Federal Register / 
mice with clinical isolates of wild type 
E. co/i carrying various polyoma- 
plasmid recombinants. In these studies, 
mice were starved for 24 hours and then 
fed bread containing 3 X 10* wild type E. 
coli or E. coli K-12 containing polyoma- 
pBR322 recombinants. Additional 
conventional and antibiotic- 
compromised mice were subcutaneously 
inoculated with 6 X 10 7 wild type E. coli 
or E. coli K-12 containing polyoma- 
pBR322 recombinants. Fecal monitoring 
indicated that the wild type E. coli used 
in these experiments was able to 
survive in the Cl tract of conventional 
mice for approximately 4 weeks. When 
the indigenous flora of the murine GI 
tract was eliminated or greatly reduced, 
colonization with high titers of wild type 
E. coli or E. coli K-12 occurred. High 
titers were found in the Cl tract of germ- 
free mice. Results also indicated that E. 
coli carrying polyoma-pBR322 DNA 
were present in high titers in the Cl tract 
of the mice. 
None of the 140 mice that were fed 
and none of the 60 mice that were 
inoculated with live E. coli (wild type or 
K-12) carrying po!yoma-pBR322 
recombinants developed antibody to 
polyoma. This was true whether the 
polyoma DNA was present in the 
recombinant as a monomer or as a 
dimer in the head-to-tail orientation. 
Absence of infectivity in mice receiving 
E. coli containing polyoma monomer- 
pBR322 DNA was not unexpected since 
it had been previously shown that 
recombinants containing a single copy 
of polyoma DNA could initiate infection 
only if mechanisms to precisely excise 
polyoma DNLA are present in either 
mouse cells or E. co/i, and it has been 
demonstrated that mouse cells and E. 
coli lack the appropriate enzymatic 
mechanisms. In contrast, specific 
excision is not required to generate 
infectious polyoma DNA from polyoma 
dimer-pBR322 recombinants because 
intramolecular recombination will yield 
infectious polyoma DNA. Therefore, 
experiments involving Co. coli carrying 
polyoma dimer-pBR322 recombinants 
are the most relevant in determining the 
extent to which potentially infectious 
viral DNA molecules can be transferred 
out of E. coli and into susceptible 
mammalian cells where intramolecular 
recombination could generate intact 
polyoma genomic DNA which would be 
infectious. The results in these 
experiments demonstrate that prolonged 
exposure to bacteria harboring doses 
equal to or greater than the minimum 
infectious dose of purified polyoma 
dimer recombinants is insufficient to 
cause a polyoma infection. 
Vol. 47, No. 235 / Tuesday, December 7, 1982 / Notices 55105 
The investigators speculate that the 
absence of polyoma infectivity following 
prolonged exposure to cumulative doses 
of polyoma dimer-pBR322 recombinants 
that were significantly greater than the 
minimum infectious dose for polyoma 
dimer recombinants could be due to two 
possibilities. The population of E. coli 
harboring polyoma dimer-pBR322 is in a 
dynamic state in the murine gut and 
large quantities of prokaryotic DNA, 
including the polyoma dimer-pBR322 
DNA, are being released into the gut 
lumen daily. Lack of infectivity due to 
free polyoma dimer-pBR322 recombinant 
DNA in the gut may be due to 
degradation of the DNA by nucleases in 
the GI tract. However, this may not be 
the whole answer. The investigators 
speculate that some of the polyoma 
dimer-pBR322 DNA may survive 
nuclease degradation in the gut and 
penetrate the mucosal epithelial cells 
lining the GI tract. They state that 
whether or not the mucosal epithelial 
cells are susceptible to polyoma 
infection is unknown, but the rapid 
shedding of these cells most likely 
precludes subsequent infection and/or 
penetration of the germinal epithelium 
where a productive infection would be 
more probable. The investigators 
conclude that, even in the “worst-case” 
scenario, it is safer working with E. coli 
containing recombinant viral DNA then 
working with intact virus particles. 
These studies have been submitted to a 
scientific journal for publication. 
B. Transmission of Vectors From E. Coli 
K-12 to Other Bacteria in Vivo 
Progress has been made on the 
evaluation of the transmission of vectors 
from E. coli K-12 to other bacteris in the 
human gastrointestinal tract. A Working 
Group had been convened on August 30, 
1979 at the NIH to review Protocols I 
and n of the Workshop on Risk 
Assessment of Recombinant DNA 
Experimentation with Escherichia co/i 
K-12 (J. Infect. Dis. 137, 704-708, 1978). 
The Working Group recommended that 
research should be supported aimed at 
gaining a better basic scientific 
understanding of bacterial colonization 
and plasmid mobilization. This is 
because although E. coli K-12, the strain 
most commonly used in recombinant 
DNA work, is debilitated to the extent 
that survival in the gut and the 
environment is minimal, there has been 
some concern that the plasmid vector 
carrying the foreign DNA could be 
transferred to another, hardier strain of 
E. coli. Members of the Working Group 
felt that such studies should be 
performed directly in humans and 
employ wild type E. coli other than 
strain K-12. The Working Group noted 
that if E. coli K-12 were used as the 
bacterial host in clinical studies very 
large numbers of volunteers (hundreds) 
would be required to assess the in vivo 
mobilization of "safe" plasmids. In 
contrast, the Working Group concluded 
that clear answers on the in vivo 
mobilizabllity of "safe” plasmids might 
bo obtained using much smaller 
numbers of subjects if an E. coli strain 
that readily colonizes the human colon 
were used in the studies. The greatly 
increased number of the colonizing 
strain in contrast to E. coli K-12 would 
provide much greater opportunity for 
transfer resulting in much greater 
sensitivity. E. coli strain HS containing 
poorly mobilizable plasmid pBR325 was 
suggested as a good initial combination. 
These studies were carried out under 
a NIAID contract (NOl AI 12GG6) at the 
Center for Vaccine Development, 
University of Maryland School of 
Medicine, Baltimore, Maryland, by Dr. 
Myron Levine, Dr. James Kaper, Mr. 
Hank Lockman, Dr. Robert Black, and 
Dr. Mary Clements, in collaboration 
with Dr. Stanley Fafkow of the 
Department of Microbiology, University 
of Washington School of Medicine, 
Seattle, Washington. (Present address: 
Department of Microbiology, Stanford 
University School of Medicine, Palo 
Alto, California.) Strain HS was chosen 
to serve as the host organism to carry 
plasmids to be tested for mobilizability. 
E. coli strain HS is a smooth, non- 
pathogenic strain which can 
successfully colonize the human 
intestine for days to weeks. 
Several separate E. co/i HS and E. co/i 
K-12 strains were constructed by 
insertion of one or two plasmids. One 
combination consisted of E. co/i HS 
containing plasmid p[BK5 which is 
easily mobilized. One strain consisted of 
E. co/i HS with poorly-mobilizable 
plasmid pBR325. Another strain 
consisted of E coli HS containing 
poorly-mobilizable plasmid pBR325 and 
conjugative plasmid F-amp which 
ordinarily mobilizes non-conjugative 
plasmids at a high rate of frequency. The 
fourth strain consisted of E coli K- 12 
containing poorly mobilizable plasmid 
pBR325 and the conjugative plasmid F- 
amp. (Non-conjugative plasmids lack the 
genetic information for transfer to 
recipient cells. Transfer functions, 
however, are found on conjugative 
plasmids which can mobilize non- 
conjugative plasmids. Biological 
containment for recombinant DNA 
experiments can be provided by the use 
of poorly mobilizable plasmids which 
are not only non-conjugative. but can be 
mobilized at frequencies of ten thousand 
