2 
germane here as part of the evaluations at 
least may Involve infectivity or penetrations. 
In the instance of a procaryotic vector, such 
a cell already has been "penetrated." 
Further, it Is not inconceivable that 
recombinant DNA-containing organisms sitting 
directly on mucosal cells of the intestine 
could penetrate directly into such cells and 
thus the nucleases in the intestinal contents 
would not be active. Direct penetration does 
not occur in bacterial enteritis. 
o Table 1: I believe Sindbis and Semliki Forest viruses 
are positive-stranded RNA viruses, not negative- 
stranded as the table indicates. 
I am In agreement that segmented genomes 
would probably cause no problems and with the 
recommendation on Baculovi ruses . 
In general, it would appear useful to do 
some experiments in an appropriate containment 
system to determine whether conmunicabil ity is 
possible or probable. 
These experiments should be performed prior to setting up the 
gui del 1 nes . 
Other examples of a similar nature are scattered through the 
document and basically rest on the optimistic feeling that animal or 
plant viral replication will not occur in procaryotic vectors. It is 
probable that this is so but caution dictates the system should be tried 
first in phase 3 or 4 containment, however onerous or relatively expen- 
sive the chore. 
cc: Dr. Armstrong, RD-683 
[Appendix A — 300] 
