Page 4 
Coalition 
infective) in intestines. 
RECOMMENDATION : Experiments involving mammalian transforming viruses 
should be performed at P4-EK2 containment levels. 
C. Shotgun Clontng ofi VNA (p. 33077) 
The proposed revised Guidelines distinguish between shotgun cloning 
of DNA from organisms which are known to produce potent polypeptide toxins 
and those which do not. Since many "normal" substances produced by any 
vertebrate (hormones, neuro-transmitters, cell recognition factors) would 
be highly toxic if presented at the wrong time or place to another animal, 
this distinction is meaningless. 
RECOMMENDATION : All vertebrate shotgun experiments should be considered 
to involve producers of polypeptide toxins, and thus be 
performed under P3-EK2 containment. 
V. "Tlzxj.blJLLty" 
The proposed revisions have a provision for "flexibility" that would 
allow, for example, experiments designated at the P3-EK1 level to be per- 
formed at the P2-EK2 level (pp. 33074, 33113). Since distinctions between 
the various P levels, up to P3, are very dependent in practice on training 
and diligence of personnel, it would seem that the validity of the trade-off 
would vary from laboratory to laboratory. Up to P3 it would seem that biolog- 
ical containment would be most reliable, but trade-offs in certain cases 
would probably be reasonable . 
However, P4 represents the closest situation we have to absolute con- 
tainment. For experiments designated at this level to be done at P3 contain- 
ment would represent a seTious falling away from the most stringent conditions 
which have been reserved for those experiments generally acknowledged to be 
in- the highest risk category. 
RECOMMENDATION : No other physical containment level should be substituted 
for the P4 designation without in-depth study by the RAC 
and the NIH Director of the specific experiment proposed. 
[A-328] 
