14 
paralysis and death apparently by damaging the endothelium of blood 
vessels in the nervous system. Shiga toxin inhibits protein synthesis 
and is, to man, one of the most potent toxic proteins. 
Dr. Gill said the Guidelines that apply to experiments involving Shiga 
toxin should be applied to the Shiga-like toxin which has been discovered 
in Eh coli . Et. Gill said under the current Guidelines human toxicity 
data are to be of paramount importance in setting containment levels. 
If no human toxicity data are available, toxicity data frcm primates 
are next most important. If neither human or other primate toxicity 
data are available, containment levels shall be determined from the 
LD 50 of the most sensitive of three small animal species (mice, guinea 
pigs, and rabbits). The available data show Shiga toxin to be highly 
toxic to primates. These data, however, are a little soft. Data 
generated using rabbits show the toxin to be extremely toxic. In 
contrast, Shiga toxin is not very toxic to mice. He said that Shiga 
toxin has the same toxicity range in the most sensitive animals 
as tetanus toxin and the botulinum toxins. 
Dr. Gill said in vitro gene expression of Shiga toxin can differ by 
factors of 10,000. How this correlates to in situ (in the gut) expres- 
sion is not known. 
Dr. Fa per said, taxonamically, Shigella , and Escherichia are so similar 
that in the future they may be classified with the same name. He added 
that the Shiga toxin gene has undoubtedly been cloned many times in the 
shotgun cloning of the Eh coli genome. 
Dr. Kaper said that studies in human volunteers have shown that invasive- 
ness in the most important virulence determinant: nontoxigenic , invasive 
Shigella strains cause disease; toxigenic, noninvasive Shigella strains 
do not cause disease in humans. 
Dr. King Holmes said certain aspects of the Shigellosis syndrane such 
as disseminated intravascular coagulation, the hemolytic uremic syndrome, 
and central nervous system symptoms may be related to toxins such as 
Shiga toxin. He asked if the E. coli Shiga-like toxin is encoded by a 
chromosomal gene. Upon receiving an affirmative answer, he questional 
whether cloning the gene on a high copy number plasmid might not result 
in increased virulence. For this reason, he argued in favor of contain- 
ment higher than PI. 
Dr. King Holmes said the investigators wish to prepare a ENA probe 
with the Shiga-like toxin gene in order to screen Eh coli isolated 
from patients with diarrhea. Dr. Gottesman suggested a probe need not 
consist of the entire toxin gene sequence; a non- toxic fragment of the 
gene would suffice. The research to generate the probe might be per- 
formed under higher containment; containment for experiments involving 
the probe could then be lowered through a multistep process. Dr. Gill 
endorsed ET. Gottesman' s suggestion of generating a partial or a 
defective gene probe. 
[201 
