9 
virus infection as monitored by the development of hemagglutination-inhibiting 
antibodies. 
These experiments were subsequently extended ( Israel et al . ) to evaluate whether 
the same PY plasmid-containing E. coli x!776 strains were tumorigenic in newborn 
hamsters. In newborn hamsters, inoculation of PY virions induces tumors in 
100% of the animals. In control experiments, purified PY virus DNA as well as 
recanbinant plasmids containing PY ENA cause tumors in 5-19% of the inoculated 
hansters. The tumor igeni city of PY DNA is at least 10^-fold lower than that 
observed for intact PY virus particles. In studies to assess whether bacteria 
could deliver potentially tumorigenic viral DNA to susceptible cells in an 
animal model system, subcutaneous inoculation of very large doses (greater 
than 10 7 organisms/animal) of E. coli K-12 strain x 1776 containing PY-plasmid 
recombinants failed to induce tumors. Thus, the results obtained frcm both 
animal model systems indicate that potentially infectious or tumorigenic recom- 
binant DNA molecules were not transferred out of the x 1776 strain of EL_ coli 
K-12 into mammalian cells. 
To more rigorously evaluate this question, the GI tract of conventional anti- 
biotic-compromised or germ-free mice were colonized with wild-type E. coli 
harboring various PY-plasmid DNAs (Smith et al . ) . Although colonization and 
prolonged excretion of large amounts of EL_ coli could be readily demons t rated, 
no evidence of PY virus infection was detected.] 
Dr. Martin thought it less likely that an infectious RNA virus would be regener- 
ated in E. coli than an infectious ENA virus since EL_ coli host-vector systems 
are not lTkely to possess the machinery needed to produce an FNA particle frcm 
a DNA template. Dr. Grodzicker said the retroviruses evolved to replicate in 
eukaryotic cells. She did not think they would function in prokaryotic cells. 
Dr. Landy asked whether the DNA provirus form of retroviruses would be more 
infectious than the retroviral particle. 
Dr. Mulligan said provirus ENA can transfect eukaryotic cells which do not 
possess the appropriate retrovirus receptor site and are otherwise nonpermis- 
sible. Upon transfection, viral particles can in seme cases be produced. 
The provirus would probably have a stable transfection frequency equivalent to 
the transfection rate of naked DNA which is approximately 10~4. 
Dr. Carol Lax Gronbeck of Genentech, Inc. , asked the working group to consider 
whether workers exposed to a host-vector system expressing retroviral sequences 
could develop an antibody response to products of these sequences. She added 
that no Genentech worker has ever converted to seropositive from product exposure 
She said this is not surprising since very low levels of worker exposure occur 
under the BLl-Large-Scale (LS) containment conditions employed by Genentech, Inc. 
Dr. Rapp said it is unlikely a worker exposed to host-vector systems carrying 
a virus but not injected with the virus itself would develop antiviral antibodies 
He said this brought up the issue, however, of whether exposure to low levels 
[ 448 ] 
