Characterization of Escherichia coli strains causing urinary tract infections in patients with transposed intestinal segments.

PURPOSE: Transposition of intestinal segments into the urinary tract predisposes to urinary tract infections. We characterized bacterial infections in these patients and examined the virulence genotype and persistence of Escherichia coli isolates. MATERIALS AND METHODS: We followed 26 patients who underwent bladder reconstructive surgery using transposed intestinal segments. E. coli strains isolated from the urine of these patients were genotyped for established virulence determinants and the frequency of carriage was compared with E. coli strains isolated from community acquired urinary infections and the fecal flora of anonymous volunteers. A longitudinal study of E. coli strains in 9 patients was also done using pulsed field gel electrophoresis. RESULTS: E. coli was the most frequently isolated organism, responsible for 59% (62 of 105) of monobacterial infections. Other bacteria isolated included Klebsiella species, Proteus species and Enterococcus faecalis. Community acquired E. coli strains were more likely to carry multiple determinants for particular adhesins (P and S fimbriae) and toxins (alpha-hemolysin and cytotoxic necrotizing factor) than fecal strains. Carriage frequency for bladder reconstruction strains was intermediary and not significantly different. The key finding was that E. coli strains persisted for prolonged periods, including 2 years in certain patients, often despite various antimicrobial treatments. CONCLUSIONS: This study highlights that further steps must be taken to prevent and treat urinary tract infections in this susceptible group. Particular attention should be given to the treatment of persistent infections.

Analysis of Escherichia coli strains causing bacteriuria during pregnancy: selection for strains that do not express type 1 fimbriae.

Escherichia coli isolates from patients with bacteriuria of pregnancy were compared by PCR with isolates from patients with community-acquired cystitis for the presence of established virulence determinants. The strains from patients with bacteriuria of pregnancy were less likely to carry genes for P-family, S-family, and F1C adhesins, cytotoxic necrotizing factor 1, and aerobactin, but virtually all of the strains carried the genes for type 1 fimbriae. Standard mannose-sensitive agglutination of yeast cells showed that only 15 of 42 bacteriuria strains (36%) expressed type 1 fimbriae compared with 32 of 42 strains from community-acquired symptomatic infections (76%) (P < 0.01). This difference was confirmed by analysis of all isolates for an allele of the type 1 fimbrial regulatory region (fim switch), which negates type 1 fimbrial expression by preventing the fim switch from being inverted to the on phase. This allele, fimS49, was found in 8 of 47 bacteriuria strains from pregnant women (17.0%) compared with 2 of 60 strains isolated from patients with cystitis (3.3%) (P < 0.05). Determination of the phase switch orientation in vivo by analysis of freshly collected infected urine from patients with bacteriuria showed that the fim switch was detectable in the off orientation in 17 of 23 urine samples analyzed (74%). These data indicate that type 1 fimbriae are not necessary to maintain the majority of E. coli bacteriurias in pregnant women since there appears to be selection against their expression in this particular group. This is in contrast to the considered role of this adhesin in community-acquired symptomatic infections. The lack of type 1 fimbria expression is likely to contribute to the asymptomatic nature of bacteriuria in pregnant women, although approximately one-third of the bacteriuria isolates do possess key virulence determinants. If left untreated, this subset of isolates pose the greatest threat to the health of the mother and unborn child.