133 
viruses whose optimal temperature for growth, the optimal functioning of 
their enzymes is at lower temperatures, and almost without exception they 
will not multiply at the body temperature of mammals. They are adapted to 
more like room temperature growth — their whole evolution is. So we feel 
that viruses of cold-blooded mammals can be put into prokaryotic systems at 
a one lower level of containment than the comparable virus from warm-blooded 
mamma 1 s . 
The only invertebrate virus, or virus of invertebrates, that even is 
specially handled are these baculovirus genomes that are being distributed 
into nature and have such a restricted host range in the first place. They 
have a lower containment. Containment for plant viruses, which would be 
reverse transcripts of RNA viruses, have not been changed from the previous 
Guidelines. That is, for the inserts into prokaryotes. 
Now, in the use of viruses as vectors we have changed — that is, animal 
viruses as vectors for infecting, carrying DNA segments into eukaryotic 
cells in tissue culture, we have made some changes, primarily in introducing 
the possibility of using adenovirus types 2 and 5 as vectors, since they 
meet the criteria that were set up by the working group that made the Guide- 
lines, the first issue of the Guidelines. Adeno 2 and 3 are so extensively 
studied now they can be worked with as restriction fragments. Pieces can be 
deleted from them by restriction enzyme technology. One can create abso- 
lutely defective viruses that can have DNA inserted into them but cannot rep- 
licate because they are missing many essential genes. They can only replicate 
if there is a normal virus in the same cell. 
We have specified containment levels for polyoma virus, SV40 virus, 
and adenovirus types 2 and 5. I should mention that types 2 and 5 adeno- 
virus are viruses of man, but as we specify, they cannot be used unless 
they are totally unable to infect — that is, the genome that is carrying the 
recombinant DNA. Also, almost everyone in the general population, as Harry 
Ginsberg and I showed many years ago, almost everybody has antibodies and 
is thoroughly protected against these viruses because they have them in 
childhood, and the chance of picking them up in adult laboratory workers is 
very minimal. Also, laboratory workers can be screened for antibody, and 
only people with antibody can work with them. 
We have not made any major changes in that section. It is mostly rede- 
fining some of the criteria on defectiveness. There is one recommendation 
that I have added that is in the supplemental comments, and that is that the 
mouse adenovirus, an exactly analogous virus that we isolated from mice some 
years ago which does not replicate in human cells, has now been developed 
technologically by Dan Nathans well enough that it is a candidate. Very 
shortly it will be a candidate vector, and it will have great advantages over 
the human adenoviruses in that it does not enter into human cells, it does 
not grow there. So it would remove any concern at all about infecting 
humans, and yet should serve the purpose. 
In summary, I think that these are very cautious recommendations on 
the use of viral genomes. They are very inhibitory for viral research. 
[ 337 ] 
