10 
Dr. Holmes agreed that research in this area should proceed. He asked how 
much toxin was produced by the noninvasive, toxigenic test strain of Shigella 
fed to human volunteers. Dr. Levine said the test strain was proliferative 
but noninvasive in the gut and excreted as much toxin as the parent strain. 
Dr. Holmes asked Dr. Levine which factors other than toxin production were 
necessary for pathogenicity. Dr. Levine replied that the properties depend 
on the particular organism. In Shigella / a 140 megadalton plasmid which 
codes for invasiveness characteristics is required for pathogenicity. In 
entercpathogenic EL_ ooli / a plasmid which encodes for the ability to attach 
to enterocytes is required for pathogenicity. 
Dr. Holmes then asked if more toxin would be produced if the Shiga-like 
toxin gene were cloned into coli K-12 host-vectors than would be produced 
by natural strains causing hemorrhagic colitis. He wondered whether high 
level production could lead to unanticipated toxicity. Dr. Levine thought 
such an outcome unlikely. Dr. Holmes said he was concerned that a dose 
effect is associated with hemorrhagic colitis. Ihe data presented by 
Dr. O'Brien suggest a concentration of 10 2 50% cytotoxic doses (CD 5 q) per 
milligram of protein in cell lysate is not associated with disease while a 
dose of lCr CDcq per milligram protein in cell lysate is associated with 
hemorrhagic colitis. 
Dr. Clowes said he was surprised that containment had been set at the P4 
level in the first review. Hie proposed experiment would not create ary 
"new" organism. He noted that ooli and Shigella are closely related 
and exchange genetic information. He suggested the experiment be conducted 
at the PI + EK2 containment level. Dr. Levine said it is difficult to work 
with the EK2 host, xl776. He suggested that use of P2 containment and an 
EK1 host with EK2 vectors would provide adequate containment. 
Dr. Gottesman said that she was a member of the Working Group on Toxins 
and had participated in (developing Appendix F at the spring 1981 meetings. 
Hie working group specifically placed the cloning of the Shiga toxin gene 
in Section III-A-1 of the Guidelines which overrides the exemption for 
self-cloning experiments. Each case is to be reviewed separately by the 
RAC to address issues such as the dosage effect of cloning chromosomal 
toxin genes on multicopy plasmids. 
Dr. Gottesman asked several questions concerning Shiga toxin. Is Shiga 
toxin excreted at high levels from EL_ coli or does it remain inside the 
cell? Do the enterocytes and other cells affected by the toxin have 
specific receptors for the toxin? Are there other mechanisms for entry 
into the target cells? What is the enzyme's structure? Is Shiga toxin 
produced by a single gene or is it synthesized from two or more separate 
genes? If the latter is the case, lcwer containment levels could be recom- 
mended for dealing the two genes in two separate host-vector systems than 
for cloning a single gene. Dr. Levine said Shiga toxin is composed of A 
and B subunits. Four to six B subunits cluster around one A subunit in 
the active enzyme. Hie B subunit is believed to possess membrane binding 
[255] 
