SECOND REPORT OF THE COGENE WORKING GROUP ON RISK ASSESSMENT 
5 
APPENDIX I 
Report of the Phage and Plasmid Subcommittee on Meeting 
with Recombinant DNA Contractors 
March 19, 1979 in Bethesda 
by Dr. J. Setlow, Chairperson 
Summary of Reports from Contractors 
Testing of host-vector systems in sewage (Bernard Sagik, James Walker and 
Charles Sorber, University of Texas at San Antonio) 
The testing was based on a model system set up in a laboratory, capable of 
taking 60 liters/day of waste. San Antonio waste was brought into the laboratory 
to be put into the system. Data were presented to show that the various parts 
of the model system behaved quantitatively like those of real sewage processing 
systems (with respect to such parameters as detention time, pH, temperature, 
removal of solids, etc). The organisms to be tested (xl776 and DP5 0supF and 
Charon 4A) were added continuously for 48 hours. Samples were taken at intervals 
up to 120 hours from nine parts of the treatment plant. For assays of DP50supF, 
the antibiotics nalidixic acid and trimethoprim were used, which kept the 
indigenous bacterial population from entering appreciably into the counts. 
These plus cycloserine were also used for the assay of xl776. The Charon 4A 
phage was distinguished from the indigenous phage by plating it on XG agar, such 
that the plaques of these phage had blue-green halos. Considerable effort was 
put into verifying that the colonies obtained from the putative xl776 that had 
passed through the treatments were really xl?76. Seventy-five colonies were 
tested for other markers such as phage susceptibility, efficiency of plating in 
the presence of bile salts, and various nutritional markers. 
D?5 0s’jpF to some extent multiplied during passage through the first parts of the 
treatment plant, although it would not be able to compete with wild type E. coli. 
There was little or no multiplication of \1776. The titer of DP5 0supF dropped 
from a high of around 10^/ml down to around 10^ in the last effluent during the 
time of the experiment, and was somewhat lower for the treated sludge. Comparable 
values for xl?76 were about an order of magnitude lower. Charon 4A phage titers 
dropped about three orders of magnitude. 
Testing of host-vector systems in mammals (Rolf Freter, University of Michigan) 
9 
Cells (1-5 x 10 ) were placed in the stomachs of mice, and at various intervals 
the stomach, small and large intestines were assayed for the presence of colony- 
forming units of the test microorganisms. Feces were also assayed. A non- 
absorbable radioactive tracer and spores of thermophilic Bacillus subtilis were 
added at the same time as the cells to monitor the passage of material through 
the digestive tract. Five types of IS. coli were tested: DP5 0supF , xl?76, a 
wild strain derived from mouse intestine, a wild-type human strain C25, and 
another wild type, yl666. The results showed that whereas the wild strains 
could multiply in the gut, especially when passage time was slowed by morphine, 
DP5 0supF and xl?76 were killed in the gut, particularly when passage time was 
increased. The amount of cells recovered in the feces was less than one per 
cent of the original inoculum of \1776 y and U P to 10% of the DP5 0supF . Germ- 
free mice were used in some experiments, which showed that even in such hosts 
X1776 could not become a resident in the intestine unless it was a revertant no 
longer sensitive to bile salts, but even then it did not readily establish 
[ 293 ] 
