REVIEWS OF INFECTIOUS DISEASES. VOL. I, NO.5 .• SEPTEMBER-OCTOBER 1979 © 1979 by The University of Chicago. 0162-0886/79/0105-0012$00.75
Effects of Subminimal Inhibitory Concentrations of Ampicillin, Chloramphenicol, and Nitrofurantoin on the Attachment of Escherichia coli to Human Uroepithelial Cells in Vitro Torsten Sandberg, Karin Stenqvist, and Catharina Svanborg-Eden
From the Departments of Clinical Immunology and Infectious Diseases, University of G6teborg, G6teborg, Sweden
An in vitro system with epithelial cells of the human urinary tract and Escherichia coli isolatedJrorn the urine of patients with acute pyelonephritis was used for tests of the effect of antibacterial agents on bacterial adhesion. E. coli treated with subminimal inhibitory amounts of ampicillin attached less than untreated control bacteria. Subminimal inhibitory amounts of chloramphenicol and nitrofurantoin had no effect on adhesion of the 15 strains of E. coli tested. The combination of bacterial pretreatment with ampicillin and addition of antibodies to the adhesion test tube decreased adhesion more than did either treatment separately.
Urinary tract infection (UTI) is usually caused by Escherichia coli that originate in the fecal flora and enter the urinary tract by the ascending route [1]. By attaching to the epithelial surfaces, bacteria avoid being washed away by the continuous flow of urine [2] and increase the possibility of their invading underlying tissues. By means of an in vitro test system employing epithelial cells of the human urinary tract and E. coli isolated from the urine of patients with various forms of UTI, the virulence of E. coli was found to correlate with adhesive capacity [3, 4]. Thus, a larger proportion of attaching E. coli strains was isolated from patients with acute pyelonephritis and acute cystitis than from patients with asymptomatic bacteriuria detected at screening. The possible effects· of antibacterial agents on bacterial attachment have not been evaluated fully. In preliminary reports we observed decreased adhesion of E. coli treated with subminimal inhibitory concentrations (sub-MICs) of ampicillin [5] and amoxicillin [6]. The present·study confirms
the findings with ampicillin for a larger number of E. coli strains but fails to demonstrate lowered adhesive capacity for E. coli treated with subMICs of either chloramphenicol or nitrofurantoin. The combination of bacterial pretreatment with ampicillin and addition of antibodies to the adhesion test system decreased attachment more than did either treatment alone. Materials and Methods
Bacteria. Fifteen strains of E. coli isolated from the urine of 15 patients with acute pyelonephritis and known to adhere well [3] were used. For testing, the strains were transferred from deep agar storage cultures to lactose-bromothymol blue agar plates. Determination ofMICs. Ampicillin, chloramphenicol base, and nitrofurantoin were dissolved in antibiotic-sensitive medium (ASM; Biodisc, Sweden). The MIC of each antibiotic for each bacterial strain was determined by serially diluting each drug twofold in ASM broth and incubating each dilution with rv 105 organisms from a 4-hr broth culture. The MIC was read visually after incubation for 24 hr at 37 C. Treatment of bacteria with antibiotics. Parts of one bacterial colony taken from a lactose-bromothymol blue agar plate were used for inoculation of 3-ml of ASM broth in tubes (rv 10 6 bacteria/ml). The tubes were incubated at 37 C without shaking for 2 hr, at which point the bacteria were in the logarithmic growth phase (figure
This study was supported by grants from the Medical Faculty, University of Goteborg, and the Swedish Medical Research Council (project no. 215); by the Ellen, Walter, and Lennart Hesselman Foundation for Scientific Research; and by Astra Uikemedel, Sweden. We thank Ann-Marie Jonsson and Ingrid Holmberg for technical assistance. Please address requests for reprints to Dr. C. SvanborgEden, Department of Immunology, Institute of Medical Microbiology, Guldhedsgatan 10, 413 46 Goteborg, Sweden.
838
E. coli Attachment to Uroepithelium
839
Bacterial numbers
o
_--CJ',
- - - , . - - - - - - - - - - r - _ l hours
r-j
o
2 t
6
Figure 1. Growth curves for Escherichia coli strain 7, either untreated (0) or treated with one-fourth the MIC of chloramphenicol (0), nitrofurantoin (~), or ampicillin (*). Antibiotics were added after growth of the bacteria in test tubes for 2 hr.
1). One of the three drugs being tested was added to the test cultures to a final concentration of onefourth the MIC, and ASM broth free of antibiotic was added to the control cultures. Incubation was then continued for 4 hr. After growth for a total of 6 hr, the bacteria were harvested by centrifugation, resuspended in phosphate-buffered saline (PBS; pH 7.1, 300 mOsmollliter), and quantitated by direct light microscopy with use of a BUrker chamber (Kebo Grave, Goteborg, Sweden). A volume of the bacterial suspension to give a final concentration of 10 8 bacteria/ml was then added to the epithelial cells for adhesion testing. The adhesive capacity of each strain, treated or untreated, was thus compared in each experiment. Adhesion testing. Adhesion testing was performed as previously described [7]. Epithelial cells of the human urinary tract were obtained from the sediment of fresh urine from a nonbacteriuric woman. The cells were washed, resuspended in PBS, and quantitated by direct light microscopy with use of a Burker chamber. To 108 bacteria were added 10 5 epithelial cells and PBS up to a volume of 1 ml. After incubation of bacteria and epithelial cells, the number of bacteria attached to the epithelial cells was counted under a light
microscope. Adhesion was expressed as the mean number of bacteria attached to 40 epithelial cells. In each experiment antibiotic-treated bacteria and identically handled controls exposed to drug-free diluent were compared. Viable counts. Counts of viable bacteria were performed on bacterial cultures after incubation for 0, 2, and 6 hr and on the mixture of bacteria and epithelial cells used for adhesion testing. Antibiotics and antibodies. Bacterial cultures either treated or untreated with ampicillin were prepared as previously described in [5]. Commercial y-globulin (165 g/liter; Kabi, Sweden) was used as a source of IgG as previously described [8]. After preincubation of bacteria for 30 min with 0.1, 0.01, 0.001, and 0.0001 ml of y-globulin, epithelial cells and PBS were added to a total volume of 1 ml, and adhesion testing was completed as described above. Bacteria unexposed to ampicillin or antibodies were used as controls. Procedural controls. (1) The range of MICs of ampicillin for the 15 strains of E. coli tested was 1.5-12.5 ~g/ml; of chloramphenicol, 3.1-12.5 ~g/ml; and of nitrofurantoin, 2-30 1Jg/ml. (2) Growth curves for one of the E. coli strains before and after exposure to the antibiotics are shown in figure 1. At one-fourth the MIC, only ampicillin limited bacterial numbers. (3) The number of bacteria registered by viable counts in the adhesion test tube agreed well with the number counted by direct light microscopy. (4) Morphologic changes in the bacteria were observed. Most bacteria were elongated after treatment with ampicillin but only a few after treatment with onefourth the MIC of nitrofurantoin. No morphologic changes were observed after treatment with chloramphenicol. (5) Preincubation of epithelial cells with either of the antibacterial agents did not affect the number of bacteria that attached to the cells. Results
Effect of antibiotics on E. coli attachment. The attachment of the 15 strains of E. coli, with or without exposure to ampicillin, nitrofurantoin, and chloramphenicol, is shown in tables 1, 2, and 3, respectively. At one-fourth the MIC, treatment with ampicillin consistently decreased adhesive capacity (figure 2). No such effect was demon-
Sandberg, Stenqvist, and Svanborg-Eden
840
Table 1. Effect of pretreatment of bacteria with one-fourth the MIC of ampicillin on the attachment of 15 strains of Escherichia coli to human uroepithelial cells. No. of viable bacteria t
Attachment*
E. coli strain 1 2 3 4 5 6 7 8 9 10 11 12
13 14 15
Control
Ampicillint
Control
85 49 47 27 81 10 58 10 85 114 0 44 67 89 73
23
7 x 108 6 x 108 6 x 108 4 x 10 8 3 x 108 4 X 108 5 x 108 5 X 108 6 x 108 4 x 108 4 X 108 3 X 108 2 x 108 1 x 108 1 x 109
6 3 10 0 2 5 0 32 59 0 0 9 39 24
Ampicillin 7 4 2 6 1 5 8 2 3 4 3 3 3 4 7
X X X X X X X X X X X X X X X
108 108 10 8 108 10 8 108 108 108 108 108 10 8 10 8 108 108 108
NOTE. Each value represents the mean of two experiments. * Mean number of bacteria attached to 40 epithelial cells. t In the adhesion test tube. t p
Effects of Subminimal Inhibitory Concentrations of Ampicillin, Chloramphenicol, and Nitrofurantoin on the Attachment of Escherichia coli to Human Uroepithelial Cells in Vitro Torsten Sandberg, Karin Stenqvist, and Catharina Svanborg-Eden
From the Departments of Clinical Immunology and Infectious Diseases, University of G6teborg, G6teborg, Sweden
An in vitro system with epithelial cells of the human urinary tract and Escherichia coli isolatedJrorn the urine of patients with acute pyelonephritis was used for tests of the effect of antibacterial agents on bacterial adhesion. E. coli treated with subminimal inhibitory amounts of ampicillin attached less than untreated control bacteria. Subminimal inhibitory amounts of chloramphenicol and nitrofurantoin had no effect on adhesion of the 15 strains of E. coli tested. The combination of bacterial pretreatment with ampicillin and addition of antibodies to the adhesion test tube decreased adhesion more than did either treatment separately.
Urinary tract infection (UTI) is usually caused by Escherichia coli that originate in the fecal flora and enter the urinary tract by the ascending route [1]. By attaching to the epithelial surfaces, bacteria avoid being washed away by the continuous flow of urine [2] and increase the possibility of their invading underlying tissues. By means of an in vitro test system employing epithelial cells of the human urinary tract and E. coli isolated from the urine of patients with various forms of UTI, the virulence of E. coli was found to correlate with adhesive capacity [3, 4]. Thus, a larger proportion of attaching E. coli strains was isolated from patients with acute pyelonephritis and acute cystitis than from patients with asymptomatic bacteriuria detected at screening. The possible effects· of antibacterial agents on bacterial attachment have not been evaluated fully. In preliminary reports we observed decreased adhesion of E. coli treated with subminimal inhibitory concentrations (sub-MICs) of ampicillin [5] and amoxicillin [6]. The present·study confirms
the findings with ampicillin for a larger number of E. coli strains but fails to demonstrate lowered adhesive capacity for E. coli treated with subMICs of either chloramphenicol or nitrofurantoin. The combination of bacterial pretreatment with ampicillin and addition of antibodies to the adhesion test system decreased attachment more than did either treatment alone. Materials and Methods
Bacteria. Fifteen strains of E. coli isolated from the urine of 15 patients with acute pyelonephritis and known to adhere well [3] were used. For testing, the strains were transferred from deep agar storage cultures to lactose-bromothymol blue agar plates. Determination ofMICs. Ampicillin, chloramphenicol base, and nitrofurantoin were dissolved in antibiotic-sensitive medium (ASM; Biodisc, Sweden). The MIC of each antibiotic for each bacterial strain was determined by serially diluting each drug twofold in ASM broth and incubating each dilution with rv 105 organisms from a 4-hr broth culture. The MIC was read visually after incubation for 24 hr at 37 C. Treatment of bacteria with antibiotics. Parts of one bacterial colony taken from a lactose-bromothymol blue agar plate were used for inoculation of 3-ml of ASM broth in tubes (rv 10 6 bacteria/ml). The tubes were incubated at 37 C without shaking for 2 hr, at which point the bacteria were in the logarithmic growth phase (figure
This study was supported by grants from the Medical Faculty, University of Goteborg, and the Swedish Medical Research Council (project no. 215); by the Ellen, Walter, and Lennart Hesselman Foundation for Scientific Research; and by Astra Uikemedel, Sweden. We thank Ann-Marie Jonsson and Ingrid Holmberg for technical assistance. Please address requests for reprints to Dr. C. SvanborgEden, Department of Immunology, Institute of Medical Microbiology, Guldhedsgatan 10, 413 46 Goteborg, Sweden.
838
E. coli Attachment to Uroepithelium
839
Bacterial numbers
o
_--CJ',
- - - , . - - - - - - - - - - r - _ l hours
r-j
o
2 t
6
Figure 1. Growth curves for Escherichia coli strain 7, either untreated (0) or treated with one-fourth the MIC of chloramphenicol (0), nitrofurantoin (~), or ampicillin (*). Antibiotics were added after growth of the bacteria in test tubes for 2 hr.
1). One of the three drugs being tested was added to the test cultures to a final concentration of onefourth the MIC, and ASM broth free of antibiotic was added to the control cultures. Incubation was then continued for 4 hr. After growth for a total of 6 hr, the bacteria were harvested by centrifugation, resuspended in phosphate-buffered saline (PBS; pH 7.1, 300 mOsmollliter), and quantitated by direct light microscopy with use of a BUrker chamber (Kebo Grave, Goteborg, Sweden). A volume of the bacterial suspension to give a final concentration of 10 8 bacteria/ml was then added to the epithelial cells for adhesion testing. The adhesive capacity of each strain, treated or untreated, was thus compared in each experiment. Adhesion testing. Adhesion testing was performed as previously described [7]. Epithelial cells of the human urinary tract were obtained from the sediment of fresh urine from a nonbacteriuric woman. The cells were washed, resuspended in PBS, and quantitated by direct light microscopy with use of a Burker chamber. To 108 bacteria were added 10 5 epithelial cells and PBS up to a volume of 1 ml. After incubation of bacteria and epithelial cells, the number of bacteria attached to the epithelial cells was counted under a light
microscope. Adhesion was expressed as the mean number of bacteria attached to 40 epithelial cells. In each experiment antibiotic-treated bacteria and identically handled controls exposed to drug-free diluent were compared. Viable counts. Counts of viable bacteria were performed on bacterial cultures after incubation for 0, 2, and 6 hr and on the mixture of bacteria and epithelial cells used for adhesion testing. Antibiotics and antibodies. Bacterial cultures either treated or untreated with ampicillin were prepared as previously described in [5]. Commercial y-globulin (165 g/liter; Kabi, Sweden) was used as a source of IgG as previously described [8]. After preincubation of bacteria for 30 min with 0.1, 0.01, 0.001, and 0.0001 ml of y-globulin, epithelial cells and PBS were added to a total volume of 1 ml, and adhesion testing was completed as described above. Bacteria unexposed to ampicillin or antibodies were used as controls. Procedural controls. (1) The range of MICs of ampicillin for the 15 strains of E. coli tested was 1.5-12.5 ~g/ml; of chloramphenicol, 3.1-12.5 ~g/ml; and of nitrofurantoin, 2-30 1Jg/ml. (2) Growth curves for one of the E. coli strains before and after exposure to the antibiotics are shown in figure 1. At one-fourth the MIC, only ampicillin limited bacterial numbers. (3) The number of bacteria registered by viable counts in the adhesion test tube agreed well with the number counted by direct light microscopy. (4) Morphologic changes in the bacteria were observed. Most bacteria were elongated after treatment with ampicillin but only a few after treatment with onefourth the MIC of nitrofurantoin. No morphologic changes were observed after treatment with chloramphenicol. (5) Preincubation of epithelial cells with either of the antibacterial agents did not affect the number of bacteria that attached to the cells. Results
Effect of antibiotics on E. coli attachment. The attachment of the 15 strains of E. coli, with or without exposure to ampicillin, nitrofurantoin, and chloramphenicol, is shown in tables 1, 2, and 3, respectively. At one-fourth the MIC, treatment with ampicillin consistently decreased adhesive capacity (figure 2). No such effect was demon-
Sandberg, Stenqvist, and Svanborg-Eden
840
Table 1. Effect of pretreatment of bacteria with one-fourth the MIC of ampicillin on the attachment of 15 strains of Escherichia coli to human uroepithelial cells. No. of viable bacteria t
Attachment*
E. coli strain 1 2 3 4 5 6 7 8 9 10 11 12
13 14 15
Control
Ampicillint
Control
85 49 47 27 81 10 58 10 85 114 0 44 67 89 73
23
7 x 108 6 x 108 6 x 108 4 x 10 8 3 x 108 4 X 108 5 x 108 5 X 108 6 x 108 4 x 108 4 X 108 3 X 108 2 x 108 1 x 108 1 x 109
6 3 10 0 2 5 0 32 59 0 0 9 39 24
Ampicillin 7 4 2 6 1 5 8 2 3 4 3 3 3 4 7
X X X X X X X X X X X X X X X
108 108 10 8 108 10 8 108 108 108 108 108 10 8 10 8 108 108 108
NOTE. Each value represents the mean of two experiments. * Mean number of bacteria attached to 40 epithelial cells. t In the adhesion test tube. t p
