NOTICES 
33125 
systems are retained within the broad- 
er host-vector systems (HV) classifica- 
tion, providing more specificity.] 
Implications of the use of EK2 con- 
tainment are elucidated in the follow- 
ing passage from a paper presented by 
Bernard D. Davis at a forum on recom- 
binant DNA held by the National 
Academy of Sciences, March 7-9, 
1977.(2) 
A very large safety factor is added by the 
provision in the present guidelines for bio- 
logical containment. All work with mamma- 
lian DNA must be carried out in EK2 
strains, which have a drastically impaired 
ability to multiply, or to transfer their plas- 
mid, except under very special conditions 
provided in the laboratory. The presently 
certified EK2 strain has several stable mu- 
tational defects (i.e., deletions) that prevent 
it from multiplying under the nutritional 
conditions of the gut. But the protection 
goes much further, and reaches a degree 
that is unprecedented in the annals of 
man’s exploration of potentially hazardous 
new materials: this material has been coded 
for self-destruction. For example, these 
mutant cells require diaminopimelate, a 
constituent of cell wall; and without it they 
can continue to grow and expand but 
cannot form more wall, and so they quickly 
burst. Accordingly, under conditions similar 
to those in the gut such an EK2 strain not 
only fails to multiply, but less than 1 in 10 8 
cells survives after 24 hours— and it would 
be an extraordinarily sloppy laboratory ac- 
cident that would result in ingestion of as 
many as 10 8 cells. In addition, while the cells 
are dying off in the absence of diaminopi- 
melate they are severely impaired in their 
ability to transfer plasmids to other, well- 
adapted cells— and this is the important 
point for the danger of spreading harmful 
genes. Finally, not only the cells but also 
the plasmids being used to carry recombin- 
ant genes are also weakened mutant deriva- 
ties, selected for severe impairment of their 
ability to be transmitted from the host cell 
to another cell. 
We thus see that, even with a strain 
known to carry the gene for a potent toxin, 
the production of disease in a laboratory 
worker would require the compounding of 
two low probabilities: that the strain will 
initiate an infection and that it will survive 
long enough to cause harm despite its sever- 
al disadvantages— that of being a labora- 
tory-adapted strain, that of carrying the 
burden of foreign DNA, and that of carry- 
ing the very large burden of being a suicidal 
EK2 strain. 
The criteria for NIH certification of 
an EK2 system have been defined and 
enlarged during the past year. Exten- 
sive data are required and very de- 
manding standards have been set. 
Such organisms are being designed 
and constructed by NIH contractors 
and other interested investigators. 
Their use in recombinant DNA experi- 
ments is not allowed until they have 
been certified by the Director, NIH, 
upon recommendation by the Recom- 
binant Advisory Committee. The NIH 
environmental impact statement de- 
scribes the criteria for certification 
and lists the certified EK2 systems as 
of July 1977.(25) It should be noted 
that the same depth of experience 
with K-12 that recommends its utility 
as a host for recombinant DNA experi- 
ments is central to the ability to ma- 
nipulate it for the purpose of improv- 
ing its safety. 
An important recent paper was pub- 
lished by two British workers, Petro- 
cheilou and Richmond on the absence 
of plasmid or E. coli K-12 infection 
among laboratory personnel.(29) In 
testimony before the Subcommittee 
on Science, Technology, and Space of 
the Senate Committee on Commerce, 
Science, and Transportation, on No- 
vember 10, 1977, Dr. Oliver Smithies, 
professor of medical genetics and ge- 
netics at the University of Wisconsin, 
interpreted the Petrocheilou and 
Richmond results as follows: 
Twice weekly for over 2 years these work- 
ers tested the feces of five laboratory per- 
sons who had been using without special 
precautions the laboratory strain of E. coli 
called K-12, together with a transmissible 
plasmid. Neither the E. coli K-12 nor the 
transmissible plasmid was ever found in the 
feces during these tests. (Transmissible plas- 
mids are naturally occurring circular pieces 
of DNA that can replicate inside bacteria 
and which, in nature, transfer genes be- 
tween them.) So, with E. Coli K-12 and a 
transmissible plasmid, the risk of the plas- 
mid or its host K-12 getting into the feces 
and surviving to any appreciable extent is 
less than one per laboratory worker per 10 
years of lab work, even when no special pre- 
cautions are taken. 
Now, under the NIH guidelines, none of 
the even conceivably hazardous experiments 
are performed in this type of E. coli, K-12. 
Such experiments require a specially weak- 
ened strain, Chi 1776, which introduces a 
safety factor for survival of greater than 
100 million. Chi 1776 has been proven by 
tests to survive 100 million times less well 
than K-12. 
In addition, such experiments require the 
use of a nontransmissible plasmid which in- 
troduces a safety factor for transfer of the 
plasmid to other bacteria of about 100 mil- 
lion. 
Let me emphasize again that this type of 
work requires a nontransmissible plasmid: 
that is, a plasmid derived from a transmissi- 
ble plasmid by eliminating the mechanisms 
for transfer of the plasmid between bacte- 
ria. 
So the risk of Chi 1776 strain of E. coli K- 
12 surviving in the feces or of the recombin- 
ant DNA plasmid being transferred to some 
other bacteria becomes less than one chance 
per 100,000 laboratory workers working for 
10,000 years without special physical pre- 
cautions. 
This is what is meant by a “negligible 
risk.” 
When we consider that the guidelines re- 
quire also very special physical precautions, 
you can see why I think the risk is no longer 
worth considering. 
References 
(1) Valenzuela, P., A. Venegas, F. Wein- 
berg, R. Bishop, and W. J. Rutter (1978). 
Structure of Yeast Phenylalanine-tRNA 
Genes: An Intervening DNA Segment Within 
the Region Coding for the tRNA. Proc. Natl. 
Acad. Sci., 75 <1):190-194. 
Tilghman, S. M., D. C. Tiemeier, J. G 
Seidman, B. M. Peterlin, M. Sullivan, J. V 
Maizel, and P. Leder (1978). Intervening Se- 
quence of DNA Identified in the Structural 
Portion of a Mouse fi-Globin Gene. Proc. 
Natl. Acad. Sci., 75 (2X725-729. 
Marx, J. L. (1978). Gene Structure: More 
Surprising Developments. Science, 799:517- 
518. 
Weinstock, R., R. Sweet, M. Weiss, H. 
Cedar, and R. Axel (1978). Intragenic DNA 
Spacers Interrupt the Ovalbumin Gene. 
Proc. Natl. Acad. Sci., 75 (3X1299-1303. 
Tilghman, S. M., P. J. Curtis, D. C. Tie- 
meier, P. Leder, and C. Weissmann (1978). 
The Intervening Sequence of a Mouse 0- 
Globin Gene Is Transcribed Within the 15S 
P-Globin mRNA Precursor. Proc. Natl. 
Acad. Sci., 75 ( 3 ): 1309-13 13. 
Tonegawa, S., A. M. Maxam, R. Tizard, O. 
Bernard, and W. Gilbert (1978). Sequence of 
a Mouse Germ-Line Gene for a Variable 
Region of an Immunoglobulin Light Chain. 
Proc. Natl. Acad. Sci., 75 (3X1485-1489. 
Doolittle, W. F. (1978). Genes in Pieces: 
Were They Ever Together^. Nature, 272:581- 
582. 
Williamson, B. (1978). Split Gene Tran- 
scription. Nature. 272:753. 
Reddy, V. B., B. Thimmappaya, R. Dhar, 
K. N. Subramanian, B. S. Zain, J. Pan, P. K. 
Ghosh, M. L. Celma, and S. M. Weissman 
(1978). The Genome of Simian Virus 40. Sci- 
ence, 200: 494-502. 
Leder, P. (1978). Discontinuous Genes. Ed- 
itorial in N. Eng. J. of Med., 298 (19)1079- 
1081. 
(2) Davis, Bernard D. (1977). Epidemiolog- 
ical and Evolutionary Aspects of Research 
on Recombinant DNA. In Research with Re- 
combinant DNA— An Academy Forum. Na- 
tional Academy of Sciences, Washington, 
D.C., March 7-9, 1977. 
(3) Workshop No. 1: Is It Likely That E. 
coli Can Become a Pathogen ? (Chairpersons: 
Richard Goldstein and Elena O. Nightin- 
gale). Op. cit., ref. 2. 
(4) Smith, H. Williams, and M. A. Ling- 
good (1971). Observations on the Pathogenic 
Properties of the K-88, HLY and ENT Plas- 
mids of Escherichia Coli with Particular 
Reference to Porcine Diarrhoea. J. Med. Mi- 
crobiol. 4:467-485. 
(5) Berg, Paul (1977). Potential Benefits. 
Op. cit., ref. 2. 
(6) Office of the Director, National Insti- 
tutes of Health (1977). Environmental 
Impact Statement on NIH Guidelines for 
Research Involving Recombinant DNA Mol- 
ecules, October 1977. DHEW Publication 
No. (NIH) 1489-90, available from Superin- 
tendent of Documents, U.S. Govt. Printing 
Office, Washington, D.C. 20402 (Stock 
Number 017-040-00413-3, $9.75 per set, 585 
pages) or in GPO depository libraries. 
(7) Sinsheimer, Robert L. (1977). Potential 
Risks. Op. cit., ref. 2. 
(8) Davis, B. D., R. Dulbeccc, H. N. Eisen, 
H. S. Ginsberg, and W. B. Wood (1973). Mi- 
crobiology. Harper and Row, Second Edi- 
tion, New York. 
(9) Falkow, S. (1975). The Ecology of Es- 
cherichia coli. Prepared for the NIH Recom- 
binant DNA Molecule Program Advisory 
Committee; available from NIH Office of 
Recombinant DNA Activities. 
(10) Smith, H. W. (1975). Survival of 
Orally-Administered Escherichia coli K-12 
in the Alimentary Tract of Man. Nature 
255:500-502. 
(11) Anderson, E. S. (1975). Viability of, 
and Transfer of a Plasmid from Escherichia 
FEDERAL REGISTER, VOL. 43, NO. 146— FRIDAY, JULY 28, 1978 
[ 86 ] 
