- Email: [email protected]
In Vitro and in Vivo Adherence of Uropathogenic Escherichia Coli Strains
0022-534 7/86/1356-1319$02.00/0
THE
Vol. 135, June Printed in U.S.A.
JOURNAL OF UROLOGY
Copyright © 1986 by The Williams & Wilkins Co.
IN VITRO AND IN VIVO ADHERENCE OF UROPATHOGENIC ESCHERICHIA COLI STRAINS WALTER J. HOPKINS, JEAN L. JENSEN, DAVID T. UEHLING*
AND
EDWARD BALISH
From the Department of Surgery (Urology), University of Wisconsin, Madison, Wisconsin
ABSTRACT
Twenty-eight Escherichia coli strains isolated from the urine of patients with urinary tract infections were assayed for fimbrial type, in vitro capacity to agglutinate guinea red blood cells, and in vivo adherence to rat bladder uroepithelium. A direct correlation was found between hemagglutinating and in vivo adherence. Strains with both Type 1 and P fimbriae showed the greatest adherence in vivo. Of the 28 strains, seven did not manifest either Type l or P fimbriae but agglutinated red blood cells and did adhere in vivo. In studies on bacterial adherence and urinary tract both in vivo and in vitro studies may contribute to understanding the relevance of bacterial adhesins in initiating urinary tract infections. Adherence of bacteria to mucosal cell surfaces is thought to be the initial step in many types of bacterial infections. The virulence of E. coli for the urinary tract is thought to be directly related to the presence of specific cell-surface structures, or "adhesins," which mediate bacterial adherence to epithelial cells. 1- 3 Among the surface adhesins that have been identified, fimbriae are thought to be particularly important in the binding of bacteria to epithelial cell membranes.4- 7 Several different types of fimbriae have been identified and characterized on the basis of structure, antigenic type, and receptor specificity. For example, Type 1, S, and P fimbriae are morphologically identical, but differ both antigenically and in their attachment sites on cell surfaces. Type l fimbriae bind to mannose-like receptors on the surface of epithelial cells and guinea pig red blood cells (GPRBC). 2 • 5 · 8 P fimbriae attach to carbohydrate moieties which are similar to human P blood group antigens,5 and S fimbriae recognize sialyl galactosides. 9 • 10 Thus, bacteria which possess these fimbriae, either singly, or in combination, can adhere avidly to mucosa! cell surfaces and agglutinate either red blood cells or particles coated with appropriate fimbrial receptors. Much of the current research on urinary tract infections (UTI) has focused on the adherence process and how it can be quantified. Many of these studies have examined adherence using in vitro systems in which bacteria bind to desquamated urinary epithelial cells, buccal epithelial cells, or mammalian red blood cells.'· 2 We have previously reported on bacterial adherence to the rat bladder in vivo and believe that this model has some advantages over in vitro methods since it measures adherence under physiologic conditions."· 12 In this report, we utilized both in vivo and in vitro systems to compare the adherence properties of E. coli strains isolated from patients with UTI. MATERIALS AND METHODS
Twenty-eight Escherichia coli strains were isolated from the urine of patients with urinary tract infections (UTI) at the University of Wisconsin Center for Health Sciences. Ten of the E. coli isolates came from patients being evaluated by the urology service. Based on the available evaluation, the patients' infections were graded as to seriousness based on abnormalities seen on intravenous pyelogram, voiding cystourethrogram and other diagnostic tests. The site and severity of the other patients' UTis were not known because + _,••• " , ••• of their medical Accepted for publication January 13, 1986. * for reprints: Dept. of Surgery, 600 Highland Ave., Madison Supported by TJ. S. Public I-iealtD. Service grant J\J\130808.
records. The 28 E. coli strains were typed for fimbriae, tested for their capacity to agglutinate RBC, and to adhere in vivo to rat bladder uroepithelium. For determination of hemagglutination (HA) titers, bacterial isolates were subcultured in Mueller-Hinton broth (Difeo Laboratories) or on colonization factor antigen (CF A) agar 13 plates and resuspended in 0.05 M phosphate buffered saline (PBS), pH 7.0 at a concentration of 5 x 108 bacteria/ml. Two-fold serial dilutions of the bacteria were made in PBS and to each dilution was added an equal volume of GPRBC resuspended at 3 per cent (v/v) in PBS. The dilutions were agitated for one minute. HA titers were assessed at one hour after incubation at room temperature and then again after 18 hours incubation at 4C. The HA titers are expressed as the lowest number of bacteria required to produce visible hemagglutination. 5 The E. coli strains were tested for the presence of both type 1 and P fimbriae. To assay for Type 1 fimbriae, each isolate was passaged three times in tryptose broth; each passage was for 48 hours at 37C without agitation. One drop of broth culture was then mixed with either a drop of 5 per cent GPRBC or 5 per cent GPRBC and 5 per cent D-mannose on a glass microscope slide. Agglutination of GPRBC which was inhibited by D-mannose indicated the presence of Type 1 fimbriae on the E. coli isolate. P fimbriae were assayed by the procedure of Svenson et al. 14 Briefly, a sample of each isolate was taken following the third passage in tryptose broth (Difeo Laboratories) and grown overnight at 37C on CF A agar plates. Several colonies from the CF A agar were then combined with 20 µl. of PBS on a microscope slide. The P particle agglutination (PP A) test was used for determination of P fimbriae by adding 20 µI. of the test or control solutions to the bacterial suspension. 15 Agglutination of the test solution but not the control solution indicated the presence of P fimbriae. Bacterial adherence in vivo was measured by previously described methods. 11 Briefly, 3H-labelled bacteria were inoculated into the rat bladder. The inoculum was allowed to remain in the bladder for two hours, after which the bladders were excised, rinsed to remove nonadhering inoculum, and processed for scintillation counting, Per cent bacterial adherence was calculated as: ([cpm adhering to the bladder/(cpm/bacterium)]/total number of bacteria in the inoculum) x 100. Mean per cent adherence to bladder uroepithelium for each E. coli isolate was calculated from a group of at least five animals. Low, intermediate, and high in vivo adherence were defined as mean adherence percentages of 0-1 per cent, 1-3 per cent, and 3-5 per Cent, >'P,m~,r+,,,cp
1319
1320
HOPKINS AND ASSOCIATES TABLE
1. Adhesin type, bladder adherence and hemagglutinating activity of twenty-eight E. coli clinical'isolates Type of Fimbriae*
Adherence Group
No. Isolates
High Intermediate Low
6 16 6
1
p
1 andP
Neither 1 orP
Per Cent Bacterial Adherence to Rat Bladder Uroepithelium Mean (Range)
3
0 0 0
2 1 2
1 4 1
3.9 (3.2-4.6) 2.0 (1.1-3.0) 0.5 (0.2-0. 7)
11
3
Minimum Number of Bacteria Required to Agglutinate RBC Without D-Mannose Mean (Range) 7.0 X 106 (6 X 106-1.2 X 107) 2.9 X 107 (1.2 X 107-5 X 107) 5 X 107 (5 X 107-5 X 107 )
* No. of isolates having Type 1 only, P only, or Type 1 and P fimbriae.
E
RESULTS
Among the patients with known urologic status, there appeared to be no correlation between the seriousness of the infection and the adherence assays. However, this patient group was small, and additional isolates would be needed to verify this observation. The results of fimbriae typing, rat bladder adherence assay, and hemagglutination titrations are presented in table 1. The majority of isolates (16/28) showed adherence in the mid-range, while both the high and low adherence groups each contained six clinical strains. Seventy-five per cent (21/28) of the E. coli strains tested possessed Type 1 fimbriae, either as the only detectable fimbrial type or in conjunction with P fimbriae. No isolates were found to possess only P fimbriae; however, four of the 28 strains did express P and Type 1 fimbriae. It was also noted that 25 per cent (7 /28) of the isolates did not express either Type 1 or P fimbriae, but some of these did adhere well to rat bladder and also agglutinated RBC. The expression of a particular fimbrial type was not restricted to any one of the adherence groups since some E. coli strains in each group expressed Type 1 alone or both Type 1 and P fimbriae. However, Type 1 fimbriae generally appeared to be the predominant fimbrial type in each adherence group. In comparing the in vitro to the in vivo assays (table 1) E. coli strains with the highest HA titer also showed the most adherence in vivo. Figure 1 shows a correlation of bacterial agglutination of GPRBC with bacterial adherence to the rat bladder mucosa. Statistical analysis of these data shows a significant correlation between increased bacterial HA and increased adherence in vivo (Spearman Rank correlation coefficient = 0.88, p = 0.0001; linear regression correlation coefficient = 0.84, p = 0.0001). DISCUSSION
Virulence factors for E. coli include specific combinations of O:K:H antigens, hemolysin production, serum bactericidal resistance, iron sequestration, and the presence of Type 1 and/ or P fimbriae. 4 • 16- 18 We examined the occurrence of both Type 1 and P fimbriae on E. coli strains isolated from patients with UTI. Type 1 fimbriae only were found on 61 per cent of the isolates, Type 1 and P were expressed jointly by 14 per cent, and 25 per cent of the E. coli strains did not express either fimbrial type. Our results are similar to the findings of Latham who studied the adhesins on E. coli isolated from adult women with cystitis and demonstrated either Type 1 or P fimbriae on 65 per cent of the isolates, Type 1 and Pon 14 per cent, and neither Type 1 nor Pon 21 per cent. 6 Both the study of Latham and ours indicate that the broad spectrum of UTis are caused most often by E. coli possessing Type 1 fimbriae only or by those strains expressing both types. Our results are somewhat different from those of Domingue who reported that 66 per cent of E. coli isolated from patients with uncomplicated cys titis were P-fimbriated. 18 Differences in patient selection, as to the type of patient studied, are the probable reasons for these differences. Further elucidation of the role of these adhesins as virulence factors will require their correlations in patients whose type of urinary tract infection is carefully categorized by appropriate localization tests. As another aspect of our present study on bacterial adherence, we examined the biological activity of adhesins present 0
:::, 5.0 Q)
e£
-=·a. 4.0 :::, Q)
oO
.:=: 0
~
3.0
••• • • •
••
-Q)
-~-0
••• • • ••
•••
~-o
~.5:! 2.0 om
•• •• m,e 1.0 -c,, • ••• 0 •• ~-.: .c 0.0 ......._ _........._ __,__ __,__ __,__ ___ 0
C:
Q)
-o
5
15
25
35
45
55
Minimum Number of Bacteria (xl06 ) Required to Agglutinate GPRBC in vitro
Fm. 1. Correlation of in vivo adherence to rat uroepithelium and in vitro agglutination of guinea pig red blood cells (GPRBC).
on each isolate by measuring each isolate's ability to adhere to the rat bladder and its efficiency in agglutinating GPRBC. Our data indicated that different E. coli strains varied in both their adherence to rat bladder mucosa in vivo and their capacity to agglutinate GPRBC in vitro. Those isolates which bound most avidly to rat uroepithelial cells also most efficiently agglutinated RBC. Since previous investigators have demonstrated a correlation between degree of piliation and hemagglutinating capacity of E. coli2• 19 we might speculate that for our E. coli isolates, the degree of Type 1 fimbriation might be as much a virulence factor as the presence of the fimbriae themselves. We must also address our observations that some isolates found in all three bladder adherence groups were not positive for either Type 1 or P fimbriae, yet attached well to GPRBC surfaces and rat bladder mucosa. One possible explanation is that these isolates possess adhesins other than the Type 1 and P for which we have tested. The recent investigations of Korhonen et al. 9• 10 on E. coli adherence mediated by S. fimbriae and the various types of E. coli fimbrial adhesins reviewed by Klemm strongly suggest that adherence to epithelial cells can be attributed to fimbriae other than the more extensively studied Type 1 and P. 20 Additionally, mutants of E. coli which completely lacked the ability to synthesize the major P fimbria subunit protein retained their binding specificity for a human bladder carcinoma cell line. 21 It is therefore conceivable that our Type 1and P-negative isolates, as well as those E. coli strains which e:l!,:pressed these fimbriae, can bind to cell surfaces using one or more of these other adhesins. In vivo as well as in vitro adherence assays can contribute to studies of bacterial adherence as it relates to the pathogenesis of and the immune response to urinary tract infections. Specifically, the rat model permits a quantitative examination of the adherence of uropathogenic E. coli to intact, living, mammalian urinary tract mucosal surfaces. While several studies have reported that the virulence of uropathogenic E. coli in humans
ADI-IESINS OF E. COLI
is most correlated with the presence of P fimbriae,rn, 22 and since it not been conclusively demonstrated that rat uroepithelial cells possess receptors for P fimbriae, it is possible to question the validity of using a rat model to study human uropathogens, In response to this objection, we can point out that in one study demonstrating that P-positive mutants did not adhere to rat uroepithelial cells, 23 chemically-derived mutants which may have any number of genes altered were used and lack of adherence was determined using voided, desquamated epithelial cells which may lack receptors for P fimbriae. These P receptors might be present on rat transitional cells or on viable uroepithelial cells of the intact bladder. We must also stress our present observations that not all human uropathogenic E. coli possess P fimbriae and that E. coli can express a variety of adhesins. Thus, any in vivo or in vitro assay system capable of detecting bacterial adherence to epithelial cells should be utilized to more clearly define E. coli virulence factors. REFERENCES
10. 11.
12. 13.
14.
15.
16.
1. Svanborg Eden, C., Lidin Janson, G. and Lindberg, U.: Adhesive-
2.
3.
4.
5.
6. 7. 8. 9.
ness to urinary tract epithelial cells of fecal and urinary Escherichia coli isolates from patients with symptomatic urinary tract infections or asymptomatic bacteriuria of varying duration. J Urol., 122: 185, 1979. Schaeffer, A. J., Amundsen, S. K. and Schmidt, L. N.: Adherence of Escherichia coli to human urinary tract epithelial cells. Infect. Immun., 24: 275, 1979. Kiillenius, G., Mollby, R., Hultberg, H., Svenson, 8., Cedergren, B. and Winberg, J.: Structure of carbohydrate part of receptor on human uroepithelial cells for pyelonephritogenic Escherichia coli. Lancet, 2: 604, 1981. Orskov, I., Orskov, F., Birch-Andersen, A., Kanamori, M. and Svanborg-Eden, C.: 0, K, Hand fimbrial antigens in Escherichia coli serotypes associated with pyelonephritis and cystitis. Scand. J. Infect. Dis (Supplement), 33: 18, 1982. Kallenius, G., Mollby, R., Svenson, S. B., Winberg, J., Lundblad, A., Svensson, S. and Cedergren, B.: The pk antigen as receptor for the haemagglutinin of pyelonephritic Escherichia coli. FEMS Microbiol. Lett., 7: 297, 1980. Latham, R. H. and Stamm, W. E.: Role of fimbriated Escherichia coli in urinary tract infections in adult women: correlation with localization studies. J. Infect. Dis., 149: 835, 1984. Robert, J. A.: Urinary tract infections. Am J. Kidney Dis., 4: 103, 1984. Ofek, I. and Beachey, I.: Mannose binding and epithelial cell adherence of Escherichia coli. Infect. Immun., 22: 247, 1978. Parkkinen, J., Finne, J., Achtman, M., Viiisiinen, V. and Korhonen,
17.
18.
19.
20. 21.
22.
23.
1321
T. K.: Eschericia coli strains binding neurarniny! a2-3 gaiactosides. Biochem. Biophys. Res. Comm., 11 L 456, 1983. Korhonen, T. K., Viiisanen-Rhenm V, Rhen, M., Pere, A., Parkkinen, J. and Finne, J. Escherichia coli fimbriae recognizing sialyl galactosides. J. Bact., 159: 762, 1984. Uehling, D. T., Mizutani, K. and Balish, E.: Effect of immunization on bacterial adherence to urothelium. Invest. Urol., 16: 145, 1978. Uehling, D. T., Jensen, J. and Balish, E.: Vaginal immunization against urinary tract infection. J. Urol., 128: 1382, 1982. Evans, D. G., Evans, D. J. and Tjoa, W.: Hemagglutination of human group A erythrocytes by enterotoxigenic Escherichia coli isolated from adults with diarrhea: correlation with colonization factor. Infect. Immun., 18: 330, 1977. Svenson, S. B., Kallenius, G., Mollby, R., Hultbere, H. and Winberg, J.: Rapid identification of P-fimbriated Escherichia coli by a receptor-specific particle agglutination test. Infection, l 0: 209, 1982. Miillby, R., Kiillenius, G., Korhonen, T. K., Winberg, J. and Svenson, S. B.: P-fimbriae of pyelonephritogenic Eschericia coli: detection in clinical material by a rapid receptor-specific agglutination test. Infection, 11: 68, 1983. Welch, R. A. and Falkow, S.: Characterization of Escherichia coli hemolysins conferring quantitative differences in virulence. Infect. Immun., 43: 156, 1984. Lomberg, H., Hellstrom, M., Jodal, U., Leffler, H., Lincoln, K. and Svanborg Eden, C. Virulence-associated traits in Escherichia coli causing first and recurrent episodes of urinary tract infection in children with or without vesicoureteral reflux. J. Infect. Dis., 150: 561, 1984. Domingue, G. J., Roberts, J., Laucirica, R., Ratner, M., Bell, D., Suarez, G., Kiillenius, G. and Svenson, S. Pathogenic significance of P-fimbriated Escherichia coli in urinary tract infections. J. Urol., 133: 983, 1985. Van den Bosch, J. F., Verbood-Sohmer, U., Postma, P., DeGraaf, J. and Macharen, D. M.: Mannose-sensitive and mannose-resistant adherence to human uroepithelial cells and urinary virulence of Escherichia coli. Infect. Immun., 29: 226, 1980. Klemm, P.: Fimbrial adhesins of Escherichia coli. Rev. Infect. Dis., 7: 321, 1985. Uhlin, B. E., Norgren, M., Baga, M. and Normark, 8.: Adhesins to human cells by Escherichia coli lacking the major subunit of a digalactoside-specific pilus-adhesin. Proc. Natl. Acad. Sci. USA, 82: 1800, 1985. Svenson, S. B., Kiillenius, G., Korhenon, T., Mollby, R., Roberts, J. A., Tullus, K. and Winberg, J.: Initiation of clinical pyelonephritis: the role of P-fimbriae-mediated bacterial adhesion. Contr. Nephrol., 39: 252, 1984. Hagberg, L., Hull, R., Hull, S., Falkow, S., Freter, R. and SvanborgEden, C.: Contribution of adhesion to bacterial persistence in the mouse urinary tract. Infect. Immun., 40: 265, 1983.
THE
Vol. 135, June Printed in U.S.A.
JOURNAL OF UROLOGY
Copyright © 1986 by The Williams & Wilkins Co.
IN VITRO AND IN VIVO ADHERENCE OF UROPATHOGENIC ESCHERICHIA COLI STRAINS WALTER J. HOPKINS, JEAN L. JENSEN, DAVID T. UEHLING*
AND
EDWARD BALISH
From the Department of Surgery (Urology), University of Wisconsin, Madison, Wisconsin
ABSTRACT
Twenty-eight Escherichia coli strains isolated from the urine of patients with urinary tract infections were assayed for fimbrial type, in vitro capacity to agglutinate guinea red blood cells, and in vivo adherence to rat bladder uroepithelium. A direct correlation was found between hemagglutinating and in vivo adherence. Strains with both Type 1 and P fimbriae showed the greatest adherence in vivo. Of the 28 strains, seven did not manifest either Type l or P fimbriae but agglutinated red blood cells and did adhere in vivo. In studies on bacterial adherence and urinary tract both in vivo and in vitro studies may contribute to understanding the relevance of bacterial adhesins in initiating urinary tract infections. Adherence of bacteria to mucosal cell surfaces is thought to be the initial step in many types of bacterial infections. The virulence of E. coli for the urinary tract is thought to be directly related to the presence of specific cell-surface structures, or "adhesins," which mediate bacterial adherence to epithelial cells. 1- 3 Among the surface adhesins that have been identified, fimbriae are thought to be particularly important in the binding of bacteria to epithelial cell membranes.4- 7 Several different types of fimbriae have been identified and characterized on the basis of structure, antigenic type, and receptor specificity. For example, Type 1, S, and P fimbriae are morphologically identical, but differ both antigenically and in their attachment sites on cell surfaces. Type l fimbriae bind to mannose-like receptors on the surface of epithelial cells and guinea pig red blood cells (GPRBC). 2 • 5 · 8 P fimbriae attach to carbohydrate moieties which are similar to human P blood group antigens,5 and S fimbriae recognize sialyl galactosides. 9 • 10 Thus, bacteria which possess these fimbriae, either singly, or in combination, can adhere avidly to mucosa! cell surfaces and agglutinate either red blood cells or particles coated with appropriate fimbrial receptors. Much of the current research on urinary tract infections (UTI) has focused on the adherence process and how it can be quantified. Many of these studies have examined adherence using in vitro systems in which bacteria bind to desquamated urinary epithelial cells, buccal epithelial cells, or mammalian red blood cells.'· 2 We have previously reported on bacterial adherence to the rat bladder in vivo and believe that this model has some advantages over in vitro methods since it measures adherence under physiologic conditions."· 12 In this report, we utilized both in vivo and in vitro systems to compare the adherence properties of E. coli strains isolated from patients with UTI. MATERIALS AND METHODS
Twenty-eight Escherichia coli strains were isolated from the urine of patients with urinary tract infections (UTI) at the University of Wisconsin Center for Health Sciences. Ten of the E. coli isolates came from patients being evaluated by the urology service. Based on the available evaluation, the patients' infections were graded as to seriousness based on abnormalities seen on intravenous pyelogram, voiding cystourethrogram and other diagnostic tests. The site and severity of the other patients' UTis were not known because + _,••• " , ••• of their medical Accepted for publication January 13, 1986. * for reprints: Dept. of Surgery, 600 Highland Ave., Madison Supported by TJ. S. Public I-iealtD. Service grant J\J\130808.
records. The 28 E. coli strains were typed for fimbriae, tested for their capacity to agglutinate RBC, and to adhere in vivo to rat bladder uroepithelium. For determination of hemagglutination (HA) titers, bacterial isolates were subcultured in Mueller-Hinton broth (Difeo Laboratories) or on colonization factor antigen (CF A) agar 13 plates and resuspended in 0.05 M phosphate buffered saline (PBS), pH 7.0 at a concentration of 5 x 108 bacteria/ml. Two-fold serial dilutions of the bacteria were made in PBS and to each dilution was added an equal volume of GPRBC resuspended at 3 per cent (v/v) in PBS. The dilutions were agitated for one minute. HA titers were assessed at one hour after incubation at room temperature and then again after 18 hours incubation at 4C. The HA titers are expressed as the lowest number of bacteria required to produce visible hemagglutination. 5 The E. coli strains were tested for the presence of both type 1 and P fimbriae. To assay for Type 1 fimbriae, each isolate was passaged three times in tryptose broth; each passage was for 48 hours at 37C without agitation. One drop of broth culture was then mixed with either a drop of 5 per cent GPRBC or 5 per cent GPRBC and 5 per cent D-mannose on a glass microscope slide. Agglutination of GPRBC which was inhibited by D-mannose indicated the presence of Type 1 fimbriae on the E. coli isolate. P fimbriae were assayed by the procedure of Svenson et al. 14 Briefly, a sample of each isolate was taken following the third passage in tryptose broth (Difeo Laboratories) and grown overnight at 37C on CF A agar plates. Several colonies from the CF A agar were then combined with 20 µl. of PBS on a microscope slide. The P particle agglutination (PP A) test was used for determination of P fimbriae by adding 20 µI. of the test or control solutions to the bacterial suspension. 15 Agglutination of the test solution but not the control solution indicated the presence of P fimbriae. Bacterial adherence in vivo was measured by previously described methods. 11 Briefly, 3H-labelled bacteria were inoculated into the rat bladder. The inoculum was allowed to remain in the bladder for two hours, after which the bladders were excised, rinsed to remove nonadhering inoculum, and processed for scintillation counting, Per cent bacterial adherence was calculated as: ([cpm adhering to the bladder/(cpm/bacterium)]/total number of bacteria in the inoculum) x 100. Mean per cent adherence to bladder uroepithelium for each E. coli isolate was calculated from a group of at least five animals. Low, intermediate, and high in vivo adherence were defined as mean adherence percentages of 0-1 per cent, 1-3 per cent, and 3-5 per Cent, >'P,m~,r+,,,cp
1319
1320
HOPKINS AND ASSOCIATES TABLE
1. Adhesin type, bladder adherence and hemagglutinating activity of twenty-eight E. coli clinical'isolates Type of Fimbriae*
Adherence Group
No. Isolates
High Intermediate Low
6 16 6
1
p
1 andP
Neither 1 orP
Per Cent Bacterial Adherence to Rat Bladder Uroepithelium Mean (Range)
3
0 0 0
2 1 2
1 4 1
3.9 (3.2-4.6) 2.0 (1.1-3.0) 0.5 (0.2-0. 7)
11
3
Minimum Number of Bacteria Required to Agglutinate RBC Without D-Mannose Mean (Range) 7.0 X 106 (6 X 106-1.2 X 107) 2.9 X 107 (1.2 X 107-5 X 107) 5 X 107 (5 X 107-5 X 107 )
* No. of isolates having Type 1 only, P only, or Type 1 and P fimbriae.
E
RESULTS
Among the patients with known urologic status, there appeared to be no correlation between the seriousness of the infection and the adherence assays. However, this patient group was small, and additional isolates would be needed to verify this observation. The results of fimbriae typing, rat bladder adherence assay, and hemagglutination titrations are presented in table 1. The majority of isolates (16/28) showed adherence in the mid-range, while both the high and low adherence groups each contained six clinical strains. Seventy-five per cent (21/28) of the E. coli strains tested possessed Type 1 fimbriae, either as the only detectable fimbrial type or in conjunction with P fimbriae. No isolates were found to possess only P fimbriae; however, four of the 28 strains did express P and Type 1 fimbriae. It was also noted that 25 per cent (7 /28) of the isolates did not express either Type 1 or P fimbriae, but some of these did adhere well to rat bladder and also agglutinated RBC. The expression of a particular fimbrial type was not restricted to any one of the adherence groups since some E. coli strains in each group expressed Type 1 alone or both Type 1 and P fimbriae. However, Type 1 fimbriae generally appeared to be the predominant fimbrial type in each adherence group. In comparing the in vitro to the in vivo assays (table 1) E. coli strains with the highest HA titer also showed the most adherence in vivo. Figure 1 shows a correlation of bacterial agglutination of GPRBC with bacterial adherence to the rat bladder mucosa. Statistical analysis of these data shows a significant correlation between increased bacterial HA and increased adherence in vivo (Spearman Rank correlation coefficient = 0.88, p = 0.0001; linear regression correlation coefficient = 0.84, p = 0.0001). DISCUSSION
Virulence factors for E. coli include specific combinations of O:K:H antigens, hemolysin production, serum bactericidal resistance, iron sequestration, and the presence of Type 1 and/ or P fimbriae. 4 • 16- 18 We examined the occurrence of both Type 1 and P fimbriae on E. coli strains isolated from patients with UTI. Type 1 fimbriae only were found on 61 per cent of the isolates, Type 1 and P were expressed jointly by 14 per cent, and 25 per cent of the E. coli strains did not express either fimbrial type. Our results are similar to the findings of Latham who studied the adhesins on E. coli isolated from adult women with cystitis and demonstrated either Type 1 or P fimbriae on 65 per cent of the isolates, Type 1 and Pon 14 per cent, and neither Type 1 nor Pon 21 per cent. 6 Both the study of Latham and ours indicate that the broad spectrum of UTis are caused most often by E. coli possessing Type 1 fimbriae only or by those strains expressing both types. Our results are somewhat different from those of Domingue who reported that 66 per cent of E. coli isolated from patients with uncomplicated cys titis were P-fimbriated. 18 Differences in patient selection, as to the type of patient studied, are the probable reasons for these differences. Further elucidation of the role of these adhesins as virulence factors will require their correlations in patients whose type of urinary tract infection is carefully categorized by appropriate localization tests. As another aspect of our present study on bacterial adherence, we examined the biological activity of adhesins present 0
:::, 5.0 Q)
e£
-=·a. 4.0 :::, Q)
oO
.:=: 0
~
3.0
••• • • •
••
-Q)
-~-0
••• • • ••
•••
~-o
~.5:! 2.0 om
•• •• m,e 1.0 -c,, • ••• 0 •• ~-.: .c 0.0 ......._ _........._ __,__ __,__ __,__ ___ 0
C:
Q)
-o
5
15
25
35
45
55
Minimum Number of Bacteria (xl06 ) Required to Agglutinate GPRBC in vitro
Fm. 1. Correlation of in vivo adherence to rat uroepithelium and in vitro agglutination of guinea pig red blood cells (GPRBC).
on each isolate by measuring each isolate's ability to adhere to the rat bladder and its efficiency in agglutinating GPRBC. Our data indicated that different E. coli strains varied in both their adherence to rat bladder mucosa in vivo and their capacity to agglutinate GPRBC in vitro. Those isolates which bound most avidly to rat uroepithelial cells also most efficiently agglutinated RBC. Since previous investigators have demonstrated a correlation between degree of piliation and hemagglutinating capacity of E. coli2• 19 we might speculate that for our E. coli isolates, the degree of Type 1 fimbriation might be as much a virulence factor as the presence of the fimbriae themselves. We must also address our observations that some isolates found in all three bladder adherence groups were not positive for either Type 1 or P fimbriae, yet attached well to GPRBC surfaces and rat bladder mucosa. One possible explanation is that these isolates possess adhesins other than the Type 1 and P for which we have tested. The recent investigations of Korhonen et al. 9• 10 on E. coli adherence mediated by S. fimbriae and the various types of E. coli fimbrial adhesins reviewed by Klemm strongly suggest that adherence to epithelial cells can be attributed to fimbriae other than the more extensively studied Type 1 and P. 20 Additionally, mutants of E. coli which completely lacked the ability to synthesize the major P fimbria subunit protein retained their binding specificity for a human bladder carcinoma cell line. 21 It is therefore conceivable that our Type 1and P-negative isolates, as well as those E. coli strains which e:l!,:pressed these fimbriae, can bind to cell surfaces using one or more of these other adhesins. In vivo as well as in vitro adherence assays can contribute to studies of bacterial adherence as it relates to the pathogenesis of and the immune response to urinary tract infections. Specifically, the rat model permits a quantitative examination of the adherence of uropathogenic E. coli to intact, living, mammalian urinary tract mucosal surfaces. While several studies have reported that the virulence of uropathogenic E. coli in humans
ADI-IESINS OF E. COLI
is most correlated with the presence of P fimbriae,rn, 22 and since it not been conclusively demonstrated that rat uroepithelial cells possess receptors for P fimbriae, it is possible to question the validity of using a rat model to study human uropathogens, In response to this objection, we can point out that in one study demonstrating that P-positive mutants did not adhere to rat uroepithelial cells, 23 chemically-derived mutants which may have any number of genes altered were used and lack of adherence was determined using voided, desquamated epithelial cells which may lack receptors for P fimbriae. These P receptors might be present on rat transitional cells or on viable uroepithelial cells of the intact bladder. We must also stress our present observations that not all human uropathogenic E. coli possess P fimbriae and that E. coli can express a variety of adhesins. Thus, any in vivo or in vitro assay system capable of detecting bacterial adherence to epithelial cells should be utilized to more clearly define E. coli virulence factors. REFERENCES
10. 11.
12. 13.
14.
15.
16.
1. Svanborg Eden, C., Lidin Janson, G. and Lindberg, U.: Adhesive-
2.
3.
4.
5.
6. 7. 8. 9.
ness to urinary tract epithelial cells of fecal and urinary Escherichia coli isolates from patients with symptomatic urinary tract infections or asymptomatic bacteriuria of varying duration. J Urol., 122: 185, 1979. Schaeffer, A. J., Amundsen, S. K. and Schmidt, L. N.: Adherence of Escherichia coli to human urinary tract epithelial cells. Infect. Immun., 24: 275, 1979. Kiillenius, G., Mollby, R., Hultberg, H., Svenson, 8., Cedergren, B. and Winberg, J.: Structure of carbohydrate part of receptor on human uroepithelial cells for pyelonephritogenic Escherichia coli. Lancet, 2: 604, 1981. Orskov, I., Orskov, F., Birch-Andersen, A., Kanamori, M. and Svanborg-Eden, C.: 0, K, Hand fimbrial antigens in Escherichia coli serotypes associated with pyelonephritis and cystitis. Scand. J. Infect. Dis (Supplement), 33: 18, 1982. Kallenius, G., Mollby, R., Svenson, S. B., Winberg, J., Lundblad, A., Svensson, S. and Cedergren, B.: The pk antigen as receptor for the haemagglutinin of pyelonephritic Escherichia coli. FEMS Microbiol. Lett., 7: 297, 1980. Latham, R. H. and Stamm, W. E.: Role of fimbriated Escherichia coli in urinary tract infections in adult women: correlation with localization studies. J. Infect. Dis., 149: 835, 1984. Robert, J. A.: Urinary tract infections. Am J. Kidney Dis., 4: 103, 1984. Ofek, I. and Beachey, I.: Mannose binding and epithelial cell adherence of Escherichia coli. Infect. Immun., 22: 247, 1978. Parkkinen, J., Finne, J., Achtman, M., Viiisiinen, V. and Korhonen,
17.
18.
19.
20. 21.
22.
23.
1321
T. K.: Eschericia coli strains binding neurarniny! a2-3 gaiactosides. Biochem. Biophys. Res. Comm., 11 L 456, 1983. Korhonen, T. K., Viiisanen-Rhenm V, Rhen, M., Pere, A., Parkkinen, J. and Finne, J. Escherichia coli fimbriae recognizing sialyl galactosides. J. Bact., 159: 762, 1984. Uehling, D. T., Mizutani, K. and Balish, E.: Effect of immunization on bacterial adherence to urothelium. Invest. Urol., 16: 145, 1978. Uehling, D. T., Jensen, J. and Balish, E.: Vaginal immunization against urinary tract infection. J. Urol., 128: 1382, 1982. Evans, D. G., Evans, D. J. and Tjoa, W.: Hemagglutination of human group A erythrocytes by enterotoxigenic Escherichia coli isolated from adults with diarrhea: correlation with colonization factor. Infect. Immun., 18: 330, 1977. Svenson, S. B., Kallenius, G., Mollby, R., Hultbere, H. and Winberg, J.: Rapid identification of P-fimbriated Escherichia coli by a receptor-specific particle agglutination test. Infection, l 0: 209, 1982. Miillby, R., Kiillenius, G., Korhonen, T. K., Winberg, J. and Svenson, S. B.: P-fimbriae of pyelonephritogenic Eschericia coli: detection in clinical material by a rapid receptor-specific agglutination test. Infection, 11: 68, 1983. Welch, R. A. and Falkow, S.: Characterization of Escherichia coli hemolysins conferring quantitative differences in virulence. Infect. Immun., 43: 156, 1984. Lomberg, H., Hellstrom, M., Jodal, U., Leffler, H., Lincoln, K. and Svanborg Eden, C. Virulence-associated traits in Escherichia coli causing first and recurrent episodes of urinary tract infection in children with or without vesicoureteral reflux. J. Infect. Dis., 150: 561, 1984. Domingue, G. J., Roberts, J., Laucirica, R., Ratner, M., Bell, D., Suarez, G., Kiillenius, G. and Svenson, S. Pathogenic significance of P-fimbriated Escherichia coli in urinary tract infections. J. Urol., 133: 983, 1985. Van den Bosch, J. F., Verbood-Sohmer, U., Postma, P., DeGraaf, J. and Macharen, D. M.: Mannose-sensitive and mannose-resistant adherence to human uroepithelial cells and urinary virulence of Escherichia coli. Infect. Immun., 29: 226, 1980. Klemm, P.: Fimbrial adhesins of Escherichia coli. Rev. Infect. Dis., 7: 321, 1985. Uhlin, B. E., Norgren, M., Baga, M. and Normark, 8.: Adhesins to human cells by Escherichia coli lacking the major subunit of a digalactoside-specific pilus-adhesin. Proc. Natl. Acad. Sci. USA, 82: 1800, 1985. Svenson, S. B., Kiillenius, G., Korhenon, T., Mollby, R., Roberts, J. A., Tullus, K. and Winberg, J.: Initiation of clinical pyelonephritis: the role of P-fimbriae-mediated bacterial adhesion. Contr. Nephrol., 39: 252, 1984. Hagberg, L., Hull, R., Hull, S., Falkow, S., Freter, R. and SvanborgEden, C.: Contribution of adhesion to bacterial persistence in the mouse urinary tract. Infect. Immun., 40: 265, 1983.