6 
Hr. Kopecko said it was important to ctone the Shiga gene from Shigella 
dysentenae 1 in order to understand how toxin gene expression is controlled 
since Shigella dysentenae 1, an epidemic strain, produces more toxin than 
other hactena. He thought 10^ CDs per mg protein is equivalent to the highest 
level of Shiga toxin acpression found m nature; therefore, the working group 
should rec amend EL coll K-12 host- vector systems expressing 10^ CDs per mg 
protein he permitted to be moved from RL3 to lower containment . K-12 strains 
producing more than 10^ CDs per mg protein will not be permitted to leave 
RL3 containment under the proposed language. 
Dr. Gottesman said the issue is whether EL_ coll K-12 producing high levels of 
Shiga toxin would be hazardous to animals. 
Dr. Habig said he did not see why EL_ coll K-12 host-vector systems carrying 
the Shiga toxin gene would pose a greater danger than Shigella dysentenae 1 . 
Dr. Kopecko pointed out that clinical laboratories routinely work with Shigella 
dysentenae 1 at RL2 conditions. He did not think Eh coll K-12 host-vector 
systems would be more dangerous than Shigella dysentenae 1. Dr. Gottesman 
said she felt comfortable with permitting Eh coll host-vector systems expressing 
10 R CDs of SLT to be used at BL2 containment since Eh_ coll K-12 does not 
colonize the g ut . 
Dr. Levine suggested a number such as 10^ CDs be chosen as the upper limit of 
toxin a host- vector system might express if it is to be removed from BL3 to a 
lcwer containment level . 
Dr. Levine said Fh_ coll K-12 is not an invasive pathogenic organism. Several 
factors are necessary for pathogenicity, and even for Shigella more is required 
than simply expression of the toxin gene. An attenuated Shiga bacillus vaccine 
strain which produces as much Shiga toxin as the parent pathogenic strain but 
lacks invasiveness factors did not cause disease when fed to human volunteers. 
In a the single case in which the strain reverted to invasive, the human volun- 
teer became ll 1 . 
Dr. Formal said he had constructed a strain by introducing half of the Shigella 
flexnen chromosome and an SLT converting phage into EL coll K-12. The intro- 
duced Shigella flexnen genes were thought to include all of the genes necessary 
for virulence. The constructed strain had no effect on monkeys in feeding 
experiments although it produced 10 ^ CDs per mg protein of SLT under cptimal 
laboratory conditions. 
Dr. Gill asked whether concern exists about transfer to other organisms of a 
high copy number plasmid carrying the Shiga toxin gene and its control elements. 
Dr. Hewlarri said the probability of transfer of the SLT or Shiga toxin gene by 
phage in nature is higher than the probability of transfer from K-12 by poorly 
rrobilizable plasmid vectors. 
Dr. Gottesman asked what laboratory conditions are necessary to obtain cptimal 
expression of the SLT gene in the E. coll K-12 host-vector system. Dr. Newlard 
replied that antibiotic pressure is necessary. In addition, several other 
procedures such as using specially treated medium are also required. If an 
[ 310 ] 
