DEPARTMENT OF MICROBIOLOGY AND MOLECULAR GENETICS 
HARVARD MEDICAL SCHOOL 
28 Shattuck Strut 
BOSTON. MASSACHUSETTS 02115 
October 26, 1978 
Dr. William Gartland 
Recombinant DNA Office 
National Institute of Allergy 
and Infectious Disease 
NIH, Bldg. 31, Rm. 7A04 
Bethesda, Md 20014 
Dear Dr. Gartland, 
I wish to express a concern about a particular class of 
recombinant DNA experiments which may be hazardous and which may not 
have been considered in the drawing up of NIH guidelines. I have 
come to consider this problem as a result of experiments being done 
by a collaborator of ours who is using certain of our strains in 
recombinant DNA experiments. 
The general problem is as follows: several laboratories (Boyer, 
UCSF; Gilbert, Harvard; Hofnung, Institut Pasteur, Taris) have used 
recombinant DNA techniques to construct strains which produce hybrid 
proteins between the peptide hormones insulin and/or somatostatin 
and bacterial proteins. In the case of the San Francisco group, the 
hybrid protein was cy topi a sm i c . In the Harvard studies, the hybrid is 
secreted into the bacterial perip^l^srn. I understand that this group is 
now attempting to isolate mutants of these strains which will secrete 
the hybrid protein into the cultu r e medium . In the case of the Paris 
group, they are using strains we constructed in which 8 -galactosidase, 
ordinarily a cytoplasmic protein, is exported to the exterior of the 
bacterial cell surface . They have fused the gene for somatostatin to 
this exported g -galactosidase in order to put somatostatin on the cell 
surface. 
In all these cases one has a situation in which, in essence, a 
peptide hormone (human or closely related) is produced in modified 
form by the bacteria. The modification is the attachment of the 
hormone to a bacterial protein. This modified form of the hormone may 
also result in a modified antigenic activity of the hormone portion of 
the hybrid protein. I have talked with a number of immunologists in 
the Boston area (Drs. Schur, Unanue and Dessein) , with infectious 
diseases experts (Drs. Gorbach and Fields) and others with knowledge 
in the recombinant DNA area (Drs. Baltimore and Benjamin). In all 
cases, these individuals agreed that there was a real possibility that 
these modified peptide hormones could break tolerance/induce an auto- 
immune response in humans to their own hormones if they reached the 
appropriate sites in the body. This, in turn, could clearly have 
severe effects on the health of the affected individuals. I believe 
that the individuals I have consulted with are not, in general, 
alarmist about recombinant DNA ratters and their concerns should be 
taken seriously. 
[ 241 ] 
