Page 26 of Attachment E 
experiments. At the moment, however, it seems that their ability to 
proceed is being bogged down with bureacratic problems. It is an- 
ticipated that once the results of this particular animal virus ex- 
periment were available, others equally interesting for example, with 
RNA tumor virus genes, other DNA tumor virus genes, etc. might be 
designed to test further implications of cloning animal virus genes. 
An experiment related to the one above, in that it attempts to an- 
swer the reverse question, was proposed by Allan Campbell. The 
attached material contains an outline of the experiments that he 
proposed. Briefly, Campbell's experiments are directed at the 
question of whether genes contained in eukaryotic cells can be 
passed into coli. He suggests a quite rational approach to this which 
would involve use of a gene initially present in E. coli and therefore 
most capable of functioning in E. coli if it could return. One example 
given was the coli gene for galactose utilization. This could be placed 
in an animal virus vector, (polyoma or SV40) and then into an animal 
cell. The test would be to grow an E. coli strain deleted for the gal 
gene in the presence of this animal cell (either in tissue culture or 
after the cell had been put into an animal where it could develop a 
tumor) and ask whether the E. coli mutant could pick up the gal 
marker from the animal cell. This seemed to me an elegant and 
simple first attempt to test for the existence of the hypothetical 
barrier to genetic exchange between eukaryotics and prokaryotes. 
In summary then, a number of specific risk assessment tests were 
proposed at this meeting. These include: 
1. Colonization and transmission test of E. coli and E. coli- 
containing plasmids as summarized by Anderson and 
contained in material already available to COGENE. 
2. Tests of shotgun clones by passage through animals and 
selection of “winners" as outlined by Botstein and Davis. 
3. In vivo (feeding) tests of transmissibility by mechanisms 
other than the known conjugative functions. 
4. The test of animal virus clones such as the Martin and 
Rowe polyoma test and 
5. The reverse of this, that is the Allan Campbell test for flow 
of genetic information from eukaryotes to prokaryotes. 
Also noteworthy was the unanimous agreement that it must be 
considered virtually impossible to convert E. coli K-12 into a 
pathogen of epidemic consequence by insertion of random bits of 
eukaryotic DNA. 
-20 - 
[ 405 ] 
