COLLEGE OF ENGINEERING 
DEPARTMENT OF HUMANITIES 
525 EAST UNIVERSITY 
ANN ARBOR, MICHIGAN 48109 
(313) 764-1420 
November 18, 1977 
Dr. Donald Fredrickson 
Director 
National Institutes of Health 
Bethesda, Maryland 20014 
Dear Dr. Fredrickson: 
I enclose some brief comments on proposed revisions to the guidelines for 
recombinant DNA research, which would have the effect of lowering containment 
levels. 
1. I see no sound scientific reasons for lowering containment levels at 
this point. Much of the data cited as a reason for relaxation is unpublished, 
and has not received the independent criticism and testing that is necessary 
before the data can be said to have achieved the status of sound scientific fact. 
Discussions at the Falmouth Workshop on assessment of the potential risks 
in this field demonstrated that uncertainty and sharp disagreement continue to 
surround critical aspects of the risks, including (a) the possibility of trans- 
fer of recombinant DNA from "safe" bacterial hosts to other strains of the host 
and other bacterial species; (b) the possibility of triggering autoimmune 
responses in humans and animals; (c) the possibility of conferring a selective 
advantage to host bacteria through routine use of antibiotic resistance genes 
as "markers"; (d) the possibility of altering properties of viruses, such as 
infectivity or host range; (e) the implications of modifying plant and animal 
species. 
For discussions of these areas, I refer you to the letters sent to you 
by Dr. Bruce Levin, Dr. Jonathan King, and Dr. Richard Goldstein, the analysis 
of the new data made by Dr. Burke Zimmerman for the House Subcommittee on 
Health and the Environment, and the transcript of the Falmouth conference, if 
available. 
the use of unpublished scientific opinions that have not been independently 
tested and verified to support relaxation of the guidelines represents an un- 
fortunate departure from accepted canons of scientific procedure. It is 
unthinkable that unverified data would be tolerated as a basis for the exploration 
of space or for the construction of buildings, bridges or tunnels. No more 
should such data be used in this case, where human lives may equally be at risk. 
2. Relaxation of present standards of containment could easily be taken 
as a signal to practitioners of gene splicing that the risks are now minimal. 
Relaxation is also a signal to other countries to drop their standards. The 
caution urged three years ago may well be thrown to the winds. Yet our ignorance 
of the implications of many fundamental aspects of gene splicing is as complete 
now as it was then. 
[Appendix A — 74] 
