Page 24 of Attachment E 
plasmids, does not seem to be a parameter of major concern in 
recombinant DNA research. This conclusion was supported by the 
evidence of studies reported by Dr. S. Levy in which he did tests in 
which both farmyard chickens and the farm family which took care of 
them were exposed to plasmid-containing bacteria resistant to 
tetracycline. It was shown that when the tetracycline resistance 
marker was transferred between animals and from animals to man 
this transfer was not via the plasmid but colonization with bacterium 
itself. It is also noteworthy that the plasmid in this test was a con- 
jugative plasmid. Curtiss also reported that over 160 E. coli strains 
isolated from Nature have now been tested for “ nonsense " sup- 
pressor function and all have been found negative. His tests included 
organisms isolated from the gut, from water, and from bacteremic 
patients in hospitals. Thus, it would appear that amber mutations will 
afford safety. 
Despite the evidence that transmissibility is probably not a matter 
of concern, and that in vitro tests probably give the results of the 
“worse possible case", it was still felt that perhaps some sort of in 
vivo test might be useful. This test would be designed not so much to 
ask whether transfer of conjugative plasmids would occur, but 
whether or not there is some other mechanism of transfer among 
organisms in the gut that we are not yet aware of'. Accordingly, it was 
suggested that the most relevant test of this particular hypothesis 
would be to place a marker on a nonconjugative plasmid and then to 
put this plasmid into various strains of E. coli. After feeding them to 
test animals, (mice or even humans) one would ask whether transfer 
of any sort can be detected. The results of the test would be ex- 
pressed as the frequency of transfer of this nontransferable, non- 
mobilizable plasmid over the frequency of a transfer positive, 
mobilizable positive plasmid. 
Discussion next centered on assessment of the likelihood of 
production of an “ effective ’’ toxic substance by a cloned gene 
product either in the gut (the reservoir of most concern) or in the 
urinary tract. It was noted by Sidney Brenner that perhaps a good deal 
of information relative to this topic is already available. One could, for 
example, obtain a partial assessment by asking whether one could 
detect an immune response to various bacterial proteins which are 
contained inside of the cells or on the surface. Detection of immune 
response would, of course, be considered evidence that the bacterial 
proteins had been “seen" by the host defense mechanism. Examples 
of markers which were suggested were the pili surface proteins, 
beta galactosidase as an internal protein and perhaps something like 
penicillinase which is excreted by some bacteria due to the presence 
of a plasmid. These sorts of tests would be rather straightforward and 
involve only serum testing of a variety of laboratory animals as well as 
human beings. 
Another topic which received a fair amount of discussion con- 
cerned ways to assess the effect of cloned fragments from shotgun 
experiments. A method of testing was proposed by Dr. David Bot- 
- 18 - 
[ 403 ] 
