A 
Science for the People 
897 MAIN STREET CAMBRIDGE, MASS. 02139 TEL.: (617) 547-0370 
September 25, 1978 
Dr. Donald Frederickson 
Director, NIH 
Bethesda, Maryland 20014 
Comments on the proposed revised guidelines for 
recombinant DNA research. 
There are three sections to these comments. First, specific 
safety procedures which should be instituted — including a require- 
ment that workers on antibiotic therapy not be allowed into 
recombinant DNA laboratories — are described. Second, the scientific 
justification of the guidelines are discussed. The document lacks 
a clear statement of the hazards and it makes changes in experi- 
mental containment that have not been based on adequate discussion 
among scientists and other interested parties. Last of all, there 
are some changes that should be made in the section of the 
guidelines on "Roles and Responsibilities" . 
The effects of antibiotics in promoting the spread of plasmids 
and bacteria carrying genes for antibiotic resistance is well 
known. The guidelines should explicitly prohibit those who 
are receiving antibiotic therapy from working in recombinant DNA 
laboratories, until seven days after the end of treatment. This 
concern was expressed in the original (june 1976) guidelines, and 
received scientific support from two participants at the Falmouth 
conference (1,2). The alteration of normal gut flora by antibiotic 
treatment might also contribute to the spread of chimeric plasmids. 
The Scientific Advisory Board of the EPA recommended strongly 
that all bacteria used in recombinant DNA experiments be tagged (3) 
so that their spread, after a spill, could be detected in the 
environment. Although this is not easy to carry out in practice, 
the possibility should be considered by the NIH. Spill or no spill, 
however, an important preventive measure is needed to guard against 
one possible first step in the escape of a recombinant DNA vector. 
Cultures of the HVl host E. coli K12 (as an example) may be 
contaminated with wild-type or other vigorous strains of E. coli 
being studied in the same or neighboring labs. E. coli containing 
a transferable R-factor can mobilize non-transf erable plasmids from 
E. coli K12 in vitro (4) . This is more likely than transfer in 
the intestine, for example. Because of this possibility of 
in vitro contamination and transfer, testing of the purity of 
HVl and HV2 host-vector systems should be a required part of 
recombinant DNA work. Although methods have been described for 
testing the purity of E. coli K12 and other certified bacterial 
strains for the research their use should be required in the guidelines' 
section on biological containment. 
[A-370] 
