Dr. Wallace Rowe 
Page 4 
of a human with such an antigen-adjuvant emulsion seems much more likely 
to induce autoantibody production than ingesting or even inadvertent 
inoculation of antigen-bearing enfeebled E^_ coli . To our knowledge, 
no documented instances of severe autoimmunity have occurred as a result 
of accidents of this type. 
In the event that autoantibody production is initiated, what is the 
likelihood it will lead to immunopathologic consequences? There are 
many instances in which antibodies to self antigens, by themselves, 
are not notably pathogenic. For example, it is not at all clear that 
the anti-insulin antibodies found in humans treated with insulin have 
any consequence, except in some instances of massive insulin resistance 
(12,13). Evaluation of renal lesions of insulin-treated diabetics with 
glomerulonephritis reveals no evidence of insulin-anti- insulin immune 
complexes (5). However, in experimental animals, immunization with 
insulin has produced glomerular lesions (14,15). There are clearly 
instances in which autoantibody, by itself, can lead to disordered 
function. The EAMG induced by transfer of serum is a notable example. 
In general, one would anticipate in most instances that the immunopathology 
due to autoantibody production or to the induction of autoreactive T 
lymphocytes would reverse when the antigenic stimulus was removed. 
However, the autoimmunity induced by syngeneic actylcholine receptors 
might be self sustaining. 
In summary, the association of a human or cross-reactive protein with 
a bacterium, particularly a strain that possessed a potent lipopoly- 
saccharide, might constitute a potential risk if the organism could 
propagate itself in the individual and if individual to individual 
spread could occur. On the other hand, oral ingestion of a small 
number of antigen-bearing bacteria incapable of growing outside the 
laboratory and not competent to transfer their eukaryotic genes to 
wild type E^_ coli should lead to no response at all or to only a small 
and probably self-limited antibody response in the exposed individual. 
In the great majority of cases, this should be of no immunopathologic 
consequence and the development of autoantibodies should be limited to 
the index case. 
We believe that the requirements for enfeeblement to exclude replication 
in a competent host as a precaution against other types of potential 
hazards should also be a very adequate safeguard against any postulated 
untoward immune consequences of recombinant DNA technology. 
[ 273 ] 
