4 
Dr. Levine said the data demonstrate that factors such as invasiveness and 
adherence must be present in addition to toxin production for an organism to 
cause illness. Dr. O'Brien added that Dr. James Farmer had cured a Shiga 
toxin producing strain of Shigella of the plasmid known to encode invasiveness 
and adhesion factors. The strain still produced Shiga toxin at its normal 
level but did not cause disease. 
Dr. Habig asked if the toxin isolated from EL_ coli by Dr. O'Brien is identical 
to Shiga toxin. Dr. O'Brien said that by the techniques currently applied the 
toxins appear to be identical. Dr. Gill agreed that for the purposes of the 
discussion the toxins should be considered identical. 
Dr. Gill said Dr. O'Brien's data suggest the EL_ coli strains used in industrial 
applications may be producing a potent toxin at a very low level (approximately 
0.004 molecules /E. coli cell if the toxin is Shiga toxin). This should be of 
concern to the Food and Drug Administration and to the pharmaceutical industry. 
Dr. Gill said he had attempted to duplicate Dr. O’Brien's data. He said in his 
hands some portion of the toxic activity from E. coli K-12 is neutralized by 
antiserum specific to Shiga toxin. However, the dose curve for lethality in 
HeLa cells is not linear. Moreover, only part of the lethal activity in the 
dose curve is neutralizable by anti-Shiga toxin antiserum. He wondered what 
E. coli K-12 is producing. He felt the coli strain tested might be pro- 
ducing toxic activity other than Shiga toxin. If it is producing Shiga toxin, 
his results suggest Shiga-like toxin is being produced at very low levels (1 
molecule/100,000 cells). He felt more data are necessary to address these 
questions. 
Dr. Formal felt that nature has already introduced the Shiga toxin gene into 
E. coli K-12. He disagreed with the approach of treating Shiga toxin as botu- 
linum toxin is treated in Appendix F; he felt Shiga toxin should be treated 
as cholera toxin is treated. Dr. Gill disagreed; he pointed out that cholera, 
LT toxin, and ST toxin were placed in Appendix F-III of Appendix F because 
data on the effects of these toxins on the gut were available when Appendix F 
was composed . He pointed out that these known effects are also treatable, and 
cholera toxin itself does not produce neurological effects. Cholera toxin is 
far more toxic enterally than parenteral ly while Shiga toxin is more toxic when 
administered parenterally than enterally. 
Dr. Gill asked if the data on Shiga toxin indicate there would be no systemic 
effect were Shiga toxin introduced into the human gut. Drs. Formal and Levine 
said in some cases of shigellosis, particularly in children, there are some 
general neurological effects such as convulsions, although there are no instances 
of neurotoxicity. Dr. Levine said these generalized convulsions, probably 
due to dehydration, can be treated with drugs. 
Dr. Gottesman suggested the working group recommend that containment conditions 
for cloning the Shiga toxin gene be lowered. This recommendation would be 
based on the observation that while the toxicity of the Shiga toxin molecule 
administered parenterally is high, data strongly suggest that the toxicity of 
the toxin presented enterally is considerably lower. 
[ 413 ] 
