Page 16 of Attach. E 
It seems reasonable to assume that persistence of foreign DNA in the 
prokaryotic or eukaryotic host genome, point c., will require selective 
pressure. Such persistence is clearly demonstrated by the expression of 
foreign DNA. 
Expression of foreign DNA, point d., is well documented in the case of 
yeast and Neurospora genes (R. Davis, J. Carbon, and others) able to 
complement nutritional deficiences in E. coli. The production of human 
chorionic gonadotrophin by bacteria isolated from cancer patients 
(quoted by Adelberg — Document No. 13) might represent, if confirmed, 
another instance of expression of eukaryotic genes in bacteria. It should 
be pointed out, however, that the complex processing, involving RNA 
splicing, which has been found in several genes from animals makes it 
very unlikely that a correct expression of such genes may take place in 
bacteria. 
We note that the expression observed to date has been with genes of 
lower eukaryotes in prokaryotes. The ability to be expressed may reflect: 
(1) their closer phylogenetic relatedness, (2) the fact that larger gene 
samples must be tested for more complex eukaryotes or (3) a basic 
difference between genes of metabolic pathways (for some of which 
expression has been documented) and genes of differentiated function 
(so far shown to contain inserts which must be processed in futher 
steps and are not present in mature mRNA). Under experimental con- 
ditions there is, in theory, no limit to the exchange of DNA between 
eukaryotes and prokaryotes. The effectiveness and biological 
significance of such exchanges will be determined not by whether the 
DNA can be exchanged, but by factors determining whether or not it can 
persist in the host-cell and be expressed in it. 
Question 4: 
Determine whether toxic recombinants (e.g. organisms bearing 
botulinum, tetanus or diptheria toxin genes) can be constructed and/or 
whether “shotgun" experiments using DNA from various sources can 
produce lethal infections, tumors, etc. . . . 
Comments: 
Concerning question 4, the construction of toxic recombinants ap- 
pears to be quite possible for prokaryotic toxin genes cloned in 
prokaryotes, and rather unlikely for eukaryotic toxin genes cloned in 
prokaryotes for the reasons given in discussion of Question 3 (see also 
discussion of Question 7). 
Whether “shotgun” experiments using DNA from various sources can 
eventually result in dangerous infections, etc. ... is not yet clear, but 
is conceivable in some cases, for example, when manipulations are 
made with dangerous pathogens. Clarification of such questions is the 
stated objective of the "polyoma” type experiments sponsored by 
EMBO and NIH and described by Drs. J. Tooze (Document No. 1) and M. 
Martin (Document No. 13). Until experience dictates otherwise, it seems 
reasonable that organisms carrying recombinant DNAs should be 
handled under the conditions appropriate for the most hazardous of the 
recombinant’s components. 
- 10 - 
[ 395 ] 
