Infectious Disease View of Recombinant DSA 
610 
On the basis of current evidence, it appears that 
both formulations arc relevant. There is no doubt 
that the organisms causing such infection must 
be implanted in the intestinal microflora in or- 
der to coloni/e the periurethral mucosa subse- 
quently. This explains the high correlation be 
tween serotypes in the urinary tract and those 
in the feces. There are. however, a multiplicity 
of serotypes in the intestinal flora, whereas only 
a single serotyjx;, or on rare occasions two. causes 
infection at any given time. Hence, there is selec- 
tivity operating, which can be further demon- 
strated by the virulence factors associated with 
urinary tract pathogens. The evidence strongly 
suggests that urinary tract pathogens must be ca- 
pable of colonizing the bowel of the host, as well 
as possessing certain special properties. 
Studies of £. t ali K 12 
The pedigTce of this classic strain starts with an 
isolation at Stanford in 1922 from the feces of a 
patient with diphtheria. Tlte strain was main- 
tained for many years in laboratory culture until 
its use in genetic studies by l.edcrberg and Ta- 
tum in 1017. Since that time, it lias been pass- 
aged on numerous agar slants and widely em- 
ployed in a variety of experimental situations. 
Actually, there are a vast number of progeny of 
the original K12 strain which possess different 
genotypic and phenotypic characteristics. The 
pedigrees of these derivatives have been summar- 
ized [39], All of them, however, are "rough" mu- 
tants with no detectable O-antigen specificity 
[40]. Their features are related to a block in the 
synthesis of O-specific chains of lipopolysaccha- 
ride clue to a mutation in the h is (histidine)- 
linked tfb region of the chromosome [41]. Sero- 
logical and phage studies show that the lipopoly- 
saccharitle core of F.. coli KI2 is different from 
all known core types of wild-type E. colt and Sal- 
mottclla [41. 12]. The M antigens (cholanic acid), 
which are common to most Enterobacteriaceae, 
are elaborated by the K12 strain; this antigen is 
responsible for the mucoid character of K.12 or- 
ganisms growing on agar plates [21]. 
Since colonization of the intestine is fell to be 
an initial event in many pathologic states involv- 
ing E. coli, it is natural that this feature has been 
the subject of several investigations. It is fair to 
state that there have been no instances in which 
the ingested strain of E. coli K12 lias been im- 
planted in the human intestine. Smith fed eight 
different E. coli K12 strains, containing various 
transmissible plasmids, to a volunteer at high do- 
ses (10") [-13]. Some strains could not be iso- 
lated at all from the feces, while others persisted 
in progressively reduced counts for a period of up 
to four days. The experiment was subsequently 
repeated, and there was again no persistence be- 
yond lour days. To increase the likelihood of im- 
planting E. colt KI2 in the gut. Smith then used 
a KI2 strain that had inserted in it the colicin V 
(ColV) plasmid of a wild-tv pc F.. coli [44]. (ColV 
promotes the survival of wild-type E. coli in the 
intestinal tract.) When a volunteer consumed the 
RI2 strains with and without the ColV plasmid, 
the strains were eliminated from the feces in an 
ecpial time frame, none persisting more than four 
days. 
Anderson attempted similar implantations 
with F.. coli K12 strains in eight volunteers who 
received doses of up to 10'" organisms [45]. The 
maximal period of fecal excretion ol the test 
strains was six days, with a mean of three days. 
These studies were repeated, again using eight 
volunteers, and the same findings were observed. 
Gorbach re|x>rtcd at this meeting unsuccessful 
attempts at implanting E. coli K12 in two pa- 
tients with defective bacterial clearing mecha- 
nisms in the small bowel In one case, the patient 
had a stricture in the ileum, and the other pa- 
tient had severe diarrhea due to cholera. The 
K12 strain was eliminated from the small bowel 
and feces within 21 hr in both patients. 
A number of investigators have tried in vain 
to implant El. coli K12 in the intestinal tract of 
laboratory animals [10. 16], Negative results have 
been noted in mice, rats, chickens, pigs, and 
calves E. coli K12 has been able to colonize the 
stomach of starved sheep. However, these animals 
possess a complex stomach consisting of a ru- 
men; Smith has referred to this situation as a 
“test-tube" experiment. Freter has told us at this 
meeting of his success in implanting £. coli K12 
into a germ-free mouse [10], The germ-free ani- 
mal will accept virtually any enteric organism 
for colonization even when the strain cannot im- 
plant in animals with a conventional microflora. 
On the other hand, Freter noted that the fur- 
1157] 
