STANFORD UNIVERSITY MEDICAL CENTER 
STANFORD, CALIFORNIA 94305 
DEPARTMENT OF BIOCHEMISTRY Area Code 415 
Stanford University School of Medicine 497-6161 
February 17, 1976 
Dr. Donald Fredrickson 
Director 
National Institutes of Health 
Bethesda, Maryland 20014 
Dear Don, 
Perhaps it would be useful to you if I commented on some of 
the suggestions that were voiced at the recent meeting of your Advisory 
Committee (Feb. 9-10, 1976) that considered the guidelines for research 
on recombinant DNA molecules. 
I. It would, in my opinion, be most unwise to transform the 
guidelines into a legalized set of regulations, as suggested by Peter 
Hutt. I doubt if he appreciates either how fast the data base for the 
guidelines can change from its present impoverished state, or how ap- 
proximate is their present formulation. The guidelines are, by their 
very nature, bound to contain discrepancies. This is true because: 
(1) they attempt to cover the entire range of possible experiments with 
recombinant DNA molecules and hence include all life forms on this planet; 
and (2) they attempt this coverage with a relatively small number of bio- 
logical boundary lines for classifying this vast area of experimentation. 
Under these conditions, it is inevitable that particular experiments in 
which DNA from a given organism is inserted into a given vector will be 
misplaced in the guidelines. Moreover, both the physical and biological 
containment components of the classification will undergo rapid change 
as new techniques and host-vector systems are developed. 
We therefore need flexible administrative mechanisms whereby the 
guidelines can be easily modified and that allow rapid consideration of 
individual experiments that do not fit the classification system. Cast- 
ing the guidelines in the legal format of regulations can only inhibit 
that flexibility and thereby retard both the experimentation and the 
adoption of rational safeguards. 
II. In advising existing or potential principal investigators 
about the guidelines for research with recombinant DNA molecules, the NIH 
should explicitly state: (1) that the guidelines are not hard and fast 
rules, but are instead guides to aid the principal investigator in esti- 
mating the risk associated with a given experiment and in designing ap- 
propriate safeguards for that experiment; and (2) that the NIH urges the 
principal investigator to design alternate safeguards, and to present in- 
formation that in his or her opinion alters the estimate of risk indicated 
in the guidelines. While reserving the right not to fund particular experi- 
ments that it deems too hazardous, the NIH should not get boxed into the 
position where it must defend a particular "cookbook of recipes" defining 
[ 460 ] 
