72 
[From Science Magazine. Sept. 30, 1977] 
Recombinant DNA: NIH Rules Broken in Insulin Gene Project 
(By Nicholas Wade) 
A breach of National Institutes of Health rules on gene splicing occurred 
earlier this year in the Department of Bqoohemfstry and Biophysics at the Uni- 
versity of California, San francisco. one of the leading ^enters for practice of 
the new technique. No hazard resulted, but the episode underlines some of the 
difficulties experienced by research fjibqi'atories iq adapting to the new rules. 
The breach was the use of a biological component, or “Vector,” before it had 
been certified by the NIH direqtqr. Tfic researchers, a team engaged in isolating 
the rat gene which codes for insulin, say they destroyed the experiment as soon 
as they realised their mistake. 
The experiment was repeated ip a certified veotor and published ia Science on 
17 June. It received considerable attention hecause the researchers had achieved, 
mtich eaflier than expected, the first step toward the goal of isolating the human 
insulin gene and using it for the manufacture of insulin protein- The UCSF team 
was in competition with a group at Harvard which was known to he working with 
a better sourc® material. 
USCF’s preeminence ia the gene-splicer's art has brought it some mixed blessings. 
Because of the practical implications of what its researchers are doing, a company 
called Geneqtech has established a relationship iHth Herbert Boyer, one of the 
pioneers of the technique. Members of the insulin team have set up a nonprofit 
corporation, the California Institute for Genetic Research. These commercial 
developments are a tribute to the department’s success, hut have also created 
interna} stresses. “Capitalism sticking its nose into the lab has tainted inter- 
personal relations — there are a numbr of people who feel rather strongly that 
there should be no commercialization of human insulin,” says UCSF micro- 
biologist David Martin. 
Another mixed blessing is fame, which has attracted press attention not only to 
the department’s achievements but also to certain internal tensions. A lengthy 
and circuihstantial article in the June issue of the Smithsonian called into ques- 
tion the respect accorded to the NIH safety rules by UCSF researchers, and in 
particular by the younger, postdoctoral workers who perform most of the experi- 
ments. Written fifter a 3-month internship in Boyer’s lab by Janet L. Hopson, 
formerly a reporter for Science News, the article observed that “half of the re- 
searchers here follow the guidelines fastidiously; others seem to care little. . . . 
Among the young graduate students and postdoctorates it seemed almost chic not 
to know the NIH rules,” Hopson noted. In a letter to the editor criticizing the ar- 
ticle, Boyer stated that “In practice [the NIH rules] are followed seriously.” 
The stresses of both commercial success and media attention came together 
this May when the insulin team announced their production of the rat gene. 
Other researchers resented not only the intrusive presence of the press but als<9 
the fact that they were hearing of their colleagues’ success for the first time. The 
team had worked in unusual secrecy, which many regarded as inappropriate in an 
academic setting as well as disruptive. “People would stop talking when you came 
into the room, or change the subject if you tried to make conversation about how 
the insulin project was going,” says UCSF biochemist Brian McCarthy. The 
ins^in team say that no secrecy was intended, and that it was the speed of ob- 
taining results that occasioned surprise. 
The secrecy and suspicion surrounding the insulin gene experiment, together 
with perhaps a touch of resentment, helped to fan rumors within the department 
alleging that the NIH rules had been broken and even that the experiment pub- 
lished in Science $night not have been performed as described. The focus of the 
rumors was the obvious fact that the whole intricate experiment had been corn- 
pitted only 3 weeks after the NIH had certified the vector which the researchers 
used. Even experienced gene splicers were surprised hy the rapidity of execution. 
“It is conceivably possible to do such an experiment }n three weeks if everything 
works perfectly the first time, but you know as well as I do that science never 
works as well as you hope,” a member of the department remarks. 
“Well, what cap I say? It did work well. We were all set to go,” says a member 
of the insulin team queried on the speed of the experiment. Yet members of the 
teasn concede that an earlier experiment took place, but say it was aborted half way 
through and before any pertinent information had been gained. 
The episode of the earlier experiment illustrates the problems experienced by 
research laboratories both in acclimatising to the NIH’s gene-splicing rules and 
in devising ways of enforcing them. “We can’t run a policing service,” says David 
[Appendix B — 121] 
