Dr. Donald S. Fredrickson 
Page Four 
expressed or if they do not change the character of the E. coli host. 
But these assurances are probably unrealistic because each recombinant 
would have to be reviewed in detail. Consequently, the guidelines would 
have to specify what properties could or could not be altered to qualify 
for lower containment; alternatively, some group would be reviewing 
such data to pass on requests. 
6. Your suggestion to require the Recombinant DNA Advisory 
Committee to examine the data and approve or disapprove lower contain- 
ment for cloned segments from shotgun experiments also worries me. 
Because the Committee is not in constant session, it could be a very 
complex business to get an opinion or ruling. An investigator in the 
midst of such an experiment would very likely be held up by a cumber- 
some machinery that would frustrate the entire process. 
I think the guidelines should try to state as clearly as possible 
what level of containment should be used to work with the recombinant 
organisms per se and to distinguish that, if appropriate, from the con- 
tainment requirements for work with the purified plasmid or phage DNA 
from such an organism. It's plausible that work with potentially hazardous 
live organisms should be confined to the containment designated for that 
type of experiment but the purified DNA can be worked on and character- 
ized using lower levels of containment. That would allow people to do the 
cloning in a P4 facility, grow up the culture under secure conditions, 
isolate its DNA and take it back to a lesser containment laboratory. Only 
if the segment of DNA being cloned is "beyond reasonable suspicion" 
should organisms harboring that segment be grown outside of the originally 
required facility. If that was spelled out in the guidelines, there would be 
no need for delaying reviews by Washington. 
7. I believe that cloning cDNA from an RNA tumor virus should 
require as stringent a containment as any DNA tumor virus; after all, 
it is the form of the viral genetic information that would interact with 
mammalian genomes. But your statement implies a ban on such experi- 
ments. It was my impression that experiments of that type were not for- 
bidden but did require the most secure containment levels - P4 + EK2 or 
EK3 . 
8. I would recommend (but how can you require) that mitochondria 
be isolated prior to extraction of its DNA for cloning, generally the way 
it's done to facilitate purification of that DNA. The levels you suggest for 
primate organelle DNA (embryonic and non-embryonic ) are stringent but 
not entirely unreasonable. I assume that non-primate organelle DNA is 
in a lower containment level, which is also reasonable. 
[ 515 ] 
