18 
in late March in accord with newly adopted test criteria for EK2 
vectors. The committee finally recommended certification of pBR322 
on June 23. 
In his testimony to the subcommittee, William Rutter, chairman of 
the UCSF Department of Biochemistry and Biophysics and a mem- 
ber of the insulin research team, attributed the premature use of 
pBR322 to information that the RAC had fully “approved” the vec- 
tor in January and to a literal interpretation of the guidelines’ refer- 
ence to certification by the RAC. He also argued that he and his col- 
leagues were convinced of the vector’s safety by the supporting data 
given to NIH : “By one criterion of ‘safety,’ this plasmid is 10,000 to 
1 million times better than EK2 plasmid vectors then approved.' r 
Herbert Boyer, a member of the UCSF department who developed 
pBR322 and applied for its certification, testified that he was aware 
that it had not been certified in January but did not immediately in- 
form members of the insulin team and did not discover that they were 
using it until early March. Director Fredrickson agreed that there 
had been understandable confusion about NIH procedures in early 
1977 but stated that NIH had subsequently improved its communica- 
tions with grantees. Boyer and Rutter assured the subcommittee that 
the university had also adopted new procedures to prevent similar 
occurrences. 
Although the use of pBR322 posed no hazard to the researchers or 
others, the subcommittee believes that the incident has serious impli- 
cations for Government and institutional regulation of recombinant 
DNA research. Documents and testimony in the record of the subcom- 
mittee’s hearing support the following additional findings concerning 
this violation : 
Contrary to the implication of its November 1976 memo, NIH 
did not promptly notify applicants or prospective users of host- 
vector systems in writing of the actions by the RAC or the Direc- 
tor. Researchers at UCSF who decided to proceed with the 
pBR322 experiments relied on information from technicians in 
their laboratory who said they had heard, indirectly, that the 
vector had been approved. No one on the research team attempted 
to confirm its status with NIH officials or consulted the bio- 
hazards committee prior to the experiments. 
Approximately 3 weeks after the experiments commenced. 
Boyer scheduled a meeting at which he informed members of 
the insulin team that pBR322 had not been certified; apparently, 
they did not reveal that they were using it. The researchers con- 
tinued to grow and examine the host organisms in which they 
had inserted the plasmid in order to determine whether they con- 
tained the recombinant DNA. 
At the meeting in early February, Boyer also announced that a 
log book would be kept to record use of the P3 laboratory. 
(Neither the guidelines nor the institutional biohazards com- 
mittee required such a log book.) No official entries were made, 
however, until mid-March. At that time the use of pCRl, the first 
certified vector, was recorded retroactively to February 1. Later, 
the use of pMB9, certified in April, was also listed. There was no 
reference, however, to the use of pBR322 in the interim. 
[Appendix B — 278] 
