15 
Dr. Randall Holmes said the approach of generating a defective gene, 
without using recombinant DNA technology, is currently not feasible; 
there are no Ek_ coli strains known to lack the genes; there are no 
probes available to screen for the genes; there is no good selection 
method to isolate mutants. He added that generating a probe is only 
one goal of the project. He said characterizing the structural genes 
and the mechanisms involved in regulating expression in E. coli are 
also objectives. 
Dr. Gottesman asked if it would be possible to perform these experiments 
using classical genetic methods. Dr. Randall Holmes replied that seme 
experiments could be performed by classical genetic techniques, others, 
however, can only be performed using recombinant DNA technology. 
Dr. Randall Holmes said additional information has become available since 
the proposal was originally submitted. He said two more high producers 
of Shiga-like toxin have been selected from strains isolated by the 
Centers for Disease Control (CDC). These E. coli strains are associated 
with human diarrheal disease, usually mild. Four independent isolates 
which produce as much Shiga-like toxin as Shigella dysenteriae are now 
available. Ihe role the Shiga-like toxin plays in the pathogenesis 
of enteric disease is unknown. 
Dr. Randall Holmes said approximately 13 Eh_ coli strains have been tested 
to determine if they produce Shiga-like toxin. All produce seme Shiga- 
like toxin. Therefore at this time, the possibility exists that all 
E. coli strains produce Shiga-like toxin. He added that Shiga-like 
toxin activity can be detected in Jk_ coli K-12 lacking detectable 
plasmids; therefore, functional chromosomal copies of these genes must 
be present in £k_ coli K-12. 
Dr. Randall Holmes said evidence suggests that in Eh_ coli infections, 
the toxin, if it has any relevance to pathogenesis, is an enterotoxin 
which functions as a cytotoxin. Injected intravenously the toxin 
would probably be highly toxic to monkeys. No data suggest that it is 
highly toxic by the aerosol route of exposure. The route of admin- 
istration is, therefore, important. Dr. Gill said that in an incident 
in which Shiga toxin was accidentally administered to monkeys' tracheas, 
the monkeys did suffer damage. 
Dr. Maas moved that the experiments be permitted at the P3 + EK1 level 
of containment. The motion was seconded. Er. Gill pointed out that 
the motion would permit Shiga-like toxin to be handled at lower contain- 
ment than diphtheria toxin, yet Shiga-like toxin is probably 100 times 
more potent than diphtheria toxin. 
Dr. Martin asked if it was possible to study regulation of expres- 
sion by in vitro mutagenesis of a fragment probe. Dr. Randall Holmes 
replied that in cases where such an approach was successfully employed, 
the structural gene was known. The structural gene for the Shiga-like 
toxin is not known. Dr. Martin suggested hi vitro complementation 
[ 21 ] 
