DEPARTMENT OF HEALTH & HUMAN SERVICES 
Public Health Service 
February 1, 1982 
National Institutes of Health 
Bethesda, Maryland 20205 
Bldg. 2, Rm. 322 
Dr. William Gartland, Director 
Office of Recombinant DNA Activities 
Building 31, Room 4A52 
National Institutes of Health 
Bethesda, Maryland 20205 
Dear Dr. Gartland: 
I want to comment on the two alternative proposed revisions of the NIH 
Recombinant DNA Guidelines that are now being considered. While decreased 
NIH supervision of large classes of non-hazardous experiments Is certainly 
desirable, the RAC proposal goes rather too far. In my view. In the direction 
of removing controls. It would seem desirable to maintain some NIH over- 
sight of experiments Involving a degree of risk, as In Dr. Gottesman's 
suggested revision. My reasons are the following: 
1. Although research to date has produced no evidence of hazard In 
recombinant DNA research, the class of experiments done so far has been 
quite limited. We know that random cloning of genomic segments Is almost 
certainly safe, but we do not have the same assurance about deliberate 
attempts to make some types of animal virus recombinants or to clone toxin 
genes, for Instance. Although the risks of such studies may be no worse 
than those which scientists In other areas regularly assess for themselves, 
the recombinant DNA field has a history that one cannot Ignore. An accident 
In this area, especially against a background of recently dissolved Guide- 
lines, could have social and political consequences of an entirely different 
scale than an ordinary laboratory accident. 
2. If one accepts that there are still some risks, it woud seem wise 
to maintain a uniform nationwide mechanism for approval. Throwing all 
decisions back on the Individual university or Investigator would mean that 
the RAC's expertise In these matters would no longer be available, and many 
of Its past deliberations would now have to be repeated with considerable 
effort at the local level, very likely with quite disparate decisions 
resulting. Since these are no primarily local Issues, some uniformity Is 
desirable. One would not want to encourage "forum-shopplng” among 
recombinant DNA researchers. 
[ 748 ] 
