Dr. William J. Gartland, Jr. 
January 20, 1987 
research process. Specific comments regarding the proposed 
changes are: 
A. Industrial experience from decades (E. coli and B. 
subtilis ) to hundreds of years (S. cerevisiae ) supports 
the contention that fermentations using nonrecombinant 
strains of E. coli , B. subtilis and S. cerevisiae are 
essentially benign. L-asparaginase , produced by Eli 
Lilly and Company in the early 1970s, was among the 
first examples of a commercially available E. coli 
fermentation product to be used clinically. (Grinnan, 
E. L. , L-Asparaginase : A Case Study of an E. coli 
Fermentation Product, In: Insulins, Growth Hormone, and 
Recombinant DNA Technology , John L. Gueriguian, ed.. 
Raven Press, New York, 1981). L-asparaginase was 
produced in conventional fermenters at the 40,000L scale 
with no untoward safety problems either to workers or 
to the environment. Furthermore, we have a long and 
distinguished record in fermentation techniques which 
utilize large-scale production of a wide range of 
organisms, most notably Streptomyces , Actinomyces , 
Penicillium , and Cephalosporium . Over the last forty 
years, with the exception of isolated cases of hyper- 
sensitivity reactions, we have experienced no health 
associated risks involving large-scale production 
processes with these organisms. These reactions when 
they occurred were always associated with the product 
from the fermentations rather than any inherent problem 
associated with the organism itself. If this proposed 
revision is approved by the NIH Recombinant Advisory 
Committee and accepted by the Director, NIAID, it would 
be the policy of this company to immediately report any 
novel and unexpected health or environmental problem 
which could be a result of this proposed revision, to 
the NIAID and the local IBC, as well as the steps taken 
to address the problem. 
B. Risk assessment studies have consistently failed to show 
any significant risk associated with any of the above- 
mentioned hosts carrying plasmids coding for peptides of 
animal or human origin. 
C. As was made clear at the Asilomar Conference, the 10L 
volume limit stipulated in the laboratory Guidelines was 
one merely of convenience and was not intended to imply 
that large volumes are significantly more hazardous than 
Recombinant DNA Research, Volume 1 1 
[261] 
