0022-534 7/86/1361-0132$02.00/0 Vol. 136, July
THE JOURNAL OF UROLOGY
Printed in U.S.A.
Copyright © 1986 by The Williams & Wilkins Co.
HEPARIN INHIBITION OF INCREASED BACTERIAL ADHERENCE FOLLOWING OVERDISTENSION, ISCHEMIA AND PARTIAL OUTLET OBSTRUCTION OF THE RABBIT URINARY BLADDER MICHAEL R. RUGGIERI,* PHILIP M. HANNO, SAEED SAMADZADEH, ERIC W. JOHNSON ROBERT M. LEVIN
From the Division of Urology and Department of Pharmacology, University of Pennsylvania, the Veterans Administration Medical Center, Philadelphia, Pennsylvania and the University of Urmia, Urmia, Iran
While it is well established clinically that urinary tract infection in the presence of outflow obstruction may be associated with difficulty in eradicating bacteria, it is not clear whether this is secondary to the presence of residual urine volume or other local effects of the obstruction such as attenuation of the intrinsic antibacterial defense mechanisms of the mucosal surface. Experiments in our laboratory and others over the past several years have demonstrated that the primary antibacterial defense mechanism of the bladder is the antiadherence effect of the bladder surface mucin layer. Additional studies have shown that heparin can duplicate this antiadherence activity of bladder mucin. The present report demonstrates that one hour of overdistension or ischemia and one week of partial outlet obstruction cause a functional defect in the intrinsic antiadherence effect of the bladder mucosa as evidenced by increased bacterial adherence. This defect can be reversed by heparin exposure prior to bacterial challenge. These results indicate that partial outlet obstruction and its potential sequelae such as overdistension and, particularly, mucosal ischemia, have dramatic adverse effects on the intrinsic antiadherence defense mechanism of the bladder. These effects can be reversed by intravesical exposure to an exogenous anionic polyelectrolyte (heparin). Under normal conditions, when bacteria become introduced into the bladder they are quickly eliminated. Early experimental work has shown that the bladder eliminates bacteria by two mechanisms: repeated dilution and evacuation which is dependent on small residual urine volumes to be effective 1 and intrinsic antibacterial defense mechanisms of the bladder mucosa. 2 The first mechanism becomes less effective in the presence of obstruction since relatively greater residual urine volumes exist. 3 Regardless of residual urine volume, in order for a microorganism to initiate an infectious process in the bladder it must first attach to the luminal epithelium. The mucin layer coating the bladder luminal epithelium prevents bacterial adherence. Removing this layer by various means leads to increased bacterial attachment. 4- 7 Adherence can be restored to the low levels of the mucin intact state by brief exposure of the mucin deficient bladder to either exogenous anionic polyelectrolytes such as heparinS--12 or to a bladder mucosal extract from a variety of species.13 Vesical overdistension, one common sequela to obstruction, has been reported to disrupt or destroy the mucin layer as evidenced by colloidal iron histochemistry. 14 Increased intravesical pressure resulting from overdistension is thought to disrupt the tissue integrity or blood supply to the bladder. 15- 17 The present report investigates the functional effects of acute overdistension, ischemia and chronic outlet obstruction on attachment of bacteria to the urothelium. Additionally, exogenous anionic polyelectrolyte (heparin) was used in an attempt to restore the antiadherence integrity of the bladder mucosa which was disrupted by these experimental pathologies. Accepted for publication January 13, 1986.
* Requests for reprints: University of Pennsylvania School of Medicine, Division of Urology, 3010 Courtyard Bldg., H.U.P., 3400 Spruce St., Philadelphia, PA 19104. Supported by grants from the Veterans Administration, NIH grant #R0-1-AM-6508-01 and the McCabe Fund. 132
MATERIALS AND METHODS
Preparation of bacteria. Lyophylized bacterial neotype E. coli 11775 purchased from the American Type Culture Collection was rehydrated in sterile 0.9 per cent NaCl and serially passed repeatedly in static brain-heart infusion broth (BHI; Difeo Laboratories) at 37C. To label the organism with 3 H a 0.4 per cent innoculum (v/v) was added to four ml. of fresh BHI containing five µCi 3 H-adenine (New England Nuclear) per ml. and incubated statically at 37C for 18 to 24 hours. The bacteria were sedimented at 3000 g for 20 minutes and resuspended in their original volume of 0.9 per cent NaCl. Colony forming units (CFU) were determined by serial dilution as previously described.11· 13 Bacterial suspensions were prepared for a given experiment with an optical density adjusted to 0.2 units so as to yield a constant 108 CFU/ml. suspension. Mucin deficient rabbit bladder model. Male New Zealand white rabbits weighing two to 2.5 kg. were sedated with an i.m. injection (0. 7 ml./kg.) of a ketamine/zylazine mixture (29.2 mg./ml. ketamine, 8.3 mg./ml. zylazine). Surgical anesthesia was induced with one ml. of 50 mg./ml. pentobarbital given over the course of surgery. Rabbits were secured, the urinary bladders were catheterized with a #8 French catheter, and the bladders were emptied and rinsed with three 10 ml. aliquots of 0.9 per cent NaCl. Bladders were rendered mucin deficient by infusion of 10 ml. of 50 per cent acetone/water through the catheter with a 10 ml. syringe. Mucin intact controls received 10 ml. of 0.9 per cent NaCl. During the one minute exposure the syringe plunger was gently pushed back and forth to increase the efficiency of extraction. The bladder was then flushed with three 10 ml. aliquots of 0.9 per cent NaCl. For heparin treated animals, 10 ml. of five mg./ml. heparin in 0.9 per cent NaCl was infused through the catheter, left in contact with the bladder lumen for 15 minutes, then removed. 1.5 X 108 CFU 3 H labeled bacteria suspended in 1.5 ml. of 0.9 per cent NaCl were introduced into the bladder and flushed in with five ml. of 0.9 per cent NaCl. After a 20 minute exposure to the
HEPARIN INHIBITS BACTERIAL ADHERENCE IN OBSTRUCTION
labeled bacteria the bladder was emptied and flushed with three 10 ml. aliquots of 0.9 per cent NaCL The animal was then sacrificed with pentobarbital, the bladder removed and the mucosa was dissected free from the underlying muscle layer and assayed for 3 H activity. Experimental pathologies. Rabbits were divided into three groups of 12 to 20 animals per group. Bacterial adherence was determined for each group in the presence and absence of exogenous heparin. Partial obstruction of the bladder (group #1) was established as previously described18 by gently securing a 00 silk ligature around the temporarily catheterized bladder neck (not including the vesical arteries) one week prior to the bacterial adherence studies. For the overdistended group (group #2), animals were anesthetized and catheterized with a #8 Foley catheter with a three cc balloon. Intravesical pressure was monitored as the bladder was filled with saline, and overdistension was defined as five ml. greater than the volume which resulted in maximum pressure as previously described. 19 This was maintained for one hour just prior to measuring bacterial attachment. The ischemic group was catheterized, the bladder was emptied and then exposed with a midline incision and a 00 silk ligature was secured tightly around the bladder base (including the vesical arteries) for one hour prior to bacterial adherence studies. 20 Sham operations in the control animals consisted of similar anesthesia and bladder exposure without inducing the experimental pathologies. The adherence assay was performed as described above. Since partially obstructed animals could not be catheterized, a suprapubic tube was inserted into the bladder through a midline abdominal incision. After emptying the bladder and rinsing with three 10 ml. aliquots of 0.9 per cent NaCl, 1.5 ml. 0.9 per cent NaCl containing 1.5 x 108 colony forming units of radiolabeled E. coli was then infused, washed into the bladder with five ml. 0.9 per cent NaCl and left in for 20 min. Heparin treated groups received 10 ml. of five mg./ml. heparin in 0.9 per cent NaCl for 15 minutes prior to the radiolabeled bacteria. The bladder was then emptied, irrigated with 10 ml. 0.9 per cent NaCl three times, removed and the mucosa was dissected off and assayed for radioactivity. Results are expressed as per cent of the sham operated control group on the given day and the per cent value averaged over the different days of experimentation. Recording of radioactivity. Bladder mucosae were digested overnight in one ml. of 1.0 N NaOH at 37C. The volume was then brought to five ml. with 0.9 per cent NaCl and 1.5 ml. triplicates were placed in 20 ml. glass scintillation vials and acidified with 0.1 ml. glacial acetic acid. The remaining 0.5 ml. of mucosal homogenate was diluted 50-fold with 0.9 per cent NaCl and assayed for protein content. 21 Eighteen ml. of hydrofluor (National Diagnostics) was added and the vials were vortexed. 0.1 ml. triplicates of the bacterial suspensions were also suspended in 20 ml. of hydrofluor. The radioactivity of the bacterial suspension (in counts per minute or CPM) and the viable count (colony forming units or CFU) were used to determine the CFU per CPM for each bacterial suspension. The radioactivity of the tissue samples was converted to the actual number of bacteria (in colony forming units) attached to the mucosa by multiplication of the tissue CPM by the bacterial suspension CFU per CPM. This then was divided by the protein concentration of the mucosal digest to yield the number of colony forming units of bacteria attached to the bladder mucosa per milligram of mucosal protein. Computer enhanced autoradiography. The adherence assay was performed as described above. However, instead of dissolving the mucosa and liquid scintillation spectrometry, the bladder mucosae were spread out on glass slides (mucosa side up) and air dried at 60C for one hour. A sheet of photographic film (LKB Ultrofilm) was placed over the mucosa in the dark and sandwiched between two boards to maintain close contact between the film and the tissue. After 48 hours of exposure the
film was developed and scanned by a densitometric television camera interfaced with an IBM personal computer (Logic Systems). This system generates pseudo-color enhancement of the black and white autoradiogram. Red color corresponds to high optical density of the autoradiogram and thus high numbers of bacteria attached to the mucosa. Conversely, blue color corresponds to low numbers of bacteria attached. Materials and data analysis. All chemicals used were purchased in crystalline form of the highest purity available (Sigma Chemical Company) and stored according to the manufacturers' specifications. The heparin preparation had a specific anticoagulant activity of 176 units/mg. Means of assays performed in triplicate were counted as single datum points and results are expressed as means +/- standard error of the mean for N greater or equal to 6. Statistical analysis was performed by analysis of variance using the F distribution of Fisher with a post hoc Newman-Keuls multiple range test for paired comparisons. A probability value of less than 0.05 was required for statistical significance. RESULTS
A summary of the results is displayed in figure 1. Heparin exposure did not decrease bacterial adherence below the already low levels obtained for mucin-intact animals. Rendering bladders mucin deficient by one minute exposure to 50 per cent acetone increased adherence 50-fold. This increase was nearly completely prevented by a 15-minute exposure to heparin. One week of partial outlet obstruction increased adherence approximately five-fold over controls and heparin exposure reduced adherence to levels not statistically different from controls. One hour of overdistension resulted in a four-fold increase in adherence which was restored to control levels by hepa;in. The six-fold increase in adherence induced by one hour of ischemia was significantly decreased by heparin; however, adherence 5000
1 WEEK PARTIAL OBSTRUCTION
1 HOUR 1 HOUR OVER ISCHEMIA DISTENSION
FIG. 1. Effect of heparin on E. coli adherence to rabbit bladder in various experimental pathologies. Adherence assay was performed as described in Methods. Heparin concentration was 5 mg./ml. throughout all treatments. Results are expressed as per cent of sham operated control group on given day and per cent value averaged over different days of experimentation. Average adherence overall for all sham operated control animals was 2.03 +/- 0.72 X 105 CFU per mg. mucosal protein. Average adherence for control mucin intact animals was 1.67 +/- 0.68 X 106 CFU per mg. mucosal protein. Open bars represent control data, hatched bars represent heparin preexposure.
RUGGIERI AND ASSOCIATES
FIG. 2. Computer enhanced autoradiogram of E. coli attachment to mucin-deficient and mucin -intact bladder mucosa. Procedure was performed as described in Methods. Red areas correspond to high optical density of autoradiogram and thus high numbers of bacteria attached to mucosa. Conversely, blue color corresponds to low numbers of bacteria attached. Black areas correspond to no detectable numbers of attached bacteria.
FIG. 4. E. coli adherence to 1 hour overdistended bladder mucosa. See figure 2.
FIG. 5. E. coli adherence to 1 hour ischemic bladder mucosa. See figure 2.
FIG. 3. E. coli adherence to 1 week partially obstructed bladder mucosa. Red = high numbers of bacteria. Blue = low numbers of bacteria attached.
levels in the ischemia plus heparin group remained significantly greater than to mucin intact controls. Results of the computer enhanced autoradiography were qualitatively similar to the results of scintillation spectrometry. Figure 2 clearly illustrates that more bacteria attach to the mucin deficient bladder mucosa than to the mucin intact control mucosa. Adherence to the mucin deficient mucosa was not uniform. Qualitatively greater adherence was detected in one part of the bladder dome as indicated by the red areas of figure 2. Autoradiograms from the experimental pathologies are displayed in figures 3 to 5. In comparison to the mucin deficient autoradiogram, bacteria attached to the experimental pathological mucosae in a much more discrete, punctate manner. No great quantitative differences in adherence could be detected between the three experimental pathologies using the autoradiographic technique.
In clinical urologic practice it is well known that urinary tract infection in patients with bladder outlet obstruction is usually more difficult to treat than in unobstructed patients. The reason for this has been previously thought to be due to an increase in post-void residual urine volume in obstructed patients. 3• 22 Adherence of bacteria to the urothelium was studied in this investigation because in order for a microorganism to become established in the bladder, and thus initiate an infection, it must first attach to the luminal epithelium. Since urine is known to be an excellent culture medium for most enterobacteria, 23 •24 the increased residual urine in obstruction would be expected to result in proliferation of organisms that gain entrance into the bladder lumen. Unless these proliferating organisms generate a toxin that can induce symptoms from inside the bladder lumen, no symptomatology would result from bacteria growing in the liquid medium of the bladder lumen. Therefore, in order for bacteria which are proliferating inside the bladder to initiate an infectious process there must be adherence and subsequent invasion of the mucosal lining. Me-
HEPARIN INHIBITS BACTERIAL ADHERENCE IN OBSTRUCTION
chanical overdistension of the bladder, a common sequela to bladder outlet obstruction, is thought to disrupt the tissue integrity or blood supply to the bladder as a result of increased intravesical pressure. 15- 17 While bladder overdistension has been previously shown to disrupt or destroy the colloidal ironpositive mucin layer of the urothelium, 14 the functional effect of overdistension on bacterial adherence has not, until now, been determined. Results of the present investigation demonstrate that chronic outlet obstruction and acute overdistension and ischemia result in a functional alteration in the urothelium which leads to increased bacterial adherence. Thus not only may obstruction and its sequelae (overdistension and ischemia) increase bacterial colony counts in the urine allowing infection to occur by overwhelming the natural antiadherence layer, it may also result in a primary insult to the bladder surface mucin, decreasing its inherent antibacterial adsorption properties. Computer enhanced autoradiography was utilized in this investigation to determine the regional distribution of bacteria attached to the urothelium as a result of the experimental pathologies. Scintillation spectrometry of 3H-K coli attachment to the urothelium yields quantitative data that can be analyzed by parametric statistical methods in order to determine whether a statistical difference in adherence exists between various experimental manipulations. However, it gives no information about where the bacteria attach on the urothelial surface. As illustrated in figures 3 to 5, K coli attached to the obstructed, overdistended and ischemic bladders in a very focal fashion. This may indicate that these experimental pathologies induce focal damage to or removal of the mucin layer, allowing the bacteria to attach to the underlying epithelium at those sites. An anionic polyelectrolyte (heparin) was able to inhibit adherence to the mucin deficient bladder, confirming earlier work. 8 - 13 Exposure of obstructed or overdistended bladders to heparin prior to bacterial challenge also reduced bacterial adherence to mucin-intact control levels. Thus the focal damage to the mucin layer induced by these experimental pathologies was reversed by subsequent exposure to heparin. It would appear that heparin binds to the areas of focal damage and thereby duplicates the antiadherence effect of the absent mucin layer. This conclusion is supported by the finding that 3Hheparin binds to the mucin-deficient bladder but not to the mucin-intact bladder (unpublished observations). The finding that heparin was not able to completely reverse the increased adherence resulting from one hour of ischemia is not surprizing in light of several unpublished observations from our laboratory. Whenever a mucin-intact rabbit inadvertently dies during the adherence assay the levels of adherence increase 10 to 100-fold. One interpretation of this observation is that the ischemia subsequent to death causes the mucin layer to become detached from the epithelium, allowing bacteria to attach. If pieces of bladder tissue which are frozen in liquid nitrogen then thawed out are exposed to a radiolabeled K coli suspension one can demonstrate significant adherence of bacteria to the tissue; however, pre-exposure to heparin does not inhibit this adherence. This may indicate that heparin does not bind to bladder tissue which has been freeze-thawed. While there are many possible explanations for this finding, an attractive interpretation is that the membrane potential of the epithelium is necessary for binding of the endogenous mucin layer and for the binding of exogenous anionic polyelectrolytes such as heparin. Acknowledgment. The authors acknowledge the excellent technical assistance of Ms. Beth Whitowsh
L Cox, C. Kand Hinman, F. Jr.: Experiments with induced bacteriuria, vesical emptying and bacterial growth on the mechanism of bladder defense to infection. J. Urol, 86: 739, 196L 2. Gillenwater, J. Y, Cardozo, N. C., Tyrone, N. 0. and Mulholland, S. G.: The effect of vesical mucosa on bacterial growth. Clin. Res., 16: 47, 1968. 3. Hinman, F. Jr.: Infection and stasis. In: Benign Prostatic Hypertrophy. Edited by F. Hinman, Jr. New York: Springer-Verlag, chapt. 76,pp. 727, 1983. 4. Mulholland, S. G.: Lower urinary tract antibacterial defense mechanisms. Invest. UroL, 1 7: 93, 1978. 5. Parsons, C. L., Shrom, S. H., Hanno, P. M. and Mulholland, S. G.: Bladder surface mucin: examination of possible mechanisms for its antibacterial effect. Invest. UroL, 16: 196, 1978. 6. Shrom, S. H., Parsons, C. Land Mulholland, S. G.: Vesical defense: further evidence for a charge related antiadherence mechanism. Surg. Forum, 29: 623, 1978. 7. Parsons, C L., Greenspan, C. and Mulholland, S. G. The primary antibacterial defense mechanism of the bladder. Invest. UroL, 13: 72, 1975. 8. Hanno, P. M., Fritz, R, Wein, A. J. and Mulholland, S. G.: Heparin as antibacterial agent in the rabbit bladder. Urology, 12: 411, 1978. 9. Hanno, P. M., Parsons, C. L, Shrom, S. ff, Fritz, Rand Mulholland, S. G.: The protective effect of heparin in experimental bladder infection. J. Surg. Res., 25: 324, 1978. 10. Hanno, P. M., Fritz, R W., Mulholland, S. G. and Wein, A. J.: Heparin-examination of its antibacterial adsorption properties. Invest. UroL, 18: 273, 198L lL Ruggieri, M. R, Hanno, P. M. and Levin, R M.: The effects of heparin on the adherence of five species of urinary tract pathogens to urinary bladder mucosa. UroL Res., 12: 199, 1984. 12. Ruggieri, M. R, Hanno, P. M. and Levin, RM.: K coli adherence to anion exchange resin: an in-vitro model for the initial screening of potential antiadherence agents. Urology, 27: 343, 1986. 13. Ruggieri, M. R, Hanno, P. M. and Levin, R M.: Further characterization of bacterial adherence to urinary bladder mucosa: comparison with adherence to anion exchange resin. J. UroL, 134: 1019, 1985. 14. Perlow, D. L., Gikas, P. W. and Horowitz, KM.: Effect of overdistension on bladder mucin. Urology, 18: 380, 198L 15. Lapides, J., Diokno, A. C., Lowe, RS. and Kalish, M. D.: Followup on unsterile intermittant self-catheterization. J. UroL, 111: 184, 1974. 16. Mehortra, R. M. L: An experimental study of the vesical circulation during distension and in cystitis. J. PathoL BacterioL, 66: 79, 1953. 17. Lapides, J.: Mechanisms of urinary tract infection. Urology, 14: 217, 1979. 18. Levin, R. M., High, J. and Wein, A. J.: The effect of short-term obstruction on urinary bladder function in the rabbit. J. UroL, 132: 789, 1984. 19. Levin, R. M., Staskin, D.R. and Wein, A. J.: The effects of acute overdistension of the rabbit urinary bladder. NeuroL Urodynam., 2: 63, 1983. 20. VanArsdalen, KN., Wein, A. J. and Levin, RM.: The contractile and metabolic effects of acute ischemia on the rabbit urinary bladder. J. UroL, 130: 180, 1983. 2L Lowry, 0. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J.: Protein measurement using the Folin phenol reagent. J. BioL Chem., 193: 265, 195L 22. Lloyd-Davies, R W. and Hinman, F., Jr.: Structural and functional changes leading to impaired bacterial elimination after overdistension of the rabbit bladder. Invest. UroL, 9: 136, 197L 23. Pasteur, M. L.: Examen du role attribue au gaz atmospherique clans la destruction des substances animales et vegetales apres la mort. C.R. Acad. Sci., 56: 734, 1863. 24. Mulholland, S. G., Perez, J. R and Gillenwater, J. Y: The antibacterial effects of urine. Invest. UroL, 6: 569, 1969.