Page 23 of Attachment E 
the unanimous conclusion that it would be virtually impossible to 
turn E. coli K-12 into a pathogen by the random insertion of a bit of 
eukaryotic DNA. 
Data presented which related to the epidemiology of E. coli in- 
fections both in the gut and urinary tract, etc., indicated that the 
majority of all E. coli infections originate from the gut. That is, the 
intestine is the most important reservoir of pathogenic E. coli. 
Consequently, the ability to colonize the gut is one of the most im- 
portant parameters to measure for any E. coli strain in assessment of 
risk. This generalization lead to discussion of the usefulness of doing 
further human feeding tests. Dr. E.S. Anderson reported on the 
outline of experiments, that had been proposed to WHO and later 
COGENE, which would involve feeding of a standard set of E. coli and 
E. coli containing plasmids to laboratory worker volunteers in various 
laboratories throuhout the world. It was generally agreed that 
although no one was terribly concerned about the pathogenicity of E. 
coli K-12 or the increase in pathogenicity as a result of a cloning 
experiment such an experiment might help to strengthen the con- 
clusion that E. coli K-12 is a very poor colonizer and would probably 
be so whether or not it contained a plasmid. 
The next topic that was discussed related to plasmids and 
estimates of their transmissibility with a goal of designing a possible 
test system to get further information on this question. Dr. Roy 
Curtiss presented evidence from in vitro testing which gave the 
mobilization frequency for various plasmid components of EK2 
vector systems. The data were taken from tri-parental matings which 
went on for about 60 minutes under conditions where the conjugative 
plasmid was already in the host into which the test plasmid was 
placed. Following is a table of these results: 
Plasmid 
Markers 
M daltons 
Mobilization Frequency 
pCRI 
ColEI, KrrY 
8.6 
10- 1 
PSC101 
Td 
6.0 
10- 4 
pMB9 
ColEI, Tcf 
3.5 1 
10- 7 — 10- 8 
PBP313 
ColEI, Apf,Td 
^6.7 (approx .) ) 
pBR322 
ColEI, Ap.Td 
2.6 
Curtiss emphasized that these tests show the highest measurable 
frequencies. In the gut, the normal environment of these organisms, 
there are many factors, e.g. low pH, presence of bile, the long 
generation time of the cells, etc. all of which could be expected to 
inhibit conjugative transfer. He showed a table which assigned a 
probability factor for all these parameters, and then expressed the 
total probability (expressed as transfer per day) for the standard K-12 
systems. This number comes out to be 10- 16 for pBR322 in wild type 
K-12 systems, and 10- 29 when the host is X 1776. It seems clear from 
these data that transmissibility, at least with nonconjugative 
- 17 - 
[ 402 ] 
