Role of type 1 pili and effects of phase variation on lower urinary tract infections produced by Escherichia coli.

Phase variation of type 1 pili (fimbriae) was studied during the in vivo growth of Escherichia coli in two animal models. In the first, a heavily piliated urinary tract isolate (strain 149) was placed in 1-cm polypropylene chambers sealed with 0.22-micron-pore-size filters. The chambers were surgically implanted intraperitoneally in mice and recovered at various times. Piliation, as determined by electron microscopy and by measuring the minimum number of bacteria needed to produce mannose-sensitive hemagglutination, gradually decreased, and by day 5, most of the organisms were nonpiliated. In the second model, piliated and nonpiliated E. coli phase variants were inoculated into the bladders of BALB/c mice via urinary catheters, and their fate in the lower urinary tract was studied. Viable counts of bladder homogenates revealed that piliated phase variants were significantly more effective in colonizing the bladder urothelium than were their nonpiliated counterparts. Specific antibody to type 1 pili prevented colonization by the piliated organisms. After inoculation of piliated variants, the bladder-associated bacteria gave rise to approximately 80% mannose-sensitive hemagglutination-positive colonies, and immunocytochemistry of bladder lavages revealed large numbers of type 1 piliated bacteria adhering to the bladder transitional cells. Electron microscopy confirmed the presence of piliated bacteria in association with the bladder urothelium. The urine of these mice, whose bladders were colonized with piliated bacteria, frequently showed no growth, and when bacteria were present, strain 149 yielded less than 30% hemagglutination-positive colonies. The results suggest that for some E. coli strains, phase variation may be a factor in determining the fate of the E. coli in the urinary tract and that the urine may not necessarily reflect the bacteriologic state of the bladder mucosa.     

Construction and expression of recombinant plasmids encoding type 1 or D-mannose-resistant pili from a urinary tract infection Escherichia coli isolate.

Isolates of Escherichia coli from human urinary tract infections frequently express adherence properties found less often among normal intestinal isolates. These properties include adherence to human uroepithelial cells and primary monkey kidney cells, as well as D-mannose-resistant hemagglutination of human erythrocytes, and they are mediated by a pilus type different from type 1. The genes encoding this pilus type (pyelonephritis-associated pili, pap) and those encoding type 1 pili have been cloned from a urinary tract infection isolate of E. coli and transferred to an E. coli K-12 derivative. The recombinant plasmids were found to express functional pili and to endow the new host with all of the adherence properties of the urinary tract infection isolate. Both pilus types were found to be genetically distinct, and unlike the adherence genes from bovine, porcine, and human diarrheal isolates, both were found to be chromosomally encoded.     

Escherichia coli pili as possible mediators of attachment to human urinary tract epithelial cells.

Presence of pili of fimbriae on Escherichia coli bacteria isolated from the urine of patients with urinary tract infection was related to the ability of the bacteria to attach to human uroepithelial cells. Piliated E. coli strains agglutinated guinea pig erythrocytes. D-Mannose and alpha-methyl-D-mannopyranoside inhibited this agglutination with all but one of the 12 strains tested. D-Mannose, D-galactose, alpha-methyl-D-mannopyranoside, and L-fucose did not afect attachment of piliated strains to uroepithelial cells. Heating as well as washing of piliated strains caused a parallel decrease of piliation and adhesive ability. Growth in glucose-enriched medium increased capsule formation but decreased piliation and adhesion. Capsulated strains retained their adhesive ability provided that pili extended outside the capsule.     

In vivo expression and variation of Escherichia coli type 1 and P pili in the urine of adults with acute urinary tract infections.

In vivo expression of pili by Escherichia coli in the urine of 41 adults with lower urinary tract infections was analyzed by immunostaining with polyclonal antiserum to type 1 and P pili. Type 1 pili were detected in 31 of 41 urine specimens, while P pili were detected in 6 of 18 specimens. The piliation status of bacterial populations in urine was heterogeneous, varying from predominantly piliated to a mixture of piliated and nonpiliated cells. Bacteria frequently adhered to exfoliated uroepithelial cells and leukocytes in urine. Expression of pili in vivo did not always correlate with the hemagglutination phenotype after growth in vitro. Strains isolated from different sites in the urogenital tract of two individuals showed phenotypic variation in the state of piliation. The results demonstrate that E. coli type 1 and P pili are expressed and are subject to variation in vivo during acute urinary tract infections in adults.     

Aerobactin production of serotyped Escherichia coli from urinary tract infections

Aerobactin production was examined by a bioassay in 467 Escherichia coli urinary strains from girls. All strains were of known O:K:H serotype. 139, 119 and 112 strains were isolates from pyelonephritis (Py), cystitis (Cy) and asymptomatic bacteriuria (ABU), respectively, and 97 were from fecal samples of healthy girls (FN). The incidence of aerobactin production was significantly higher among Py strains than among ABU and FN strains (P less than 0.001) and also significantly higher than among Cy strains (P less than 0.01). Aerobactin production was associated with serotype, e.g. the majority of O6:K2:H1 strains and of O16:K1:H6 were positive while e.g. the O6:K13:H1 strains were negative. There was no consistent pattern of coappearance of aerobactin and hemolysin.     

Aerobactin production of serotyped Escherichia coli from urinary tract infections

Aerobactin production was examined by a bioassay in 467 Escherichia coli urinary strains from girls. All strains were of known OKH serotype. 139, 119 and 112 strains were isolates from pyelonephritis (Py), cystitis (Cy) and asymptomatic bacteriuria (ABU), respectively, and 97 were from fecal samples of healthy girls (FN). The incidence of aerobactin production was significantly higher among Py strains than among ABU and FN strains (P<0.001) and also significantly higher than among Cy strains (P<0.01). Aerobactin production was associated with serotype, e.g. the majority of 06K2H1 strains and of 016K1H6 were positive while e.g. the 06K13H1 strains were negative. There was no consistent pattern of coappearance of aerobactin and hemolysin.     

Distribution of type 1 and P pili on uropathogenic Escherichia coli O6.

The distribution of type 1 and P pili on individual cells of an O6 uropathogenic Escherichia coli strain, 6260, was determined immunologically with pilus-specific monoclonal antibodies by indirect immunofluorescence and immunogold electron microscopy. Variations in pilus expression under different culture conditions were monitored with an indirect immunofluorescence assay; 63% of piliated cells expressed type 1 pili when grown on agar at 37 degrees C versus 14 to 38% when grown in broth at 37 degrees C. In contrast, generally fewer cells with P pili (18 to 44%) were detected on agar than when grown in broth (up to 86%). Both type 1 and P pili were absent from cells cultured at 20 degrees C. Immunogold and immunofluorescence double labeling techniques with monoclonal antibodies 11-2 and 91-1 were used to study subpopulations of cells with type 1 and P pili; 39 to 41% of the piliated cells demonstrated only type 1 pili, and 12 to 16% of the cells showed only P pili. The immunogold method proved more sensitive than the immunofluorescence technique for detecting subpopulations expressing both pili types simultaneously, 19 versus 7%. We observed variations between type 1 and P pili, both in expression on individual cells and in the distribution of subpopulations of cells.     

Six widespread bacterial clones among Escherichia coli K1 isolates.

Variable properties among Escherichia coli isolates include serotype, electrophoretic migration of major outer membrane proteins, metabolic properties, production of hemolysin or colicin or both, and plasmid content. These characteristics were compared in E. coli strains of capsular types K1, K5, K92, and K100 and in non-encapsulated isolates. The 234 bacterial strains from the United States and Europe which we studied had been isolated from healthy or diseased individuals recently or as long ago as 1941. Regardless of source, most O7:K1, O16:K1, and O75:K100 isolates could be assigned to three unique, serotype-specific groups, which were interpreted as representing three bacterial clones. Two bacterial (sub)clones each were discerned among the O18:K1 and O18:K5 isolates, and two further, distinct clones were discerned among the O1:K1 isolates. The implications of these results for epidemiological analyses and for virulence are discussed.     

Relationship of type 1 pilus expression in Escherichia coli to ascending urinary tract infections in mice.

The role of type 1 pili and P adhesins during the in vivo growth of Escherichia coli inoculated into the urethras of BALB/c mice was studied. Strains which produced type 1 pili when grown in broth but lost this trait when grown on agar (regulated variants) were tested. Broth-grown organisms colonized the bladder of every animal tested, with counts of 10(3) to 10(4) viable organisms recovered from bladder homogenates. Agar-grown organisms gave lower rates of infection and the number of viable organisms recovered from bladders was significantly reduced. The degree of inoculum piliation influenced bladder colonization in a direct way: as piliation increased, the number of bacteria recovered from bladders also increased. After intraurethral inoculation, all of the bladders and 44% of the kidneys were colonized on day 1, and by day 5, 94% of the bladders and 16% of the kidneys were positive. Hemagglutination titers remained high for the bladder isolates, but the organisms colonizing the kidneys became significantly less piliated with time. Bacteriuria was unrelated to bladder or renal colonization. Strains that demonstrated random phase variation of type 1 pili during growth on agar produced similar colonizations of the urinary tract with broth- and agar-grown inocula. Strains that produced only P adhesins were less effective in colonizing the urinary tract than were type 1 piliated organisms. Other strains which did not produce pili only minimally colonized the bladder. The results suggest that type 1 pili play an essential role in ascending infections of the urinary tract.     

Stimulation of human polymorphonuclear leukocyte oxidative metabolism by type 1 pili from Escherichia coli.

We compared the degree to which Escherichia coli phase variants which do (T1P+ E. coli) or do not (T1P- E. coli) express type 1 pili (T1P) stimulate human polymorphonuclear leukocyte (PMN) oxidative activity. Unopsonized T1P+ E. coli stimulated the release of 0.20 to 0.24 nmol of H2O2 per 10(6) PMN per min and the consumption of 1.4 to 4.0 nmol of O2 per 10(6) PMN per min; no measurable PMN oxidative activity was stimulated by unopsonized T1P- E. coli. In the presence of serum opsonins, T1P+ E. coli stimulated the release of 1.12 to 1.16 nmol of H2O2 per 10(6) PMN per min and the consumption of 5.0 to 6.0 nmol of O2 per 10(6) PMN per min, whereas T1P- E. coli stimulated the release of 0.42 to 0.43 nmol of H2O2 per 10(6) PMN per min and the consumption of 0.6 to 2.0 nmol of O2 per 10(6) PMN per min. Although unaggregated T1P did not stimulate PMN, latex beads coated with T1P (T1P-latex) stimulated alpha-methylmannoside-inhibitable, opsonin-independent PMN oxidative activity. The activity stimulated by either T1P+ E. coli or T1P-latex was susceptible to inhibition by cytochalasin B. Latex particles coated with bovine serum albumin or mannose-resistant pili did not stimulate PMN. These data indicate that T1P+ E. coli stimulate PMN oxidative metabolism more effectively than do T1P- E. coli and that a similar PMN oxidative response follows cellular stimulation by either unopsonized T1P+ or opsonized T1P- E. coli. Furthermore, T1P-latex faithfully mimics the ability of T1P+ E. coli to stimulate PMN oxidative metabolism. Such particles may be useful in further analyses of cellular responses to T1P+ E. coli.     

Iron-suppressible production of hydroxamate by Escherichia coli isolates.

A total of 476 strains of Escherichia coli isolated from humans, pigs, cattle, poultry, potable water, or effluent were examined for iron-suppressible ability to produce hydroxamate. Isolates able to produce such material (Hyd+ isolates) are presumed to be able to carry out hydroxamate-dependent transport of iron. The percentages of Hyd+ isolates found among E. coli isolated from the feces of breast-fed babies (71%), adults (46%), milk-fed calves (32%), or poultry (28%) were significantly greater (P less than 0.01) than the percentages isolated from potable water and effluent (6%) or from the feces of suckling piglets (6%), weaned pigs (6%), or weaned cattle (4%). The percentages of Hyd+ isolates found among E. coli associated with diarrhea in humans (51%), weaned pigs (7%) or calves (25%) were not significantly different (P greater than 0.1) from those found among strains isolated from corresponding nondiarrheic hosts. Many of the E. coli isolated from cases of E. coli bacteremia in humans and poultry were Hyd+ (64% and 83%, respectively). We conclude that ability to carry out hydroxamate-mediated transport of iron is widely distributed among natural isolates of E. coli but that the distribution of Hyd+ E. coli is not random. E. coli isolated from sources where levels of available iron might be expected to be low tend to be Hyd+. It seems that a link may exist between prevalence of Hyd+ E. coli and active host-defense based on restricted availability of iron.     

Virulence factors in Escherichia coli urinary tract infection.

Uropathogenic strains of Escherichia coli are characterized by the expression of distinctive bacterial properties, products, or structures referred to as virulence factors because they help the organism overcome host defenses and colonize or invade the urinary tract. Virulence factors of recognized importance in the pathogenesis of urinary tract infection (UTI) include adhesins (P fimbriae, certain other mannose-resistant adhesins, and type 1 fimbriae), the aerobactin system, hemolysin, K capsule, and resistance to serum killing. This review summarizes the virtual explosion of information regarding the epidemiology, biochemistry, mechanisms of action, and genetic basis of these urovirulence factors that has occurred in the past decade and identifies areas in need of further study. Virulence factor expression is more common among certain genetically related groups of E. coli which constitute virulent clones within the larger E. coli population. In general, the more virulence factors a strain expresses, the more severe an infection it is able to cause. Certain virulence factors specifically favor the development of pyelonephritis, others favor cystitis, and others favor asymptomatic bacteriuria. The currently defined virulence factors clearly contribute to the virulence of wild-type strains but are usually insufficient in themselves to transform an avirulent organism into a pathogen, demonstrating that other as-yet-undefined virulence properties await discovery. Virulence factor testing is a useful epidemiological and research tool but as yet has no defined clinical role. Immunological and biochemical anti-virulence factor interventions are effective in animal models of UTI and hold promise for the prevention of UTI in humans.     

Identification of urovirulence traits in Escherichia coli by comparison of urinary and rectal E. coli isolates from dogs with urinary tract infection

Spontaneously occurring urinary tract infection (UTI) in dogs was exploited as an experiment of nature to gain insights into UTI pathogenesis in humans. Concurrent urinary and rectal Escherichia coli isolates from 37 dogs with UTI were compared with respect to phylogenetic background, O antigens, and extended virulence genotype. In 54% of the UTI episodes, the dog's urinary and rectal isolates represented the same strain. Urinary isolates differed dramatically from rectal-only isolates in that they derived predominantly from E. coli phylogenetic group B2, expressed typical (human) UTI-associated O antigens, and possessed many virulence-associated genes, most notably pap elements (P fimbriae), papG (adhesin) allele III, sfa/foc and sfaS (S fimbriae), hly (hemolysin), fyuA (yersiniabactin), iroN (siderophore), and ompT (outer membrane protease T). The 20 urinary isolates that corresponded with the host's predominant rectal strain were no less virulent according to the markers analyzed than were the 17 urinary isolates that differed from the host's predominant rectal strain. These findings suggest that UTI pathogenesis is similar in dogs and humans, provide added support for the special-pathogenicity over the prevalence hypothesis of UTI pathogenesis, and identify numerous specific virulence-associated factors as significant correlates of urovirulence.     

Clonal relationships among bloodstream isolates of Escherichia coli.

The clonal relationships among 187 bloodstream isolates of Escherichia coli from 179 patients at Boston, Mass., Long Beach, Calif., and Nairobi, Kenya, were determined by multilocus enzyme electrophoresis (MLEE), analysis of polymorphisms associated with the ribosomal operon (ribotyping), and serotyping. MLEE based on 20 enzymes resolved 101 electrophoretic types (ETs), forming five clusters; ribotyping resolved 56 distinct patterns concordant with the analysis by MLEE. The isolates at each study site formed a genetically diverse group and demonstrated similar clonal structures, with the same small subset of lineages accounting for the majority of isolates at each site. Moreover, two ribotypes accounted for approximately 30% of the isolates at each study site. One cluster contained the majority (65%) of isolates and, by direct comparison of the ETs and ribotypes of individual isolates, was genetically indistinguishable from the largest cluster for each of two other collections of E. coli causing pyelonephritis and neonatal meningitis (R. K. Selander, T. K. Korhonen, V. Väisänen-Rhen, P. H. Williams, P. E. Pattison, and D. A. Caugent, Infect. Immun. 52:213-222, 1986; M. Arthur, C. E. Johnson, R. H. Rubin, R. D. Arbeit, C. Campanelli, C. Kim, S. Steinbach, M. Agarwal, R. Wilkinson, and R. Goldstein, Infect. Immun. 57:303-313, 1989), thus defining a virulent set of lineages. The isolates within these virulent lineages typically carried DNA homologous to the adhesin operon pap or sfa and the hemolysin operon hly and expressed O1, O2, O4, O6, O18, O25, or O75 antigens. DNA homologous to pap was distributed among isolates of each major cluster, whereas hly was restricted to isolates of two clusters, typically detected in pap-positive strains, and sfa was restricted to isolates of one cluster, typically detected in pap- and hly-positive strains. The occurrence of pap-positive isolates in the same geographically and genetically divergent lineages suggests that this operon was acquired early in the radiation of E. coli, while hly and sfa were acquired subsequently, most likely by pap-positive and pap- and hly-positive precursors, respectively.     

Conservation of plasmids among Escherichia coli K1 isolates of diverse origins.

Escherichia coli K1 isolates of various O types were previously assigned to different clonal groups. Members of the two clones defined by membrane pattern 9 (MP9) and serotypes O18:K1 and O1:K1 had been found to be very similar to each other. The plasmid contents of these bacteria confirmed this conclusion. Both groups carried a self-transmissible plasmid of the FI incompatibility group that coded for colicin production and a major outer membrane protein called the plasmid-coded protein (PCP). The size of this plasmid varied from 76 to 96 megadaltons, but restriction endonuclease digestion and DNA heteroduplex analysis revealed that these plasmids were highly related. O18:K1 bacteria of MP6 had previously been determined to represent a subclone, related to but different from O18:K1 MP9 bacteria. These MP6 bacteria carried a different, smaller IncFI plasmid which did not code for colicin production or the PCP protein. This smaller plasmid was primarily related to the larger plasmid within the regions of DNA encoding incompatibility, replication, and conjugation. O1:K1 bacteria of MP5 contained other unrelated plasmids in agreement with the previous conclusion that they are unrelated to O1:K1 bacteria of MP9. The bacteria examined had been isolated from two continents over a time span of 38 years, and the results attest to conservative inheritance of plasmids within bacteria of common descent.     

Expression of type 1 fimbriae may be required for persistence of Escherichia coli in the catheterized urinary tract.

Long-term urinary catheterization results in polymicrobial bacteriuria and is complicated by fever, bacteremia, acute pyelonephritis, and death. Escherichia coli is a common urine isolate from catheterized patients and can persist for months. We hypothesized that fimbria-mediated adherence contributes to its persistence. For 1 year, urine specimens were collected from 51 patients greater than or equal to 65 years of age who were catheterized for greater than or equal to 30 days. E. coli was isolated at greater than or equal to 10(5) CFU/ml from 447 (36%) of 1,230 weekly urine specimens from 26 patients. Week 1 isolates from 52 definable episodes were tested for hemagglutination, hybridization with gene sequences from the pil and pap operons, in vitro adherence to catheter material, binding of 125I-labeled Tamm-Horsfall protein, hemolysin and colicin V production, and serum resistance. The proportions of isolates of short (1 week only), medium (2 to 11 weeks) and long (greater than or equal to 12 weeks) episodes of bacteriuria which expressed type 1 fimbriae as assayed by mannose-sensitive hemagglutination were 59, 65, and 92%, respectively. Isolates with the pil operon (the genome for type 1 fimbriae) from episodes lasting greater than 1 week expressed mannose-sensitive hemagglutination more frequently (P = 0.011) than pil-positive isolates from episodes of less than or equal to 1 week. Isolates from episodes of greater than 1 week also bound significantly more Tamm-Horsfall protein than isolates from episodes of less than or equal to 1 week (P = 0.044). Although nearly half of the isolates produced P fimbriae, an important virulence factor for the development of pyelonephritis, no correlation with persistence could be made. Overall, the E. coli isolates expressed traits similar to those of strains that caused cystitis. Type 1 fimbriae appear to be important for the persistence of E. coli in the long-term-catheterized urinary tract.