149 
but no virus coming out. So it is very narrow, and it is almost totally 
restricted to Mus musculus for replication of virus as infectious particles. 
DR. SINSHEIMER: You see what I am getting at is, Are there other 
species which we might endanger by these experiments inadvertently? 
DR. TOOZE: I think that is a very reasonable position, because it 
seems to me the whole business is enormously anthropocentric, and we could 
well argue that when we go out of E. col i we have more reason to be cautious 
than when we stop in it, because the chief target for E . col i will be our- 
selves, but we have little idea what the ecological significance of non-coli 
forms may be. I would agree. And I think when we look at viruses we have 
to look at not just whether they are infectious for man, but what they are 
infectious for. I have the thought that if you released harmful recombinant 
— if you could make such a thing, which is questionable, which devastated 
some plant species, you might do more harm than any other way. 
The point, however, is that if you have a small characterized part of — 
let us take any virus, and you have sequenced the damn thing, and you know 
that it is the 100 base pairs of the coat protein of a virus, which as far 
as you can tell has no significant pathological effect on its natural host 
species, and you are asking someone to keep that thing P3-EK2 because it 
just came from a virus, that I do not believe is a rational guideline. 
DR. SINSHEIMER: I just ha-'e one other technical question. 
DR. FREDRICKSON: Yes, Dr. Sinsheimer. 
DR. SINSHEIMER: This relates to the use of polyoma as a vector. You 
indicated that in any experiment where you infect mice, many of the parti- 
cles will come out with fragments of mouse genome. Are those random? 
DR. TOOZE: That is difficult to say. I am not sure enough effort has 
been put into characterizing it. I don't think it is totally random, but 
on the other hand I don't think it is totally very specific either. Maxine 
Singer is doing the same sort of experiment, I think, with SV40, which has 
the same thing, and as far as I can gather from sequencing, there are no 
clear sequence homologies that were constant so that it was going into one 
site all the time and coming out at one site all the time. On the other 
hand, whether it is totally random I don't know. Mai Martin could probably 
answer better than I. 
DR. MARTIN: Well, it is not clear that it is totally random, but both 
repetitive and nonrepet it ive host sequences are present in polyoma. 
DR. SINSHEIMER: Obviously what I am getting at is, again, the question 
of if you do shotgun experiments even of mouse DNA with polyoma, are you 
creating some novel forms? 
DR. TOOZE: We would propose, with using the viruses as vectors, that 
the most rational way to make a--here we are using animal viruses as vectors 
[ 353 ] 
