independently to the overall containment (i.e., 
either one alone must be demonstrated to pro¬ 
duce 10-8 survival within 24 hours under non- 
permissive conditions). At least one trait must be 
non-revertible, and the reversion rate of the others 
must be shown to be equal or less than 1 x 10-® 
per cell per generation. 
II. Transmission 
Although only non-conjugative plasmids may be 
used in EK2 systems, the flora of the intestinal 
tract include bacteria that harbor conjugative 
plasmids; transfer of the latter to an EK2 cell may 
result in the mobilization of the EK2 plasmid* for 
transfer to a third, recipient bacterium. 
The probability of this occurring will be the pro¬ 
duct of four independent probabilities: 
(1) The probability (P,) that the EK2 cell will 
encounter a bacterium that harbors a mobilizing 
plasmid and is capable of conjugating under the 
prevailing environmental conditions. 
(2) The probability that the mobilizing plasmid 
will be transferred to the EK2 cell. This can be 
measured experimentally under optimal mating 
conditions in vitro, and expressed as the transfer 
rate, R,. 
(3) The probability (P 2 ) that an EK2 cell which 
has received a mobilizing plasmid will encounter a 
competent recipient bacterium, and conjugate 
with it under prevailing environmental conditions. 
(4) The probability that the EK2 plasmid will be 
mobilized and transferred to the recipient. This 
can be measured under optimal mating conditions 
in vitro, and expressed as the transfer rate, R 2 . 
In actual experimentsf on transfer between E. 
coll strains in vivo using host-plasmids for which 
R, and R 2 are both very high (> 10-2), the observed 
transfer frequencies (roughly corresponding to P, 
X R,, and P 2 x R 2 , respectively) were less than 1 x 
10-9, suggesting that P, and P 2 are each less than 
1 X 10-T Since, however, much higher values 
might obtain under other conditions (e.g., in sew¬ 
age, or in animals fed antibiotics), it would seem 
prudent to demand that a significant degree of 
containment be contributed by the factors R, and 
R2. 
A maximal value of 1 x 10-® for the product R, x 
R 2 , for example, would give a figure of less than 1 x 
10-8 for the overall rate of transfer (P, x R, x P 2 x 
R 2 ), even if P, x P 2 were as high as 1 x 1 0-2, rather 
than the observed value of less than 1 x 10-'^, 
In view of these considerations, the investigator 
is required to furnish data from the following types 
of experiments: 
(A) The rate of transfer (R,) of each of several 
mobilizing plasmids (see below) from their respec¬ 
tive donors to the proposed EK2 system must be 
measured under optimal, permissive conditions, 
with 90-minute matings in vitro. 
(B) The rate of transfer (Rj) of the EK2 plasmid 
from the EK2 system harboring each mobilizing 
plasmid must be measured under optimal, permis¬ 
sive conditions with matings of at least three hour 
duration. 
(C) The above tests must be done with one 
mobilizing plasmid from each of the following 
compatibility groups: Fll, N, I, and P. A genetically 
derepressed plasmid should be used in each 
case, if available. 
(D) The product, R, x R 2 , must be equal to or 
less than 1 x 10-6 in the case of derepressed 
mobilizing plasmids, and must be equal to or less 
than 1 X 1 0-8 in the case of repressed plasmids. 
(E) For each mobilizing plasmid, a positive con¬ 
trol must be carried out under permissive condi¬ 
tions to confirm the efficacy of the test. In these 
control experiments, the donor-mobilizing plasmid 
system and the recipient bacterial strain must be 
the same as used in the above experiments. In 
these matings, R, x R 2 (separately measured) 
must be equal to or greater than 1 x 10-3 in the 
case of derepressed plasmids and must be equal 
to or greater than 1 x 10-® for repressed plasmids. 
(F) Matings should be carried out at initial cell 
concentrations of 10® cells/ml and a donor to 
recipient ratio of 1. Survival of the EK2 and control 
host during the three-hour mating period must be 
measured, and the transfer rates must be calcu¬ 
lated in terms of the number of viable EK2 cells at 
the end of the mating period. 
*We will refer to the separate components of the EK2 system as 
the “EK2 plasmid" and the "EK2 cell”. 
tS Falkow, personal communication. 
18 
