TUFTS UNIVERSITY 
MolecuUr B,ology “ d School of Medicine 
Microbiology 
December 21, 1979 
Dr. Donald S. Fredrickson 
Director 
National Institutes of Health 
Bethesda, Maryland 20205 
Dear Dr. Fredrickson, 
I am writing to express my approval, in principle, of your designated 
change in the classification of experiments using E^. coli K12 host-vector 
systems as described in Section III-O in the revised Guidelines and published 
in the Federal Register, November 30, 1979. Our experiments lend strong 
support to much of the already accumulated data showing that E^. coli K12 
cannot colonize the human intestinal tract. Of critical importance was our 
inability to detect transfer of the plasmid vector pBR322 to any endogenous 
host bacteria, despite the survival of the original E). coli K12 host-vector 
system in the human for up to six days at a peak titer of 10^ - 10^ organisms 
per gram of feces. This finding greatly diminished my concern that, if there 
were chance survival in the environment of an EK1 host-vector system, the 
recombinant vector would be transferred out of its enfeebled host. 
It is in regard to biologic containment of the plasmid vector that I 
do have a specific recommendation. The plasmid we studied was pBR322 which 
is a nonconjugative , poorly mobilizable plasmid. I would suggest that a 
distinction be made in the descriptive portions of the Guidelines between 
"nonconjugative" and the nonconjugative plasmids which are also "poorly 
mobilizable". Some nonconjugative plasmids can be mobilized at a frequency 
as high as 10“^ - 10~3 when assisted by conjugative-proficient plasmids 
present in the same cell. Data on this point have been provided by Dr. Roy 
Clowes, Dallas. These rates of mobilization are as high as transfer rates of 
some conjugation-proficient plasmids. While I realize that mobilization re- 
quires a conjugation-proficient plasmid to enter the cell bearing the noncon- 
jugative plasmid (rate of acceptance being 10~1 - 10~3) > nevertheless there is 
an important additional biologic containment in having a resident plasmid-vecto 
such as pBR322 which is not only nonconjugative, but also mobilizable at very 
low frequency (lO - ^ - 10“^) under the best conditions. We are currently examin 
ing the frequency of transfer of pBR322 to endogenous fecal bacteria iur vivo 
when the plasmid is present in the same cell with two conjugative plasmids. 
These studies should be important in evaluating the potential spread of re- 
combinant vectors in the extreme situation of sharing a cell with two conjuga- 
tion-proficient plasmids. 
h6 Harrison Avenue 
liston, Massachusetts 02111 
1*7 956-6750 
[ 567 ] 
