Studies of the pathogenesis and treatment of urinary tract infections using a model of the human bladder

Urinary tract infections are generally preceded by transfer of organisms from the distal urethra to the bladder (20, 148). However, although urinary infections are predominantly due to pure cultures of Escherichia coli, the distal urethra contains a mixed flora in which E. coli is relatively uncommon and anaerobes predominate (73, 103). This discrepancy between the bladder and distal urethral flora may be due to differential adhesion or differential growth rates. In this dissertation I have tested the hypothesis that differential growth rates of urethral organisms in urine explains the predominance of E. coli as a pathogen. These experiments showed that the balance between bacterial growth and washout may have a pivotal role in the pathogenesis of infection and perhaps therefore in treatment. A model of the human bladder used for the pathogenesis studies was then used to study the activity of mecillinam and ampicillin under conditions simulating human urinary infection. The model proved realistic especially for synergy studies where shortcomings in conventional in vitro methods are a cause for concern. The following topics were studied. 1. Urine was chosen as a test medium for definitive experiments because growth rates of organisms other than E. coli were different in broth and in urine. A method for sterilizing urine in bulk was developed which did not affect growth supporting properties. 2. E. coli was shown to grow faster and to have a shorter lag period than almost all other organisms when studied in shake culture. 3. A continuous culture model of the human urinary bladder was employed for differential growth studies of organisms in sterilized human urine. This model reproduced many of the characteristics of the human lower urinary tract and enabled study of the balance between bacterial growth and the tendency of urine to wash organisms out of the tract. 4. Mixed cultures of approximately equal numbers of E. coli and a second potential urinary' pathogen were introduced into the bladder model and quantitative cultures performed at intervals up to 24 h. In 15 experiments E. coli eventually dominated the second pathogen which was sometimes undetectable at 24 h. Similar changes in bacterial populations seen in infected patients indicate that differential growth rates may be an important determinant of the pathogenicity of E. coli. 5. The use of the bladder model was then extended to investigations of antibiotic activity under realistic conditions. The value of the model for synergy studies with ampicillin and mecillinam was assessed by parallel conventional in vitro tests and an animal infection protection test*. The bladder model gave similar results to mammalian studies and appeared to be far superior to conventional methods. This model may be valuable in the initial assessment of new urinary antibiotics. 6. A representative array of organisms for the above study was selected following a survey of resistance patterns of 2000 clinical isolates of Enterobacteriaceae. An incidental by-product of this survey was the establishment of a breakpoint for mecillinam susceptibility in the Kirby-Bauer antibiotic disk test. 7. Work on the effect of mecillinam and/or ampicillin upon bacterial viability was extended to investigations of the relative contribution of permeability barriers and 3-lactamases to antibiotic susceptibility. Unlike ampicillin, mecillinam resistance of 77 clinical isolates of bacteria appeared to be independent of intracellular 3-lactamase levels, suggesting that the barrier effect may be more pronounced in bacterial resistance to mecillinam than to ampicillin. Kinetic studies using urine as a growth medium, and in particular the use of a bladder model have provided a unifying explanation of many features of both the pathogenesis and treatment of urinary infections. * Carried out by Dr. R.C. Cleeland.Science, Faculty ofMicrobiology and Immunology, Department ofGraduat