Department of Molecular Biology and 
Microbiology 
TUFTS UNIVERSITY 
School of Medicine 
January 7, 1983 
Director 
Office of Recombinant DNA Activities 
Building 31, Room 4A52 
National Institutes of Health 
Bethesda, Maryland 20205 
Dear Bill: 
I'm responding to your notice about risk assessment specifically to suggest 
that you remove from the proposal the section G that concerns the expression 
of the diphtheria toxin gene (Federal Register 47:55108). 
The feeding to a few guinea pigs of an _E. coli clone carrying the diphtheria 
toxin gene will tell us if there is a high risk of continuing research with 
that clone. That is all. By the nature of the experiment it cannot tell us 
of any infrequent event that might cause the clone or its descendants to be 
dangerous. Either way, the experiment will not generate data that can be 
generally applied to other clones. 
If the clone is created, then of course its safety must be tested, aL least 
to the extent proposed. But we should eliminate the notion that this would be 
a "risk assessment experiment" of general interest. There is very little chance 
that these experiments will provide information "on the handling and absorption 
of polypeptides from the lower gastrointestinal tract and the effect of cyto- 
toxins on the GI tract." Certainly this should not be advanced as a justification 
for cloning this gene. 
I know that Myron Levine and I suggested that the full danger of toxin production 
in the gut could be estimated only if we could get the toxins made by bacteria 
that were resident in the gut but I now feel that it would be absolutely wrong to 
deliberately construct new pathogens in order to gain this information. Risk 
assessments of this nature will have to be done solely by administering purified 
toxins to the gut. 
Sincerely 
DMG : ah 
D. Michael Gill, Ph.D. 
Professor 
136 Harrison Avenue 
Boston, Massachusetts 02111 
617 956-6750 
[ 564 ] 
