Analysis of Arguments Developed in Section III-B 
1. What is the likelihood of E.coli K-12 escape from a PI laboratory? 
The assumption that recombinant DNA work with E.coli K-12 
host-vector systems will always be carried out under PI conditions 
in the absence of required training and of oversight mechanisms 
is discussed in my letter. 
The discussion following this heading implies that microorganisms 
do not "escape" from laboratories in which containment is supposed 
to be practiced. Experience does not support this claim. A recent 
example is the smallpox case at Birmingham University where an 
employee not directly working in the rooms where the smallpox virus 
was stored died of the disease. An earlier incident occurred at 
the London School of Tropical Medicine in 1973 as a result of infection 
of a laboratory worker who in turn transmitted the infection to 
two outside contacts, both of whom died. (Department of Health 
and Soiial Security, Report of the Working Party on the Laboratory 
use of Dangerous Pathogens (London: H.M. Stationary Office, 1975) , 
pTT 
The discussion does not address infection of laboratory personnel. 
There is considerable evidence on this possibility. See, e.g. 
the report of A.G.Wedum, "The Detrick Experience as a guide to the 
probable efficacy of P4 microbiological containment facilities 
for studies on microbial recombinant DNA molecules," (sponsored 
by the National Cancer Institute under contract no. NOl-CO-25423 
with Litton Bionetics, Inc.). This study documents 423 cases of 
infection and three deaths over some 25 years under a variety of 
containment conditions at the Army's biological warfare laboratories 
at Fort Detrick, MD. 
2. What is the probability of E.coli causing an epidemic by person- 
to-person spread? 
3. Is E.coli K-12 pathogenic? 
4. Can E.coli K-12 be made pathogenic by the insertion of recombinant 
DNA? 
The arguments used under questions 2,3, and 4 were used to 
justify revision of the guidelines in 1978. 
5. Does the introduction of eukaryotic shotgun DNA into E.coli 
alter its pathogenicity? 
The data cited, the results of an unpublished experiment of 
Chan, Botstein et al. pertain to a single source of DNA, 
Saccharomyces cerevisiae (yeast) . While these results seem to be 
reassuring, they are quite limited in implication because of the 
use of a single source of DNA. It is unlikely that these experi- 
ments provide any information concerning the effects of prokaryotic 
[ 631 ] 
