Article
Vol. 10. No. 3
157 '1
Eur. J. Clin. Microbiol. Infect. Dis., March 1991, p. 157-162 0934-9723/91/03 0157-06 $ 3.00/0
' Activity of Antiseptics against Biof'dms of Mixed Bacterial Species Growing on Silicone Surfaces
D. Stickler*, P. H e w e t t
As part of a programme to establish the relative merits of antiseptics that are used as bladder instillations to control urinary tract infectious in patients with indwelling catheters, the activity of five such formulations were tested against dense (109 cfu/ cm 2) mixed biofilms composed of Citrobacter diversus, Pseudomonas aeruginosa and En terococeusfaecalis growing on silicone discs. All three species were resistant to chlorhexidine (200 rag/l) and povidone-iodine (1% v/v) in the biofilm mode of growth, whereas these agents rapidly eliminated viable cells from urine suspensions of the mixed community. Lactic acid (1% v/v) produced a 1 log reduction of the biofilm population within 30 min of exposure. The mandelic acid (1% w/v) and mandelic acid (0.5 % w/v)/iactic acid (0.5 % v/v) mixture proved to be the most effective in eliminating the biofilm organisms. It is suggested that these latter solutions should now be tested for efficacy in bladder washouts against urinary tract infectious in catheterized patients.
Indwelling bladder catheters taken from patients with chronic urinary tract infections are colonized by films of adherent bacteria (1-3). It is becoming clear that in this biofilm mode of growth, bacteria are resistant to a variety of antibacterial agents (4, 5). Costerton (6) suggested that during antibiotic treatment of infections of the catheterized urinary tract, while the antibacterial agent might kill planktonic organisms in urine suspension, cells in the biofilms on the catheter and the bladder epithelium could survive and be responsible for the rapid re-establishment of these infections that occurs after the course of treatment is completed
(7). As an alternative to systemic antibiotic therapy, washing the bladder with instillations of antiseptics has been advocated as a method for controlling urinary infections in patients with long-term indwelling catheters (8). While regular bladder washouts with antiseptics such as chlorhexidine have been reported to prevent infections (9, 10), once the infection has become established in the catheterized urinary tract many of the antiseptics currently used as instillations are unable to eliminate the bacteria from the bladder (11-13). This failure to clear established infections was also observed in experiments performed in a physical School of Pure and Applied Biology, University of Wales College of Cardiff, PO Box 915, Cardiff CF1 3TL, UK.
model of the catheterized bladder (14, 15). Cells that had colonized the surfaces of the model were particularly resistant to the antiseptics and initiated the rapid recovery of the urine cultures after the instillation period. Biofilms of Escherichia coli, Citrobacter diversus, Pseudomonas aeruginosa, Proteus mirablis and Klebsiella pneumoniae growing on silicone discs have also been shown to resist the concentrations of chlorhexidine commonly used in bladder washouts (16, 17). Mandelic acid and a mandelic/lactic acid formulation, however, proved to be markedly bactericidal against these biofilms. In these experiments single species biofilms composed of monolayers of ceils w ere produced at cell densities of ca. 107 cfu/cm 2. While these simple pure culture biofilms exist in vivo, it is also clear that mixed populations of up to four species can form thick multi-layered films on catheter surfaces (2, 3). In attempting to refine a laboratory model [or testing the activity of antibacterials against biofilms we have sought to mimic the characteristics of some of the biofilms that we have observed on catheters taken from paraplegic patients on long-term chatheterization. In this paper we report some observations on the activity of antiseptics on mixed community biofilms composed of Citrobacter diversus, Pseudomonas aeruginosa and Enterococcusfaecalis at ceil densities of> 109 cfu/cm2.
158
Materials and Methods Test Organisms. The strains of bacteria used in this investigation, Citrobacter diversus R25.I, Pseudomonas aeruginosa R1811 and Enterococcus faecalis R812, were all isolated from catheters removed from patients undergoing long-term indwelling catheterization at the Welsh Spinal Injuries Unit, Rookwood Hospital, Cardiff. Silicone Discs. Cosmesil elastomer (Cosmedia, UK) was prepared as a polymer sheet 0.2 era in thickness. Discs (0.7 cm diameter) were cut from the sheet and autoclaved in deionized water (121 °C for 15 rain) prior to use.
Model of the Catheterized Bladder. The catheterized bladder is a continuous flow system in which a reservoir of urine is maintained by the presence of the catheter. The details of our simple physical model of such a system have been described previously (14). In summary, sterile pooled urine is delivered by a peristaltic pump at a rate of i ml/min to a small fermentation flask which acts as the bladder. The flask is maintained at 37 °C by a water jacket and the residual volume of urine is maintained at i0 ml by a catheter drainage system connected to a vacuum pump via a Buchner flask (21), which represents the urine drainage bag.
Biofilm Formation. Mixed population biofilms were established with the three bacterial species using a modified version of the procedure described by Prosser et at. (18). The bacteria were grown overnight on Mueller-Hinton Agar (Oxoid, UK) at 37 °C. Cells were then scraped from the agar and suspended in Hanks-Hepes (N-2-hydroxyethylpiperazine-N'-2-ethane sulphonic acid)(H-H) buffer pH 7.4, to an optical density of 0.8 at 540 nm. Aliquots (5 ml) of each bacterial suspension were added to the sterile discs in a plastic petri dish. The discs were incubated for 1 h at 37 °C to permit attachment of suspended cells. They were then removed from the suspension and washed once in H H buffer. Sets of discs were transferred to the bladder model, being placed on a nylon mesh support in a growth medium of pooled sterile urine. After a preliminary incubation period of 2 h, the supply of fresh sterile urine to the model was switched on and the model run for 48 h.
Bactericidal Activity against Biofllms. Sets of 30 discs colonized by the three membered biofilm were prepared for each experiment. The discs were taken from the model and washed twice in H H buffer to remove planktonic cells. Nine discs (controls) were placed into a petri dish containing 20 ml of sterile pooled urine. Two further sets of nine discs were then placed into dishes containing sterile pooled urine and antibacterial agents at concentrations normally used for bladder washouts. The remaining three discs were used to determine the initial number of viable cells present in the biofilms at the start of the experiment. The three petri dishes were incubated at 37 °C and triplicate discs removed from each at 30, 60 and 120 man. Discs were washed twice in HH buffer and placed individually in NIH broth (10 ml) containing Tween 80 (3 % v/v) to neutralize the biocides. Disruption of the biofilms was achieved by sonication for 5 rain in a Trans-Sonic T310 sonicating bath (Camlab, UK) and agitation for 2 man on a vortex mixer. Discs removed at zero time were treated in the same way and viable cell counts were performed by plating out serial dilutions of the resulting cell suspensions onto CLED agar (Oxoid) and overnight incubation at 37 °C.
Eur. J. Clin, Microbiol. Infect. Dis.
Bactericidal Activity of Antiseptics on Suspended Cells. Sets of colonized silicone discs were prepared as described above. Sets of nine discs were then placed individually into 10 ml of sterile urine (controls) and sterile urine containing antiseptics. Three discs were also taken to determine the initialnumber of viable cells present in the biofilm (t = 0). The biofilms were disrupted by sonication and vortex mixing and the discs then removed from the suspensions. The ceils from the disrupted biofilms were incubated in urine and urine containing antiseptics at 37 °C. Triplicate samples were removed at 30, 60 and 120 min for viable cell counting on CLED agar.
Antibacterial Agents. The antiseptic agents used in this study were chlorhexidine digluconate (ICI, UK), mandelic and lactic acids (BDH, UK), and povidone-iodine (Napp Laboratories, UK). Scanning Electron Microscopy. Silicone discs colonized by biofilm were washed gently in HH buffer and placed in a fixative solution consisting of 3 % glutaraldehyde in phosphate buffer (pH 7.4) for 2 h at 4 °C. Samples were then exposed to osmium tetroxide (1% w/v) in phosphate buffer (pH 7.4) for 1 h and dehydrated in a series of aqueous ethanol solutions (20-100 %). Critical point drying was performed in liquid COz, and finally samples were coated with gold i n a sputter coater before examination in a scanning electron microscope (JEOL JSM5200).
Results
The 48 h growth period in the bladder model consistently produced biofilms composed of about 109 cfu/cm2. The composition of the three-membered community in the biofilm was also relatively uniform. Citrobacter diversus was characteristically the major component and Enterococcus faecalis the minority member, the ratio of the numbers of Citrobacter diversus to Pseudornonas aeruginosa to Enterococcusfaecalis being 10:4:1. The activity of chlorhexidine, mandelic acid,lactic acid, a mandelic/lactic acid mixture and povidoneiodine against these biofilms and control cell suspensions prepared from biofilms are shown in Figures 1 to 5. It can be seen that chlorhexidine (200 txg/ml) and povidone-iodine (1% v/v) had negligible activity against the three species in the biofilm mode of growth. Lactic acid ( 1 % ) produced after 2 h exposure a uniform 2 log reduction in the viable numbers of each organism. Mandelic acid and the mandelic/lactic acid mixture proved to be the most active, essentially eliminating all three organisms within 2 h. The scanning electron micrograph (Figure 6) confirms the multi-layered nature of these biofilms. The effect of mandelic acid (Figure 7) is clearly elimination of the biofilm from the silicone surface.
Vol. 10, 1991
159
8.0 _
C
B
A
6.0 m
'~E O
4.0 o
q 2.0
!
I
l
I
tlJ
_z
it
o
7.
6.0
--~3
4.0
.q 2.0
I ....
0
I
I
60
0
120
60
120
o
60
120
TIME (mln) Figure 1: Activity of chlorhexidine (200 I.tg/ml) against biofilms and suspensions of Citrobacter diversus (o), Pseudomonas aeruginosa (A) and Enterococcus faecalis (•). T h e corresponding d o s e d symbols indicate the viability of control biofilms and cell suspensions in normal urine. The vertical bar represents + SD.
A
8.0
B
C
..a ii
6.0
o
"E t)
4.0
9 2•0
•
I
I
I
I.U
z
o
~.__.--.--o
6.0
o
4.0
o 2.0
v
(I
60
120
0
60 TIME (min)
120
0
v
60
w
120
Figure 2: Activity of mandelic acid ( 1 % w/v) against biofilms and suspensions of Citrobacter diversus (o), Pseudomonas aeruginosa (z~) and Enterococcus faecalis (C3). The corresponding closed symbols indicate the viability of control biofilms and cell suspensions in normal urine. The vertical bar represents +- SD.
160
Eur. L Clin. Microbiol. Infect. Dis.
LI..
_.9_ rn
C
B
A
8.0
6,0
"E
4.0
it
o
9 2,0
v
f
Ill
z
r~
6.0
@ ¢o,=,
4.0
o, 2.0
•
w 0
.__...O_.__..-~
_.._...--o
'
v
v
60
120
v
60
o
........U
120
0
60
120
TIME (min)
l~gure 3: Activity of lactic acid ( 1 % v/v) against biofilms and suspensions of Citrobacter diversus (o), Pseudomonas aeruginosa (A) and Enterococcus faecalis (0). The corresponding closed symbols indicate the viability of control biofilms and cell suspensions in normal urine. The vertical bar represents +_SD.
8.0
__. an
6.0
~ 0
4.0
A
C
B
q 2.0
I
I
r
I
I
I
W
~
6.0
,p 0
e
4.0
0 2.0
v
0
60
120
0
w
60
v
v
120
0
v
60
120
TIME (min}
l,igure 4: Activity of mandelic (0.5 % w/v) and lactic acid (0.5 % v/v) against biofilms and suspensions of Citrobacter diversus (o), Pseudomonas aeruginosa (A) and Enterococcus faecalis (O). The corresponding dosed symbols indicate the viability of control biofilms and cell suspensions in normal urine. The vertical bar represents _+SD.
VO1.10, 1991
161
Discussion
Previous studies on the sensitivity of urinary tract organisms growing on the surface of catheter materials have used pure cultures and biofilms with cell densities which represent simple layers one or a few cells thick (16-19). Ramsay et al. (3) reported that indwelling catheters often carry mixed communities of organisms and our own unpublished observations on long-term indwelling catheters removed from paraplegic patients have also revealed the presence of thick multi-layered biofilms populated by several bacterial species. Figure 5: Activity of povidone-iodine (1% v/v) against biofilms and suspensions of Citrobacterdiversus (o), Pseuaomonos aeruginosa (A) and Enterococcus faecalis (~). The corresponding dosed symbols indicate the viability of control biofilms and cell suspensions in normal urine. The vertical bar represents + SD.
Figure 6: Scanning electron micrograph of the mixed community biofilm after 48 h growth in urine in the mode/of a catheterized bladder.
~°
I~KU
HI
!~
OBOIL@2
Figure 7: The biofilm after exposure to mandelic acid (I % w/v) for 120 rain.
The Citrobacter diversua, Pseudomonas aeruginosa, Enterococcus faecalis mixture used in this study grew as a stable, reproducible community with relatively constant proportions of each species. Incubating the discs in urine in the model of the catheterized bladder for 48 h resulted in thick multi-layered biofilms with a characteristic cell density of i09 cfu/cm2. The cells in these biofilms respond quite differently from the same cells in simple urine suspension to antiseptics. While the suspended cells are essentially eliminated from the urine by chlorhexidine, the biofilm community survives well, even after exposure periods of 120 min (Figure 1). Previous work had shown that simple monolayers of this strain of Citrobacter diversus were sensitive to chlorhexidine (17). The thick multi-layered three membered community obviously provides protection for an organism which would normally be considered extremely sensitive to this antibacterial agent. These observations also highlight the need for more information on the nature of biofilms growing on indwelling catheters. The testing for agents active against biofilms can then be performed on laboratory biofilms that are models of the natural biofilms. While mandelic acid was rapidly and effectively bactericidal against the cell suspension, it took 120 min to eradicate the three species from the biofilm (Figure 2). The normal times of exposure to antiseptics in the bladder washout procedure are generally 20-30 rain and this period produced a 34 log reduction in the viability of the biofilm. Mandelic acid has previously been shown to be effective against simple biofilms of Escherichia coli, Pseudomonas aeruginosa, Proteus mirabilis, Citrobacter diversus and Klebsiella pneumoniae (16-17). It would appear therefore that this solution has useful activity against a range of biofilms likely to be found on indwelling bladder catheters.
162
The results presented in Figure 3 indicate that a 20-30 min exposure to lactic acid (1% v/v) would produce a 1 log reduction in the viability of the biofilm. A similar effect has been observed against simple Escherichia coli films (16). The combination of mandelic and lactic acids proved as effective as mandelic alone against both the biofilm and urinary suspension (Figure 4). In experiments conducted in a physical model of the catheterized bladder (15), povidone-iodine was shown to be ineffective against simulated established infections with urinary tract pathogens including Pseudomonaz" aeruginosa and Enterococcus faecalis. The results presented in Figure 5 confirm the poor activity of povidone-iodine against organisms colonizing the surface of catheter material. In conclusion, it seems that ceils in mixed culture biofilms at cell densities of > 109 cfu/cm2 are resistant to the concentrations of chlorhexidine and povidone-iodine that are used in the bladder washout procedure. Mandelic and mandelic/lactic mixtures, however, show appreciable activity against the thick biofilms and it would now be worthwhile investigating the efficacy of these formulations as bladder instillations against urinary tract infections in catheterized patients.
References 1. Ladd TI, Schmiel D, Nickel JC: Rapid method for detection of adherent bacteria on Foley urinary catheters. Journal of Clinical Microbiology 1985, 21: 1004-1006. 2. Nickel JC, Gristina AG, Cosl~erton JW: Electron microscope study of an infected Foley catheter. Canadian Journal of Surgery 1985, 28: 50--52. 3. Ramsay JWA, Garnham A J, Mulhall AB, Crow RA, Bryan JM, Eardley I, Vale JA, Whitfield HN: Biofilms, bacteria and bladder catheters. British Journal of Urology 1989, 64: 395-398. 4. Costerton JW, Cheng KJ, Geesey GG, Ladd TI, Nickel JC, Dasgupta M, Marrle TJ: Bacterial biofilms in nature and disease. Annual Reviews of Microbiology 1987, 41: 435-464. 5. Nichols WW: Susceptibility of biofilms to toxic compounds. In: Characklis WG, Wilderer PA (ed): Structure and function of biofflms. John Wiley, Chichester, 1989, p. 321-331.
Eur. J. Clin. Microbiol. Infect. Dis.
6. Costerton JW: The aetiology and persistence of cryptic bacterial infections: a hypothesis. Reviews of Infec16ous Disease 1984, Supplement 3: 608--612. 7. Clayton CL, Chawla JC, Stickler D J: Some observations on urinary tract infections in patients undergoing long-term bladder catheterization. Journal of Hospital Infection 1982, 3: 39-47. 8. Guttman L: Spinal cord injuries. Comprehensive management and research. BlackweU Scientific Publications, Oxford, 1973, p. 230. 9. Kirk D, Dunn M, Bullock DW, Mitchell JP, Hobbs SJF: t-Iibitane bladder irrigation in the prevention of catheter-assodated urinary infection. British Journal of Urology 1979, 51: 528-531, 10. BaH A J, Carr TW, Gillesple WA, Kelly M, Simpson RA, Smith PJB: Bladder irrigation with chlorhexidine for the prevention of urinary infection after transurethral operations: a prospective controlled study. Journal of Urology 1987, 138: 491-494. 11. Brocklehurst JC, Brocklehurst S: The management of indwelling catheters. British Journal of Urology 1978, 50: 102-105. 12. Stickler D J, Plant S, Bunni Nil, Chawla JC: Some observations on the activity of three antiseptics used as bladder irrigants in the treatment of urinary tract infection in patients with indwelling catheters. Paraplegia 1981, 19: 325-333. 13. Davies A J, Desai HN, Turton S, Dyas A: Does instillation of chlorhexidine into the bladder of catheterized geriatric patients help reduce bacteriuria? Journal of Hospital Infection 1987, 9: 72-75. 14. Stickler D J, Clayton CL, Chawla JC: The resistance of urinary tract pathogens to chlorhexidine bladder washouts. Journal of Hospital Infection 1987, 10: 219228. 15. Stickler D J, Clayton CL, Chawla JC: Assessment of antiseptic bladder washout procedures using a physical model of the catheterized bladder. British Journal of Urology 1987, 60: 413-418. 16. Stickler D, Dolman J, Rolfe S, Chawla J: Activity of antiseptics against Escherichia coli growing as biofilms on silicone surfaces. European Journal of Clinical Microbiology and Infectious Diseases 1989, 8: 974-978. 17. Stickier D, Doiman J, Rolfe S, Chawla J: Activity of some antiseptics against urinary tract pathogens growing as biofilms on silicone surfaces. European Journal of Clinical Microbiology and Infectious Diseases 1991, 10. 18. Prosser BTL, Taylor D, Dix BA, Cleeland R: Method of evaluating effects of antibiotics on bacterial biofilm. Antimierobial Agents and Chemotherapy 1987, 31: 1502-1506. 19. Nickel JC, Ruseska 1, Wright JB, Costerton JW: Tobramycin resistance of Pseudomonas aeruginosa cells growing as a biofilm on urinary catheter material. Antimicrobial Agents and Chemotherapy 1985, 27: 619--624.
Vol. 10. No. 3
157 '1
Eur. J. Clin. Microbiol. Infect. Dis., March 1991, p. 157-162 0934-9723/91/03 0157-06 $ 3.00/0
' Activity of Antiseptics against Biof'dms of Mixed Bacterial Species Growing on Silicone Surfaces
D. Stickler*, P. H e w e t t
As part of a programme to establish the relative merits of antiseptics that are used as bladder instillations to control urinary tract infectious in patients with indwelling catheters, the activity of five such formulations were tested against dense (109 cfu/ cm 2) mixed biofilms composed of Citrobacter diversus, Pseudomonas aeruginosa and En terococeusfaecalis growing on silicone discs. All three species were resistant to chlorhexidine (200 rag/l) and povidone-iodine (1% v/v) in the biofilm mode of growth, whereas these agents rapidly eliminated viable cells from urine suspensions of the mixed community. Lactic acid (1% v/v) produced a 1 log reduction of the biofilm population within 30 min of exposure. The mandelic acid (1% w/v) and mandelic acid (0.5 % w/v)/iactic acid (0.5 % v/v) mixture proved to be the most effective in eliminating the biofilm organisms. It is suggested that these latter solutions should now be tested for efficacy in bladder washouts against urinary tract infectious in catheterized patients.
Indwelling bladder catheters taken from patients with chronic urinary tract infections are colonized by films of adherent bacteria (1-3). It is becoming clear that in this biofilm mode of growth, bacteria are resistant to a variety of antibacterial agents (4, 5). Costerton (6) suggested that during antibiotic treatment of infections of the catheterized urinary tract, while the antibacterial agent might kill planktonic organisms in urine suspension, cells in the biofilms on the catheter and the bladder epithelium could survive and be responsible for the rapid re-establishment of these infections that occurs after the course of treatment is completed
(7). As an alternative to systemic antibiotic therapy, washing the bladder with instillations of antiseptics has been advocated as a method for controlling urinary infections in patients with long-term indwelling catheters (8). While regular bladder washouts with antiseptics such as chlorhexidine have been reported to prevent infections (9, 10), once the infection has become established in the catheterized urinary tract many of the antiseptics currently used as instillations are unable to eliminate the bacteria from the bladder (11-13). This failure to clear established infections was also observed in experiments performed in a physical School of Pure and Applied Biology, University of Wales College of Cardiff, PO Box 915, Cardiff CF1 3TL, UK.
model of the catheterized bladder (14, 15). Cells that had colonized the surfaces of the model were particularly resistant to the antiseptics and initiated the rapid recovery of the urine cultures after the instillation period. Biofilms of Escherichia coli, Citrobacter diversus, Pseudomonas aeruginosa, Proteus mirablis and Klebsiella pneumoniae growing on silicone discs have also been shown to resist the concentrations of chlorhexidine commonly used in bladder washouts (16, 17). Mandelic acid and a mandelic/lactic acid formulation, however, proved to be markedly bactericidal against these biofilms. In these experiments single species biofilms composed of monolayers of ceils w ere produced at cell densities of ca. 107 cfu/cm 2. While these simple pure culture biofilms exist in vivo, it is also clear that mixed populations of up to four species can form thick multi-layered films on catheter surfaces (2, 3). In attempting to refine a laboratory model [or testing the activity of antibacterials against biofilms we have sought to mimic the characteristics of some of the biofilms that we have observed on catheters taken from paraplegic patients on long-term chatheterization. In this paper we report some observations on the activity of antiseptics on mixed community biofilms composed of Citrobacter diversus, Pseudomonas aeruginosa and Enterococcusfaecalis at ceil densities of> 109 cfu/cm2.
158
Materials and Methods Test Organisms. The strains of bacteria used in this investigation, Citrobacter diversus R25.I, Pseudomonas aeruginosa R1811 and Enterococcus faecalis R812, were all isolated from catheters removed from patients undergoing long-term indwelling catheterization at the Welsh Spinal Injuries Unit, Rookwood Hospital, Cardiff. Silicone Discs. Cosmesil elastomer (Cosmedia, UK) was prepared as a polymer sheet 0.2 era in thickness. Discs (0.7 cm diameter) were cut from the sheet and autoclaved in deionized water (121 °C for 15 rain) prior to use.
Model of the Catheterized Bladder. The catheterized bladder is a continuous flow system in which a reservoir of urine is maintained by the presence of the catheter. The details of our simple physical model of such a system have been described previously (14). In summary, sterile pooled urine is delivered by a peristaltic pump at a rate of i ml/min to a small fermentation flask which acts as the bladder. The flask is maintained at 37 °C by a water jacket and the residual volume of urine is maintained at i0 ml by a catheter drainage system connected to a vacuum pump via a Buchner flask (21), which represents the urine drainage bag.
Biofilm Formation. Mixed population biofilms were established with the three bacterial species using a modified version of the procedure described by Prosser et at. (18). The bacteria were grown overnight on Mueller-Hinton Agar (Oxoid, UK) at 37 °C. Cells were then scraped from the agar and suspended in Hanks-Hepes (N-2-hydroxyethylpiperazine-N'-2-ethane sulphonic acid)(H-H) buffer pH 7.4, to an optical density of 0.8 at 540 nm. Aliquots (5 ml) of each bacterial suspension were added to the sterile discs in a plastic petri dish. The discs were incubated for 1 h at 37 °C to permit attachment of suspended cells. They were then removed from the suspension and washed once in H H buffer. Sets of discs were transferred to the bladder model, being placed on a nylon mesh support in a growth medium of pooled sterile urine. After a preliminary incubation period of 2 h, the supply of fresh sterile urine to the model was switched on and the model run for 48 h.
Bactericidal Activity against Biofllms. Sets of 30 discs colonized by the three membered biofilm were prepared for each experiment. The discs were taken from the model and washed twice in H H buffer to remove planktonic cells. Nine discs (controls) were placed into a petri dish containing 20 ml of sterile pooled urine. Two further sets of nine discs were then placed into dishes containing sterile pooled urine and antibacterial agents at concentrations normally used for bladder washouts. The remaining three discs were used to determine the initial number of viable cells present in the biofilms at the start of the experiment. The three petri dishes were incubated at 37 °C and triplicate discs removed from each at 30, 60 and 120 man. Discs were washed twice in HH buffer and placed individually in NIH broth (10 ml) containing Tween 80 (3 % v/v) to neutralize the biocides. Disruption of the biofilms was achieved by sonication for 5 rain in a Trans-Sonic T310 sonicating bath (Camlab, UK) and agitation for 2 man on a vortex mixer. Discs removed at zero time were treated in the same way and viable cell counts were performed by plating out serial dilutions of the resulting cell suspensions onto CLED agar (Oxoid) and overnight incubation at 37 °C.
Eur. J. Clin, Microbiol. Infect. Dis.
Bactericidal Activity of Antiseptics on Suspended Cells. Sets of colonized silicone discs were prepared as described above. Sets of nine discs were then placed individually into 10 ml of sterile urine (controls) and sterile urine containing antiseptics. Three discs were also taken to determine the initialnumber of viable cells present in the biofilm (t = 0). The biofilms were disrupted by sonication and vortex mixing and the discs then removed from the suspensions. The ceils from the disrupted biofilms were incubated in urine and urine containing antiseptics at 37 °C. Triplicate samples were removed at 30, 60 and 120 min for viable cell counting on CLED agar.
Antibacterial Agents. The antiseptic agents used in this study were chlorhexidine digluconate (ICI, UK), mandelic and lactic acids (BDH, UK), and povidone-iodine (Napp Laboratories, UK). Scanning Electron Microscopy. Silicone discs colonized by biofilm were washed gently in HH buffer and placed in a fixative solution consisting of 3 % glutaraldehyde in phosphate buffer (pH 7.4) for 2 h at 4 °C. Samples were then exposed to osmium tetroxide (1% w/v) in phosphate buffer (pH 7.4) for 1 h and dehydrated in a series of aqueous ethanol solutions (20-100 %). Critical point drying was performed in liquid COz, and finally samples were coated with gold i n a sputter coater before examination in a scanning electron microscope (JEOL JSM5200).
Results
The 48 h growth period in the bladder model consistently produced biofilms composed of about 109 cfu/cm2. The composition of the three-membered community in the biofilm was also relatively uniform. Citrobacter diversus was characteristically the major component and Enterococcus faecalis the minority member, the ratio of the numbers of Citrobacter diversus to Pseudornonas aeruginosa to Enterococcusfaecalis being 10:4:1. The activity of chlorhexidine, mandelic acid,lactic acid, a mandelic/lactic acid mixture and povidoneiodine against these biofilms and control cell suspensions prepared from biofilms are shown in Figures 1 to 5. It can be seen that chlorhexidine (200 txg/ml) and povidone-iodine (1% v/v) had negligible activity against the three species in the biofilm mode of growth. Lactic acid ( 1 % ) produced after 2 h exposure a uniform 2 log reduction in the viable numbers of each organism. Mandelic acid and the mandelic/lactic acid mixture proved to be the most active, essentially eliminating all three organisms within 2 h. The scanning electron micrograph (Figure 6) confirms the multi-layered nature of these biofilms. The effect of mandelic acid (Figure 7) is clearly elimination of the biofilm from the silicone surface.
Vol. 10, 1991
159
8.0 _
C
B
A
6.0 m
'~E O
4.0 o
q 2.0
!
I
l
I
tlJ
_z
it
o
7.
6.0
--~3
4.0
.q 2.0
I ....
0
I
I
60
0
120
60
120
o
60
120
TIME (mln) Figure 1: Activity of chlorhexidine (200 I.tg/ml) against biofilms and suspensions of Citrobacter diversus (o), Pseudomonas aeruginosa (A) and Enterococcus faecalis (•). T h e corresponding d o s e d symbols indicate the viability of control biofilms and cell suspensions in normal urine. The vertical bar represents + SD.
A
8.0
B
C
..a ii
6.0
o
"E t)
4.0
9 2•0
•
I
I
I
I.U
z
o
~.__.--.--o
6.0
o
4.0
o 2.0
v
(I
60
120
0
60 TIME (min)
120
0
v
60
w
120
Figure 2: Activity of mandelic acid ( 1 % w/v) against biofilms and suspensions of Citrobacter diversus (o), Pseudomonas aeruginosa (z~) and Enterococcus faecalis (C3). The corresponding closed symbols indicate the viability of control biofilms and cell suspensions in normal urine. The vertical bar represents +- SD.
160
Eur. L Clin. Microbiol. Infect. Dis.
LI..
_.9_ rn
C
B
A
8.0
6,0
"E
4.0
it
o
9 2,0
v
f
Ill
z
r~
6.0
@ ¢o,=,
4.0
o, 2.0
•
w 0
.__...O_.__..-~
_.._...--o
'
v
v
60
120
v
60
o
........U
120
0
60
120
TIME (min)
l~gure 3: Activity of lactic acid ( 1 % v/v) against biofilms and suspensions of Citrobacter diversus (o), Pseudomonas aeruginosa (A) and Enterococcus faecalis (0). The corresponding closed symbols indicate the viability of control biofilms and cell suspensions in normal urine. The vertical bar represents +_SD.
8.0
__. an
6.0
~ 0
4.0
A
C
B
q 2.0
I
I
r
I
I
I
W
~
6.0
,p 0
e
4.0
0 2.0
v
0
60
120
0
w
60
v
v
120
0
v
60
120
TIME (min}
l,igure 4: Activity of mandelic (0.5 % w/v) and lactic acid (0.5 % v/v) against biofilms and suspensions of Citrobacter diversus (o), Pseudomonas aeruginosa (A) and Enterococcus faecalis (O). The corresponding dosed symbols indicate the viability of control biofilms and cell suspensions in normal urine. The vertical bar represents _+SD.
VO1.10, 1991
161
Discussion
Previous studies on the sensitivity of urinary tract organisms growing on the surface of catheter materials have used pure cultures and biofilms with cell densities which represent simple layers one or a few cells thick (16-19). Ramsay et al. (3) reported that indwelling catheters often carry mixed communities of organisms and our own unpublished observations on long-term indwelling catheters removed from paraplegic patients have also revealed the presence of thick multi-layered biofilms populated by several bacterial species. Figure 5: Activity of povidone-iodine (1% v/v) against biofilms and suspensions of Citrobacterdiversus (o), Pseuaomonos aeruginosa (A) and Enterococcus faecalis (~). The corresponding dosed symbols indicate the viability of control biofilms and cell suspensions in normal urine. The vertical bar represents + SD.
Figure 6: Scanning electron micrograph of the mixed community biofilm after 48 h growth in urine in the mode/of a catheterized bladder.
~°
I~KU
HI
!~
OBOIL@2
Figure 7: The biofilm after exposure to mandelic acid (I % w/v) for 120 rain.
The Citrobacter diversua, Pseudomonas aeruginosa, Enterococcus faecalis mixture used in this study grew as a stable, reproducible community with relatively constant proportions of each species. Incubating the discs in urine in the model of the catheterized bladder for 48 h resulted in thick multi-layered biofilms with a characteristic cell density of i09 cfu/cm2. The cells in these biofilms respond quite differently from the same cells in simple urine suspension to antiseptics. While the suspended cells are essentially eliminated from the urine by chlorhexidine, the biofilm community survives well, even after exposure periods of 120 min (Figure 1). Previous work had shown that simple monolayers of this strain of Citrobacter diversus were sensitive to chlorhexidine (17). The thick multi-layered three membered community obviously provides protection for an organism which would normally be considered extremely sensitive to this antibacterial agent. These observations also highlight the need for more information on the nature of biofilms growing on indwelling catheters. The testing for agents active against biofilms can then be performed on laboratory biofilms that are models of the natural biofilms. While mandelic acid was rapidly and effectively bactericidal against the cell suspension, it took 120 min to eradicate the three species from the biofilm (Figure 2). The normal times of exposure to antiseptics in the bladder washout procedure are generally 20-30 rain and this period produced a 34 log reduction in the viability of the biofilm. Mandelic acid has previously been shown to be effective against simple biofilms of Escherichia coli, Pseudomonas aeruginosa, Proteus mirabilis, Citrobacter diversus and Klebsiella pneumoniae (16-17). It would appear therefore that this solution has useful activity against a range of biofilms likely to be found on indwelling bladder catheters.
162
The results presented in Figure 3 indicate that a 20-30 min exposure to lactic acid (1% v/v) would produce a 1 log reduction in the viability of the biofilm. A similar effect has been observed against simple Escherichia coli films (16). The combination of mandelic and lactic acids proved as effective as mandelic alone against both the biofilm and urinary suspension (Figure 4). In experiments conducted in a physical model of the catheterized bladder (15), povidone-iodine was shown to be ineffective against simulated established infections with urinary tract pathogens including Pseudomonaz" aeruginosa and Enterococcus faecalis. The results presented in Figure 5 confirm the poor activity of povidone-iodine against organisms colonizing the surface of catheter material. In conclusion, it seems that ceils in mixed culture biofilms at cell densities of > 109 cfu/cm2 are resistant to the concentrations of chlorhexidine and povidone-iodine that are used in the bladder washout procedure. Mandelic and mandelic/lactic mixtures, however, show appreciable activity against the thick biofilms and it would now be worthwhile investigating the efficacy of these formulations as bladder instillations against urinary tract infections in catheterized patients.
References 1. Ladd TI, Schmiel D, Nickel JC: Rapid method for detection of adherent bacteria on Foley urinary catheters. Journal of Clinical Microbiology 1985, 21: 1004-1006. 2. Nickel JC, Gristina AG, Cosl~erton JW: Electron microscope study of an infected Foley catheter. Canadian Journal of Surgery 1985, 28: 50--52. 3. Ramsay JWA, Garnham A J, Mulhall AB, Crow RA, Bryan JM, Eardley I, Vale JA, Whitfield HN: Biofilms, bacteria and bladder catheters. British Journal of Urology 1989, 64: 395-398. 4. Costerton JW, Cheng KJ, Geesey GG, Ladd TI, Nickel JC, Dasgupta M, Marrle TJ: Bacterial biofilms in nature and disease. Annual Reviews of Microbiology 1987, 41: 435-464. 5. Nichols WW: Susceptibility of biofilms to toxic compounds. In: Characklis WG, Wilderer PA (ed): Structure and function of biofflms. John Wiley, Chichester, 1989, p. 321-331.
Eur. J. Clin. Microbiol. Infect. Dis.
6. Costerton JW: The aetiology and persistence of cryptic bacterial infections: a hypothesis. Reviews of Infec16ous Disease 1984, Supplement 3: 608--612. 7. Clayton CL, Chawla JC, Stickler D J: Some observations on urinary tract infections in patients undergoing long-term bladder catheterization. Journal of Hospital Infection 1982, 3: 39-47. 8. Guttman L: Spinal cord injuries. Comprehensive management and research. BlackweU Scientific Publications, Oxford, 1973, p. 230. 9. Kirk D, Dunn M, Bullock DW, Mitchell JP, Hobbs SJF: t-Iibitane bladder irrigation in the prevention of catheter-assodated urinary infection. British Journal of Urology 1979, 51: 528-531, 10. BaH A J, Carr TW, Gillesple WA, Kelly M, Simpson RA, Smith PJB: Bladder irrigation with chlorhexidine for the prevention of urinary infection after transurethral operations: a prospective controlled study. Journal of Urology 1987, 138: 491-494. 11. Brocklehurst JC, Brocklehurst S: The management of indwelling catheters. British Journal of Urology 1978, 50: 102-105. 12. Stickler D J, Plant S, Bunni Nil, Chawla JC: Some observations on the activity of three antiseptics used as bladder irrigants in the treatment of urinary tract infection in patients with indwelling catheters. Paraplegia 1981, 19: 325-333. 13. Davies A J, Desai HN, Turton S, Dyas A: Does instillation of chlorhexidine into the bladder of catheterized geriatric patients help reduce bacteriuria? Journal of Hospital Infection 1987, 9: 72-75. 14. Stickler D J, Clayton CL, Chawla JC: The resistance of urinary tract pathogens to chlorhexidine bladder washouts. Journal of Hospital Infection 1987, 10: 219228. 15. Stickler D J, Clayton CL, Chawla JC: Assessment of antiseptic bladder washout procedures using a physical model of the catheterized bladder. British Journal of Urology 1987, 60: 413-418. 16. Stickler D, Dolman J, Rolfe S, Chawla J: Activity of antiseptics against Escherichia coli growing as biofilms on silicone surfaces. European Journal of Clinical Microbiology and Infectious Diseases 1989, 8: 974-978. 17. Stickier D, Doiman J, Rolfe S, Chawla J: Activity of some antiseptics against urinary tract pathogens growing as biofilms on silicone surfaces. European Journal of Clinical Microbiology and Infectious Diseases 1991, 10. 18. Prosser BTL, Taylor D, Dix BA, Cleeland R: Method of evaluating effects of antibiotics on bacterial biofilm. Antimierobial Agents and Chemotherapy 1987, 31: 1502-1506. 19. Nickel JC, Ruseska 1, Wright JB, Costerton JW: Tobramycin resistance of Pseudomonas aeruginosa cells growing as a biofilm on urinary catheter material. Antimicrobial Agents and Chemotherapy 1985, 27: 619--624.
