134 
There is one piece of information which I think is quite relevant that 
has come up since we prepared these Guidelines, that to me and many others 
is a very impressive feature on the possible safety of these experiments. 
That is the phenomenon of splicing discovered by Phil Sharp and George 
Khoury and some of the Cold Spring Harbor people, and that is that viral 
genomes, adenovirus and SV40, are the ones on which it was discovered, they 
do not just transcribe off linear copies of the RNA which then makes linear 
copies in the protein. There are extensive processes that go on. The RNA 
is transcribed out and then enzymes chop out pieces of the RNA, paste it 
together again so that you have the start and the synthetic sequences hooked 
up together. Now, this happens beautifully in animal cells, but bacteria do 
not have these enzymes as far as it is a control mechanism that bacteria 
don't use. So it is a way that we feel sure these viruses can never be put 
together as whole viruses in a prokaryotic system, because it is a whole 
immensely complex and important enzymatic system that bacteria just don't use 
and don ' t have . 
Thank you. 
DR. FREDRICKSON: Thank you, Dr. Rowe. Would you remain there at 
the podium? 
Dr. Rowe has discussed this very important, almost excruciatingly 
important area of the Guidelines. We will have Dr. Tooze talking about 
some of these aspects again later, but I think maybe the Committee would 
like to ask Dr. Rowe some questions at this time. 
Dr. DeRoos. 
DR. DE ROOS : When you mentioned the use of the antibody testing in 
relation to adenoviruses, this implies to me somewhat specific medical 
monitoring, an approach to laboratory employees. I am wondering, do you 
feel that the Guidelines, later on where they mention medical monitoring, 
do you feel that the Guidelines can be more specific in relation to medical 
monitoring for different groups of experiments? 
DR. ROWE: For the most part not, in that 99 percent of the real 
interest in recombinant DNA research is in E. col i and similar bacterial 
systems, where serological responses are just not good. You don't develop 
an antibody response to K-12 or most other E. colis. E. col is have so many 
antigens, there are hundreds and hundreds of different combinations in the 
organism, but they change as they exchange information. I don't think 
serologic monitoring for E. coli has any possibilities. For the adenovirus 
it is very easy and very clearcut. 
DR. DE ROOS: So this would be one of the few instances where you 
could be specific. 
DR. ROWE: Yes, very clearly. SV40 also, either as a prerequisite 
for working with them or as a way of monitoring lab infections. 
[ 338 ] 
