23 
B. SOME EXPERIMENTS WHICH HAVE YET TO BE PERFORMED 
§22. The testing of a modified Escherichia coli EK-2 host-vector 
system must assess its viability in other than specifically designed and 
carefully regulated laboratory environments. These environments should 
represent , -.most reasonably, naturally occurring environments which E. 
coli is known to inhabit. These environments include soils, raw sewage, 
arthropod intestinal tracts, and the human gut. Since there is' a v/ide 
rangfe of passage response to Escherichia coli among the human populati on , 
any EK-2 host should be tested for survival in the human CI_ tract prior 
to certification . In addition, given the data presented cn the survival 
of yXll 6 and xl^76 in tap water, the likelihood of Xl?76 reaching the 
sewer were it accidentally poured down a drain, is quite high. Raw 
sewage is a rich environment harboring a plethora of possibilities for 
genetic exchange between coliform and other types of bacteria, as 
evidenced by the diversity and number of coliphages one finds in a small 
volume of raw sewage (R. Goldstein, unpublished results). 
§23. The genetic construction of an EK-2 strain from a parent strain 
should be done as isogenically as possible, so that comparisons of 
phenotype do not rely on markers other than those intentionally intro- 
duced. The presence of secondary metabolic systems such as ebg (Campbell, 
Lengyel, and Langridge, 1973) which can rescue function of a particular 
allele by bypassing the normal biochemical pathways employed by E. coli- 
in the laboratory, in other than a laboratory environment, should be 
assessed directly by testing in a number of different natural environments. 
§24. If 'a strain as complex as X1776 (in the sense of having many 
mutational markers) were distributed among many laboratories wishing to 
[Appendix C — 211] 
