61 
Excerpted from The Recombinant-DNA Debate by Dr. Clifford Grobstein. 
Copyright © 1977 by Scientific American, Inc. All rights reserved. 
BIOLOGICAL CONTAINMENT (FOR E. COLI HOST SYSTEMS ONLY) 
EK 1 
DNA from nonpathogenic prokaryotes that 
naturally exchange genes with E. coli 
Plasmid or bacteriopha g e DNA from host 
cells that natural ly exchan ge g enes with 
E. coli. I It plasmid or bacterio p hag e 
g enome contains harmful g enes or if 
DNA se g ment is less than 99 percent 
pure and characterized, h ig her levels of 
containment are re q uired .) 
DNA from embryonic or germ- line cells of 
cold-blooded vertebrates 
DNA from other cold-blooded animals and 
lower eukaryotes (except insects 
maintained in the laboratory for fewer 
than 10 generations) 
DNA from plants (except plants containing 
known pathogens or producing known 
toxins) 
DNA from low-risk pathogenic prokaryotes 
that naturally exchange genes with E. coli 
Org anelle DNA from nonprimate 
eukar y otes. ( For or g anelle DNA that is 
less than 99 percent p ure h ig her levels ot 
containment are required.) 
DNA from nonpathogenic prokaryotes 
that do not naturally exchange genes 
with E. coli 
DNA from plant viruses 
Plasmid or bacterio phag e DNA from host 
cells that do not naturall y exchang e g enes 
with E. coli. (If there is a risk that 
recombinant will increase patho g enicit y .or 
ecolo gical potential of host , hig her levels 
of containment are reguired .l 
EK 2 
DNA from nonembryonic cold-blooded 
vertebrates 
DNA from moderate-risk pathogenic 
prokaryotes that naturally exchange 
genes with E. coli 
DNA from nonpathogenic prokaryotes that 
do not naturally exchange genes 
with E. coli 
DNA from plant viruses 
Org anelle DNA from primates. ( For 
org anelle DNA that is less than 99 p ercent 
pure hi g her levels of containment 
are required.) 
Plasmid or bacteriopha g e DNA from host 
cells that do not natural ly exchan ge 
g enes with E. col i . (I f there is a risk that 
recombinant will increase patho g enicit y 
or ecolo g ical p otential of hos t, higher 
levels of containment are required. ) 
DNA from embryonic primate-tissue or 
germ-line cells 
DNA from other mammalian cells 
DNA from birds 
DNA from embryonic, nonembryonic or 
germ-line vertebrate cells (if vertebrate 
produces a toxin) 
DNA from moderate-risk pathogenic 
prokaryotes that do not naturally exchange 
genes with E. coli 
DNA from animal viruses ( if cloned DNA 
does not contain harmful genes ) 
•DNA from nonembryonic primate tissue':-: 
• DNA' from animal viruses (if cloned DNA • 
( contains harmful g enes ) 
EK 3 
DNA from nonembryonic primate tissue : 
DNA from animal viruses (i f cloned DNA ' 
contains harmful g enes ) 
“SHOTGUN" EXPERIMENTS USING E. COLI K- 12 OR ITS DERIV- 
ATIVES AS THE HOST CELL AND PLASMIDS, BACTERIOPHAGES 
OR OTHER VIRUSES AS THE CLONING VECTORS 
EXPERIMENTS IN WHICH PURE , CHARACTERIZED "FOREIGN' ' 
GENES CARRIED BY PLASMIDS, BACTERIOPHAGES OR OTHER 
VIRUSES ARE CLONED IN E COLI K- 12 OR ITS DERIVATIVES 
SOME EXAMPLES of tbe physical and biological containment re- 
quirements set forth in tbe NIH guidelines for research involving re- 
combinant-DNA molecules, issued in June, 1976, are given in this ta- 
ble. The guidelines, which replaced the partial moratorium that lim- 
ited such research for the preceding two years, are based on “worst 
case” estimates of the potential risks associated with various classes 
of recombinant-DNA experiments. Certain experiments are banned, 
such as those involving DNA from known high-risk pathogens; oth- 
er experiments, such as those involving DNA from organisms that 
are known to exchange genes with E. coli in nature, require only the 
safeguards of good laboratory practice (physical-containment level 
PI) and tbe use of the standard K - 12 laboratory strain of E. coli (bio- 
logical-containment level EK 1). Between these extremes the NIH 
guidelines prescribe appropriate combinations of increasing physi- 
cal and biological containment for increasing levels of estimated risk. 
(In this table containment increases from upper left to lower right) 
Thus physical-containment levels PI, Pi and P4 correspond respec- 
tively to minimum isolation, moderate isolation and maximum isola- 
tion. Biological-containment level EK2 refers to the use of new “crip- 
pled” strains of K-12 incorporating various genetic defects designed to 
make the cells’ survival outside of laboratory conditions essentially 
impossible. Level EK3 is reserved for an £K2-level host-vector sys- 
tem that has successfully passed additional 6eld-testing. Because of 
the very limited availability of P4 facilities and because no bacterial 
host-vector system has yet been certified by tbe NIH as satisfying 
the EK3 criteria, the recombinant-DNA experiments now in prog- 
ress in tbe U.S. with E. coli host systems are with a few exceptions 
limited to those in the unshaded boxes. Experiments with animal- 
virus host systems (currently only the polyoma and SV40 viruses) 
require either the P3 or the P4 level of physical containment Ex- 
periments with plant-virus host systems have special physical-con- 
tainment requirements that are analogous to the /’1-I«-M system. 
