Efficacy of silver-releasing rubber for the prevention of Pseudomonas aeruginosa biofilm formation in water

KRISTOF DE PRIJCK, HANS NELIS & TOM COENYE

Laboratorium voor Farmaceutische Microbiologie, Universiteit Gent, Gent, Belgium

Abstract

The aim of the present study was to evaluate the efficacy of silver-releasing rubber in preventingPseudomonas aeruginosa biofilm formation in water.Biofilm formation by P. aeruginosaunder various conditions in anin vitro model system were compared for silver-releasing and conventional rubber. Under most conditions tested,the numbers of sessile cells attached to the silver-releasing rubber were considerably lower with reference to conventional rubber, although the effect diminished with increasing volumes. The release of silver also resulted in a decrease in planktonic cells. By exposing both materials simultaneously to conditions for biofilm growth, it became obvious that the antibiofilm effect is due to a reduction in the number of planktonic cells, rather than to contact-dependent killing of sessile cells.Our data demonstrate that the use of silver-releasing rubberreducesP. aeruginosa biofilm in water and reduces the number of planktonic cells present in the surrounding solution.

Key words : Biofilm, silver, Pseudomonas aeruginosa, disinfection

Running title : Inhibition of P. aeruginosa biofilm formation by silver

Correspondence : Tom Coenye, Laboratorium voor Farmaceutische Microbiologie, Universiteit Gent, Harelbekestraat 72, 9000 Gent, Belgium. E-mail :

Introduction

Biofilms are consortia of micro-organisms that are formed on various surfaces, including industrial surfaces and various household surfaces. Biofilm formation is a multi-stage process in which microbial cells adhere to the surface, and the subsequent production of an exopolysaccharide matrix results in a more irreversible attachment(Donlan, 2001 ; Donlan & Costerton, 2002 ; Hall-Stoodley et al. 2004). Depending on the conditions, biofilms can further develop into complex and differentiated communities. Biofilms can be a serious threat to human health, as sessile cells are often extremely resistant to antimicrobial treatment and biofilms are known to be involved in indwelling medical device-related infections, endocarditis, otitis media, periodontitis and various airway infections (Lewis, 2001 ; Donlan, 2001 ; Donlan & Costerton, 2002 ; Fux et al. 2005). In addition, the impact of biofilms on industrial processes can hardly be overestimated as losses caused by biofilm formation during food processing, in drinking water distribution systems and in various industries are significant (Wong, 1996 ; Cloete et al. 1998 ; Szewzyk et al. 2000 ; Fleming et al. 2002). Pseudomonas aeruginosa is an important human pathogen, causing a wide range of diseases, including potentially life-threatening pulmonary infections in cystic fibrosis patients (Govan & Deretic, 1996). Recent data have shown that P. aeruginosa can form biofilms on many different surfaces, including whirlpools, waterlines in dental units and various parts of drinking water systems(see for example Price & Ahearn, 1988 ; Barbeau et al. 1996 ; Chaidez & Gerba, 2004 ; Kohnen et al. 2005). Its presence can be associated with complaints about taste, odour and turbidity and may pose a threat to the health of susceptible individuals (WHO, 2004).

Silver (Ag) has a long history of use as antimicrobial agent, and is currently still used for the prophylaxis of conjunctivitis of the newborn and the topical treatment of burn wounds (Weber & Rutala, 2001). Other medical applications include the use of silver impregnated urinary or vascular catheters and bandages for trauma and diabetic wounds (Weber & Rutala, 2001 ; Silver, 2003). The non-medical applications of silver include its use in textile products (sleeping bags, sport socks), paint and as a disinfectant for particular water systems (e.g. hospital water systems) (Rogers et al. 1995 ; Weber & Rutala, 2001 ; Silver, 2003).Silver-containing disinfectants for surface decontamination have also been described (Brady et al. 2003 ; Surdeau et al. 2006).The exact mechanism of action of silver is not known, but it is thought that, due to its reactive nature, silver will interact with thio, amino,imidazole, carboxylate, and phosphate groups in biological molecules. The binding of silver ions to bacterial DNA results in DNA condensation and blocking of replication (Feng et al. 2000). In addition, silver can interact with proteins involved in cellular oxidation processesas well as the respiratory chain (Feng et al. 2000 ;Weber & Rutala, 2001). More recently, it was demonstrated that silver ions can interact with the bacterial ribosome and may exhibit their bactericidal action by inhibiting the production of essential proteins (Yamanaka et al. 2005). Chaw et al. (2005) proposed that silver can also exert a specific antibiofilm effect by destabilising the biofilm matrix. This was shown for Staphylococcus epidermidis biofilms and was thought to be the result of binding to electron donor groupsof biological molecules, leading to reduction in the number of binding sites for H-bonds and electrostatic and hydrophobic interactions.There are numerous studies in which silver-impregnated or –coated medical materials have been compared with their native counterparts in terms of preventing device-related infections, often with apparently conflicting results. For example, a meta-analysis on the use of silver-impregnated central venous catheters revealed that their use results in significant reduction of risk of bloodstream infection (Weber & Rutala, 2001) while a large-scale study (1309 patients) failed to demonstrate efficacy of a silver-oxide coated urinary catheter in prevention of catheter-associated bacteriuria (Riley et al. 1995). Studies with P. aeruginosa and silver-coated or silver-releasing devices showed that the use of modified materials often resulted in reduced P. aeruginosa biofilm formation (Liedberg et al. 1990 ; Stickler et al. 1996 ; Kumon et al. 2001 ; Balazs et al. 2004), although in some cases a lack of efficacy in preventing P. aeruginosa biofilm formation was noted as well (Biedlingmaier et al.1998 ; Kampf et al. 1998 ; Berry et al. 2000 ; Jarret et al. 2002).

The goal of the present study was to investigate whether the incorporation of a silver-releasing compound in rubber would result in reduced P. aeruginosa biofilm formation on this material.

Materials and methods

Surfaces, strains and growth conditions

The surface tested (provided by Milliken Europe, Gent, Belgium) was aheat-cured rubber containing a zirconium phosphate-based ceramic ion-exchange resin. The latter is responsible for the release of silver ions in exchange for monovalent cations like K+ and Na+. Rubber without the silver-loaded ion exchange resin was used as a control surface. Disks were cilinder-shaped andwere 6.8 mm in diameterand 2.6 mm in height (total surface area appr. 120 mm²). The test strain used was Pseudomonas aeruginosa ATCC 9027 (a clinical isolate recovered from an ear infection and used in many standardised susceptibility tests). Other tests strains included in this study wereP. aeruginosa ATCC 27853 (a clinical isolate recovered from blood), ATCC 15442 (isolated from an animal room water bottle), CIP A22 (isolated from a wound) and reference strain PAO-1. Strains were routinely cultured aerobically on Tryptic Soy Agar (TSA)(Oxoid, Drongen, Belgium) at 37°C, unless otherwise mentioned.

Determination of the minimal inhibitory concentration (MIC) of silver

MIC determinations were carried out in modified asparagine broth (MAB), containing 3% (w/v) DL-asparagine, 0.1% (w/v) KH2PO4and 0.05% MgSO4.7H2O (pH 6.9 – 7.2),by making serial dilutions of a silver nitrate (AgNO3) solution in the wells of a round-bottom 96-well microtiter plate (TPP, Trasadingen, Switzerland) and adding 100 µl of a standardised cell suspension. The final Ag+ concentrations ranged from 0.00128 µg ml-1 to 128 µg ml-1. Following 24h incubation at 37°C, growth was assessed by determining the absorbance at 690 nm using a Wallac Victor2 (PerkinElmer Life And Analytical Sciences, Waltham, MA, USA) microtiter plate reader.

Biofilm formation on silver-releasing and conventional rubber in microtiterplates

Biofilms were formed on disks in 24-well microtiter plates (TPP) (for 700µl and 2 ml volumes) or in Petri dishes (for 5 ml and 10 ml volumes). Disks were placed in P. aeruginosa suspensions and biofilms were allowed to be formed. During biofilm formation at room temperature the microtiterplate or petri dish was placed in a humidified environment (to minimise evaporation) on a rotary shaker (300 rpm)(Titramax 1000, Heidolph, Nurnberg, Germany). Unless otherwise mentioned, an inoculum density of 106 CFUml-1 was used. Biofilm experiments were carried out in MAB and in milliQ water. In order to determine the efficacy of silver-releasing rubber in preventing biofilm formation in commercial waters, we also tested five different commercial waters (labelled A through E).

Enumeration of planktonic and sessile cells

After biofilm formation, each disk was transferred to 10 ml 0.9% (v/w) NaCl. Tubes were subjected three times to 30 s of sonication (Branson 3510, 42 kHz, 100 W, Branson Ultrasonics Corp.) and 30 s of vortex mixingto remove the biofilm cells from the disks. From these suspensions serial tenfold dilutions were made and the number of CFU per disk was calculated by plating on TSA and counting colonies on the plates following incubation. For the enumeration of planktonic cells,supernatant was diluted 100-fold in Tryptic Soy Broth (TSB)(Oxoid, Drongen, Belgium). From this suspension serial tenfold dilutions were made and the number of CFU per ml was calculated by plating on TSA and counting colonies on the plates following incubation. Parallel enumerations using membrane filtration (0.22 µm pore diameter) (Microfil, Millipore Corp., Bedford, CT, USA) showed that the supernatant was diluted enough for the silver not to interfere with the actual enumeration.

Determination of the amount of silver released

The amount of silver released was determined by Inductively Coupled Plasma-Optical Emission Spectrometry (ICP-OES) using a Vista-MPX ICP-OES (Varian, Palo Alto, CA, USA). Prior to the analysis the samples were filtered by filtration (0.22 µm pore diameter) (Microfil) to remove microbial cells. We made several attempts to quantify the amounts of silver released using ICP-OES but we consistently observed that the levels of silver were below the detection limit (appr. 1 µg l-1) of the assay. We assume that this was due to the fact that the silver is bound to cellular components (e.g. proteins): as the microbial cells had to be removed from the supernatant prior to the analysis, an accurate quantification of the silver was not possible.

Interpretation of the data and statistical analysis

We used the standard deviation of a measurement across repetitions to indicate repeatability and based on this repeatability standard deviation, the coefficient of variation was calculated (Pitts et al. 2001). Currently no standardised tests or guidelines are available to test the efficacy of antimicrobial surfaces under the conditions used in the present study and for that reason we defined efficient treatments as those resulting in at least 99.9% reduction (3 log) of P. aeruginosa microorganisms compared to the control. Whenever appropriate, statistical tests were performed using the SPSS 12.0 software package (SPPS).

Results

Minimal inhibitory concentration of silver against P. aeruginosa

The MIC of silver for P. aeruginosa ATCC 9027 was determined using a broth microdilution method. In order to determine the potential effect of the presence of rubber on the MIC, MICs were determined in the absence and presence of a conventional rubber disk. As can be seen from Fig. 1, the MIC was appr.0.26 µg Ag+ml-1. MICs were appr. the same in MAB and water (data not shown). In addition, the presence of a rubber disk had no meaningful influence on the MIC, indicating that it did not interfere with the antimicrobial effect of silver. In order to determine whether the susceptibility of P. aeruginosa ATCC 9027 towards silver was representative for the susceptibility of other P. aeruginosa strains, MICs were also determined for P. aeruginosa ATCC 27853, ATCC 15442, CIP A22 and PAO-1, and were found to be in the same range as for ATCC 9027 (data not shown). These MIC values are also in agreement with previously determined values (Berger et al. 1976).[INSERT FIGURE 1 HERE]

Repeatability assessment of biofilm formation

As expected, P. aeruginosa readily formed biofilms on rubber disks immersed in different growth media, including TSB, MAB and water. In order to determine the reproducibility of our assay, biofilms were formed on conventional and silver-releasing disks (belonging to two different production batches) by three different operators, on multiple occasions. Experiments were carried out in 700 µl MAB and the bacterial biomass was quantified following 24 h of incubation (Table I). For rubber disks, the repeatability standard deviations were 0.20 CFU per disk for sessile cells and 0.21 CFU per ml for planktonic cells. Slightly higher values (0.23 CFU per disk for sessile cells and 0.50 CFU per ml for planktonic cells) were obtained for silver-releasing disks. Repeatability standard deviations for observed log reductions were also low (0.30 CFU per disk for sessile cells and 0.54 CFU per ml for planktonic cells).[INSERT TABLE I HERE]

Effect of silver incorporation on P. aeruginosa biofilm formation

In order to determine the effect of the release of silver ions on the numbersof sessile and planktonic cells, biofilms were formed on conventional and silver-releasing rubberunder various conditions. An overview of the results is shown in Tables II and III. When using MAB as growth medium there was considerably less biofilm formation (at least 99.9% reduction) on the silver-releasing disks at all time points sampled for the 700 µl, 2 ml and 5 ml experiments (Table II).[INSERT TABLE Ii HERE]The silver released from the disks drastically reduced the numbers of planktonic cells as well. When a larger volume was tested (10 ml) the efficacy dropped, especially with longer incubation periods (> 24 h). When water was used as growth medium, similar results were obtained, with prolonged incubation (72 hours and more) of disks in a small (up to 5 ml) volume even resulting in near-sterility of the disk and/or the surrounding solution (Table III).[INSERT TABLE III HERE]All experiments were carried out with rather high inocula (appr. 106 CFUml-1)and using MAB as growth medium, higher reductions were obtained with more dilute inocula (105, 104, 103 or 102 CFUml-1)than with the standard inoculum (Fig. 2).[INSERTFIGURE 2 HERE]The use of silver-releasing disks led to sterile or near-sterile surfaces and solutions as neither sessile nor planktonic cells could be recovered. When lower inocula (≤ 105 CFU ml-1) were used with water as growth medium, similar reductions were observed as with the standard inoculum (106 CFU ml-1) (Fig. 2).With these inocula, neither sessile nor planktonic cells could be recovered, indicating sterile (or near-sterile) surfaces and solutions.

Mechanism of the antibiofilm effect

In order to determine which mechanismunderlies the antibiofilm effect of silver-releasing rubber we determined the effect of simultaneous incubation of two disks in the same well of a microtiterplate. As can be seen from Table IV, the presence of a silver-releasing disk also resulted in reduced biofilm formation on the conventional rubber disk, suggesting that surface contact-dependent killing is not the mechanism by which the silver exerts its action. This is confirmed by the consistently observed reductions in planktonic cell numbers and by the decreasing efficacy in higher volumes (Tables II and III).[INSERT TABLE IV HERE]

Application of silver-releasing rubber in commercial waters

We determined whether silver-releasing rubber could be used to control biofilm formation in water. For this test we used five different commercial waters available on the Belgian market, which varied significantly in their composition (Table V). As can be seen from Fig. 3, the presence of a silver-releasing disk resulted in a marked decrease in numbers of both sessile and planktonic cells when commercial waters were used as growth media. [INSERT TABLE V HERE][INSERT FIGURE 3 HERE]

Discussion

The aim of the present study was to evaluate the efficacy of silver-releasing rubber in preventing P. aeruginosa biofilm formation in various conditions, using a microtiter plate model system. The results from our experiments clearly show that our biofilm model system can provide repeatable assays of the efficacy of silver-releasing rubber disks against P. aeruginosa biofilms, as relatively small repeatability standard deviations were obtained for log reduction measurements (Table I) which are within the range of standard deviations observed for standard suspension and surface disinfection assays (0.2 – 1.2) (Tilt & Hamilton 1999). It was also obvious that the use of silver-releasing rubber results in a marked decrease in biofilm formation on the surface, as well as in a marked reduction in the number of planktonic cells in the surrounding suspension, in a volume-, time- and inoculum-dependent way (Table II, Fig. 2).

There are two possible mechanisms of action that could explain the observed reductions in cell numbers. A first possible mode of action is “contact-dependent killing”, in which silver ions continuously released from the surface kill bacterial cells adhering to this surface. A second possible mechanism is that the number of sessile cells is reduced because the release of silver ions into the culture medium reduces the number of planktonic cells.We tried to determine this mechanism by determining the effect of simultaneous incubation of two disks. The rationale behind these experiments is that when the antibiofilm effect would arise from contact-dependent killing, there would be a significant difference in biofilm biomass on a conventional and a rubber-releasing disk that have been incubated together. However, if the killing of planktonic cells is the underlying reason for the antibiofilm effect, no meaningful differences should be observed between the biomass on both disks, and our data (Table IV) strongly suggest that the latter is the case.

We also tested the applicability of silver-releasing rubber in commercial waters with different compositions and our data clearly indicated that its use also results in a marked decrease in biofilm build-up under those conditions (Fig. 3). In addition, there appeared to be no correlation between the chemical composition of the waters tested and the observed reduction in sessile and planktonic cells, indicating that inorganic ions (including chloride) and/or other compounds present in these waters did not interfere with the antimicrobial activity of the released silver ions.

In conclusion, our study presents an evaluation of the effect of a silver-releasing rubberin a microtiter plate model system against P. aeruginosa biofilms and the data clearly demonstrate that, compared to conventional rubber, the use of the silver-releasing rubber prevents the build-up of these biofilm in water, under various conditions. We also showed that the antibiofilm effect is due to a reduction in the number of planktonic cells, rather than to a contact-dependent killing. Together, the present study indicates that silver-releasing rubber may be applicable in various situations where surfaces come into contact with small volumes of water (e.g. tubing, valves and faucets of water dispensers) to reduce or even prevent biofilm formation. As such its use may be a valuable addition to existing cleaning/disinfection procedures.