Director, National Institutes of Health 
September 22, 1978 
Page 2 
3) Page 33077, Section III-A-2-a-(l)-(B)-(2)-b 
An EK-2 host should be used with recombinants of the whole viral genome. 
4) Page 33078, Section III-A-2-a-(2)-a-(l)-(a) 
1) I have the same thoughts in this section as those expressed in item 2 
regarding our ability to determine "non-transformation". 
2) In the (b) section — an EK-2 host should be used with the entire 
transforming gene. 
3) In the (c) section — the EK-1 host alternative should be deleted. 
5) Page 33078, Section III-A-2-a-(2)-(a)-(2)-(b) 
An EK-2 host should be used with entire transforming gene and with whole 
virus genome. 
6) Page 33078, Section III-A-2-a-(2)-(b)- Negative Strand RNA viruses 
No special provision is made for rabies virus. 
In addition, no special provision is made for human encephalitis viruses, 
some of which are RNA(+) strand and others of which are (maybe) RNA(-) strand. 
In closing, I should add that I have read the report of the U.S.-E.M.B.O. 
workshop on pages 33159-33169 but think that our knowledge of these structural and 
replicative aspects are too primitive to be as certain as the authors indicate. 
This opinion is mine and not necessarily that of other members of the campus 
biosafety committee of which I am a member. 
Sincerely yours. 
Warren Levinson, M.D., Ph.D. 
Professor of Microbiology and Immunology 
WLrjh 
cc: Dr. J. Cleaver 
Chair 
UCSF Biosafety Committee 
[A-262] 
