10 
(5) E. col i ir> the etioJ.ogic agent for 30 - 40% of the 
~T7 
cases of sepsis (often fatal). This i.s also independent of the 
ability of £. Coli to colonize llic gut. 
It has been argued, in comments received by the Director 
of NIH during and after the February, ]976, meeting of the 
Advisory Committee to the Director of NIII, that E. coli , because 
it is ubi.quitous in nature and resides in and infects humans 
in a variety of ways, is an inherently poor choice as a safe 
bacterial host for recombinant DNA research. The fact that it 
is already a human pathogen would argue that modifications which 
could possibly alter its infectivity or virulence should be 
avoided. Yet there is no guarantee that the recombinant DNA's 
which are now being added to E. coli via plasmids or bacteriophage 
will not alter the infectious properitics of E. coli , the human 
pathogen . I consider this to be a much more serious potential 
threat to human health than the possibility of developing a 
strain of E. coli which produces a speci.fic toxic gene product 
in quantity in the human colon. 
The drafters of the Guidelines considered the latter possibility, 
and strict adlierence to the Guidelines should make the possibility 
of E. coli producing a toxin from a foreign donor highly unlikely, 
simply by avoiding the use of DNA donors containing such genes. 
But E. col i as a human pathogen was not considered adequately, 
hence the impact statement underestimates the full spectrum of 
untoward events which could result from genetic modifications of 
JL... Clill. 
1/ Dr. Ilalstod Holman — Testimony before a hearing of the Svila- 
committoe on Health of the Senate Comniitteo on Labor and Public 
Welfare, Sept. 22 , 1976. 
Appendix K — 64 
