently available animal viruses as vectors. 
4. The introduction into E. col i_ of the whole genome or any "purified" 
segment of a viral DNA which is oncogenic in any species. Also the 
formation of hybrids of total or partial oncogenic virus DNA molecules 
with _E_. _co 1 i vectors, whether or not these hybrid molecules are 
subsequently intentionally cloned in £. col i . 
5. Formation of recombinant DNA molecules (and, clearly, 
subsequent cloning) where any part of the recombinant DMA is derived 
from highly pathogenic organisms in classes 3, 4 and 5 of /'Classification 
of Etioloqic Agents on the Basis of Hazard" (1). ( 
6. formation of recombinant DNA molecules with DNA derived 
from organisms containing genes for the synthesis of toxins of very 
high toxicity (e.g. botulinum, diphtheria toxins). 
7. Formation of recombinant DNA molecules where part of the 
recombinant is from plant pathogens whose virulence or host range 
might likely be increased. 
8. Introduction into E.. col_i of DNA from higher plants which are 
known or suspected to make substances, wi th potent physiological effects, 
e.g. alkaloids, toxins. 
I 
9. Any recombinant DNA molecules or strains carrying them 
mentioned in items 1-8, which have already been made, should no longer 
be used, but rather be digested, autoclaved, etc., 
B. The second item in this overview of our specific proposals involves a 
i 
discussion of the col i K-12 host-vector system and physical containment 
levels. The suggestion of a "disarmed", low-survival, E. col i host made in 
the Asilomar report, carrying possible specific mutations enumerated in a foot- 
note of the Woods Hole Guidelines, deserves particularly careful consideration. 
[ 354 ] 
