6 
A. M. Skalka 
itself. DP5 0supF similarly only became implanted when it had lost its bile 
sensitivity . 
Testing of host-vector systems in mice and humans (Stuart Levy and Bonnie 
Marshall, Tufts University) 
Two strains of E. coli were tested yl? 76 and y2236 (the latter x!776 plus the 
plasmid pBR322) . The mouse results were similar to those of the Freter study, 
in that no colonization was observed, even x2236 in germ-free mice. No survival 
of y2236 was detected after 24 hours in either germ-free or conventional mice. 
The human study involved groups of four individuals «?ho were kept free of other 
human contacts for the ten days of the experiment, in a room previously prepared 
for bone marrow transplants. In addition to assay of stool samples, hand, 
throat and nose swabs were tested for the presence of the microorganism previously 
ingested in milk (around 10^ cells) . No recovery of the test organisms was 
detected from skin, and there was recovery from throat and nose swabs only the 
first day after ingestion. In the case where three of the individuals were 
given yl776 but not the fourth, no transfer was observed to the fourth. After 4 
days y2236 survived only to the extent of 5 x 10 in its passage through the 
digestive tract, and x!776 somewhat less. No recoveries were detected after 4-5 
days. It was not clear whether the plasmid pBR322 really slightly increased the 
survival of x!776 in the human gut (nor was it certain that the plasmid-contain- 
ing x 2236 had no properties different from those of xl776 in addition to the 
presence of the plasmid) . Two hundred isolated colonies obtained from stools 4 
days after ingestion were tested by the contractor and also colonies obtained 
from stools 4 days after ingestion were tested by the contractor and also 
colonies were sent to Roy Clowes and John Donch for further checking that the 
cells indeed had the characteristics of the test organisms. The presumed y2236 
isolates contained pBR322 and no other plasmid by the criterion of restriction 
analysis, and in no case was there detectable mobilizability of the plasmid. In 
all other respects (antibiotic resistances, nutritional requirements) the isolates 
had the expected properties. 
Testing of host-vector systems in situations simulating accidental spills 
(Melvin Hatch and Mark Chatigny, Naval Biosciences Laboratory) 
Survival of ylllS, DP50 supF , wild type x!666, an E. coli strain 980 isolated 
from an infant with diarrhea and the EK2 Charon 4A phage was tested at various 
relative humidities and temperatures on various types of laboratory surfaces, 
and also in aerosols. The latter measurements were made by obtaining the ratio 
of survival of the test organisms to that of ]3. subtilis spores resistant to 
aerosolization. As expected, xl?76 survives better under conditions not permit- 
ting any attempts at growth in the absence of diaminopimelic (dap) acid and 
thymine or thymidine (which could result in dap-less or thymineless death) . The 
aerosol survival of xl?76 is only slightly lower than that of strain 980 at 50% 
relative humidity and 22°C, but much lower than that of xl666. The two EK2 
hosts, x! 776 and DP5 0supF , survived considerably less well on laboratory surfaces 
than did the wild type x!666. Charon 4A phages were even more readily killed 
under comparable conditions. 
Conclusions of the Phage and Plasmid Subcommittee 
Concerning the Contractors' leports 
1) The data show no gross discrepancy between natural environmental decay of 
the EK2 microorganisms and the in vitro testing performed earlier (in which it 
was required that "no more than one in 10® host cells should be able to perpetu- 
ate a cloned DNA fragment under specified nonpermissive laboratory conditions") . 
[ 294 ] 
