202 • Impacts of Applied Genetics— Micro-Organisms, Piants, and Animals 
For some, one carefully planned experiment 
using the most sensitive tests is sufficient to 
allay fears. But for others, significant doubt 
about safety remains, regardless of how many 
viruses are examined. The criteria depend on 
an individual’s perception of risk. 
Many experiments carried out for purposes 
other than risk assessment have provided 
evidence that scenarios of doom or catastrophe 
are highly unlikely. This is the general consen- 
sus of specialists, not only in molecular biology, 
but in population genetics, microbiology, infec- 
tious diseases, epidemiology, and public health. 
Experiments have revealed that the structure 
of genes from higher organisms (plants and 
animals) differ from those of bacteria. Con- 
sequently, those genes are unlikely to be ex- 
pressed accidentally by a bacterium; the original 
fears of ‘‘shotgun’’ experiments have become 
less well-founded. Hence, data gathered to date 
have made the accidental construction of a new 
epidemic strain more unlikely. 
Conference discussions have also contributed 
to a better understanding of the risks. At one 
such conference,® which was attended by 45 ex- 
perts in infectious diseases and microbiology, it 
was concluded that: 
• E. coli K-12 (the weakened form of E. coli, 
used in experiments) does not flourish in 
the intestinal tract of man; 
• the type of plasmid permitted by the Guide- 
lines has not been shown to spread from E. 
coli K-12 to other E. coli in the gut; and 
• E. coli K-12 cannot be converted to a harm- 
ful strain even after known virulence fac- 
tors were transferred to it using standard 
genetic techniques. 
A workshop sponsored by NIH^ provided a 
forum for scientists to discuss the risks posed by 
viruses in rDNA experiments. They concluded 
that the risks were probably less when a virus 
was placed inside a bacterium in rDNA form 
“ "Workshop on Studies tor Assessment ol Potential Risks As.soei- 
ated With Recomhinant DNA experimentation," I'almouth, Mass., 
June 20 - 21 . 1977. 
'"Workshop to Assess Risks for Recomhinant UNA experiments 
Imolving \'iral Cenomes," cosponsored hv the National Institutes 
of Health and the european Molecular Biology Organization, 
Ascot, england, Jan. 26-28, 1978. 
than when it existed freely.* Experts in infec- 
tious disease have stressed repeatedly that the 
ability of a micro-organism to cause disease 
depends on a host of factors, all working togeth- 
er. Inserting a piece of DNA into a bacterium is 
unlikely to suddenly transform the oi'ganism 
into a virulent epidemic strain. 
Careful calculations can also allay fears about 
the damage a genetically engineered micro-or- 
ganism might cause. Doomsday scenarios of 
escaped E. coli that carry insulin or other 
hormone-producing genes were recently exam- 
ined in another workshop.® I’rior to this work- 
shop, newspaper accounts raised the possibility 
that an E. coli carrying the gene for human in- 
sulin production might colonize humans and 
thus upset the hormonal balance of the body. 
The participants calculated how much insulin 
could be produced. First, it was assumeil that a 
series of highly unlikely events would occur— 
accidental release, ingestion by humans, stable 
colonization of the intestine by E. coli K-12. E. 
coli constitutes approximately 1 percent of tlu? 
intestinal bacterial population, and it was 
assumed that all the normal E. coli would h(* 
replaced by the insulin-producing E. coli. Insulin 
is made in the foi-m of a precursor moh'cule, 
proinsulin. It was assumed that 50 p(>rcent of all 
bacterial protein [jroduction would h(> dexoted 
to this single pi'Otein, anotluM- highly unlikely 
situation. If so, 30 micrograms (;ig)— or 0 0 
units— would then he made in the inlestiiuv 
Although proteins are \’(M’v |)oorly ai)sorhed 
from the intestinal ca\ ity, it w as assumed foi' 
the sake of argument that 100 percent of the 
proinsulin would h(f absorbed into the circula- 
tion. Thus, 0.0 units of insulin would he added 
to the noi’mal dail\’ human production of 25 to 
30 units— an imperceptible difference. 
Calculations like these ha\c been cai i ied «)Ut 
for several other' hoi inones. I',\(*n with the most 
implausible seri(\s of (wents, leading to the 
gi'eatest oppoi’tunity for hormone pi'oduction. 
'On Iho ()lh(‘r hand. Il h.is hrnn .irgucil lhal ihiv ha\ |iiii\idi-a 
vim.so.s with a new mule lor (li.sM'ininalion Nf\ ri ihrlcss thi-n- i-. 
no cvidiMU'C that v irn.scs can icadih cm .ipr li om the li.n in i.i .mil 
.snh.s(HHi('nllv cau,s(‘ inicclion 
"'"National Inslilulc ol Mlcrg^v .ind Inicc lions I Iim .im-s v\ m k .liup 
on Recomhinant l).\ \ Risk \sscssincnl I’.is.idcn.i ( .ilil \|n 
11-12. 1981) 
